part b speech-language pathologyaiish.ac.in/images/pdf/09-10slp.pdf · 1e-mail:...

STUDENT RESEARCH AT A.I.I.S.H. MYSORE

(ARTICLES BASED ON DISSERTATION DONE AT AIISH)

VOLUME VIII: 2009- 10

PART – B

Speech-Language Pathology

Compiled by

Dr. Vijayalakshmi Basavaraj Director

Dr. Swapna N Lecturer in Speech Pathology

All India Institute of Speech & Hearing

Manasagangothri, Mysore 570 006

© 2011 A Publication of the All India Institute of Speech and Hearing Under the title: „Student Research at A.I.I.S.H Mysore‟

Articles based on dissertation done at AIISH: Vol. VIII Part B: Speech-Language Pathology

Telephone : 0821 – 2514449, 2515410, 2515805, 2515218, 2514313 Guest House : 2515786, Security (after office hours): 2514449 Fax : 0821 – 2510515 E-mail : [email protected], [email protected] Website : http://www.aiishmysore.in, http://aiish.ac.in Working hours : 9:00 am to 5:30 pm – Monday through Friday Holidays : Central Government Holidays Price: 150/- Published by Dr. Vijayalakshmi Basavaraj

Director, AIISH, Mysore

Table of Contents

Sl. No.

Title Page No.

1 Treatment Manual for Persons with Anomic Aphasia - Anil Kumar & Jayashree C. Shanbal

1

2 Cognitive Communicative Assessment Protocol for Persons with Dementia in Malayalam (CCAPD-M) - Anjana A.V. & Jayashree C. Shanbal

12

3 Benchmark for Speaker Identification Using Glottal Source Parameters in Hindi Speakers - Aparna V.S. & S.R. Savithri

23

4 Agrammatism in Children with Hearing Impairment - Aswathy A.K. & Shyamala K.C.

30

5 Development of Revised Token Test in Oriya - Bijoyaa Mohapatra & S.P.Goswami

43

6 Re-Standardization of Kannada Articulation Test - Deepa Anand & S.R.Savithri 53 7 Articulatory Acquisition in Typically Developing Malayalam Speaking Children:

2-3 Years - Divya P. & N. Sreedevi 66

8 A Self Therapy Manual in English for Speech and Swallowing Difficulties in Persons with Parkinson‟s disease(PsWPD) - Gayathri Krishnan & R. Manjula

77

9 Attitudes, Anxiety and Coping Strategies in Persons with Stuttering - Jasmine Mallik & Y.V. Geetha

81

10 Benchmark for Speaker Identification Using Long Term Average Spectrum in Kannada Speaking Individuals - Jyothi S. & S. R. Savithri

90

11 Cognitive-Linguistic Assessment Protocol in Telugu: An Adaptation of CLAP in Kannada - D. K. Veena & S. P. Goswami

98

12 Cognitive-Linguistic Assessment Protocol in Malayalam: An Adaptation of CLAP in Kannada – Lakshmi S.Mohan & S. P. Goswami

106

13 Development of Revised Token Test in Malayalam - Lincy Mary Varghese & S. P. Goswami

121

14 Agrammatism in Children with Mental Retardation - Liveem Mariam Tharakan & Shyamala. K.C.

131

15 Performance of PWS on Self Select Reaction Time Paradigm Using Speech and Non Speech Tasks - Mahesh B.V.M. & R. Manjula

140

16 Verbal Perseveration in Malayalam-English Bilinguals - Manju Mohan & Swapna.N

150

17 Some Aspects of Syntax in 4-7 year Old Typically Developing Malayalam Speaking Children: A Computerized Approach - Maria P.R. & Swapna. N

159

18 Computer Based Assessment of Phonological Process (CAPP-M) in Malayalam- A Preliminary Attempt - Merin John & N. Sreedevi

167

19 Spoken Language Proficiency (discourse level) in Bilingual Children with Learning Disability - Mohana P. & Shyamala. K.C.

176

20 English Language Test for Indian Children (ELTIC) - Namala Bhuvaneswari & Jayashree C. Shanbal

182

21 Gender Difference in Nature of Disfluencies in Children with Stuttering - Nisha Sudhi & Y.V.Geetha

190

22 Resource Manual for Children with Communication disorders (3-8 years) on Multi Sensory Skills - Priyashri S. & R.Manjula

199

23 The Effect of Training on the Voice of Carnatic Classical Singers - Ranjini Mohan & S. R. Savithri

205

24 Treatment Manual in English for Indian Children with Dyslexia - Ranjini R. & Jayashree C. Shanbal

210

25 Developing a Protocol for Measuring Participation of Persons with Aphasia - Rupali Mathur & S. P. Goswami

220

26 Agrammatism in Malayalam speaking Children with Autism Spectrum Disorders - Shafna Jahan & Shyamala K.C.

227

27 Written Language Skills in Children with Dyslexia - Sheetal R. & Sangeetha Mahesh

236

28 Effects of Aging on Naming in Bilingual Older Adults - Sharon Susan Sam & Jayashree C. Shanbal

244

29 Vocal Load in Undergraduate Students of Speech-Language Pathology - Shwetha C. & S. R. Savithri

253

30 Nonword Repetition in Children with Language Impairment: An Exploratory Study – Shylaja K. & Swapna. N

263

31 Development of Western Aphasia Battery in Telugu (WAB-T) - Sri Pallavi & Shyamala K.C.

273

32 Usefulness of Cognitive Cueing in Eliciting Vocal Variability and Vocal Naturalness - Tiffy George Roy & K. Yeshoda

281

33 Articulatory Acquisition in Typically Developing Telugu Speaking Children: 2-3 years - Usha Rani K. & N. Sreedevi

291

34 Linguistic and Metalinguistic skills in Children with Stuttering - Yashaswini R. & Y.V. Geetha

302

35 Comparison of Signed Expression of Indian Sign Language (ISL) Users and Markers of Signed English (SE) - Yashomathi & R. Manjula

310

Treatment manual for persons with anomic aphasia

1

Treatment Manual for Persons with Anomic Aphasia Anil Kumar 1 & Jayashree C. Shanbal 2

Abstract

Anomia is synonymously described using several terms like naming impairments, word finding problems, word- recall deficits, and word-retrieval difficulties. In one usage, the term is synonymous with naming disorder which means in a broader sense all aphasic patients are anomic. There have been investigations and assessments attempted to classify them into different types based on their homogeneity of features in the heterogeneous aphasia group. Hence, a treatment procedure which attempts to cater to these difficulties in different types of anomia is essential. Based on the available cognitive neuropsychological models that recognize and explain the mechanism of naming in retrieval as well as production, management procedures were developed for the treatment of anomia. Four subjects with aphasia (all males) following a language-dominant hemisphere stroke with an age range of 47-72 years participated in the present study. The target stimuli and the test stimuli to be used during therapy sessions were developed with the help of two experienced Speech-Language Pathologists. The potential categories selected were nouns and verbs under the category of daily objects, fruits, colors and body parts. The treatment techniques that benefitted most subjects with aphasia were the cueing hierarchy technique and the semantic treatment technique. However, the efficacies of these techniques over the other techniques are required to be investigated on a larger population with anomia. Hence, the present study was only a preliminary investigation of the treatment techniques for anomia.

Key words: anomia, treatment, aphasia, semantics, cueing

nomia has been 1found to be one of the most common problems in patients with aphasia who often have a variety of language

formation problems (Raymer & Ellsworth, 2002). Anomia is synonymously described using several terms like naming impairments, word finding problems, word-recall deficits, and word-retrieval difficulties. In one usage, the term is synonymous with naming disorder which means in a broader sense all aphasic patients are anomic. Schuell, Jenkins and Carroll (1962) found anomia to be the most prevalent general factor in aphasic disorders, and naming difficulties have been found to be the most permanent residual deficit in chronic aphasics. In another of a word-finding disorder (e.g., decreased performance on a confrontation naming task), anomia is not considered of a localizing value (Benson & Geschwind, 1985).

Anomia has also been synonymously used as anomic aphasia (Kertesz, 1982; Goodglass & Kaplan, 1983; Benson & Geschwind, 1985), nominal aphasia (Head, 1926), or amnesic aphasia (Luria, 1976). Naming difficulties often called anomia are present in all aphasics but the term anomia is also used to refer to a particular aphasia syndrome (anomic aphasia). Naming difficulties can result from a deficit at different stages of the naming process; perception (decoding), storage, selection, retrieval, or actual production of the word (encoding) (Barton, Maruszewski & Urrea, 1969; Benson, 1979). Severity of anomia can range from mild to severe, including difficulty remembering a person‟s own

1e-mail: [email protected]; 2Lecturer in Language Pathology, AIISH, Mysore, [email protected].

name. Severity of anomia also depends upon the extent of lesion and site of lesion. Damage to left inferior temporal cortex has been associated with naming deficits resulting either from impaired access to phonological word forms (pure anomia) or from degraded semantic knowledge (semantic anomia). Anomia has been classified in various ways (e.g., Weinstein & Keller, 1963; Geschwind, 1967; Brown, 1972; Benson, 1979; Kremin, 1988). All the schemes are based on clinical distinctions between subtypes of naming disorders.

Aphasia therapy has always been a

challenging task for the Speech–Language Pathologists (SLPs). Aphasia as a disorder itself is not homogenous or unitary due to its varied nature in terms of characteristic features and severity. There have been investigations and assessments attempted to classify anomia into different types based on their homogeneity of features in the heterogeneous aphasia group. Hence, a treatment procedure which attempts to cater to these difficulties in different types of anomia is essential. Based on the available cognitive neuropsychological models that recognize and explain the mechanism of naming in retrieval as well as production, management procedures will be designed for the treatment of anomia. The treatment manual was prepared to be used systematically so that they may encourage an evidence based practice in the treatment of anomia. This would help the clinicians to select the technique appropriate to the type of anomia that they encounter in the clinic. There are various treatment techniques available to treat individuals with anomic aphasia. For example, the Helm elicited language program for syntax stimulation (Helm-Estrabrooks & Ramsburger, 1986)

A

Dissertation Vol. VIII, 2009-10, Part-B, SLP, AIISH, Mysore

2

which target the sentence types in syntax. Likewise some therapy procedures believe in stimulation for the overall facilitation of the language, for example, LOT- Language Oriented Treatment by Bandur & Shewan (2001). Some of the techniques selected for the present study was from a thorough review of literature which included Cueing Hierarchy (CH), Facilitation Therapy (FT), Circumlocution Induced Naming (CIN), Semantic Treatment (ST), and Space Retrieval therapy (SR).

The aim of the present study was to develop a

treatment manual for the treatment of anomic aphasia. The study also attempted to explore the possible techniques that would benefit different types of anomic aphasia.

Method

Participants: Four subjects with aphasia (all males) following a language-dominant hemisphere stroke with an age range of 47-72 years participated in the present study. The demographic details of the participants are given in Table 1. All the participants selected for the present study were enrolled for speech-language therapy at the Institute. All participants provided written informed consent to participate in this study. The participants were assigned letter names as A1, A2, A3 and A4 for confidentiality and the data were analyzed and reported accordingly. The onset of aphasia for all participants was at least 6 months prior to inclusion in the study. All of the participants had attended individualized speech-language therapy program post-stroke. All participants were initially administered the Western Aphasia Battery (WAB; Kertesz, 1982) and they presented with severe naming problems or anomia during the time of testing. Development of stimuli: The target stimuli and the test stimuli to be used during therapy sessions were developed with the help of two experienced SLPs. The potential categories selected were nouns and verbs under the category of daily objects, fruits, colors and body parts. A total of sixty stimuli were prepared to be used for the treatment purposes. These stimuli material were selected from the list of pictures from the UNICEF picture cards (Karanth, Manjula, Geetha & Prema, 1999). A picture (5 × 7 inches) representing each stimulus was made in such a way that each picture card had one stimuli on the card. The stimuli selected were subjected for

familiarity, frequency and imageability. Stimuli which were rated by two SLPs as highly familiar, high frequent and imageable were selected for the therapy program. Not all the stimuli were taken up during therapy session as targeted stimuli. A maximum of 5 stimuli formed the targets during the therapy sessions. The stimuli selected were all line drawings presented initially for confrontation naming before beginning the therapy sessions. The present manual was designed based on the documented principles and guidelines prescribed in the literature for persons with aphasia. The illustrations of various activities are based on the principles of aphasia management. The collected information from these resources are compiled and organized. Procedure: Prior to the therapy program, the pre-test scores of WAB were recorded. During the period of therapy program, the persons with aphasia (PWA) participated for twenty sessions, with each session lasting approximately 45 minutes. Speech-language therapy sessions: Depending on the naming subsection of the WAB test the words were prepared along with a list of other functional words. The words from categories like common objects, fruits, body parts, animals, vehicles, flowers, colors etc. were taken up during therapy sessions. The number of words used for each session was however variable due to the nature of responses by the PWA. Most of these words were not specifically trained/targeted during the therapy sessions; however, a few commonly-used words were included as targets during the therapy sessions. The therapy program was a one-on-one basis with the clinician and PWA. The number of correct responses was recorded for each session for each PWA. After reviewing a vast range of therapy techniques in literature, techniques widely recommended for anomia were considered for the present study. The techniques selected for the present study are listed and described briefly in Table 2. Since the objective of the study was only to prepare therapy manual for anomia and not to assess the efficacy of the therapy techniques, the techniques were used together as additional techniques to facilitate better naming in the above PWA in the present study. The responses were coded and analyzed for correct and incorrect responses. Every correct response was given a score of „1‟ and scored „0‟ for an incorrect response. Subjective remarks of each patient were also recorded in the response sheet for each session.


3

Table 1. Participants‟ demographic information

Table 2. Techniques used in the present study for

treatment of anomia

Results and Discussion

The aim of the present study was to develop a treatment manual for the persons with anomic aphasia. Due to a small sample size, the results of the study are described qualitatively. Non-parametric statistics was employed to compare the pre-therapy and the post-therapy performance of subjects with aphasia. However, Wilcoxon Signed Rank test revealed that there was no significant difference in the performance of subjects with aphasia. The results are presented and discussed under different sections.

I. Performance of subjects with aphasia on cueing hierarchy technique: Cueing hierarchy technique described by Linebaugh and Lehner (1977) was applied for the present treatment program for naming. This technique involved arranging of cues

by increasing the explicitness of the cues provided for each stimulus items. Cues were arranged according to increasing stimulus power and presented until the patient produced the target word. When the patient responded accurately, cues were presented in the order of decreasing stimulus power until the patient again produces the correct response. The cues selected were orthographic, visual and auditory cues. Pre-therapy scores were considered as the baseline score. After the baseline scores were established the post-therapy scores were calculated from responses in each session. Table 3 shows pre-therapy scores and total post-therapy scores calculated for a total of twenty sessions. The Wilcoxon signed rank test results did not reveal any significant difference between the two conditions i.e., between pre-therapy and post-therapy conditions (p>0.05).

Analysis of results for pre-therapy and post-therapy condition for the cueing hierarchy technique revealed that there was an improvement in the performance of PWA. Table 3 shows that the mean scores on post-therapy (Mean=53.25, SD=20.88) was greater than the pre-therapy scores (Mean=32.00, SD=15.68). A qualitative descriptive analysis of individual data was also done and the results are discussed further. In the case of cueing hierarchy technique, partial cues, phonetic cues, description, demonstration of an action and functional description of the target were used as cues. This means that both phonological as well as semantic cues were used when the cueing hierarchy technique was employed. It was found that Subject A1 produced more number of phonological naming errors than the semantic errors. So, the errors observed were more of phonemic paraphasias than semantic paraphasias.

Information A1 A2 A3 A4 Age (in years) 47 55 50 72

Primary language Hindi Hindi Hindi Hindi Months post stroke 9 5 11 1

Lesion site (Hemisphere) Left Left Left Left Premorbid Handedness Right Right Right Right WAB aphasia quotient 30.2 45.6 15 23.7

Aphasia type Broca‟s Broca‟s Global Broca‟s Associated deficits Hypertension Diabetes Diabetes Diabetes

Sl. No.

Type of technique Stimuli modality

1. Cueing hierarchy

Orthographic visual

2. Facilitation

therapy

Orthographic visual

auditory 3. Circumlocution

induced naming (CIN) therapy

Auditory visual

4. Semantic treatment

Orthographic auditory visual

5 Spaced retrieval therapy (SRT)

Orthographic auditory visual


4

Table 3. Pre-therapy and Post-therapy performance of individuals with aphasia

.

Figure 1. Performance of subjects on WAB.

Table 4. Mean and SD of subjects with aphasia on various techniques (N=4)

For e.g., the subject A1 produced /...en/ for /pen/ on a confrontation naming task. A few semantic errors were those of neologisms and circumlocutions in the subject A1. However relatively better performance was seen on post-therapy assessment (scores= 44.00) compared to pre-therapy (scores=25.00) for naming. Comparison of pre-therapy (scores=22.00) and post-therapy (scores=44.00) for subject A2 revealed that there was an improvement in the performance of the subject from pre-therapy to post-therapy (scores=44.00) (see Figure 1). However, the performance of subject A2 was relatively poorer in comparison to subject A1 for cueing hierarchy technique. Analysis of the errors made by subject 2 revealed that there were both phonological as well as semantic errors on confrontation naming task before therapy. After therapy using the cueing hierarchy technique it was found that the errors retained in subject A2 were phonological types of errors.

Analysis of scores for subject A3 revealed that the performance was better post-therapy (scores=35) than pre-therapy (scores=20), however, compared to subjects A1 and A2, subject A3 showed severe naming errors thus affecting verbal expression. Naming errors in subject A3 included semantic as well as phonologic type of errors. Both semantic and phonologic types of errors were evident even after cueing hierarchy therapy. Analysis of scores for subject A4 revealed that similar to subjects A1, A2 and A3, there was an improvement in the performance of subject A4 from pre-therapy (scores=55) to post-therapy (scores=83) performance. Subject A4 exhibited the least number of naming errors compared to subjects A1, A2 and A3. It was also observed that a combination of phonologic, orthographic and semantic cues benefitted subject A4 to the maximum extent. Subject A4 revealed improvement in the overall verbal expression compared to the other subjects.

Performance Participants

A1 A2 A3 A4 Performance on WAB (Pre-therapy) 30.2 45.6 15 23.7

Type of Aphasia based on WAB (Pre-therapy) Broca‟s Broca‟s Global Broca‟s Performance on WAB (Post -therapy) 34.5 47.4 17.4 25

Type of Aphasia based on WAB(Post-therapy) Broca‟s Broca‟s Global Broca‟s

Techniques Maximum score Mean SD

Cueing hierarchy Pre-therapy 60 32.00 15.68 Post-therapy 60 53.25 20.88

Facilitation Pre-therapy 60 28.75 6.34 Post-therapy 60 48.50 11.38

Circumlocution induced naming Pre-therapy 60 24.00 7.25 Post-therapy 60 42.25 14.36

Semantic therapy Pre-therapy 60 28.50 13.77 Post-therapy 60 51.00 26.84

Spaced retrieval Pre-therapy 60 24.50 8.22 Post-therapy 60 45.50 18.33


5

Figure 2. Comparison of scores of A1 in pre and post

therapy conditions.


therapy conditions. Note: CH- cueing hierarchy, FT- facilitation treatment, CIN- circumlocution induced naming treatment, ST- semantic treatment, SR- Spaced Retrieval

Overall, it was found that on cueing hierarchy technique; there was an improvement in the performance of all the four subjects with aphasia. Also within a given period of time, subjects appeared to show variable degrees of improvement on cueing hierarchy technique. II. Performance of subjects with aphasia on facilitation treatment: Facilitation technique described by Hickin, Best, Herbert, Howard and Osborne (2002) was applied for the present treatment program for naming. Facilitation is similar to priming—it examines the effect of performing a task once on naming. For example, someone more likely to be able to name a picture accurately was asked to repeat the name of the picture, than if they simply had a second attempt at naming.

Figure 3. Comparison of scores of A2 in pre and post therapy conditions.


therapy conditions.

Words which are functionally useful were selected for facilitation technique. Cues were arranged according to increasing stimulus power and presented until the patient produced the target word. When the patient responded accurately, cues were presented in the order of decreasing stimulus power until the patient again produced the correct response. The cues selected were orthographic, visual and auditory. Pre-therapy scores were considered as the baseline scores. After the baseline scores were established the post-therapy scores were calculated from responses for each session. Table 4 shows pre-therapy scores and total post-therapy scores calculated for a total of twenty sessions.

Analysis of results for pre-therapy and post-therapy condition for the facilitation treatment technique revealed that there was an improvement in the performance of individuals with Aphasia.


6

Table 4 shows that the mean scores on post-therapy (Mean=48.50, SD=11.38) was greater than the pre-therapy scores (Mean=28.75, SD=6.34). A qualitative descriptive analysis of individual data was also done and the results are discussed further. In the case of facilitation treatment, partial cues, phonetic cues, description, demonstration of an action and functional description of the target were used as cues. This means that both phonological as well as semantic cues were used when the cueing hierarchy technique was employed. It was found that Subject A1 produced more number of phonological naming errors than the semantic errors. So, the errors observed were more of phonemic paraphasias than semantic paraphasias.

For e.g., the subject A1 produced /ki__ / for /kita:b/ on a confrontation naming task. A few semantic errors were those of neologisms and circumlocutions in the subject A1. However at the end of the sessions their relatively better performance was seen on post-therapy assessment (Scores=45.00) compared to pre-therapy (scores= 29.00) for naming. Similar to analysis of group means, facilitation treatment was found to show improvement for subject A1 as indicated in Figure 3. Comparison of pre-therapy (scores=31.00) and post-therapy (scores=52.00) for subject A2 revealed that there was an improvement in the performance of the subject from pre-therapy to post-therapy (scores=52.00) (see Figure 4). However, the performance of subject A2 was relatively poorer in comparison to subject A1 for facilitation treatment. Analysis of the errors made by subject A2 revealed that there were both phonological as well as semantic errors on confrontation naming task before therapy. After therapy using the facilitation treatment it was found that the errors retained in subject A2 were phonological types of errors.

Analysis of scores for subject A3 revealed

that the performance was better post-therapy (scores=35.00) than pre-therapy (scores=20.00), however, compared to subjects A1 and A2, subject A3 showed severe naming errors thus affecting the verbal expression. Naming errors in subject A3 included semantic as well as phonologic type of errors. Both semantic and phonologic types of errors were evident even after facilitation treatment. Analysis of scores for subject A4 revealed that similar to subjects A1, A2 and A3, there was an improvement in the performance of subject A4 from pre-therapy (scores=35.00) to post-therapy (scores=62.00) performance. Subject A4 exhibited the least number of naming errors compared to subjects A1, A2 and A3. It was also observed that a combination of phonologic, orthographic and semantic cues benefitted subject A4 to the maximal extent. Subject A4 revealed improvement in the

overall verbal expression compared to the other subjects.

Overall, it was found that on facilitation

treatment, there was an improvement in the performance of all the four subjects with aphasia. Also within a given period of time, subjects appeared to show variable degrees of improvement on facilitation treatment. III. Performance of subjects with aphasia on circumlocution induced naming treatment: Circumlocution induced naming (CIN) treatment described by Francis, Clark and Humphreys (2002) was applied for the present treatment program for naming. CIN therapy refers to accessing a word when the patients themselves circumlocute until they finally access the name by themselves without the aid of external cueing. Table 4 shows pre-therapy scores and post-therapy scores calculated for a total of twenty sessions.

Analysis of results for pre-therapy and post-therapy condition for the circumlocution induced treatment revealed that there was an improvement in the performance of individuals with Aphasia. Table 4 shows that the mean scores on post-therapy (Mean=42.25, SD=14.36) was greater than the pre-therapy scores (Mean=24.00, SD=7.25). A qualitative descriptive analysis of individual data was also done and the results are discussed further. In the case of circumlocution induced naming treatment, partial cues, phonetic cues, description, demonstration of an action and functional description of the target were used as cues. This means that both phonological as well as semantic cues were used when the cueing hierarchy technique was employed. It was found that subject A1 produced more number of phonological naming errors than the semantic errors. So, the errors observed were more of phonemic paraphasias than semantic paraphasias.

For e.g., the subject A1 produced /bi__/ for /billi/ on a confrontation naming task. A few semantic errors were those of neologisms and circumlocutions in the subject A1. However relatively better performance was seen on post-therapy assessment (scores=49.00) compared to pre-therapy (scores=27.00) for naming. Similar to analysis of group means circumlocution induced naming treatment was found to show improvement for subject A1. Comparison of pre-therapy (scores= 24.00) and post-therapy (scores=43.00) for subject A2 revealed that there was an improvement in the performance of the subject from pre-therapy to post-therapy scores. However, the performance of subject A2 was relatively poorer in comparison to subject A1 for circumlocution induced naming treatment. Analysis of the errors made by subject A2 revealed that there were both phonological as well as semantic


7

errors on confrontation naming task before therapy. After therapy using the facilitation treatment it was found that the errors retained in subject A2 were phonological types of errors.

Analysis of scores for subject A3 revealed that the performance was better post-therapy (scores=22.00) than pre-therapy (scores=14.00), however, compared to subjects A1 and A2, subject A3 showed severe naming errors thus affecting verbal expression. Naming errors in subject A3 included semantic as well as phonologic type of errors. Both semantic and phonologic types of errors were evident even after circumlocution induced naming treatment. Analysis of scores for subject A4 revealed that similar to subjects A1, A2 and A3, there was an improvement in the performance of subject A4 from pre-therapy (scores=31.00) to post-therapy (scores=55.00) performance. Subject A4 exhibited the least number of naming errors compared to subjects A1, A2 and A3. It was also observed that a combination of phonologic and semantic cues benefitted subject A4 to the maximal extent. Subject A4 revealed improvement in the overall verbal expression compared to the other subjects.

Overall, it was found that on circumlocution induced naming treatment, there was an improvement in the performance of all the four subjects with aphasia. Also within a given period of time, subjects appeared to show variable degrees of improvement on circumlocution induced naming treatment. IV. Performance of subjects with aphasia on semantic treatment: Semantic treatment described by Howard, Patterson, Franklin, Orchard-Lisle and Morton (1985) was applied for the present treatment program for naming. Word-picture matching, featuring distracters that are semantically related to the target were selected. Davis and Pring (1991) and Hillis and Caramazza (1995) suggested that the target may be spoken or written words, and the task may or may not require a patient to actually produce the word. The rationale here is that these tasks force the patient to focus on the semantic features that distinguish related items, thereby encouraging him or her to relearn the complete semantic description. Semantic tasks include printed and auditory word-to-picture matching, sorting words and pictures by semantic category, and making semantic judgments. The cues selected were orthographic, visual and auditory cues. Pre-therapy scores were considered as the baseline score. After the baseline scores were established the post-therapy scores were calculated from responses in each session. Table 1 shows pre-therapy scores and post-therapy scores calculated for a total of twenty sessions.

Analysis of results for pre-therapy and post-therapy condition for the facilitation treatment technique revealed that there was an improvement in the performance of individuals with Aphasia. Table 2 shows that the mean scores on post-therapy (Mean=51.00, SD=26.84) was greater than the pre-therapy scores (Mean=28.50, SD=13.77). A qualitative descriptive analysis of individual data was also done and the results are discussed further. In the case of semantic treatment, partial cues, phonetic cues, description, demonstration of an action and functional description of the target were used as cues. This means that both phonological as well as semantic cues were used when the cueing hierarchy technique was employed. It was found that Subject A1 produced more number of phonological naming errors than the semantic errors. So, the errors observed were more of phonemic paraphasias than semantic paraphasias.

For e.g., the subject A1 produced /ka…/ for /ka:r/ on a confrontation naming task. A few semantic errors were those of neologisms and circumlocutions in the subject A1. However relatively better performance was seen on post-therapy assessment (scores=46.00) compared to pre-therapy (scores= 27.00) for naming. Similar to analysis of group means, semantic treatment was found to show improvement for subject A1 as indicated in Figure 3. Comparison of pre-therapy (scores=24.00) and post-therapy (scores=43.00) for subject A2 revealed that there was an improvement in the performance of the subject from pre-therapy to post-therapy scores (see Figure 4). However, the performance of subject A2 was relatively poorer in comparison to subject A1 for facilitation treatment (see Figures 3 and 4). Analysis of the errors made by subject 2 revealed that there were both phonological as well as semantic errors on semantic treatment task before therapy. After therapy using the semantic treatment it was found that the errors retained in subject A2 were phonological types of errors.

Analysis of scores for subject A3 revealed that the performance was better post-therapy (scores=17.00) than pre-therapy (scores=10.00), however, compared to subjects A1 and A2, subject A3 showed severe naming errors thus affecting verbal expression (see Figure 4). Naming errors in subject A3 included semantic as well as phonologic type of errors. Both semantic and phonologic types of errors were evident even after semantic treatment. Analysis of scores for subject A4 revealed that similar to subjects A1, A2 and A3, there was an improvement in the performance of subject A4 from pre-therapy (scores=35.00) to post-therapy (scores=62.00) performance (see Figure 5). Subject A4 exhibited the least number of naming errors compares to subjects A1, A2 and A3. It was also observed that a combination of phonologic,


8

orthographic and semantic cues benefitted subject A4 to the maximal extent. Subject A4 revealed improvement in the overall verbal expression compared to the other subjects.

Overall, it was found that on semantic

treatment, there was an improvement in the performance of all the four subjects with aphasia. Also within a given period of time, subjects appeared to show variable degrees of improvement on semantic treatment. Boyle (2004) studied the effect of semantic feature analysis (SFA) treatment on

confrontation naming with anomic aphasia. The findings indicated that confrontation naming of treated nouns improved and generalized to untreated nouns for both participants, who appeared to have different lexical access impairments. Both the participants demonstrated improvement in some aspects of discourse production associated with the confrontation naming SFA treatment. These findings support previous work regarding improved and

generalized naming associated with SFA treatment and indicated a need to examine effects of improved confrontation naming on more natural speaking situations. The findings of the present study also support Boyle (2004) as an improvement was observed in the performance of individuals with aphasia on semantic treatment technique, however the treatment efficacy and generalization cannot be explained through the present study.

V. Performance of subjects with aphasia on spaced retrieval (SR) therapy: Spaced retrieval therapy described by Bourgeois, Camp, Rose, Blanche, Malone and Carr (2003) was applied for the present treatment program for naming. Spaced retrieval treatment approach was developed to facilitate recall of information by individuals with anomia. Essentially spaced retrieval therapy was considered as an errorless learning procedure that can be used to facilitate recall of a variety of information. Spaced retrieval therapy was administered by gradually increasing the interval between correct recall of target items. Spaced retrieval therapy opined to be an alternative for managing naming impairment. The cues selected were orthographic, visual and auditory cues. Pre-therapy scores were considered as the baseline score. After the baseline scores were established the post-therapy scores were calculated from responses in each session. Table 4 shows pre-therapy scores and post-therapy scores calculated for a total of twenty sessions.

Analysis of results for pre-therapy and post-therapy condition for the spaced retrieval technique revealed that there was an improvement in the performance of individuals with Aphasia. Table 4 shows that the mean scores on post-therapy (Mean=45.50, SD=18.33) was greater than the pre-therapy scores (Mean=24.55, SD=18.33). A

qualitative descriptive analysis of individual data was also done and the results are discussed further. In the case of facilitation treatment, partial cues, phonetic cues, description, demonstration of an action and functional description of the target were used as cues. This means that both phonological as well as semantic cues were used when the spaced retrieval technique was employed. It was found that Subject A1 produced more number of phonological naming errors than the semantic errors. So, the errors observed were more of phonemic paraphasias than semantic paraphasias.

For e.g., the subject A1 produced /b】…./ for

/bus/ on a confrontation naming task. A few semantic errors were those of neologisms and circumlocutions in the subject A1. However relatively better performance was seen on post-therapy assessment (scores=43.00) compared to pre-therapy (scores=26.00) for naming. Similar to analysis of group means, spaced retrieval treatment was found to show improvement for subject A1. Comparison of pre-therapy (scores=24.00) and post-therapy (scores=51.00) for subject A2 revealed that there was an improvement in the performance of the subject from pre-therapy to post-therapy scores. However, the performance of subject A2 was relatively poorer in comparison to subject A1 for spaced retrieval treatment. Analysis of the errors made by subject A2 revealed that there were both phonological as well as semantic errors on confrontation naming task before therapy. After therapy using the spaced retrieval treatment it was found that the errors retained in subject A2 were phonological types of errors.

Analysis of scores for subject A3 revealed that

the performance was better post-therapy (scores=22.00) than pre-therapy (scores=14.00), however, compared to subjects A1 and A2, subject A3 showed severe naming errors thus affecting verbal expression. Naming errors in subject A3 included semantic as well as phonologic type of errors. Both semantic and phonologic types of errors were evident even after spaced retrieval treatment. Analysis of scores for subject A4 revealed that similar to subjects A1, A2 and A3, there was an improvement in the performance of subject A4 from pre-therapy (scores=34.00) to post-therapy (scores=66.00) performance. Subject A4 exhibited the least number of naming errors compares to subjects A1, A2 and A3 (Figures 2, 3, 4 and 5). It was also observed that a combination of phonologic, orthographic and semantic cues benefitted subject A4 to the maximal extent. Subject A4 revealed improvement in the overall verbal expression compared to the other subjects.

Overall, it was found that on spaced retrieval

treatment, there was an improvement in the performance of all the four subjects with aphasia.


9

Also within a given period of time, subjects appeared to show variable degrees of improvement on spaced retrieval treatment. Limited data are available to address whether SR could be used with disorders other than dementia. Recent work by Bourgeois and colleagues (2003) revealed that they used the SR paradigm in individuals with TBI, another problem that is accompanied by memory difficulties (Melton & Bourgeois, in press). Only one published study reported using SR with a person with aphasia. Brush and Camp (1998b) employed SR with two persons defined as having had a stroke (one of whom appeared to have aphasia) and seven individuals with dementia. They found that SR facilitated recall of clinician‟s name, patient‟s room number, and a compensatory technique for naming difficulty. The persons with stroke learned the clinician‟s name faster than the persons with dementia (three vs. nine sessions). However, only one aphasic individual participated in the study, and only limited documentation of the study design was provided.

The primary purpose of the present study was

to develop a treatment manual for anomic aphasia. As a preliminary try out of the manual, it was administered on a small sample of four participants. The therapy programs utilized a combined semantic and phonological treatment approach to improve word naming through the use of various techniques which used cues, descriptions, and cognitive related tasks. The general naming condition within which these techniques were carried out was naming at the single-word level. Individual performance for the treatment program, results of assessments, as well as qualitative analysis for each participant are discussed below. Group performance and results are also discussed in the sections below.

Previous studies have shown positive effects

of cueing treatments on naming. In many single and multiple case studies the long-term effect of semantic cueing treatment on naming was established, not only on trained but also on untrained items (Drew & Thompson, 1999; Coelho, McHugh & Boyle, 2000; Wambaugh, 2003). The positive effects of phonological techniques have also been noted (Hickin et al., 2002). In a review of word finding therapy, Nickels (2002) concluded that semantic and phonological techniques are effective, and suggested that a combination of both may prove to be most effective. As orthographic cues are found to assist in retrieving the phonological word form, these cues were also used in treatment and better response to phonologic, orthographic and semantic cues were found in the present study. The effect of treatment based on orthographic cues is reported to be equally effective as (Hickin et al., 2002) or more effective than (Basso, Marangolo, Piras & Galluzzi, 2001) treatment based on phonological cues. Use of orthographic cues is found to be effective in

individuals with better written than spoken naming. Subject A4 was found benefit maximum with the use of orthographic cues along with phonologic and semantic cues. This could be because once grapheme-phoneme conversion is relearned, the person with aphasia may use the available orthographic information to generate his own phonological cues. Studies also support that generalization to untrained words is expected with use of multiple cues while training (Nickels, 2002).

The effect of cueing treatments on verbal

communication is unknown. It is often implicitly assumed that improved performance on a naming task brings about improved verbal communication, but this is not well supported by research. Of over 50 studies investigating the efficacy of impairment-oriented word finding treatment, only a handful explicitly consider generalization to spontaneous speech, with contradictory results (Boyle & Coelho, 1995; McNeil, Robin & Schimdt, 1997; Franklin, Buerk & Howard, 2002; Doesborgh, van de Sandt-Koenderman, Dippel, van Harskamp, Koudstaal & Visch-Brink, 2003). Although much is known about the efficacy of different cueing techniques on naming, it is not fully understood which cues are suitable for which individuals. There is no simple one-to-one relationship between the loci of impairment and the cues that will facilitate word finding: semantic techniques can improve naming for individuals with good semantic processing (Nickels & Best, 1996) and phonological tasks can improve naming for individuals with semantic impairments (Raymer, Thompson, Jacobs & LeGrand, 1993; Nickels, 2002).

Conclusions

The treatment manual containing five well known therapy techniques for treatment of anomia was administered on four subjects with aphasia. The techniques included cueing hierarchy, facilitation, circumlocution induced naming technique, semantic treatment and spaced treatment techniques. The pre-therapy and post-therapy scores of these individuals with aphasia were compared qualitatively. Analysis of scores across individuals with aphasia revealed that subject A4 showed greater improvement than the subjects A1, A2 and A3. This could be attributed to the less severe condition of anomia in subject A4. Similarly subject A3 showed lesser improvement compared to the other subjects with aphasia as subject A3 had greater severity of aphasia compared to the other types of aphasias in subjects A1, A2 and A4. This indicates that the type of aphasia and severity are yet other factors contributing to improvement of anomia in aphasic conditions. The treatment techniques that benefitted most subjects with aphasia were the cueing hierarchy technique and the semantic treatment technique. However, the

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1486764/#R19






10

efficacies of these techniques over the other techniques are required to be investigated on a larger population with anomia. Hence, the present study was only a preliminary investigation of the treatment techniques for anomia. The above techniques adopt both phonological treatment and semantic treatment techniques in a broader sense and need to be used according to the nature of deficit in the individuals with anomia. To conclude, specific and theoretically motivated treatment methods can result in improvement in individuals with naming difficulty, however a larger sample size and inclusion of generalization ability to spontaneous speech will facilitate beneficial management strategies for anomic aphasias.

The treatment manual can be used for

individuals with anomia in order to regain their naming ability through restitutive or substitutive approaches. However, the generalization of naming using the above techniques could not be assessed hence, generalization effect needs to be studied and then the technique can be used for treatment of anomia. Even though single subject designs are excellent for demonstration of experimental control, only limited assumptions can be drawn about external validity. Therefore, it is important that the effects demonstrated here be investigated further.

Although the present study does not prove the efficacy of any of the treatment techniques, the treatment-specific effects found at the impairment level suggest that there may be routes that lead to improved verbal communication: a semantic route and a phonological route and a further analysis of the data and increase in sample size may support to explain the efficacy of a specific treatment technique.

References Bandur, D. L., & Shewan C. M. (2001). Language-

orientated treatment: A psycholinguistic approach to aphasia. In R. Chapey, (Eds.), Language intervention strategies in aphasia and related neurogenic communication disorders (7th Edn.), (pp. 629-656). Williams and Wilkins, Lippincott Maryland.

Barton, M., Maruszewski, D., & Urrea. (1969). Variations of stimulus context and its effect on word finding ability in aphasics. Cortex, 5, 351-364.

Basso, A., Marangolo, P., Piras, F., & Galluzzi, C. (2001). Acquisition of new "words" in normal subjects: A suggestion for the treatment of anomia. Brain and Language, 77, 1, 45-59.

Benson, D. F. (1979). Neurologic correlates of anomia. Studies in Neurolinguistics. New York: Academic Press.

Benson, D. F., & Geschwind, N. (1985). The aphasias and related disturbances. In A. B. Baker and R. J. Joynt (Eds.), Clinical neurology (pp. 1-34). New York: Harper & Row.

Bourgeois, M. S., Camp, C., Rose, M., Blanche, W., Malone, M., & Carr, J. (2003). A comparison of training strategies to enhance use of external aids by

persons with dementia. Journal of Communication Disorders, 36, 5, 361–378.

Boyle, M. (2004). Semantic feature analysis treatment for anomia in two fluent aphasia syndromes. American Journal of Speech-language Pathology, 13 (3), 236–249.

Boyle, M., & Coelho, C. (1995). Application of semantic feature analysis as a treatment for aphasic dysnomia. American Journal of Speech-Language Pathology, 4, 94–108.

Brown, J. (1972). Aphasia, apraxia and agnosia. Springfield IL: Charles C Thomas.

Brush, J. A., & Camp, C. J. (1998). Using spaced retrieval as an intervention during speech-language therapy. Clinical Gerontologist, 19, 1, 51–64.

Coelho, C. A., McHugh, R. E., & Boyle, M. (2000). Semantic feature analysis as a treatment for aphasic dysnomia: A replication. Aphasiology, 14(2), 133–142.

Davis, A., & Pring, T. (1991). Therapy for word finding deficits: More on the semantic and phonological approaches to treatment with anomic patients. Neuropsychological rehabilitation, 1, 135-145.

Doesborgh, S. J. C., van de Sandt-Koenderman, W. M. E., Dippel, D. W. J., van Harskamp, F., Koudstaal, P. J., & Visch-Brink, E. G. (2002). The impact of linguistic deficits on verbal communication. Aphasiology, 16, 413–423.

Drew, R. L., & Thompson, C. K. (1999). Model-based semantic treatment for naming deficits in aphasia. Journal of Speech, Language, and Hearing Research, 42, 4, 972.

Francis, D. R., Clark, N., & Humphreys, G. W. (2002). Circumlocution-induced naming (CIN): A treatment for effecting generalization in anomia? Aphasiology, 16, 243-260.

Franklin, S., Buerk, F., & Howard, D. (2002). Generalised improvement in speech production for a subject with reproduction conduction aphasia. Aphasiology, 16, 1087-1116.

Geschwind, N. (1967). The varieties of naming errors. Cortex, 3, 97-112.

Goodglass, H., & Kaplan, E. (1983). Boston Diagnostic Aphasia Examination. (2nd Ed.), Media, PA: Williams and Wilkins.

Head, H. (1926). Aphasia and kindred disorders of speech. New York: Macmillan.

Helm-Estabrooks, N., & Ramsburger, G. (1986). Treatment of agrammatism in long term Broca‟s aphasia. British Journal of Disorders of Communication, 21, 39-45.

Hickin, J., Best, W., Herbert, R., Howard, D., & Osborne, F. (2002). Phonological therapy for word-finding difficulties: A re-evaluation. Aphasiology, 16, 981-999.

Hillis, A., & Caramazza, A. (1995). The compositionality of lexical semantic representation: Clues from semantic errors in object naming. Memory, 3, 333-358.

Howard, D., Patterson, K. E., Franklin, S., Orchard-Lisle, V., & Morton, J. (1985). The treatment of word retrieval deficits in aphasia: A comparison of two therapy methods. Brain, 108, 817–829.

Karanth, P., Manjula, R., Geetha, Y.V., & Prema, K, S. (1999). Production of Language Training Materials in Major Indian Languages. A UNICEF funded project submitted to All India Institute of Speech and Hearing, Mysore.

Kertesz, A. (1982). Western Aphasia Battery. New York: Grune & Stratton.


11

Kremin, H. (1988). Naming and its disorders. In F. Boiler, & J. Grafman (Eds.), Handbook of neuropsychology. New York: Elsevier.

Linebaugh, C. W., & Lehner, L. (1977). Cueing hierarchies and word retrieval: A therapy program. In R. H. Brookshire (Eds.), Clinical aphasiology conference proceedings. Minneapolis, MN: BRK.

Luria, A. R. (1976). Basic problems of neurolinguistics. The Hague: Mouton.

McNeil, M. R., Robin, D. A., & Schmidt, R. A. (1997). Apraxia of speech: definition, differentiation, and treatment. In M. R. McNeil, (Eds.), Clinical management of sensorimotor speech disorders. New York: Thieme.

Melton, A., & Bourgeois, M. (in press). Training compensatory memory strategies via the telephone for persons with TBI. Aphasiology.

Nickels, L. (2002). Therapy for naming disorders: Revisiting, revising and reviewing. Aphasiology, 16(10/11), 935–979.

Nickels, L., & Best, W. (1996). Therapy for naming disorders (Part II): Specifics, surprises, and suggestions. Aphasiology, 10, 2, 109–136.

Raymer, A. M., Thompson, C. K., Jacobs, B., & LeGrand, H. R. (1993). Phonological treatment of naming deficits in aphasia: model-based generalization analysis. Aphasiology, 7, 1, 27–53.

Raymer, A. M., & Ellsworth, T. A. (2002). Response to contrasting verb retrieval treatments: A case study. Aphasiology, 16, 1031–1045.

Schuell, H. M., Jenkins, J. J., & Carroll, J. B. (1962). A factor analysis of the Minnesota Test for the differential diagnosis of aphasia. Journal of Speech and Hearing Research, 5, 349-369.

Wambaugh, J. L. (2003). A comparison of the relative effects of phonologic and semantic cueing treatments. Aphasiology, 17(5), 433–441.

Weinstein, E. A., & Keller, N. J. A. (1963). Linguistic patterns of misnaming in brain injury. Neuropsychologia, 1, 79-90.


12

Cognitive Communicative Assessment Protocol for Persons with Dementia in Malayalam (CCAPD-M)

Anjana A. V.1 & Jayashree C. Shanbal 2

Abstract

Dementia is an umbrella term for a group of pathological conditions or syndromes that occur with unsuccessful five areas of mental activity (i) Language (ii) Memory; (iii) Visuospatial skills; (iv) Emotion or personality and (v) Cognition (ex: abstraction, calculation, and judgment). Persons with Dementia (PWD) have trouble producing linguistic information because they have trouble thinking and generating and ordering ideas, in part because information processing capabilities of declarative and working memory systems are disturbed. The aim of the current study was to develop an assessment protocol for PWD in Malayalam that can be useful for Speech-Language Pathologist. The developed assessment protocol called „Cognitive Communicative Assessment Protocol for Persons with Dementia - in Malayalam (CCAPD-M)‟ consists of six domains - (i) Memory (ii) Linguistic comprehension (iii) Linguistic expression (iv) Problem solving (v) Organization and (vi) Visuo-spatial construction with a total of twenty four sub domains. CCAPD-M was administered on sixty normal individuals (two groups i.e. between 40-60years and between 60-80years) and ten PWD. Results of the study revealed significant deterioration in the performance of normal individuals as the age advanced. PWD performed poorly on delayed recall, generative naming, picture naming, categorization, problem solving etc. compared to other domains. This indicated that cognitive decline as well as semantic decline was more in PWD, while the syntactic abilities were better. Executive dysfunction and working memory deficits can account for many of the linguistic deficits in dementia.

Key words: dementia, cognition, aging, alzheimer’s disease, parkinson’s disease

he term 1dementia refers to the clinical syndrome in which there is an impairment of memory, deterioration of intellect sufficient to

interfere with social or occupational functioning, unclouded state of consciousness, the presence of an organic factor related to the disturbance, as well as one of the following impairment of abstract thinking, impairment of judgement, personality change, impairment of other cortical functions as evidenced by the presence of aphasia, apraxia, agnosia, or constructional difficulty (DSM III, American Psychiatric Association, 1980). Dementia is an umbrella term for a group of pathological conditions or syndromes that occur with unsuccessful five areas of mental activity viz. (i) Language (ii) Memory (iii) Visuospatial skills (iv) Emotion or personality and (v) Cognition (e.g., abstraction, calculation, and judgment). The prevalence of dementia in Kerala was found to be 33.6 per 1000. Alzheimer's disease (AD) was the most common type (54%) followed by vascular dementia (39%), and 7% of cases were due to other causes such as infection, tumour and trauma (Shaji & Bose, 2005).

Communication and cognition in Dementia: Persons with dementia have trouble producing linguistic information because they have trouble thinking and generating and ordering ideas, in part because information processing capabilities of declarative and working memory systems are disturbed. These individuals have difficulty in comprehending language because of deficits in 1e-mail: [email protected]; 2Lecturer in Language Pathology, AIISH, Mysore, [email protected]

perception, recognition, attention, inferencing, memory, and degradation of knowledge.

Memory problems are the defining feature of dementia and is the first sign of cognitive decline. Sensory memory: Many of the functional deficits that results from sensory problems appear in the late stages of dementia, and evidence exist for preserved sensory processing at early stages of perceptual priming tasks (Salmon & Fennema-Notestine, 1996) and on perceptually based repetition priming tasks (Fleischman, Gabrieli, Reminger, Rinaldi, Morrell, & Wilson, 1995). Working memory : Working memory processes are particularly vulnerable to the effects of dementia, which may be due to the failure in the executive control system or as with sensory memory, due to types of tasks used to document the process. There is evidence of reduced memory span and short term memory capacity in AD (Morris, 1986) and performance deficits in short term memory tasks with divided attention conditions (Morris, 1996). Access to semantic memory has been identified as the cause of working memory deficits in persons with AD, whereas disrupted inhibitory processes are thought to explain working memory in persons with Parkinson‟s disease (PD). Declarative memory and explicit memory: Semantic memory is the most central of all cognitive processes and is fundamental to language production and comprehension, reading, and writing, object and face perception. Persons with AD have significant difficulty in recalling recent and current events but demonstrate good retrieval of childhood memories. Remote memory performance was correlated with a

T

Cognitive communicative assessment protocol for PWD-Malayalam

13

measure of semantic fluency. There are differences across dementia etiologies in recognition versus recall tasks. Recall memory is more impaired than recognition memory in dementia. Nondeclarative or procedural memory: Procedural memory is relatively preserved in AD, but impaired in Parkinson‟s disease (PD) (Zgaljardic, Borod, Foldi & Mattis, 2003) and Vascular Dementia (VaD) (Libon, Bogdanoff, Cloud, Skalina, Giovannetti & Gitlin, 1998). Executive function: Executive dysfunction is observed to some extent in all dementias, and accounts, to certain extent for the difficulty in performing instrumental activities of daily living. Across dementia types, executive functioning for complex tasks and problem solving declines with increasing dementia severity. Initiation and planning problems are the earliest signs of executive dysfunction in PD (Zagaliadric et al., 2003). The executive dysfunction deficits seen early in Fronto temporal dementia (FTD) are also more severe than those in AD (Rosen, Hartikainen, Jagust, Kramer, Reed & Cummings, 2002). Abstraction or problem solving process: An impairment of the abstract thinking and problem solving and a deficient ability to shift or maintain set is often a prominent clinical feature of Dementia of Alzhemier‟s Type (DAT). These deficits are usually ascribed to the neuropathologic changes that occur in the prefrontal association cortex of patients with DAT. Although these deficits may emerge at different stages in various patients, they are invariably present by the middle stages of the disorder. Freedman and Oscar-Berman (1986) have reported that the impaired performance of the DAT patients on these tasks presumably results from deficiency in cognitive flexibility that is required to alternate responses and shift mental set. Visuospatial processes: Visuoconstructional tasks (e.g., clock drawing) rely in part on linguistic output and comprehension, as well as praxis, memory and visuo-motor coordination. Adequate performance also depends on reasonably intact attention. Effect of frontal system damage can affect performance. According to Albert, Blacker, Moss, Tanzi and McArdle (2001) some vision related cognitive deficits should be investigated early in dementias. Visual perceptual as well as visual constructional deficits are seen in patients with dementia (Bayles & Kaszniak, 1987). Language characteristics in dementia: The presence of memory impairment in any form will interfere with language production and comprehension to some extent, and will vary over the course of illness. Semantic and pragmatic language systems have been found to be more impaired in

dementia than do syntax and phonology. A number of studies have documented that an increase in the severity of the language deficit parallels an increase in severity of dementia (Albert, 1981; Bayles, 1982; Cummings, Benson, Hill & Read, 1985). The most frequently cited communication problems in the literature include difficulty with verbal memory, word finding (i.e., anomia), disruptive vocalizations, and understanding of spoken language (i.e., auditory comprehension deficits). Phonologic impairments in dementia: The selection and sequencing of individual phonemes for speech production remain intact throughout most of the duration of AD. Violations of phonotactic constraints of native language and errors in prosody rarely occur (Appel, Kertesz & Fishman, 1982). Semantic impairments in dementia: The most common early symptoms of dementia are word finding, naming, and verbal description difficulties, due to semantic memory impairment. Difficulty in naming or word retrieval has been observed to be the most obvious early symptom of dementia, regardless of cause, and has been found to occur before other language changes associated with the syndrome are measurable. Studies investigating categorization skills in adults with dementia have revealed that these individuals show significant deterioration in the structure and/or contents of semantic and conceptual knowledge as compared to their peer age matched healthy cohorts. Grossman, Smith, Koenig, Glosser, Rhee and Dennis (2003) found significant difficulty with rule based semantic categorization of familiar object descriptions is found in AD and FTD, although similarity categorization does not differ from control subjects. Syntactic deficits in dementia: Syntax appears to be less impaired when the context cues or structures the syntactic task. A working memory deficit in AD would contribute to syntactical errors of processing. Kemper, LaBarge, Ferraro, Cheung and Storandt, (1993) found that declines in sentence length, grammatical complexity, verbal fluency, and propositional content were seen in the interview transcript of persons with mild dementia compared to persons without dementia. Pragmatic deficits: Pragmatics is assessed through discourse production. Some aspects of discourse are clearly impaired in DAT though the mild and moderate DAT patients take conversational turns when appropriate and often produce socially ritually parts of the conversations with appropriate timing affect and linguistic structure. These deficits could be secondary to existing and document problems such as anomia, decreased attention for poor memory, etc.


14

Oral reading: The ability to read aloud remains relatively intact until the late stages of AD. Single word reading ability was evident on the Functional Linguistic communication inventory (Bayles & Tomoeda, 1994), even in late stage AD. Written language production: Writing severity has been correlated with dementia severity across a variety of tasks: spelling performance (Aarsland, Hoien, Larsen & Oftedal, 1996), single written sentence performance (Kemper, LaBarge, Ferraro, Cheung & Storandt, 1993). Auditory comprehension: Bayles and Tomoeda (1994) suggested that individuals with severe dementia occasionally answer multiple-choice and yes-no questions accurately, although overall accuracy is low. They also found that patients at this stage can follow one step commands with 60% accuracy. Reading comprehension: Reading deficits may be attributed to memory encoding deficits, difficulty keeping multiple ideas in mind and making inferences, or long term and semantic memory deficits. Working memory deficits are thought to account for the deficits in reading comprehension of sentences (Kempler, Van Lancker & Read, 1988). Assessment: Assessment of cognitive communicative function serves many purposes (Tomoeda, 2001). These include detection of dementia, identification of cognitive communicative deficits, identification of retained abilities, establishment of a baseline of cognitive communicative functioning for planning intervention and measuring response to treatment, counselling caregivers and predicting skills vulnerable to future decline.

A complete evaluation for dementia should include (1) a careful and through case history; (2) neurologic and medical diagnostic studies and examination; (3) behavioral assessment; and (4) communication assessment. There are a variety of tools available for assessment, including formal and informal measures, and observational and interview protocols.

However, the role of Speech-Language

Pathologist (SLP) in the assessment of individuals with dementia remains ill-defined. American Speech, Language and Hearing Association (2005) issued a position statement which highlights the SLPs‟ role in identification/assessment of dementia. SLPs have a primary role in screening, assessment and treatment, as well as a role in caregiver training and counselling. Appropriate and specific assessment and therapy material for people with dementia are limited.

Most of the neuropsychological tests developed for persons with dementia focus on cognitive skills. The rest of them focus on memory and executive functioning. There are only very few test materials available for assessing the communicative deficits in persons with dementia. There are few assessment tools available in western countries such as Functional Linguistic Communication Inventory (FLCI), ABCD (Arizonna Battery for Communication Disorders), COMFI (Communication Outcome Measure of Functional Independence) etc. No suitable tests are available for dementia in Indian context. Cultural and educational background can influence the performance of persons with dementia in various cognitive and linguistic tasks. So, we need to have appropriate tests to identify the persons with dementia in Indian context. Thus the aim of the study was to develop an assessment protocol for cognitive communicative deficits of persons with dementia and to test the efficacy of the protocol using normal and dementia population.

Method

The aim of the study was to develop a cognitive communicative assessment protocol for persons with dementia in Malayalam (CCAPD-M).

Participants: Broadly the participants for the study included two groups: Group1- control group: normal individuals; Group 2- clinical group: individuals with dementia (IWD). Group 1 (Control group): The participants in the control group were divided into two groups, older adults (in the age range of 40-60 years) and geriatrics (in the age range of 60 years and above). Thirty normal participants were included in each of the two groups. These participants were selected based on the following inclusion criteria: native Malayalam speakers, with no history of neurological or psychiatric illness or of alcoholism or drug abuse, with at least 10 years of education in Malayalam were included for the study. All the participants were initially assessed using Mini Mental Status Examination (MMSE) (Folstein, Folstein & McHugh, 1975). It was found that all the individuals in this group passed with a score of above 23 which indicated that they do not have any cognitive impairment. Group 2: (Clinical group: Individuals with dementia): The clinical group consisted of ten individuals with dementia (IWD) with varying severity. Individuals with dementia were identified through local hospitals, dementia associations and other dementia clinics. These participants were native Malayalam speakers with adequate hearing and vision (corrected) and who had a history of gradual deterioration in cognitive abilities. All the


15

participants in the clinical group were also evaluated for their mental status on Mini Mental Status Examination (Folstein, Folstein & McHugh, 1975) and all failed with a score of below 23 which indicated presence of cognitive impairment.

Test material and Procedure for testing: The study was conducted in two phases: Phase 1: Development of the assessment tool Phase 2: Test administration Phase 1: Development of the assessment tool

This cognitive communicative assessment protocol in Malayalam (CCAPD-M) comprised of 6 main domains which were memory, linguistic comprehension, linguistic expression, problem solving, organization and visuo-spatial construction.

These 6 domains consisted of 24 subtests. The

subtests were selected from different language tests. The test was developed by adopting various other tools such as the Dementia Assessment Battery in Kannada (Sunil & Shyamala, 2009), Arizona Battery of Communication Disorders for Dementia (Bayles & Tomoeda, 1993), and Cognitive-Linguistic Assessment Protocol (CLAP; Kamath, 2002).

Developed assessment material was given for

rating. Rating was done by 20 postgraduate SLPs on the basis of 8 parameters on a 3 point scale (0-poor, 1- average, 2- good). Items that had 90% agreement across the raters were considered for the test protocol. All the parameters were rated above 90%.

Domain-1: Memory: This domain consisted of four subtests within it. Total score of this domain was 50. Episodic memory: This subtest consisted of 10 items. The questions may be given orally or in writing. Each correct answer was assigned with a score of 1 and for incorrect answer a score of 0 was given. The maximum possible score of this subtest is 10. Working memory: This subtest consisted of 10 items, 5 in each of digit forward task and digit backward task. In the digit forward task, the list of 3- 4-, 5-, 6-, 7- digits were presented to the participants and they were instructed to hear and repeat those numbers in the same order. In the digit backward task, the list of digits was presented to the participants and participants were instructed to hear and repeat those numbers in the reverse order. Every correctly repeated sequence was assigned a score of 1. The maximum possible score of this subtest was 10. Semantic memory: This subtest consisted of two tasks: co-ordinate naming and super ordinate naming. In co-ordinate naming, participants were instructed to name two items for the given activity. Each correct response was assigned a score of 1and the total score was 5. In super-ordinate naming, the participants were given a list of items belonging to a particular

class and they were instructed to identify the class/ category to which the given items may be classified. A score of 1 was assigned for each correctly named class. The maximum possible score of this subtest was 10. Delayed story recall: The participants were presented a story and were asked to answer five questions after 45 minutes of the presentation of story. Each correct response were assigned score of 2. The maximum possible score of this subtest was 10. Domain-2: Linguistic expression: Six subtests are included in this domain. The total maximum score of this domain was 50. Picture naming: In this subtest, participants were presented with 10 pictures and were asked to name the pictures. Each correct response was assigned score of 1 and mild paraphasias were accepted. The maximum possible score of this subtest was 10. Generative naming: In this subtest, participants were asked to name as many animals as possible in one minute time. Each response was assigned 0.5 score. The maximum score of this subtest was 10. Sentence completion: This subtest consisted of 5 items and the participant was instructed to fill in the blanks with suitable answer. Each correct response was assigned with a score of 1. The maximum score for this subtest was 5. Responsive speech: In this subtest, participants were instructed to answer 5 questions with suitable answers. Each correct response was assigned a score of 1. The maximum score of this subtest was 5. Spontaneous speech: In this test, participants were asked to tell about him/her and his/her family. Both information content and fluency in patient‟s response were checked. The responses were rated on a 5 point scale (with 1 being least and 5 being maximum score) for both information content and also on fluency aspects. The maximum possible score was 20. Repetition: In this subtest, the participants were asked to repeat the given words and sentences. Clinician repeated items once if the patient asked or did not seem to hear. Minor errors in articulation were scored as correct. Take 1 point off for errors in order of word sequence or for each literal paraphasia. Domain-3: Linguistic comprehension: The third domain of CCAPD-M was linguistic comprehension and this domain consisted of four subtests within it. The maximum score of this domain was 50. Comparative questions: This subtest consisted of 5 items which were presented to the participants and they were instructed to answer either yes or no. Each correct answer was assigned a score of 2 and the maximum possible score is 10. Following commands: This subtest consisted of 5 items which were arranged in increasing complexity. Participants were instructed to listen and follow the commands given. Score for each command is given


16

in the test material. The maximum possible score was 20. Reading comprehension of sentences: This subtest consisted of 5 items, which were presented to the participants in orthographic mode and the participants were asked to choose the correct answer from four answers. Each correct response was assigned a score of 2 and the maximum possible score was 10. Reading commands: This subtest consisted of 5 commands and were presented to the participants in orthographic mode and the participants were instructed to read and follow the commands. Score for each command is given in the test material and the maximum possible score was 10. Domain-4: Problem solving: This domain consisted of five subtests that assess reasoning abilities to aid in problem solving. The maximum possible score was 60. Sentence formulation: This is a word order unscrambling task to form a grammatically correct sentence. Each correct response was assigned a score of 2. The maximum score of this subtest was 10. Predicting the outcome: In this subtest a situation was given and the outcome of that situation had to be predicted by the participants. For each correct response a score of 2 was assigned. The maximum possible score of this subtest was10. Comparing and contrasting two objects: The participants had to give one similarity and one difference between a pair of objects named. For each correct response a score of 2 was assigned. The maximum possible score of this subtest was 10. Predicting cause of a described situation: A situation was given and the participants had to predict the cause of that situation. For each correct response a score of 2 was assigned. The maximum possible score of this subtest was 10. Answering why questions: Each correct response was assigned a score of 2. The maximum possible score of this subtest was 10. Sequential task analysis: The steps involved in carrying out a named task, was required to be listed by the participants. The participants were required to analyze the task into at least four steps for a full score to be given. The maximum possible score was 10. Domain- 5: Organization Categorization abilities: Participants have to select the item which is in the same category of the already told item. Each correct response in this domain carries a score of 2. The maximum possible score of each subtest was 10. Analogies: This task consisted of items to test ability to recognize word concept to meet task demands. This task also involves logical reasoning processes. Each correct response in this domain carries a score

of 2. The maximum possible score of each subtest was 10. Domain – 6: Visuospatial construction Generative drawing: This subtest consisted of 2 items, where the participants were asked to draw pictures of eight items free handedly and the total score of this subtest was 30. Figure copying: This subtest consisted of 2 figures and was presented to the participants and the participants were instructed to copy the figures. Each correct response was assigned a score of 5 and the maximum possible score was 10. Phase 2: Test administration

A Pilot study was carried out in which the developed protocol was administered on 5 participants in the age group of 40-60years, 5 participants in the age group of 60-80years, and 5 participants with dementia. Further revision of the test items was done due to lengthy test duration and ease of performance. By following the above procedure the final version of Cognitive Communicative Assessment Protocol for Persons with Dementia in Malayalam was prepared.

All the participants were first screened using

MMSE-Malayalam. After the above preliminary assessment procedures, the developed assessment protocol CCAPD-M was administered. All the participants were tested in a quiet room wherein they were seated comfortably. The administration took around one and a half to two hours for normal individuals in a span of two sessions and their responses were quantitatively recorded. IWD were tested with breaks in between. They took around 2-4 sessions for the test administration, each session was around 40 minutes. Initially IWD were very uncooperative to the test administration. For mild group of individuals the test administration was easier. IWD with moderate severity was used to topic shift frequently during the administration of test material. All of them were very reluctant to answer the questions, even though they know the correct answer. It took lot of effort on the part of examiner to come out with answer. All the responses were recorded for proper analysis.

All the participants scores were coded and was

analyzed using SPSS-16version software. Raw scores and percenatge scores were calculated for each individual and for each subtest. Mean and standard deviation for each age group, for each subtest, each domain and for the clinical group was also calculated. For the group wise comparison MANOVA (Multi Variate Analysis of Varience) was done i.e. between two age groups and control and clinical group. Within age groups comparison was done across domains using Friedman test to find any significant difference in performance across domains


17

Table 1. Mean (in %) and Standard Deviation (SD) scores for all sub domains across age groups for both the groups

Domain Subtest

Control group Clinical group Younger group1

(N=30) Older group 2

(N=30) Group 3 (N=5)

Mean SD Mean SD Mean SD

Memory

MEPI 98.67 3.46 91.33 7.30 54.00 18.17 MWR 72.67 7.85 50.17 13.03 26.00 11.40 MSEM 98.83 2.15 84.67 15.92 36.00 15.17 MDEL 98.67 3.46 73.00 22.15 24.00 16.73 MEM 93.53 2.96 77.20 13.38 35.20 14.87

Linguistic expression

LEPIC 99.00 30.05 77.33 19.99 18.00 14.83 LEGN 98.33 3.79 66.00 18.68 16.00 11.40 LESC 98.67 5.07 78.00 14.24 68.00 26.83 LERS 98.00 6.10 79.33 17.80 72.00 22.80 LESS 99.33 2.54 71.00 15.61 52.00 26.83 LERT 98.00 80.05 65.33 13.83 38.00 13.04 LETO 97.33 7.15 71.67 15.48 34.80 15.21

Linguistic comprehension

LCCQ 98.00 6.10 75.33 17.95 36.00 16.73 LCFC 98.50 4.38 69.50 15.78 34.00 15.17

LCRCS 97.33 6.91 77.33 22.73 52.00 17.89 LCRCC 96.33 9.99 77.33 22.73 56.00 21.91 LCTO 97.73 6.10 74.60 16.16 42.40 16.76

Problem solving

PSSF 98.67 5.07 81.33 16.55 48.00 17.89 PSPO 99.33 3.65 80.67 17.80 40.00 24.49 PSCC 98.67 5.07 77.33 18.18 48.00 26.83 PSPC 98.00 6.10 78.00 18.46 48.00 17.89 PSY 100.00 0.00 79.33 17.80 52.00 17.89

PSSA 97.33 6.91 77.33 19.64 40.00 24.49 PSTO 98.56 3.78 78.77 17.67 45.60 20.56

Organization OC 96.67 9.22 73.00 22.15 44.00 16.73 OA 96.67 9.22 70.33 24.14 40.00 24.49

OTO 96.67 9.22 71.67 22.91 42.00 19.24

Visuospatial skills

VSGN 97.33 6.91 69.33 23.18 32.00 17.89 VSFC 98.67 5.07 71.33 23.00 44.00 16.73 VISS 98.67 5.07 70.33 22.97 60.00 24.49

Total TS 97.05 4.55 70 20.00 39.80 16.56 Note: MEPI-memory -episodic memory, MWR- memory- working memory, MSEM- memory-semantic memory, MDEL- memory-delayed recalling, MEM- memory, LEPIC –linguistic expression-picture naming, LEGN- linguistic expression-generative naming, LESC- linguistic expression-sentence completion, LERS- linguistic expression-responsive speech, LESS- linguistic expression- spontaneous speech, LERT- linguistic expression-repetition, LETO- linguistic expression –total, LCCQ- linguistic comprehension-comparative questions, LCFC- linguistic comprehension-following commands, LCRCS- linguistic comprehension- reading comprehension of sentences, LCRCC- linguistic comprehension- reading comprehension of commands, LCTO- linguistic comprehension- total, PSSF-problem solving-sentence formation, PSPO- problem solving-predicting outcome, PSCC- problem solving- compare and contrast, PSPC- problem solving-predicting cause, PSY- problem solving- why questions, PSSA- problem solving- sequential analysis, PSTO- problem solving-total, OC- organization- categorization, OA- organization- analogies, OTO-organization- total, VSGN- visuospatial- generative naming, VSFC- visuospatial-figure copying, VISS- visuospatial-total, TS- total scores. if any. If significant differences were noticed Wilcoxen test was administered to find the pairwise difference across domains.


The results have been presented and discussed under separate sections given below. Domain 1: Memory: The data was statistically analyzed for total memory score and each of the sub domains separately. Analysis of results on total

scores for memory domain revealed that mean score was better for group 1 compared to group 2. The results also revealed that there was a significant difference between the performance of two groups [F (1, 58) = 43.147, p<0.001]. Analysis of results showed that the performance of older group was poorer than the younger group and this indicated that there was a decline in the memory skills with increase in age. The results revealed that there was a decline in the subdomains of memory including


18

episodic memory, semantic memory and working memory. Salthous and Meinz (1995) said that changes in working memory during the course of ageing are pivotal determinant of more general age related decline in cognitive performance. Current study findings are consistent with the above findings that age related decline is seen in all domains of memory. Domain 2: Linguistic expression

Analysis of results on total scores for linguistic expression domain revealed that mean score was better for group 1 compared to group 2. The results also revealed that there was a significant difference in performance between the two groups [F(1, 58)=67.95,p<.001]. This indicated that as age increased there was a decline in linguistic expression. Results revealed that across age groups performance decreases i.e. as age advances mean scores for linguistic expression were found to be decreasing. Naming skills was scored less compared to other sub domains. Picture naming was better compared to generative naming as age advances. Generative naming required more cognitive load as compared to picture naming. Responsive speech and sentence completion scored better than other domains.

Researchers say that decline in linguistic

expression may be due changes in neuronal structures (Raz, 2000). Decline in linguistic expression can also be explained on the basis of transmission deficit hypothesis (Burke & Shafto, 2004), which explains that ageing may cause reduction in activation between phonological, semantic, and orthographic system, as a result of which word finding difficulties can emerge. So in the current study also decline in linguistic expression may be due to reduction in activation between phonology, semantics, orthography, which causes word finding problems. Domain 3: Linguistic comprehension

Analysis of results on total scores for linguistic comprehension domain revealed that mean scores was better for younger group 1 compared to older group 2. The results also revealed that there was a significant difference in performance between the two groups [F(1, 58) =53.79,p<.001]. This indicated that as age increases there is a decline in linguistic comprehension.

Overall the result showed that there was a

significant difference in performance across age groups. Decrease in mean scores was noticed as age advanced. Main differences were noticed in comparative questions and following commands. For the second group poor performance may be because of their short term memory deficit. Reading comprehension was relatively preserved compared to other sub domains.

Domain 4: Problem solving Analysis of results on total scores for problem

solving domain revealed that mean scores was better for younger group 1 compared to older group 2. The results also revealed that there was a significant difference between the two groups at [F (1, 58) = 35.93, p<.001]. This indicated that as age increased there was a decline in problem solving. Mean values for the entire sub domains were less for older group 2 compared the younger group 1. Sentence formulation was less impaired as compared to other subtest, indicating the intactness of syntactic skills. Performance of compare and contrast and sequential analysis was less compared to other domains. This may be because compare and contrast depends on semantic knowledge, and sequential analysis on procedural memory, which declines as age advances. Davis and Klebe (2001) and Kamath (2002) reported that there is decline in problem solving skills in elderly adults as compared to young adults but the result was not significant. They suggest that decline in problem solving may be due to decrease in working memory or executive function with ageing. Domain 5: Organization

Analysis of results on total scores for organization domain revealed that mean scores was better for younger group 1 (Mean = 96.67) compared to older group 2 (Mean = 71.67). The results also revealed that there was a significant difference between the two groups [F (1, 58) =32.93, p<.001]. This indicated that as age increased there was a decline in organization skills. Categorization (OC) was found to be declining as age increased. This difference was found to be significant [F(1,58)=29.18,p<0.001]. Within organization, analogies were performed poorer by older group 2 compared to younger group 1. Difference was found to be statistically significant. Age related decrease in performance was found for organization i.e. as age increased organization skills are decreasing. Performance of analogies was poor compared to categorization because the task involved in analogies required the participant‟s lexical retrieval skills. Lexical retrieval was difficult because of interference happening due to poor inhibition.

Domain 6: Visuospatial skills

Analysis of results on task for visuospatial tasks revealed that mean scores were better for younger group 1 (Mean = 98.67) compared to older group 2 (Mean = 70.33). Difference between the two age groups were statistically significant at [F (1, 58) =43.53, p<.001]. This indicated that as age increased there was a decline in visuospatial skills. Results indicated that visuospatial skills showed a decrease in performance across age groups. Visuospatial test performance declined with age, whereas verbal test performance remained fairly constant. This pattern has been attributed to an age-related decline in either


19

right-hemisphere functioning or executive functions (EFs), which may be associated with prefrontal cortical decline. Poor performance of participants in the current study may be due to more change in the right hemisphere rather than left hemisphere. Comparison of normal and clinical group across domain: Data was analyzed for each age groups and group 3 across all domains. Mean and standard deviation (SD) values were calculated for each domain for three groups. Each domain will be discussed in detail in following sections.

Memory

The analysis of results for the memory domain revealed that the mean scores were found to be less for group 3 compared to the two other groups. Scores for episodic memory, working memory, semantic memory and delayed recall was poor compared to other sub domains in group 3. Within working memory, digit forward task was performed better than digit backward task. Semantic and delayed recall was less compared to other two domains. Within semantic memory both super ordinate naming and coordinate naming was performed equally. Since most of the participants in this study were individuals with moderate dementia all memory domains were impaired.

Research evidences shows that individuals with AD have a distribution of disease that includes modality- neutral association cortex in the temporal and frontal lobes, implicating a categorization deficit in AD patient‟s semantic memory difficulties (Grossman et.al, 2003), IWD have damage to the frontal lobes, and the structures that input to the frontal lobes. As a consequence of this may be the functioning of working memory is compromised in them. Linguistic expression

Results revealed that performance of group 3 was poor in all sub domains of linguistic comprehension compared to two normal groups. Significantly poor scores were found for picture naming and generative naming for IWD compared to older group 2. So this task can be used to differentiate between IWD and elderly individuals. Paraphasias, perseverations were noticed in spontaneous speech of IWD. Repetition was better compared to generative naming and picture naming. Responsive speech and spontaneous speech was better compared to other domains. So this cannot be used for differentiating between dementia and normal ageing. Discourse analysis is needed for the assessment for IWD, as it assesses language functions at a higher level. In this current study discourse analysis was included.

According to Bayles (1982) to name an object requires the perception and recognition of the item and retrieval of its referent from the mental lexicon. Language tasks such as naming, linguistic disambiguation, correction of phonologic, syntactic, and semantically anomalous, sentences, verbal expression, and story re-telling, which interface with cognitive process of abstraction, memory, attention, perception, and reasoning may have clinical potential for use with dementia patients. In the current study for picture naming tasks equal number of living and non living items were included. Performance of persons with dementia was better for nonliving items compared to living things. This result indicated that there was a selective deficit in the knowledge of living items. Linguistic comprehension

The results also revealed that there was a significant difference between the first and third group at |Z| = 4.602 at 0.05 level of significance. At the same level of significance difference between second and third group was found to be statistically significant (|Z| = 3.328).

Group 3 was compared with the normal group

(first and second) to find the differences. Overall the result showed that there was significant difference across these groups. Decrease in mean scores was noticed as age advanced. Significant differences were noticed between second and third group. Reading comprehension was better compared to other sub domains in this age group. IWD were able to follow 1-2 step commands. But for long step commands performance was poor because of the short term memory loss.

Poor comprehension of grammatical

component of sentences in IWD is associated with little recruitment of the inferior frontal area. So for IWD in this study reduced activation in the dorsal frontal area may be attributed to their poor performance in linguistic comprehension. Cummings, Houlihan and Hill (1986) reported that the reading aloud was intact in all except the most severely impaired cases and was found to be relatively independent of intellectual deterioration. Reading comprehension declined progressively with increasing dementia severity and correlated well with quantitative mental status assessments. Albert (1981) written comprehension is relatively preserved in dementia than auditory comprehension.

Domain 4: Problem solving

The results also revealed that there was a significant difference between the first and third group at |Z| = 4.604 at 0.05 level of significance. Mean values for the entire sub domains were less for


20

third group compared the first and older group 2. Analysis of results revealed that performance of third group i.e. group 3 was significantly lower than the performance of first and older group 2. Sentence formulation was less impaired as compared to other subtest, indicating the intactness of syntactic skills in group 3 as well as older group 2. The deficits in the frontal/executive functions also suggest that a disruption of cortical pathways to the frontal lobes and the pathological changes in this region occur early in the disease. These changes in executive functions may be the cause of problem solving skills impairment for IWD in the current study.

Domain 5: Organization

In this domain performance of third group (IWD) is poor compared to first and older group 2. AD patients suffer a gradual deterioration of the organization and content of semantic memory as the disease progresses. Lots of literature shows that semantic memory is impaired in dementia (Chertkow & Bub, 1990).

Domain 6: Visuospatial tasks

Analysis of data revealed that there was significant difference across groups in this domain. This may because of significant change in the brain structure in dementia. Newcombe (1985) suggested that temporal-parietal regions play an important role in visuo spatial tasks. Visuoconstructional dysfunction in AD patients is significantly correlated with a lower glucose metabolism in the right parietal cortex or in the bilateral occipital and temporoparietal regions (Ober, Jagust, Koss, Delis & Friedland, 1991). In the current study also may be because of poor glucose metabolism visuospatial skills are impaired compared to normal elderly.

Comparison of performance of IWD across domains

Analysis of results revealed that MEM and LETO was not statistically different at |Z| = 3.68, p>0.05. No significant difference was noticed across LCTO and MEM. Difference between PSTO and MEM was also insignificant. Difference between MEM and VISS was found to significant at |Z| = 2.02, p>0.05. LETO was found to be significantly different from LCTO and VISS. VISS was significantly different from OTO at |Z| = 2.041, p>0.05. VISS was different from PSTO at |Z| = 2.02, p<0.05. Picture naming and generative naming was significantly different from other sub domains within the linguistic comprehension domain. For linguistic comprehension domain performance scores were poor for following commands and sequential analysis. Within the problem solving domain performance was poor for predicting outcome and sequential analysis.

Conclusions

The aim of the current study was to develop an assessment protocol for persons with dementia in Malayalam that can be useful for speech language pathologist. The developed assessment protocol called – Cognitive Communicative Assessment Protocol for Persons with Dementia - in Malayalam (CCAPD-M) consists of six domains - (i) Memory (ii) Linguistic comprehension (iii) Linguistic expression (iv) Problem solving (v) Organization and (vi) Visuo-spatial construction with a total of twenty four sub domains. CCAPD-M was administered on sixty normal individuals (2 groups i.e. between 40-60years and between 60-80 years) and ten IWD. Results of the study revealed significant deterioration in the performance of normal individuals as the age advanced. Across age groups performance of memory, linguistic expression, linguistic comprehension, problem solving, organization, and visuospatial tasks decreased. Episodic memory, semantic memory, picture naming, responsive speech, spontaneous speech, linguistic comprehension etc. were better compared to other sub domains. Generative naming, delayed recall etc. were performed poorly by elderly population. So these skills cannot be used for differentially diagnosing IWD from elderly normal. Performance of IWD was very less compared to elderly participants across all domains. Delayed recall, generative naming, picture naming, categorization, problem solving etc. were performed poorly compared to other domains. This indicated that cognitive decline as well as semantic decline was more in IWD, whereas syntactic abilities were better. Executive dysfunction and working memory deficits can account for many of the linguistic deficits in dementia.

References Aarsland, D., Hoien, T., Larsen, J. P., & Oftedal, M.

(1996). Lexical and nonlexical spelling deficits in dementia of the Alzheimer‟s type. Brain and Language, 52, 551- 563.

Albert, M. L. (1981). Language in elderly aphasics and the dementic patients. Acquired aphasia. New York: Academic Press.

Albert, M., Blacker, D., Moss, M. B., Tanzi, R. & McArdle, J. J. (2007). Longitudinal change in cognitive performance among individuals with mild cognitive impairment. Neuropsychology, 21 (2), 158-169.

American Speech – Language – Hearing Association. (2005). The roles of speech - language pathologists working with individuals with dementia-based communication disorders: Technical Report [Technical Report].

Appel, J., Kertesz, A., & Fishman, M. (1982). A study of language function in Alzheimer patients. Brain and Language, 17, 73-91.


21

Bayles, K. A. (1982). Language function in senile dementia. Brain and Language, 16, 265 – 280.

Bayles, K. A. (2006). Assessment of cognitive communication disorders of dementia. Boston: College Hill Press.

Bayles, K. A., & Kaszniak, A. W. (1987). Communication and cognition in normal aging and dementia. Boston: College Hill Press.

Bayles, K. A., Boone, D. R., Tomoeda, C. A., Slauson, T. J., & Kaszniak, A. W. (1989). Differentiating Alzheimer‟s patients from the normal elderly and stroke patients with aphasia. Journal of Speech and Hearing Disorders, 54, 74 – 87.

Bayles, K. A., Tomoeda, C. K., & Trosset, M. W. (1990). Naming and categorical knowledge in Alzheimer‟s Disease: The process of semantic memory deterioration. Brain and Language, 39, 498 – 510.

Bayles, K. A., Tomoeda, C. K., & Trosset, M. W. (1992). Relation of linguistic communication abilities of Alzheimer‟s patients to stage of disease. Brain and Language, 42, 454 – 472.

Bayles, K. A., & Tomoeda, C. K. (1993). Arizona battery for communication disorders of dementia. Tucson, Arizona: Canyonlands Publishing, Inc.

Bayles, K. A., & Tomoeda, C. K. (1994). Functional linguistic communication inventory. Tucson, AZ: Canyonlands Publishing.

Boller, F., Pillon, B., Levy, R., & Dubois, B. (1996). Cognitive deficits and dementia in Parkinson‟s disease. In F. Boller & J. Grafman (Eds.), (2001). Handbook of neuropsychology. Aging and dementia, (pp. 311–371) Vol 6. Amsterdam: Elsevier.

Bourgeois, M. S., & Hickey, E., M. (2009). Dementia from diagnosis to management a functional approach. Psychology press, New York.

Burke, D. M., & Shafto, M. A. (2004). Ageing and language production. Current direction in Psychological science, 13, 21-24.

Chertkow, H., & Bub, D. (1990). Semantic memory loss in dementia of alzheimer's type, what do various measures measure? Brain, 113, 397-417.

Cummings, J. L., Benson, D. F., Hill, M. A., & Read, S. (1985). Aphasia in dementia of Alzheimer type. Neurology , 35, 394-397.

Davis, H. P., & Klebe, K. J. (2001). A longitudinal study of the performance of the elderly and young on the Tower of Hanoi puzzle and Rey recall. Brain and Cognition, 46, 95–99.

Diagnostic and Statistical Manual of mental disorders, 3rd edn. (1980). Washington, D. C. American Psychological Association.

Fleischman, D. A., Gabrieli, J. D. E., Reminger, S. L., Rinaldi, J. A., Morrell, F., & Wilson, R. S. (1995). Conceptual priming in perceptual identification for patients with Alzheimer's disease and a patient with right occipital lobectomy. Neuropsychology, 9, 187–197.

Folstein, M., Folstein, S.E., & McHugh, P.R. (1975). “Mini-Mental State” a practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12(3), 189-198.

Freedman, M. & Oscar-Berman, M. (1986). Selective delayed response deficits in Parkinson‟s and Alzheimer‟s disease. Archives of Neurology, 43, 886 – 890.

Grossman, M., Smith, E. E., Koenig, P., Glosser, G., Rhee, J., & Dennis, K. (2003). Categorization of object description in Alzheimer‟s disease and fronto temporal dementia: limitation in rule based processing. Cognitive, Affective, and Behavioral Neuroscience, 3, 120-132.

Kamath, A. (2002). Cognitive Linguistic Assessment Protocol (CLAP). Unpublished masters‟ dissertation submitted to the University of Mysore, Mysore.

Kaplan, E., Goodglass, H., & Weintraub, S. (1983). Boston Naming Test. Philadelphia: Lee & Febiger.

Kemper, S., LaBarge, E., Ferraro, R., Cheung, H, T., & Storandt, M. (1993). On the perseveration of syntax in Alzhiemer‟s disease: Evidence from written sentences. Achieves of Neurology, 50, 81-86.

Kempler, D., Curtiss, S., & Jackson, C. (1987). Syntactic preservation in Alzheimer's disease. Journal of Speech and Hearing Research, 30, 343-350.

Kempler, D., Van Lancker, D., & Read, S. (1988). Proverb and idiom interpretation in Alzhemeir‟s disease. Alzhemeir‟s disease and associated disorders, 2, 38-49.

Kertesz, A. (1982). Western aphasia battery. New York: Grune & Stratton.

Meudell, P. R. & Greenhalgh, M. (1987). Age related differences in left and right hand skill and in visuo-spatial performance: Their possible relationships to the hypothesis that the right hemisphere ages more rapidly than the left. Cortex, 23, 431–445.

Morris, R. G., (1986). Short term forgetting in dementia of the Alzheimer type. Cognitive neuropsychology, 377-87.

Morris, R. G., (1996). Attentional and executive dysfunction. In R. Morris (Eds.) The cognitive neuropsychology of Alzhemier‟s disease. Oxford: Oxoford University Press, 49-70.

Newcombe, N. (1985). Methods for the study of spatial representation. In R. Cohen (Eds.), The development of spatial cognition, (pp. 277-300). Hillsdale, NJ: Lawrence Erlbaum Associates.

Ober, B. A., Jagust, W. J., Koss, E., Delis, D. C., & Friedland, R. P. (1991). Visuoconstructive performance and regional cerebral glucose metabolism in Alzheimer's disease. Journal of Clinical Experimental Neuropsychology, 13, 752-772.

Obler, L. K. (1983). Language and brain dysfunction in dementia. In S. Segalowitz, (Eds.), Language functions and brain organization. New York: Academic press.

Raz, N. (2000). Aging of the brain and its impact on cognitive performance: Integration of structural and functional findings. In F. I. M. Craik, & T. A. Salthouse (Eds.), Handbook of Aging and Cognition-II (pp. 1–90). Mahwah, NJ: Erlbaum.

Rosen, H. J., Hartikainen, K. M., Jagust, W., Kramer, J., Reed, B., & Cummings, J. (2002). Utility of clinical criteria in differentiating frontotemporal lobar degeneration (FTLD) from AD. Neurology, 58, 1608- 1615.

Salmon, D. P., Fennaema- Notestine, C. (1996). Implicit memory in Alzheimer‟s disease: Priming and skill learning. In R. Morris (Eds.), The cognitive neuropsychology of Alzhemier‟s disease (pp. 105-127). Oxford: oxford university press.


22

Salthouse, T. A., & Meinz, E. J. (1995). Aging, inhibition, working memory, and speed. Journal of Gerontology, 50B, 297–306.

Shaji & Bose (2005). Prevalence of dementia in an urban population in Kerala, India. The British Journal of Psychiatry, 186, 6-140.

Silveri, M. C., Daniele, A., Giustolisi, L., & Gainotti, G. (1991). Dissociation between knowledge of living and nonliving things in dementia of Alzheimer type. Neurology, 41, 545-546.

Sunil, K. R., & Shyamala, K. C. (2009). Dementia Assessment Battery in Kannada. Unpublished

Masters‟ dissertation submitted to the University of Mysore, Mysore.

Tomoeda, C. K. (2001). Comprehensive assessment for dementia (A necessity for differential diagnosis and management). Seminars in Speech & Language, 22, 275-289.

Zgaljardic, D., Borod, J., Foldi, N., & Mattis, P. (2003). A review of cognitive behavioural sequeale of Parkinson‟s disease: Relationship to fronto striatal circuitry. Cognitive and Behavioural Neurology, 16, 193-210.

Forensic speaker identification

23

Benchmark for Speaker Identification Using Glottal Source Parameters in Hindi Speakers

Aparna V. S.1 & S. R. Savithri2

Abstract

Studies concerned with establishing parameters for speaker verification are important because of the legal ramifications and because of the forensic involvements associated with the application of these studies. Success of identifying the speaker depends on extracting speaker-dependent features from speech signals that can effectively distinguish one speaker from another. It is not known as to what percent matching would indicate similarity/dissimilarity of speaker, or benchmarking of various features is not established. In this context the aim of the present study was to determine the benchmark for speaker identification using glottal source parameters in direct recording condition. Ten normal Hindi speaking male subjects in the age range of 21-38 years participated in the study. The material used was nine commonly occurring, meaningful Hindi words containing the long vowels /a:/, /i:/, and /u:/ in the word-medial position embedded in sentences. The vowels were displayed as waveform and were acoustically zoomed to extract the source and filter parameters using Acophone I and SSL software (Voice and Speech Systems, Bangalore). Glottal source parameters open quotient (OQ), leakage quotient (LQ) and speed quotient (SQ) were extracted in 10 steady state point of each of the vowels. The results of the present study showed that the glottal source doesn‟t remain the same even in normal mode of speaking. Hence these parameters don‟t serve as a good measure for speaker identification. In general it could be concluded that OQ*LQ, OQ*SQ, and LQ*SQ cannot be considered as an efficient parameter for speaker identification in field conditions in Hindi speakers.

Key words: open quotient, leakage quotient, speed quotient, euclidian distance

uman beings have many characteristics that make it possible to distinguish one individual from another. Some individual characteristics can be perceived very rapidly such as facial

features vocal quality and behaviors. Voice is the very emblem of the speaker, indelibly woven into the fabric of speech. In this sense each of our utterances of spoken language carries, not only its own message, but through ones accent, tones of voice and habitual voice quality it is also an audible declaration of our membership of a particular regional group, of our individual physical and psychological identity, and of our momentary mood. Thus the voice of an individual is said to be having its own characteristics and distinct distinguishable quality.

Speaker recognition is any decision making

process that uses speaker dependent features of the speech signal (Hecker, 1971). Atal (1976) suggests that speaker recognition is any decision making process that uses some features of the speech signal to determine if a particular person is the speaker of a given utterance. Nolan (1983) identified two classes of speaker recognition - speaker identification and speaker verification. Speaker recognition includes two sub-fields

1e-mail: [email protected]; 2Professor of Speech Sciences, AIISH, Mysore, [email protected]

(a) naive speaker recognition and (b) technical speaker recognition. Technical speaker recognition is usually called as “Speaker Identification by expert” which uses specialized techniques (Nolan, 1983). Hecker (1971) and Bricker and Pruzansky (1976) identified three methods of speaker recognition (a) by listening (b) by visual inspection of spectrograms, and (c) by machine.

In speaker verification an identity claim from an

individual is accepted or rejected by comparing a sample of his speech against a stored reference sample by the individual whose identity he is claiming (Nolan, 1983). An utterance from an unknown speaker has to be attributed, or not, to one of a population of known speakers for whom reference samples are available. Here only two types of decision are possible, either the unknown sample is correctly identified or it is not. The goal of speaker identification is to determine which one of a group of known speakers‟ best matches the test speech sample. Speaker identification can be constrained to a known phrase (text-dependent) or totally unconstrained (text-independent).

Of the three methods speaker identification by

machine has received greater interest in recent past. Extracting speaker dependent parameter from signals and analyzing them by machines is an objective method which is classified into automatic and semiautomatic

H


24

method. In the semi-automatic method, there is extensive involvement of the examiner with the computer, whereas in the automatic method, this contact is limited. A combination of subjective and objective method is usually used. In the past pitch, intensity, phonemic voicing patterns (Hecker, 1971), long-term speech spectra (Hollien & Majewski, 1977), fundamental frequency (Abberton & Fourcin, 1978), cepstral parameterization (Plumpe, Quatieri & Reynolds, 1999), fundamental frequency, the third and fourth formants, and the closing phase of the glottal wave (Lavner, Gath & Rosenhouse, 2001), four formants (F1, F2, F3, F4), the amount of periodic and aperiodic energy in the speech signal, the spectral slope of the signal and the difference between the strength of the first and second harmonics (Carol Epsy-Wilson, Sandeep & Vishnubhotla, 2006), first three formants, word duration, closure duration, transition duration in disguised speech (Savithri, 2008) have been used. Sharma, Jain and Sharma (2009) in their study found that other supralaryngeal parameters like formant frequencies may shift during disguise but the open quotient and glottal leakage were found to occur in certain range for normal and disguised modes as the degree of glottal opening remains similar in normal mode but varies appreciably for disguised mode. Pamela (2002) studied the reliability of voice prints. Within the preview of her study, it was suggested that two samples can be considered to be from different speakers when more than 67% of measurements are different in natural speaking condition. But the validity of this method is still in question. Jakhar (2009) used quefrency for benchmarking and result obtained was a mean percentage of 88.33 (5 speakers), 81.67 (10 speakers) and 60 (20 speakers) in live v/s live condition, 81.67 (5 speakers), 68.33 (10 speakers) and 50 (20 speakers) in mobile v/s mobile, and 78.33 (5 speakers), 68.33 (10 speakers) and 43.33 (20 speakers) in live v/s mobile condition. The results indicate that speaker identification was higher when mode of recording was same and when the number of the speakers was less in the group. Lakshmi (2009) used formants F1- F2 for benchmarking and obtained benchmark of 70% for vowel /i:/, 65% for vowel/a:/ and benchmarking for other vowels were below chance level when 5 speakers were considered, and below chance level for ten and twenty speakers for all three vowels.

However, the question regarding the most

appropriate speech parameter for semi automatic/automatic speaker identification in real forensic condition are still far from being answered. To prove that the suspect is a criminal, it needs to be verified beyond reasonable doubt that the voice of the criminal and voice of the suspect are the same. Success

in this task depends on extracting speaker-dependent features from the speech signal that can effectively distinguish one speaker from another. Ideally, the features chosen for speaker recognition must satisfy the following criteria (Wolf, 1972): have lower within-speaker (within source) variability and relatively higher between speaker (between sources) variability, be stable over time, be difficult to disguise or mimic, be robust to transmission and noise, be relatively easy to extract and measure, and should occur frequently in the speech samples.

The glottal source is an important component of

voice as it can be considered as the excitation signal to the voice apparatus. The use of the glottal source for pathology detection or the biometric characterization of the speaker is an important objective in the acoustic study of the voice now a days. The likely shape of the vocal tract can be approximately estimated from the analysis of the spectral shape of the voice signal. In automatic speaker recognition, coefficients representing the sounds, taking into consideration the vocal tract shape and excitation, are parameterized and used as features. It is not known as to what percent matching would indicate similarity/dissimilarity of speaker or benchmarking of various features is not established. In this context, the present study evolved a benchmark for speaker identification using glottal source parameters, specifically open quotient, leakage quotient and speed quotient extracted from glottal source in direct recording condition.

Method

Participants: Ten normal Hindi speaking male subjects in the age range of 21-38 years with no history of neurological or psychological illness participated in the study.

Material and Procedure: The material used was nine commonly occurring, meaningful Hindi words containing the long vowels /a:/, /i:/ and /u:/ in the word- medial position embedded in sentences. Direct (live) recording of the four repetitions of these sentences by the participants as done by Jakhar (2009) was taken for the present study. The words and in turn the target vowels were truncated from the samples and stored in folders D1, D2, D3 and D4. The vowels were displayed as waveform and were acoustically zoomed to extract the source and filter parameters using Acophone I and SSL software (Ananthapadmanabha, 2008; Voice and Speech Systems, Bangalore). Glottal source parameters open quotient (OQ), leakage quotient (LQ) and speed quotient (SQ) were extracted in 10 steady state point of each of the vowels.


25

For each vowel thirty values (3 * 10 observations) were obtained in a single recording. Each recording was considered as that of unknown speaker and subsequent recording as that of the known speaker and percentage correct identification was noted for five speakers for vector OQ*LQ. Percentage correct identification for 10 speakers were calculated taking the average of D1, D2 and D3, D4. The study was extended to find out percent correct identification for vectors OQ*SQ and LQ*SQ in five speakers.

The OQ*LQ was plotted with OQ on horizontal

axis and LQ on vertical axis for a group of known speaker versus one unknown speaker. Euclidian distance (ED) was calculated as the distance between unknown speaker (reference sample) and the known speakers (test sample) using the following formula: In a plane with p1 at (x1, y1) and p2 at (x2, y2)

D (p, q) = √ ((x1 - x2)² + (y1 - y2)²). Where p=reference subject, and q= test subject X and Y belong to one of the parameters (OQ, LQ and SQ). If the distance between reference sample and test sample is least it is considered as correct identification. Percentage of correct identification was calculated using the following formula:

Results

Inter-speaker identification OQ-LQ: The OQ-LQ was plotted with OQ on horizontal axis and LQ on vertical axis for a group of known speakers versus one unknown speaker. A total of 210 figures (3 vowels X 2 group of speakers X 6 combinations of 4 recordings X 10 speakers) were plotted for speaker identification. The Euclidian distances between selected unknown speakers and the corresponding five known speaker were calculated. The lowest Euclidian distance value is highlighted in the table. If the lowest value and the Euclidian distance value corresponding to the actual speaker are the same it was considered as correctly identified. That is if the distance between the unknown and corresponding known speaker was the lowest, then speaker was deemed to be correctly identified. If unknown speaker is closer to some other known speaker in terms of Euclidian distance it as deemed to as wrong identification. That is if the distance between unknown speaker and corresponding known speaker is more, then the speaker was deemed to be not identified. The results indicated percent correct identification of 48.3, 34.6, and 33.3 for vowels /a:/, /i:/, and /u:/,

respectively when five subjects were considered. The percent correct identification reduced drastically when 10 subjects were considered. Table 1 show the percent correct identification for three vowels when five and ten subjects are considered. Figure 1 shows the benchmarking for three vowels.

Table 1. Percent correct identification of vowels /a:/, /i:/

and /u:/ for two groups of speaker for OQ-LQ

Groups % correct identification

/a:/ /i:/ /u:/ 5 speakers (A) 48.3% 34.6% 33.3%

10 speakers (B) 10% 20% 30%

Figure 1. Percent correct identification of vowels /a:/, /i:/ and /u:/ for two groups of speaker for OQ-LQ. Tables 2-3 shows the Euclidian distance of five

speakers when the recordings D1 V/S D2 were considered for vowel /a/, and figures 2-3 represent the correct / false identifications.

Table 2. Correct identification in a group of 5 speakers for OQ-LQ on vowel /a:/ when reference speaker was

US1 (Lowest Euclidian distance is in bold)

For 10 speakers, average of two recordings D1

and D2 (D1) was considered as the reference and average of recordings D3 and D4 (D2) was considered as the test sample. As reported earlier the percent correct identification was less than chance level. Table

Reference speaker

Reference sample

Test sample ED

OQ LQ OQ LQ US1 0.687 0.011 S1 0.682 0.016 0.007

S2 0.604 0.044 0.089 S3 0.559 0.023 0.128 S4 0.605 0 0.082 S5 0.645 0.057 0.062


28

when 5 speakers were considered and benchmarking for other vowels were below chance level. Of the three vectors, identification was above chance level for only for vowel /u:/ for LQ*SQ. Plumpe, Quatieri and Reynolds (1999) reported that while traditional speaker identification systems rely on the vocal tract dynamics, addition of source information can prove to be valuable speaker-specific information. They suggested the use of parameters obtained from the time-domain glottal source description in speaker identification experiments. The results of the present study are not in consensus with the observations of Plumpe et al., (1999) as the benchmarking obtained was poor. Sharma et al., (2009) found that laryngeal measures (open quotient, leakage quotient) are less subjected to change compared to the supralaryngeal measures in disguise condition. So these parameters are supposed to provide better benchmarking. The results in the present study show that all the three parameters had poor benchmarking and therefore the result is not in consensus with the finding.

The results obtained for all three vectors in this

study was poorer compared to that of study by Lakshmi (2009) who used formants F1-F2 for benchmarking and obtained a benchmark of 70% for vowel /i:/, 65% for /a:/ and benchmarking for other vowels were below chance level when 5 speakers were considered. This shows that both laryngeal and supralaryngeal measures change even in normal condition.

Jakhar (2009) used quefrency for benchmarking in Hindi speakers. Benchmarking obtained was 88.33 (5 speakers), 81.67 (10 speakers) and 60 (20 speakers) in live v/s live condition, 81.67 (5 speakers), 68.33 (10 speakers) and 50 (20 speakers) in mobile v/s mobile, and 78.33 (5 speakers), 68.33 (10 speakers) and 43.33 (20 speakers) in live v/s mob condition. With respect to the number of speakers, the percent correct identification was higher when the number of the speakers was less in the group. The present study showed very poor benchmarking on all three vectors than Jakhar for 5 speakers and 10 speakers. However, it is in consensus with Jakhar on the finding that percent correct identification decreased as the number of speakers increased. The results of the present study show that the glottal source doesn‟t remain the same even in normal mode of speaking. Hence these parameters don‟t serve as a good measure for speaker identification.

Conclusions

The result of the present study has contributed to the field of speaker identification. In general, it could be

concluded that OQ*LQ, OQ*SQ and LQ*SQ cannot be considered as an efficient parameter for speaker identification in field conditions. However, the results cannot be generalized to other conditions and disguised speech. The results cannot be generalized as it depends on vowel, language and recording conditions. The present study used samples of field recording which might have added on to the disadvantage. However, future studies in laboratory recording, inclusion of more subjects, and speaker identification under disguise conditions are warranted.

Acknowledgements

The authors wish to express their gratitude to Dr. Vijayalakshmi Basavaraj, Director, AIISH for permitting to carry out this study. They also thank all the subjects for their cooperation for the study.

References Abberton, E., & Fourcin, A. J. (1978). Intonation and speaker

identification. Language and Speech, 21, 305-318. Ananthapadmanabha, T.V. (2008). Voice and Speech Systems,

Bangalore. Atal, B. S. (1976). Automatic recognition of speaker from

their voices, Proc.IEEE, 64, 4, 460-75. Bricker, P. S., & Pruzansky, S. (1976). Speaker recognition:

Experimental Phonetics. London: Academic Press. Carol, Epsy-Wilson, Sandeep, M., & Vishnubhotla, S.A.

(2006). New Set of Features for Text-Independent Speaker Identification. Institute for Systems Research and Dept. of Electrical & Computer Engineering, University of Maryland, College Park, MD, USA.

Hecker, M. H. L. (1971). Speaker Recognition: Basic considerations and methodology. Journal of the Acoustical Society of America, 49,138-138.

Hollien, H., & Majewski, W. (1977). Speaker identification by long-term spectra under normal and distorted speech conditions. Journal of the Acoustical Society of America, 62, 975-980.

Jakhar, S. S. (2009). Bench mark for speaker Identification using Cepstrum. Unpublished project of Post graduate Diploma in Forensic Speech Science and Technology submitted to University of Mysore, Mysore.

Lakshmi, P. (2009). Benchmark for speaker Identification using Vector F1 ~ F2. Unpublished project of Post graduate Diploma in Forensic Speech Science and Technology submitted to University of Mysore, Mysore.

Lavner, Y., Gath, I., & Rosenhouse, J. (2001). The prototype model in speaker identification by human listeners. International Journal of Speech Technology, 4, 63-74.

Nolan, F. (1983). Phonetic bases of speaker recognition. Cambridge: Cambridge University Press.

Pamela, S. (2002). Reliability of voice prints, Unpublished dissertation submitted to the University of Mysore, Mysore.


29

Plumpe, M., Quatieri, T., & Reynolds, D. (1999). Modeling of the glottal flow derivative waveform with application to speaker identification. IEEE Trans. Speech and Audio Proc, 1, 569-586.

Savithri, S. R. (2008). Acoustic similarities and differences within and between speakers. AIISH Research fund Project.

Sharma, S., Jain, S. K., & Sharma, R. M. (2009). Characterization of temporal and acoustic parameters for speaker identification in disguise speech. XX All India Forensic Science Conference, Jaipur.

Wolf, J. J. (1972). Efficient acoustic parameter for speaker recognition. Journal of the Acoustical Society of America,2044-2056.


30

Agrammatism in Children with Hearing Impairment Aswathy A. K.1 & Shyamala K.C. 2

Abstract

In the realm of speech and language production, children with hearing loss produce ungrammatical sentences and have difficulties in the acquisition of syntactic structures which gives the impression of agrammatic kind of speech. The study attempted to account for agrammatism in Malayalam speaking children with hearing impairment using the Index of Productive Syntax (IPSyn) (Scarborough, 1990). The natural speech samples of thirty children with normal hearing were compared with thirty children with hearing impairment and analyzed using the Index of Productive Syntax. Malayalam Language Test (MLT) (Rukmini, 1994) was used for assessing the language age of the participants. Correlation between MLT and IPSyn scores was checked. The results revealed comparatively better scores for children with normal hearing for some of the items in IPSyn. Also a positive correlation was noticed between the two syntactic assessment tools, MLT and IPSyn. The test results lead to the conclusion that there are some missing grammatical structures in speech and an overall delay in the performance of children with hearing impairment in comparison to the language age matched normally hearing children revealing the presence of agrammatism.

Key words: agrammatism, syntax, morphology, index of productive syntax, malayalam language test

anguage is a 1systematic conventional use of sounds (or signs or written symbols) for the purpose of communication or self expression

(Crystal, 1995). Given the complex nature of language and communication, there are a number of structural and functional aspects of the language system to be assessed. Among the subcomponents of language (phonology, lexicon, morphology, syntax, pragmatics, sociolinguistics, literacy), syntax and morphology plays a main role in the language development and together called as grammar are the most researched. The study of words, its parts and rules that govern their combination is called morphology. Syntax is the component of grammar that governs the ordering of words in sentences. However the syntax and morphology are the systems for combining units of meaning. All speakers of a language know the rules of syntax and morphology as well as the categories and structures these rules operate over. Language development in typically developing children

By seeing the language milestones of a child, one can notice that from birth to one year children show dramatic change in the sounds they produce and in the communicativeness of their behavior. As children start putting words together in longer sentences, two important things can be noted, that is the presence of affirmative, declarative type of sentences and the tendency towards missing certain words and bound morphemes which makes children‟s utterances sound li�e telegraphic speech (Brown & Fraser, 1963). This kind of telegraphic speech in Adult Language Disorders (Brocas Aphasia) is often termed as agrammatism. As per the

1e- mail: [email protected]; 2Professor of Language Pathology, AIISH, Mysore, [email protected].

definition agrammatism refers to the restricted use of grammar (Webster, Morris & Franklin, 2005). In agrammatism the function words are either wrongly used or missed out completely. These kinds of features are evident in typically developing children which is a normal process during the acquisition of grammar. The determiners, prepositions, auxiliaries, bound morphemes etc. are the elements the child constantly omits and are called as grammatical morphemes. Unlike in English, the morphological development is a larger component of language acquisition for languages with a rich morphological system such as the Finno-Ugric languages (Finnish, Estonian, Hungarian, Turkish), Dravidian languages (Tamil, Telugu, Malayalam, Kannada), and Indo-Aryan languages (Hindi, Bengali, Marathi, Gujarati).

Language development in children with hearing impairment

It has been clear since long that the acquisition of spoken-language syntax and morphology poses a particular challenge for hearing impaired children with severe and profound hearing impairments (Moores, 1972; Swisher, 1976). It can generally be said that although hearing impaired children learning spoken English show similar patterns to normal hearing children in terms of order of acquisition of inflectional morphemes, the age of acquisition is substantially later (Cleary, Schwartz, Wechsler – Kashi & Madelll, 2006).

Assessment of the language disorder is an

important aspect in the field of Speech-Language Pathology. The use of language functions by a child (language disordered) needs constant and keen observation. Only by constant observation a Speech- Language Pathologist can succeed in the intervention of language disorders. Because of the great concern about this, many assessment tools have been developed in different languages.

L

Agrammatism in children with hearing impairment

31

In western countries, many tests have been developed to assess language development both in normals and in children with hearing impairment. These include Northwestern Syntax Screening Test (NSST) which was developed by Lee in (1969); Test of syntactic ability by Quigley, Steinkamp, Power and Jones, (1978); Developmental Sentence Types (DST) by Lee (1966, 1974) and Teacher Assessment of grammatical structures by Moog and Kozak (1983). Researchers have attempted to adapt the western tests to other languages including Indian languages.

The Dravidian family of languages is

mainly spoken in southern India and parts of eastern and central India as well as in northeastern Sri Lanka, Pakistan, Nepal, Bangladesh, Afghanistan, Iran, and overseas in other countries such as Malaysia and Singapore. Among them Tamil, Telugu, Kannada and Malayalam are the languages with the most speakers. In case of Dravidian languages the syntax is different. Malayalam is a Dravidian language which is commonly spoken by people in the south- eastern part of India. Many of the features common to the Dravidian languages are present in Malayalam also. Much research has been conducted in the area of Malayalam grammar.

Moreover now the research is focusing more

on the differences in the language abilities, styles and rate of language development in normally developing preschool aged children and children with other language disorders. For this, other than the traditional method of language assessments children‟s natural language samples are used. This natural language samples has to be analyzed and scored. For this purpose a few scoring methods were available for consideration. Among these the most widely used is the Mean Length of Utterance (MLU) by Brown (1973). Other scoring methods available were Developmental Sentence Score (DSS; Lee, 1974), and Assigning Structural stage (ASS; Miller, 1981). Each one has proved to be useful when analyzing relatively few corpora. One another measure similar to these is the Index of Productive Syntax (IPSyn), (Scarborough, 1990) which is widely used nowadays because of its higher content validity.

Index of Productive Syntax was developed

to serve as a readily obtained summary scale of grammatical complexity that would be appropriate for the study of individual differences in language acquisition. IPSyn was developed by Scarborough (1989) as a response to the specific needs that arose within a longitudinal study of individual differences in language abilities in relation to later reading abilities. The syntactic and morphological complexity of language samples was analyzed using

IPSyn in the study. It was originally conceived as a quantification of Miller‟s ASS to convert it to a type – based scoring system. It consisted of fifty six grammatical items under four main sections such as, Noun Phrases, Verb phrases, Questions / Negations, Sentence structures.

While all the above studies are encouraging as

a whole, in the literature regarding the use of an assessment tool like Index of Productive Syntax in Indian languages, especially Dravidian language family is sparse. Hence the need for the current study arises as a preliminary attempt to account for agrammatism in Malayalam speaking children with hearing impairment using the Index of Productive Syntax.

The objectives of this study were (a) to adapt

the Index of Productive Syntax to Malayalam (b) to account for the agrammatism seen in orally trained Malayalam speaking children with hearing impairment in three different age groups (3-4years, 4-5 years and 5-6 years) (c) to compare the syntactic development seen in these children with their language age matched normals, by using the index of productive syntax (Scarborough, 1990) and (d) to check the correlation between MLT and IPSyn scores.

Method

In the present study two groups of Malayalam speaking participants were selected. One group included 30 children with normal hearing and the other group contained 30 language age matched children with hearing impairment. The age range of children with normal hearing was 3-6 years. The participants were grouped into 3 age groups, such as each group containing 10 children (3-4 years, 4-5 years, 5-6 years). The subject selection criteria are given below in the Table 1. Procedure: The present study consisted of two different phases: Phase I and Phase II. Phase I: This was carried out on children with normal hearing in three stages, data collection, transcription, translation and adaptation of IPSyn. Data collection: Fifty to seventy spontaneous utterances were collected through audio recording for 30 children with normal hearing in school set up (kindergarten and primary school). The main task given to children for collecting the speech sample was story narration using picture cards. For story narration task, Panchatantra stories were used. These stories were part of a fluency test which was developed and standardized as a part of master‟s dissertation at AIISH by Nagapoornima (1990), Yamini (1990) and Srinivasan (1992). It consists of pictures (3-4 years), connected pictures


32

Table 1. Subject selection criteria Children with normal hearing

Children with hearing impairment

Children with normal speech and language

skills. Children with congenital hearing impairment who were

language age matched with the children with normal hearing.

Language assessment conducted using Malayalam Language test (Rukmini, 1994).

Language assessment done with Scales of Early Communication Skills (SECS) for hearing impaired children

(Moog & Geers, 1975).

Children attending regular school. Language age assessment using Malayalam Language Test (Rukmini, 1994) for those children who had language age

above 3 years in SECS. Children from same type of socio - economic background as that of children with hearing

impairment.

Children with mild-profound sensory neural hearing loss using binaural behind the ear hearing aids.

Children who are undergoing training for speech, language, listening skills & oral communication from trained Speech- Language Pathologists & Special Educators from the time of

identification.

Children who were regularly using binaural behind the ear

hearing aids.

Children without any history of psychological / neurological

impairment. (4-5 years), cartoons (5-6 years), and Panchatantra stories (6-7 years) to be administered to children in the age group of 3-7 years.

Conversation (about their daily routines, school picnic, and play activities) was also included. Open ended and closed ended questions were asked during the conversation. It was noticed that all through the conversation the interaction was investigator-directed in character. Along with this all possible types of utterances elicited at the time of administration of MLT were also considered for speech sample.

In the present study MLT was used to assess

the language age of the child and was scored at the time of administration of test. MLT is a test which assess the language development and language acquisition (to an extent). It consists of two main sections; Syntax & Semantics which includes 11 subsections in each.

Transcription: The recorded speech samples were transcribed by the investigator and then subjected to IPSyn analysis and scoring. While transcribing the sample, the repeated as well unintelligible items were removed. Translation and adaptation of IPSyn: The 56 items were translated to Malayalam with the help of a Linguist. As part of adaptation of the manual some modifications were made. Out of the fifty six items present in the original manual five items were not considered during the analysis as these were not applicable in Malayalam language. The items not considered were two word noun phrase after verb (N6), catenative preceding a verb (V5), negative

morpheme between subject and verb (Q5), bitransitive predicate (S14), fronted or center embedded subordinate clause (S19).

In Malayalam the word order is different, i.e., two word noun phrase before verb or preposition which is already present as another sub item under noun phrase. Hence the item two word noun phrase after verb or preposition (N6), the catenative (pseudo- auxiliary) preceding a verb (V5), cannot be considered as a strict grammatical usage, rather than a dialectal variation. In the sentence structure section the occurrence of negative morpheme between subject and verb (Q5), bitransitive predicate (S14), fronted or center embedded subordinate clause (S19) are unusual and these were not included in the manual.

The transcribed normative speech samples

were analyzed and scored using the fifty one items in the IPSyn manual, following the scoring protocol outlined by Scarborough (1990). Each of the transcribed language samples were reviewed line by line and a check mark was placed next to each item as it was encountered. While scoring zero, one, or two points per item were awarded, so that the total score would be the sum of these points over all items. It was scored in such a way that if two examples of each form of each item was encountered it was given a score of two, if only one example was encountered it was given a score of one and a score of zero was awarded if it was absent. Then the sub items were totaled and sum of the sub items gave the total score of test for a particular sample.

One modification was also made in regard of

scoring as suggested by Scarborough, (1990) that the


33

sampling context was taken into consideration. He pointed out that the pragmatic contextual factors may constrain a child‟s production of particular forms. Phase II: The second phase which was carried out on children with hearing impairment consisted of three stages, data collection, transcription, grammatical analysis and statistical analysis. Data collection: Thirty orally trained Malayalam speaking children with hearing impairment who were matched for language age with normally hearing peers were considered for data collection. The speech sample was collected in therapy and (special) school situations. Task (story narration and conversation) given for children with normal hearing was adopted for children with hearing impairment also. Transcription: As described in phase I recorded speech samples were transcribed by the investigator and then subjected to analysis and scoring. While transcribing the sample, the repeated as well unintelligible items were removed. Grammatical analysis: The transcribed speech samples were subjected to analysis. The modified IPSyn consisting of 51 items were used for scoring. The scoring procedure was same as that conducted for normative speech sample. Statistical analysis: Following the grammatical analysis, the IPSyn and MLT scores were subjected to statistical analysis.


As a part of adaptation of IPSyn to Malayalam, the language sample elicited from the children with hearing impairment were analyzed using IPSyn under four main sections, Noun Phrase (NP), Verb Phrase (VP), Question / Negation (Q/N), and Sentence Structure (S) with a total of fifty one sub items.

The total percentage score of the sub items for

the three groups were calculated. Among this the items which scored less than 50% were removed from the manual. Then it was administered to the children with hearing impairment.

The following sub items were scored poorly

by children with normal hearing, (N5) article, (N7) plural suffix, (N11) three word noun phrase, (V9) modal preceding verb, (V11) past tense modal, (V14) modal adverb, (V15) copula, modal / auxiliary / ellipses, (Q6) wh question with inverted modal, copula or auxiliary, (Q7) negation of copula, (Q8) yes/no question with inverted modal, copula or auxiliary, (Q10) why/when/whose/which questions, (S5) conjunction, (S9) introducer, (S11)

propositional complement, (S13) wh clause, (S16) relative clause, (S17) infinitive clause, and (S18) gerund. Account of agrammatism: In order to account for the agrammatism seen in orally trained Malayalam speaking children with hearing impairment and children with normal hearing, descriptive statistics was used. The mean and standard deviation values were computed for each of the main sections, noun phrase, verb phrase, questions/ negations and sentence structures separately which are given in Table 2.

Subsequent to the descriptive statistics, the non parametric tests Mann Whitney and Kruskal Wallis were employed to estimate the significant difference for each of the items for both groups and ages respectively.

The following items were found to be significantly different across groups; N3, which represents the modifiers (/z/ = 3.082, p < 0.05); N4, which represents the two word noun phrase (/z/ = 2.60, p < 0.05); N8, two word noun phrase before verb (/z/ = 3.188, p < 0.05); N9, three word noun phrase (/z/ = 2.528, p < 0.05); N10, adverb (/z/ = 3.042, p < 0.05) in the section noun phrase. The items, V2, particle or preposition (/z/ = 3.53, p < 0.05); V12, regular past tense suffix (/z/ = 2.56, p < 0.05); V13, past tense auxiliary (/z/ = 2.47, p < 0.05); and V16, past tense copula (/z/ = 2.31, p < 0.05) in the verb phrase section. Q1, intonationally marked question (/z/ = 2.56, p < 0.05); Q2, routine or existence or wh pronoun question (/z/ = 1.76, p < 0.05); Q3, simple negation (/z/ = 4.031, p < 0.05); Q4, initial wh pronoun question followed by verb (/z/ = 6.846, p < 0.05); Q9, why, when, whose, which questions (/z/ = 6.272, p < 0.05) in the question/ negation section. In the section, sentence structure, S6, sentence with two verb phrase (/z/ = 2.65, p < 0.05); S10, adverbial conjunction (/z/ = 8.38, p < 0.05); S12, conjoined sentences (/z/ = 5.86, p < 0.05); and S15, sentence with three or more verb phrases (/z/ = 4.62, p < 0.05) are found to be significant.

Significant difference was not present for the items, N1, proper nouns (/z/ = 0.00, p < 0.05); N2, pronoun (/z/ = 1.416, p < 0.05); N5, article (/z/ = 1.844, p < 0.05); N7, plural suffix (/z/ = 0.083, p < 0.05); V1, verb (/z/ = 0.00, p < 0.05); V3, prepositional phrase (/z/ = 0.58, p < 0.05); V4, copula linking two nominals (/z/ = 0.35, p < 0.05); V6, auxiliary (/z/ = 1.61, p < 0.05);V7, progressive suffix (/z/ = 0.66, p < 0.05); V8, adverb (/z/ = 1.72, p < 0.05); V10, third person singular present tense suffix (/z/ = 1.92, p < 0.05); S1, two word combination (/z/ = 0.00, p < 0.05); S2, subject – verb sequence (/z/ = 0.00, p < 0.05); S3, object verb


34

Table 2. Mean (M) and Standard Deviation (SD) of noun phrase, verb phrase, questions and negations and sentence structures across groups and age (years)

ITEMS NL Hl

3-4 4-5 5-6 3-4 4-5 5-6 M SD M SD M SD M SD M SD M SD

N1 2.0 0 2.0 0 2.0 0 2.0 0 2.0 0 2.0 0 N2 1.6 0.5 1.1 0.5 1.0 0.6 1.6 0.5 1.4 0.5 1.4 0.5

N3 1.8 0.4 1.8 0.6 1.4 0.6 0.6 0.5 1.0 0.8 1.7 0.4 N4 2 0 1.6 0.5 1.6 0.7 1.2 0.6 1.3 0.8 1.5 0.5 N5 1.8 0.4 1.8 0.6 1.5 0.7 1.0 0.7 1.5 0.7 1.8 0.4

N7 0 0 1.0 0.8 0.2 0.4 0.1 0.3 0.6 0.8 0.6 0.8

N8 2.0 0 1.7 0.4 1.8 0.4 1.3 0.4 1.6 0.5 1.4 0.5 N9 0.7 0.5 1 0 1 0.8 1.4 0.5 0.2 0.4 0 0 N10 0.8 0.4 0.7 0.8 1.3 0.6 0.1 0.3 0.8 0.4 0.3 0.6

N11 0.0 0.4 0.7 0.8 1.3 0.7 - - - - - - V1 2.0 0 2.0 0 2.0 0 2.0 0 2.0 0 2.0 0 V2 1.6 0.8 1.1 0.5 1.4 0.8 1.8 0.4 2.0 0 2.0 0

V3 2.0 0 2.0 0 1.9 0.3 1.8 0.4 2.0 0 2.0 0 V4 1.6 0.5 1.9 0.3 2.0 0 1.6 0.5 2.0 0 2.0 0 V6 0.9 0.3 1.6 0.8 1.3 0.6 0.2 0.4 1.7 0.4 0.9 0.7

V7 1.1 0.9 1.8 0.4 1.0 0.8 0.6 0.8 1.4 0.5 1.5 0.8 V8 1.3 0.4 1.4 0.5 1.3 0.9 1.1 0.7 2.0 0.0 1.7 0.4 V9 0.5 0.8 0.0 0.0 0.2 0.4 - - - - - -

V10 0.0 0.0 1.8 0.6 1.2 0.9 0.6 0.5 0.5 0.5 0.4 0.5 V11 0.3 0.4 0.0 0.0 0.2 0.4 - - - - - - V12 1.8 0.4 1.6 0.5 2.0 0.0 2.0 0.0 2.0 0.0 2.0 0.0

V13 0.2 0.4 1.8 0.6 1.3 0.9 0.2 0.4 0.5 0.5 0.8 0.7 V14 0.0 0.0 0.1 0.3 0.2 0.4 - - - - - - V15 0.0 0.0 0.2 0.4 0.0 0.0 - - - - - -

V16 1.8 0.6 2.0 0.0 1.4 0.8 2.0 0.0 2.0 2.0 0.0 2.0 Q1 1.0 0.0 1.2 0.4 2.6 4.0 0.6 0.5 1.1 0.7 0.7 0.6 Q2 1.8 0.4 2.0 0.0 1.8 0.6 2.0 0.0 2.0 2.0 0.0 2.0

Q3 1.4 0.5 1.8 0.6 1.4 0.5 2.0 0.0 2.0 2.0 0.0 2.0

Q4 0.0 0.0 0.9 0.7 1.0 0.0 2.0 0.0 2.0 2.0 0.0 2.0 Q6 0.5 0.5 0.3 0.4 0.3 0.4 - - - - - - Q7 0.4 0.5 1.7 0.4 0.4 0.5 - - - - - -

Q8 0.4 0.5 0.0 0.0 0.2 0.4 - - - - - Q9 1.5 0.5 0.5 0.5 0.8 0.6 2.0 0.0 2.0 2.0 0.0 2.0 Q10 0.0 0.0 0.0 0.0 0.0 0.0 - - - - - -

S1 2.0 0.0 2.0 2.0 0.0 2.0 2.0 0.0 2.0 2.0 0.0 2.0 S2 2.0 0.0 2.0 2.0 0.0 2.0 2.0 0.0 2.0 2.0 0.0 2.0 S3 2.0 0.0 2.0 2.0 1.8 0.4 2.0 0.0 2.0 2.0 0.0 2.0

S4 2.0 0.0 2.0 2.0 0.0 2.0 2.0 0.0 1.9 0.3 2.0 0.0 S5 0.7 0.4 0.2 0.4 1.0 0.8 - - - - - - S6 1.1 0.7 1.9 0.3 2.0 0.0 1.4 0.8 0.7 0.6 1.4 0.6

S7 1.9 0.3 2.0 0.0 1.0 0.8 1.5 0.5 2.0 0.0 1.7 0.4 S8 0.7 0.4 1.3 0.6 1.7 0.4 1.0 0.0 1.4 0.5 1.3 0.4 S9 0.3 0.4 0.9 0.3 1.0 0.3 - - - - - -

S10 1.5 0.5 1.3 0.6 1.4 0.6 0.8 0.4 1.0 0.0 1.0 0.4 S11 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 S12 1.0 0.0 0.9 0.3 1.0 0.6 0.0 0.0 0.4 0.5 0.0 0.0


35

S13 0.0 0.0 0.4 0.8 0.0 0.0 - - - - - -

S15 0.0 0.0 1.8 0.6 1.8 0.4 0.0 0.0 0.0 0.0 0.3 0.6

S16 0.0 0.0 0.0 0.0 0.0 0.0 - - - - - -

S17 0.0 0.0 0.0 0.0 0.0 0.0 - - - - - -

S18 0.5 0.5 1.0 0.0 0.7 0.8 - - - - - - N1- Proper noun, N2 - pronoun, N3 - modifier, N4 – two word noun phrase, N5 - article, N7 – plural suffix, N8 – two word noun phrase before verb, N9 – three word noun phrase, N10 - adverb, N11 – other bound morphemes, V1 – verb, V2 - particle or preposition, V3 – prepositional phrase, V4 – copula linking two nominals, V6 - auxiliary, V7 – progressive suffix, V8 - adverb, V9 – modal preceding verb, V10 – third person singular present tense suffix, V11- past tense modal, V12 – regular past tense suffix, V13 – past tense auxiliary, V14 – modal adverb, V15 – copula, modal or auxiliary or ellipses, V16 – past tense copula, Q1 – intonationally marked question, Q2 – routine/ existence/ wh pronoun question, Q3 – simple negation, Q4 – initial wh pronoun followed by verb, Q6 – wh question with inverted modal, copula or auxiliary, Q7 – negation of copula, modal or auxiliary, Q8 – yes/no question with inverted modal, copula, or auxiliary, Q9 – why, when, which, whose questions, Q10 – tag question, S1 – two word combination, S2 – subject verb sequence, S3 – object verb sequence, S4 – Subject object verb sequence, S5 - conjunction, S6 – sentence with two verb phrase, S7 – conjoined phrases, S8 – infinitive marked with to, S9 - introducer, S10 – adverbial conjunction, S11 – propositional complement, S12 –conjoined sentences, S13 - wh clause, S15 – sentence with three or verb phrase, S16 – relative clause, S17 – infinitive clause, S18 - gerund, NL – children with normal hearing, HI – children with hearing impairment.

sequence (/z/ = 1.42, p < 0.05); S4, SOV sequence (/z/ = 1.00, p < 0.05); S7, conjoined phrase (/z/ = 0.23, p < 0.05); S8, infinitive verb (/z/ = 0.24, p < 0.05). In the section of question/ negation the other items were not considered in the manual, as the significant difference could not be established. As mentioned in the results, the increased frequency of occurrence of the items in the descending order and the comparable items are given in the Table 3. Several factors can be attributed to the above mentioned results. One such important factor is the properties of input which includes input frequency, perceptual salience, and semantics of the functors.

The increased frequency of occurrence of the

items modifier, two word noun phrase, two word noun phrase before verb, three word noun phrase, and adverbs, in children with normal hearing than children with hearing impairment could be accounted to the properties of input such as input frequency. This is supported by the studies conducted by Brown (1958), Naigles and Hoff-Ginsberg (1998). Although the input frequencies for both the groups are almost same, the children with hearing impairment mainly focus to the content words than the functors, because of their impairment. This will be reflected in their speech output also.

The second property of input could be the

perceptual salience which involves several factors such as phonetic substance, susceptibility to heavy stress and pitch, possibility of occurring in utterance final position. Blandell and Jensen (1970) claimed that these factors may help in the imitation skill of children. Brown (1976) reported that in English the inflections such as plurals, possessives, and regular past tense on verbs are not even fully syllabic and that there is a tendency for the above mentioned

features to be lost. Thus the chances of occurrence of these items in their speech will be less. He also suggested about some variations in these factors which can interfere with the development such as, the language of the child who is acquiring the grammar, the stage of the development etc. While in certain languages like German, some of these inflections are syllabic. As Malayalam is one such syllabic language, these parameters are stressed and thus there is a chance for early development of the inflections in both the groups.

Pressnell (1973) supported this in terms of

differences in the relative perceptual salience of morphological endings for children with normal hearing and for children with hearing impairment, who must rely on limited hearing in the low frequencies and on lip-reading.

The use of pronouns, proper nouns and

articles is not significantly different in either of the group which may be because of the over learning of these items in natural context. Brown and Fraser (1963) in their study of imitated morphemes in various syntactic classes showed that functors were omitted more often than content words. The scores for the items pronoun, proper noun, article and plural suffix usage can be supported by this study. Both groups showed higher scores for pronouns and proper nouns lower scores for article and plural suffix. The interaction of all the above mentioned factors can also be accounted for the differences and similarities observed across the groups.

In the verb phrase section, the mean values

were found to be more for items participle or preposition, regular past tense suffix, past tense copula in children with hearing impairment than the


36

Table 3. Frequency of occurrence of items in the descending order Noun phrase Verb phrase Questions / negation Sentence structure

N1 (proper noun) V1 (verb), V12 (regular past tense suffix), V16 (past

tense copula)

Q2 (routine / existence / wh pronoun questions), Q3 (simple

negation), Q4 (initial wh pronoun followed by verb), Q9 (why, when,

which whose questions)

S1(two word combination), S2 (subject

verb sequence), S3 (object verb sequence)

N2 (pronoun) V2 (participle or preposition),V3 (prepositional

phrase)

Q1 (intonationally marked question) S4 (subject object verb sequence)

N8 (two word noun phrase before verb)

V8 (adverb) S7 (conjoined phrases)

N5 (article) V7 (progressive suffix)

S8 (infinitive marked with to)

N4 (two word noun phrase)

V6 (auxiliary) S6 (sentence with two verb phrases)

N3 (modifier) V10 (third person singular present

tense suffix)

S10 (adverbial conjunction)

N9 (three word noun phrase)

V13 (past tense auxiliary)

S11(propositional complement)

N7 (plural suffix) S15 (sentence with three or more verb phrases)

N10 (adverb)

language age matched normals while the mean score for the item past tense auxiliary was less for the hearing impaired group than the other group. This could be due to fact that was described by Keren – Potnoy (2006) as children practice the use of new verbs by repeated trial and occasional error. That is they generate a large number of structures, most of them would be erroneous. These erroneous forms would be shaped into the correct form by the parental and environmental input, while children with developmental delay such as hearing impairment do not receive adequate resources for language development through the natural process of trial and error. The same reason can be accounted for the absence of a significant difference in the other items such as, prepositional phrase, copula linking two nominals, auxiliary, progressive suffix, adverb, third person singular present tense suffix for the two groups.

These findings were supported by Elbers and

Wijnen (1992) who suggested that language improves through practice. They also suggested two variables, intensive use and errors as the evidence for practice, which would be accounted for the difficulty and gradualness with which learning first occurs in most of the structures.

The above mentioned findings were

contradicted by Keren-Potnoy (2006). According to these authors the modern theories ignore the factors such as practice and problem solving in the

development of syntactic structures. They posit rote learning and learning triggered by innate knowledge as the major processes through which syntax is initially acquired. Rote learning could be attributed as a factor for the presence of the items participle or preposition, regular past tense suffix, and past tense copula in children with hearing impairment than the children with normal hearing.

In the third section, viz, question/ negation,

the item simple negation follows negation + sentence structure (IPSyn manual, Scarborough, 1990), which is not applicable to the syntactic structure of Malayalam grammar. In Malayalam, the structure is different when compared with English. The pattern of acquisition of negation is seemed to be different with respect to languages. That is in Malayalam the negation succeeds sentence (e.g. /aviťe po:ka٣da/) whereas in English negation precedes sentence (e.g. don‟t go). Further studies done by Bloom (1970) and Bowerman (1973) supported this view and he questioned the universality of acquisition of negation (negation + sentence form) reported by McNeill (1970). Formal training provided to the group of children with hearing impairment probably facilitated the early acquisition of negation.

The results showed reduced mean values for

some of the items, viz, routine/existence/wh pronoun questions, wh pronoun followed by verb questions, why/when/which/whose questions in children with normal hearing. This may be due to the factors like personality type, environmental constraints, culture etc. These factors can interfere with the performance


37

of children in both groups, but the formal training given to the children with hearing impairment facilitates their better performance.

The scores for the intonationally marked

questions were found to be less in the hearing impaired group than the normal group. This can be attributed to the fact that improper functioning of the auditory system, may deprive the linguistic input from others speech. Auditory deprivation arising from hearing loss in the early stages of life affects the different aspects of language development, including the patterns of speech production (Lee & Canter, 1971; Pressnell, 1973; McGarr & Osberger, 1978; Oller, Jensen & Lafayette, 1978; Quigley & King, 1982; Wood, 1984; Levitt, McGarr & Geffner, 1987; Madison & Wong, 1992; Tobin, 1997).

Among the items which are significantly

different across the two groups, conjoined phrases was found to have higher scores for children with hearing impairment. This could be due to the fact that functors such as conjunctions are often omitted in the spoken utterance of this population but the formal training provided places greater emphasis on the correct use of these functors. Factually however lacks of these functors are acceptable as elliptical

utterances when produced by children with normal hearing.

Similarly the item S3 (object – verb) was

found to be significantly more in children with hearing impairment, owing to the fact that, these children try to limit their speech output through the use of limited number of words, mostly content words like nouns and verbs.

Kruskal Wallis test was employed for the age

wise comparison of the two groups (children with normal hearing and children with hearing impairment) separately and the results are given in the Tables 4, 5, 6, and 7.

Following the Kruskal Wallis, Mann Whitney

test was carried out for those sub items, (which were significantly different) to see which age group is significantly different from the other for the two groups. The results showed that for the noun phrase, among the children with normal hearing, the performance of the age range of 4-5 years was significantly different (p< 0.05 = 0.002) from 3-4 and 5-6 years, for the sub item plural suffix. There was no significant difference noticed between the age groups of 3-4 and 5-6 years.

Table 4. Age wise comparison of noun phrase for normal and children with hearing impairmentGroups N1 N2 N3 N4 N5 N7 N8 N9 N10 N11

NL 1.00 0.07 0.14 0.10 0.32 0.00 * 0.19 0.29 0.11 1.00

HI 1.00 0.59 0.00 * 0.60 0.01* 0.21 0.39 0.00* 0.00*

*p<0.05 N1- Proper noun, N2 - pronoun, N3 - modifier, N4 – two word noun phrase, N5 - article, N7 – plural suffix, N8 – two word noun phrase before verb, N9 – three word noun phrase, N10 - adverb, N11 – other bound morphemes

Table 5. Age wise comparison of verb phrase for normal and children with hearing impairment

Groups V1 V2 V3 V4 V6 V7 V8 V9 V10 V11 V12 V13 V14 V15 V16 NL 1.00 0.15 0.36 0.04* 0.03* 0.06 0.89 0.18 0.00* 0.19 0.08 0.00* 0.34 0.12 0.05 HI 1.00 0.12 0.12 0.01* 0.00* 0.03* 0.00* - 0.67 - 1.00 0.13 - - 1.00

*p<0.05 V1 – verb, V2 - particle or preposition, V3 – prepositional phrase, V4 – copula linking two nominals, V6 - auxiliary, V7 – progressive suffix, V8 - adverb, V9 – modal preceding verb, V10 – third person singular present tense suffix, V11- past tense modal, V12 – regular past tense suffix, V13 – past tense auxiliary, V14 – modal adverb, V15 – copula, modal or auxiliary or ellipses, V16 – past tense copula.

Table 6. Age wise comparison of questions/negation for normal and children with hearing impairment *p<0.05 Q1 – intonationally marked question, Q2 – routine/ existence/ wh pronoun question, Q3 – simple negation, Q4 – initial wh pronoun followed by verb, Q6 – wh question with inverted modal, copula or auxiliary, Q7 – negation of copula, modal or auxiliary, Q8 – yes/no question with inverted modal, copula, or auxiliary, Q9 – why, when, which, whose questions, Q10 – tag question.

Groups Q1 Q2 Q3 Q4 Q6 Q7 Q8 Q9 Q10 NL 0.12 0.36 0.08 0.00* 0.57 0.00* 0.08 0.00* 1.00 HI 0.24 1.00 1.00 1.00 - - - 1.00 -


38

Table 7. Age wise comparison of sentence structure for normals and children with hearing impairment

*p<0.05 S1 – two word combination, S2 – subject verb sequence, S3 verb object sequence, S4 – Subject verb object sequence, S5 - conjunction, S6 – sentence with two verb phrase, S7 – conjoined phrases, S8 – infinitive marked with to, S9 - introducer, S10 – adverbial conjunction, S11 – propositional complement, S12 –conjoined sentences, S15 – sentence with three or verb phrase, S16 – relative clause, S17 – infinitive clause, S18 – gerund.

For the section on verb phrase, the test results

showed that auxiliary, third person singular present tense suffix and past tense auxiliary was significantly different (p < 0.05 = 0.011, 0.000, 0.000 respectively) for the age groups, 3-4 and 4-5years; the sub items, copula linking two nominals, third person singular present tense suffix and past tense auxiliary was significantly different (p < 0.05 = 0.029, 0.002, 0.009) for the age groups, 3-4 and 5-6years; and the results did not show any significant difference between the age groups, 4-5 and 5-6 years.

Mann Whitney test for the section questions/

negation, the result revealed a significant difference (p < 0.05 = 0.002, 0.000, 0.002) for initial wh pronoun followed by verb, negation of copula, modal, or auxiliary and why, when , which , whose questions between the age groups, 3-4 and 4-5 years; among the age groups 3-4 and 5-6 years, a significant difference (p < 0.05 = 0.000, 0.002) was noted for initial wh pronoun followed by verb & why, when, which whose questions; the age groups 4-5 and 5-6 years, was found to be significantly different (p < 0.05 = 0.000) for the items negation of copula, modal, or auxiliary.

Comparison between the age groups, for the

section sentence, showed a significant difference (p < 0.05 = 0.028, 0.007, 0.037, 0.008, 0.000) for the sub items, conjunction, sentence with two verb phrases, infinitive mar�ed with „to‟, introducer, sentence with three or more verb phases between the age groups, 3-4 and 4-5 years; sentence with two verb phrases, conjoined phrases, infinitive mar�ed with „to‟, and sentence with three or more verb phases was found to be significantly different (p < 0.05 = 0.002, 0.007, 0.001, 0.000) for the age groups, 3-4 and 5-6 years; and conjunction, conjoined phrase, and introducer was significantly different (p < 0.05 = 0.018, 0.002, 0.000) between the age groups, 4-5 and 5-6 years.

Similarly the Mann Whitney test results

conducted for children with hearing impairment revealed that, for the development of noun phrase, the age group 3-4 years showed a significant difference with 5-6 years (p < 0.05 = 0.001, 0.006, 0.000) for the items, modifier, article, and three word noun phrase; for the sub items, modifier and

adverb, 4-5 years age group was found to be significantly different from 5-6 years (p < 0.05 = 0.042, 0.027).

In the section verb phrase, for the items,

copula linking two nominals, auxiliary, progressive suffix, and adverb 3-4 years age group is significantly different from 4-5 years at the level of significance p < 0.05 = 0.029, 0.000, 0.024, 0.002 and 3-4 years is significantly different from 5-6 years at the level of significance p < 0.05 = 0.029, 0.022, 0.032 for the items, copula linking two nominals, auxiliary, and progressive suffix. The item auxiliary verb was found to be significantly different for the age groups 4-5 and 5-6 years at the level of significance p < 0.05 = 0.015.

In the third section questions/ negations, as

there was no significant difference found in the Kruskal Wallis test during age wise comparison, the Mann Whitney test was not administered.

In the last section sentence, the sub items,

conjoined phrases, and conjoined sentences was found to be significantly different (p < 0.05 = 0.012, 0.029) between the age groups of 3-4 and 4-5 years. No significant difference was present between the age groups of 3-4 and 5-6 years. S12 was found to be significantly different (p < 0.05 = 0.029) among the age groups of 4-5 and 5-6years.

From the above findings, it is clear that, the

age group 4-5 years is significantly different from 3-4 and 5-6 years in most of the items and the significant difference noted between the age groups 4-5 and 5-6 years is very less when compared with that of 3-4 years. This pattern of development is seen in the group of children with hearing impairment also. This can be explained by the reason that most of the linguistic development is occurring in the age range of 4-5 years and above. According to Tunmer and Bowey (1984) this age period is called as the middle childhood (the period from 4-8 years). He suggested that, while there is continuation of earlier development, an emergence of a new set of linguistic functions is also evident in this period. He referred to this new kind of linguistic functions as metalinguistic development.

Group S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S15 S16 S17 S18 NL 1.00 1.00 0.12 1.00 0.02* 0.00* 0.0* 0.00* 0.00* 0.81 1.00 0.82 0.12 0.00* 1.00 1 0.1 HI 11.00 1.00 1.00 0.36 - 0.06 0.04* 0.09 - 0.36 - 0.01 0.12 - - -


40

Table 8. Mean and standard deviation of the total score of, noun phrase, verb phrase, question/ negation, sentence structure and overall total scores across the variables hearing aid, severity of hearing loss and

duration of therapy taken

Variable Mean (M) &

Standard Deviation(SD)

NT VT QT ST IPST

HA PROG

M 9.96 16.26 8.84 13.38 48.46

SD 2.02 2.47 0.67 1.67 4.97

TRIM M 11.25 17.50 8.50 13.00 50.25 SD 2.50 2.88 0.57 0.81 4.57

SEV

MS M 9.66 16.33 9.00 13.33 48.33 SD 2.08 3.05 1.00 1.52 6.80

SEVE M 9.88 16.76 8.76 13.41 14.82 SD 2.08 2.10 0.66 1.69 4.33

PFD M 10.70 15.90 8.80 13.20 48.60 SD 2.21 3.14 0.63 1.54 5.75

DUR

0-1 years M 10.00 16.00 8.50 13.50 48.00 SD 1.41 4.24 0.70 1.11 6.30

1-2 years M 10.66 16.55 9.00 13.33 49.55 SD 2.59 3.28 0.70 1.11 6.30

2-3 years M 9.25 16.66 8.66 12.50 47.08 SD 176 2.10 0.77 1.24 4.33

3-4 years M 11.00 16.00 8.85 14.71 50.57 SD 1.82 2.16 0.37 1.97 3.77

NT – total score of noun phrase, VT – total score of verb phrase, QT – total score of questions/ negations, ST – total score of sentence structures, IPST – overall total score of IPSyn, HA – type of hearing aid, SEV – severity of hearing loss, DUR – duration of therapy taken, PROG – programmable digital hearing aid, TRIM – trimmer digital hearing aid, MS – moderately severe HL, SEVE – severe HL, PFD – profound Hl.

Table 9. Results of Mann Whitney test Table 10. Results of Kruskal Wallis test

Figure 2. Scatter plot depicting correlation between IPSyn and MLT.

the normal group. This is probably to do with the greater stress on these by oral as well as written mode in language therapy given to children with hearing impairment.

Conclusions

From the statistical test results, it was evident that the children with hearing impairment showed significantly poorer scores than the children with normal hearing in the items such as modifier, noun phrase, two word noun phrase, three word noun phrase, adverb, past tense auxiliary, intonationally marked question, sentence with two verb phrase, adverbial conjunction, conjoined sentences, sentence with three or more verb phrase.

The IPSyn scores revealed that the children with hearing impairment in the age range of 4-5 years showed a significant difference with the other two age groups, 3-4 years and 5-6 years. The performance of 5-6 years age group was better than

Variables NT VT QT ST IPST

HA 0.29 0.42 0.34 0.59 0.13

Variables NT VT QT ST IPST

SEV 0.72 0.78 0.88 0.85 0.94

DUR 0.12 0.82 0.47 0.01 0.25


41

the children with hearing impairment in the age range of 3- 4years. The results also revealed the presence of a positive correlation between the two syntactic assessment tools, Malayalam Language Test (MLT) and Index of Productive Syntax (IPSyn).

The test results lead to the conclusion that

there are some missing grammatical structures in the speech and an overall delay in the performance of children with hearing impairment in comparison to the language age matched normally hearing children. A lack of use of grammatical structures already developed and effectively used by the typically developing children was seen in children with hearing impairment. Thus the study revealed the presence of telegraphic speech in these children. Hence this resulted in a kind of restricted use of grammar, in the speech of children with hearing impairment which gives a picture of agrammatism.

References Berman, R. A. (1986). A crosslinguistic perspective:

Morphology and syntax. In P. Fletcher, & M. Garman, (eds.), Language acquisition (2nd ed.). Cambridge: Cambridge University Press.

Blandell, R., & Jensen, P. (1970). Stress and word position as determinants of imitation in first language learners. Journal of Speech and Hearing Research, 13, 193 – 202.

Bloom, L. (1970). Language and development: Form and function in emerging grammars. Cambridge, MA: MIT Press.

Bowerman, M. (1973). Early syntactic development: a cross – linguistic study with special reference to Finnish. London: C.U.P.

Brown, R. (1958). How shall a thing be called?. Psychological Review, 65, 14 – 21.

Brown, R. (1973). A first language: The early stages. Cambridge, MA: Harvard University Press.

Brown, R. (1976). A first language. Harmondsworth: Penguin Books.

Brown, R., & Fraser, C. (1963). The acquisition of syntax. In C. N. Cofer & B. S. Musgrave (Eds.), Verbal Behavior and learning. New York: McGraw-Hil1.

Cleary, M., Schwartz, R. G., Wechsler – Kashi, D., & Madelll, J. R. (2006). Hearing impairment in children associated with slower retrieval from memory during recall. Poster presented at the association for Psychological science 18th Annual Convention. New York, NY, May 25-28.

Crystal, D. (1995). Cambridge encyclopedia of the English Language. Cambridge: Cambridge University Press.

David, M. (2005). Dictionary of communication disorders (4th ed.). USA: Whurr Publications.

Elber, L., & Wijnen, F. (1992). Effort production skill and language learning. In C. A., Ferguson, L., Menn, & Stoel – Gammon (Eds.), Phonological development: models, research implications. Timonium: York Press.

Geffner, D. S., & Freeman, L. R. (1980). Assessment of language comprehension of 6-year-old deaf children. Journal of Communication Disorders, 13, 455-471.

Hakes, D. T. (1980). The development of metalinguistic abilities in children. Berling: Springer Verlag.

Keren-Potnoy, T. (2006). Facilitation and practice in verb acquisition. Journal of Child Language, 33, 487-518.

Lee, L. (1966). Developmental sentence types: A method for comparing normal and deviant syntactic development. Journal of Speech and Hearing disorders, 31, 311-330.

Lee, L. (1969). Northwestern syntax screening test. Evanston: Northwestern University Press.

Lee, L. L. (1974). Developmental sentence analysis. Evanston: Northwestern University Press.

Lee, L. L., & Canter, S. M. (1971). Developmental sentence scoring: a clinical procedure for estimating syntax development in children‟s spontaneous speech. Journal of Speech and Hearing Disorders, 36, 315- 340.

Levitt, H., Mcgarr, N. S., & Geffner, D. (1987). Development of language and communication skills in hearing impaired children. ASHA Monographs, 26, 1- 158.

Madison, C. L., & Wong, E. Y. (1992). Use the Clarck-Madison test of oral language with the hearing impaired: a content validity and comparative study. Journal of Communication Disorders, 25, 241 – 250.

Mcgarr, N., & Osberger, M. (1978). Pitch deviancy and intelligibility of deaf speech. Journal of Communication Disorders, 11, 237 – 247.

McNeill, D. (1970). The acquisition of language: The study of developmental psycholinguistics. New York: Harper & Row.

Miller, J. F. (1981). Assessing language production in children. Baltimore: University Park Press.

Moog, J. S., & Geers, A. E. (1975). Scales of early communication skills for hearing-impaired children. Central Institute for the Deaf; St. Louis, MO.

Moog, J. S., & Kozak, V. J. (1983). Teacher assessment of grammatical structures. St Louis, MO: Central Institute for the Deaf.

Moores, D. (1972). Language abilities of hearing-impaired children. In J. V., Irwin & M. Marge (Eds.), Principles of childhood language disabilities. Appleton-Century-Crofts, New York.

Nagapoornima, M. (1990). Disfluencies in children. An unpublished Master‟s dissertation to the University of Mysore, Mysore.

Naigles, L., & Hoff – Ginsberg, E. (1998). Why are verbs learned before other verbs? Effects of input frequency and structure on children‟s early verb use. Journal of Child Language, 25, 95-120.

Oller, D. K., Jensen, H. T., & Lafayette, R. H. (1978). The relatedness of phonological processes of a hearing-impaired child. Journal of Communication Disorders, 11, 97-105.

Pressnell, L. (1973). Hearing impaired children comprehension and production of syntax in oral language. Journal of Speech and Hearing Research, 16, 12-21.

Quigley, S. P., & King, C. M. (1982). The language development of deaf children and youth. In S. Rosenberg (Eds.), Handbook of applied psycholinguistics. (pp. 429 – 475). Hillsdale: Erlbaum Associates.

Quigley, S., Steinkamp, M., Power, D., & Jones, B. (1978). Test of Syntactic Abilities: Guide to administration and interpretation. Beaverton: Dormac.

Rukmini, A. P. (1994). Malayalam Language Test. An unpublished Master‟s dissertation submitted to the University of Mysore, Mysore.


42

Scarborough, H. S. (1990). Index of productive syntax. Applied Psycholinguistics, 11, 1-22.

Scarborough, H.S. (1989). Prediction of reading disability from familial and individual differences. Journal Educational Psychology, 81, 101 – 108.

Srinivasan, S. (1992). Disfluencies in children. An unpublished Master‟s dissertation to the University of Mysore, Mysore.

Swisher, L. (1976). The language development of the oral deaf. In H. Whitaker and H. A. Whitaker. Studies in Neurolinguistics. Academic Press, New York.

Tobin, Y. (1997). Phonology as human behavior: Theoretical implications and clinical applications. Durham: Duke University Press.

Tunmer, W. E., & Bowey, J. A. (1984). Metalinguistic awareness and reading acquisition. Part III (3.2). In W. E., Tunmer, C., Pratt, & M. L., Herriman, (Eds.), Metalinguistic awareness in children. New York: Berling: Springer – Verlag.

Webster, J., Morris, J., & Franklin, S. (2005). Effects of therapy targeted at verb retrieval and the realization of the predicate argument structure: A case study. Aphasiology, 19, 748-764.

Wood, D. (1984). Aspects of the linguistic competence of deaf children. British Journal of Audiology, 18, 23- 30.

Yamini, B. K. (1990). Disfluencies in children. An unpublished Master‟s dissertation to the University of Mysore, Mysore.

Revised Token Test in Oriya

43

Development of Revised Token Test in Oriya Bijoyaa Mohapatra 1 & S. P. Goswami 2

Abstract

The present study is aimed at adapting the Revised Token Test into Oriya (RTT-O) language based on the principles of the original English version (Mc Neil & Prescott, 1978) and Kannada version (Veena, 1982). The adaptation of the test was done in three phases- listing of the test stimuli, development of test material and the administration of the test. The scores obtained on the test were subjected to appropriate statistical analysis. The level of performances in terms of comprehension ability on the RTT-O in normal population across age group 20-60 years was documented and then compared with the performance of the aphasic group. The findings of the present study revealed that the performances of males and females across subtests in different age groups did not differ significantly. The younger age groups (20-30 years) were better in their performance in comparison to the older age groups (30-40, 40-50 and 50-60 years). Obvious differences were observed among the persons with respect to the response patterns such as promptness, responsiveness, self-correction and sub-vocal rehearsals. Aging, memory span, attention are described in context to this. Hence, a reorganization of the test is recommended thereof. A hierarchy in the magnitude of comprehension deficits was exhibited by the normal participants on the various subtests. The aphasic participants performed relatively poorer than the normal participants. Cognitive and/or auditory processing deficits have been implicated from the responses of most of the aphasics. This study underscores the importance of a thorough assessment of auditory comprehension and its implication in drawing a profile of a person with aphasia and documenting even the subtle deficits in auditory comprehension among the different aphasic types and that research should be orientated at ameliorating the language specific test development in a multilingual country as India. Key words: Revised Token Test, aphasia, auditory comprehension

phasia 1has been described as a multi faceted problem involving deficits in the communicative modalities of speaking,

reading, writing and listening; and manifesting greater impairment in language areas than other intellective or mental functions. Due to the complexity of the problem, diversity in the approaches and opinion can be seen at the level of definition. Hence, it is necessary that it has to be studied from different dimensions. Of the various views on aphasia, it has also been described as a „disorder of comprehension‟ with an inability to understand linguistic utterances that cannot be attributed to deficient sensory input or generalized cognitive deficits (Rosenbek, La Pointe & Wertz, 1989). Schuell and Jenkins (1961) and Smith (1971) even reported of auditory comprehension deficits that exist in all the cases of aphasia.

There are many factors which contribute to this comprehension difficulty and these factors are interacting. Some of these factors (speech sound and word meaning recognition) may be selectively impaired as a result of focal lesions, and may contribute to clearly defined aphasic syndromes. Other factors such as attention and short term auditory memory problems are more difficult to isolate and also they interact with the other two factors mentioned.

1e-mail: [email protected]; 2Reader in Speech Pathology, AIISH, Mysore, [email protected]

Aphasia research has shown that majority of the expressive problems are often associated with impairments of comprehension. However, the comprehension deficits in aphasia have been studied lesser than the expressive deficits. It can be possibly due to the fact that the analysis of comprehension needs to be based on observations of overt responses, which is usually confounded by the observed output deficits. A set of tests in common currency facilitates communication among clinicians and researchers, enabling them to compare and contrast persons under investigation from a known baseline. Kertesz (1988) has suggested that standardized assessments are necessary to diagnose the type of aphasia, allow a prognosis, and help in planning therapy. Thus, it should be taken care to devise a crisp and adequate test tool which would avoid the non-essential or frankly dubious aspects, and follow the essential components in the test structure. In the assessment of comprehension ability, the stimulus material will be presented to the subject, and his ability to comprehend will be inferred based on the response. A test of comprehension, thus, consists of carefully worked out administrative procedures, and stimulus materials that are designed to elicit relatively simple responses. The individuals with severe expressive deficits can also express their degree of comprehension using simple responses. The complexity of stimulus materials presented can also

A


44

be systematically varied across test batteries, and the stimulus material can be increased by increasing the number of items to be presented (such as phonemes, morphemes, phrases and sentences) in a test. The level at which the performance is disrupted due to information overload can be determined by systematically increasing the number of items to be processed both in terms of semantic content as well as syntactic complexity.

Keeping the above attributes in mind, it was decided upon to construct a test that would prove effective at assessing the comprehension ability of a person with aphasia.

Despite India having vast ethno-cultural

variations, still there have been very few attempts at studying and developing language test tools to assess brain damaged individuals in the Indian context. As far as concerned, the aphasia tests developed in Indian languages are intended mostly at testing the expressive ability of a person with aphasia in comparison to his comprehension ability. Of these various assessment tools developed in the Indian languages, Revised Token Test-Kannada (Veena, 1982) is the only test that has been developed to assess Kannada speaking persons with aphasia on their receptive abilities. With these aspects in mind, there developed a need to study and effectively assess the comprehension deficits in Oriya speaking persons with aphasia, Oriya being another Indian language. This led to the development of the test tool, Revised Token Test in Oriya.

The present study aimed at adapting the

Revised Token Test into Oriya language based on the principles of the original English version (Mc Neil & Prescott, 1978) and Kannada version (Veena, 1982) into Oriya language. The objectives of the study were to (1) determine the performances of normal participants on the Revised Token Test-Oriya (RTT-O), (2) to identify the level of performances in terms of comprehension ability in different age groups across the test (RTT-O), (3) to assess the variability in performance of participants across the complexity of the test (on various subtests) and (4) to compare the normal and the persons with aphasia on their comprehension abilities.

Method

Procedure: The „adaptation‟ of test was done in three phases. First phase included listing of the test stimuli, second phase included development of test material in Oriya and in the third phase, the test battery was administered on normals and four persons with aphasia.

Phase-I: Listing of the test stimulus: The first phase involved the development of the test stimuli in Oriya. Basically this test is a modified version of the already existing Revised Token Test in Kannada (Veena, 1982). For the present study, 5 objects, 2 colours, and 2 sizes were incorporated. The list of objects included were flower, comb, tumbler, pencil, and bangle. Each of the object was in either of the colour- red or yellow and of varying sizes as big and small. A total of twenty objects were considered as the test material. Phase-II: Development of test material: The Revised Token Test- Oriya (RTT-O) was developed after taking input from the Revised Token Test-English and Kannada. The RTT was translated into Oriya language by experienced Speech–Language Pathologists (SLPs) and linguists while considering the information from various sources like the RTT- English and RTT- Kannada. Syntactic and semantic aspects of the Oriya language was kept in mind while preparing the test material. The developed test was given to five SLPs who were proficient in reading and writing Oriya and have been exposed and trained on the language as a subject during the primary study career. The SLPs were asked to rate the test on various parameters on a scale, “Feedback Questionnaire for Aphasia Treatment Manuals” (Field Testing of Manual for Adult Non-fluent Aphasia Therapy in Kannada, MANAT-K; Goswami, Shanbal, Samasthitha & Navitha, 2010). Description of the test: The RTT-Oriya consisted of ten subtests which comprehensively assessed different command lengths and different sentence types involving the twenty test stimuli. Each subtest further included ten homogeneous items ranging in difficulty and linguistic construction. A total of twenty objects (tokens) were considered for the test, RTT-Oriya involving five objects of two colours and two sizes each. Distribution of subtest difficulty: The test consisted of ten subtests with increasing order of difficulty with each subtest. The test contained four different sentence types. The four sentence lengths in the Revised Token Test varied between 3, 4, 6 and 8 critical stimulus units in the commands (Park, McNeil & Tompkins, 2000). The sentence types included imperative prepositional phrases, adverbial clauses, simple imperatives and compound imperatives. The conjunction used in the RTT-O was and. The spatial pospositions used were on/above, below/under, by the side of, behind and in front of. The two left/right directional postpositions in Oriya were- left and right. The subordinate clauses used were- unless, either-or, instead of, if-then, if you have not-then.


45

Phase-III: Administration of the test: The normal participants were tested to form a baseline which was considered as normative for this test. The participants were seated comfortably; the objects were arranged according to the demands of the task of each subset and the instructions were given verbally. The test was conducted in a quiet environment. Participants: The test was administered on all the group of participants whosoever were available in colleges, clinics, work stations, and/ or houses. The details of the participants are provided in Table 1 and 2.

Table1. Details of the participants of the study

Type of Population

Age Range (in yrs)

Males Females

Normal participants

20-30 5 5 30-40 5 5 40-50 5 5 50-60 5 5

Persons with Aphasia

20-60 3 1

Table 2. Details of persons with aphasia

Procedure followed in test administration arrangement and placement: The tokens were arranged on a standard table in front of the participant, and the order of arrangement was kept the same always, for all groups of participants and all subtests. For all the odd-numbered subtests only big tokens (objects) were arranged and for the even numbered subtests both the small and the big tokens (objects) were arranged as per the subtest commands. Appropriate distance was maintained between the objects and rows. Each participant was seated in front of the table at a comfortable distance from where it was easy for him/her to reach and pick up the test material. The examiner sat to the left of the participant and slightly behind clearly out of his/her

working area and field of vision to avoid distractions that he/she might receive.

Ethical considerations were maintained and adhered to while selecting the participants (or family members/ care takers in case of persons with aphasia) were explained the procedure and the purpose of the study and an informal verbal/written consent was taken. They were randomly selected based on the following inclusionary criteria/s. (1) All the participants selected in the study were native speakers of Oriya. (2) The participants in the normal group had no past/present history of any neurological, psychological problems and sensory deficits and of any history of alcoholism or drug/abuse. (3) The persons with aphasia were identified through local hospitals, neurological clinics and/or speech and hearing centers. All of them had adequate sensory abilities (hearing and vision) and reported to be having no history of gradual deterioration in cognitive abilities. Test instructions and parameters: The participants were instructed to point out to the appropriate items, when named by the examiners. Prior to the actual test the participants were given pre-test instructions. If any part of the test trial was not performed correctly, the instructions were repeated. Introduction of subtest: The examiner administered the subtests while giving the commands. The behavior of the participant was observed during the command following which the participant‟s behavior and response was rated. Scoring pattern: A 15-point multidimensional scoring system was used to describe performance, and quantify deficits and differences among normal and pathological groups. McNeil and Prescott (1978) used the five dimensions of the PICA to construct a 15-point multidimensional scale in the Revised Token Test (RTT), in which each of the 15 scores represents a different combination of response adequacy on all five dimensions. This 15-point scale allowed for quantification of each response on the following dimensions: (a) Accuracy, (b) Responsiveness, (c) Completeness, (d) Promptness, and (e) Efficiency.

While administering the test, provision for

repeat cue was provided and while scoring this was taken into consideration. Repeats and cues were judged to be appropriate only by the verbs. The command was repeated when the participant did the wrong task (judged by verb only), or did nothing for 30 seconds, indicated that he/she has not understood the command, or requested for repetition of the command.

Participant No.

Age/Sex Cause and Aphasia type

1 35 yrs/M Cause-Thalamic hemorrhage Diagnosis- Global Aphasia

2 46 yrs/ M

Cause-Cerebro-vascular accident Diagnosis-Wernicke‟s Aphasia

3 57 yrs/M

Cause- Ischaemic infarcts of the thalamus Diagnosis-Transcortical aphasia

4 52yrs/ F Cause-Brain tumor Diagnosis- Sensory Aphasia


46

Table 3. RTT scale score category descriptions Score Description of Response

15 Complete 14 Vocal-Subvocal Rehearsal 13 Delay 12 Immediacy 11 Self-correction 10 Reversal 9 Repeat 8 Cue 7 Error 6 Perseveration 5 Intelligible/Rejection 4 Unintelligible (Differentiated) 3 Unintelligible (Perseveration) 2 Omission 1 No Response

The score sheet included a place to

accommodate the participant‟s demographic data (name, age, sex), handedness, diagnosis, age of onset etc. A place for the summarization and accessibility of the overall test time, mean overall score for all subtests, and the mean for each individual subtest is provided.

The time taken to complete the test was

approximately 20 minutes for normal participants and approximately one and a half hours for the persons with aphasia.

Statistical analysis: The normative values for each group were calculated separately and the mean scores were compared in all the age groups and between normal adults and the aphasic groups across all subtests. Statistical analysis for the above was done using SPSS software (Statistical Package for the Social Sciences package, version 16.0).


An „item analysis‟ for the test was conducted from the ratings of the five SLPs on the „Feedback Questionnaire for Aphasia Treatment Manuals‟. The SLPs rated the test to be „good‟ to „excellent‟ on various parameters as Simplicity, Proverbiality, Size of the picture, Colour and Appearance, Arrangement etc. indicating that the test has enough implications in its scope of practice and generalization of the test commands for intervention purposes in persons with aphasia. The findings of the present study have been broadly presented and discussed under different headings.

I. Performances of males and females across subtests in different age groups: Mean and standard deviation (SD) were calculated for all the dependant variables i.e. subtests of RTT-O. Table 4 illustrates these scores for all the participants. From the mean values it can be inferred that, on an average

most of the participants obtained either a mean of 15.00 or 14.00, indicating that the responses were either complete or in the form of sub-vocal rehearsals. Based on the quantitative data, , a Mann-Whitney U-test was conducted within each age group for all the subtests and the overall mean to detect the differences between them, if any. The results revealed that, there was no significant difference between the males and females on any of the subtest or on groups at p<.05 (i.e. the males and females performed equally well on all subtests). Owing to this very reason, males and females were combined and considered as one single group, for further analyses i.e. consideration of ten participants instead of five males and five females in each age group. Moreover, there were only five participants of each gender in each age group (which is not a large sample to be considered). Hence, gender was not regarded as an independent variable on all advanced analyses.

Table 4 summarizes the qualitative responses in male and female participants across ages. It is evident that the younger age groups 20- 30 and 30- 40 years responded mostly in a normal manner without needing any extra information to perform the task. The responses of these participants were mostly complete type and few of them on rare occasions showed sub-vocal rehearsals, whereas in the older age groups (40- 50 and 50- 60 years), few of the participants even scored as poorer as delayed, immediate and self correction type of responses. To generalize, it can be put down that the normal responses across the test varied between complete and subvocal rehearsal, thereby paving the idea that the Revised Token Test involves tasks that are well suited to identify an individuals‟ auditory comprehension levels.

Table 5. Types of responses of participants across

gender in various age groups

Many authors have reported that there is a

difference in the performance between the two tasks in comprehension and reception tasks i.e females are considered to perform better on comprehension task

Age Group

Gender Type of responses in various age groups

C SVR D I SC 20-30 M

F 30-40 M

F 40-50 M

F 50-60 M

F Note. C=complete, SVR=Sub-vocal rehearsals, D=Delayed, I=Immediate, SC= Self Correction


47

better than males (Ramer, 1976). Moreover, the test encapsulates a series of cognitive processes in addition to auditory comprehension, including working memory (Kitson, 1985; Emery, 1986), analysis of the whole into a series of items, or the ability to adequately ignore automatically evoked, distracting stimuli. Such factors are evidently distinct in both the gender. The present study, thus, provides corroborative evidences to the existing research that gender variation in normal persons may not be a major variable in comprehension task in auditory mode only. Either there are no differences in the performances of the two groups, or if present, may be subtle in nature.

II. Overall performance of the participants in the various subtests: A mixed ANOVA was administered to find out the differences in subtests (where age served as the independent factor) and across age groups (where subtests was considered as the independent factor). a. Comparison of the performance on subtests: Table 5 and Figure 1 give the mean scores of age groups across each subtest. It is evident that 20- 30 years age group scored the highest and the 50-60 years age group scored comparatively the lowest, with 30-40 years and 40-50 years preceding it. However, the meager mean differences were evident in the performances of the participants.

Figure 1. Mean scores of the different age groups on

various subtests.

The results of the present study are in accordance to the results of various studies on Token Tests where researchers have confirmed that age-related effects in normal population although not very distinct, were still observed with elderly participants scoring comparatively low than the others (De Renzi & Faglioni, 1978; Emery, 1986; Pena-Casanova, Quinones-Ubeda, Gramunt-Fombuena, Aguilar, Casas, Molinuelo & Blesa,

2009). Researchers have also reported that the differences in scores could be that older individuals face difficulty in retaining the auditory stimuli for a longer duration in comparison to the younger individuals. They even face problems in retaining verbal memory span (Albert & Bear, 1974). It is also evident that the scores of subtests in each age group indicate that there is deterioration in comprehension from the S-I to S-X subsections in a gradual fashion. This is because of its minimal redundancy, where the participants are required to understand the significance of each word in a series of increasingly complex commands (Goswami, 2004). Moreover, on RTT the linguistic stimuli are presented only in the verbal mode, hence the participants do not get any additional cues, graphic or visual to aid in comprehension. Next to complete responses, sub-vocal rehearsals were noticed in few normal participants, which is an indication that they rely on their auditory feedback for comprehension.

On mixed ANOVA (repeated measures

ANOVA with age as independent factor), a significant difference was observed between the subtests [F (9,234) =10.937, p<0.001]. Therefore, subsequently, a post- hoc Bonferroni test for pairwise comparison was done to find out the subtests which differed significantly.

The results obtained are attributed to various reasons, first of all, subtest I involves easiest of the commands hence, it differs from subtest S-III, S-IV, S-V, S-VI and S-X. The increasing complexity of the tasks in terms of grammaticality brings about a difference in the overall scores. Similarly, the significant difference of other subtests from the rest is because of the increase in sentence length and linguistic complexity. The second fact evident is that the even numbered subtests are only an increase in the linguistic level of the immediately preceding odd subtests and is hence, predictable by the participant after performing on the odd ones. Moreover subtests (except I and II) integrate two step tasks which demands an extra cognitive load on the participant and demands on the performance thereby are expected to increase (Goswami, 2004).

b. Comparison of the performance on age groups: A significant difference on Mixed ANOVA was observed between the performances of various age groups for F(3, 36)=5.895, p<0.001 as the participants of the various age groups exhibited comprehension deficits to varying degrees of severity. Post-hoc Duncan‟s mean range test (5% level of significance) (DMRT) elaborated on the differences between the various age groups.


48

Although a gross view of the Figure2 reveals no evident differences yet a critical analysis

MALES FEMALES TOTAL SUBTESTS GROUPS Mean SD Mean SD Mean SD

S I 20-30 15.00 0.00 15.00 0.00 15.00 0.00 30-40 15.00 0.00 15.00 0.00 15.00 0.00 40-50 15.00 0.00 15.00 0.00 15.00 0.00 50-60 14.99 0.00 14.80 0.44 14.89 0.31

TOTAL 14.99 0.00 14.95 0.22 14.97 0.15 S II 20-30 15.00 0.00 15.00 0.00 15.00 0.00

30-40 14.80 0.44 14.88 0.15 14.84 0.31 40-50 15.00 0.00 14.60 0.54 14.80 0.42 50-60 15.00 0.00 14.40 0.54 14.70 0.48

TOTAL 14.95 0.22 14.72 0.43 14.83 0.36 S III 20-30 14.99 0.00 14.94 0.10 14.97 0.07

30-40 14.99 0.00 14.99 0.00 14.99 0.00 40-50 14.45 0.79 13.85 0.78 14.15 0.80 50-60 14.25 0.75 14.05 0.94 14.15 0.80

TOTAL 14.67 0.60 14.46 0.77 14.56 0.69 S IV 20-30 14.62 0.60 14.66 0.52 14.64 0.53

30-40 14.97 0.05 14.13 0.90 14.55 0.74 40-50 14.24 0.75 13.91 0.90 14.07 0.80 50-60 14.24 0.75 14.10 1.02 14.17 0.85

TOTAL 14.52 0.64 14.20 0.84 14.36 0.75 S V 20-30 14.91 0.19 14.97 0.05 14.94 0.14

30-40 14.44 0.56 14.79 0.45 14.61 0.51 40-50 14.75 0.43 14.36 0.63 14.55 0.55 50-60 14.12 0.83 14.36 0.63 14.24 0.71

TOTAL 14.56 0.59 14.62 0.53 14.59 0.56 S VI 20-30 14.72 0.42 14.79 0.44 14.75 0.41

30-40 14.70 0.64 14.55 0.61 14.63 0.59 40-50 14.70 0.44 14.11 0.55 14.40 0.56 50-60 14.29 0.84 14.11 0.55 14.20 0.68

TOTAL 14.60 0.59 14.39 0.58 14.50 0.59 S VII 20-30 15.00 0.00 15.00 0.00 15.00 0.00

30-40 15.00 0.00 14.99 0.00 14.99 0.00 40-50 14.96 0.08 14.51 0.50 14.73 0.41 50-60 14.96 0.08 14.51 0.50 14.73 0.41

TOTAL 14.98 0.06 14.75 0.40 14.86 0.31 S VIII 20-30 14.97 0.15 14.74 0.49 14.86 0.35

30-40 14.80 0.44 14.75 0.54 14.77 0.47 40-50 15.00 0.00 14.95 0.10 14.97 0.07 50-60 15.00 0.00 14.95 0.10 14.97 0.07

TOTAL 14.94 0.22 14.85 0.35 14.89 0.29 S IX 20-30 14.78 0.45 14.72 0.44 14.75 0.42

30-40 14.68 0.68 14.92 0.15 14.80 0.48 40-50 14.70 0.53 14.57 0.58 14.63 0.53 50-60 14.50 0.58 14.49 0.62 14.49 0.56

TOTAL 14.66 0.53 14.67 0.47 14.67 0.50 S X 20-30 14.79 0.45 14.79 0.45 14.79 0.42

30-40 13.41 0.56 13.36 0.45 13.38 0.48 40-50 14.94 0.10 13.88 0.78 14.41 0.77 50-60 14.59 0.90 13.88 0.78 14.23 0.88

TOTAL 14.43 0.81 13.97 0.79 14.23 0.82 OVERALL 20-30 14.88 0.12 14.86 0.09 14.87 0.10

30-40 14.68 0.13 14.63 0.20 14.66 0.16 40-50 14.77 0.15 14.37 0.27 14.57 0.29 50-60 14.59 0.15 14.36 0.28 14.18 0.24

TOTAL 14.73 0.16 14.56 0.29 14.64 0.25

Table 5. Mean and SD values for male and female participants (normals) across subtests


49

on the post hoc test reveals that the age groups 30-40, 40-50 and 50-60 years perform similarly on the test (without any significant differences) whereas they differ significantly from the age group of 20-30 years. A study by Blumenfeld, Schroeder, Ali and Marian (2009) on inhibition and facilitation in auditory comprehension across the lifespan shows that both accuracy rates and response latencies in word comprehension decline with age. Age-related effects were reported as few in a series of papers on Token Test and its versions (De Renzi & Faglioni, 1978; Ivnik, Malec, Smith, Tanglos & Peterson, 1996).

c. Interaction of the subtests and groups: On repeated measures ANOVA, interaction effect between the subtests and groups was evident {4 (Group) x 10 (Subtests)}. Results revealed that a significant interaction was present between them at F (27,324) =2.89 for p<0.001.

III. Comparison of normals in various subtests of RTT in the different age groups: Since repeated measures ANOVA revealed a significant interaction effect between age groups and subtests; hence comparisons were made to determine the age group differences varying in magnitude within each subtest and comparison of subtests within each age group.

Table 6. Comparison of overall mean of the part Subtests F (3,36) Sig.

S-I 1.04 0.38 S-II 1.21 0.31 S-III 7.04 0.00* S-IV 1.37 0.26 S-V 2.99 0.04* S-VI 1.82 0.16 S-VII 2.70 0.06 S-VIII 1.04 0.38 S- IX 0.72 0.54 S-X 7.85 0.00*

Overall Mean 5.89 0.00*

a. Comparison of age groups within each subtest: Collapsing results across all subtests, a Multivariate Analysis of Variance (MANOVA) was performed. The MANOVA revealed statistically significant differences on subtests. Table 6 lists the ten subtests for which a significant difference was found for the total scores.

With age as independent variable and dependent variables as subtests S-I to S-X and overall mean, a MANOVA depicts that subtests S-III, S-V, S-X and overall mean present with a statistically significant difference at p<0.05 for F(3, 36) equaling to 7.04, 2.99, 7.85 and 5.89 respectively. Various inferences are attributable to the above results. From the RTT-O it is seen that every subtest is a transition to the preceding pair of

even and odd subtests, such as, subtests III (and IV), V (and VI), and IX (and X) have a slightly different complexity in their grammatical structure than the rest (I and II, VII and VIII). Unlikely the subtest VII (and VIII) did not show any evident difference in scores from the simple imperative sentences as in subtest I (and II), which could be likely that the left-right postpositional scores have not been encountered by the participants to be as tough as the III, V, X subtests and thus has elicited scores following that of subtests I and II. This trend in scores is likely to suggest that a reorganization of the subtests could be effective, i.e. the subtests VII and VIII dealing with left-right postpositional commands could succeed immediately after the subtests I and II and take over places of III and IV in terms of complexity. Thus, it is suggested that the sequence of subtests in the RTT-O can be reorganized as– I, II, VII, VIII, followed by III, IV, V, VI, IX and X.

Further post- hoc Duncan Mean Range test

elaborated on the differences between the age groups on each of the significant difference of the RTT-O. The Duncan‟s mean range test at 5% level of significance was conducted for S-III, S-IV, and S-V subtests and overall mean scores of the RTT-O. On different subtests adjacent groups were similar in their performance while in others younger age groups were significantly different from the older ones. b. Comparison of subtests within each age group: Repeated measure of ANOVA was conducted for each of the age groups to determine the difference in subtest scores, if any. Appropriate statistical tests like tests of within-participants, post-hoc pairwise comparison on Bonferroni test and paired t- test was conducted.

On a careful observation of the findings

revealed from the tests, it can be summarized that, in the older age groups (40-50 years and 50-60 years),

Figure 2. Overall mean scores on all the subtests produced by the different age groups.


50

the scores on the various subtests follow a similar pattern of differences. The fact that the higher age groups perform similarly and that the complexity of the task encountered by them is the same can be explained by the fact that language comprehension and production plays a major role in aging (Gutbrod, Meger, Meter & Cohen, 1985). IV. Comparison of normal population and the brain damaged individuals (persons with aphasia) on comprehension abilities: From the Table 7 it is viewed that, the normal participants exhibited significantly better comprehension as compared to the aphasics (persons with aphasia) on all the linguistic levels (subtests I to X) at which comprehension was assessed. It is evident from the Table 7 that the mean scores in each subtest in the normal participants‟ decreased with complexity of tasks. From the scores of both the groups of participants (the normals and brain damaged), it is observed that highest mean scores were observed in subtest I and lowest in subtest X. The mean scores in subtest I and subtest X were 14.97 and 14.0 and 8.25 and 0.00 respectively in normal and brain damaged participants. The mean scores of other subtests fell within these ranges. However, the limitation of the comparison is that in one group there were forty participants (normals) and in the other group only four participants (persons with aphasia).

The overall mean scores on all the subtests by

the normals and aphasics and their individual scores on the subtests is depicted in Figures 3 and 4 respectively.

The above attributions can be intended at the brain damage in aphasics, which causes a deficit in individual‟s ability to comprehend linguistic stimuli (Caramazza & Zurif, 1976; Peach, Canter & Gallaher, 1988).

A further comparison on Mann– Whitney U-

test (Table 8) showed that RTT scores for the persons with aphasia were significantly poorer than those of the normal participants at 0.05 levels. Thus, the persons with aphasia performed significantly less in auditory comprehension capabilities than normal population of the same age range.

The influence of stimulus length on

comprehension has also been reported by Shewan and Canter (1971) and Curtis, Jackson, Kempler, Hanson & Metter, 1986. These authors reported that other factors held constant, the sentence comprehension tends to decrease as length increases. The deterioration of comprehension with increase in sentence length is indicative of retention deficits in persons with aphasia. Comprehension of grammatical elements requires intact reasoning skills and good attention and memory span (as the length

and complexity of the stimuli increases), and deficits in all of these cognitive processes are implicated in persons with aphasia (Wright & Newhoff, 2004). Aphasic individuals do better on single word comprehension tasks when written and auditory stimuli are used instead of auditory stimuli alone (Schuell & Jenkins, 1961). Goswami (2004) reported that the repetition of linguistic command also improved the performances of these persons in the form of sub-vocal rehearsals. Table 7. Summaries of mean and standard deviation

(SD ) scores on subtests in normal and aphasic groups

Table 8. Results of Mann- Whitney U-test

(comparison of normals and aphasics) Subtests |Z| Sig.

S-I 5.47 0.00* S-II 4.17 0.00* S-III 3.53 0.00* S-IV 3.35 0.00* S-V 3.46 0.00* S-VI 3.38 0.00* S-VII 4.05 0.00* S-VIII 4.29 0.00* S- IX 3.34 0.00* S-X 3.30 0.00* Overall Mean 3.26 0.00*

The results of the following study can be

summarized as follows (1) The performances of males and females across subtests in different age groups did not differ significantly. (2) On a comparison of the performances on the various subtests, it was evident that subtests III, IV, V, VI and X differed significantly from Subtest I. (3) Critical analysis of the age groups reveal that 30-40, 40-50 and 50-60 years performed similarly and were different in their scores from the 20-30 years. (4) An interaction effect was observed between the subtests and groups which was present at 0.05 level. (5) In a

Normals (N=40) Aphasics (N=4) Subtests Mean S.D Mean S.D

S-I 14.97 0.15 8.25 3.30 S-II 14.83 0.36 7.75 3.30 S-III 14.56 0.69 6.07 3.17 S-IV 14.36 0.75 2.24 1.25 S-V 14.59 0.56 2.55 1.44 S-VI 14.50 0.59 1.58 1.33 S-VII 14.86 0.31 2.82 1.90 SVIII 14.89 0.29 1.88 1.45 S-IX 14.67 0.50 1.50 1.00 S-X 14.20 0.82 1.00 1.00

Overall Mean

14.64 0.25 3.36 1.16


51

Figure 3. Overall mean scores on all the subtests as produced by the two groups (normals and aphasics).

Figure 4. Mean scores of normals and persons with aphasia on various subtests.

comparison of subtests across age groups, S-III, S-V, S-X and overall mean present with a statistically significant difference where adjacent groups performed comparably than the non adjacent ones (except on S-X). (6) The older age groups (40-50 years and 50-60 years presented a similar pattern of differences, that is, in both the age groups the following pairs showed significant differences: S-I from S-III and S-V, S-II from S-VI, SIII from S-VII and SVIII from S-III, S-IV, S-V, S-VI, S-X. (7) The normal participants exhibited significantly better comprehension as compared to the aphasics (persons with aphasia) on all the linguistic levels (subtests I to X).

Therefore the present test satisfies the criteria of a good and useful test for aphasia. These facts indicate RTT to be one of the most elaborate clinical tools that help in terms of assessing an individual‟s auditory comprehension with stimuli of gradual complexity, and thereby identifying the auditory deficits in a person with aphasia. This study (adapting RTT in Oriya), thus, underscores the importance of a thorough assessment of comprehension and its implication in drawing a profile of a person with aphasia and documenting even the subtle deficits in auditory comprehension among the different aphasic types.

Conclusions

The present study was aimed at drawing some theoretical conclusions as obtained on the RTT-O which is listed below: (1) On the auditory comprehension test, the performance of the males and females was similar across all the subtests of RTT-O. The gender of a person seems not to influence his comprehension of spoken messages to any significant extent. (2) On the RTT-O, which assesses syntactic comprehension in the auditory modality alone, the deficits varied to different degrees, and also had qualitative differences in the responses of the normal participants. (3) The scores of normals on the RTT-O fell mostly in complete (score of 15) or in the vocal-sub vocal rehearsal (score of 14) range. This indicates that on assessment, if a native Oriya speaking participant scores 14 or 15 on commands, mean values and/or subtests, then it can be inferred that the person is performing in a normal manner; and a score less than 14 would account that there is a deficit in the auditory comprehension skills of the person and a treatment in that area is intended thereof. (4) The sentence length, linguistic complexity, grammatical usage and number of critical stimulus in the commands were factors that contributed to the test results. The scores on the third, fourth, fifth, sixth and tenth subtests were evidenced to differ from that of subtest one. (5) The younger age groups (20-30 years) were better in their performances in comparison to the older age groups (30-40, 40-50 and 50-60 years). Obvious differences were observed among the persons with respect to the response patterns such as promptness, responsiveness, self-correction and sub-vocal rehearsals. Aging, memory span, attention are described in context to this. Hence, a reorganization of the test is recommended thereof. (6) There was a hierarchy in the magnitude of comprehension deficits exhibited by the normal participants on the subtests. The trend of scores was highest for subtest I followed by II, VII, VIII, III, IV, V, VI, IX and X. (7) The aphasic participants performed relatively poorer than the normal participants. Cognitive and/or auditory processing


52

deficits have been implicated from the responses of most of the aphasics.

The RTT-O results are conclusive of the

fact that the difference in performances within normals and in persons with aphasia intend towards the effectiveness of the test in fulfilling its purpose of assessing auditory comprehension; and the variation in each category of stimuli can be considered indirectly to be measuring the linguistic competence of normals as well as the persons with aphasia.

Results of this study present the importance of

having testing tools that assess a person‟s skills in one of the modalities (or dimensions), like, RTT-O is projected at assessing the comprehension in terms of auditory modality only, and moreover out of all linguistic levels it is focused at measuring solely the syntactic level. Results even underscore the fact that research should be orientated at ameliorating the language specific test development in a multilingual country as India, and that it caters to the needs of all the assessors within a broad work culture. The normatives obtained on the RTT-O, as put forth, have been collected from a group of persons who belong to a part of eastern India, and thus acknowledge the fact that these performances (scores) can be accepted and generalized to the wider dimension of inhabitants residing in this region of the country.

References Blumenfeld, H. K., Schroeder, S. R., Ali, Z. R., & Marian,

V. (2009). Inhibition and facilitation in auditory comprehension across the lifespan. Paper presented at the COGSCI 2009-annual meeting of the cognitive science society. Retrieved from http:// csjarchive.cogsci.rpi. edu/ Proceedings/ 2009/ papers/ 49/ paper49.pdf

Caramazza, A., & Zurif, E. B. (1976). Denunciation of algorithmic and heuristic processes in language comprehension: Evidence from aphasia. Brain and Language, 3, 572-582.

Curtis, S., Jackson, C. A., Kempler, D., Hanson, W. R., & Metter, E. J. (1986). Length vs. structural complexity in sentence comprehension in aphasia. In R. H. Brookshire (Eds.). Clinical aphasiology (pp. 45-53), Minneapolis, MN: BBK.

De Renzi E., & Faglioni P. (1978). Development of a shortened version of the Token test. Cortex, 14, 41–49.

Emery, O. B. (1986). Linguistic decrement in normal aging. Language and Communication, 6, 47–64.

Goswami, S. P. (2004). Comprehension deficits in aphasics. Unpublished Doctoral Thesis, University of Mysore, Mysore, India.

Goswami, S. P., Shanbal, J. C., Samasthitha, S., & Navitha, U. (2010). Field testing of Manual for Adult Non-fluent Aphasia therapy- Kannada (MANAT-K). Research Project, All India Institute of Speech and Hearing Research Fund, Mysore, India.

Gutbrod, K., Meger, B., Meter, E., & Cohen, R. (1985). Cognitive processing of tokens and their description in aphasia. Brain and Language, 25, 37–51.

Ivnik, R. J., Malec, J. F., Smith, G. E., Tangalos, E. G., & Petersen, R. C. (1996). Neuropsychological tests norms above age 55: COWAT, BNT, MAE Token, WRAT-R, Reading, AMNART, Stroop, TMT, and JLO. The Clinical Neuropsychologist, 10, 262–278.

Kertesz, A. (1988). What do we learn from recovery of aphasia? Advances in Neurology, 47, 277-292.

Kitson, D. L. (1985). Comparison of the Token test of language development and the WISC-R. Perceptual and Motor Skills, 61, 532–534.

McNeil, M. R., & Prescott, T. E. (1978). Revised Token Test. Austin, TX: Pro-Ed.

Peach, R., Canter, G., & Gallaher, A. (1988). Comprehension of sentence structure in anomic and conduction aphasia. Brain and Language, 35, 119-135.

Pena-Casanova, J., Quinones-Ubeda, S., Gramunt-Fombuena, N., Aguilar, M., Casas L., Molinuelo, J. L., & Blesa, R. (2009). Spanish Multicenter Normative Studies (NEURONORMA Project): Norms for Boston Naming Test and Token Test. Archives of Clinical Neuropsychology, 24, 343–354.

Rosenbek, J., LaPointe, L., & Wertz, R. (1989). Aphasia: A clinical approach. Boston: College Hill Press.

Schuell, H., & Jenkins, J. (1961). Reduction of vocabulary in aphasia. Brain, 84, 243-261.

Shewan, C.M., & Canter, G.J. (1971). Effects of vocabulary, syntax and sentence length on auditory comprehension in aphasic patients. Cortex, 7, 209-226.

Smith, A. (1971). Objective indices of severity of chronic aphasia in stroke patients. Journal of Speech and Hearing Disorders, 36(2), 167–207.

Veena, N. R. (1982). Revised Token Test in Kannada. Unpublished Masters Dissertation, University of Mysore, Mysore, India.

Wright, H. H., & Newhoff, M. (2004). Primary auditory comprehension in aphasia: Facilitation and interference effects. Aphasiology, 18 (5/6/7), 555-565.

Kannada articulation test

53

Re - Standardization of Kannada Articulation Test Deepa Anand1 & S.R. Savithri2

Abstract

The present study developed, re-standardized and validated the test of articulation in Kannada. Specifically photographs were familiarized and a photo articulation test of Kannada was developed. This was administered on 240 typically developing children in the age range of 2-6 years. The effect of age and gender on the articulation abilities in children was investigated. Further the test was administered on 10 clients with misarticulation to find the validity. The results indicated main effects of age and gender on the score of the articulation test. The scores increased with increase in age and girls had significantly better scores compared to boys. Key words: articulation, re-standardization, misarticulation, validity

rticulation 1refers to the totality of motor processes involved in the planning and execution of sequences of overlapping

gestures that result in speech (Fey, 1992). Articulation errors are typically classified according to the child‟s age, which translates into stages within this developmental process. A child is said to have an articulation disorder when their error patterns and/or sound acquisition sequence deviate from those seen in most children of their age. The deviances in articulation could be due to organic factors, emotional conflicts, perceptual deficiencies, and difficulties in phonetic discrimination, poor motor coordination, poor model or functional. Evaluation of an individual‟s articulation involves description of his or her speech sound production and relating this to the normal or standard in the language and community. In order to evaluate the articulation of clients with articulation disorders, tests of articulation are essential. An articulation test is an evaluation that yields information about the nature, number and characteristics of articulatory errors as they occur in a person‟s speech. It is a technique employed to measure the general phonemic capacity of an individual. The purpose of articulation test is to compare the phonemes that are actually used by an individual with the phonemic structure of his language group. The test of articulation is a basic tool of the speech pathologist.

The purpose of articulation test varies and

hence the nature and scope of the articulation test inventory varies. Traditionally, Van Riper and Irwin (1958) define an articulation test as a technique employed to measure the general phonemic capacity of an individual. Articulation tests can be used in screening, diagnosing, predicting articulation disorders or deep testing. Over the years, many investigators have developed and established norms

1e-mail: [email protected]; 2Professor of Speech Sciences, AIISH, Mysore, [email protected].

using various articulation tests as shown in the Tables 1 and 2.

Babu, Rathna and Bettagiri (1972) developed

Kannada diagnostic articulation test. It is a four part test with 52 and 49 picturable words in part I and II, respectively. Part III has 10 clusters. Part IV has a reading passage which is administered to subjects who can read. Part I tests 10 vowels in initial position and not in medial and final position as they can be influenced by the consonants preceding them. Part II is similar to Part I but has different words with the same sounds being tested. Phonemes misarticulated in part I are tested in Part II.

Tasneem (1977) standardised the Kannada

diagnostic articulation test. However, it has been more than 30 years that norms were established for Kannada articulation test. Children are acquiring proficiency in articulatory skills at an earlier age than would be expected from previously established

norms. Recent studies (Roberts, Burchinal & Footo, 1990; Bharathy, 2001; Rahul, 2006; Sreedevi & Shilpashree, 2008) focusing on phonological processes have revealed suppression of most of the processes by the age of 3- 4 years.

Results of an exploratory study on articulation

showed that children were acquiring proficiency in articulatory skills at an earlier age than would be expected from previously established norms. These results indicate a need for new normative data

consistent with the performance of children seen at the present time (Arlt & Goodban, 1976).

As children of present generation are more

exposed to different environments at a very early age due to advancement in technology, education, nurture, awareness, increased speech stimulation, it is very important to re-standardize and validate the test. In this context the present study was planned. The aim of the study was to develop, re-standardize and

A


54

Table 1. Age levels for the speech sound development according to different authors in English

“-” indicates the sound not acquired; Empty space indicates speech sounds not tested (IP: Initial position, FP: final position)

Table 2. Age levels for the speech sound development according to different Indian authors

*criteria for the sound to be considered as acquired. “-” indicates sound not met the criteria

Speech sounds

Wellman et al.,

(1931)

Poole (1934)

Templin (1957)

Mecham (1962)

Sander (1972)

Prather et.al., (1975)

Arlt & Goodban (1976)

Irwin et.al., (1983)

Smit (1990)

Fudala & Reynolds (2000)

IP FP /m/ 3 3 ½ 3 3.5 Below 2 2 3 1.5 3 2 2 /n/ 3 4 ½ 3 3.5 2 2 3 2 3 2 2.5 /h/ 3 3 ½ 3 3.5 Below 2 2 3 2 3 2 - /p/ 4 3 ½ 3 3.5 Below 2 2 3 3 3 2 3 /f/ 3 5 ½ 3 4.5 3 2-4 3 3 3 3 3 /w/ 3 3 ½ 3.5 Below 2 2-8 3 2 3 2.5 - /b/ 3 3 ½ 4 3.5 Below 2 2-8 3 1.5 3 2 3 /┕/ -- 4 ½ 3 3.5 2 2-8 3 3 7-9 - 4 /j/ 4 4 ½ 3 ½ 4.5 3 2-4 3 4-5 5 - /k/ 4 4 ½ 4 4.5 2 2-4 3 3 3.5 3 3 /g/ 4 4 ½ 4 4.5 2 2-4 3 3 3.5-4 3 3 /l/ 4 6 ½ 6 5.5 3 3-4 4 3 5-7 5 5.5 /d/ 5 4 ½ 4 4.5 2 2-4 3 4 3-3.5 3 3 /t/ 5 4 ½ 6 5.5 2 2-8 3 3 3.5-4 3 4 /s/ 5 7 ½ 4 ½ 5.5 3 3 4 3 7-9 6 6 /r/ 5 7 ½ 4 5.5 3 3 5 3 8 6 - /Ƶ/ 5 4 ½ 4 ½ 5.5 4 3-8 4 4 6-7 5 - /v/ 5 6 ½ 6 5.5 4 4 3 ½ 3.5 5.5 5 5 /z/ 5 7 ½ 7 7.5 4 4 4 3 7-9 6 6 /ࣝ/ 6 6 ½ 7 7.5 6 4 4 3 - - - /し/ 7 ½ 6 5.5 5 4 5 4 6-8 5.5 - /d/ 7 4 4.5 4 4 4 - - - /Ƒ/ 6 ½ 4 ½ 5.5 4 3-8 4 ½ 3 5 5 1.5

Speech sounds

Tasneem (1977)(Kannada)

*75%

Usha (1986) (Tamil) *75%

Padmaja (1988)

(Telugu)*75%

Arun Banik (1988)(Bengali)

*90%

Maya(1990) (Malayalam)

*75%

Prathima(2009) (Kannada) *90%

/m/ 3 3 2.6 2.5 3-3.6 3-3.6 /n/ 3 3 2.6 2.5 3-3.6 3-3.6

/┕/ - - - 2.5 3-3.6 3-3.6 /p/ 3 3 2.6 2.5 3-3.6 3-3.6 /f/ - - 2.9 - 3-3.6 /h/ - - 2.6 3 3-3.6 - /k/ 3 3 2.6 2.7 3-3.6 3-3.6 /b/ 3 3 2.6 2.5 3-3.6 3-3.6 /d/ 3.6 3 2.6 3 3-3.6 3.6-4 /g/ 3 3 2.6 3 3-3.6 3-3.6 /r/ 4.6 - 3.9 4 3.7-4 - /s/ 3 3 3.3 - 3.6-4 3-3.6 /Ƒ/ 5.1 6 3.6 3 5-5.6 3.6-4

/tƑ/ 3.7 3 2.6 3 3-3.6 3-3.6 /t/ - 3 2.6 3 3-3.6 3-3.6 /v/ - 3 2.6 - 3-3.6 3-3.6 /l/ 3 3 2.6 3 3-3.6 3-3.6 /j/ 3 3 2.6 3 3-3.6 3-3.6


55

validate the test of articulation in Kannada (Kannada is one of the major Dravidian languages of India, spoken predominantly in the state of Karnataka. Native speakers are called Kannadigas, number roughly 38 million, making it the 27th most spoken language in the world. It is one of the scheduled languages of India and the official & administrative language of the state of Karnataka.( Kannada (n.d) In Wikipedia Online. Retrieved from http:// www. wikipedia .com). Specifically photographs were familiarized and a photo articulation test of Kannada was developed. This was administered on 240 typically developing children in the age range of 2-6 years. The effect of age and gender on the articulation abilities in children was investigated. Further the test was administered on 10 clients with misarticulation to find the validity.

Method

The study was conducted in two phases. Phase I included development of Kannada diagnostic articulation test and phase II included standardization of the test.

Phase I: Construction of Diagnostic articulation test in Kannada

A list of four hundred and eighty five words having all phonemes in all naturally occurring positions in Kannada (initial and medial) were given to three judges to check the familiarity of words on a three point familiarity rating scale as follows: Very familiar (> 95%); Familiar (90 ≤ 95%); Unfamiliar (75 ≤ 90%).

Out of 485 words, 259 words were rated very

familiar, 219 words were rated familiar, and 7 words as unfamiliar. 259 words which were rated as very familiar, picturable, and unambiguous were photographed and were again given for familiarity rating to extract final 114 picturable words. The final test material (Kannada Diagnostic Photo Articulation Test) comprised of a total of 114 words which were divided into two parts. Part I consisted of a total of 52 words with 10 vowels, 2 diphthongs in initial position, 19 consonants in initial and medial position and /┕/ and /宜/ in medial position. Part II consisted of a total of 62 words with 10 vowels, one diphthong (/ai/) in initial position, 19 consonants in initial and medial position, /┕/ and /宜/ in medial position and eleven clusters in naturally occurring positions. Appendix I shows the words of the Kannada diagnostic photo articulation test.

Phase II: Re-standardization of Kannada diagnostic articulation test Subjects: 240 typically developing children (15 males and 15 females in each age range) in the age range of 2 to 2.6 years, 2.6 to 3 years, 3 to 3.6 years,

3.6 to 4 years, 4 to 4.6 years, 4.6 to 5 years, 5 to 5.6 years, and 5.6 to 6 years participated in the study. They were selected based on the teachers/parental interview and the informal screening for speech, language, hearing, cognitive or any other motor difficulties.

Material: Kannada diagnostic photo articulation test was used for the study. The test was re-standardized by administering it on two hundred and forty pre-school and school going children in the age range of two to six years.

Procedure: The photographs were presented visually one after the other through the use of laptop and the children were instructed to name the photograph. Oral responses were audio recorded using the same laptop.

Data transcription and Scoring: The data obtained from all the 240 children was transcribed using broad and narrow IPA transcription. All the responses of each child were analyzed sound-by-sound on a response sheet. A score of 1 was given for each correct response, 0.75 for distortion, 0.5 for substitution and 0 for omission.

Reliability: Six children (3 boys and 3 girls) in each age group were re-tested within 8-15 days for test-retest reliability.

Validity: Validity of test was checked by administering the test on ten subjects with misarticulation. The responses were audio recorded and later transcribed and compared to the scores of typically developing children without misarticulation.

Data analyses: Phonemes which were produced correctly by 90% of the children in each age group in the initial and medial positions for both part I and part II were identified. The order of acquisition of speech sounds was determined and is given in Appendix II. Statistical analyses: A commercially available SPSS version 16.0 was used for all statistical analyses. The test was scored on the basis of the frequency of correct responses. The data was statistically treated by obtaining the mean and standard deviation. A two-way MANOVA and paired t-test was carried out to find the significant difference in articulatory scores across age groups, across gender within the group, as well as between the sub parts of the test and also to check the interactions among gender and age group.


56

Results

The results indicated main effects of age and gender on the score of the articulation test. The scores increased with increase in age and girls had significantly better scores compared to boys. Also, within gender the main effect of age was observed. Figure 1 shows the scores in boys and girls.

Figure 1. Mean percent score in part I and II in boys and girls.

The mean scores increased from 2 years to 6

years of age in both part I and II. However, children obtained higher score in part I compared to part II. Girls had higher scores compared to boys. Results indicated no significant interaction between age * gender for any part of the test. Figures 2 – 4 shows the raw scores in both genders.

Results of paired t-test showed significant

difference between part I and II in the age groups of 2 - 4 years (with 6 month age interval); in boys at p<0.01 and in the age groups of 2 to 3.6 years in girls (p<0.01). It was also observed that most of the vowels, diphthongs, semivowels, dentals, bilabials were acquired by more than 90% the children in the age range of 2.6 to 3years. However, another salient observation was that the glottal fricative /h/ was not mastered even by 75% of the children by 6 years. Appendix III shows cut off scores of Kannada diagnostic photo articulation test. Test-retest reliability: Cronbach‟s coefficient reliability test showed an g of 0.947 for part I and 0.972 for part II, respectively. The overall correlation coefficient was 0.970. All correlations were high and significant.

Figure 2. Raw scores on part I in boys and girls.

Figure 3. Raw scores on Part II in boys and girls.

Figure 4. Overall scores in boys and girls.


57

Table 3. Comparison between the scores obtained for subjects with misarticulations and the average scores of typically developing children (total, Part I, Part II)

Validity : Scores obtained by the misarticulation cases were much lesser than that of normal children as shown in Table 3. Thus, the test can differentiate normal children from misarticulation cases.

Discussion

The results revealed several interesting points. First of all, results indicated that the scores significantly increased with increase in age in both boys and girls and in both parts. These findings are in agreement with those of Wellman et al., (1931), Poole (1934), Templin (1957), Tasneem (1977), Usha (1986), Padmaja (1988) and Arun Banik (1988). It is evident that because of the increasing maturity of all motor skills, articulation skill increases as age increases.

Second, girls scored significantly higher than boys in both parts of the test. The result is in consonance with that of Templin (1963) who reported that “in articulation development, girls consistently are found to be slightly accelerated; in all instances the differences are relatively small and often are not statistically significant”. In the present study, the difference between boys and girls was significant. The result is also in consonance with those of Usha (1986), Arun Banik (1988), and Maya (1990). However, it is not in consonance with those of Padmaja (1988) and Tasneem Banu (1977), and Prathima (2009).

Third, some sounds were acquired earlier than

the others. It was generally observed that all the vowels and most of the consonants except /r/, /h/, /宜/, /Ƒ/, and /s/ were acquired by the age of 3-3.6 years. The results of the present study were compared with the earlier studies (Wellman et al., 1931; Templin, 1957 in English and Tasneem, 1977; Usha, 1986 in

Tamil; Padmaja, 1988 in Telugu and Prathima, 2009 in Kannada) to observe whether the order of acquisition was similar. The present study revealed that some sounds were acquired earlier than others. At the outset it appears that the age of acquisition of different speech sounds in Indian languages are relatively faster compared to Western studies. However this observation needs to be interpreted with much caution because all the reported studies in the western context that are available are carried out from early thirties to the seventies or so.

Fourth, the results indicated good test-retest

reliability and the test was valid to differentiate typically developing children from atypical children. A definite pattern in the acquisition of articulation was found. All the vowels and diphthongs were found to have been acquired by 90% of the children by the age of 2 years. Most of the consonants were acquired by the age of 2.6 years; except /Ƶ/, / Ʋ/, /居 /, /危 /, /┕/, /r/, /l/, /Ƒ/, /s/, /h/, /宜/ and clusters.

Among boys, most of the consonants were

acquired by 90% of the children by the age of 3.6- 4 years. However, /r/ and /h/ were exceptional, in that /r/ was acquired by 90% of the children in initial position and not in the medial position and /h/ was not acquired even by 75% of the children by the age of 6 years. Among girls, most of the consonants were acquired by the age of 3 years except /Ƶ/, /Ʋ/, /居/, /危/, /┕/, /r/ and /h/ in both the positions tested. Among the clusters only /sk/ was acquired.

Another salient feature observed was that, the

children in the present study seemed to acquire most of the sounds at a younger age compared to the earlier reports in English and Kannada. The early articulatory acquisition in the present study compared to earlier report by Tasneem Banu (1977) may be attributed to the cultural differences or a

Sl. No.

Subjects with misarticulation Age in years

(Gender)

Score obtained (score of typically developing children)

Total raw score

Score obtained (score of typically developing children)

Part I raw score

Score obtained (score of typically developing

children) Part II raw score

1. 4.0 (F) 82.75 (110.45) 41.25 (50.3) 41.50 (58.55) 2. 4.9 (M) 93.50 (112.03) 46.00 (50.65) 47.50 (59.79) 3. 4.9 (M) 82.50 (112.03) 42.25 (50.65) 40.25 (59.79) 4. 6.3 (M) 97.75 (111.53) 45.00 (50.93) 52.75 (60.59) 5. 6.9 (F) 90.75 (111.53) 45.75 (50.93) 45.00 (60.59) 6. 7.3 (F) 99.75 (111.53) 47.50 (50.93) 52.25 (60.59) 7. 9.0 (M) 95.00 (111.53) 43.00 (50.93) 52.00 (60.59) 8. 9.3 (M) 102.75 (111.53) 47.50 (50.93) 55.25 (60.59) 9. 15. (M) 105.50 (111.53) 48.00 (50.93) 57.50 (60.59) 10. 16 (M) 96.25 (111.53) 46.75 (50.93) 49.50 (60.59)


58

change in norms, over years because of greater exposure to speech environment.

Conclusions

It can be concluded that Kannada Photo Articulation test is useful as a diagnostic tool and also to aid in prescribing the nature of speech correction required. It is recommended that a detailed phonological process analyses be done with the data collected. It is time that all articulation tests in Indian languages are re-standardized. Speech samples of children below two years of age can be recorded and analyzed for phoneme acquisition. India is a multilingual country where people speak more than one language. Therefore, the effect of bilingualism and multilingualism on acquisition of phonology is warranted. The present study was restricted to elicitation of phoneme in only one context. Deep test can be developed to investigate the correct articulation of phoneme or otherwise in several phonetic contexts.

Acknowledgements

The authors would like to thank Dr. Vijayalaksmi Basavaraj, Director, AIISH, Mysore for granting permission to take up the study. They thank all the subjects who participated in this study.

References

Arlt, P. B., & Goodban, M. T. (1976). A comparative study of articulation acquisition as based on a study of 240 normals, aged three to six. Language, Speech, and Hearing Services in Schools, 7, 173-180.

Arun Banik (1988). Articulation test in Bengali. Unpublished Master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing. University of Mysore, Mysore.

Babu, R. M., Rathna, N., & Bettagiri, R. (1972). Test of Articulation in Kannada. Journal of All India Institute of Speech and Hearing, 3, 64-79.

Bharathy, R. (2001). Development of phonological processes of 3-4 year old age normal Tamil speaking children. Unpublished Master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing. University of Mysore, Mysore.

Fey, M. E. (1992). Clinical forum: Phonological Assessment and Treatment. Articulation and phonology: Inextricable constructs in speech pathology. Language, Speech, and Hearing Services in Schools, 23, 225-232. (Reprinted from Human Communication Canada, 1985, 9, 7-16).

Fudala, J., B., & Reynolds, W. M. (2000). In A. B. Smit (2004). Articulation and phonology resource guide for school-age children and adults. Thomson Learning, Inc.

Irwin, J. W., & Wong, S. P. (1983). Phonological development in children 18 to 72 months. Journal of Speech and Hearing Disorders, 12, 402-404.

Maya (1990). An articulation test battery in Malayalam. Unpublished Master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing. University of Mysore, Mysore.

Mecham, M. J. (1962). Combined Wellman, Poole and Templin norms. Salt Lake City. Unpublished Paper.

Padmaja, B. (1988). Telugu Articulation Test. Unpublished Master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing. University of Mysore, Mysore.

Poole, E. (1934). Genetic development of articulation of consonant sounds in speech. Elementary English Review, 11, 159-161.

Prather, E., Hedrick, D., & Kern, C. (1975). Articulation development in children aged two to four years. Journal of Speech and Hearing Research, 40, 55-63.

Prathima, S. (2009). Articulatory acquisition in typically developing native Kannada speaking children. Unpublished Master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing. University of Mysore, Mysore.

Rahul (2006). Study of phonological processes in 2-3 year old Hindi speaking children. Unpublished Master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing. University of Mysore, Mysore.

Roberts, J., Burchinal, M., & Foote, M. (1990). Phonological process decline from 2½ to 8 years. Journal of Communication Disorders, 23, 205-217.

Sander, E. (1972). When are speech sounds learned? Journal of Speech and Hearing Disorders, 37, 55-63.

Smit (1990). In M. N. Hegde (2000). Assessment and treatment of articulation and phonological disorders in children. United States of America. Pro-ED Inc.

Sreedevi, N., & Shilpashree, H. N. (2008). Phonological processes in typically developing Kannada speaking children. Journal of All India Institute of Speech and Hearing, 27, 83 – 88.

Tasneem, B. (1977). Articulatory acquisition in Kannada A study of normal children 3-6.6 years. Unpublished Master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing. University of Mysore, Mysore.

Templin, M. C. (1957). Spontaneous versus imitated verbalization in testing articulation in preschool children. Journal of Speech and Hearing Disorders, 12, 293-300.

Templin, M. (1963). Development of speech. Journal of Pediatrics, 62, 11 – 14.

Usha, D. (1986). Tamil Articulation Test. Unpublished Master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing. University of Mysore, Mysore.

Van Riper, C., & Irwin, J. V. (1958). Voice and articulation. Englewood Cliffs, N.J.:Prentice Hall.

Wellman, B., Case, I., Mengert, I., & Bradbury, D. (1931). Speech sounds of young children. University of Iowa Studies in Child Welfare, 5 (2).

Wikipedia Online. Retrieved from http://www.wikipedia.com.

http://www.wikipedia.com/

59

Appendix I

Kannada Diagnostic Photo Articulation Test: Part I

Sl. No.

Phoneme Tested

Position (Initial [I] or Medial [M])

Stimulus in Kannada

IPA

1. a I CfÓ /a限限I/ 2. a: I D£É /a:ne/ 3. i I E° /ili/ 4. i: I FgÀÄ½î /i:rU称称傑/ 5. u I GAUÀÄgÀ /użgura/ 6. u: I Hl /u:職a/ 7. e I J¯É /ele/ 8. e: I K¼ÀÄ /e:称u/ 9. ai I LzÀÄ /aid作 u/ 10. o I MAmÉ /on職e/ 11. o: I N¯É /o:le/ 12. au I OµÀzsÀ /au検ad作 a/ 13. k I PÀvÀÛj /ka触触ari/ 14. k M ¨ÉPÀÄÌ /b計kku/ 15. g I UÀrAiÀiÁgÀ /ga召傑ja:ra/ 16. g M ªÀÄÆUÀÄ /mu:gu/ 17. 古 I ZÀ¥Áw /古apa:触傑/ 18. 古 M ¨ÁZÀtÂUÉ /ba:古aż傑ge/ 19. 限 I fAPÉ /限傑nke/ 20. 限 M ¥ÀÇeÁj /pu:限a:r傑/ 21. 職 I mÉÆÃ¦ /職o:p傑/ 22. 職 M amÉÖ /古i職職計/ 23. 召 I qÁPÀÖgï /召a:k職ar/ 24. 召 M CAUÀr /ażga召i/ 25. ż M PÀtÄÚ /każżu/ 26. 触 I vÀmÉÖ /触a職職e/ 27. 触 M PÉÆÃw /ko:触i/ 28. 商 I zÁgÀ /商a:ra/ 29. 商 M PÀÄzÀÄgÉ /ku商ure/ 30. n I £À°è /nalli/ 31. n M zÉÃªÀ¸ÁÜ£À /devas触a:na/ 32. p I ¥ÀÇj /pu:ri/ 33. p M PÀ¥Éà /kappe/ 34. b I ¨ÁV®Ä /ba:gilu/ 35. b M PÀ§Äâ /kabbu/ 36. m I ªÀÄ£É /mane/ 37. m M JªÉÄä /jamme/ 38. j I AiÀÄPÀëUÁ£À /jak検aga:na/ 39. j M vÉAV£ÀPÁ¬Ä /t計ngina�賠i/ 40. r I gÉÊ®Ä /ra傑lu/ 41. r M ªÀÄgÀ /mara/ 42. l I ¯ÉÆÃl /lo:職a/ 43. l M ºÀ®Äè /hallu/ 44. v I «ªÀiÁ£À /v傑ma:na/ 45. v M Q« /kivi/

60

Kannada Diagnostic Photo Articulation Test: Part II

Sl. No.

Phoneme Tested


Stimulus in Kannada

IPA

1. a I CrUÉªÀÄ£É /a召igemane/ 2. a: I D¸ÀàvÉæ /a:spa触re/ 3. i I EgÀÄªÉ /Iruve/ 4. i: I F½UÉªÀÄuÉ /i:称igemaże/ 5. u I GAiÀiÁå¯É /ujja:le/ 6. u: I HgÀÄ /u:ru/ 7. e I JgÀqÀÄ /era召u/ 8. e: I KtÂ /e:żI/ 9. ai I LªÀvÀÄÛ /aiva触触u/ 10. o I MAzÀÄ /o:n商u/ 11. o: I NqÀÄ /o:召u/ 12. k I PÁgÀÄ /ka:ru/ 13. k M ¸ÉÊPÀ®Äè /saIkallu/ 14. g I UÁ½¥Àl /ga:称pa職a/ 15. g M PÁUÉ /ka:ge/ 16. 古 I ZÀPÀÄÌ° /古akkulI/ 17. 古 M ªÀÄAZÀ /man古a/ 18. 限 I dqÉ /限a召e/ 19. 限 M ¸ÀÆf /su:限I/ 20. 職 I mÉÆªÉÆÃmÉÆÃ /職omo:職o/ 21. 職 M QlQ /ki職覗kI/ 22. 召 I qÀ©â /召abbI/ 23. 召 M PÀ£ÀßqÀPÀ /kanna召禽ka/ 24. ż M VtÂ /giżi/ 25. 触 I vÀ§® /触abala/ 26. 触 M PÀvÉÛ /ka触触e/ 27. 商 I zÉÆÃ¸É /商o:se/ 28. 商 M UÉÆÃ¢ /go:商I/ 29. n I £Á¬Ä /na:ji/ 30. n M «ÄÃ£ÀÄ /mi:nu/ 31. p I ¥É£ÀÄß /pennu/ 32. p M ZÀ¥Àà° /古app賠li/ 33. b I §¸ÀÄì /bassu/ 34. b M ¢A§Ä /dImbu/ 35. m I ªÀÄÆgÀÄ /mu:ru/ 36. m M DªÉÄ /a:me/ 37. j I AiÀÄAvÀæ /jantra/ 38. j M PÀqÀ¯ÉPÁ¬Ä /�禽召禽le�禽I/

46. 検 I ±ÀlÄð /検ar職u/ 47. 検 M §æµï /bra検/ 48. s I ¸ÉÃ§Ä /se:bu/ 49. s M «ÄÃ¸É /mi:se/ 50. h I ºÁªÀÅ /ha:vu/ 51. h M ¹AºÀ /simha/ 52. 称 M §¼É /ba称e/

61

39. r I gÀAUÉÆÃ° /rango:lI/ 40. r M ¸ÀgÀ /s禽ra/ 41. l I ®AUÀ /langa/ 42. l M UÉÆÃ° /go:lI/ 43. v I «ÃuÉ /vi:że/ 44. v M ºÀÆªÀÅ /hu:vu/ 45. 検 I ±ÀAR /検禽n�禽/ 46. 検 M UÀuÉÃ±À /g禽że:検a/ 47. s I ¹ÃgÉ /si:r計/ 48. s M ºÀ¸ÀÄ /hasu/ 49. h I ºÀÄ° /hulI/ 50. h M ¨Á¼ÉºÀtÄÚ /ba:称ehażżu/ 51. 称 M PÉÆÃ½ /ko:称I/ 52. st I ¸ÁÖA¥ÀÅ /s職a:mpu/ 53. st M ¥ÉÇÃ¸ïÖ¨ÁPïì /po:s職ba:ks/ 54. sku I ¸ÀÆÌlgï /sku:職禽r/ 55. ske M ©¸ÉÌlÄÖ /bIske職職u/ 56. dra I zÁæQë /商rak検I/ 57. dra M ZÀAzÀæ /古禽n商ra/ 58. r古i M PÀÄað /kur古I/ 59. kra M ZÀPÀæ /古akra/ 60. ble I ¨ÉèÃqÀÄ /ble:召u/ 61. k検a M DmÉÆÃjPÀë /a:職orik検a/ 62. skru ¸ÀÆÌç /skru/

62

Appendix II

Kannada Diagnostic Photo Articulation Test

Ordered (according to age)

Part I

Sl. No. Age range in years

Phoneme Tested


Stimulus in Kannada

IPA

1. 2-2.6 a I CfÓ /a限限I/ 2. a: I D£É /a:ne/ 3. i I E° /ili/ 4. i: I FgÀÄ½î /i:rU称称i/ 5. u I GAUÀÄgÀ /użgura/ 6. u: I Hl /u:職a/ 7. e I J¯É /ele/ 8. e: I K¼ÀÄ /e:称u/ 9. ai I LzÀÄ /aid作 u/ 10. o I MAmÉ /on職e/ 11. o: I N¯É /o:le/ 12. au I OµÀzsÀ /au検ad作 a/ 13. k I PÀvÀÛj /ka触触ari/ 14. k M ¨ÉPÀÄÌ /bekku/ 15. g I UÀrAiÀiÁgÀ /ga召Ija:ra/ 16. g M ªÀÄÆUÀÄ /mu:gu/ 17. 触 I vÀmÉÖ /触a職職e/ 18. 触 M PÉÆÃw /ko:触i/ 19. 商 I zÁgÀ /商a:ra/ 20 商 M PÀÄzÀÄgÉ /ku商ure/ 21. n I £À°è /nalli/ 22. n M zÉÃªÀ¸ÁÜ£À /devas触a:na/ 23. p I ¥ÀÇj /pu:ri/ 24. p M PÀ¥Éà /kappe/ 25. b I ¨ÁV®Ä /ba:gilu/ 26. b M PÀ§Äâ /kabbu/ 27 m I ªÀÄ£É /mane/ 28. m M JªÉÄä /jamme/ 29. j I AiÀÄPÀëUÁ£À /jak検aga:na/ 30. j M vÉAV£ÀPÁ¬Ä /t計ngina�賠i/ 31 v I «ªÀiÁ£À /v傑ma:na/ 32. v M Q« /kivi/ 33. 2.6 - 3 l I ¯ÉÆÃl /lo:職a/ 34. l M ºÀ®Äè /hallu/ 35. 3 – 3.6 限 I fAPÉ /限傑nke/ 36. 限 M ¥ÀÇeÁj /pu:限a:r傑/ 37. 召 I qÁPÀÖgï /召a:k職ar/ 38. 召 M CAUÀr /ażga召i/ 39. 3.6 - 4 古 I ZÀ¥Áw /古apa:触i/ 40. 古 M ¨ÁZÀtÂUÉ /ba:古aż傑ge/ 41. 職 I mÉÆÃ¦ /職o:p傑/ 42. 職 M amÉÖ /古i職職計/ 43. ż M PÀtÄÚ /każżu/

63

44. r I gÉÊ®Ä /ra傑lu/ 45. r M ªÀÄgÀ /mara/ 46. 称 M §¼É /ba称e/ 47. 検 I ±ÀlÄð /検ar職u/ 48. 検 M §æµï /bra検/ 49. s I ¸ÉÃ§Ä /se:bu/ 50. s M «ÄÃ¸É /mi:se/ 51. 4 - 4.6 NIL 52. 4.6 - 5 NIL 53. 5 – 5.6 NIL 54. 5.6 - 6 NIL

Part II

Sl. No.

Age range in years

Phoneme Tested


Stimulus in Kannada

IPA

1. 2-2.6 a I CrUÉªÀÄ£É /a召igemane/ 2. a: I D¸ÀàvÉæ /a:spa触re/ 3. i I EgÀÄªÉ /Iruve/ 4. i: I F½UÉªÀÄuÉ /i:称igemaże/ 5. u I GAiÀiÁå¯É /ujja:le/ 6. u: I HgÀÄ /u:ru/ 7. e I JgÀqÀÄ /era召u/ 8. e: I KtÂ /e:żI/ 9. ai I LªÀvÀÄÛ /aiva触触u/ 10. o I MAzÀÄ /o:n商u/ 11. o: I NqÀÄ /o:召u/ 12. k I PÁgÀÄ /ka:ru/ 13. k M ¸ÉÊPÀ®Äè /saIkallu/ 14. g I UÁ½¥Àl /ga:称pa職a/ 15. g M PÁUÉ /ka:ge/ 16. 触 I vÀ§® /触abala/ 17. 触 M PÀvÉÛ /ka触触e/ 18. 商 I zÉÆÃ¸É /商o:se/ 19. 商 M UÉÆÃ¢ /go:商I/ 20 n I £Á¬Ä /na:ji/ 21. n M «ÄÃ£ÀÄ /mi:nu/ 22. p I ¥É£ÀÄß /pennu/ 23. p M ZÀ¥Àà° /古app禽li/ 24. b I §¸ÀÄì /b禽ssu/ 25. b M ¢A§Ä /dImbu/ 26. m I ªÀÄÆgÀÄ /mu:ru/ 27 m M DªÉÄ /a:me/ 28. j I AiÀÄAvÀæ /jantra/ 29. j M PÀqÀ¯ÉPÁ¬Ä /�禽召禽le�禽I/ 30. v I «ÃuÉ /vi:że/ 31 v M ºÀÆªÀÅ /hu:vu/ 32. 2.6 – 3 l I ®AUÀ /langa/ 33. l M UÉÆÃ° /go:lI/ 34. 3 – 3.6 限 I dqÉ /限a召e/ 35. 限 M ¸ÀÆf /su:限I/ 36. 召 I qÀ©â /召禽bbI/

64

37. 召 M PÀ£ÀßqÀPÀ /kanna召禽ka/ 38. 3.6 - 4 古 I ZÀPÀÄÌ° /古akkulI/ 39. 古 M ªÀÄAZÀ /man古a/ 40. 職 I mÉÆªÉÆÃmÉÆÃ /職omo:職o/ 41. 職 M QlQ /ki職覗kI/ 42. ż M VtÂ /giżi/ 43. r I gÀAUÉÆÃ° /rango:l傑/ 44. r M ¸ÀgÀ /s禽ra/ 45. 称 M PÉÆÃ½ /ko:称I/ 46. 検 I ±ÀAR /検禽n�禽/ 47. 検 M UÀuÉÃ±À /g禽że:検a/ 48. s I ¹ÃgÉ /si:r計/ 49. s M ºÀ¸ÀÄ /hasu/ 50. 4 – 4.6 sku I ¸ÀÆÌlgï /sku:職禽r/ 51. kra M ZÀPÀæ /古akra/ 52. ble I ¨ÉèÃqÀÄ /ble:召u/ 53. 4.6 - 5 st I ¸ÁÖA¥ÀÅ /s職a:mpu/ 54. st M ¥ÉÇÃ¸ïÖ ¨ÁPïì /po:s職ba:ks/ 55. ske M ©¸ÉÌlÄÖ /bIske職職u/ 56. k検a M DmÉÆÃjPÁë /a:職o:rik検a/ 57. skru ¸ÀÆÌç /skru/ 58. 5 – 5.6 dra I zÁæQë /商rak検I/ 59. dra M ZÀAzÀæ /古禽n商ra/ 60. 5.6 - 6 NIL

65

Appendix III

Cut off scores of the Kannada Diagnostic Photo Articulation test

Age group

in years

Part I Part II

Mean % Raw

score

Mean % Raw

score

2 – 2.6 87.51 45.50 74.15 45.97

2.6 – 3 90.89 47.26 81.65 50.62

3 - 3.6 93.17 48.45 84.71 52.52

3.6 – 4 95.11 49.45 89.79 55.67

4 – 4.6 96.73 50.3 94.44 58.55

4.6 – 5 97.41 50.65 96.43 59.79

5 – 5.6 98.38 51.15 98.18 60.87

5.6 – 6 97.94 50.93 97.72 60.59


66

S

Art iculatory A cquisition in Typically Developing Malayalam Speaking Children: 2-3 Years

Divya P. 1 & N. Sreedevi 2

Abstract

The purpose of the present study was to obtain norms for the Malayalam Articulation Test (MAT; Maya, 1990) in the age range of 2-3 years and to determine those phonemes which are acquired by 75% and 90% of the children. MAT was administered to 120 (60 males and 60 females) children in the age range of 2 - 3 years subdivided into four age groups (2 - 2.3 years, 2.3 - 2.6 years, 2.6 - 2.9 years & 2.9 - 3 years) and the responses were transcribed using IPA. The statistical analysis of the data indicated that there was a significant difference across the age but not across gender. As age increased, the scores also increased indicating improved articulatory abilities due to neuromuscular maturation. All the vowels tested were mastered by 2.6 years of age. The phonemes as /┕, /n/, /b/, /m/, /p/, /-nt/, /-拒/, /-nth/, /-cj/, /-nd/ and /Ƶ/ reached 90% criteria in all the three positions by 2.3 years itself /k/, /Ʋ/,

palatal nasal/ƀ/, /g/, /v/, /-ty/, /宜/, /t/, /危/, /t/, /d/, /j/, /Ƶȯ/, /-nk/, /tȯ/ and /t/ reached 90% criteria by 3 years. None of the clusters reached even the 75% criteria by this age. Bilabials, labiodentals, dentals and velars were acquired earlier than alveolar, palatal, retroflex and glottal sounds. Unaspirated sounds were acquired earlier compared to aspirated sounds. Earlier acquisition of sounds in the present study is attributed to differences in lifestyle and to greater exposure to speech and language environment.

Key words: arti culatory acquisition, malayalam, initial/ medial/ final position

peech sound development refers primarily to gradual mastery of speech sound form within in a given language. The order of acquisition of

sounds and age norms pertains to the area of phonology. Many investigators have studied the acquisition of phonology describing the patterns that takes place in children at dif ferent age levels. The availabili ty of normative data is essential to clinical assessment in child phonology. The data should include different groups of children acquiring the same target language because the norms are sensitive to sociolinguistic factors such as gender, socioeconomic status and language backgrounds.

The acquisition of phonemes in the speech of

young English-speaking children has received considerable attention by speech language pathologists since 1930‟s (Wellman, Case, Mengert, & Bradbury, 1931; Poole, 1934; Templin, 1957; Sander, 1972; Prather, Hedrick & Kern, 1975; Arlt & Goodban, 1976; Fudala & Reynolds, 1986; Dyson, 1988; Mowrer & Burger, 1991; Robb & Bleile 1994).

Smit, Hand, Freil inger, Bernthal and Bird (1990)

provided some normative data on the acquisition of speech sounds in children residing in Iowa and Nebarska. The study included 1049 children in the age range of 3 - 9 years of age. They considered 90% level

1e-mail:[email protected]; 2Lecturer in Speech Sciences, AIISH, Mysore, [email protected].

of acquisition for the mastery of sounds. The results indicated that girls appeared to acquire sounds earlier than boys, although this effect reached statistical significance only at age of 6 years and below. The consonants /m/, /n/, /h/, /p/, /f/, /w/ and /b/ were

mastered by 3 years of age, the phonemes /l/, /tȒ/, /k/, /g/, /j/, /d/ and /t/ were acquired by 7 years of age and the phonemes /┕/ and /s/ were acquired by 7 to 9 years.

Mowrer and Burger (1991) stated that glides /j/ and /w/ are generally mastered before 3 years of age. Bauman-Waengler (1994) reported that the early sounds that are developing are nasals, stops, glides and liquids. Later sounds that are developing are fricatives and affricates.

Robb and Bleile (1994) studied the speech sample over a 12 month period of seven children, aged 8-14 months at the beginning of the study and 19 - 26 months at the end of the study. The findings showed that the number of consonants in their inventories increased over time, the number of consonants used in initial position was greater than in the final positions, stops and nasals emerged earlier than fricatives and bilabial, alveolar, glottal place of articulation predominated and produced were earlier than velars.

The literature on vowel development suggests that the acquisition is earlier than consonants (Templin, 1957). Fudala and Reynolds (1986) in their normative

mailto:e-mail:[email protected]

mailto:[email protected]

Articulatory acquisition in Malayalam speaking children

67

data reported that by the age of 1.6 to 1.11 years,

vowels and diphthongs such as /殴/, /】/, /i/, /æ/, /ǣ/, /ș/, /a/, /I/, /i/, /u/, /ou/, /ai/, /ei/, and /au/ reached 90%

criterion. However the mid central vowel /Ǭ/ and /ǩ/ did not reach the mastery criterion of 90% until the 5.6 to 5.11years of age. The rhotic diphthong also reached the mastery criterion at the 5.6 to 5.11 years with the exception of /ir/, which reached 90% acquisition level at the 4.6 to 4.11 years.

Extensive studies on articulatory acquisition

have been carried out in the Indian context also (Kumudavalli, 1973; Tasneem, 1977 and Prathima & Sreedevi, 2009 in Kannada; Usha, 1986 in Tamil; Padmaja, 1988 in Telugu; Arun Banik, 1988 in Bengali and Maya, 1990 in Malayalam). They concluded that the acquisition followed same pattern as in English but generally it was found that most of the sounds were acquired earlier in the Indian studies compared to the western context.

Padmaja (1988) studied the acquisition of

sounds in 160 Telugu speaking children in the age range of 2½ - 4½ years and found that all vowels and most of the consonants except /r/, /s/, /∫/, /居/, /危/ and aspirated stops are acquired by 2½ years. /s/, /r/ and aspirated consonants are acquired by 3.3 years and /居/, /∫/ and some clusters by 3½ years.

Maya (1990) studied 240 Malayalam speaking

children in the age range of 3-7 years. She reported that they acquired /s/, /r/, /l/, /f/, /tȒ/, /j/ and unaspirated stops at an earlier age of 3-3.6 years, while aspirated stops were acquired as late as 6-6.6 years. The criterion of acquisition considered was 75%.

Prathima and Sreedevi (2009) studied 120

Kannada speaking children in the age range of 3-4 years and reported that all the vowels, diphthongs and most of the consonants were mastered by 90% of the children by 3-3.6 years except /r/ and /h/ in initial and medial positions. By 4 years of age, /r/ was acquired in medial but not in initial positions by 90% of children and /h/ was not acquired even by 75% of the children. Among the clusters only /ski/ met the 90% criteria by 4 years.

More recently it has been the observation of

speech-language pathologists that the children of this generation are acquiring speech sounds at an earlier age than their earlier counter parts. In view of this fact, the previous norms were obtained 20 years back in Malayalam by Maya (1990). Hence there is an immediate need to test much younger children for

determining the exact age of acquisition and to update the previously obtained norms for both clinical and research purposes. Thus this study aimed at obtaining norms for the mastery of articulatory skills in typically developing Malayalam speaking children in the age range of 2-3 years.

The specific objectives of the study were (a) to administer the Malayalam Articulation Test (Maya, 1990) and to establi sh the ages at which 75% and 90% of the children produce the phonemes of Malayalam correctly (b) to compare the articulatory skills across age and gender (c) to compare the difference in the articulatory acquisition of phonemes in initial, medial and final positions of the words (d) to compare the data obtained with that of the earlier reported studies in both English and other Indian languages.

Method

Subjects: 120 typically developing Malayalam speaking children in the range of 2 - 3 years were selected randomly from different localit ies of Wayanad district in Kerala. The children were subdivided into four groups with an age interval of three months (2 - 2.3 years, 2.3 - 2.6 years, 2.6 - 2.9 years & 2.9 - 3 years). Each group comprised of a total of 30 children including 15 girls and 15 boys. The subjects reared in an ambient environment of Malayalam and belonging to middle socio economic status was selected. The subjects were exposed to some amount of English language also in the kindergarten set up. The children who did not have any speech, language, hearing, cognitive or any other motor diff iculties were considered. They were selected based on the parental interview and the informal screening for speech, language, hearing, cognitive or any other developmental dif ficulties. Test material: The diagnostic test of articulation “The articulation test battery in Malayalam” by Maya (1990) was used as the test material. The Articulation Test consists of 82 target words for testing 10 vowels, 38 consonants and eight consonant clusters. The vowel sounds were tested only in the initial position. Most of the consonants were tested in the initial and medial positions, two consonants in medial and final positions

(/宜/ and /吃/) and three consonants (/r/, /R/ and /m/) were tested in all the positions. The consonant clusters were tested in initial and medial positions (two clusters in initial and six in medial positions). The material consisted of 82 colored photographs of the real objects which were highly representative of the target stimuli. Each target picture was designed to elicit the target sound as a single phoneme or cluster in each position.


68

Procedure: Each subject was tested individually in a quiet room. Once the rapport was established, the examiner presented the target pictures one after another. The target picture was presented one at a time to the subjects on a laptop computer (Compaq - C 770 TU) screen. Subjects were encouraged to name the pictures spontaneously. If any of the subjects failed to identify a target picture, additional cues were presented by the examiner. In spite of additional cues, if the child failed to name the target picture, the child was asked to repeat the target word after the examiner. The response elicited was audio recorded using Cenix digital recorder VR – P - 217 at high quality mode with an external mini microphone placed approximately 10 inches away from the subject‟s mouth.

Scorin g: The data obtained from all the 120 subjects were transcribed using broad and narrow IPA transcription. All the responses of each subject were analyzed sound-by-sound on a response sheet. Correct responses (CR), substitutions (S), omissions (O), distortions (D), additions (A) or any other type of articulatory deviation (Ao) were recorded on the response sheet. A score of‟1’ was given to each correct response; a score of „¾’ for distortion error; a score of „½’ for substitution error and score of‟0’ for omission error were assigned. The maximum score was 82, i.e. when all the test phonemes were correctly produced. Finally the total score for each subject was calculated. The score sheet is provided in the Appendix.

Inter-judge Reliability: To examine inter-judge reliabili ty, 10% of the total samples were selected randomly from each of the four groups and it was transcribed and analyzed by two experienced speech- language pathologists who were native speakers of Malayalam. The transcribed samples of the two judges were compared and the mean percentage of consonant agreement was calculated.

Data analysis: From the scores obtained for the articulatory production, the mean, standard deviation and range of scores were calculated for each age group and separately for boys and girls in initial, medial positions and final positions. The data obtained were subjected to suitable statistical analysis. Phonemes which were produced correctly by 75% and 90% of the subjects in each age group in the initial, medial and final positions were separately identified. An analysis of the phonemes substituted or distorted in production is discussed in detail.


Two-way ANOVA was carried out to analyze the significant difference in articulatory scores between the age groups as well as across gender. Their results indicated that in general there was no significant difference between gender but there was a significant difference across age groups. The results are presented and discussed under the following sections.

I Age vs. Ar ticulatory acquisition: Two-way ANOVA was carried out to find the significant difference between different age groups and the result indicated that the articulatory development was steady from 2 to 2.6 years of age. There seemed to be a sudden increase in the development of articulatory skills in the period of 2.6 to 2.9 years following which there was no significant change in the developmental scores till 3 years of age. All t h e vowels a n d some of the consonants were acquired by 75 % of the children by 2.3 years of age itself. The exceptional consonants were /g/, /Ʋ/, /ƀ/, /吃/, /l/, /宜/,

/v/, /pȯ/, /s/, /球/, /r/, /Ƒ/,/R/, /Ƶȯ/, /dȯ/, /欺/, /Yȯ/, /h/ and

/k ȯ /. Table 1 and Figure 1 depict the articulatory scores across the age groups over the years.

Table 1. Mean articulation scores & standard deviation in different age groups

Gender Groups Age in years N Mean (Std.

Deviation)

Boys

Group I 2-2.3 15 63.07 (1.87) Group II 2.3-2.6 15 64.40 (2.74) Group II I 2.6-2.9 15 68.55 (2.77) Group IV 2.9-3 15 69.40 (5.21) Mean Total Scores 60 66.35(4.26)

Girls

Group I 2-2.3 15 63.30 (2.23) Group II 2.3-2.6 15 64.42 (3.19) Group II I 2.6-2.9 15 68.72 (3.18)

Group IV 2.9-3 15 69.55 (1.94) Mean total score 60 66.49(3.77)

Combined Scores

Group I 2-2.3 30 63.18 (2.02) Group II 2.3-2.6 30 64.40 (2.92) Group II I 2.6-2.9 30 68.63 (2.93) Group IV 2.9-3 30 69.47(3.86)

The findings of the earlier Western classical

studies (Wellman et al., 1931; Poole, 1934; Templin, 1957, Mecham, 1962; Arlt & Goodban, 1976) and some of the Indian studies (Tasneem, 1977; Usha, 1986; Padmaja, 1988; Arun Banik, 1988; Maya, 1990 and Prathima & Sreedevi, 2009) indicated that


70

Table 2. Comparison of the result of the present study with western studies

„*‟ indicates criteria for the sound to be considered as acquired. „-‟ indicates consonants not acquired, empty space indicate speech sound not tested

However during data collection it was observed that 60% of the children were substituting one of the consonants of the clusters by 2.9 years. Therefore it can be inferred that by 2.9 years children begin to produce clusters but they have substitution errors. The clusters with substitution errors seen in this study were /Ƒta/ for /sta/ and /Ƒka/ for /ska/, that is, a palatal fricative was used for a dental fricative. This finding is convincing because by 2.9-3 years, palatal /Ƒ/ was achieved by 73.33% of the children whereas dental /s/ was achieved by only 53.33% of the subjects in the present study. Hence it can be inferred that in the process of acquisition of clusters, the early achieved palatal was used as a substitution. The results of the present study agree with most of the western studies (Templin, 1957 and Smit et al., 1990) and Indian studies (Tasneem, 1977; Arun Banik, 1988; Maya, 1990 and Prathima & Sreedevi, 2009) which say that the clusters were acquired later compared to consonants and vowels. Maya (1990) reported that the consonant clusters were emerging o n l y at t h e a g e o f 4 . 7 years and the acquisition continued till the age of seven years.

In contrast, the present finding was that clusters emerged by 2.9-3 years itself however with substitution errors. Acquisition based on place, voicing and manner features Place feature: When considering 90% criteria, the bilabials (/p/, /b/, /m/), dentals (/t/, /d/, /n/), labiodentals (/v/), and velars (/k/, /g/, /┕/) except their aspirated counterparts were acquired by 2.6 years. Alveolars (/t/, /s/, /l/, /r/, /R/), retroflex (/危, /吃, /宜/, /球/) and palatals (/Ƶ/,/Ʋ/, /j/, /Ƒ/, /欺/, /ƀ/), were the late acquiring phonemes. Among alveolars, /t/ was acquired earlier (2.6 years), while other phonemes were not mastered by 90% of the children by 3 years. Palatal /Ƶ/ and /j/ (2.3 years) were acquired earlier followed by /Ʋ/ (3years), whereas /欺/ and /Ƒ/ did not reach 90% criteria by 3 years. Retroflex /宜/, /危/ and /吃/ (2.9 years) were achieved earlier than retroflex /球/ and none of these retroflex sounds reached 90% criteria even at the age of 3. Glottal /h/ was not acquired by 3 years even by 75% children.

Speech sounds

Well man et

al., 1931) 75%*

Poole (1934) 100%*

Templin (1957) 75%*

Mecham (1962)

Sander (1972) 75%*

Prather (1975) 75%*

Arlt (1976)

Irwin & Wong (1983)

Smith et al., (1990) 75%*

Fudala & Reynolds

(2000), 90% *

I F

Present study (2010)

75 % 90%

/m/ 3 3 ½ 3 3.5 < 2 2 3 1.5 3 2 2 2-2.3 2-2.3 /n/ 3 4 ½ 3 3.5 2 2 3 2 3 2 2.5 2-2.3 2-2.3 /h/ 3 3 ½ 3 3.5 < 2 2 3 2 3 2 - - - /p/ 4 3 ½ 3 3.5 < 2 2 3 3 3 2 3 2-2.3 2-2.3 /f/ 3 5 ½ 3 4.5 3 2-4 3 3 3 3 3 - - /b/ 3 3 ½ 4 3.5 < 2 2-8 3 1.5 3 2 3 2-2.3 2-2.3 /┕/ -- 4 ½ 3 3.5 2 2-8 3 3 7-9 - 4 2-2.3 2-2.3 /j/ 4 4 ½ 3 ½ 4.5 3 2-4 3 4-5 5 - 2-2.3 2-2.3 /w/ 3 3 ½ 3 - 2 2-8 3 1.5 /k/ 4 4 ½ 4 4.5 2 2-4 3 3 3.5 3 3 2-2.3 2.3-2.6 /g/ 4 4 ½ 4 4.5 2 2-4 3 3 3.5-4 3 3 2.3-2.6 2.6-2.9 /l/ 4 6 ½ 6 5.5 3 3-4 4 3 5-7 5 5.5 - - /d/ 5 4 ½ 4 4.5 2 2-4 3 4 3-3.5 3 3 2-2.3 2.3-2.6 /t/ 5 4 ½ 6 5.5 2 2-8 3 3 3.5-4 3 4 2-2.3 2.3-2.6 /s/ 5 7 ½ 4 ½ 5.5 3 3 4 3 7-9 6 6 - - /r/ 5 7 ½ 4 5.5 3 3 5 3 8 6 - 2.6-2.9 - /Ƶ/ 5 4 ½ 4 ½ 5.5 4 3-8 4 4 6-7 5 - 2-2.3 2-2.3

/v/ 5 6 ½ 6 5.5 4 4 3 ½ 3.5 5.5 5 5 2.3-2.6 2.6-2.9 /z/ 5 7 ½ 7 7.5 4 4 4 3 7-9 6 6 - - /識/ 6 6 ½ 7 7.5 6 4 4 3 - - - - - /し/ 7 ½ 6 5.5 5 4 5 4 6-8 5.5 - 2-2.3 2-2.3 /Ƒ/ 6 ½ 4 ½ 5.5 4 3-8 4 ½ 3 5 5 1.5 - -


71

When compared with Maya‟s (1990), it was observed that all the phonemes in the present study were acquired earlier. The order of acquisition was similar for all phonemes, except phoneme /h/. She reported the early acquisition of bilabials, dentals, labiodentals and glottal (3-3.6 years) followed by alveolar, palatal and retroflex phonemes. In the present study the same order of acquisition was observed, however glottal /h/ was found to be acquired at a much later stage than the other phonemes.

Voicing feature: In the present study, the unvoiced stop /k/ and affricate /Ƶ/ were acquired by 2.6 years considering the 90% criteria, while voiced stop /g/ (2.9 years) and affricate /Ʋ/ (3 years), were acquired later. Dyson (1988) supports this finding by stating that in word final inventories voiceless stops were always present but voiced stops appeared to be emerging. Prather et al., (1975) reported early acquisition of voiceless stop /p/ (2 years) compared to voiced stop /b/ (2-8 years). Smit et al., (1990) also supports the earlier acquisition of voiceless /k/ compared to voiced /g/.

Manner of articulation: Generally when considering the manner of acquisition it was observed that nasals, unaspirated stops, frictionless continuant were acquired

first compared to laterals, fricatives, affricates, flap and trill. Plosives: Some of the plosives were acquired by the age of 2.3 years itself, where as few plosives did not reach 90% criteria even by 3 years. The sounds /p/, /b/ and /居ȸ/ were the first acquired plosives by the age of

2.3 years, followed by /t/, /d/, /拒/, /危/, / tȸ/ and /k/ by 2.6

years, /g/ by 2.9 years and /kȸ/ by 3 years of age.

Aspirated /pȸ/ did not meet 75% of criteria even by 3 years. The present findings indicated an early acquisition of unaspirated phonemes (2.9 years) compared to aspirated phonemes. All the unaspirated plosives reached the 90% criteria by 2.9 years itself compared t o 3.6 years reported by Maya (1990).

According to her, the aspirated retroflex /居ȸ/ and dental

/tȸ/ were acquired by 6.6 years, and / kȸ/ by 7 years. However she considered children above the age of 3 years. Nasals: All the nasals achieved 90% criteria by 2 - 2.3 years itself. When comparing the present results with Western (Irwin & Wong, 1983; Fudala & Reynolds, 2000) and Indian studies (Padmaja, 1988 and Arun Banik, 1988) similar pattern of acquisition was seen, in which the nasals were acquired much earlier than other sounds. When compared with the results obtained by

Table 3. Comparison of the result of the present study with other Indian studies

„*‟ indicates criteria for the sound to be considered as acquired. „-‟ indicates consonants not acquired, empty space indicate speech sounds not tested

Speech sounds

Tasneem, 1977 (Kannada)

75%*

Usha, 1986

(Tamil) 75%*

Padmaja,1988 (Telugu)

75%*

Arun Banik, 1988 (Bengali)

90%*

Maya, 1990 (Malayalam)

75%*

Prathima & Sreedevi,

2009 (Kannada) 90%*

Present study (Malayalam)

2010 75% * 90 %*

M 3 3 2.6 2.5 3-3.6 3-3.6 2-2.3 2-2.3 N 3 3 2.6 2.5 3-3.6 3-3.6 2-2.3 2-2.3 └ 2.5 3-3.6 2-2.3 2-2.3 P 3 3 2.6 2.5 3-3.6 3-3.6 2-2.3 2-2.3 F 2.9 3-3.6 - - H 2.6 3 3-3.6 - - - K 3 3 2.6 2.7 3-3.6 3-3.6 2-2.3 2.3-2.6 B 3 3 2.6 2.5 3-3.6 3-3.6 2-2.3 2-2.3 D 3.6 3 2.6 3 3-3.6 3-3.6 2-2.3 2.3-2.3 G 3 3 2.6 3 3-3.6 3-3.6 2.3-2.6 2.6-2.9 R 4.6 3.9 4 3.7-4 - 2.6-2.9 - S 3 3 3.3 3.6-4 3-3.6 - - Ȓ 5.1 6 3.6 3 5-5.6 3.6-4 - -

tȒ 3.7 3 2.6 3 3-3.6 3-3.6 2-2.3 2-2.3 t 3 2.6 3 3-3.6 3-3.6 2-2.3 2.3-2.6 v - 3 2.6 3-3.6 3-3.6 2.3-2.6 2.6-2.9 l 3 3 2.6 3 3-3.6 3-3.6 - - j 3 3 2.5 3 3-3.6 3-3.6 2-2.3 2.3-2.6


72

Maya (1990) all the nasals were acquired one year and three months earlier. Fricatives: None of the fricatives tested (/s/, /Ƒ/, /球/ and /h/) met 75% criteria of acquisition by 3 years. These results are in consonance with the Western study by Wellman et al., 1931; Poole, 1934; Sander, 1972 and Fudala & Reynolds, 2000), which indicated the late acquisition of fricatives. Similarly the results are also in agreement with Maya‟s (1990) findings, that the fricatives /s/, /Ƒ/, /球/ and /h/ were acquired much later.

Aff ricates: The results indicated that affricate /Ƶ/ was acquired at the age of 2.3 years, followed by aspirated

/Ƶȸ/ by 2.6 years, whereas voiced affricate /Ʋ/ was achieved only by 3 years. The present results indicated that the acquisition of affricates was much earlier compared to the western studies, as per Sander (1972) by 4 years, Prather et al., (1975) by 3-3.8 years and Fudala and Reynolds (1986) by 5.6 years. On comparison with Maya‟s (1990), the unaspirated affricates (/Ʋ/, /Ƶ/) were acquired 6 months earlier and aspirated affricates were acquired 4 years earlier in the present study.

Laterals: The results revealed that voiced retroflex lateral (/宜/) was acquired earlier (2.9 years) by 90% of the children, where as voiced alveolar lateral (/宜/), voiced palatal lateral (/欺/) did not reach 90% criteria by 3 years. On comparison with Maya‟s (1990) findings, voiced retroflex lateral was acquired almost one year earlier in the present study.

Flaps: The Flap /r/ was acquired by only 75% of the children by 3 years of age. In Maya‟s (1990) findings /r/ was acquired only by 4 years of age.

Trill : The trill /R/ was not mastered even at the age of three years. Also the observation was that the percent of acquisition of trills was higher in the final positions (80%), followed by medial (53%) and then the initial position (26.6%) for both boys and girls. On comparison with Maya‟s (1990), /R/ in final position was acquired almost two years earlier in the present study.

Frictionless continuant (Semivowels): /j/ reached 90% criteria by 2.3 years itself, whereas /v/ reached only by 2.9 years. This finding is well supported by Mowrer and Burger‟s (1991) study, which indicated the early acquisition of glides before 3+ years itself.

On comparison with Maya‟s (1990), both the phonemes were acquired much earlier, /j/ by one year and /v/ by 6 months. Figure 2 indicates the manner of consonant acquisition by 90% of the children in the present study. IV Speech sound acquisition versus word position: In the present study the phonemes /l/, /R/ and /m/ were tested in all the three positions, where as retroflexes /宜/ and /吃/ only in medial and final positions, as it occurs only in these two positions. All other consonants were tested in both initial and medial positions. Results indicated that the phoneme /m/ was mastered in all the three positions by the age of 2.3 years itself. The phonemes /l/, /R/ and /吃/ were first acquired in final, followed by medial and initial positions by 90% of the children. However phoneme /宜/ was achieved first in medial position compared to final position. When considering the phonemes that were tested in initial and medial positions, the results found that /g/, /Ʋ/, /ƀ/, /危/,

/r/, /kȸ/, /Ƒ/, /球/ and /s/ were first acquired in medial position of words whereas /v/, /d/, /t/ and /j/ were acquired in the initial position. The present findings are similar to the study by Prathima and Sreedevi in Kannada and Stoel-Gammon‟s (1985) in English which indicated the phoneme /r/ appeared earlier in medial position than in the initial position of words. Inter-judge reliability: Inter-judge reliability for phoneme transcription was assessed by comparing the percentage of phoneme agreement between the transcriptions of the investigator and those of the two judges on 10% of the samples randomly selected across the total 120 samples. The inter-judge agreement for phoneme transcription was 94%.

Based on the scoring obtained from the present study, typically developing Malayalam speaking children in the age range of 2-3 years should obtain mean scores as shown in Table 4. As no significant gender difference was observed, the articulation scores are combined for each age group.

Table 4. Mean articulation scores expected for typically developing children in the age range of 2-3

years

Age in years Mean articulatory scores Maximum score - 82

2.0 - 2.3 years 66.18 + 2.02 2.3 - 2.6 years 66.40+ 2.92 2.6 - 2.9 years 68.63+ 2.93 2.9 – 3.0 years 69.47+ 3.86


73

References

Manner of articulation Boys and Girls

2-2.3 years 2.3-2.6 years 2.6-2.9 years 2.9-3 years

Plosives /k/ /g/ /t/ /d/ /p/ /b/

/しȸ/ /kȸ/ /居ȸ/ /pȸ/

Fricatives /s/ /Ƒ/ /球/ /h

Affricates /Ƶ

/Ƶȸ/ /Ʋ/

Nasals

/m/ /n/ /┕/

Glides /j/ /v/

Trills /l/ /宜/ /欺/

Flaps

/r/ Trill /R/

Frictionless continuent /v/ /j/

Figure 2. Manner of consonants acquisition by 90% of children.

Indicates sounds acquired by 90% of the children. Indicates sounds not acquired by 90% of the children


74

Conclusions

The purpose of the study was to obtain articulatory scores for Malayalam speaking children in the age range of 2-3 years using Malayalam articulation test (MAT). The results revealed that all the vowels tested were acquired by the age of 2.3 years itself except /u/ and /u:/. The children acquired bilabials, labiodentals, dentals and velars earlier compared to alveolars, palatal, retroflex and glottal sounds. Unaspirated sounds were acquired earlier compared to aspirated sounds. The early articulatory acquisition in the present study compared to some earlier studies may be attributed to differences in lifestyle and a change in norms, over years because of greater exposure to speech and language environment.

References

Arlt, P. B., & Goodban, M. T. (1976). A comparative study of articulation acquisition as based on a study of 240 normals, aged three to six. Language, Speech, and Hearing Services in Schools, 7, 173-180.

Arun Banik (1988). Articulation test in Bengali. Unpublished master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing. University of Mysore, Mysore.

Bauman-Waengler, J. A. (1994). In J. A. Bauman-Waengler. (2004). Articulatory and phonological impairments: A clinical focus. Boston. Allyn and Bacon.

Dyson, A. T. (1988). Phonetic inventories of 2 and 3 year old children. Journal of Speech and Hearing Disorders, 53, 402-410.

Fudala, J. B., & Reynolds, W. M. (1986). In M. N. Hegde. (2000). Assessement and treatment of articulation and phonological disorders in children. United states of America. Pro-ED Inc.

Fudala, J. B., & Reynolds, W. M. (2000). Arizona articulation proficiency scales (2nd Ed.). Los Angeles: Western psychological Press.


Kumudavalli (1973). The relationship between articulation and discrimination of Kannada speech sounds in terms of distinctive features. Unpublished Master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing. University of Mysore, Mysore.

Maya, S. (1990). An articulation test battery in Malayalam. Unpublished Master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing. University of Mysore, Mysore.

Mowrer, D., & Burger. S. (1991). A comparative analysis of the phonological acquisition of consonants in the speech

of two and a half and six year old Xhosa and English speaking children. Clinical Linguistics and Phonetics, 5,139-64.

Mecham, M. J. (1962). In J. A. Bauman-Waengler. (2004). Articulatory and phonological impairments: A clinical Focus. Boston. Allyn and Bacon.

Padmaja, B. (1988). Telugu test of articulation and discrimination (TTAD). Unpublished Master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing. University of Mysore, Mysore.

Poole, I . (1934). In M. L. Hanson. (1932). Articulation. United States of America. W. B. Saunders Company.


Prathima, S., & Sreedevi, N. (2009). Articulatory acquisition i n typically developing native Kannada speaking children. Students research at AIISH. VII. University of Mysore, Mysore.

Robb, M. P., & Bleile, K. M. (1994). In M. E. Gordon- Brannan, & C. E. Weiss, (2007). Clinical management of articulatory and phonologic disorders. Philadelphia LippincottvWillaims & Wikims.

Roe, V., & Mil isen, R. (1942). In B. Padmaja. (1988). Unpublished Master‟s dissertation submitted in part- fulfillment for the master‟s degree in Speech and Hearing. University of Mysore, Mysore.


Smit, A. B., Hand. L., Freilinger, J. J., Berthal, J. E. & Bird, A. (1990). The Iowa articulation norms project and its Nebraska replication. Journal of Speech and Hearing Disorders, 55, 779-798.

So, L. H., & Dodd, B. J. (1995). The acquisition of phonology by Cantonese speaking children. Journal of Child language, 22,473 – 495.

Stoel-Gammon, C. (1985). Phonetic inventories, 15-24 months: A longitudinal study. Journal of Speech and Hearing Research, 28, 505-512.

Tasneem, B. (1977). Articulatory acquisition in Kannada A study of normal children 3-6.6 years. Unpublished Master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing. University of Mysore, Mysore.

Templin, M. C. (1957). Spontaneous versus imitated verbalization in testing articulation in preschool children. Journal of Speech and Hearing Disorders, 12, 293-300.

Usha, D. (1986). Tamil Articulation Test. Unpublished master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing. University of Mysore, Mysore.



75

APPENDIX ( Score sheet)

Name: Age/ sex: Place:

Sl. No. Phoneme Position Check word CR* S* O* D* A* Ao* Score

1 a initial a吃吃a:n 2 a: initial a:na 3 i initial ila 4 I: initial I:Ƶa 5 u initial uRi 6 u: initial u: ƀƀa:l 7 e initial eli 8 e: initial e:吃i 9 o initial onnŨ 10 o: initial o:la 11 k initial kuda 12 medial ta:ko:l 13 g initial gada 14 medial bagŨ 15 ┕ medial ma: ┕a 16 Ƶ initial Ƶi:ppŨ 17 medial Pu:Ƶa 18 Ʋ initial Ʋannal 19 medial gaƲam 20 ƀ initial ƀandŨ 21 ƀa:ƀƀu:l 22 t initial tapŨ 23 危 initial 危o:kta:r 24 medial rodŨ 25 吃 medial ki吃ar 26 final tu:吃 27 t initial tata 28 medial mo: tiram 29 d initial di:pam 30 medial maddalam 31 n initial nak球atram 32 p initial Pu:vŨ 33 medial uduppŨ 34 b initial bassŨ 35 medial Riban 36 m initial ma:la 37 medial am:a 38 final maram 39 j initial Je:Ƒu 40 medial mujal 41 l initial lo:ri 42 medial alama:ra 43 final viral


76

44 宜 medial va宜a 45 final va: 宜 46 v initial vi:dŨ 47 medial Ƶevi 48 h medial simham 49 -nt medial pa:ntŨ 50 -拒 medial pu:mpa:拒a 51 -nt medial pantŨ 52 -nj medial sanji 53 -nd medial tivandi 54 nk medial panka 55 s initial su:rjan 56 medial kas:era 57 pȸ medial pȸalam 58 r initial ra:Ʋa:vŨ 59 medial ƵeruppŨ 60 -ty medial intya 61 Ƒ initial Ƒankȸǩ 62 medial me:Ƒa 63 欺 medial ko: 欺 64 R initial Re:dio 65 medial uRumpŨ 66 final caR 67 ndra medial Ƶandran 68 pra initial pravŨ 69 kra medial Ƶakram 70 tra medial pa:tram 71 球 initial 球artŨ 72 medial bra球Ũ 73 sta medial pustakam 74 ska initial sku:tar 75 biskattŨ 76 Ƶȸ initial Ƶȸ:ja 77 tȸ medial ka tȸaka宜i 78 居ȸ medial mi 居ȸa:ji 79 stra medial vastram 80 kȸ initial kȸagam 81 medial mukȸam 82 k球 initial nak球atram

* ‟CR‟-Correct responses, „S‟-substitutions, „O‟-omissions, „D‟-distortions, „A‟-additions & „Ao‟-any other type of articulatory deviation

Self therapy manual for persons with Parkinson‟s Disease

77

A Self Therapy Manual in English for Speech and Swallowing Difficulties in Persons with Parkinson’s Diseases (PsWPD)

Gayathri Krishnan 1& R.Manjula 2

Abstract

The advancements in the field of medical tourism in India have not yet found its roots in the remote society of the country. The motoric, cognitive, psychological, social and financial limitations of Persons with Parkinson‟s disease (PsWPD) and the requirement on the part of the family members to attend to regular long term management has given rise to the need for a self therapy manual for speech and swallowing difficulties of PsWPD. This self therapy manual is offered as a practical guide that includes general information about Parkinson‟s disease, answers to the most frequently asked questions by the PsWPD and their family members and also includes sections of most useful, practical, effective and easy to do muscle training exercises to improve the speech and swallowing problems in this population of clients. Also, information on augmentative and alternative communication methods and compensatory swallowing maneuvers are provided. The final section of this manual consists of a list of hospitals and specialized centers for medical and non medical management of Parkinson‟s disease and the contact numbers of these centers. Key words: self therapy, manual, Parkinson’s disease, speech and swallowing difficulties

iagnosis 1of the progressive neurological condition of Parkinson‟s disease (PD) is always difficult for the client and the family to accept.

After a phase of initial denial, when the client and family members ultimately come to terms with the diagnosis of the condition, there arise a number of questions, doubts, and worries in the members of the family and the Person with Parkinson disease (PWPD). The anxiousness to know what is the first best step to get rid of or cope with this disease is always on top of their thoughts. Lack of awareness among the general public creates situations of social ridicule, non acceptance and many other factors which in turn subjects the PWPD as well as his/her family members to undue stress. Lack of a number of super specialty treatment centers specifically for the PsWPD in all cities of India is another big challenge, so much so that in rural areas, PD might get passed off as any other aging problem. Even for the PWPD whose diagnosis is made in multispecialty hospitals, the guidance received from medical professional stops at the level of rehabilitative measures in the first tier of the disease, which is about medical line of treatment. Not much is focused on rehabilitative services offered by Speech-Language Pathologist for swallowing and speech disorders, Physiotherapist and Occupational therapists for the physical issues, Dietician for management of the diet and other team members involved in the management of Persons with Parkinson‟s disease (PsWPD).

1email:[email protected]; 2Professor of Speech Pathology, AIISH, Mysore, [email protected]

Sometimes, the medical professional themselves in many of the hospitals are not aware of the type of help a non medical professional such as speech-language pathologists, physiotherapists and dieticians can offer to PsWPD. Hence the PsWPD are in a blind spot and they do not seek management services from such professionals which in turn will have a bearing on the lifestyle and coping behaviors of the PWPD. Compared to the western countries where many clinical facilities are available in almost all major hospitals for movement disorders such as PD, there are only a few centers offering specific treatments to PsWPD in India. These few clinics or hospitals are spread out in different states of the country and may not be accessible to all PsWPD and this means that the PWPD may not receive holistic management package to overcome his/her problems.

The medical treatment is the first step any family would take up for treatment of the PWPD. The doctor dealing with the client prescribes the needed dosage of medicines. Speech and swallowing difficulties may arise and become pronounced towards the middle and later stages of the disease. Most of the clients with PD do not get referred to speech-language pathologists due to lack of awareness on the part of the clients, their caregivers as well as medical professionals. A speech-language pathologist often deals with a PWPD when the disease has progressed and has significantly started affecting the speech and swallowing functions. At this stage of disease, the intervention strategies may not yield maximum benefits. A small percentage of them may be referred in the initial stages but the often discontinue the long term treatment due to the physical disabilities due to

D




78

PD as well as its medications. Thus, PsWPD who receive long term treatments for speech, communication and swallowing functions under the direct supervision of a speech-language pathologist are very few. In order to counter these limitations a self therapy manual for PsWPD and their caregivers is proposed to help and guide them with respect to speech and swallowing functions that appear in the course of the disease.

The resources for treatment of PD clients are often available in the form of text books for clinicians. A practical guide or manual for this purpose has never been attempted by speech-language pathologists in India. While textbooks about Parkinson‟s disease, its assessment, diagnosis and management (Biziere & Kurth, 1997; Duvoisin & Sage, 2001; Christensen, 2005; Tuite & Fernandez, 2009) reaches many clinicians very few reaches the client and family of PsWPD (Weiner & Schullman, 2001; Foltynie, Lewis & Barker, 2003; Schwarz, 2006). A self help guide for patients with PD was developed by Jahanshahi and Marsden in 2000 to bridge the gap between patient- professional communications. But the information provided in this self help guide is now outdated as the science has improved over the decade. There is none that is adapted to the Indian scenario and none focusing on speech and swallowing functions in PsWPD. Hence this self therapy manual is written for the benefit of clients with PD in India.

The aim was to prepare a self therapy manual in English that specifically deals with the behavioral management of speech and swallowing difficulties in PsWPD and the family members along with a section of Frequently Asked Questions (FAQs) about the disease and treatment. This manual consists of three parts: Part I includes section on FAQs for PWPD and their family members. Part II includes material for therapeutic management of speech and swallowing problems in PsWPD. In this section, exercises and strategies are suggested to cope and improve speech, communication and swallowing abilities in a PWPD. Part III includes a list of centers and professionals who deal specifically with the management of Parkinson‟s disease in India.

Method

A self therapy manual in English was prepared for persons with PD in India, which can easily be comprehended and used to improve their speech,

communication and swallowing functions at all stages of the disease. The manuscript of this manual was prepared and verified for its sensitivity in six stages.

Stage 1: Preparation for the manual: Information from the literature on different aspects of Parkinson‟s disease (PD) was collected and compiled. Taking this as the baseline, the contents of the manual was written, instructions and supportive material such as photographs, list to resources etc. was collated to develop the manual in English.

Stage 2: Development of the structure of the manual: The manual mainly consists of three parts.

Part I: Frequently asked questions: This part consists of three sections. Section 1: Frequently asked questions about PD: This section consists of a set of frequently asked questions about the Parkinson‟s disease. It provides information on Parkinson‟s disease, especially speech and swallowing difficulties, the factors that affect these, the changes that might be expected with the progression of disease, the treatment options available to PsWPD in a very simple and easy to understand language. Section 2: Frequently asked questions by PsWPD: This section consists of “frequently asked questions” by the PsWPD. Section 3: Frequently asked questions by the family members or caregivers of PsWPD: This section was developed by keeping in mind, the family of a PWPD. The questions that may arise in the minds of every family member of a PWPD, their concerns was compiled and answers were provided.

Part II: Management of speech and swallowing disorders in PsWPD: This section was developed by collecting the best information on the speech and swallowing treatment options for PsWPD. It further consists of two sections. Section A: Management of speech and communication problems in PsWPD: An introduction to normal aspects of speech and swallowing functions was included as the first section, in order to help a PWPD to understand the rationale behind every exercise that is recommended for improving speech and swallowing systems. This section is further subdivided into 5 subsections. Section A.1: Questionnaire for speech and communication difficulties in PsWPD: This section consists of a questionnaire on speech production mechanism and communication systems. This questionnaire was developed by including the most common problems a PWPD encounters in his/her activities of daily living. The questionnaire included bipolar questions (Yes or No questions). The PWPD can read and answer the questions by himself/herself and

Self therapy manual for persons with Parkinson‟s Disease

79

also consult family members to answer the questions. From the answers to the questionnaire a three point rating scale of severity was developed with equal intervals, on which a PWPD can rate his/her difficulties in different speech production systems as mild, moderate or severe. Section A.2: Management of speech and communication difficulties in PsWPD: This section includes exercises suggested for strengthening the four speech subsystems that are eventually affected in PsWPD: Respiration, Phonation, Articulation and Prosody. Exercises for each of these subsystems were given as separate subsections under the headings: A.2.1. Exercises to improve breathing abilities in

PsWPD. A.2.2. Exercises to improve quality of voice in

PsWPD. A.2.3. Exercises to improve production of speech

sound in PsWPD. A.2.4. Exercises to improve prosody/ tune of speech in

PsWPD.

Exercises for improving resonatory system were not included in the manual because the literature on hypokinetic dysarthria reported that difficulties with this system are not common in PD. The suggested sequence in which these exercises need to be practiced is based on the recommendation by Dworkin (1991). The exercises recommended for improving functions of these speech subsystems are those recommended in literature and the most effective exercises are adapted to the Indian home scenario. Care was taken to recommend the use of common materials that are found in every home in India for these exercises.

The last subsection of section A.2 (section

A.2.5.) consists of information about Augmentative and Alternative forms of Communication (AAC) to improve communication in PsWPD. Section B: Management of swallowing difficulties in PsWPD: This section consists of two major sections. Section B.1. Questionnaire on swallowing difficulties in PsWPD: This section included a questionnaire that facilitates rating of the swallowing difficulties of PsWPD in different phases of swallowing. Each phase of swallowing is addressed separately. The questions include bipolar questions (Yes or No questions), in a simple language so that it could be well understood and easily answered by the PWPD him/her self and if needed with the help of the family members or care givers. At the end of each section of the questionnaire, a three point rating scale of equal interval is provided to rate the severity of difficulty the PWPD faces in that particular phase of swallowing. Based on this, the exercises were recommended in the further sections.

Section B.2. Management of swallowing difficulties in PsWPD: This section consists of the basic exercises to improve swallowing. This section is further subdivided into the following subsections. B.2.1. Exercises to improve the strength of muscles of

mouth (Oral preparatory and Oral phase of swallow).

B.2.2. Exercises to improve the movement of food from pharynx to food pipe (pharyngeal phase of swallow).

B.2.3. Postures that help in improving swallowing ability in PsWPD.

B.2.4. Diet modifications for different stages of PD.

The exercises included in all these sections are based on literature survey. The most effective techniques are provided with adaptations to the Indian context. The modifications in diet are also recommended based on a review of literature on dysphagia diet and nutrition status specifically for PsWPD.

Part III: Addresses and contacts of hospitals and professionals who deal with PsWPD in India: This section aimed at providing further guidance and support that might be required to the client and family. It included information about centers of services for medical, speech and swallowing function, information on self help groups or associations that function for the social, personal and financial welfare of PsWPD. The details about these were collected by running a wide search on the internet about medical tourism in India.

Stage 3: Check for content validity: The manual was given to three Speech- language pathologists who were experienced in rehabilitation of individuals with Parkinson‟s disease for their feedback on the contents of the manual. The feedback was collected in the form of a three point rating scale on various sections of the manual. The rating scale provided was as follows: 1- The contents are not very valid. 2- Some contents are valid but some are not. 3- All contents are valid. Stage 4: Modification of the content based on the feedback obtained in Stage 3: The suggestions obtained from the judges of stage 3 were incorporated and the modified manual was subjected to further procedures.

Stage 5: Sensitivity of the manual as a resource material for PsWPD: Three individuals suffering from Parkinson‟s disease were provided with the manual for their use. Their suggestions and feedback was obtained on a three point rating scale. The rating scale provided was as follows:


80

1- Difficult to comprehend the content of all sections 2- Easy to comprehend few sections only. 3- Easy to comprehend all the sections.

Stage 6: Modification of the content based on the feedback from the client and family members and finalization: After incorporating all the changes that were suggested by the PsWPD and their family members or caregivers who may use this manual, the manual was finalized for its content and structure.

Conclusions

This self therapy manual was proposed to help PsWPD to cope with their speech and swallowing difficulties during the course of the disease. This manual was prepared based on the best knowledge of literature, clinical evidences and reports. The sections of this manual were rated to be well understood by three PsWPD and approved for its contents by experienced speech-language pathologists. Thus, this manual is expected to be useful for PsWPD.

Limitations of this manual

The manual is in English and thus may be useful only for literate PD clients and their families. Although it is well attempted to offer as a comprehensive manual for self therapy, it may not satisfy the needs of all PD clients because it is known that PD clients exhibit a wide heterogeneity in their characteristics. Hence in some clients, the manual may only serve as a rudimentary guide.

Recommendations

The manual needs to be clinically validated for its usefulness and suitable modifications incorporated based on the feedback obtained from the target groups such as PD clients, their caregivers and others in the society. Translation of the information in this guide to other native Indian languages will help in reaching out to more families who need such help. Also, this manual can be digitalized to help families access the information almost instantaneously.

References Biziere, K., & Kurth, M. C. (1997). Living with parkinsons

disease. Demos Medical Publishing. New York. Christensen, J. H. (2005). Parkinsons disease: an essential

guide for the newly diagnosed. Marlowe & Co. USA. Duvoisin, R. C., & Sage, J. (2001). Parkinsons disease: a

guide for patient and family. Lippincott Williams & Wilkins, Newyork.

Dworkin, J. P. (1991). Motor speech disorders: a treatment guide. Mosby-Year Book. University of Michigan.

Foltynie, T., Lewis, S., & Barker, R. A. (2003). Parkinson‟s disease: your questions answered. Elseiver Health Sciences, UK.

Jahanshahi, M., & Marsden, D. C. (2000). Parkinson‟s disease: a self help guide. Demos, UK.

Schwartz, S. P. (2006). Parkinsons disease: 300 tips for making life easier. Demos. UK

Tuite, P. J. & Fernandez, H. (2009). Parkinsons disease: a guide to patient care. Springer Publishing Company, New York.

Weiner, W. J. & Schullman, L. M. (2001). Parkinsons disease: a complete guide for patients and families. Johns Hopkins University Press, UK.

Attitudes, anxiety & coping strategies in PWS

81

Attitudes, Anxiety and Coping Strategies in Persons with Stuttering

Jasmine Mallik 1 & Y.V.Geetha 2

Abstract

People who stutter (PWS) tend to have increased levels of anxiety compared to people who do not stutter (PWNS), particularly in social situations. Coping behaviors in the developing or chronic stuttering problem can take many forms and change with experience. The present study has the following objectives as to how: (a) attitudes and coping strategies differ with respect to severity of stuttering, (b) attitudes and coping strategies differ with respect to chronicity of stuttering, (c) attitudes and coping strategies differ with respect to relapses, and to study (d) the relationship between attitude, anxiety and coping in PWS. This was carried out in two phases: in phase I a questionnaire containing thirty-five questions was prepared in English pertaining to attitudes, anxiety and coping strategies seen in PWS which was administered to all the participants individually. The participants included were thirty individuals in the age range of 10-40 years, diagnosed as having stuttering by qualified speech language pathologists. These included 10 new PWS who had not taken therapy earlier, 10 PWS who had undergone therapy and had improved and 10 PWS who had undergone therapy and have had relapses. It was seen that the scores of attitudes in PWS in the new and relapse groups were much higher than the post therapy group where there was increased anxiety about speaking situation when meeting new people/superiors and hence avoided speaking. Regarding the coping strategies in PWS it was found that there was no significant difference (p>0.05). The participants included in one month follow up after therapy and relapse group had changes in attitudes, anxiety and coping strategies. It was also seen that most of the PWS had negative feelings such as fluent periods may not last long. PWS differed in attitudes, anxiety and coping strategies in mild and moderate severity groups. Therefore it can be pointed out that PWS do have attitudes, anxiety problem and adopt various coping strategies. This can also be seen at various severity levels. Therefore it may be concluded that PWS do suffer from negative feelings, inferiority complexes, anxiety related to stuttering and personality changes which could be changed with treatment. All these above mentioned factors cause fear of stuttering and this in turn leads to avoidance of speaking situations. Key words: anxiety, attitudes, coping strategies

tuttering is an 1involuntary, intermittent and debilitating speech disorder that afflicts approximately 1% of the population. Its

primary manifestations include aberrant sound prolongations and syllabic repetitions that are interspersed with otherwise perceptually normal speech patterns (Bloodstein, 1995). In other words, a person who stutters may begin oral communication normally without disruption and then suddenly and uncontrollably, begins to produce unexpected rapid oscillatory syllabic repetitions (Kalinowski, Saltuklaroglu, Guntupalli & Stuart, 2004). Stuttering is described as unusually frequent disruptions in the flow of speech (Guitar, 2006). These disruptions include phoneme, syllable or word repetitions, phoneme prolongations, and airflow or voicing blocks. Additional symptoms include facial grimacing, fixed articulatory postures, and obvious fear during speech attempts, or anticipation of speech failure prior to speech attempts (Sheehan, 1975). Nevertheless, these overt symptoms of stuttering are only a small part of the disorder, resulting in the analogy of stuttering as an iceberg (Sheehan, 1975). The audible and visible signs of stuttering are likened to the tip of an iceberg that rises above the water level. Yet, far greater and more detrimental is its submerged portion, which when likened to stuttering,

1e-mail: [email protected]; 2 Professor of Speech Sciences, AIISH, Mysore, [email protected]

comprises feelings of fear, shame, guilt, anxiety, hopelessness, isolation, and denial.

People who stutter (PWS) tend to have increased levels of anxiety compared to people who do not stutter (PWNS), particularly in social situations (Messenger, Onslow, Packman & Menzies, 2004).

Of particular interest to the present study is the

role of anxiety in stuttering. There are several well-known theories that have been developed which focus on anxiety and stuttering. For example, the Two-Factor Theory of stuttering (Brutten & Shoemaker, 1967) suggests that the listeners‟ negative reaction to the speech of PWS conditions a link between speech and anxiety. An individual‟s consequent avoidance of phonemes and words they perceive as difficult, or even avoidance of speech situations due to apprehension of stuttering, results in stuttering, and thus reinforces the link between speech and anxiety. A similar theory, known as the Anticipatory struggle hypothesis (Bloodstein, 1987), suggests that some children simply consider speech a demanding task. This is primarily due to experiencing difficulty and frustration. On the other hand, the Approach-Avoidance Conflict Theory, proposed by Sheehan (1953), is based on the notion of internal conflict. Although PWS desire to speak in social situations, they are also afraid of speaking for

S


82

fear of stuttering. The Approach Avoidance Conflict theory was further developed by Sheehan (1975), based on earlier work by Miller (1944), and described as the Double Approach-Avoidance Conflict Theory (Miller, 1944; Sheehan, 1975). In this there are approach and avoidance tendencies for both speaking and remaining silent. Firstly, when PWS desire to approach speaking to fulfill their social obligations, they are simultaneously faced with a fear of stuttering during their speaking attempts (Johnson & Knott, 1936). The alternative to speaking is silence, which appears an appealing approach tendency, since it bypasses the potential risk of stuttering that is associated with speaking. However, silence is also a threat to social standing. Struggling between the possibilities of speaking and remaining silent, together with an inability to resolve this inherent conflict, consequently results in stuttering. An additional facet of the Double Approach-Avoidance Conflict Theory is the association of negative emotions to either speaking or remaining silent. In the event of speaking, the trade-off is shame and guilt, whereas in remaining silent, feelings of frustration and guilt are experienced. Because both choices result in guilt, a choice must be made between experiencing either shame or frustration.

It is commonly believed that anxiety is related

to stuttering, despite conflicting evidence in the literature with regard to the precise nature of this relationship (Craig, 1990; Weber & Smith, 1990; Miller & Watson, 1992; Blood, Blood, Bennett, Simpson & Susman, 1994; Poulton & Andrews, 1994; Ezrati-Vinacour & Levin, 2004). PWS often report anxiety related to producing particular sounds or words, or participating in certain communicative situations (Blood et al., 1994). In addition, stuttering severity appears to be dependent on factors such as communication partner status or the number of addressees, novelty, formality, and familiarity with the speaking situation, and feelings of conspicuousness (Porter, 1939; Siegel & Haugen, 1964; Buss, 1980). Because stuttering severity is associated with emotions such as embarrassment, frustration and apprehension of negative social evaluation, greater anxiety levels in PWS compared to PWNS are to be expected (Craig, Hancock, Tran & Craig, 2003). Nevertheless, it remains unclear at present, whether PWS are more anxious in general than PWNS.

The act of communication happens always in

the social context, involving one or more listeners. Hence, communication disorders are always entangled with the attitudes of the listeners towards that disorder and the person who possess the disorder. Such attitudes are influenced by the level of adequacy of communication. People with communication disabilities, especially stuttering,

develop a negative personality stereotypes maintained by different groups of people.

When stuttering, PWS will often use coping

strategies such as nonsense syllables or less-appropriate (but easier to say) words to ease into the flow of speech. They also may use various personal tricks to overcome stuttering or blocks at the beginning of a sentence, after which their fluency can resume. Finger-tapping or head-scratching are two common examples of tricks, which are usually idiosyncratic and may look unusual to the listener. Hence, it is very important to assess the attitudes, anxiety and coping strategies in PWS and to intervene if required. Stuttering is a heterogeneous group of disorders, and hence it is necessary to study it in different cultural and linguistic perspective.

There is a need to study the attitudes, anxiety

and coping strategies in PWS with regard to various groups of PWS such as: new PWS, PWS after a month of therapy and in relapse cases. This is in view of the general observation that; because of the negative attitudes PWS are resistant to changes in their fluency even after fluency therapy.

According to Personal Construct Theory (Kelly, 1955), “A person's unique psychological processes are channeled by the way s/he anticipates events”. There are relapses seen in at least 50% of PWS, Assessing the attitudes and coping strategies in PWS will be helpful in overall management of PWS.

Hence, the present study was planned by taking

into account all these factors. It was also desirable to study if the attitudes, anxiety and coping strategies in PWS vary with respect to severity, chronicity and family history which might later help in intervention of PWS. The present study was hence planned with the following objectives as to how (1) attitudes and coping strategies differ with respect to severity of stuttering, (2) attitudes and coping strategies differ with respect to chronicity of stuttering and (3) attitudes and coping strategies differ with respect to relapses and to study the relationship between attitude, anxiety and coping in PWS.

Method

The present study was conducted to find out attitudes, anxiety and coping strategies in PWS. This was carried out in two phases.

Phase I: A questionnaire was prepared in English through literature survey, consisting of statements to gather information related to attitudes, anxiety and coping strategies seen in PWS. It contained thirty-five questions pertaining to attitudes, anxiety and coping strategies seen in PWS (See Appendix).


83

Phase II: The questionnaire was administered to all the participants individually. Participants: Thirty individuals in the age range of 10-40 years, diagnosed as having stuttering by qualified speech-language pathologists, were considered as participants of the study. These included 10 new PWS who had not taken therapy earlier, 10 PWS who had undergone therapy and had improved and 10 PWS who had undergone therapy and have had relapses. The participants were chosen regardless of language, gender and severity. Exclusion criteria: PWS were excluded from the study if they had any associated central neurological problems, language problems, psychiatric problems, any sensory-motor deficits. Materials: The Anxiety, Attitudes and Coping questionnaire developed as part of the study consisted of 2 parts; Part I included demographic data, questions regarding the participants‟ age, background, family history, stuttering, and therapy experiences. Part II of the questionnaire included 10 questions to investigate the attitudes, 10 questions to investigate anxiety behaviors and 15 questions to investigate coping strategies in PWS.

Each item in the questionnaire was expressed as a statement. The subjects responded to each statement by putting a tick on the appropriate response options. All the components of the questionnaire are rated on a 5-point rating scale ranging from 0 to 4 by the participants (0-no/never/not at all; 1- sometimes [<25%]; 2- medium/average amount [25-50%]; 3- usually/a good deal/rather often [50-75%]; 4- practically always/entirely [>75%]). Procedure: Prior to the interview written consent was obtained from the participants and the questionnaire was administered through interviewing each of the participants individually. The data obtained from the participants were tabulated and analyzed using SPSS 16 software to answer the research questions.


I. Attitudes in PWS: Figure 1 depicts mean scores of attitudes across groups. The mean scores of relapse group were much higher than the other two groups.

That is, they had increased negative attitudes and poor self-esteem, may be because the participants‟ confidence level in speaking situations had decreased in spite of attending therapy. Participants belonging to post-therapy had lesser mean scores indicating that they had decreased negative attitudes after attending therapy, may be

Groups

RelapseAfter 1 monthNew

30.0

28.0

26.0

24.0

22.0

20.0

18.0

16.0

14.0

12.0

10.0

22.9

14.0

16.8

Figure 1. Mean scores of attitudes and groups.

because the participants‟ self-perception of attitudes had improved after therapy. The new PWS exhibited scores in between these two groups which is also as to be expected. MANOVA was done to compare the overall scores of the subtest-attitude across groups which revealed that for F (2, 27) at p<0.05, the score of the subtest was 2.622. There was no significant difference in attitudes in the three groups of PWS. This may be because of the fact that stuttering is a heterogeneous group of disorders and also probably due to limited number of participants (ten participants in each of the three groups) with many individual differences as seen in mean and standard deviation values (Table 1). Van Riper (1970), Manning, Dailey and Wallace (1984) and Knott (1935) too came to the same conclusion. The findings are also in consonance with the study conducted by Mulcahy, Hennesey, Beilby and Byrnes (2008) who concluded that PWS do suffer from self-esteem problems and anxiety and psychosocial conflicts.

Since an overall score on the subtests did not

reveal a significant difference in the groups, subsequently the scores of individual questions in each subtest were compared across the groups. A Kruskal-Wallis test was done to find out the significant difference for individual scores of questions in all the groups. The results of Kruskal-Wallis test for individual scores of questions for subsets on attitudes revealed no significant difference (p>0.05). This may also be probably because of individual differences and the small number of subjects, in addition to the number of sessions and its duration in post therapy group which could not be controlled.

II. Anxiety in PWS: Figure 2 gives mean scores of anxiety across groups. The mean scores of new group and relapse group were much higher than the post therapy group, as expected. The new group, since they had not taken therapy showed more anxiety features because of their speech difficulty.

The relapse group had more anxiety which may be because the participants‟ confidence level in speaking situations had decreased in spite of


84

attending therapy previously due to relapse, leading to more anxiety. Participants belonging to post-therapy had lesser mean scores indicating that they had decreased anxiety after attending therapy, may be because the participants had increased confidence towards speaking situations after therapy with new techniques.

Groups


30.0

28.0

26.0

24.0

22.0

20.0

18.0

16.0

14.0

12.0

10.0

18.7

12.1

17.0

Figure 2. Mean scores of anxiety and groups.

MANOVA was done to compare the overall scores of the sub items of anxiety across groups. The results revealed that for F (2, 27) at p<0.05, the score of the subtest were 2.055. There was no significant difference in anxiety in PWS among the groups although it appears so from the graph, especially compared to the post therapy group. Since the overall scores on the sub items did not reveal a significant difference in the groups, subsequently the scores for individual items were compared on the groups. A Kruskal-Wallis test was done to find out the significant difference for individual scores of questions in all the groups. The Table 1 gives results of Kruskal-Wallis test for individual scores of questions for anxiety revealed that there was significant difference (p<0.05) only in question number A7 which elicited anxiety about speaking situation when meeting new people / superiors and

Table 1. Result of Kruskal-Wallis test for scores of questions (individual)

Items Chi-Square df Asymp. Sig.

a1 0.946 2 0.623

a2 3.348 2 0.187

a3 0.272 2 0.873

a4 3.075 2 0.215

a5 1.270 2 0.530

a6 0.750 2 0.687

a7* 6.012* 2 0.049*

a8 4.003 2 0.135

a9 2.982 2 0.225

a10 1.514 2 0.469

there was no significant difference in any of the other 9 questions. This is one of the commonest problems faced by most PWS as observed in clinical practice by most clinicians.

This result is in agreement with the study by Bloodstein (1950) in which he concluded that the less the anxiety about speech difficulties, the less the effort to avoid it, and consequently the less the stuttering. Blumgart, Tran and Craig (2010) too came to the same conclusion that the AWS had significantly raised trait and social anxiety, as well as significantly increased risk of social phobia which led to increased stuttering in PWS comparison to the PWNS.

III. Coping strategies in PWS: Figure 3 depicts mean scores of coping strategies across groups. It is evident from the Figure 3 that mean scores of new PWS and relapse group were much higher than the post therapy group which are as to be expected. That is, the relapse group had increased use of coping strategies which may be because the participants‟ confidence level in speaking situations had decreased after attending therapy previously due to relapse and hence resorted to the reuse of the same. Similarly for new PWS it is to be expected that the coping strategies are higher as reported by many authors.

A MANOVA was done to compare the overall scores of the subtest-coping across groups. The results revealed that for F (2, 27) at p<0.05, the score of the subtest was 1.505.

Groups


30.0

28.0

26.0

24.0

22.0

20.0

18.0

16.0

14.0

12.0

10.0

29.4

19.2

29.4

Figure 3. Mean scores of coping strategies among the three groups.

There was no significant difference in coping strategies in PWS among the three groups although figure showed differences in the post therapy group compared to the other two groups. The result of the study is in consonance with Carver, Scheier and Weintraub (1989) who developed a multidimensional coping inventory to assess the different ways in which people respond to stress reported correlations between the various coping scales and several theoretically relevant personality measures which emphasize effective use of coping.

The findings of Plexico, Manning and Levitt

(2009) are also in agreement with the results obtained in the present study for the post therapy


85

group. The results revealed that participants moved from emotion-based avoidant patterns of coping that focused on protecting the self and the listener from experiencing discomfort associated with stuttering to cognitive-based approach patterns that focused on the needs of the speaker. As the participants chose to approach rather than avoid or escape stuttering, they experienced many positive social, physical, cognitive, and affective results. Since an overall score on the subtests did not reveal a significant difference in the groups, subsequently the scores of individual questions in subtest on coping were compared on the groups. A Kruskal-Wallis test was done to find out the significant difference for individual scores of questions in all the groups which revealed no significant difference (p>0.05). This may be because of small sample size and also probably because of individual differences seen in PWS.

IV. Comparison of scores across groups: The mean and standard deviation scores across groups on attitudes, anxiety and avoidance or coping were in general better than for the other two groups as expected. The standard deviation scores were high for all the groups for all the three variables indicating high variability among the subjects. This is also as expected in the population of PWS in general.

Since an overall score on the subtests did not

reveal a significant difference in the groups, a post-hoc analysis using Mann-Whitney test was done across group for question A7 which consisted anxiety and avoidance about speaking situation when meeting new people/superiors since there was no significant difference in any of the other 9 questions. - Group 1 (new PWS) and Group 2 (Post therapy) - Groups 1 (new PWS) and Group 3 (Relapse) - Groups 2 (1 month post therapy) and 3 (relapse)

The results of post-hoc analysis are given below in Table 2. (i) Group 1 (new) and Group 2 (Post therapy): Post-hoc analysis done by using Mann-Whitney test across group 1 and 2 revealed no significant difference (p>0.05) between the two groups for question A7, although the raw scores and the graphs reveal differences. This may be because of limited number of participants and limitations regarding number of sessions and duration of sessions and individual differences. This result is not in consonance with study done by Guitar (1976) who contended that the clinicians may predict therapy outcomes from pre treatment attitudes. Guitar‟s Table 2. Mann-Whitney results between groups for

question A7

Groups A7 Z Asymp. Sig. (2-tailed)

1 & 2 -1.126 0.260 1 &3 -1.094 0.274

2 &3* -2.551 0.011*

conclusions were further supported by other researchers who believed the process of therapeutic change involves not only the development of smooth speech production but also modification in negative speech-related attitudes (Guitar & Bass, 1978; Andrew & Craig, 1988; Kraaimaat, Janssen & Brutten, 1988; Feinberg, Griffin & Levey, 2000).

(ii) Groups 1 (new PWS) and Group 3 (Relapse): Post-hoc analysis done by using Mann-Whitney test across group 1 and 3 revealed no significant difference (p>0.05) between the two groups for question A7 (Table 5). (iii) Groups 2 (Post therapy) and 3 (relapse): The results of post-hoc analysis are given in Table 3. Post-hoc analysis done by using Mann-Whitney test across group 2 & group 3 revealed a significant difference (p<0.05) between the two groups for question A7. This may be because after a month of therapy the participant‟s attitudes and their anxiety levels in speaking situations improved and they also tend to have increased self-esteem and also probably due to changes in participant‟s self-perceptions, similarly in relapse group.

This is in agreement with the study conducted

by Andrews and Cutler (1974) where it was concluded that through therapy the attitudes of PWS can be changed to some extent. Similarly, Guitar (1976) investigated the relationship between pre therapy attitudes of adult PWS and post therapy treatment outcomes.

V. Comparison of scores on severity (very mild, mild and moderate) in groups 1, 2 & 3: The participants were divided into four groups based on severity levels which included (a) very mild, (b) mild, (c) moderate and (d) severe. The mean and standard deviation for various severity groups were higher for subtest C (coping strategies) in relapse group at moderate level of severity. This may be because at moderate severity level the attitudes and anxiety levels may be much lesser than severe level and also probably because the problem is little compared to severe severity level.

A Kruskal-Wallis test was done to compare

the overall scores of severity levels across groups. The results revealed no significant difference in severity across groups in PWS at p>0.05 which is in agreement with the study conducted by Craig, Hancock, Tran and Craig (2003) showed that PWS had increased anxiety levels regardless of condition.

Since an overall score on the severity level

did not reveal a significant difference in the groups, subsequently the scores of individual questions in each severity levels were compared on the groups. The mean scores for different items in different subtests do not indicate differences across various


86

severity groups may be due to wider individual variations as seen in standard deviation scores and because of lesser number of subjects in each of the severity groups.

A Kruskal-Wallis test was done to find out

the significant difference for individual scores of questions in all the subtests. The following table gives results of Kruskal-Wallis test. The results of Kruskal-Wallis test for individual scores regarding questions on attitudes for three severity levels, namely very mild, mild and moderate revealed that there was significant difference (p<0.05) in question numbers N3, which included negative feelings such as fluent periods may not last long and may begin to stutter sooner and N9, which involves complexes about the PWS‟s way of speaking and what others may think about PWS.

Table 3. Result of Kruskal-Wallis test for scores on questions regarding attitudes

Questions Chi-Square df Asymp. Sig.

n1 2.538 2 0.281

n2 0.752 2 0.687

n3* 7.911* 2* 0.019*

n4 2.641 2 0.267

n5 2.491 2 0.288

n6 2.060 2 0.357

n7 1.863 2 0.394

n8 2.072 2 0.355

n9* 8.123* 2* 0.017*

n10 0.807 2 0.668

There was no significant difference (p>0.05) in any of the other 8 questions. Several studies conducted by Baumgartner and Brutten (1983), Bloodstein (1975) and Vanryckeghem and Brutten (1996) have confirmed the presence of negative communication attitudes in PWS. The results of Kruskal-Wallis test for individual scores regarding anxiety questions for three severity levels, namely very mild, mild and moderate (Table 3) revealed that there was significant difference (p<0.05) in question number A6, which includes rigid articulatory postures which makes one unable to move the articulators and there was no significant difference (p>0.05) in any of the other 9 questions.

Similar result was obtained in a study conducted by Vinacour and Levin (2004) where there was no difference in anxiety levels in PWS as a function of stuttering severity. However, there was a difference in state anxiety levels specific to social situations. Craig, Hancock and Tran (2003) too came to the conclusion that stuttering severity is associated with the anxiety levels in PWS which is in agreement with the findings of this study. The results of

Kruskal-Wallis test for individual scores regarding anxiety questions for three severity levels, namely very mild, mild and moderate revealed that there was significant difference (p<0.05) in question number A6, about rigid articulatory postures which makes one unable to move the articulators and there was no significant difference (p>0.05) in any of the other 9 questions.

Table 4. Result of Kruskal-Wallis test for scores of questions regarding anxiety

Questions Chi-Square df Asymp. Sig.

a1 2.236 2 0.327

a2 0.017 2 0.991

a3 1.321 2 0.517

a4 5.244 2 0.073

a5 1.063 2 0.588

a6* 6.463* 2* 0.039*

a7 1.755 2 0.416

a8 0.964 2 0.617

a9 1.148 2 0.563

a10 5.680 2 0.058

The results of Kruskal-Wallis test for individual scores regarding coping questions for three severity levels, namely very mild, mild and moderate revealed that there was no significant difference (p>0.05) in any of the questions.

There are no studies comparing the severity of

stuttering to coping strategies and hence no comparison could be made. Post-hoc analysis was done by using Mann-Whitney test across severity levels:(a) very mild and mild (b) very mild and moderate and (c) mild and moderate The results of Post-hoc analysis was done by using Mann-Whitney test is given below in Table 5.

Post-hoc analysis done by using Mann-

Whitney test across very mild & mild severity level and very mild & moderate severity level revealed no significant difference between any of the two groups.

Table 5. Mann-Whitney for different severity levels

Severity Very mild & mild

Very mild & moderate

Mild & moderate*

Sub tests

N3 N9 A6 N3 N9 A6 N3 N9 A6

Z -1.87 -1.46 -1.41 -.22 -.75 -.18 -2.78 -2.82 -2.58 Asymp.

sig .062 .144 .160 .826 .454 .860 .005* .005* .010*

But mild and moderate severity level revealed

a significant difference (p<0.05). This may be because of limited number of participants in very mild which consisted of only four participants and


87

mild severity level which consisted of only seven participants when compared with moderate severity level which included of eighteen participants and also because of the individual differences which PWS showed. These findings are in consonance with the study conducted by Miller and Watson (1992) where it was found that PWS with mild and moderate stuttering severity exhibited a significant positive correlation between measures of communication attitudes and both state and trait anxiety. Conversely, PWS with severe stuttering showed no significant correlations between anxiety and communication attitudes. Severe level was not used in analysis because only one subject was available with severe stuttering. Therefore it was excluded from the analysis.

Therefore from analysis it is evident that PWS showed anxiety about speaking situation when meeting new people/superiors, had negative feelings, inferiority complexes and rigid articulatory postures. PWS also had improved results after attending therapy. PWS with mild and moderate severity levels exhibited a significant positive correlation between measures of communication attitudes and both state and trait anxiety.

Conclusions

It can be pointed out that PWS do have attitudes, anxiety problems and adopt various coping strategies. This can also be seen at various severity levels. Therefore it may be concluded that PWS do suffer from negative feelings, inferiority complexes, anxiety related to stuttering, personality changes which could be changed with treatment. All these above mentioned factors cause fear of stuttering and this in turn leads to avoidance of speaking situations.

References

Andrew, G., & Cutler, J. (1974). Stuttering therapy: The relation between the changes in symptom level and attitudes. Journal of Speech and Hearing Disorders, 39, 312-319.

Andrews & Craig (1988). Prediction of outcome after treatment for stuttering. The British Journal of Psychiatry, 153, 236-240. Retrieved from http://bjp.rcpsych.org/cgi/content/abstract/153/2/236

Baumgartner, J. M., & Brutten, G. J. (1983). Expectancy and heart rate as predictors of the speech performance of stutterers. Journal of Speech and Hearing Research, 26, 383-388.

Blood, G. W., Blood, I. M., Bennet, S., Simpson, K. C., & Susman, E. J. (1994). Subjective anxiety measurements and cortical responses in adult who stutter. Journal of Speech and Hearing Research, 37, 760-768.

Blood, G. W., Blood, I, M., Maloney, K., Meyer, C., & Qualls, C. D. (2007). Anxiety levels in adolescents who stutter. Journal of Communication Disorders, 40, 452-469.

Bloodstein, O. (1950). A rating scale study of conditions under which stuttering is reduced or absent. Journal of Speech and Hearing Disorders, 15, 29-36.

Bloodstein, O. (1975). Stuttering as tension and fragmentation. In J. Eisenson (Eds.), Stuttering: A second symposium. New York: Harper and Row.

Bloodstein, O. (1987). A handbook on stuttering. Chicago: National Easter Seal Society.

Bloodstein, O. (1995). A handbook on stuttering (5th Edn.). San Diego, CA: Singular Publishing Group.

Blumgart, E., Tran, Y., & Craig, A. (2010). Social anxiety disorder in adults who stutter. Depression & Anxiety, 0, 1-6.

Brutten, & Shoemaker. (1967). The modification of stuttering. Englewood Cliffs. NJ: Prentice-Hall.

Carver, C. S., Scheier, M. F., & Weintraub, J.K. (1989). Assessing coping strategies: A theoretically based approach. Journal of Personality and Social Psychology, 56(2), 267-283.

Craig, A. (1990). An investigation into the relationship between anxiety and stuttering. Journal of Speech and Hearing Research, 55, 290-294.

Craig, A., Hancock, K., Tran, Y., & Craig, M. (2003). Anxiety levels in people who stutter: A randomized population study. Journal of Speech, Language and Hearing, 46, 1197-1206.

Feinberg, A. Y., Griffin., B, P., & Levey, M., (2000). Psychological aspects of chronic tonic and clonic stuttering: Suggested therapeutic approaches. American Journal of Orthopsychiatry 70 (4), 465-473.

Guitar, B. (2006). Stuttering: An integrated approach to its nature & treatment (3rd edition). United States of America: Library of congress cataloging-in-publication data.

Guitar, B. (1976). Pretreatment factors associated with the outcomes of stuttering therapy. Journal of Speech and Hearing Research, 19, 590-600.

Guitar, B., & Bass, C. (1978). Stuttering therapy: The relation between attitude change & long-term outcome. Journal of Speech and Hearing Disorders, 43, 392-400.

Johnson, W., & Knott, J. R., (1936). The moment of stuttering. Journal of Genetic Psychology, 48, 475-480.

Kalinowski, J., Saltuklaroglu, T., Guntupalli, V.K., & Stuart, A. (2004). Gestural recovery and the role of forward and reversed syllabic repetitions as stuttering inhibitors in adults. Neuroscience Letter, 363, 144–149.

Kelly, G. A. (1955). The psychology of personal constructs vol. 1. New York, NY: W. W. Norton and Company.

Kraaimaat, F., Janssen, P., & Brutten, G. J. (1988). The relationship between stutterer‟s cognitive and autonomic anxiety and therapy outcome. Journal of Fluency Disorders, 13, 107-113.

Manning, W. H., Dailey, D., & Wallace, S. (1984). Attitude and personality characteristics of older stutterers. Journal of Fluency Disorders, 9, 207-215.

Messenger, M., Onslow, M., Packman, A., & Menzies, R. (2004). Social anxiety in stuttering: Measuring negative social expectancies. Journal of Fluency Disorders, 29, 201-212.

Miller, N. E. (1944). Experimental studies of conflict. In J. McV. Hunt (Ed.). Personality and the behavior disorders. New York: Ronald Press.

Miller, S., & Watson, B. C. (1992). The relationship between communication attitude, anxiety and


88

depression in stutterers and non-stutterers. Journal of Speech and Hearing Research, 55, 789-798.

Mulcahy, K., Hennessey, N., Beilby, J., & Brynes, M. (2008). Social anxiety and the severity and typography of stuttering in adolescents. Journal of Fluency Disorders, 33, 306-319.

Porter, H. (1939). Studies in psychology of stuttering:XIV. Stuttering phenomena in relation to size & personnel of audience. Journal of Speech Disorders, 4, 323-333.

Plexico, L., Manning, W. H., & Levitt, H. (2009). Coping responses by adults who stutter: Part II. Approaching the problem and achieving agency. Journal of Fluency Disorders, 34, 108-126.

Poulton, R. G., & Andrews, G. (1994). Appraisal of danger & proximity in social phobics. Behaviour, Research & Therapy, 32, 639-642.

Siegel, G. M., & Haugen, D. (1964). Audience size and variations in stuttering behavior. Journal of Speech and Hearing Research, 7, 381-388.

Sheehan, J. G. (1953). Theory and treatment of stuttering as an approach-avoidance conflict. Journal of Psychology, 36, 27-49.

Sheehan, J. G. (1975). Conflict theory and avoidance-reduction therapy. In J. Eisenson (Eds.), stuttering: A second symposium (pp. 97-198). New York: Harper and Row.

Van Riper, C. (1970). Stuttering and cluttering. Folia Phoniatrica, 22, 347-353.

Vanryckeghem, M., & Brutten, G. J. (1996). The relationship between communication attitude and fluency failure of stuttering and non-stuttering children. Journal of Fluency Disorders, 21, 109-118.

Vinacour, R. E., & Levin, I., (2004). The relationship between anxiety and stuttering: a multi-dimensional approach. Journal of Fluency Disorders, 29, 135-148.

Weber, C. M., & Smith, A. (1990). Autonomic correlates of stuttering and speech assessed in a range of experimental tasks. Journal of Speech and Hearing Research, 33, 690-706.

APPENDIX PART I : CHECKLIST FOR PWS

I. General information

a. Case name: b. Case number: c.Age/gender: d. Phone number: e. Address: f. Mail: g. Education: h. Occupation: i. Languages used: At home: Kannada/English/Hindi/Tamil/Malayalam/Telugu j. Mother tongue: Kannada/English/Hindi/Tamil/Malayalam/Telugu/others k. Severity of the problem (as per SSI): v. mild/ mild/ moderate/moderately severe/ severe/ v. severe l. Fluency therapy if taken before: yes/no; If yes, duration (approx): m. Relapse of the problem: yes/no; If yes, specify reasons if any:

II. Brief history/ onset &development of problem

a. Onset: Sudden/Gradual b. Age of onset: c. Duration: d. Has the dysfluency been consistent or intermittent? Increased/reduced/remain same/fluctuating e. Associated problems: articulation/language/auditory/motor/cognitive f. Medical history: nil/affected

III. Stuttering history/reaction

a. Family history: no/yes; If yes: parents/siblings/grandparents/uncle/aunt b. Reaction towards stuttering: Self: Family members: Friends:

IV. Variation in stuttering

a. Situations b. Individuals c. Do you avoid speaking situations- yes/no d. Coping mechanisms: As reported: As observed:


89

PART II: Questionnaire for attitudes, anxiety & coping strategies in PWS Therapy: (if yes) Duration: Instructions: Please read the following statements and answer with appropriate options as noted below: 0 - No/never/not at all; 1 –Sometimes (<25%); 2 – medium/average amount (>25-50%); 3 – Usually/a good deal/rather often (>50-75%); 4 - Practically always/entirely (> 75%)

Sl.No. Questions Ratings 0 1 2 3 4

N1. I feel/anticipate interruptions in speech (e.g..repetitions, prolongations or blocks)

N2. I expect certain sounds, letters or words are be particularly “hard” to say N3. I feel fluent periods are unusual, cannot last and that sooner or later I will stutter N4. Even though knowing the right answer, I have often failed to give it because of

fear to speak out.

N5. Sometimes I feel embarrassed by the way I talk. N6. Sometimes wish that I could say things as clearly as others do N7. Worry if I make a fool of myself, or feel I have been made to look foolish? N8. Feel that other people are better than me? N9. I have complexes about the way I speak & what others think about me. N10. I feel self-conscious about my appearance even when I am well-dressed and

groomed.

A1. I have general body tension during speech attempts (e.g., shaking. trembling or feeling knotted up inside)

A2. I breath noisily or with great effort while trying to speak A3. I feel the face getting warm and red (as if blushing) while struggling to speak A4. I run out of “breath” while speaking A5. I strain to talk without being able to make a sound A6. I hold lips, tongue or jaw in a rigid position before speaking or when getting

“stuck” on a word

A7. I feel uncomfortable when meeting new people/superiors (teachers, employers, authorities)

A8. I often feel nervous while talking. A9. Even the idea of giving a talk in public makes me feel afraid A10. I make sudden jerky or forceful movements with my head, arms or body during

speech attempts (e.g., clinching of fist or jerking head to one side)

C1. I avoid talking to people in authority (e.g., teacher, employer, or clergyman) C2. I avoid asking for information (e.g., asking for directions or inquiring about a

train schedule)

C3. I avoid choosing a job or a hobby because speaking would be required C4. I avoid making new acquaintances (e.g., not visiting with friends, not dating, or

not joining social, civic, or church groups)

C5. I avoid introducing self, giving my name, or making introductions C6. I avoid speaking situations – eg., before an audience, telephone C7. I omit a word, part of a word or a phrase planned to say (e.g., words with

certain sounds or letters)

C8. Having another person speak for me in a difficult situation (e.g., having someone make a telephone call or order food in a restaurant)

C9. I hesitate to volunteer in a discussion or debate with a group of people C10. I reply briefly using the fewest words possible. C11. I act in a manner intended to keep out of a conversation or discussion (e.g.,

being a good listener, pretending not to hear what was said, acting bored or pretending to be in deep thought)

C12. I try to give excuses to avoid talking (e.g., pretending to be tired or pretending lack of interest in a topic)

C13. I make my voice louder or softer when stuttering is expected

C14. I say words slowly or rapidly preceding the word on which stuttering is expected

C15. I try to look away while speaking

Dissertation Vol. VIII, 2009-10, Part- B, SLP, AIISH, Mysore

90

Benchmark for Speaker Identification Using Long Term Average Spectrum in Kannada Speaking Individuals

Jyothi S. 1 & S. R. Savithri 2

Abstract

The identification of people by their voices is a common practice in everyday life. In the last four decades, speaker recognition research has advanced a lot. The aim of the study was to generate benchmarking for speaker identification using Long Term Average Spectrum of speech in Kannada speaking individuals. Ten female Kannada speaking normal subjects in the age range of 18-25 years participated in the study. Material included two standard sentences in Kannada developed such that it embedded most of the phonemes in Kannada. Subjects were informed to speak the sentence in a normal modal voice. Samples were analyzed using Long Term Average Spectrum (LTAS) of speech of Computerized Speech Lab (CSL). From the LTAS, kurtosis and skewness were extracted and noted for each speaker. The results revealed several interesting points. Skewness and kurtosis appear to be robust when the number of subjects is limited. Its efficiency drops when the number of subjects increased. However, a 90% benchmarking was obtained for a group of 5 speakers. The results of the present study are restricted to female speakers and Kannada language. Hence generalization of the results to other languages and gender is questionable. Future studies with five speakers in other Indian languages, indirect or mobile recording and disguise conditions are warranted. Key words: speaker recognition, kurtosis, skewness, Long Term Average Spectrum (LTAS)

1poken language is the most natural way used by humans to communicate information. The speech signal conveys several types of information.

From the speech production point of view, the speech signal conveys linguistic information (e.g., message and language) and speaker information (e.g., emotional, regional, and physiological characteristics). Most of us are aware of the fact that voices of different individuals do not sound alike. This important property of speech of being speaker-dependent is what enables us to recognize a friend over a telephone. The ability of recognizing a person solely from his voice is known as speaker recognition. Hecker (1971) suggests that speaker recognition is any decision-making process that uses the speaker-dependent features of the speech signal. Hecker (1971) and Bricker and Pruzansky (1976) recognize three major methods of speaker recognition - (1) by listening (2) by visual inspection of spectrograms, and (3) by machine. More recently, with the availability of digital computers, automatic and objective methods can be devised to recognize a speaker uniquely from his voice.

Speaker identification by listening is entirely a subjective method. Hecker (1971) reported that speaker recognition by listening appears to be the most accurate and reliable method at that time. The second method of speaker recognition is based upon the visual

1e-mail: [email protected]; 2Professor of Speech Sciences, AIISH, Mysore, savithri2k@ gmail. com.

examination and comparison of the spectrograms. Kersta (1960) coined the word “voice print” in a report discussing identification of speaker by visual inspection of spectrograms and concluded this method seemed to offer good possibility. Stevens (1968) compared aural with the visual examination of spectrogram using a set of eight talkers and found that error rate for listening is 6% and for visual is 21%. These scores depended upon the talker, phonetic content and duration of the speech material. In speaker identification by machine, acoustic parameters from the signals are extracted and are analyzed by the machines. The objective methods can be further classified into (a) semi-automatic method, and (b) automatic method. In the semi-automatic method, there is extensive involvement of the examiner with the computer, whereas in the automatic method, this contact is limited. In the last four decades, speaker recognition research has advanced a lot. The applications of speaker recognition technology are quite varied and continually growing. Some commercial systems have been applied in certain domains. Speaker recognition technology makes it possible to use a person‟s voice to control the access to restricted services (automatic banking services), information (telephone access to financial transactions), or area (government or research facilities). It also allows detection of speakers, for example, voice-based information retrieval and detection of a speaker in a multiparty dialog.

S


91

There have been several studies on the choice of acoustic features in the speech recognition tasks. In these methods first and second formant frequencies (Stevens, 1971; Atal, 1972; Nolan, 1983; Hollien, 1990; Kuwabara & Sagisaka, 1995 and Lakshmi & Savithri, 2009), higher formants (Wolf, 1972), Fundamental frequency (Atkinson, 1976), F0 contour (Atal, 1972), LP coefficients (Markel & Davis, 1979; Soong, Rosenberg, Rabiner & Juang, 1985) , Cepstral Coefficients & MFCC (Atal, 1974; Fakotakis, Anastasios & Kokkinakis, 1993; Rabiner & Juang, 1993; Reynold, 1995), LTAS (Kiukaanniemi, Siponen & Mattila, 1982), Cepstrum (Luck, 1969; Atal, 1974; Furui, 1981; Li & Wrench, 1983; Higgins & Wohlford, 1986; Che & Lin, 1995; Jakkar, 2009) & glottal source parameters (Plumpe, Quatieri & Reynolds, 1999), and long-term average spectra (Hollien & Majewski, 1977 among others) have been used in the past.

Long Term Average Spectrum (LTAS) is

computed by calculating consecutive spectra across the chosen segment and then taking the average of each frequency interval of the spectra. However, it may be unstable for short segments (Pittam & Rintel, 1996). A range of factors have been correlated or found to be important in speaker recognition. These are all related to the original set of indices that was defined by Abercrombie (1967). The features presented include the speaker‟s gender, age, and regional or foreign accent. In addition, other factors not related to the voice production impact upon the listeners‟ ability to detect speaker identity. These include retention interval, sample duration and speaker familiarity. Further, acoustic features that are immediately available from the voice signal can be used to separate speakers. These include LTAS, fundamental frequency and formant transitions.

Hollien and Majewski (1977) concluded that n-dimensional Euclidian distance among long-term speech spectra (LTS) can be utilized as criteria for speaker identification at least under laboratory conditions. Its power as identification tool is somewhat language dependent. The LTS technique constitutes a reasonable robust tool in the laboratory but its efficiency is quickly reduced when distorting effect of the type found in more realistic environment impinge on the process. It has been argued to be effective in speaker discrimination processes (Hollien & Majewski, 1977; Doherty & Hollien, 1978; Kiukaanniemi, Siponen & Mattila, 1982; Hollien, 2002). It has however, also been argued to display voice quality differences (Hollien, 2002; Tanner, Roy, Ash & Buder, 2005), to successfully differentiate between genders (Mendoza, Valencia, Muñoz & Trujillo, 1996) and to display talker ethnicity (Pittam & Rintel, 1996).

The advantage of LTAS from a forensic perspective is that it has more or less direct physical interpretation, relating to the location of the vocal tract resonances. This makes LTAS more justified as evidence than Mel Frequency Cepstral Coefficients (MFCC). LTAS vectors of the questioned speech sample and the suspect‟s speech sample can be plotted on top of each other for visual verification of the degree of similarity. The advantages of LTAS from automatic speaker recognition perspective would be simple implementation and computational efficiency. In particular, there is no separate training phase included; the extracted LTAS vector will be used as the speaker model directly and matched with the test utterance LTAS using a distance measure. In view of this, and in view of the lack of benchmark of LTAS for Kannada speakers, the present study was undertaken. The aim of the study was to generate benchmarking for speaker identification using Long Term Average Spectrum of speech in Kannada speaking individuals. Specifically, skewness and kurtosis were extracted from LTAS for which the percent correct identifications were determined.

Method

Subjects: Ten female Kannada speaking normal subjects participated in the study. The subjects were in the age range of 18-25 years. They had passed at least 10th standard and all speakers belonged to the same dialect. The inclusion criteria of subjects were (a) no history of speech, language and hearing problem (b) normal oral structures and (c) no other associated psychological and neurological problems.

Material: Two standard sentences in Kannada formed the material. These sentences were developed such that it embedded most of the phonemes in Kannada. The sentences were written on a separate card. The sentences are given below.

Namma u:ru karnataka ra:dzjadalliruva shivamogga dzilleja chikkada:da thirthahalli. illi dzo:ga dzalapa:thavu bahu rabasava:gi entunu:ra ippathombattu adi etharadinda dhumukuttade.

Recording procedure: The testing was done in a laboratory condition. Speech samples were collected individually. The sentences were presented visually to the participants. Subjects were informed about the nature of the study and were instructed to speak the sentence in a normal modal voice. Four repetitions of the sentences were recorded. Thus forty samples were recorded from 10 speakers. The recordings were done using Computerized Speech Lab CSL Model 4500 software (Kay Pentax, New Jersey). All these were recorded on a computer memory using a 12-bit A/D


92

(Analog to Digital) converter at a sampling frequency of 16,000 Hz.

Acoustic analysis: The pauses and noises were edited from the sample using Adobe Audition software (version 2.00, Syntrillium Software Corporation). All the four recordings of each subject were stored in separate folders. Long Term Average Spectrum (LTAS) of CSL was used to analyze the samples. A Hamming window with a Nyquist frequency sampling, and pre-emphasis of 0.8 was used to extract LTAS. Figure 1 shows the waveform and LTAS from a speech sample.

Figure 1. Waveform (upper window) and LTAS (lower window) of a speech sample.

From the LTAS, kurtosis and skewness were extracted and noted for each speaker. The data was normalized using the following formula.

In this study all the voice samples were contemporary, as all the four recordings were carried out in one sitting. Closed-set speaker identification tasks were performed, in which the examiner was aware that the “unknown” speaker was among the “known” ones. The speakers recorded in first and second trails were considered as “known” and those done in thirds and fourth trials were considered as “unknown” speakers. All the “known” speakers were numbered from KS1 to 10 and corresponding “unknown” speakers were numbered as US1 to 10 For example, speaker KS1 (known) and speaker US1 (unknown) represent the same speaker in different trials of recording.

Two conditions were considered. In the first condition, one “unknown” speaker was compared with all the ten “known” speakers. An illustration is provided in the Table 1.

Table 1. Unknown speaker (speaker 1) is compared with ten known speakers on skewness and kurtosis

The Euclidean distance was calculated in Microsoft Excel. Euclidean Distance is the most common use of distance. Euclidean distance or simply 'distance' examines the root of square differences between coordinates of a pair of objects. The formula to calculate Euclidean distance was as follows: Euclidean distance = √(x2-x1)

2 + (y2-y1)2

where X and Y, in this study, refer to skewness and kurtosis. In Table 2, Euclidian distance is least for KS4. Therefore, US1 is likely to be KS4.

In the second condition, all the ten speakers

were grouped into two sub-groups of five speakers. Only five speakers were considered in each group and one “unknown” speaker was compared with all the five “known” speakers. For example, in Table 3, the least Euclidian distance is for KS4. Therefore, it implies that US1 is likely to be KS4.

The graphs were plotted with skewness on the

horizontal axis and kurtosis on vertical axis for group of different number of speakers. The unknown speaker was compared with the known speakers. Positive and negative speaker identifications were based on the Euclidian distance between the unknown and the known speakers. If the distance between unknown speaker and the respective known speaker was less, then speaker identification was deemed to be correct; if the distance between unknown speaker and any other known speaker was less, then speaker was deemed to be falsely identified or not correctly identified. The percentage correct identification was calculated by using the following formula:

The mean and SD of skewness and kurtosis were calculated.

Speaker Unknown speaker Known speakers Skewness Kurtosis Skewness Kurtosis

KS1 0.180 0.299 0.318 0.437 KS2 0.397 0.336 KS3 0.453 0.486 KS4 0.247 0.320 KS5 0.445 0.414 KS6 0.114 0.162 KS7 1 1 KS8 0 0 KS9 0.858 0.805 KS10 0.542 0.567


93

Table 2. Euclidian distances for US1 with KS1-KS10

.

Table 3. Euclidian distances for US1 with KS1- KS5

Table 4. Mean and Standard Deviation (SD) of normalized skewness and kurtosis

Table 5. Unknown speaker (US1) is compared with ten known speakers and is identified with KS4 (false

identification)

Unknown speaker Skewness Kurtosis Skewness Kurtosis Known speakers Euclidean distance US1 0.180 0.30 0.318 0.437 KS1 0.195

0.397 0.336 KS2 0.220 0.453 0.486 KS3 0.331 0.247 0.320 KS4 0.070 0.445 0.414 KS5 0.289 0.114 0.162 KS6 0.153 1 1 KS7 1.078 0 0 KS8 0.349 0.858 0.805 KS9 0.845 0.542 0.567 KS10 0.450

Unknown speaker Skewness Kurtosis Known speakers Skewness Kurtosis Euclidean distance US1 0.180 0.3 KS1 0.318 0.437 0.195

KS2 0.397 0.336 0.220 KS3 0.453 0.486 0.331 KS4 0.247 0.320 0.070 KS5 0.445 0.414 0.289 KS6 0.114 0.162 0.153 KS7 1 1 1.078 KS8 0 0 0.349 KS9 0.858 0.805 0.845 KS10 0.542 0.567 0.450

Unknown speaker Skewness Kurtosis Known speakers Skewness Kurtosis Euclidean distance US1 0.180 0.3 KS1 0.318 0.437 0.195

KS2 0.397 0.336 0.220 KS3 0.453 0.486 0.331 KS4 0.247 0.320 0.070 KS5 0.445 0.414 0.289

Subject No.

Skewness Kurtosis Skewness Kurtosis

Trials 1,2 Trials 1,2 Trials 3,4 Trials 3,4

1. 0.180 0.299 0.318 0.437 2. 0.146 0.116 0.397 0.336 3. 0.237 0.303 0.453 0.486 4. 0.167 0.219 0.247 0.320 5. 0.192 0.195 0.445 0.414 6. 0.176 0.228 0.114 0.162 7. 1 1 1 1 8. 0 0 0 0 9. 0.551 0.543 0.858 0.805 10. 0.422 0.491 0.542 0.567 Mean 0.307 0.339 0.438 0.453

SD 0.288 0.282 0.308 0.291


94

Table 6.Unknown speaker (US6) is compared with ten known speakers and is identified with KS6 (correct identification)

Unknown speaker

Skewness Kurtosis Skewness Kurtosis Known speakers Euclidean distance

0.318 0.437 KS1 0.253 0.397 0.336 KS2 0.247 0.453 0.486 KS3 0.379 0.247 0.320 KS4 0.116 0.445 0.414 KS5 0.328

US6 0.176 0.228 0.114 0.162 KS6 0.090 1 1 KS7 1.129 0 0 KS8 0.288 0.858 0.805 KS9 0.893 0.542 0.567 KS10 0.499

Table 7. Mean (M) and Standard Deviation (SD) of skewness and kurtosis in groups of 5 subjects

S. No

Skewness Kurtosis Skewness Kurtosis S

No. Skewness Kurtosis Skewness Kurtosis

Trials 1,2 Trials 1,2 Trials 3,4 Trials 3,4 Trials 1,2 Trials 1,2 Trials 3,4 Trials 3,4 1) 0.180 0.230 0.318 0.437 1) 0.176 0.228 0.114 0.162 2) 0.146 0.116 0.397 0.336 2) 1 1 1 1 3) 0.237 0.303 0.453 0.486 3) 0 0 0 0 4) 0.167 0.219 0.247 0.320 4) 0.551 0.543 0.858 0.805 5) 0.192 0.195 0.445 0.414 5) 0.422 0.491 0.542 0.567 M 0.185 0.227 0.372 0.399 0.430 0.452 0.503 0.507 SD 0.034 0.078 0.088 0.070 0.384 0.376 0.441 0.422

Table 8. Unknown speaker US5 is compared with ten known speakers and is identified with KS5 (Correct identification)

Unknown speaker

Skewness Kurtosis Skewness Kurtosis Known speakers

Euclidean distance

0.114 0.162 KS1 0.450 1 1 KS2 0.770 0 0 KS3 0.647 0.858 0.805 KS4 0.537

US5 0.422 0.491 0.542 0.567 KS5 0.142

Table 9. Unknown speaker US4 is compared with ten known speakers and is identified with KS5 (false identification)

Unknown speaker

Skewness Kurtosis Skewness Kurtosis Known speakers

Euclidean distance

0.114 0.162 KS1 0.580 1 1 KS2 0.640 0 0 KS3 0.774

US4 0.551 0.543 0.858 0.805 KS4 0.403 0.542 0.567 KS5 0.025


95

Figure 2. False identification of US1 as KS4 in a group of ten speakers.

Figure 3. Correct identification of US6 as KS6 in a group of ten speakers.

Results

Condition I: The data showed high variations in skewness and kurtosis. Subject 8 had normalized skewness and kurtosis of „0‟ and subjects 7 had normalized skewness and kurtosis of „1”. Table 4 shows the mean and SD of normalized skewness and kurtosis in ten subjects across trials. Tables 5 and 6 and figures 2 and 3 show the Euclidian distances and correct/ false identification, respectively. The overall percentage of the correct responses was found to be only 30%.

Condition II: The results indicated variability among subjects. Mean kurtosis was higher than mean skewness. Table 7 shows the mean (M) and Standard Deviation (SD) of skewness and kurtosis in groups of 5 subjects. Table 8 and 9 shows the Euclidian distances and figures 4 and 5 show an example of correct and false identification. The percentage of the correct identification was 90%. To summarize, in the first

Figure 4. Correct identification of US5 as KS5 in a group of five speakers.

Figure 5. False identification of US4 as KS5 in a group of five speakers.

condition, the correct identification was 30% and in the second condition it was 90%.

Discussion

The results supports the earlier studies in that the percent correct identification reduced with increase in the number of subjects. Hollien and Majewski (1977) found 100% and 88% identification using LTAS in normal speech in full band and limited band conditions. However, Hollien and Majewski (1977) used the power spectra but the present study used skewness and kurtosis extracted from LTAS. Skewness and kurtosis appear to be robust when the number of subjects is limited to 5. Its efficiency drops when the number of subjects increased. However, a 90% benchmarking was obtained for a group of 5 speakers.

It appeared that some speakers were very distinct (subjects 7, 8) and others were not. Because of


96

subjects 7 and 8 who had a skewness and kurtosis 1 and 0, all the Euclidian distances were affected. However, removing these subjects resulted in poorer percent identifications.

The results of the present study are restricted to female speakers and Kannada language. Hence generalization of the results to other languages and gender is questionable. Future studies with five speakers in other Indian languages, indirect or mobile recording and disguise conditions are warranted.

Acknowledgements

The authors wish to express their gratitude to Dr. Vijayalakshmi Basavaraj, Director, AIISH for granting permission to carry out this study. They also thank all the subjects for their cooperation for the study.

References

Abercrombie, D. (1967). Elements of general phonetics. Edinburgh: University Press.

Atal, B. S. (1972). Automatic speaker recognition based on pitch contours. Journal of the Acoustical Society of America, 52, 1687-1697.

Atal, B. S. (1974). Effectiveness of linear prediction characteristics of the speech wave for automatic speaker identification and verification. Journal of the Acoustical Society of America, 55, 1304-1312.

Atikinson, E. J. (1976). Inter and Intra Speaker variability in Fundamental voice frequency. Journal of the Acoustical Society of America, 440-445.

Bricker, P. D., & Pruzansky, S. (1976). Speaker recognition. In N. J. Lass (Eds.), Contemporary issues in experimental phonetics. (pp.295-326). New York: Academic Press.

Che, C., & Lin, Q. (1995). Speaker recognition using HMM with experiments on the YOHO database. In Eurospeech, 625-628.

Doherty, E. T., & Hollien, H. (1978). Multiple-factor speaker identification of normal and distorted speech. Journal of Phonetics, 6, 1 – 8.

Fakotakis, N., Anastasios, T., & Kokkinakis, G. (1993). A text independent Speaker recognition system based on vowel spotting. Speech Communication, 57-68.

Furui, S. (1981). Cepstral analysis technique for automatic speaker verification. IEEE Transactions on Acoustics, Speech and signal Processing, 29, 254-272.

Hecker, M. H. L. (1971). Speaker recognition: basic considerations and methodology. Journal of Acoustical Society of America, 49, 138.

Higgins, A., & Wohlford, R. E. (1986). A new method of text independent speaker recognition. In International Conference on Acoustics, Speech and Signal processing in Tokyo, IEEE, 869-872.

Hollien, H. (1990). The acoustics of crime. The New Science of Forensic Phonetics, Plenum, Nueva York.

Hollien, H. (2002). Forensic voice identification. San Diego, CA: Academic.

Hollien, H., & Majewski, W. (1977). Speaker identification by long-term spectra under normal and distorted speech conditions. Journal of the Acoustical Society of America, 62, 975 – 980.

Jakkar, S. S. (2009). Bench mark for speaker identification using Cepstrum. Unpublished project of Post graduate Diploma in Forensic Speech Science and Technology submitted to University of Mysore, Mysore.

Kersta, L. G. (1960). Voiceprint identification. Nature, 196, 1253–1257.

Kiukaanniemi, H., Siponen, P., & Mattila, P. (1982), Individual differences in the long term speech spectrum. Speech Communication, 21-28.

Kuwabara, H., & Sagisaka, Y. (1995). Acoustic characteristics of speaker individuality: control and conversion. Speech Communication, 16, 165-173.

Lakshmi, P., & Savithri, S. R (2009). Benchmark for speaker identification using vector F1 & F2. Proceedings of the international symposium, Frontiers of Research on Speech & Music, FRSM-2009, 15-19.

Li , K. P., & Wrench, E. H. (1983). Text independent speaker recognition with short utterances. In international Conference on Acoustics, Speech and Signal Processing in Boston, IEEE, 555-558.

Luck, J. E. (1969). Automatic speaker verification using cepstral measurements. Journal of the Acoustical Society of America, 46, 1026-1032.

Markel, J. D., & Davis, S. B. (1979). Text independent speaker recognition from a large linguistically unconstrained time spaced data base, IEEE Transactions on Acoustics. Speech and Signal Processing ASSP-27, 74-82.

Mendoza, E., Valencia, N., Muñoz, J., & Trujillo, H. (1996). Differences in voice quality between men and women: Use of the long-term average spectrum (LTAS). Journal of Voice, 10, 59 – 66.

Nolan, F. (1983). Phonetic bases of speaker recognition. Cambridge, Cambridge university.

Pittam, J., & Rintel, E. S. (1996). The acoustics of voice and ethnic identity. In P. McCormack & A. Russell, (Eds.), Proceedings of the sixth Australian International Conference on Speech Science and Technology (pp. 115 – 120). Adelaide, Australia: Australian Speech Science and Technology Association.

Plumpe, M. D., Quatieri, T F., & Reynolds, D. A. (1999). Modeling of the glottal flow derivative waveform with application to speaker identification. IEEE Trans. on Speech and Audio Processing, 7(5), 569– 586.

Rabiner, L., & Juang, B.H. (1993), Fundamentals of speech recognition, Prentice Hall PTR.

Reynold, D.A. (1995). Speaker identification and verification using Gaussian mixture speaker models, Speech Communication, 17, 91-108.

Soong, F., Rosenberg, A. E., Rabiner, L., & Juang, B.H. (1985). A vector quantisation approach to speaker recognition. In International Conference on Acoustics, Speech and Signal Processing in Florida, IEEE, 387-390.

Stevens, K. N. (1968). Speaker authentication and identification: A comparison of spectrographic and auditory presentations of speech material. Journal of the Acoustical Society of America, 44, 1596–1607.


97

Stevens, K. N. (1971). Sources of inter and intra speaker variability in the acoustic properties of speech sounds, Proceedings 7th International Congress. Phonetic Science, Montreal, 206-227.

Tanner, K., Roy, N., Ash, A., & Buder, E. H. (2005). Spectral moments of the long-term average spectrum: Sensitive

indices of voice change after therapy? Journal of Voice, 19, 211 – 222.

Wolf, J. J. (1972). Efficient acoustic parameter for speaker recognition. Journal of the Acoustical Society of America, 2044–2056.


98

Cognitive-Linguistic Assessment Protocol in Telugu - An Adaptation of CLAP in Kannada

Didi Karthika Veena 1 & S. P. Goswami 2

Abstract

The current study mainly aimed at adapting the Cognitive Linguistic Assessment Protocol for adults (CLAP) - Kannada (Kamath, 2001) into Telugu language (CLAP-T). Normal elderly persons in the age range of 65-80 years were considered for the study. The participants were classified based on the classification given by Nuegarten (1974); young-old, 75-85 as old-old, above 85 years as oldest-old. All male and female participants performed almost equal mean and SD values in all domains. Few subtests showed variations in performance of participants in three age groups. All the four domains showed a general trend where in, as the age increases there is degradation in cognitive performance. Reaction time for the visual task was considered and showed a delineating performance across three age groups. This assessment tool will help the professionals to find out cognitive-linguistic performance of Telugu speaking individuals across different age, gender and four domains i.e. attention, perception and discrimination, memory, problem solving and organization. Key words: cognitive linguistic domain, assessment, elderly

ognition 1means „to know‟. Knowledge can be thought of as memories formed from the manipulation and assimilation of raw

input, perceived via our senses of sight, hearing, taste, touch and smell. Using knowledge to direct and adapt action towards goals is the foundation of the cognitive process. Past experiences and trends inform our sense of what the future might hold and help us to act accordingly (Kellett, 2008).

Importance of cognitive-linguistic tools Cognitive tools are generalizable computer or

screening tools that are intended to engage and facilitate cognitive processing. Cognitive tools help to screen the complex cognitive learning activities and critical thinking. These tools are person‟s controlled in the sense that they construct their knowledge themselves using the tools rather than memorizing knowledge. Cognitive tools are medium that helps transcend the limitations of the mind, such as memory, in activities of thinking, learning, and problem solving (Dimok, 1999).

Studies in Indian context are restricted and one such study by Kamath (2001) in Kannada, a classical language spoken in Karnataka, resulted in Cognitive-linguistic Assessment Protocol (CLAP) for adults. Nevertheless, assessment of persons from varied ethnic, cultural, and linguistic backgrounds are the current significant challenges for clinicians, because most of the tests have not comprised a representative number of people from diverse backgrounds in their consistency. Therefore, these all variables should be considered throughout assessment procedure.

The current study mainly aimed at adapting the Cognitive Linguistic Assessment Protocol for adults (CLAP)-Kannada (Kamath, 2001) into Telugu language (CLAP-T). The specific objectives were (1)

1e-mail:[email protected]; 2Reader in Speech Pathology, AIISH, Mysore, [email protected].

to investigate age related changes in performances on various cognitive-linguistic skills on the CLAP-T, (2) study gender contingent variations in performances on different cognitive-linguistic skills on the test (3) to estimate reaction time measures of the participants (4) to determine the effect of age and gender on various domains of the test (viz. Attention, perception and discrimination; Memory; Problem solving; and Organization) and (5) to find out the interaction effect of age and gender on the domains.

Method

Participants: Normal elderly persons in the age range of 65-80 years were considered for the study. The participants were classified based on the classification given by Nuegarten (1974): 55-75 as young-old; 75-85 as old-old; above 85 years as oldest-old. The existing classification system given by Nuegarten (1974) was modified by adding two more groups as: 65-75 years: young-old adults, 75-80 as old-old adults.

Age groups: A total of 60 Telugu speaking participants (30 males and 30 females) were included in the study. The participants were divided in three age groups, which is shown in Table 1. Equal number of males and females were considered for the study.

Table 1. Demographic data of subjects sampled Age Number of

Males Number of Females

65-70 10 10 70-75 10 10 75-80 10 10

The participants were selected by adhering to

the appropriate ethical procedures. Participants were explained the purpose and procedures of the study, and informed verbal and or written consent was taken.

C

CLAP in Telugu

99

Inclusionary criteria considered for selection of participants were as follows: (1) the participant should not have any obvious or known history of neurological and/or psychological disorders as reported (2) he/she should be able to speak, read and write in Telugu, and (3) he/she should have a minimum of primary schooling. The exclusionary criteria were as follows: 1) No noticeable sensory deficits, (2) No history of drug/alcohol abuse as reported, (3) No difficulty in communication.

Mini Mental State Examination-MMSE (Folstein, Folstein & McHugh, 1975) was used for screening. All participants scoring greater than or equal to 25 points on the MMSE (Folstein, 1975) were considered for the study. Participants were selected from the residential areas and old age homes in the city of Hyderabad (Andra Pradesh state). Procedure: The study was carried out in 3 phases Phase 1: The existing CLAP in Kannada (Kamath, 2001) was translated into Telugu. The material was prepared by reviewing the literature in Telugu from books, journals and web-based sources. Preparation of the material: Adaptation of CLAP-T test battery based on the original test item of CLAP in Kannada (Kamath, 2001) was carried out. The steps involved in the adaptation of the test were (a) syntactic structure and semanticity in Telugu language was studied and compared to Kannada (b) the commonalities and differences were studied to choose the appropriate test items (c) the test item in Telugu version was seen for semanticity, familiarity and ambiguity (d) then the Kannada version was translated to Telugu language after suitable modification (e) the translation was done with the help of a linguist and speech-language pathologist (f) materials for various test items were prepared in Telugu. Phase 2: The translated materials were given to five experienced Speech-Language Pathologist (SLP) who served as judges. SLP‟s were asked to rate the material on a “Feedback questionnaire for Aphasia Treatment Manuals” (Field Testing of Manual for Adult Non-fluent Aphasia Therapy in Kannada-MANAT -K, Goswami, Shanbal, Samasthitha & Navitha, 2010), which included 17 parameters rating from very poor to excellent. The three parameters excluded for judgment were volume, size of the picture and color of the picture. Judges were also asked to give suggestions regarding the rating scale used and any other changes in the protocol. Phase 3: The final test item was modified by incorporating the suggestions of SLP‟s and administered on 60 participants in the age range of 65-80 years. Instructions specific to the task were given in Telugu. The scoring was carried out simultaneously for each task as per the scoring procedure scheduled for each item.

Scoring: Various test items and their respective scoring under each domain was same as stated in the CLAP- Kannada version (Kamath, 2001). Domain I: Attention, Perception and Discrimination

Two main modes were considered for this domain viz. visual and auditory. Two types of attention processes were evaluated, viz. selective attention and sustained attention. The cognitive processes of perception and discrimination were contingent on attention and three were parallel processes.

Table 2. Cognitive-linguistic measures used in the protocol

Source: Cognitive-linguistic Assessment Protocol for adults (Kamath, 2001) Visual category included (a) The cancellation at a letter level (b) The cancellation at word level (c) Contingent cancellation

Domain Test item Max. score

I: Attention, Discrimination and Perception Visual

Auditory

Letter Cancellation Contingent Letter Cancellation Word Cancellation Sound count Letter-Pair discrimination Word –Pair discrimination Months-backwards naming

10 10 10 10 5 5 10

II: Memory Episodic Memory

Orientation and recent memory questions

0

Working memory

Digit Forward Digit backward

5 5

5Semantic memory

Co-ordinate naming Super ordinate naming Word-naming fluency Generative naming Sentence repetition Carry out commands

5 5 5 5 10 10

III: Problem solving

Sentence disambiguation Sentence formulation Predicting outcome Compare and contrast Predicting cause Why Questions Sequential Analysis

10 5 10 10 10 5 10

IV: Organization

Categorization Analogies Sequencing events

10 10 40


100

Auditory category (a) Sustained auditory attention was evaluated

using a task where the participants were required to count mentally how many times a particular letter was read out in a list.

(b) & (c) The second and third subtests involved the participant‟s ability to discriminate amongst a pair of letters and words read out by the tester (same/different task) respectively. This is a predominantly a discrimination task.

(d) The last sub test required the participants the names of the months in the backward direction (i.e. December to January). This task requires the attention and involves recall processes too.

Scoring: Performance on each task was scored for number of correct answers. Subtest (a) was given one point for every correct count. One point was deducted for every count above the correct number of total occurrence of stipulate letters. Domain II: Memory

Three main types of memory processes tested were (1) Episodic memory (2) Working memory and (3) Semantic memory Domain III : Problem solving

This domain includes various tests that assess reasoning abilities to aid in problem solving. The following tasks considered were Sentence disambiguation, Sentence formulation, Comparing and contrasting two objects, Predicting cause of a described situation, Answering why questions, Sequential task analysis. Domain IV: Organization

Organization of available information to result in coherent communication is an important metalinguistic and metacognitive task. This included Categorization Analogies and Sequencing of events.

Statistical analysis was carried out using

(SPSS 16). MANOVA was administered, to check for any significant difference across age in all male and female participants. Post hoc Duncan‟s test was administered.


The results obtained from the data were analyzed on various aspects. The findings of the present study have been broadly presented and discussed under the following headings.

I. (a) Performance of males and females in different age groups in attention

In this section visual and auditory, two modalities were assessed. The following were the various tasks in visual and auditory modes. The visual mode included Letter Cancellation (AVLC), Contingent Letter Cancellation (ACLC); Word

Cancellation (AVWC). The auditory mode Sound count (AASC); Letter-Pair discrimination (AALPD); Word-Pair discrimination (AAWPD); Months-backwards naming (AAMBN).

The mean and standard deviation scores of

three test items attention, visual letter cancellation (AVLC), Attention visual contingent letter cancellation (AVCLC), attention visual word cancellation (AVWC) in domain I.

Both male and female participants had equal

mean scores on AVLC and AVWC which required sustained attention. On the other hand, correspondingly equal scores were obtained by both the male and female participants in AVCLC test item (rely on selective attention) across the three age groups. In general, although the age increased the performance of both the male and female participants were equal in all the above test-items.

To summarize apart from the AAMBN test

item, the other test items (AASC, AALPD, and AAWPD) show equal mean scores on performance across age and gender.

I. (b) Performance of males and females in different age groups in memory across gender

The following test items were considered in this domain: Orientation and recent memory questions (EMOQ); Digit Forward (WMDF); Digit backward (WMDB); Co-ordinate naming (SMCN); Super ordinate naming (SMON); Word-naming fluency (SMWN); Generative naming(SMGN); Sentence repetition(SMSR); Carry out commands (SMCC). Working Memory Digit Forward (WMDF)

A mean score ranging between 3.70 and 3.60 was secured by the participants in the age range of 65-70 years. An almost equal performance was contributed by all the participants in the age group of 70-75 with a mean value of 3.00 and 3.10 respectively, in the age group of 75-80years who scored mean values of 2.50 and 2.20. Finally, both the male and female participants achieved a mean score of 3.06 (SD=0.73) and 2.96 (SD=0.76) respectively across three age groups. The total mean values for all the participants was 3.01 (SD=0.74) for the test item WMDF in domain II. Working Memory Digit Backward (WMDB)

In this test item, the male and female participants of age group 65-70years secured a mean value of 2.80 and 2.60 with comparatively better mean score value of 2.00 (SD=0.47) and 2.30 (SD=0.48) by the male and female participants in the age group of 70-75 years. A lowest mean score value of 1.90 (SD=0.56) and 1.70 (SD=0.48) was secured by the male and female participants in the age group

CLAP in Telugu

101

of 75-80 years. A total mean score was 2.23 (SD=0.62) and 2.20 (SD=0.61) in all male and female participants of the three age groups and a total mean score of 2.21 with (SD=0.61) for the domain II in WMDB.

Semantic memory Co-ordinate naming (SMCN), Semantic Memory Super Ordinate Naming (SMSN), Semantic memory Word Naming Fluency (SMWNF), Semantic memory Generative Naming (SMGN)

In exception to the above test items in the domain II, all the participants irrespective of age and gender got a maximum score of 5.00 for the test item in, SMCN, SMSN, SMWNF, and SMGN. Semantic memory Sentence Repetition (SMSR) and Semantic memory carry out Commands (SMCC)

The same trend was repeated in the next two test items, SMSR and SMCC where the participants showed a complete score of 10.00 across the three age groups and gender. I. (c) Performance of males and females in different age groups in Problem solving across gender

The following test items were considered in this domain of problem solving. They were Sentence disambiguation (PSSD); Sentence formulation (PSSF); Predicting outcome (PSPO); Compare and contrast (PSCAC); Predicting cause (PSPC); Why Questions (PSWQ) and Sequential Analysis (PSSA).

Problem Solving Compare and Contrast (PSCAC) A mean value of 9.90 (SD=0.31) and 9.50 (SD=0.70) was scored by the male and female participants of age group 65-70years. However, the participants in the age group 70-75 years had a mean of 9.20 (SD=0.78) and 9.20 (SD=0.63) in males and females respectively. The performance of male and female participants was shown in the age group of 75 -80 years with a mean score of 9.30 (SD=0.82) and 8.70 (SD=0.94). A score of 9.46 (SD=0.73) and 9.13 (SD=0.81) was obtained by all male and female participants of the three age groups and a total mean score of 9.30 (SD=0.78) for the domain III, PSCAC was observed in all the participants.

Table 3. Mean and SD values for male and female participants (normals) in Domain III

*PSCAC=problem solving compare and contrast

Table 4. Mean (M) and standard deviation (SD) values of male and female participants (normals)

*AVLCRT=attention visual letter cancellation reaction time; AVCLCRT=Attention visual contingent reaction time letter cancellation; AVWCRT=attention visual word cancellation reaction time. I. (d) Performance of males and females in different age groups in organization across gender

The following test items were considered in this domain: Organization, Categorization (OC), Analogies (OA) and Sequencing events (OSE) Organization Sequential Events (OSE)

In OSE, the primary age group explicitly 65-70 years attained a mean value of 35.80 (SD=1.47) and 35.20 (SD=1.03) in males and females correspondingly. On the other side, the subsequent participants in the group 70-75 years confirmed a mean 32.60 (SD=1.89) and 32.8 (SD=1.03) in males and females. At the same instance, the supplementary participants of age group 75-80 years attained a mean of 26.20 and 23.40 in males and females. A score of 31.53 (SD=4.53) and 30.46 (SD=5.45) was acquired by all male and female participants of the three age groups and an entire mean score of 31.00 with (SD=5.00) intended for the domain IV, OSE was experiential in all the participants. II. Reaction time of males and females across the only first three test items in the domain of Attention, Perception and Discrimination, in different age groups a) Attention visual letter cancellation reaction time (AVLCRT)

Mean scores (19.00 and 18.20) for reaction time of male and female participants in the age range of 65-70years. Both the male and female participants in the age group of 70- 75 years attained an equal mean scores (23.50 and 23.50) for reaction time. The following age group i.e.75-80years of males and females participants achieved a means score of 24.20 and 25.30 for reaction time respectively. The total mean score for reaction time 22.23 and 22.33 was attained by all males and females correspondingly.

Males Females Total Domain

III * Group Mean SD Mean SD Mean SD

PSCAC

65-70 9.90 0.31 9.50 0.70 9.70 0.57 70-75 9.20 0.78 9.20 0.63 9.20 0.69 75-80 9.30 0.82 8.70 0.94 9.00 0.91

TOTAL 9.46 0.73 9.13 0.81 9.30 0.78

Domain I* Groups Males Females Total


AVLCRT

65-70 70-75 75-80

TOTAL

19.00 23.50 24.20 22.23

1.15 0.84 0.78 2.51

18.20 23.50 25.30 22.33

0.42 0.70 1.25 3.17

18.60 23.50 24.75 22.28

0.94 0.76 1.16 2.84

AVCLCRT

65-70 70-75 75-80

TOTAL

21.00 29.70 34.50 28.40

2.94 0.67 1.17 5.96

19.50 30.30 33.80 27.86

0.97 0.67 0.91 6.24

20.25 30.00 34.15 28.13

2.26 0.72 1.08 6.06

AVWCRT

65-70 70-75 75-80

TOTAL

23.60 35.30 39.10 32.66

1.64 0.94 0.99 6.81

22.50 34.50 39.50 32.16

1.43 0.70 0.70 7.32

23.05 34.90 39.30 32.41

1.605 0.911 0.86 7.016


102

All participants acquired a total mean score for reaction time was 22.28 (SD=2.84). b) Attention visual contingent letter cancellation reaction time (AVCLCRT)

The male and female participants in the age group of 65-70years attained almost equal mean score (21.00 and 19.50) for reaction time correspondingly. The other side, following age group 70-75 years, male and female participants acquired a mean score (29.70 and 30.00) for reaction time respectively. The third age group that is 75-80years of male and female participants showed a mean score (34.50 and 33.80) for reaction time. The total mean score of 28.40 and 27.86 for reaction time was achieved by all female and male participants respectively. Overall mean scores of participants was 28.13 (SD=6.06) for reaction time. c) Attention visual word cancellation reaction time (AVWCRT)

The male and female participants of age group 65-70years achieved an equal mean score (23.60 and 22.50) for reaction time. In the age group of 70-75years male and female participants showed a mean score (35.30 and 34.50) for reaction time. A mean score of 39.10 showing a better performance (SD=0.99) for reaction time was obtained by male participants compare to females with a mean score of 39.5 (SD=0.70). The total mean score of males and female participants was 32.66 (SD=6.81) and 32.10 (SD=7.32) respectively. Finally resulting a total mean score of 32.41 (SD=7.01) for the above mentioned test item.

To summarize it is evident from the above results that the overall performances of male and female participants w.r.t reaction time across all the three age groups showed a lower mean score values for the test item attention visual contingent letter cancellation reaction time (AVWCRT) compared to other two test items attention visual letter cancellation reaction time and attention visual word cancellation reaction time (AVLCRT) & (AVCLCRT).

III. Effect of age and gender performance on various domains Domain I-Visual mode Attention visual letter cancellation (AVLC), Attention visual contingent letter cancellation (AVCLC), Attention visual word cancellation (AVWC)

All the participants in these test items scored a full mean score valves. Hence, further statistical analysis (post hoc and Duncan‟s) was not administered. This finding was supported by the Madden (1997) in a study, where he reported young and older adults showed a equivalent performance in attention visual items, i.e. their ability to distinguish

relevant from irrelevant information, but an increase in the number of relevant display items either in the target set or in the display impairs search performance more for older adults than for young adults.

The AVCLC is a task which is based on selective attention. Result of the present study fails to find an age related decline in this particular task, this is in concurrence with the findings of Groth and Allen (2000). Attention visual letter cancellation reaction time (AVLCRT)

The male and female participants showed a significant difference for reaction time with F (2, 54) = 257.58, p<0.05 across age.

The Duncan‟s test was carried out to see the

significant difference across age, the test results showed that there was a decline in the reaction time of task performance as the age increases. In age group of 65-70 years performed with a mean valve of 18.60, 70-75years performed with mean score of 23.50 and 75-80 years with a mean of 24.70 according to mean score there was not much decline in the age group of 70-75 and 75-80years. Attention visual contingent letter cancellation reaction time (AVCLCRT)

From Table 4, it can be inferred that there is a significant difference with F (2, 54) =478.99, p<0.05 in male and female participants respectively.

Duncan‟s test was done to show the effect of AVCLCRT on each age group individually, it revealed there is a decline in the performance of participants as the age increased i.e. 65-70years showed low reaction time to perform a task with a mean of 20.25 than other age groups with a mean of 30.00 and 34.10, as the age increased they required more reaction time to finish the task. Attention visual word cancellation reaction time (AVWCRT)

A significant difference with F (2, 54) = 110.70, p<0.05 by each male and female participant across age. Following the MANOVA to see the significant difference across each age group separately, post hoc analysis and Duncan‟s test was carried out. Duncan‟s test results showed 65 -70years performed better with a mean of 23.05 compared to 70-75 and 75-80years with a mean of 34.90 and 39.30, this showed that the 75-80 years age group was away from the 65-70 and 70-75years. Gender effect

No significant gender effect was found in all participants in above mentioned test items for reaction time i.e. attention visual letter cancellation

CLAP in Telugu

103

reaction time (AVLCRT), attention visual contingent letter cancellation reaction time (AVCLCRT), attention visual word cancellation reaction time (AVWCRT). Age and gender interaction

No obvious age and gender interaction was observed among the participant on two tasks of domain I, which includes attention visual contingent letter cancellation reaction time (AVCLCRT), and attention visual word cancellation reaction time (AVWCRT). The test item Attention visual letter cancellation reaction time (AVLCRT) showed an age with gender interaction across the three age groups of all participants.

As with the advancement of age there could be a gradual degradation in fine motor skill as well as eye-hand co-ordination which are important for performing the subtle tasks like the one mentioned in the above test items. This will in turn result in longer reaction time in elderly populations. In addition the elderly participants experience difficulty in inhibiting the irrelevant (distractor) and facilitating the relevant ones (target). This is in congruence with finding of Redfern, Muller, Jennings and Furman (2002) where they reported that in the presence of distracters, older people tend to devote their exclusive attention to one stimulus, and ignore another stimulus than younger people. Literature also reports that older adults were as skilled as younger people at assimilating information, but they did take longer time to react Myerson, Robertson and Hale (2007). Attention auditory sound count (AASC), Attention auditory letter pair discrimination (AALPD), Attention auditory word pair discrimination (AAWPD)

Apart from the attention auditory month background naming (AAMBN) test item, all the test items were performed equally and achieved a full mean score by all male and female participants of three age groups, as a result further statistical analysis were not carried out for the same. Attention auditory month background naming (AAMBN)

The male and female participants showed significant difference across age with [F(2,54)= 47.63, p<0.05]. Duncan‟s test was carried out to show the performance of each age group separately which revealed that 70-75 and 75-80years showed a almost equal mean of 7.35 and 6.85 i.e. 70-75 and 75-80years showed a significant difference which was totally away from 65-70 years with a mean of 9.40, this showed that 70-75 and 75-80years performed poorer compared to 65-70 years.

Gender effect There was no significant difference among

male and female participants in below mentioned tasks [attention auditory sound count (AASC), attention auditory letter pair discrimination (AALPD), attention auditory word pair discrimination (AAWPD), attention auditory month background naming (AAMBN)]. Age and gender interaction

No interaction effect was found between age and gender in all participants of above mentioned age groups in each tasks [attention auditory sound count (AASC), attention auditory letter pair discrimination (AALPD), attention auditory word pair discrimination (AAWPD), attention auditory month background naming (AAMBN)]. Episodic memory orientation and recent memory questions (EMORMQ), Semantic memory coordinates naming (SMCN), Semantic memory subordinate naming (SMSN), Semantic memory word naming fluency (SMWNF), Semantic memory Generative naming (SMGN), Semantic memory Sentence repetition (SMSR), Semantic memory Carry out commands (SMCC)

All the above mentioned test items confirmed complete mean scores of all the male and female participants across three age groups, therefore further statistical analysis i.e. post hoc analysis and Duncan‟s test was administered only for the two test items, not intended for remaining test items. Working memory digit forward (WMDF)

The male and female participants showed a significant difference with [F (2, 54) =29.49, p<0.05] respectively using MANOVA. The Duncan‟s test was carried out to show the significant effect of above test item in each age group and the results showed a significant difference between the age range of 75-80 years with a mean of 2.35 which significantly away from 70-75 and 65-70 years with a mean score of 3.05 and 3.65 .It revealed a significant difference of 75-80 from other age groups.

Working memory digit backward (WMDB)

There was no significant difference found with [F (2, 54) = 16.96, p<0.05] across age in males and females correspondingly. The results of the post hoc analysis revealed that 75-80years showed a significant difference from 65-70years with a mean value of 1.80. There was no significant difference between 65-70 and 70-75 years with a mean valve of 2.7 0and 2.15. 65-70 and 70-75 years age group performed better compared to 75-80years.


104

Gender effect There was no significant difference across

gender in above mentioned test items in all male and female participants. Age and gender interaction

Interaction effect was not found across age and gender in any participant above mentioned test items. Literature attributes one more reason for the declined performance in digit span task to greater storage load than the executive function deficit. Though, digit span involved relatively little complex processing, the maximal verbal memory span depended on both the phonological loop and central executive. As the digit load increase, demands made on central executive also increases (Baddeley, 2001). In some cases, older adults seem to be using different strategies than younger adults. For example, brain imaging studies have revealed that older adults are more likely to use both hemispheres when completing memory tasks than younger adults (Cabeza, 2001).

That there was a significant decline in the two

test items across age group in male and female participants respectively, although domain II includes 9 test items. As all the participants in the seven test items i.e. SMSR, SMSN, SMCC, SMCN, SMWNF, EMORMQ and SMGN showed complete responses. Domain III

All participants showed a significant difference in performance across the age group on MANOVA for the test items, problem solving sentence disambiguation (PSSD), problem solving compare and contrast (PSCAC), and problem solving predicting cause (PSPC) in domain III. Gender effect

The males and female participants of all age groups showed no significant difference across gender in all test items.

Age and gender

No interaction was found between age and gender in any test item of all participants of three age groups.

Most studies of problem-solving are cross-

sectional. Thus, age is confounded with cohort factors such as education. The two longitudinal studies by Arenberg (1974, 1982) showed that age declined only appear to be significant in the 60s and 70s. It appears that two factors can account for much of the decline in problem solving performance, changes in speed of execution of elementary information processes and changes in the capacity of working memory. It is also possible that speed alone

can account for all the changes, to the extent that working memory relies on the speed of processes such as rehearsal.

Denney (1990) study reports the most general

conclusion of this study was that, older subjects universally perform less well on problem solving tasks than do younger subjects, and the decline in performance is a linear function with respect to age. Domain IV a) With age as independent variable and various

test items organization categorization (OC), organization sequencing events (OSE) as dependent variable, on the above domain showed a significant difference of [F(2,54)=33.95 for p<0.05]; [F(2,54)=163.64] for respectively.

b) Subsequent to this, post hoc Duncan‟s test revealed all three age group were significantly different from each other for all three test items. i. In organization categorization (OC): The age

group 65-70 and 70-75 years was significantly different from 75-80 years with a mean value of 10.00 and 9.90. The performance of participants in 75-80 years age group showed a significant difference with the other two age group with a mean value of 8.00.

ii. Organization analogies (OA): All the participants acquired full mean scores across three age groups, therefore further statistical analysis was not done.

iii. Organization sequencing events (OSE): The older age group 75-80 years was significantly different from 65-70 and 70-75 years with a mean value of 24.80,that is to say 65-70 and 70-75 years scored almost equal mean score valves of 32.70 and 35.50 which was away from 75-80years.

c) No significant difference was seen on statistical analysis in gender and various test items

d) Similarly no interaction effect is seen between age and gender among various test items.

Conclusions

This assessment tool will help the professionals to find out cognitive-linguistic performance of Telugu speaking individuals across different age, gender, and four domains i.e. attention, perception and discrimination, memory, problem solving and organization. The present study further corroborates the evidence to research in cognitive-linguistic performance and the factors contributing to it and opines to carry out extensive research in this area. Assessment of cognitive skills across age will give an insight into the senile versus senescence in older group. Profiling of various cognitive skills in elderly population which in turn useful in therapeutic intervention.

CLAP in Telugu

105

References American Association of Retired Persons (AARP). (2010).

In encyclopedia Britannica. Retrieved April 21, 2010, from http:// www. britannica. com/ EBchecked/ topic/19286/American-Association-of-Retired-Persons.

Anuroopa, L. (2006). Development of cognitive linguistic assessment protocol for children (CLAP-C). Unpublished Masters Dissertation, University of Mysore, Mysore, India.

Baddeley, A. (2001). Is working memory still working? American Psychologist, 56, 851-864.

Baddeley, A. (2003). Working memory: looking back and looking forward. Nature Reviews Neuroscience, 4 (10), 829-839.

Cabeza, R. (2001). Cognitive neuroscience of aging: Contributions of functional neuroimaging. Scandinavian Journal of Psychology, 42, 277-286.

Cabeza, R., Anderson, N. D., & McIntosh, A. R. (2002). Aging gracefully: Compensatory brain activity in high-performing older adults. Neuroimage, 17(3), 1394-1402.

Denney, N. W., Pearce, K. A., & Palmer, A. M. (1982). A developmental study of adults, performance on traditional and practical problem solving tasks. Experimental Aging Research, 8, 115 -118.

Goswami, S. P., Shanbal, J. C., Samasthitha, S, & Navitha, U. (2010). Field testing of Manual for Adult Non-fluent Aphasia Therapy in Kannada. (MANAT-K).Research Project, All India Institute of Speech and Hearing Research Fund, Mysore, India.

Gregoire, J., & Van der Linden, M. (1997). Effects of aging on forward and backward digit span. Aging, Neuropsychology and Cognition, 4,140-149.

Gunning-Dixon, F. M., & Raz, N. (2000). The cognitive correlates of white matter abnormalities in normal

aging: A quantitative review. Neuropsychology, 14(2), 224-232.

Gunning-Dixon, F. M., & Raz, N. (2003). Neuroanatomical correlates of selected executive functions in middle-aged and older adults: A prospective MRI study. Neuropsychologia, 41(14), 1929-1941.

Kamath, A. (2001), Cognitive Linguistic Assessment Protocol for Adults. Unpublished Masters Dissertation, University of Mysore, Mysore, India.

Kavya, V. (2007). Development of cognitive linguistic assessment protocol for children with learning disability. Unpublished Masters Dissertation, University of Mysore, Mysore, India.

Kennard, C., (2005). The aging brain. Retrieved March 2010,fromhttp://alzheimers.about.com/od/research/a/aging_brain.htm.

Kumar, V. (2007). Cognitive linguistic flexibility and aging. Unpublished Masters Dissertation, University of Mysore, Mysore, India.

Madden, D. J., Turkington, J. M., & Provenzale, L. L. (1999). Adult age differences in the functional neuroanatomy of verbal recognition memory. Human Brain Mapping, 7, 115-135.

Mayerson, J. S., Robertson., & Hale, S. (2007). Aging and intra individual variability in performance: Analysis of response time distributions. Journal of the Experimental Analysis of Behaviour, 88(3),319-337

Nuegarten, B. (1974). The aging central nervous system: clinival aspects. In H. Brody, D.Harman & S. M. Ordy (Eds.), Aging. (pp 1-10). New York: Roven Press.

Redfern, M. S., Muller, J. R., Jennings, J. M., & Furman (2002). Attentional dynamics in postural control during perturbations in young and older adults. The Journal of Gerontology Series A, 57 (8), B 298.


106

Cognitive-Linguistic Assessment Protocol in Malayalam: An Adaptation of CLAP in Kannada

Lakshmi S. Mohan 1 & S. P. Goswami 2

Abstract

The current study aimed at adapting the Cognitive-Linguistic Assessment Protocol for Adults (CLAP) - Kannada, (Kamath, 2001) to Malayalam speaking elderly population. Normal elderly persons in the age range of 65-80 years were considered for the study. The modified version of the questionnaire was prepared and given to five speech- language pathologists for judging the content semanticity and other aspects of the language using a feedback questionnaire rating scale. The modified version, CLAP-M was administered on participants across the four domains, which were Attention Perception and Discrimination, Memory, Problem-solving, and Organization. Statistical analysis revealed a significant difference in all the four domains as the age advances. A general trend of declined cognitive performance was noted across the various test items in each domain. All test items in domain IV were found to be sensitive to depict a significant cognitive decline as the age advances. No significant gender effect was seen across the domains as the age progresses.

Key words: cognition, attention, memory problem-solving, organization

ognition 1can be defined as the processes an organism uses to organize information. This includes acquiring information (perception),

selecting (attention), representing (understanding) and retaining (memory) information, and using it to guide behavior (reasoning and coordination of motor outputs). Intervention to improve cognitive function may be directed at any one of these core faculties (Bostrom & Sandberg, 2009).

Cognitive process is the main source which is very important in every human being. Most of the studies in aging and cognition points to the existence of a direct relationship between the two. As aging is a universal process, it affects various cognitive processes of elderly population irrespective of language. Aging has shown a major impact on communication skills too.

Aging can also be defined as a progressive functional decline or a gradual deterioration of physiological function with age, including a decrease in fecundity (Partridge & Mangel, 1999), or the intrinsic, inevitable, and irreversible age-related process of loss of viability and increase in vulnerability (Comfort, 1964).

Many older people may lose their independence and quality of life because of unrecognized and untreated chronic disabilities such as visual impairment, hearing impairment, cognitive impairment and physical immobility. The three prevalent chronic conditions of old age are arthritis, hypertension and hearing impairment (Weinstein, 2003). The question whether aging affects language production in only

1e-mail: lachu3692gmail.com; 2Reader in Speech Pathology, AIISH, Mysore, [email protected].

specific functions or across the board is important for evaluating two different types of theories of cognitive aging: information universal and information-specific theories which focus on word retrieval. Changes in language production in old age have a practical as well as theoretical significance as language production is a critical component of interpersonal communication. If aging impairs language production, it will disrupt interpersonal communication contributing to social isolation. (Ryan, See, Meneer & Trovato, 1994)

Language is the expression of human communication through which knowledge, belief, and behavior can be experienced, explained, and shared. This sharing is based on systematic, conventionally used signs, sounds, gestures, or marks that convey understood meanings within a group or community (National institute of deafness and other communication disorder, 2002).

Foremost among the studies in the linguistic literature, an age-related decrease in linguistic comprehension is reported are those of (Ulatowska, Cannito, Hayashi & Fleming, 1985), studied the comprehension of discourse. These studies were done by considering discourse as “representing that level of communicative function wherein the interaction between linguistic abilities and cognitive abilities is most clearly manifested, and where the complexity of language is quite high” (Ulatowska et al., 1985). Linguistic and cognitive deficits observed in aging: Research investigating aspects of language in the elderly has mainly focused on whether their language is different from that of younger age groups and, if it is, whether the changes show only increased difficulty or whether different processing strategies are used by the

C

http://www.nidcd.nih.gov/health/voice/whatis_vsl.htm

CLAP in Malayalam

107

elderly (Light 1990). Cohen (1979) reported that elders have difficulty in making inferences based on story content and recalling gist information from narratives. Riegel (1968) reported age-related changes in word-association responses specifically increased variability and longer reaction time for elderly.

Van der Linden, Hupet & Feyereisen, 1999 tested

young and older adults on their ability to understand texts and recall sentences and words. Further, they concluded that age-related differences in language, memory and comprehension were explained by a reduction of the capacity of working memory, which was itself influenced by reduction of speed, increasing sensitivity to interference.

Cognitive decline is not uniform across various domains (attention, memory, problem solving, organization) as well as there exists some interactions among them. These various aspects of cognition are vulnerable to aging in different individual. Researchers are always keen to study those changes in cognition in relation to aging which they would believe, helps them to differentiate between normal and pathological aging. Age-related changes in the attentional network: Studies reported that deficit in sustained attention as well as selective attention accompanies normal aging. Isella, Mapelli, Morielli, Pelati, Franceschi and Appollonio ,2008 opinioned a possible decrease in the ability to maintain sustained attention during complex and prolonged tasks to a known deficit underlying impaired decision making in normal aging. Cancellation Test (CT) is used in the literature as a measure of visuo-motor performance and of vigilance and sustained attention. (Lezak, 1995). Aging and reaction time: Literature also says older adults were as skilled as younger people at assimilating information, but they did take longer time to react Myerson, Robertson and Hale, (2007). The elderly participants experience difficulty in inhibiting the irrelevant (distractor) and facilitating the relevant ones (target). This is in concurrence with finding of Redfern, Muller, Jennings and Furman.(2002), where they reported that in the presence of distracters, older people tend to devote their exclusive attention to one stimulus, and ignore another stimulus than younger people. Aging and memory: Nilsson (2003) did a study where he compared the effects of aging on episodic memory, semantic memory; short term memory and priming and reported that episodic memory is primarily impaired in normal aging.

Backward recall of automatic word sequences involves declarative and working memory abilities found to be impaired in the early stages of cognitive decline (Osterberg, Fernaeus, Bogdanovic & Wahlund, 2008). Babcock and Salthouse (1990) in their study revealed an age related decreased performance for backward digit span to about 14 to 8% compared to forward. This decline was probably due to the central executive function involvement (Pearson, Logie & Gilhooly, 1999). Aging and problem solving: In (1982), Kausler reported, problem solving activity stretches over minute, if not hours and it is important to consolidate states and their values in long term memory. Denney (1990) gave a conclusive statement regarding the problem solving task and aging that, older subjects universally perform less well on problem solving tasks than do younger subjects, and the decline in performance is a linear function with respect to age. Aging and organization: Sloutsky (2003) explained three reasons why categorization is important in perception, problem-solving and memory. First, with regard to cognitive resources, categorization is an efficient way to incorporate large number of objects in to smaller number of categories. Secondly, it helps to create hierarchies about the knowledge of objects which makes the cognitive processing less demanding. Finally, categorization promotes induction process as every member of single category often possesses unobserved common properties. Smiley and Brown (1979) gave a conclusive statement regarding the comparative study of categorization skills in elderly and young adults that a decrease in the use of taxonomic relation appears after 60 years of age.

As there exists some interaction and independence among various cognitive domains and are susceptible to aging in individuals differently, a cognitive-linguistic assessment protocol which accounts for this variability is of great interest to researchers and to the growing numbers of older people who want to make sure that their cognitive functioning remains intact well into their later years.

There are several tools which assess the

cognitive-linguistic abilities which include Cambridge Cognition Examination (CAMCOG) by Roth, Tym, Mountjov, Huppert, Hendrie, Verma and Goddard, 1986; American Speech-Language-Hearing Association Functional Assessment of Communication Skills for adults (ASHA-FACS) by Frattali, Holland, Thompson, Wohl and Ferketic, 1995; Measures of Cognitive-Linguistic Abilities (MCLA) by Graser,

Krchnavek, Hauck, Calabrese and Ellmo, (1995); Ross


108

Information Processing Assessment (RIPA-2) by Deborah Ross-Swain, (1996); Assessment of Motor and Process Skills (AMPS) by Bray, Fisher and Duran, 2001.(cited in StrokEngine,2009)

The available Cognitive linguistic assessment

tools in Kannada language include Cognitive-Linguistic Assessment Protocol for Adults (CLAP) in Kannada (Kamath, A, 2001).Cognitive-Linguistic Assessment Protocol for Children (CLAP-C) (Anuroopa, 2006). Cognitive-Linguistic Assessment Protocol for Children with Learning Disability (Kavya, 2007).The available tools to assess cognitive communication abilities in Malayalam speaking population include Malayalam Mini Mental Status Examination (M-MMSE) by Mathurnath, Hodges, Mathew, Cherian and George, 2004 and Addenbrookes Cognitive Examination (ACE) by Mathurnath et al, 2004 etc

India, being a multilingual country, a lots of linguistic and ethno cultural issues will arise when we use the western assessment tools in our population. Hence, there arises a need to develop a tool which can assess cognitive-linguistic abilities in elderly population.

Considering the above as basis, the current study mainly aimed to adapt the Cognitive-Linguistic Assessment Protocol for Adults (CLAP)-Kannada (Kamath, 2001) to native Malayalam speaking, normal elderly persons. The following were the objectives considered for the present study: (1) Assessment of cognitive-linguistic abilities in normal elderly persons in the age range of 65-80years through the administration of adapted (CLAP-M). (2) To see the age related changes if any, in performances in various cognitive-linguistic skill and (3) To see the gender related changes if any, in performance in various cognitive-linguistic skills.

Method

The current study primarily aimed at adapting the Cognitive-Linguistic Assessment Protocol for Adults (CLAP) - Kannada (Kamath, 2001) to Malayalam speaking persons in the age range of 65-80 years. Participants: A total of 60 Malayalam speaking participants (30 males and 30 females) in the age range of 65-80 years were included in the study. Participants were selected from the residential areas and old age homes in the city of Trivandrum (Kerala state).They were classified based on the age classification given by Nuegarten (1974): 55-75years as young-old, 75-85years as old-old, above 85 years as oldest-old. The

existing classification system given by Nuegarten (1974) was modified by adding two more groups as: 65-75 years as young-old adults, 75-80 years as old-old adults. The participants were selected by adhering to the appropriate ethical procedures. They were explained the purpose and procedures of the study, and informed verbal and or written consent was taken. The following criteria were considered for selection of participants: The participants should not have any obvious or known history of neurological and/or psychological disorders as reported. The participant should be able to speak, read and write in Malayalam. He/she should have a minimum of primary schooling. Equal number of males and females were considered for the study. The participant should not have any noticeable sensory deficits. The participant should not have a history of drug/alcohol abuse as reported. The participant should not have any deficit in communication. Mini Mental State Examination (MMSE) was used for screening. All participants scoring greater than or equal to 25 points on the MMSE were considered for the study. Procedure: The study was carried out in 3 phases: Phase 1: The existing CLAP in Kannada (Kamath, 2001) was translated in to Malayalam. The material was prepared by reviewing the literature in Malayalam from books, journals and web-based sources.

The Syntactic structure and semanticity in Malayalam language was studied and compared to Kannada. The commonalities and difference were studied to choose the appropriate test items. The test item in Malayalam version was seen for semanticity, familiarity adapt and ambiguity. Then the Kannada version was translated to Malayalam language after suitable modification. The translation was done with the help of a linguist and speech-language pathologist. Materials for various test items were prepared in Malayalam.

The test items consisted of the following domains which formed the basis for CLAP in Malayalam. These were: a) Attention, Discrimination and Perception b) Memory c) Problem solving and d) Organization. Phase 2: The translated materials were given to five experienced speech language pathologist (SLP) who served as judges. Professionals (SLP) were asked to rate the material on a Feedback questionnaire rating scale (Goswami, Shanbal, Samasthitha & Navitha, 2010), which included 17 parameters rating from very poor to excellent. Judges were also asked to give

CLAP in Malayalam

109

suggestions regarding the rating scale used and any other changes in the protocol. Phase 3: The final test item was modified by incorporating the suggestions of SLP‟s and administered on 60 participants in the age range of 65-80 years. Instruction specific to the task were given in Malayalam. The scoring was carried out simultaneously for each task as per the scoring procedure scheduled for each item. Various test items and their respective scoring under each domain was same as stated in the CLAP- Kannada version (Kamath, 2001). Domain I: Attention, Perception and Discrimination

Two main modes were considered for this domain: Visual and Auditory Two types of attention processes were evaluated, via, selective attention and sustained attention. The cognitive processes of perception and discrimination were contingent on attention and three are parallel processes. Visual category a) The cancellation at a letter level: The cancellation of

the entire stipulated letter was considered as the task required sustained attention.

b) The cancellation at word level: Similar to the above test, this is also a sustained attention task.

c) Contingent cancellation: This task requires the fulfillment of a pre-requisite contingency before cancellation of the letter. This task was done to evaluate the selective attention.

Scoring: Performance of each subtest was scored based on the accuracy of responses. The time taken to complete the task was also noted. Auditory category a) Sound count : Sustained auditory attention was

evaluated using a task where the participant were required to count mentally how many times a particular letter was read out in a list.

b) & c) Letter-Pair discrimination and Word –Pair discrimination: The second and third subtest involved the participant‟s ability to discriminate amongst a pair of letters/ words read out by the tester (same/different task) respectively. This is a predominantly a discrimination task.

d) Months-backwards naming: The last sub test required the participants the names of the months in the backward direction (i.e. December to January). This task requires the attention and involves recall processes too.

Scoring: Performance on each task was scored for number of correct answers. Subtest (a) was given one point for every correct count. One point was deducted

for every count above the correct number of total occurrence of stipulate letters. Domain II: Memory

Three main types of memory processes were tested: Episodic memory, Working memory and Semantic memory. Episodic memory: was tested by asking questions that tested orientation of self with respect to place, self and time and also a few questions of general knowledge. Scoring: A score of one was given for each correct answered question. Working memory : was evaluated using digit forward and digit backward repetition tasks. A maximum of seven digits were included in the list Scoring: A score of one was given if all the digits in the list were repeated in the correct expected order. Semantic memory: Tasks included items to test for language –knowledge such as: a) Co-ordinate naming: The participants were given

a noun-class and asked to name at least five objects which may be included with in that class.

b) Super-ordinate naming: This is a task complementary to coordinate naming. A list of items belonging to a particular class was given to the subject to identify the class to which the given items may be classified.

c) Word naming fluency: This is a task to evaluate recall, and involved the participants to name five words that begin with a specified letter.

d) Generative naming: The participants were asked to name the target word, the description of which in terms of use was given.

e) Sentence repetition tasks: A phrase /sentence were read out to the participants and an immediate repetition was expected. The target utterance was not very long, keeping in mind that very long and multiple transformation utterances were not used often in everyday verbal language use.

f) Carrying out commands: Two objects; a pen and a book were placed in front of the subject. Commands of various level of complexity, which required manipulation of these objects, were given.

Scoring: All items were allotted a score of one for each correctly answered question. They were untimed task.

Domain III: Problem solving

This domain includes various tests that assess reasoning abilities to aid in problem solving. The following tasks were considered. a) Sentence disambiguation: An ambiguous sentence

was given to the participants, and was instructed to


110

explain the two interpretations that could be made from the sentence.

Scoring: One point was scored for each correct explanation of the meaning of the sentence. Two points were scored only if both the meaning interpretation were clearly stated. b) Sentence formulation: This was a word order

unscrambling task to form a grammatically correct sentence.

c) Predicting the outcome of a desired situation d) Comparing and contrasting two objects: Here, the

participants were required to give one similarity and one difference between a pair of objects named.

e) Predicting cause of a described situation f) Answering why questions g) Sequential task analysis: The step involved in

carrying out a named task was required to be listed by participants. The participants were required to analyze the task in to at least four steps for a full score to be given.

Scoring: Each of the above tasks (b-g) was given a score of one for a correct answer. They were untimed task. Domain IV: Organization Organization of available information to result in coherent communication is an important met linguistic and metacognitive task. a) Categorization abilities were tested as a measure

of word-class organization abilities. b) Analogies: This task consisted of items to test the

ability to recognize word concept to meet task demands. This task also involves logical reasoning processes.

c) Sequencing of events in a temporal order to form a coherent story was also taken up as a task of organization skills.

Scoring: A score was given for each correct answer. The task was not timed.

All the items within a subtest were placed in an order of increasing complexity. All tasks were given only one trial each.


The results obtained from the data were analyzed on various aspects. The findings of the present study have been broadly presented under the following headings:

Domain I: Attention, Discrimination and Perception (a) Performances of males and females in different age groups in Attention, Discrimination and Perception

(domain I) across in gender. In this section visual and auditory, two modalities were assessed.

As mentioned in the Table1, it was found that, among the three test items, the participants obtained better mean scores on AVLC (M=9.31; SD=0.74) and AVWC (M=9.28;SD=0.88) which requires sustained attention. On the other hand, comparatively poorer scores were obtained by the participants in AVCLC test item (M=8.63; SD=1.02) (rely on selective attention) across the three age groups.

Table 1. Mean(M) and SD values for male and female

participants (normals) in domain I- Visual mode

*AVLC=attention visual letter cancellation: AVCLC=Attention visual contingent letter cancellation; AVWC=attention visual word cancellation female participant (normals) in domain I-visual mode b) Effect of age and gender performance on various domains

Table 2.F values of age and gender

As depicted, the above Table 2 shows a

significant difference [F (2, 54) =31.53, p< 0.05] across each age group but no significant difference across gender and no age and gender interaction were noted in all test items of domain I visual mode. As there existed a significant difference across all age groups, the post hoc analysis Duncan‟s was administered to show the performance of each age group separately in all the test items of domain 1-visual mode.

Domain I*

Group Males Females Total

M SD M SD M SD

AVLC

65-70 70-75 75-80

TOTAL

9.70 9.80 8.60 9.36

0.48 0.42 0.69 0.76

9.70 9.60 8.50 9.26

0.48 0.51 0.52 0.73

9.70 9.70 8.55 9.31

0.47 0.47 0.60 0.74

AVCLC

65-70 70-75 75-80

TOTAL

9.60 8.50 7.80 8.60

0.51 0.84 0.63 0.99

9.70 8.40 7.80 8.63

0.48 1.07 0.42 1.06

9.65 8.45 7.80 8.63

0.48 0.94 0.52 1.02

AVWC

65-70 70-75 75-80

TOTAL

9.90 9.60 8.50 9.30

0.31 0.69 0.84 0.88

9.90 9.50 8.30 9.23

0.31 0.52 0.82 0.89

9.90 9.55 8.40 9.28

0.30 0.60 0.82 0.88

Domain I-Visual

Age Gender Age & Gender

AVLC F Sig F Sig F Sig

1.53 0.00 0.53 0.46 0.17 0.83 AVCLC 5.76 0.00 0.00 1.00 0.10 0.90 AVWC 1.22 0.00 0.38 0.54 0.12 0.88

CLAP in Malayalam

111

Attention visual letter cancellation (AVLC): In Duncan‟s test it was found the age group of 65-70 and 70-75 years performed equally and which was away from 75-80years with a mean of 9.70. Attention visual contingent letter cancellation (AVCLC) : The male and female participants showed a significant difference [F (2,54)=35.7, p<0.05] corresponding with age. On post-hoc analysis, results showed that there was a decline in the means scores as the age increases (i.e. 75-80 years totally away from 65-70years (9.65) and 70-75 years (8.45) with a mean score of 7.80). Attention visual word cancellation (AVWC): In this task obvious significant difference was shown by all female and male participants across age [F (2, 54) = 31.22, p<0.05]. The Duncan‟s test revealed there was a significant decline from 65-70 years to 75-80years i.e. the age group 65-70 and 70-75 years performed almost equally with a mean of 9.90 and 9.50 showing no significant difference between these two groups where as it was away from 75-80years with a mean of 8.40.

The Continuous Performance Test has been widely used to measure sustained attentional deficits. In the above mentioned visual attention task it was found that the performance of participants on contingent letter cancellation which relay on sustained attention with a greater memory load and selective attention (Kamath, 2001) showed a poorer performance compared to other two test items, this could be due to deficit in discriminating (target) relevant from irreverent information. In AVCLC where the participants were asked to cancel the target letter following a particular letter which requires a pre mind set for the target letter. Cancellation Test (CT) is used in the literature as a measure of visuo-motor performance and of vigilance and sustained attention (Lezak, 1995). The Six Letter Cancellation Task (SLCT) measures sustained attention (Rangan, Nagendra & Bhatt, 2009).

Two common error patterns were observed in participants of 70 years onwards which may be contributing factors in poor performances in AVCLC. In one type of error the participants merely canceled the target letter without paying much attention to observe whether the target letter is followed by a particular letter. As a frequency of occurrence of both the target and particular letter is same, some participants adopt a strategy where they make use of the following particular letter to cancel the target letter assuming that the target letter always followed by a

particular letter. So, this task demands the selective attention for a better performance. As the cognitive load is more in selective attention compare to sustain attention, all the age groups perform differently.

A similar finding, which talks about an age related deficit in selective attention, was mentioned in the study of Allen, Madden, Groth and Crozier, (1992). A refuting study was reported by Madden (1990) who found a preservation of selective attention in older adults. Rabbit (1964) gave a conclusive statement that the older subjects had difficulty ignoring irrelevant information and were at a disadvantage when searching complex stimuli.

Table 3. Mean(M) and SD values for male and female

participants (normals) in domain I- auditory mode

*AASC=attention auditory sound count: AALPD=attention auditory letter-pair discrimination: AAWPD=attention pair discrimination; AAMBN=attention auditory memory backward naming Table 4. F values of age and gender for each male and

female participant (normals) in domain I-auditory mode

Domain I*

Group Males Females Total M SD M SD M SD

AASC 65-70 70-75 75-80

TOTAL

9.60 9.10 7.80 8.83

0.84 0.99 0.91 1.17

9.70 8.70 8.10 8.83

0.48 0.82 0.73 0.94

9.65 8.90 7.95 8.83

0.67 0.91 0.82 1.06

AALPD 65-70 70-75 75-80

TOTAL

4.90 4.80 4.70 4.80

0.31 0.42 0.48 0.40

5.00 4.90 4.60 4.83

0.00 0.31 0.51 0.37

4.95 4.85 4.65 4.81

0.22 0.36 0.48 0.39

AAWPD 65-70 70-75 75-80

TOTAL

5.00 5.00 4.80 4.93

0.00 0.00 0.42 0.25

5.00 5.00 4.70 4.90

0.00 0.00 0.48 0.30

5.00 5.00 4.75 4.90

0.00 0.00 0.40 0.27

AAMBN 65-70 70-75 75-80

TOTAL

8.40 7.30 5.80 7.16

0.84 0.94 0.78 1.36

8.70 6.70 5.70 7.03

0.82 0.82 0.94 1.51

8.50 7.00 5.75 7.10

0.82 0.91 0.85 1.43

Domain I-auditory

Age Gender Age & Gender

F Sig F Sig F Sig AASC 21.77 0.00 0.00 1.00 0.97 0.38

AALPD 3.19 0.04 0.11 0.73 0.45 0.63 AAWPD 6.08 0.00 0.24 0.62 0.24 0.78 AAMBN 52.59 0.00 0.35 0.55 1.35 0.26


112

It can be inferred from the Table 4 that there is significant difference of each test item individually. Attention auditory sound count (AASC): In this task across all age groups each and every participant showed a significant difference [F (2, 54) =21.77, p<0.05]. Attention auditory letter pair discrimination (AALPD) The male and female participants attained a value of F (2, 54)= 3.19, p<0.05 with significant difference across age. The Duncan‟s test was carried out to see the particular test item performance of each age group which reveals that the performance of participants in the age range of 70-75 were not significantly difference from 75-80 and 65-70years with a mean valve of 4.85, but the 75-80 and 65-70 years showed significant difference during the performance with a mean score of 4.65 and 4.95. Attention auditory word pair discrimination (AAWPD ) A significant difference [F (2, 54) =6.08, p<0.05] across age group was shown by male and female participants. On post hoc Duncan‟s test, the participants of 75-80years showed significant difference which is almost away from 65-70 and 70-75years with a mean of 4.75. At the same time, the participants of 65-70 and 70-75 years performed equally on test item resulting no significant difference with an equal mean value of 5.00. At tention auditory month background naming (AAMBN) : The male and female participants showed significant difference across age [F(2, 54)= 52, p<0.05]. Duncan‟s test was carried out to show the performance of each age group separately which revealed 65-70 and 70-75 years showed a significant difference with a mean of 8.55 and 7.0; 75-80years showed a significant difference which is totally away from 65-70 and 70-75years with a mean of 5.75, this led to the conclusion that participants of 75-80 years performed poorer compare to 65-70 and 70-75years. In these domains, the auditory sound count task requires the participant to pay sustained attention whenever she/he hears a target sound. Hochandel and Kaplan (2004) reported that deficit in sustained attention as well as selective attention accompanies normal aging. This notion was supported by Isella et al., (2008), where they opinioned a possible decrease in the ability to maintain sustained attention during complex and prolonged tasks to a known deficit

underlying impaired decision making in normal aging. There was a decline as the age progresses in this particular task and the decline was so evident in 75-80years. Though the two test items AALPD and AAWPD are almost same in terms of cognitive processes involved i.e. the participants need to discriminate letters as in AALPD and words in AAWPD. As age advances the discrimination ability in elderly reduces this can be attributed to their impaired temporal processing ability. Difficulty in discriminating speech sounds in the elderly can be reasoned to the deterioration in the ability to process dynamic aspects of speech such as formant transitions. In the aging auditory system, this deterioration in temporal processing speed may be a manifestation of a deficit in encoding time-varying sounds that contain rapidly changing frequencies such as formant transitions (Mendelson & Ricketts, 2001). All the participants across three age group show significant difference with each other and also scored poor in AMBN task. This can be attributed to the deficit in executive function where they have to inhibit the irrelevant and promote relevant information. Executive aspects of working memory such as inhibition typically decline in normal aging as reported by Treitz, Heyder and Daumand (2007). Taking in to account of that, the AMBN can predict some amount of cognitive decline in elderly. Reaction time: MANOVA was done to see the significant difference in reaction time across age and gender in domain I-visual mode items. No significant difference was seen on statistical analysis across gender and various test items. Similarly no interaction effect is seen between age and gender among various test items. Attention visual letter cancellation reaction time (AVLCRT ): The male and female participants showed a significant difference for reaction time [F (2, 54) = 166.78, p<0.05] across age. Table 5. F values of age and gender for each male and

female participant (normals) for reaction time in domain I-visual mode

Domain I- Visual

Age Gender Age & gender F Sig F Sig F Sig

AVLCRT 166.78 0.00 0.76 0.38 0.32 0.72

AVCLCRT 134.94 0.00 0.08 0.77 0.33 0.71

AVWCRT 114.23 0.00 0.33 0.56 0.43 0.65

CLAP in Malayalam

113

The Duncan‟s test was carried out to see the significant difference across age, the test results shows there was a decline in the reaction time of task performance as the age increases. In this task, age group of 65-70years performed with a mean value of 25.60, 70-75years performed with a mean score of 36.30 and 75-80 years with a mean value of 56.70. Attention visual contingent letter cancellation reaction time (AVCLCRT) : From Table 5 it can be inferred that there is a significant difference with [F (2, 54) = 134.94, p<0.05] in male and female participants respectively. Duncan‟s test was done to show the effect of AVCLCRT on each age group individually, it revealed there was a decline in the performance of participants as the age increased .i.e. 65-70years showed low reaction time to perform a task with a mean of 33.10 than other two age groups with a mean of 44.10 and 65.10 respectively and as the age increased they required more reaction time to finish the task. Attention visual word cancellation reaction time (AVWCRT) : A significant difference with [F (2, 54) = 114.23, p<0.05] was shown by each male and female participant across age. To see the significant difference across each age group separately, post hoc Duncan‟s test was carried out. Duncan‟s test results showed that 65-70years performed better with a mean of 28.70 compared to 70-75 and 75-80years with a mean of 39.70 and 61.40, this showed the performance of participants in 75-80 years age group was away from the 65-70 and 70-75years.

It is vivid from the results that the reaction time of elderly participants increased as the age progressed. Moreover, across the three test items, participants took more time for performing the task of attention visual contingent letter cancellation (AVCLC), with almost equal time shared between the other two tasks, attention visual letter cancellation (AVLC) and attention visual word cancellation (AVWC). This could be due to the difference in the cognitive demand they encounter while performing the task. All the three tasks make use of a recognition reaction time pattern, where they have to recognize the target letter or word from the distractor items. Even though, attention visual contingent letter cancellation (AVCLC) also make use of a recognition reaction time pattern, the cognitive demand it impose is more compared to other in two ways; a) It‟s a task which depends on selective attention whereas attention visual letter cancellation (AVLC) and attention visual word cancellation (AVWC) depend on sustained attention. b) Working memory load is more as participants has to remember the complex instruction.

This could be one of the reasons why participants perform differently to test items showing a clearly descending fashion of performance in reaction time with the increment of age across the three age groups. The above reason could also be speculated for the similar pattern of performance by the younger two age group taken up in the study compared to the older one (greater than 75years) for the task attention visual letter cancellation (AVLC) and attention visual word cancellation (AVWC). Familiarity and simplicity of the task could also contribute to their mixed performance across the test items which mean to say that the letter cancellation and word cancellation are much simpler one compared to contingent letter cancellation. This indirectly points out that the amount and type of cognitive process involved is same.

As with the advancement of age there could be a

gradual degradation in fine motor skill as well as eye-hand co-ordination which are important for performing the subtle tasks like the one mentioned in the above test items. This will in turn result in longer reaction time in elderly populations.

In addition the elderly participants experience

difficulty in inhibiting the irrelevant (distractor) and facilitating the relevant ones (target). This is in concurrence with finding of Redfern (2002), where they reported that in the presence of distracters, older people tend to devote their exclusive attention to one stimulus, and ignore another stimulus than younger people. Literature also says older adults were as skilled as younger people at assimilating information, but they did take longer time to react Myerson, Robertson and Hale (2007). Domain II: Memory Episodic memory orientation and recent memory questions (EMORMQ): The male and female participants obtained a significant difference with [F (2, 54) = 19.89, p<0.05] respectively. Following MANOVA, the Duncan‟s test was administered to show the performance of each participant across the three age group and the test results revealed a significant difference with a mean value of 10.00, 9.30and 8.60 by the participants of 65-70, 70-75 and 75-80 years correspondingly. The participants of 65-70years showed a better performance compare to 70-75 and 75-80years which shows as the age increases the performance declined. Working memory digit forward (WMDF): The male and female participants showed a significant difference with [F (2, 54) =13.71, p<0.05] respectively. Following MANOVA, the Duncan‟s test was carried out to show the significant effect of above test item in each age


114

group, results showed a significant difference between the age range of 75-80 years with a mean of 2.00 which was away from 70-75 and 65-70 years with a mean score of 2.70 and 2.95. Working memory digit backward (WMDB): There was no significant difference found with [F (2, 54) = 5.77, p<0.05] across age in males and females correspondingly. The results of the post hoc analysis can be listed as: The participants of 75-80years showed a significant difference from 65-70years with a mean value of 1.25. There was no significant difference between 65-70 and 70-75 years with a mean value of 1.60 and 1.85 respectively. Semantic memory coordinates naming (SMCN): There was a significant difference across age with [F (2, 54) =14.2, p<0.05]. Following MANOVA, post hoc analysis Duncan‟s test was administered to see the performance of each age range for the particular test item, the test showed a significant difference in age group of 75-80 years with mean score of 3.25 which was perfectly away from the age range of 70-75 and 65-70 years with a mean valve of 4.10 and 4.30 it shows the 65-70 and 70-75 years performed almost equally but it was away from 75-80 years. Semantic memory subordinate naming (SMSN): A significant difference was found with [F (2, 54) =9.0, p<0.05]. across age groups. On Duncan‟s post hoc analysis, it was found that the participants of age group 65-70 years (5.00) showed a significant difference to both the 70-75 and 75-80 years age group with a mean value of 4.60 and 4.45 respectively. On the other hand, participants of 70-75 and 75-80 years obtained an almost equal mean value, failed to show a significant difference between them. Semantic memory word naming fluency (SMWNF): The participants across all age group showed a significant difference of [F (2, 54) =23.62, p<0.05] for the current test item. On Duncan‟s post hoc analysis, a significant difference was seen between all the three age group with the participants in the first age group, 65-70 years (4.75) showed a significant difference with 70-75 (4.30) and 75-80 years (3.50) respectively. A significant difference was also found in the performance of participants between 70-75 and 75-80 years age group. All male and female participants of each age group showed a significant decline in performance of episodic memory task with the progression of age. This finding is well supported with the study done by Nilsson (2003) who reviewed the various studies on memory function in normal aging and concluded that

episodic memory is more susceptible for aging. It could be predicted from the performance of participants for digit forward and backward tasks which intend to assess a working memory showed an age related decline. This statement maintains agreement with the study done by Gregoire and Van der Linden (1997); Baeckman et al., (2000). Moreover from the mean score value of these tasks it was evident that participants perform poorer in backward digit span compared to forward. The same finding was reported by Babcock and Salthouse (1990) as their study revealed an age related decrease performance for backward digit span to about 14 to 8% compared to forward. This decline was probably due to the central executive function involvement (Pearson et al., 1999). Literatures attribute one more reason for the declined performance in digit span task to greater storage load than the executive function deficit. Though digit span is involve relatively little complex processing, the maximal verbal memory span depended on both the phonological loop and central executive. As the digit load increase, demands made on central executive also increases (Baddeley, 2001). Studies comparing the effects of aging on episodic memory, semantic memory, short- term memory and priming find that episodic memory is especially impaired in normal aging (Nilsson, 2003). These deficits may be related to impairments seen in the ability to refresh recently processed information (Johnson, Reeder, Raye & Mitchell, 2002). In addition, even when equated in memory for a particular item or fact, older adults tend to be worse at remembering the source of their information, a deficit that may be related to declines in the ability to bind information together in memory during encoding and retrieve those associations at a later time (Mitchell, Johnson, Raye & D‟Esposito, 2000; Naveh-Benjamin, 2000). Semantic memory is related to experience. As people grow older, semantic memory typically remains stable or even improves. It is true that memory changes with age. It is not true that all memory declines with age. It really depends on the memory type referred to and the life circumstance of the individual (e.g., medical condition; level of activity). Older adults exhibit major declines in episodic memory, performing more poorly that involve episodic recall or recognition of virtually any stimuli, for example, single words or prose passages, spatial locations, pictures, faces and activities (Burke & Light1981). Age-related declines in episodic memory performance are inversely related to years of education, but nevertheless they usually remain when education is controlled statistically (Nyberg, Cabeza & Tulving, 1996).

CLAP in Malayalam

115

Most research on memory and aging has focused on how older adults perform less well at a particular memory task. However, recently researchers have also discovered that simply saying that older adults are doing the same thing, only less of it, is not always accurate. In some cases, older adults seem to be using different strategies than younger adults. For example, brain imaging studies have revealed that older adults are more likely to use both hemispheres when completing memory tasks than younger adults (Cabeza,

2002). In addition, older adults sometimes show a positivity effect when remembering information, which seems to be a result of the increased focus on regulating emotion seen with age (Mather & Carstensen, 2005). For instance, eye tracking reveals that older adults showed preferential looking toward happy faces and away from sad faces (Goren & Wilson, 2006). No significant difference was seen on statistical analysis in gender and various test items.

Table 6. Mean (M) and Standard Deviation (SD) values for male and female participants (normals) in Domain II

*EMORMQ= episodic memory orientation and recent memory questions: WMDF= working memory digit forward: WMDB=working memory digit backward SMCN=semantic memory coordinate naming: SMSN=semantic memory super ordinate naming: SMWN=semantic memory word naming fluency: SMGN=semantic memory generative naming: Semantic memory sentence repetition: SMCC=semantic memory carry out commands

DOMAIN II *

Males Females Total Groups M SD M SD M SD

EMORMQ 65-70 70-75 75-80

TOTAL

10.0 9.40 8.70 9.36

0.00 0.84 0.82 0.85

10.009.20

8.50 9.20

0.00 1.13 0.52 0.93

10.00 9.30 8.60 9.30

0.00 0.97 0.68 0.88

WMDF 65-70 70-75 75-80

TOTAL

2.90 2.70 2.00 2.53

0.56 0.67 0.47 0.68

3.00 2.70 2.00 2.56

0.66 0.48 0.66 0.72

2.95 2.70 2.00 2.55

0.60 0.57 0.56 0.69

WMDB 65-70 70-75 75-80

TOTAL

1.80 1.60 1.20 1.53

0.63 0.51 0.42 0.57

1.90 1.60 1.30 1.60

0.73 0.51 0.48 0.62

1.85 1.60 1.25 1.56

0.67 0.50 0.44 0.59

SMCN 65-70 70-75 75-80

TOTAL

4.30 4.00 3.30 3.80

0.67 0.47 0.82 0.77

4.30 4.20 3.20 3.90

0.67 0.63 0.63 0.80

4.30 4.10 3.25 3.88

0.65 0.55 0.71 0.78

SMSN 65-70 70-75 75-80

TOTAL

5.00 4.60 4.50 4.70

0.00 0.51 0.52 0.46

5.00 4.60 4.40 4.60

0.00 0.51 0.51 0.47

5.00 4.60 4.45 4.68

0.00 0.50 0.51 0.46

SMWNF 65-70 70-75 75-80

TOTAL

4.80 4.30 3.60 4.23

0.42 0.67 0.69 0.77

4.70 4.30 3.50 4.16

0.48 0.48 0.52 0.69

4.70 4.30 3.50 4.20

0.44 0.57 0.60 0.73

SMGN 65-70 70-75 75-80

TOTAL

5.00 5.00 5.00 5.00

0.00 0.00 0.00 0.00

5.00 5.00 5.00 5.00

0.00 0.00 0.00 0.00

5.00 5.00 5.00 5.00

0.00 0.00 0.00 0.00

SMSR 65-70 70-75 75-80

TOTAL

10.0 10.0 10.0 10.0

0.00 0.00 0.00 0.00

10.00 10.00 10.00 10.00

0.00 0.00 0.00 0.00

10.00 10.00 10.00 10.00

0.00 0.00 0.00 0.00

SMCC 65-70 70-75 75-80

TOTAL

10.0 10.0 10.0 10.0

0.00 0.00 0.00 0.00

10.00 10.00 10.00 10.00

0.00 0.00 0.00 0.00

10.00 10.00 10.00 10.00

0.00 0.00 0.00 0.00


116

Similarly no interaction effect is seen between age and gender among various test items.

Domain III: Problem solving

In the test item, problem solving sentence disambiguation (PSSD), participants showed a significant difference [F (2, 54) =30.31, p<0.05]. A significant difference [F (2, 54) = 18.81] was obtained by all the participants for the test-items, PSSF. All the male and female participants showed a significant difference in performance for the test items, problem solving predicting outcome (PSPO), problem solving compare and contrast (PSCAC ) across age group [F(2,54) = 11.04, p<0.05] and [F(2 54)= 24.36, p<0.05] respectively.

A significance of [F (2, 54) =21.18, p<0.05] and

[F (2, 54) =31.38, p<0.05] was found in all the participants for the test item problem solving predicting cause (PSPC) and problem solving sequential analysis (PSSA) in the above domain correspondingly. Following to this, Duncan‟s test was administered to see which age group is showing significant difference across various test items in the above domain. It was found that no significant difference was seen on statistical analysis in gender and various test items. Similarly no interaction effect is seen between age and gender among various test items. Problem Solving Sentence Disambiguation (PSSD): The participants 75-80 years showed a significant difference with a mean of 6.40 which was away from 70-75 and 65-70 years with a mean of 7.50 and 8.65 respectively. With the mean value of 4.55, the participants of 65-80 years age group showed a significant difference with the participants of 70-75 years and 75-80 years having a mean value of 4.05 and 3.55 correspondingly. At the same time, the participants of 70-75 years were found to be significantly differing from 75-80 years for the test item PSSF. Problem Solving Predicting Outcome (PSPO): The participants of 70-75 years with a mean value of 8.85 were significantly different from the participants of the age group of 65-70 and 75-80 years with a mean value of 9.50 and 8.20 respectively. Being 9.50as the mean value, the participants of age group 65-70 years showed a significant difference from 75-80 years. Problem Solving Predicting Cause (PSPC): All the participants of 65-70 years with a mean of 9.30 showed a significant difference with the age group of 70-75 years (8.60) and 75-80years (7.70).At the same time, the participants of 70-75 years showed a significant

difference in performance with the participants of 75-80 years. Problem Solving Sequential Analysis: By having a mean value of 9.45, participants of 65-70 years age group attained a performance which is significantly different from 70-75 years (8.85) and 75-80 years (7.45). Similarly all the participants in the age group

Table 7. Mean (M) and SD values for male and female

participants (normals) in Domain III

*PSSD=problem solving sentence disambiguation: PSSF=problem solving sentence formulation: PSPO= problem solving predicting outcome: PSCAC=problem solving compare and contrast: PSPC=problem solving predicting cause: PSWQ=problem solving why questions: PSSA= problem solving sequential analysis

70-75 years showed a significant difference in performance compared to 75-80 years. In most of the tasks the participant has to think about routine events and perform a task, as in problem solving compare and contrast (PSSD), problem solving sentence formulation (PSSF), problem solving predicting outcome (PSPO), problem solving predicting cause (PSPC), and problem

Domain

III *

Group

Males Females Total M SD M SD M SD

PSSD

65-70 70-75 75-80

TOTAL

8.90 7.60 6.10 7.53

0.99 0.51 0.87 1.40

8.4 7.4 6.7 7.5

0.96 0.84 1.15 1.19

8.65 7.5 6.4 7.51

0.98 0.68 1.04 1.29

PSSF

65-70 70-75 75-80

TOTAL

4.70 4.00 3.50 4.06

0.48 0.47 0.70 0.73

4.4 4.1 3.6 4.03

0.51 0.31 0.51 0.55

4.55 4.05 3.55 4.05

0.51 0.39 0.60 0.64

PSPO

65-70 70-75 75-80

TOTAL

9.70 9.00 8.30 9.00

0.67 0.94 1.15 1.08

9.3 8.70 8.10 8.70

0.94 0.82 0.56 0.91

9.5 8.85 8.20 8.85

0.82 0.87 0.89 1.00

PSCAC

65-70 70-75 75-80

TOTAL

9.10 8.30 7.20 8.20

0.87 0.82 0.63 1.09

8.90 8.60 6.90 8.13

1.10 1.17 0.73 1.33

9.00 8.45 7.05 8.16

0.97 0.99 0.68 1.20

PSPC

65-70 70-75 75-80

TOTAL

9.20 8.60 7.80 8.53

0.91 0.84 0.63 0.97

9.40 8.60 7.60 8.53

0.69 0.96 0.51 1.04

9.30 8.60 7.70 8.53

0.80 0.88 0.57 0.99

PSWQ

65-70 70-75 75-80

TOTAL

5.00 5.00 5.00 5.00

0.00 0.00 0.00 0.00

5.00 5.00 5.00 5.00

0.00 0.00 0.00 0.00

5.00 5.00 5.00 5.00

0.00 0.00 0.00 0.00

PSSA

65-70 70-75 75-80

TOTAL

9.50 8.90 7.50 8.63

0.70 1.10 0.52 1.15

9.40 8.80 7.40 8.53

0.84 0.91 0.69 1.16

9.45 8.85 7.45 8.58

0.75 0.98 0.60 1.15

CLAP in Malayalam

117

solving sequential analysis (PSSA). All these tasks where the participant require memory to perform, like when the participants were asked to formulate a sentence which needs organization capacity as well long term and short term memory to carry out the task. In case of PSCAC where the participants require to compare and contrast between two meaningful words, so as the age degrades the problem solving skills starts decline where in PSCAC task 65-70 and 70-75 years performed equally, where in 75-80 years performance was noticed to be poorer. But in case of PSSD, PSSF, PSPO, and PSPC the decline was greater compared to the PSCAC task where the problem solving methods are processing delayed which was supported by many studies. The poor performance of participants in PSSD can be reasoned to their decline activity of cognitive flexibility with the increment of age.

Older adults are less proficient at recalling previous activities, though this is less true for problem-solving activities. Since most problem-solving activity stretches over minutes, if not hours (years for scientific problem-solving), it is important to consolidate states and their values in long-term memory (Kausler, 1982).

Most studies of problem-solving are cross-sectional. Thus, age is confounded with cohort factors such as education. The two longitudinal studies by Arenberg (1974, 1982) show that age decline only appear to be significant in the 60s and 70s. It appears that two factors can account for much of the decline in problem solving performance: changes in speed of execution of elementary information processes and changes in the capacity of working memory. It is also possible that speed alone can account for all the changes, to the extent that working memory relies on the speed of processes such as rehearsal. Denney (1990) study reports the most general conclusion of this study was that, older subjects universally perform less well on problem solving tasks than do younger subjects, and the decline in performance is a linear function with respect to age. Domain IV: Organisation

With age as independent variable and various test items organization categorization (OC), organization analogies (OA), organization sequencing events (OSE) as dependent variable, on the above domain showed a significant difference of F(2,54)=44.34 for p<0.05; F(2,54)=34.75 for p<0.05:f (2,54)=116.77 for p<0.05 respectively. Subsequent to this, post hoc Duncan‟s test revealed all the three age groups are significantly differ from each other for all three test items. Organization categorization (OC): The age group 65-70 years was significantly different from 70-75 years

with a mean value of 9.75 which in turn significantly different from 75-80 years with a mean value of 8.50. The performance of participants in 75-80 years age group showed a significant difference with the other two age groups with a mean value of 6.90. Organization analogies (OA): A similar trend was seen, with the age group 65-70 years was significantly different from 70-75 years and 75-80 years age group Organization sequencing events (OSE): The participants of older age group 75-80 years and 70-75 years were significantly different from 65-70 years with a mean value of 23.05 and 28.60. No significant difference was seen on statistical analysis in gender and various test items. Similarly no interaction effect is seen between age and gender among various test items.

In both sequencing and categorizing events, the

cognitive decline was evident in all participants due to the decline in memory and attention aspect though the task was simple to perform. According to Sloutsky (2003), in most cognitive activities such as perception, problem-solving and memory, for three reasons: first, with regard to cognitive resources it is more efficient to incorporate a potentially infinite number of individual objects into a smaller number of categories; secondly, categorizing allows knowledge of objects to be organized, particularly by creating hierarchies, which again makes processing less demanding cognitively; and finally, categorization enables induction processes,

Table 8. Mean (M) and SD values for male and female

participants (normals) in Domain IV

*OC=organization categorization; OA = organization Analogies; OSE=organization sequencing events

since members of a single category often possess unobserved common properties.

Domain IV*

Group Males Females Total M SD M SD M SD

OC

65-70 70-75 75-80

TOTAL

9.9 8.4 7.0 8.4

0.31 1.26 1.05 1.52

9.6 8.6 6.8 8.3

0.84 0.96 1.03 1.49

9.75 8.50 6.90 8.38

0.63 1.10 1.02 1.49

OA

65-70 70-75 75-80

TOTAL

9.4 8.4 7.0 8.2

0.96 0.84 1.05 1.36

9.4 8.4 6.8 8.2

0.96 0.84 1.03 1.42

9.40 8.40 6.90 8.23

0.94 0.82 1.00 1.38

OSE

65-70 70-75 75-80

TOTAL

35.2 28.1 23.2 28.8

2.52 1.44 2.78 5.49

34.6 29.2 22.9 28.9

3.09 1.60 2.70 5.40

34.9 28.6 23.0 28.8

2.77 1.63 2.68 5.42


118

Literature reveals that effect of aging depends on the type of categorical organization, i.e., taxonomic or thematic organization. Taxonomic relations refer to groupings of objects of the same kind belonging to a semantic category (i.e. Dog & Cat as animals), whereas thematic relations correspond to an organization of knowledge in terms of familiar scenes or events (i.e. Dog & Bone since the dog usually eats bones). Taxonomic organization is often considered as more elaborated, thus being acquired later (Nelson, 1983) and having a greater inductive power (Markman, 1989). In older people, taxonomic categorization seems also to be less available than in younger adults .Studies which compared the categorization skills of elderly persons with those of young adults reflects a decrease in the use of taxonomic relations after 60 years of age (Smiley & Brown, 1979).

Two explanations could be proposed by various

authors that the decline would be functional before 75 years of age since learning increased taxonomic categorization (Denney & Denney, 1974), but would be structural after this age, being related to neurophysiological modifications (Pennequin & Fontaine, 2000). The second explanation refers to developmental changes in conceptual (Smiley & Brown, 1979) or categorical (Fontaine & Toffart, 2000) preferences, which would reflect the differential accessibility of taxonomic and thematic relations at a given age.

Conclusions

The prime focus of the current study was to adapt (CLAP) in Kannada, (Kamath, 2001) to Malayalam. Cognitive decline is not uniform across domains. Although there are some interactions among cognitive domains, it seems evident that they have some degree of independence too. These aspects of cognition are susceptible to aging in different individuals and gender differences can also be a variable. So, a cognitive-linguistic assessment protocol which accounts for this variability is of great interest to researchers and to the growing numbers of older people who want to make sure that their cognitive functioning remains intact well into their later years.

Hence, this study was taken up to adapt the Cognitive-Linguistic Assessment Protocol for Adults developed in Kannada by Kamath (2001) to document the age and gender related variation across the cognitive domain in normal aging population and also to predict the cognitive decline in senile aging. This assessment tool will help the professionals to find out cognitive-linguistic performance of Malayalam speaking individual across different age, gender and four domains i.e. attention, perception and

discrimination, memory, problem solving and organization. The present study further corroborates the evidence to research in cognitive-linguistic performance and the factors contributing to it and opines to carry out extensive research in this area. Assessment of cognitive skills across age will give an insight into the senile versus senescence in older group. Profiling of various cognitive skills in elderly population may be useful in therapeutic intervention.

References

Allen, P. A., Madden, D. J., Groth, K. E., & Crozier, L. (1992). Impact of age, redundancy, and perceptual noise on visual search. Journal of Gerontology: Psychological Sciences, 47(2), 69-74.

Anuroopa, L. (2006). Development of Cognitive- Linguistic Assessment Protocol for Children (CLAP-C). Unpublished Masters Dissertation, University of Mysore, Mysore, India.

Arenberg, D. (1974). A longitudinal study of problem solving in adults. Journal of Gerontology, 29, 650-658.

Arenberg, D. (1982). Changes with age in problem solving. In F. 1. M. Craik and S. Trehub (Eds). Aging and cognitive processes. New York: Plenum Press.

Babcock, R. L., & Salthouse, T. A. (1990). Effects of increased processing demands on age differences in working memory. Psychology and Aging, 5, 421- 428.

Baddeley, A. (2001). Is working memory still working? American Psychologist, 851-864.

Baeckman, L., Small, B. J., & Larsson, M. (2000). Cognitive functioning in very old age. In F. I. M. Craik (Eds.), Handbook of Age and Cognition (2nd ed) (pp. 499-558). Mahwah, NJ: Lawrence Erlbaum Associates.

Bostrom, N., & Sandberg, A. (2009). Cognitive enhancement: methods, ethics, regulatory challenges. Science and Engineering Ethics, 15(3), 311-341.

Burke, D. M., & Light, L. L. (1981). Memory and aging: the role of retrieval processes. Psychological Bulletin, 90(3), 513-514.

Cabeza, R., Anderson, N. D., Locantore, J. K., & McIntosh, A. R. (2002). Aging gracefully: Compensatory brain activity in high-performing older adults. Neuroimage, 17(3), 1394-1402.

Cohen, G. (1979). Language abilities in elderly. In H. K. Ulatowska, M. P. Cannito, M. M. Hayashi, & S. G. Fleming (1985). In H. K. Ulatowska, (Eds.), The aging brain: Communication in elderly (pp. 125-140). London: Taylor & Francis.

Comfort, A. (1964). Aging: The Biology of Senescence. Routledge & Kegan Paul, London.

Connelly, S. L., Hasher, L., & Zacks, R. T. (1991). Age and Reading: The Impact of Distraction. Psychology and Aging, 6, 533-541.

Denney, N. W., & Denney, D. R. (1974). Modeling effects on the questioning strategies of the elderly. Developmental Psychology, 10(3), 458.

Denney, N, W. (1990). Adult age differences in traditional and practical problem-solving. In E.A Lovelace (Eds.), Aging and cognition: Mental processes, self-awareness

CLAP in Malayalam

119

and intervention (pp. 329-350). Amsterdam North-Holland: Elseiver Science Publishers,

Fontaine, R., & Toffart, L. (2000). Les prédicteurs des capacités de réserve cognitive chez lapersonne âgée. In D. Brouillet & A. Syssau. Le vieillissement cognitif normal, DeBoeck, Bruxelles.

Goswami, S. P., Shanbal, J. C., Samasthitha, S. & Navitha, U. (2010). Field testing of Manual for Adult Non-fluent Aphasia Therapy in Kannada. (MANAT-K). Research Project, All India Institute of Speech and Hearing Research Fund, Mysore, India.

Gregoire, J., & Van der Linden, M. (1997). Effects of aging on forward and backward digit span. Aging, Neuropsychology and Cognition, 4, 140-149.

Hochandel, G., & Kaplan, E. (2004). Theory and practice of neuropsychological assessment. In M. D. Lezak (3rd ed.). Neuropsychological assessment (pp. 292). Madison Avenue: Oxford University Press US.

Isella, V., Mapelli, C., Morielli, N., Pelati, O., Franceschi, M., & Appollonio, I. M. (2008). Age-related quantitative and qualitative changes in decision making ability. Behavioural Neurology, 19(1-2), 59-63.

Johnson, M, K., Reeder, J. A., Raye, C. L., & Mitchell, K. J. (2002). Second thoughts versus second looks: An age-related deficit in selectively refreshing just active information. Psychological Science, 13, 64-67.

Kamath, A. (2001). Cognitive-linguistic assessment protocol for adults. Unpublished Masters Dissertation, University of Mysore, Mysore, India.

Kausler, D. H. (1982). Experimental psychology and human aging. New York: John Wiley and Sons.

Kavya, V. (2007). Development of cognitive-linguistic assessment protocol for children with learning disability. Unpublished Masters Dissertation, University of Mysore, Mysore, India.

Lezak, M. D. (1995). Neuropsychological assessment (3rd Ed.). New York: Oxford University Press.

Light, L. (1990) . Interaction between memory and language in old age. In J. E. Birren & K. W. Schaie (Eds.), Handbook of the psychology of the aging (pp. 275-290). New York: Academic Press.

Madden, D. J. (1990). Adult age differences in attentional selectivity and capacity. European Journal of cognitive psychology, 2, 229-252.

Markman, E. M. (1989). Categorization and naming in children: Problems of induction. Cambridge, MA: MIT Press.

Mathuranath, P. S., Hodges, J. R., Mathew, R., Cherian, P. J., George, A., & Bak, T. H. (2004). Adaptation of the ACE for a Malayalam speaking population in southern India. International Journal of Geriatric Psychiatry, 19, 1188-1194.

Mendelson, J. R., & Ricketts, C. (2001). Age-related temporal processing speed deterioration in auditory cortex. Hearing Research, 158(1), 84-94.

Mitchell, K. J., Johnson, M. K., Raye, C. L., & D‟Esposito, M. (2000). FMRI evidence of age related hippocampus dysfunction in feature binding in working memory. Brain Research: Cognitive Brain Research, 10(1-2), 197-206.

Myerson, J. S., Robertson., & Hale, S. (2007). Aging and intra individual variability in performance: Analysis of

response time distributions. Journal of the Experimental Analysis of Behavior, 88(3), 319-337.

National Institute on Deafness and Other Communication Disorders. (2008). What is voice? What is speech? What is Language? Retrieved April 9, 2010,fromhttp://www.nidcd.nih.gov/health/voice/whatis_vsl.htm

Naveh-Benjamin, M. (2000). Adult age differences in memory performance: Tests of an associative deficit hypothesis. Journal of Experimental Psychology: Learning, Memory and Cognition, 26(5), 1170-1187.

Nelson, K. (1983). The derivation of concepts and categories from events representations. In E. K. Scholnick, New trends in conceptual representations: Challenges to Piaget's theory. (pp. 129-149). Hillsdale (N. J.): Erlbaum.

Nilsson, L. G. (2003). Memory functions in normal aging. Acta neurologica Scandinavica. Supplementum, 179, 7-13.

Nuegarten, B. (1974). The aging central nervous system: clinical aspects. In H. Brody, D. Harman & S. M. Ordy (Ed.), Aging (pp. 1-10). New York: Roven Press.

Nyberg, L., Cabeza, R., & Tulving, E. (1996). PET studies of encoding and retrieval the HERA model. Psychonomic Bulletin Review, 3(2), 135-148.

Ostberg, P., Fernaeus, S., Bogdanovic, N., & Wahlund, L. (2008). Word sequence production in cognitive decline: Forward ever, backward never. Journal of Logopedics Phoniatrics Vocology, 33, 126-135.

Partridge, L., & Mangel, M. (1999). Messages from mortality: the evolution of death rates in the old. Trends in Ecology and Evolution, 14(11), 438-442.

Pearson, D., Logie, R., & Gilhooly, K. J. (1999). Verbal representations and spatial manipulation during mental synthesis. European Journal of Cognitive Psychology, 11, 295-314.

Pennequin, V., & Fontaine, R. (2000). Training for older adults: the example of class inclusion. Journal of Adult Development, 2(2), 68-88.

Rabbitt, P. (1964). Set and age in a choice-response task. Journal of Gerontology, 19, 301-306.

Rangan, R., Nagendra, H. R., & Bhatt, R. (2009). Effect of yogic education system and modern education system on sustained attention. International Journal of Yoga, 2(1), 35-38.

Redfern, M. S., Muller, J. R., Jennings, J. M., & Furman (2002). Attentional dynamics in postural control during perturbations in young and older adults. The Journal of Gerontology, Series A, 57 (8), B 298.

Riegel, K. F. (1968). Changes in psycholinguistic performances with age. In G. A. Talland (Eds.), Human aging and behavior. New York: Academic Press.

Ryan, E. B., See, S. K., Meneer, W. B., & Trovato, D. (1994). Age-based perceptions of conversational skills among younger and older adults. In M. L. Hummert, J. M. Wiemann, & J. N. Nussbaum (Eds.), Interpersonal communication in older adulthood (pp.15-39). Thousand Oaks, CA: Sage Publications.

Sloutsky, V. M. (2003). The role of similarity in the development of categorization. Trends in Cognitive Sciences, 7, 246-251.

Smiley, S. S., & Brown, A. L. (1979). Conceptual preference for thematic or taxonomic relations: A non monotonic


120

age trend from preschool to old age. Journal of Experimental Child Psychology, 28, 249-257.

StrokEngine (2009). Assessment tools. Retrieved March 5, 2010 Retrieved from http://www.strokengine.ca/

Treitz, F. H., Heyder, K., & Daum, I. (2007). Differential course of executive control changes during normal aging. Aging Neuropsychology and Cognition,14, 370-393.

Ulatowska, H. K., Cannito, M. P., Hayashi, M. M., & Fleming, S. G. (1985). Language abilities in elderly. In H. K. Ulatowska, (Eds.), The aging brain: Communication in elderly (pp.125-140). London: Taylor & Francis.

Van der Linden, M., Hupet, M., & Feyereisen, P. (1999). Cognitive mediators of age-related differences in language comprehension and verbal memory performance. Aging, Neuropsychology, and Cognition, 6, 32-55.

Weinstein, B. E. (2003). Theoretical foundations of communication disability in aging. In L. E. Worrall & L. M. Hickson (Eds.), Communication disability in aging: from prevention to intervention: Healthy Aging (pp. 8-9).Clifton Park, NY: Cengage Learning.

http://www.strokengine.ca/

Development of RTT-M

121

Development of Revised Token Test in Malayalam Lincy Mary Varghese 1 & S.P.Goswami 2

Abstract

Aphasia is a multifaceted disorder. Many authors classified aphasia from different perspectives. Though mostly considered as a disorder of expression, aphasia encompasses a disorder of comprehension as an inability to understand linguistic utterances, which cannot be attributed to deficient sensory input or generalized cognitive deficits. Token test is a test which can be used to assess the subtle comprehension deficits in brain damaged individuals. It is an accurate and sensitive indicator of the presence of aphasia. The current study focused on developing Revised Token Test in Malayalam (RTT-M). Experimental participants were brain damaged adults (n=9) and controls were normal adults (n=40). Results revealed that as the complexity of the commands increased the performance decreased. Similarly older adults performed poorer compared to younger adults. Normal participants were better in comprehension as compared to the persons with aphasia on all subtests I to X. Factors such as sentence length, linguistic complexity, grammatical usage and number of critical stimulus in the commands influenced the test results. Key words: aphasia, Revised Token Test- Malayalam

phasia is 1the loss or impairment of language function caused by brain damage (Benson & Ardila, 1996). Many authors describe the same phenomena from a different angle and in fact they rather than contradict each other

(Kertesz, 1979). Different investigators classified aphasia differently. Rosenbek, La Pointe and Wertz (1989) defined disorder of comprehension as an inability to understand linguistic utterances, which cannot be attributed to deficient sensory input or generalized cognitive deficits. The most obvious aspect of a person with aphasia is the lack of oral, written or gestural output. However, the comprehension deficits in aphasia have been studied lesser than the expressive deficits. It can be possibly due to the fact that the analysis of comprehension needs to be based on observations of overt responses, which is usually confounded by the observed output deficits. Auditory comprehension can be impaired to varying degrees in each individual depending on the severity and type of the problem.

It is difficult to see an aphasic patient who can comprehend the spoken language with normal speed and accuracy. There are many factors which contribute to this comprehension difficulty and these factors are interacting. Some of these factors (speech sound and word meaning recognition) may be selectively impaired as a result of focal lesions, and may contribute to clearly defined aphasic syndromes. Other factors such as attention and short term auditory memory problems are more difficult to isolate and also they interact with the other two factors mentioned. Recent reports suggest working memory may account for language comprehension deficits in persons with aphasia (Caspari, Parkinson, LaPointe & Katz, 1998; Wright, Downey, Gravier, Love & Shapiro, 2007; Martin, Kohen & Kalinyak-

1e-mail: [email protected]; 2Reader in Speech pathology, AIISH, Mysore, [email protected].

Fliszar, 2008). Situational variables such as personal relevance and emotional significance of the subject matter may also contribute to the comprehension deficits. Basso, Capitani and Laiacona (1988) and Schuell, Jenkins and Landis (1961) reported that auditory comprehension deficits exists in all the cases of aphasia. It is difficult to detect the mild deficits in conversation.

In the assessment of comprehension the

stimulus material will be presented to the subject, and his ability to comprehend will be inferred based on the response. Generally, a test of comprehension consists of carefully worked out administrative procedures, and stimulus materials designed to elicit relatively simple responses.

Several tests for assessing the comprehension

abilities were developed by many authors over the years. Auditory Comprehension Test for sentences (Shewan, 1980), Discourse Comprehension Test (Brookshire & Nicholas, 1997), Reading Comprehension Battery for Aphasia-2 (LaPointe & Horner, 1998), Token Test (DeRenzi & Vignolo, 1962), Revised Token Test (McNeil & Prescott, 1978) are few among them. Token test is a test which can be used to assess the subtle comprehension deficits in brain damaged individuals. Token Test is an accurate and sensitive indicator of the presence of aphasia.

Though many tests are available for the

assessment of aphasia, most of them cannot be directly used in Indian population due to ethno-cultural barriers. Moreover, most of the existing tests cannot be used for the assessment of subtle deficits in auditory comprehension in persons with aphasia. Keeping this in view it was decided to adapt a test to assess the comprehension ability in Malayalam speaking persons with aphasia.

A


122

The current study aimed at developing Revised Token Test in Malayalam (RTT-M) and investigating the performances of the normal participants on the developed RTT-M. The main objectives of the study were to determine the overall performance of the normal participants in the various subtests, to examine the performances of males and females across subtests in different age groups and levels in various subtests of RTT-M in the different age groups and comparing normal population and the brain damaged individuals (persons with aphasia) on comprehension abilities on the test.

Method

Procedure: The „adaptation‟ of the test was done in three phases which included listing of the test stimuli, development of test material in Malayalam, and finally administration of the test battery on normals and persons with aphasia. Participants: The performance of a group of population on the RTT-Malayalam was the main aim. Individuals (normals and persons with aphasia) within the age range of 20-60 years were considered as the participants. The normal participants considered were native Malayalam speakers with no past/present history of any neurological, psychological problems, sensory deficits or any history of alcoholism or drug/abuse. The persons with aphasia were identified through local hospitals, neurological clinics and/or speech and hearing centers. All of them were native Malayalam speakers. No history of deterioration in cognitive abilities or sensory abilities was present. Procedure of test administration Arrangement and Placement: For all groups of participants and all subtests, the tokens were arranged on a standard table in front of the participant, and the order of arrangement was kept the same always. Arrangement of Seating: During testing the participant was seated in front of the table at a comfortable distance from where it was easy for the participant to reach and pick up the test material. The examiner sat to the left of the participant and slightly behind to avoid distractions that he/she might receive. Introduction of Subtest: Instructions were given to the participants prior to each subtest. The test was administered by giving the commands. The participant‟s behavior was observed during the command and response was rated based on a multidimensional 15-point scoring system taken from the original Revised Token Test developed by Mc Neil and Prescott (1978).

Scoring pattern: A multidimensional 15-point scoring system (McNeil & Prescott, 1978) was used to describe performance, and quantify deficits and differences among normal and pathological groups. Score Sheet: The participant‟s demographic data (name, age, and sex), handedness, diagnosis, age of onset etc. were included in the score sheet. The summarization and accessibility of the overall test time mean overall score for all subtests, and the mean for each individual subtest was included in the score sheet. Statistical analysis: SPSS (Statistical Package for the Social Sciences, version 16.0) software was used for statistical analysis. The tabulated scores were used for obtaining the mean (M) and standard deviation (SD). Parametric tests were utilized to obtain the significant difference measures. Mixed ANOVA was used to obtain significant difference between various subtests and across various age groups. Interaction effect between the subtests and groups were compared using the repeated measures ANOVA. A Multivariate Analysis of Variance (MANOVA) was performed to compare subtest across groups.


The results obtained from the data were analyzed on various aspects. The findings of the present study have been broadly presented under the following headings.

(I) Performance of males and females across subtests in different age groups

Subtest I (SI) The mean scores of male and female

participants in S I subsection of RTT is shown in Table 1. The mean values of both male and female participants of the age group 20-30, 30-40, 40-50 and 50-60 years was 15.00. In the subtest I, the male and female participants scored a mean of 15.00. The total mean score was also found to be 15.00. It is indicated that the task on this subtest was easy and did not require enough effort from the participant‟s side in its comprehension. All the participants executed a complete response.

Subtest II (SII) In the male participants the lowest value was

obtained for the 30-40 age group with a mean of 14.97 (SD= 0.05), whereas the other three groups, 20-30, 40-50 and 50-60 age groups obtained a score of 15. While considering the female participants it is observed that all the age groups scored 15. All the male participants secured a mean of 14.99 (SD=0.02), parallel to which all the females obtained a mean score of 15 in all the age groups for Subtest II. Their total mean score on this subtest was 14.99 (SD=0.02). The participants were mostly complete in


123

their responses, with a few of them going down to the level of sub-vocal rehearsal.

Subtest III (S III) A minimum of 14.18 (SD=0.51) and a

maximum of 15.00 in the age ranges of 50-60 and 20-30 years of age was scored by the male participants and 13.81 (SD=0.70) and 14.95 (SD=0.09) in the age ranges 50-60 and 20-30 years of age were scored by the female participants. The means of other groups fell between these boundaries. A total score of 14.63 (SD=0.44) and 14.52 (SD=0.61) was obtained by male and female participants of all the age groups and they obtained a total mean score of 14.57 (SD=0.53) for the Subtest III. The responses of the male and female participants were mainly in the form of sub-vocal rehearsals and complete ones with the female participants of the oldest age group of 50- 60 years showing some delayed responses.

Subtest IV (S IV) A score of nearly equal to 15 was obtained in

all the age groups in the male participants, except for 50- 60 years, (41.11) which signifies that the responses were between sub-vocal rehearsals and complete. The female participants in the subtest scored between 13.76 (SD=0.67) to 14.65 (SD=0.24), which indicated that the responses were sub-vocal rehearsals in almost all of the participants. The male participant of all the age groups across Subtest IV (S IV) had a mean score of 14.56 (S=0.46) while for female participants mean value was 14.23 (SD=0.59). The total mean score for the participants across S IV was 14.39 (SD=0.55). The responses of the male participants were scored as sub-vocal rehearsals and those of females as sub-vocal rehearsals and delayed.

Subtest V (S V) In S V subtest of RTT the males obtained a

highest mean score of 14.74 (SD=0.26) in the age group of 20-30 years and a least of 13.87 (SD=0.51) in the age ranges of 50-60 years. Similarly, the females in the age group of 50-60 years scored poorer to the males of the same age group with a score of only 13.78 (SD=0.77). The female participants in the age group of 30-40 years scored the highest mean of 14.64 (SD=0.27). The female participants seemed to perform comparatively poorer than the male participants. The responses of the male and female participants were mainly subvocal rehearsals, whereas all the participants in the age group of 50-60 years exhibited responses which were in the form of delayed ones.

Subtest VI (S VI) The mean scores of 20-30, 30-40, 40-50 and 50-60 years group was 14.43 (SD=0.25), 14.41 (SD=0.50), 14.56 (SD=0.26), 13.85 (SD=0.38) respectively with the least being scored by the oldest

age group. The mean scores of females in the age groups 20-30, 40-50 and 50-60 years (14.14, SD=0.24; 14.30, SD=0.57; 13.52, SD=0.74) was poorer in comparison to the males and that of the 30-40 years (14.40; SD=0.33) was better. The male and female participants of all the age groups on subtest VI pulled up an overall total mean score of 14.31 (SD=0.43) and 14.09 (SD= 0.58) respectively. They secured an overall total mean score of 14.20 with a SD of 0.52. This depicts that the responses were predominantly sub-vocal rehearsal type with few of them with a delayed response. Subtest VII (S VII) The male participants in the subtest scored between 13.57 (SD=0.57) to 14.56 (SD=0.46), which indicated that the responses obtained were sub-vocal rehearsals in almost all of the participants. A score of 14 was attained in all the age groups, except the older group (where the score is 13) in the female participants, which signified that the responses were between delayed responses and sub-vocal rehearsals. The male participants of all the age groups across subtest VII (S VII) had a mean score of 14.11 (S=0.58) while for female participants mean value was 14.02 (SD=0.69). The total mean score for the participants across S VII was 14.07 (SD=0.63). Most of the participants in both the genders performed sub-vocal rehearsals on this subtest which demanded left and right postpositional task that comparatively puts greater load on the participants‟ comprehension than that of above subtests (except subtest I and II).

Subtest VIII (S VIII) In the S VIII subtest, the males scored a mean score of 14.24 (SD=0.38) and 14.45 (SD=0.56) in age group of 20-30 and 30- 40 years respectively and an even lower score of 13.79 (SD=0.51) and 13.19 (SD=0.42) in the higher age groups, 40- 50 and 50- 60 years. Similarly in the female participants the mean score was equivalent to the score of 20-30 and 30-40 years male participants and the scores of other two age groups were 13. All the female participants of all the age groups across the subtest VIII (S VIII) scored a mean of 13.72 (SD=0.68), while the male participants obtained a mean score of 13.92 (SD=0.66). 13.82 (SD=0.67) was the total mean score of all the participants of all the age group on S VIII. The female participants seemed to perform comparatively poorer than the male participants. The responses of the male and female participants were mainly delay and subvocal rehearsals.

Subtest IX (S IX) A score of nearly equal to 15 was obtained in the age groups of 30-40 and 40-50 years in the female participants, which signifies that the responses were between sub-vocal rehearsals and complete. The scores of male participants in the age range 50-60 years were poorer (mean=13.99;


124

SD=0.50) than the lower age groups. Likewise the mean values of age groups 20-30, 30-40 and 40-50 years in case of females performed with sub-vocal rehearsal to complete responses, scoring a mean ranging from14.48 (SD=0.54) to14.75 (SD=0.33). The female participants in the 50-60 years age groups were much lower than this (mean 11.21; SD=5.41). The male participant of all the age groups across Subtest IX (S IX) had a mean score of 14.44 (SD=0.48) while for female participants mean value was 13.75 (SD=02.92). The total mean score for the participants across S IX was 14.09 (SD=2.10). The responses of the male and female participants were mainly in the form of complete type, subvocal rehearsal and delayed with few of the older female participants (50-60 years) giving self correction responses.

Subtest X (S X) The lowest value in the male participants was obtained for the 50-60 age group with a mean of 13.92 (SD=0.41). While the 30-40 age group scored the highest of 14.71 (SD=0.52). At the same time the female participants it was observed that the younger age groups 20-30, 30-40, and 40-50, (14.62, SD=0.28; 14.72, SD=0.34; 14.50, SD=0.42 respectively) obtained better scores than the older age group of 50-60 years (13.53, SD= 0.58) (which is a much higher score in comparison to the same age group scoring 11.21 (SD=5.41) on subtest IX. On a careful examination it was evident that the higher groups performed better on subtest X with respect to subtest IX. All the male participants secured a mean of 14.47 (SD=0.48), parallel to which all the females obtained a mean score of 14.34 (SD=0.62) in all the age groups for Subtest X. Their total mean score on this subtest was 14.40 (SD=0.55). Most of the male and female participants in the other age groups (20-30, 30-40, and 40-50 years) except 50-60 age group were scored as sub-vocal rehearsal and complete on the adverbial clauses. At the same time the male and female participants in the higher age groups were scored as delayed. Overall On the overall mean the male and female participants were comparable across age groups with the scores of male participants‟ ranging between 14.07 (SD=0.24) (50- 60 years) to 14.69 (SD= 0.37) (20- 30 years) and that of female participants ranging between 13.58 (SD=0.73) (50- 60 years) to 14.69 (SD=0.37) (20- 30 years). For the middle age groups 30- 40 and 40-50 years the scores were between these ranges (in males 14.65 (0.2); 14.53 (0.15) and in females 14.45 (0.27), 14.43 (0.27) for age groups respectively, thus revealing the fact that the performances on the auditory comprehension tasks declined as age increased. For all the subtests the overall total scores were summed up and the mean and standard deviation (SD) were calculated for all

the dependant variables i.e. subtest of RTT-M. An overall mean score of 14.48 (SD=0.34) was obtained for the male participants of all the age groups. Similarly a mean score of 14.29 (SD=0.56) was illustrated by the female participants. For both males and females a total mean score was 14.39 with a SD of 0.47 was obtained. It can be inferred from the above mean values that, on an average most of the participants obtained either a mean of 15.00 or 14.00, indicating that the responses were either complete or in the form of sub-vocal rehearsal with very few participants showing delayed or immediate responses. It is clear from the responses of the participants that their performances differed qualitatively across the age groups. The younger age groups 20- 30 and 30- 40 years responded mostly in a normal manner without needing any extra information to perform the task. Most of the participants demonstrated a complete type and few of them on rare occasions showed sub-vocal rehearsals. Whereas in the older age groups (40- 50 and 50- 60 years), few of the participants even scored as poorer as delayed and immediate type of responses. In general the normal responses across the test varied between complete and subvocal rehearsal, thereby paving the idea that the Revised Token Test involves tasks that are well suited to identify an individuals‟ auditory comprehension levels. A Mann-Whitney U-test was conducted within each age group for all the subtests and the overall mean to detect the differences between them. Mann-Whitney U-test revealed that there was no significant difference between the males and females on any of the subtest or on groups at p<0.05 (i.e. the males and females performed equally well on all subtests). Due to this reason, males and females were combined and considered as one single set, for further analyses i.e. consideration of ten participants instead of five males and five females in each age group. Furthermore, there were only five participants of each gender in each age group (which is not a large sample to be considered). Hence gender was not regarded as an independent variable on all advanced analyses. The test encapsulates a series of cognitive processes as well to auditory comprehension, including working memory (Lesser, 1976; Smith, Mann & Shankweiler, 1986), analysis of the whole into a series of items, or the ability to adequately ignore automatically evoked, distracting stimuli. Such factors are evidently distinct in both the gender. Parietotemporal metabolism (Karbe, Herholz, Szelies, Pawlik, Wienhard & Heiss, 1989) could have affected the performance of the participants in this test. Another factor is the measures of language production (Gutbrod, Meger, Meter & Cohen, 1985).


125

Hence, the present study gives evidences regarding the gender variation in normal persons that may not be a major variable in comprehension task in auditory mode only. No differences in the performances of the two groups were seen, and if present, may be subtle in nature.

(II) Overall performance of the participants in the various subtests

The performances of female and male participants were not significantly different across groups and subtests; hence the following comparisons were made using repeated measure analysis of variance:

(a) Comparison of the performance on subtests A significant difference was observed

between the subtests [F (9,234) =10.937, p<0.05] on mixed ANOVA (repeated measures ANOVA with age as independent factor). The mean and standard deviations of age groups across each subtest are given in Table 1 and Figure 1. 30-40 years and 50-60 years age group scored the highest and the lowest respectively. The age groups 20-30 years and 40-50 years scored in between.

The results of the present study receives

support from Swisher and Sarno (1969); DeRenzi and Faglioni (1978); Emery (1986); Ivnik, Malec, Smith, Tanglos and Peterson (1996). Many

researchers suggested that the differences in scores exhibited by the older individuals could be due to the difficulty in retaining the auditory stimuli for a longer duration in comparison to the younger individuals.

The deterioration in comprehension from the

S I to S X subsections in a gradual fashion is due to minimal redundancy, where the participants are required to understand the significance of each word in a series of increasingly complex commands (Goswami, 2004). Moreover, the linguistic stimuli are presented only in the verbal mode on RTT; hence the participants need to rely completely on the auditory mode for comprehending the stimuli. Sub-vocal rehearsals were observed beside the complete responses in few of the normal participants, which indicate that they rely on their auditory feedback and even sub-vocal rehearsals also help in retaining the linguistic stimuli for a longer duration.

On mixed ANOVA, a significant difference

was observed between the performances of various subtests [F (9, 324) =12.68, p<0.05]. Therefore, subsequent to this a post- hoc Bonferroni test for pairwise comparison was done to find out the subtests which differed significantly. This is being depicted in Table 2.

Table 1. Mean and standard deviation (SD) of groups across

Subtests 20-30 years 30-40 years 40-50 years 50-60 years Mean

I Mean 15.00 15.00 15.00 15.00 15.00 SD 0.00 0.00 0.00 0.00 0.00

II Mean 15.00 14.98 15.00 15.00 14.99 SD 0.00 0.04 0.00 0.00 0.02

III Mean 14.97 14.84 14.49 14.00 14.57 SD 0.06 0.29 0.36 0.61 0.53

IV Mean 14.63 14.67 14.35 13.93 14.39 SD 0.36 0.40 0.54 0.59 0.55

V Mean 14.48 14.60 14.39 13.82 14.33 SD 0.35 0.33 0.30 0.62 0.50

VI Mean 14.28 14.41 14.43 13.69 14.20 SD 0.27 0.40 0.44 0.58 0.52

VII Mean 14.23 14.57 14.08 13.39 14.07 SD 0.29 0.36 0.56 0.62 0.63

VIII Mean 13.93 14.40 13.84 13.10 13.82 SD 0.49 0.45 0.57 0.47 0.67

IX Mean 14.41 14.74 14.63 12.60 14.09 SD 0.37 0.44 0.32 3.90 2.10

X Mean 14.59 14.72 14.58 13.72 14.40 SD 0.29 0.40 0.35 0.51 0.55

Overall Mean 14.55 14.69 14.48 13.82 14.39 SD 0.13 0.28 0.18 0.57 0.47


126

On a pairwise comparison between subtests the following results were evident. The scores of:- 1. S I was significantly different from S III, S IV, S

V, S VI, S VII, S VIII and SX; 2. S II from S III, S IV, S V, S VI, S VII, S VIII

and SX; 3. S III from S V, S VI, S VII, and S VIII; 4. S IV from S VII and S VIII; 5. S V from S VIII; 6. S VI from S VIII; 7. S VII from S VIII, and S X; 8. S VIII from S X; and vice versa.

The obvious differences could be ascribed to the fact that, the increasing complexity of the tasks in terms of grammaticality brings about a difference in the overall scores.

Figure 1. Mean scores of the different age groups on various subtests.

It is evident from the table that the adjacent

subtests are not significantly different from each other except for S VII and S VIII. The difference in the complexity of tasks, increase in sentence length are some of the very reasons the scores of the pair of odd and even subtests are not differing, except for S VII and S VIII, which means that the commands of S VII (which involve left-right postpositional phrase) demands the participant to comprehend only six critical items, while S VIII involves eight critical items. The level thus becomes the toughest of all these subtests.

Several studies have reported that auditory

comprehension in individuals is affected by factors such as effect of sentence length (Levy & Holland, 1971; Brookshire, 1974; Pierce & Wagner, 1985) and grammatical complexity (Lasky, Weidner & Johnson, 1976). Goswami (2004) even puts down the fact that the more complex a sentence is grammatically, the more difficult is its comprehension.

(b) Comparison of the performance on age groups A significant difference on Mixed ANOVA

was observed between the performances of various age groups for [F(3, 36)=12.71, p<0.05] as the participants of the various age groups exhibited comprehension deficits to varying degrees of severity. Post-hoc Duncan‟s mean range test elaborated on the differences between the various age groups. Figure 2 depicts the difference in the overall scores between the age groups.

Figure 2. Overall mean on all the subtests as produced by the different age groups.

On the post hoc test the age groups 20-30, 30-

40, and 40-50 years performed similarly on the test (no significant difference in all the three age groups) whereas they differ significantly from the older age group of 50-60 years.

Older adults show a greater reliance on target

word activation for word selection while in younger adults, activation and inhibition mechanisms are tightly linked during auditory word comprehension in a dual mechanism context. It appears that inhibitory mechanisms become less involved in auditory comprehension with aging. Age-related effects were reported as few in a series of papers on Token Test and its versions (Swisher & Sarno, 1969; DeRenzi & Faglioni, 1978; Ivnik et al., 1996).

(c) Interaction of the subtests and groups Interaction effect between the subtests and groups on repeated measures ANOVA, was evident. Results revealed that a significant interaction was present between them at [F (27, 324) =1.96 for p<0.05[. (III) Levels in various subtests of RTT-M in the different age groups

Repeated measures ANOVA revealed a significant interaction effect between age groups and subtests, hence comparisons were made to determine the age group differences varying in magnitude across the subtests and vice versa.

(a) Comparison of subtest across groups

Multivariate Analysis of Variance (MANOVA) was performed across all subtests. The


127

Table 2. Pairwise comparison between subtests Note: Cells marked with a “+” indicate a condition in which significant difference was obtained on a given subtest from the corresponding subtest. Cells marked with a “-“ indicate a condition in significant difference was not obtained on pair wise comparisons. MANOVA revealed statistically significant differences on subtests. Table 4 lists the ten subtests for which a significant difference was found for the total scores. It is clear from Table 4 that all the subtests and the overall mean, except subtest II and IX present with a statistically significant difference at p<0.05 (with age as independent variable and dependent variables as subtests S I to S X and overall mean). The distributions of grammatical complexity suggest that all the subtests except I and II, have a slightly different complexity in their grammatical structure, which could have possibly contributed to the significant difference in scores in comparison to the others.

Further post- hoc Duncan Mean Range test elaborated on the differences between the age groups on each of the significant difference of the RTT. It can be summarized as follows:

On Subtest III, age groups 20-30 years and

30-40 years and age groups 30-40 and 40-50 years performed similarly whereas the performance of higher age group 50-60 years, varied from the lower age groups significantly i.e. the adjacent age groups were similar in their performance. Age groups 20-30 years, 30-40 years and 40-50 years performed comparably with no significant difference in scores among them, on subtest IV. The scores of age groups 40-50 and 50-60 also did not differ significantly at 0.05 levels. The scores of all the age groups except the higher age group (50-60 years) performed similarly with a statistically significant score than the 50-60 age group on subtest V and subtest VI.

Participants in the 20-30 and 30-40 years age groups and 20-30 and 40-50 years age groups performed similarly. While the scores obtained for the older age group (50-60 years) differed significantly from the

Table 3. Tests of between subjects effects others

younger age groups on subtest VII. On subtest VIII, the performance of 20-30 and 40-50 years age groups were comparable with no significant difference among them. The scores obtained for 30-40 years age group was significantly different from the other age groups. The scores of the age group 50-60 years was not comparable to the younger age groups i.e. the scores of 50-60 years were significantly different from the younger age groups on subtest X as well as on overall mean. (b) Comparison of groups across subtests

Repeated measure of ANOVA was conducted for each of the age groups to determine the difference in subtest scores, if any: 1. Tests of within-subjects effects for 20-30 years age group revealed that the scores across subtests differed at 0.05 level of significance. Hence, post-hoc pairwise comparison on Bonferroni test was performed to find out the differences between the subtests. The scores on various subtests in 20-30 years age group:

Subtests S I S II S III S IV S V S VI S VII S VIII S IX S-X

S I - + + + + + + - + S II - + + + + + + - + S III + + - + + + + - - S IV + + - - - + + - - S V + + + - - - + - - S VI + + + - - - + - - S VII + + + + - - + - + SVIII + + + + + + + - + S IX - - - - - - - - - S-X + + - - - - + + -

Subtests F (3,36) Sig. S I - - S II 1.00 0.40 S III 12.57 0.00* S IV 4.77 0.00* S V 6.76 0.00* S VI 6.24 0.00* S VII 10.62 0.00* S VIII 11.40 0.00* S IX 2.58 0.06 S X 13.02 0.00*

Overall Mean 12.71 0.00* *p<0.05


128

1. S I was significantly different from S VI, S VII, S VIII, and S IX; 2. S II from S VI, S VII, S VIII, and S IX; 3. S III from S VI, S VII, S VIII, and S IX; 4. S IV from S VII; and vice versa. 2. A test of within-subjects effects for 30-40 years was done in order to find out the difference within the subtests at 0.05 level of significance. The scores were statistically significant [F (9, 81) = 9.36 at p<0.05]. In order to find out the differences between the subtests in the 30-40 years age group, the data was subjected to Bonferroni test. The scores on various subtests in 30-40 years age group: 1. S III was significantly different from S VI; 2. S VIII from S IX and S X; and vice versa. 3. A significant difference was identified within the subtests in the age group of 40-50 years on the repeated measure ANOVA [F (9, 81) =9.14, p<0.05]. Differences in the scores were identified on Post hoc Bonferroni test. 1. S I was significantly different from S V, S VII and S VIII . 2. S II from S V, S VII and S VIII; and vice versa. 4. In the age group of 50-60 years a significant difference was identified within the subtests on the repeated measure ANOVA [F (9, 81) = 3.52, p<0.05]. Subsequently a Post hoc Bonferroni test was done and identified significant differences in the scores, which is shown in Table 3. 1. S I was significantly different from S III, S IV, S V, S VI, S VII, S VIII, and S-X; 2. S II from S III, S IV, S V, S VI, S VII, S VIII, and S-X; 3. S IV from S VII, and S VIII; 4. S VII from S III; 5. S VII from S III; and vice versa. It can be summarized that there was a significant difference between the subtests in all the our age groups. It can be reasoned that the linguistic complexity and sentence length increases gradually from S I to S X. As discussed above, another

Figure 3. Mean scores of normals and persons with aphasia.

attributable factor to this result is that the subtests I (and II) contribute to lesser comprehension demand on the participants in comparison to the other subtests

in terms of the sentence length, complexity (Goswami, 2004) and number of critical stimulus units. (IV) Comparison of normal population and the brain damaged individuals (persons with aphasia) on comprehension abilities

The overall mean scores on all the subtests by the normals and aphasics and their individual scores on the subtests is depicted in Figure 3. Normal participants were better in comprehension as compared to the persons with aphasia on all subtests I to X. The mean scores in each subtest in the normal participants‟ decreased with complexity of tasks. The mean scores in subtest I and subtest VIII were 15.00 and 13.82 and 10.16 and 5.92 respectively in normal and brain damaged participants. From the scores of both the groups of participants (the normals and brain damaged), it was observed that highest mean scores were observed in subtest I and lowest in subtest VIII.

Table 4. Results of Mann Whitney- U test (Comparison of normals and aphasics)

Geschwind (1965) and Caramazza and Zurif (1976) have attributed reasons to the brain damage in aphasics, which causes a deficit in individual‟s ability to comprehend linguistic stimuli. Mann-Whitney U-test (Table 4) showed that RTT scores for the persons with aphasia were significantly poorer than those of the normal participants.

The sentence comprehension tends to decrease

as length increases (Goswami, 2004). Improvements in sentence comprehension of aphasics following training to improve verbal memory span was reported by Francis, Clark & Humphreys (2003). Deficits in all of these cognitive processes in comprehending grammatical elements have been implicated in persons with aphasia (Schuell, Jenkins & Jimenz-Pabon, 1964; Brookshire, 1974). The repetition of linguistic command also improved the performance of these persons in the form of sub-vocal rehearsals, Goswami (2004).

Subtests |Z| Sig. S I 6.88 0.00* S II 6.60 0.00* S III 4.70 0.00* S IV 4.65 0.00* S V 4.64 0.00* S VI 4.64 0.00* S VII 4.64 0.00* S VIII 4.65 0.00* S IX 4.42 0.00* S-X 4.65 0.00*

Overall Mean 4.64 0.00* *p<0.05


129

Conclusions

The results of the study can be summarized in the following manner. There was no difference in performances of males and females across subtests. It was evident that all the subtests except Subtest II and IX differed significantly from the other subtests on the RTT-M. The age groups 20-30, 30-40, and 40-50 years (younger age groups) performed almost similarly and were different in their scores from the 50-60 years (older age group). A reduction in the performance of the participants was observed as the linguistic complexity and sentence length increased. Normal participants showed better comprehension as compared to the aphasics (persons with aphasia) on all the subtests (subtests I to X).

The RTT proves to be one of the most useful clinical tools that help in terms of assessing an individual‟s auditory comprehension. It also assesses the comprehension on stimuli of gradual complexity. Adaptation of RTT in Malayalam, thus, put-forth the importance of a thorough assessment of comprehension abilities and its implication in assessing comprehension among the different aphasic types.

It can be concluded that the performance of

the males and females was similar across all the subtests on RTT-M. The gender of a person seems not a crucial factor in the comprehension of spoken messages to any significant extent. The deficits in auditory comprehension varied to different degrees, depending on the linguistic length and complexity on the RTT-M. Qualitative differences in the responses of the normal participants were also observed. The responses of normal participants on the RTT-M were mostly complete (score of 15) or vocal-sub vocal rehearsals (score of 14). It can be inferred that on assessment, if a scores 14 or 15 on commands for a native normal participant, mean values and/or subtests, his performance is in a normal manner. A score less than 14 indicate a deficit in the auditory comprehension skills of the person. Factors such as sentence length, linguistic complexity, grammatical usage and number of critical stimulus in the commands contributed to the test results. All the scores except for subtest II and subtest IX were evidenced to differ from that of subtest I. Aging, memory, and attention span of the participant also influence the responses. The younger age groups were better in their performances in comparison to the older age groups. The performances of aphasic participants were relatively poorer than the normal participants. Cognitive and/or auditory processing deficits have been implicated from the responses of most of the aphasics.

It can be concluded from the results of RTT-

M that the difference in performances within normals and within aphasics points to the effectiveness of the

test in fulfilling its purpose of assessing auditory comprehension. The linguistic competence in normals as well as in persons with aphasia can be assessed using the test.

References

Basso, A., Capitani, E., & Laiacona, M. C. (1988). Progressive language impairment without dementia: A case of isolated category specific naming defect. Journal of Neurology, Neurosurgery and Psychiatry, 51, 1201-1207.

Benson, F. D., & Ardila, A. (1996). Aphasia: a clinical perspective. New York, Oxford University Press.

Brookshire, R. H. (1974). Differences in responding to auditory verbal materials among aphasic patients. Acta Symbolica, 1, 1-18.

Caramazza, A., & Zurif, E. B. (1976). Denunciation of algorithmic and heuristic processes in language comprehension: Evidence from aphasia. Brain and Language, 3, 572-582.

Caspari, I., Parkinson, S. R., LaPointe, L. L., & Katz, R. C. (1998). Working memory and aphasia. Brain and Cognition, 37, 205-223.

DeRenzi, E., & Faglioni, P. (1978). Development of a shortened version of the Token test. Cortex, 14, 41-49.

Emery, O. B. (1986). Linguistic decrement in normal aging. Language and Communication, 6, 47–64.

Francis, R. D., Clark, N., & Humphreys, W. H. (2003). The treatment of an auditory working memory deficits and the implications for sentence comprehension abilities in mild “receptive” aphasia. Aphasiology, 17 (8), 723-750.

Geschwind, N. (1965). Disconnexion syndromes in animals and man. Brain, 88, 585-644.

Geschwind, N., & Fusillo, M. (1966). Colour naming deficits in association with alexia. Archieves of Neurology, 15,137-146.


Gutbrod, K., Meger, B., Meter, E., & Cohen, R. (1985). Cognitive processing of tokens and their description in aphasia. Brain and Language, 25, 37–51.

Ivnik, R. J., Malec, J. F., Smith, G. E., Tangalos, E. G., & Petersen, R. C. (1996). Neuropsychological tests norms above age 55: COWAT, BNT, MAE Token, WRAT-R, Reading, AMNART, Stroop, TMT, and JLO. The Clinical Neuropsychologist, 10, 262–278.

Karbe, H., Herholz, K., Szelies, B., Pawlik, G., Wienhard, K. K., & Heiss, W. D. (1989). Regional metabolic correlates of Token test results in cortical and sub-cortical left hemisphere infarction. Neurology, 39, 1083–1088.

Kertesz, A. (1979). Aphasia and associated disorders: Taxonomy, localization and recovery. New York: Grune and Stratton.

LaPointe, L. L., & Horner, J. (1998). Reading comprehension battery for aphasia (2nd Ed.). Austin, TX: Pro-Ed.

Lasky, E. Z., Weidner, W. E., & Johnson, J. P. (1976). Influence of linguistic complexity, rate of presentation and inter-phrase pause time on auditory verbal comprehension of adult aphasic patients. Brain and Language, 3, 386-395.


130

Lesser, R. (1976). Verbal and non-verbal memory components in the Token test. Neuropsychologia, 14, 79–85.

Levy, C., & Holland, A. (1971). Influence of grammatical complexity and sentence length on comprehension with adult aphasics. In A. Holland (Ed.), Psycholinguistics and behavioural variables in recovery from aphasia. Final report submitted to Social and Rehabilitative Service, DHEW, 1-26.

Martin, N., Kohen, F., & Kalinyak-Fliszar, M. (2008). A diagnostic battery to assess language and shortterm memory deficits in aphasia. Poster presentation at Clinical Aphasiology Conference, Teton Village, WY.

McNeil, M., & Prescott, T. (1978). Revised Token Test. Austin, TX: ProEd.

Pierce, R. S., & Wagner, C. (1985). The role of context in facilitating syntactic decoding in aphasia. Journal of Communication Disorders, 18, 203-213.

Rosenbek, J. E., LaPointe, L., & Wertz, T. R. (1989). Aphasia: a clinical approach, Austin, Texas: Pro-ed.

Schuell, H. M., Jenkins, J. J., & Jimenz-Pabon, E. (1964). Aphasia in adults. New York: Harper & Row.

Schuell, H., Jenkins, J. J., & Landis, L. (1961). Relationship between auditory comprehension and word frequency in aphasia. Journal of Speech and Hearing Research, 4, 30-36.

Shewan, C.M. (1980). Auditory comprehension test for sentences. Chicago, Biolinguistics Clinical Institutes.

Smith, S. T., Mann, V. A., & Shankweiler, D. (1986). Spoken sentence comprehension by good and poor readers: A study with the Token test. Cortex, 22, 627–632.

Swisher L. P., & Sarno M. T. (1969). Token test scores on three matched patient groups: Left brain-damaged with aphasia; right brain-damaged without aphasia; non brain-damaged. Cortex, 5, 264–273.

Wright, H. H., Downey, R. A., Gravier, M., Love, T., & Shapiro, L. P. (2007). Processing distinct linguistic information types in working memory in aphasia. Aphasiology, 21, 802–813.

Agrammatism in children with mental retardation

131

Agrammatism in Children with Mental Retardation Liveem Mariam Tharakan 1 & Shyamala K. C2

Abstract

Children master the intricacies of their native language before they are able to tie a knot, jump rope, or draw a circle. This achievement is so expected that most people rarely give a second thought about how children accomplish this remarkable feat. The child must be aware of and use the rules of syntax before one can be said to have adequate command of language. One group of children in whom the development of language does not occur as expected are children with mental retardation. These children fail to acquire language in the normal course, are bound to be at disadvantage in many aspects of their living. The present study is on the grammatical structures used by children with mental retardation. Language samples of thirty typically developing children and language age matched children with mental retardation were analyzed using Index of Productive Syntax (Scarborough, 1990). While adapting the manual to Malayalam language few grammatical constructions were removed and few modifications were made based in the grammatical structure of Malayalam language. It could be concluded from the present study that in children with mental retardation a number of grammatical structures fail to emerge but develop in the same way as seen in typically developing peers. A period of agrammatic language production is evident in the spoken output of children with mental retardation.

Key words: agrammatism, Index of Productive Syntax, mental retardation

anguage is so 1intricately woven in to human life that it is expected to be present throughout and we seldom think about its emergence and

development. Language development is a process starting early in human life. Language starts off as recall of simple words without associated meaning, but as children grow, words acquire meaning, with connections between words being formed. As a person gets older, new meanings and new associations are created and vocabulary increases as more words are learned.

Children begin to combine words to create

simple two word utterances, also called telegraphic speech. Gradually, children‟s sentences increase in length as they add more semantic and grammatical elements. This growth is measured in terms of child‟s mean length of utterances (MLU; Brown, 1973). At the early stages of language development, open class categories dominate the child‟s vocabulary (nouns, verbs, adjectives, adverbs) but as MLU grows, children use increasing closed-class terms such as pronouns, prepositions, conjunctions. At around 3 years, children engage in simple sentences, which are 3 word sentences. Simple sentences follow adult rules and get refined gradually. Grammatical morphemes get added as these simple sentences start to emerge. By 3–5 years, children continue to add grammatical morphemes and gradually produce complex grammatical structures. By 6–10 years, children refine the complex grammatical structures such as passive voice.

1e-mail: [email protected]; 2Professor of Language Pathology, AIISH, Mysore, [email protected].

The current view of the language performance of retarded children holds that they acquire language in the same sequence as normal children but at a slower rate (Miller, Chapman & Bedrosian, 1977). The complexities of grammar tend to be difficult for mentally retarded children.

Given the complex nature of language and

communication, there are a number of structural and functional aspects of the language system to be assessed. Language sample analysis is an appealing assessment tool since it has superior sensitivity and specificity in identifying children with language impairments (Aram, Morris & Hall, 1993).

Language sampling in natural situations and

descriptive analysis of the same are yet to be attempted with disordered population like mental retardation. Various assessment methods have been used to account for grammar development in the English language. There is a dearth of such studies with regard to Malayalam language. This study will describe and try to account for the agrammatism in Malayalam speaking mentally retarded children. Hence the study was planned.

The first aim of the study was to adapt Index of Productive Syntax (Scarborough, 1990) into Malayalam and to use this broad descriptive analysis of the Index of Productive Syntax (Scarborough, 1990) to account for agrammatism in Malayalam speaking mentally retarded children in 3 different age groups: 3- 4 years, 4-5 years and 5-6 years. Also we compared the syntactic development seen in these children with their typically developing peers. This study also aimed at studying the correlation of Malayalam Language Test (MLT, Rukmini, 1994) and IPSyn (Scarborough, 1990).

L


132

Method

Participants: Language samples were collected from thirty native Malayalam speaking typically developing children in the age range of 3-6 years and thirty language age matched children with mental retardation. The children were placed into three groups with 10 children in each group according to their language age. The groups were: Group 1: 3-4years, Group 2: 4-5 years and Group 3: 5-6 years.

The subject selection criteria followed for typically developing children was that they should have no sensori-motor deficits, no cognitive impairments, no social or emotional deficits, no history of any sort of speech and language disturbances and also non-incidence of dyslexia in their families. Language age matched thirty Malayalam speaking children with mental retardation constituted the other group. Only those children in whom the data on psychometric evaluation was available were selected. Children in this group had IQ between 70 to 35 based on standard scores of intelligence tests, no associated conditions like cerebral palsy, no neurologic problems and no impairment in vision and hearing.

Out of the thirty children with mental retardation, eighteen of them were diagnosed to have mild mental retardation while twelve children had moderate mental retardation. Thirteen children had attended therapy for more than one year while seventeen children attended therapy for less than a year.

Test material: MLT (Rukmini, 1994) was used to identify the language age of the participants. The test consists of two sections: semantics and syntax. Reception and expression of every item under each section could be estimated. The language sample was analysed using Index of Productive Syntax (Scarborough, 1990).

Procedure: The present study consisted of five different phases:

Phase 1: Data collection: A portable digital sound recorder (Sony Digital Recorder ICD- 320) was used to record the language sample. Few of the typically developing children were tested at schools in a classroom away from the noisy environment and few children were from the neighbourhood. Children with mental retardation were tested in the clinical settings. Once rapport was established the child was asked to narrate a familiar story. Children were then engaged in conversation about their family, interests and hobbies. Conversation was open ended and the interaction was not strictly adult directed. Phase 2: Language sample transcription and analysis: Each utterance of the tester and the subject was transcribed verbatim within a few hours of recording on the same day. The clues which were provided during

recording sessions were noted. Within each transcript the language corpus to be analyzed was defined as the child‟s first 100 successive, intelligible utterance excluding imitations, self repetitions and routines.

Phase 3: Adaptation of the test material (IPSyn) to Malayalam: The 56 items in IPSyn were translated to Malayalam with the help of a Linguist. Several items in the IPSyn had to be modified according to Malayalam grammatical structure and few items which were not in accordance with the Malayalam grammar had to be removed. Changes were made for the following items: noun phrases: two word noun phrase after verb or preposition (N6), this is not in agreement with the Malayalam syntactic structure where the verb follows the object and not vice-versa, catenative (V5) the occurrence of which is rare and varies with different dialects, bitransitive predicate (S14)and fronted or centre embedded subordinate clause (S19). The occurrence of such kind of grammatical units are unusual and it was not included in the manual. The transcribed language sample was analyzed and scored following the scoring protocol outlined in Index of Productive Syntax (Scarborough, 1990). The IPSyn adapted to Malayalam consisted of 51 sub items under 4 sections: Noun phrases: Proper, mass, or count noun, Pronoun or prolocative, (excluding modifiers), Modifier (including adjectives, possessives, and quantifiers), Two word noun phrase (nominal preceded by article or modifier), Article used before a noun, Plural suffix, Two word noun phrase before verb, Three word noun phrase (det/mod + Mod + N), Adverb modifying adjective or nominal, any other bound morpheme on N or adjective. Verb phrases: Verb, Particle or preposition, Prepositional phrase (prep + NP), Copula linking two nominals, Auxiliary be, do, have in V5 VP,Progressive suffix, Adverb, Modal preceding verb, Third person singular present tense suffix, Past tense modal, Regular past tense suffix, Past tense auxiliary, Medial adverb, Copula, modal, or auxiliary for emphasis or ellipses (uncontractible context), Past tense copula. Questions/ Negations: Intonationally marked question, Routine do/go or existence /name question or wh pronoun alone, Simple negation (neg + X) : neg = no(t), can‟t, don‟t ; X= NP,VP,PP, adj, Initial wh - pronoun followed by verb, Wh- question with inverted modal, copula, or auxiliary, Negation of copula, modal, or auxiliary, Yes/ no queston with inverted modal, copula, or auxiliary, Why, when, which, whose, Tag question. Sentence structures: Two word combination, Subject – verb sequence, Verb- object sequence, Subject – verb – object sequence, Conjunction, Sentence with two verb phrases, Conjoined phrases, Infinitive without catenative, marked with to Let/ make/help/


133

watch introducer, Adverbial conjunction, Propositional complement, conjoined sentences (except for imperatives, will usually have sub + predicate in each clause), Wh-clause, Sentence with three or more VPs, Relative clause, marked or unmarked, Infinitive clause : new subject, Gerund. Phase 4: Grammatical analysis of the speech sample of the children with mental retardation: The transcribed samples were analyzed and scored using IPSyn. While scoring zero, one, or two points per item could be awarded, so that the total score become the sum of these points over all items. It was scored in such a way that if two examples of each from each item was encountered it will be given a score of two, one point if only one example is encountered and zero if it is absent. Then the sub items were totalled and sum of the sub items gave the total score of test for a particular sample. The IPSyn manual which was adapted to Malayalam was used for analyzing the language sample of children with mental retardation in the age range of 3- 6 years. The same scoring procedure which was used for normative samples was conducted. Phase 5: Correlation of MLT and IPSyn scores: The total percentage scores of Malayalam Language Test and The Index of Productive Syntax was used to find out the correlation between the two tests.


Adaptation of IPSyn to Malayalam: As a part of adaptation of Index of Productive Syntax to Malayalam, the language sample elicited from the children with mental retardation were analyzed using Index of Productive Syntax under four main sections, Noun Phrase (NP), Verb Phrase (VP), Question / Negation (Q/N), and Sentence Structure (S) with a total of fifty one sub items. The total percentage score of the sub items for the three groups were calculated. Among this the items which scored less than 50% were removed from the manual. The following sub items were scored poorly by typically developing children: (N5) article, (N7) plural suffix, (N11) three word noun phrase, (V9) modal preceding verb, (V11) past tense modal, (V14) modal adverb, (V15) copula, modal/auxiliary/ellipses, (Q6) wh question with inverted modal, copula or auxiliary, (Q7) negation of copula, (Q8) yes/no question with inverted modal, copula or auxiliary, (Q10) why/when/whose/which questions,(S5) conjunction, (S9) introducer, (S11) propositional complement, (S13) wh clause, (S16) relative clause, (S17) infinitive clause, (S18) gerund. These sub items were not considered while analyzing IPSyn to children with mental retardation.

Account of agrammatism: In order to account for the agrammatism in orally trained Malayalam speaking children with mental retardation and typically developing children, descriptive statistics was used. The mean and standard deviation values were computed for each of the main sections (NP, VP, Q/N, and S) separately.

Following the descriptive statistics, non parametric tests Mann-Whitney test and Kruskal- Wallis test were employed to estimate the significant difference for each of the items for both groups and ages respectively. In this section, the following items were found to be significantly different across the two groups: modifiers, article used before nouns, plural suffix, adverb, particle or preposition, prepositional phrase, copula linking two nominals, progressive suffix, adverb, modal, third person singular present tense suffix, past tense modal, past tense auxillary, medial adverb, copula, modal or auxillary for emphasis or ellipsis, past tense copula, simple negation, initial wh- pronouns followed by verb, wh questions with inverted modal, sentence with two noun phrase, conjoined phrases, infinitive without catenative, let/make/help introducer, adverbial conjunction, propositional complement, conjoined sentences, sentence with three or more noun phrase and gerund.

Significant difference were not found for the following items: proper nouns, pronoun, two word noun phrase, two word noun phrase before verb ,three word noun phrase and other bound morphemes, verbs (V1), auxiliary, regular past tense suffix, intonationally marked question, routine questions, negation of copula, yes/no question, why when which whose , tag questions, two word combination, subject verb sequence), verb object sequence, subject- verb- object sequence, conjunction, wh clause, relative clause, infinitive clause.

From Table 1, it is evident that in children with mental retardation three word noun phrases, bound morphemes, adverbs and plural suffixes are acquired at a later stage when compared to proper nouns, pronouns, modifiers and articles. Under verbs, they are found to acquire copula or modal auxillaries for emphasis or ellipsis, third person singular forms, past tense modal, past tense auxillary, medial adverb and past tense copula only at a later stage. Negation of copula, yes/no questions, wh- questions, tag questions were found to be more difficult for children with mental retardation when compared to intonationally marked questions and routine questions. Wh-clause, relative clause and infinitive clause conjunctions, adverbial conjunctions and gerunds were the difficult structures observed in children with mental retardation.


134

Table 1. Frequency of occurrence of items in decreasing order

Several factors can be attributed to the above

mentioned results. One such factor is the property of the input which includes: input frequency, perceptual salience and semantics of the functors. The results can also be due to the interaction between these factors. The increased frequency of occurrence of the items such as articles, adverbs, two word and three word noun phrases and other bound morphemes in normal children than children with mental retardation could be attributable to the properties of the input such as input frequency. This is supported by the studies conducted by Brown (1958) and Naigles and Hoff-Ginsberg (1998).

The second property of input could be the perceptual salience which involves several factors such as phonetic substance, susceptibility to heavy stress and pitch, possibility of occurring in utterance final position. Blandell and Jensen (1970) claimed that this may help in the imitation of children. Brown (1976) reported that in English the inflections such as plurals, possessives, regular past tense on verbs are not even fully syllabic and that there is a tendency for the above mentioned features to be lost. Thus the chances of occurrence of these items in their speech will be less. While in German, some of these inflections are syllabic. He also reported that there can be variations in these factors depending on the language of the child who is acquiring the grammar and the stage of the development. As Malayalam is a syllabic language, these parameters are stressed and thus there is a chance for early development of these in both the groups.

Brown and Fraser (1963) in their study of

imitated morphemes in various syntactic classes showed that functors were more often omitted than

content words. Occurrence of proper nouns and pronouns in both the groups can be attributed to this.

In the verb section, the mean values were found

to be more for the items V2 (particle or preposition), V14 (medial adverb) and V15 (copula, modal for emphasis or ellipsis) in children with mental retardation than typically developing children. Mean values for V3 (prepositional phrase), V4 (copula linking two nominals), V7 (progressive suffix), V8 (adverb),V9 (modal), V10 (third person singular), V11(past tense modal), V13 (past tense auxillary) and V16 (past tense copula) were found to be more in the typically developing group. This can be attributed to the fact described by Tamar Keren- Portnoy (2005) as children practice the use of new verbs by repeated trial and occasional error they generate a large number of structure most of them which would be erroneous. These erroneous forms will be shaped into the correct form by parental and environmental input while children with developmental delays such as mental retardation do not receive adequate resources for language development through the natural process of trial and error. The same reason can be attributed to the absence of a significance difference in the other items such as verbs, auxiliary and regular past tense suffix for the two groups.

These findings were supported by Elber and

Wijnen (1992) who suggested that language improves through practice. He also suggested two variables, intensive use and errors as the evidence for practice, which can be accounted for the difficulty and gradualness with which learning first occurs for most of the structures. Above mentioned findings were contradicted by Tamar Keren-Portnoy and Tamar Parush (2005). According to these authors the modern theories ignore the factors such as practice

Noun Phrase

Verb Phrase Questions/ Negation Sentences

Proper Noun , Pronoun Verb, Particle, Auxillary

Intonationally marked questions and Routine questions

Two Word Combination, Subject Verb Sequence

Modifier, Two Word noun phrase, article

Prepositional Phrase, Progressive Suffix, Adverb, Modal Preceding Verb, Regular Past Tense Suffix, Past Tense Auxiliary

Simple negation, Initial Wh Pronoun, Wh- Question with Inverted Modal,.

Object –Verb sequence, S-O-V sequence, Conjoined Phrases, Conjunction

Three word noun phrase, Bound morphemes, Adverb, Plural Suffix

Copula or Modal Auxillaries for emphasis or ellipsis, Third Person Singular Forms,Past Tense Modal, Past Tense Auxillary, Medial Adverb And Past Tense Copula

Negation of Copula, Yes/No Questions, Wh- Questions, Tag Questions

Wh- Clause, Relative Clause And Infinitive Clause Conjunctions, Adverbial conjunctions, Gerund


135

and problem solving in the development of syntactic structures. They posit rote learning and learning triggered by innate knowledge as the major processes through which syntax is initially acquired. Rote learning could be attributed as a factor for the presence of the items particle or preposition, copula linking two nominals, copula or modal for emphasis or ellipsis in children with mental retardation when compared to that of the typically developing children.

In the question / negation section the item Q3,

simple negation follows negation + sentence structure (IPSyn, Scarborough, 1990) which is not applicable to the syntactic structure of the Malayalam language. In Malayalam the structure used is sentence + negation which is an adult form and is hence developed at a later stage. This accounts for the decreased score of the item in the typically developing group. This is supported by Bloom (1970) and Bowerman (1973) who questioned the universality of acquisition of negation +sentence

structure reported by McNeill (1970). Formal training provided to the children with mental retardation probably facilitated the greater use of the adult form of negation.

Factors like personality type, environment,

culture etc. may interfere with the performance of the children in both the groups. But in the group of children with mental retardation, the formal training given to this group facilitates their better performance. The reduced mean values of the other items such as Q4 (initial wh- pronoun) and Q6 (wh- question with inverted modal) in typically developing children may be because of the same reason.

In the study by Tamar Keren-Portnoy and Tamar

Purush (2005) where they studied naturalistic production samples of six children, the order of acquisition of sentence structures are as follows: SV, VO and SVO. The similar order of acquisition is found in the present study.

Table 2. Age wise comparison of NP for typically developing (TD) and children with mental retardation (MR)

*p< 0.05 N1-proper noun; N2-pronoun, N3-modifier, N4-two word noun phrase, N5-article,N7- plural suffix, N8- two word noun phrase, N9- three word noun phrase, N10- adverb, N11- other bound morpheme.

Table 3. Age wise comparison of Verb Phrase for typically developing (TD) and children with mental

retardation (MR)

*p< 0.05 V1-verb, V2- particle, V3- prepositional phrase,V4- copula, V6- auxiliary, V7- progressive suffix, V8- Adverb,V9- modal preceding verb, V10- third person singular present tense,V11- past tense modal, V12- regular past tense suffix, V13- past tense auxiliary, V14- medial adverb, V15- copula, modal or auxiliary for emphasis or ellipsis, V16- past tense copula.

Table 4. Age wise comparison of Questions / Negation for typically developing (TD) and children with mental

retardation (MR)

*p< 0.05 Q1: Intonationally marked questions, Q2: Routine questions, Q3: simple negation, Q4: initial wh pronoun, Q6: wh- question with inverted modal, Q7: negation of copula, Q8: yes/no questions, Q9: wh- questions, Q10: tag questions.

Groups N1 N2 N3 N4 N5 N7 N8 N8 N9 N10 N11

TD 1.0 1.0 0.1 1.0 0.0* 0.08 0.0* 0.6 0.0* 0.5 1.0

MR 1.0 1.0 1.0 1.0 0.0* 0.0* 0.0* 0.1 0.0* 0.1 1.0

Groups V1 V2 V3 V4 V6 V7 V8 V9 V10 V11 V12 V13 V14 V15 V16 TD 1.00 0.00* 0.00* 0.00* 0.00* 0.03* 0.00* 0.04* 0.00* 0.00* 0.00* 0.00* 1.00 0.09 0.05

MR 1.00 1.00 0.00* 1.00* 0.00* 0.00* 0.20 1.00 0.00* 1.00 0.10 0.00* 0.03* 0.00* 1.00

Groups Q1 Q2 Q3 Q4 Q6 Q7 Q8 Q9 Q10 TD 1.00 1.00 0.00* 0.04* 1.00 1.00 1.00 1.00 1.00 MR 1.00 1.00 0.00* 0.00* 0.01* 1.00 1.00 1.00 1.00


136

Table 5. Age wise comparison of Sentence structure for typically developing (TD) and children with mental retardation (MR)

*p<0.05 S1- two word combination, S2- subject verb sequence, S3- verb object sequence, S4- subject verb object sequence, S5- conjunctions, S6- sentences with two verb phrases,S7- conjoined phrases, S8- infinitive, S9- introducer, S10- adverbial conjunction, S11- propositional complement, S12- conjoined sentences, S13- Wh- clause, S15- sentence with three or more verb phrases, S16- relative clause, S17- infinitive clause, S18- gerund.

Kruskal-Wallis test was employed for the age wise comparison of the two groups (TD and children with mental retardation) separately. Following the Kruskal Wallis, Mann Whitney test was carried out for those subitems which are significantly different to see which age group is significantly different from the other for both the groups. When typically developing children were compared across age groups 3-4 years and 4-5 significant difference were found for the following items: N5 (article before noun), N8 (two word noun phrase before verb), N9 (three word noun phrase), V2 (particle or preposition), V3( prepositional phrase), V4 (copula linking two nominals),V8 (adverb),V10 (third person singular), V11(past tense modal), V13 (past tense auxillary), V12, S6, (sentence with two noun phrase), S7 (conjoined phrases),S8 (infinitive without catenative), S9 (let/make/help introducer),S10 (adverbial conjunction), S11 (propositional complement), S12 (conjoined sentences), S15 (sentence with three or more noun phrase (/Z/=2.3, p< 0.05), S6 (sentences with two verb phrases), S7 (conjoined phrases), S8 (infinitive), S9 (introducer), S10 (adverbial conjunction), S11 (propositional complement), S12 (conjoined sentences), S15 (sentence with three or more verb phrases).

Across 3-4 years and 5-6 years significant difference were found for the following items: N5-article, , N8- two word noun phrase, N9- three word noun phrase,V2- particle, V3- prepositional phrase,V4- copula, V6- auxiliary, V7- progressive suffix, V8- Adverb,V9- modal preceding verb, V10- third person singular present tense,V11- past tense modal, V12- regular past tense suffix, V13- past tense auxiliary, Q3: simple negation, Q4: initial wh pronoun followed by verb S5- conjunctions, S6- sentences with two verb phrases,S7- conjoined phrases, S8- infinitive, S9- introducer, S10- adverbial conjunction, S11- propositional complement, S12- conjoined sentences, S15- sentence with three or more verb phrases. Between 4-5 and 5-6 years the following items were found to be significantly different: V10 (third person singular present tense), V11 (past tense modal), V12 (regular past tense suffix), V13 (past tense auxiliary), Q3 (simple negation), S5 (conjunction), S10(adverbial conjunction) and S11(propositional complement).

In the case of children with mental retardation, between ages 3-4 and 4-5 the following items were significantly different: three word noun phrase, auxiliary, conjoined phrases, progressive suffix, past tense auxiliary, simple negation, initial wh pronoun, conjunctions. Between 3-4 and 5-6 the significantly different items were: article, plural suffix, two word noun phrase, three word noun phrase, prepositional phrase, auxiliary, progressive suffix, third person singular present tense, past tense auxiliary, simple negation, initial wh pronoun followed by verb conjoined phrases. Between 4-5 and 5-6 years, article, plural suffix, prepositional phrase, third person singular present tense, past tense auxiliary, medial adverb, questions with initial wh pronoun, wh- question with inverted modal, conjunctions and conjoined phrases were significantly different.

In above findings it is clear that, the age group 4-5 years was significantly different from 3-4 and 5-6 years in most of the items and the significant difference noted between the age groups 4-5 and 5-6 years is very less when compared with that of 3-4 years. This pattern of development was seen in the group of children with mental retardation also. That is, the performance was better for the children of 4-5 years compared to the other age groups. This can be explained by the reason that most of the major linguistic development is occurring in the age range of 4-5 years and above. According to Tunmer and Bowey (1984) this age period is called as the middle childhood (the period from 4-8 years). He suggested that there is continuation of earlier development, besides an emergence of a new kind of linguistic functions evident in this period. He referred this new kind of linguistic functions as metalinguistic development. This was also supported by Hakes (1980). This finding holds good for both the groups, but there will be delay in case of children with, as suggested by Quigley, Power and Steinkamp (1977) and Geffner and Freeman (1980).

Mann Whitney test and Kruskal Wallis was conducted to check for the significant difference of the total scores of sub items and overall score of IPSyn with respect to the variables which can affect the performance of the children with mental retardation such as severity of retardation and the duration of therapy taken. The mean and standard

Groups S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S15 S16 S17 S18 TD 1.00 1.00 0.10 0.50 0.00* 0.00* 0.00* 0.00* 0.00* 0.00* 0.00* 0.00* 1.00 0.00* 1.00 1.00 0.09 MR 1.00 1.00 0.04 0.01 0.00* 0.72 0.00* 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00


138

the statistical measures carried out, it was evident that the children with mental retardation exhibit difficulty in the following grammatical structures: bound morphemes, articles, copula, modal auxillaries, third person singular forms, past tense modal, past tense auxillary, medial adverb and past tense copula. In the question and negation section only intonationally marked questions and routine questions were correctly produced by children with mental retardation. In sentence structure section, only two word combination, subject verb sequence, conjunction, wh- clause, relative clause and infinitive clause were seen to be produced similar to the language age matched typically developing children. Other forms including sentences with two noun phrases, conjoined phrases, adverbial conjunctions and gerunds were found to be poorly developed in children with mental retardation.

The performance of children with mental

retardation was below the performance of typically developing children across all the age groups but followed a similar trend. As age increased, the performance on both MLT and IPSyn increased. There was also high correlation between IPSyn and MLT for both the groups of children.

It could be concluded from the present study

that in children with mental retardation a number of grammatical structures fail to emerge but develop in the same way as seen in typically developing peers. Hence a period of agrammatic language production is evident in the spoken output of children with mental retardation.

References

Aram, D., Morris, R., & Hall, N. (1993). Clinical and research congruence in identifying children with specific language impairment. Journal of Speech and Hearing Research, 36, 580–591.

Bedrosian, J., & Prutting, C. (1978). Communicative performance of mentally retarded adults in four conversational settings. Journal of Speech and Hearing Research, 21, 79-95.

Blandell, R., & Jensen, P. (1970). Stress and word position as determinants of imitation in first language learners. Journal of Speech and Hearing Research, 13, 193- 202.

Bloom, L. M. (1970). Language development: Form and Function in Emerging Grammars. Cambridge, MA: MIT Press.

Bloom, L., & Lahey, M. (1978). Language development & Language Disorders. New York.

Bowerman, M. (1973). Early syntactic development: a cross-linguistic study with special reference to Finnish. London: C.U.P.

Bowerman, M. (1973). Early syntactic development: a cross – linguistic study with special reference to Finnish. London: C.U.P.

Brown, R. (1958). How shall a thing be called? Psychological Review, 65, 14 – 21.

Brown, R. (1973). A first language. Boston, MA: Harvard University Press.

Brown, R. (1976). A first language. Harmondsworth: Penguin Books.

Brown, R., & Fraser, C. (1963). The acquisition of syntax. In C. N. Cofer & B. S. Musgrave (Eds.), Verbal Behavior and learning (pp.158-196). New York: McGraw-Hill.

Chapman, R. (1981). Exploring children's communicative intents. In J. Miller (Eds.), Assessing language production in children (pp. 111-138). Baltimore, MD: University Park Press.

Crystal, D., Garman, M., & Fletcher, P. (1989). The grammatical analysis of language disability: A procedure for assessment and remediation (2nd ed.). London: Cole and Whurr.

Elber, L., & Wijnen, F. (1992). Effort, production skill and language learning. In C. A., Ferguson, L., Menn, & Stoel – Gammon (Eds.), Phonological Development: models, research implications. Timonium: York Press.

Fraser, C., Bellugi, U., & Brown, R. (1963). Control of grammar in imitation, comprehension, and production. Journal of Verbal Learning and Verbal Behavior, 2, 121-135.

Geffner, D. S., & Freeman, L. R. (1980). Assessment of language comprehension of 6-year-old deaf children. Journal of Communication Disorders, 13, 455-47.

Hakes, D. T. (1980). The development of metalinguistic abilities in children. Berling: Springer Verlag.

Lee, L. L. (1974). Developmental sentence analysis: A grammatical assessment procedure for speech and language clinicians. Evanston, IL: Northwestern University Press.

Loban, Walter, (1976). Language development: kindergarten through grade twelve. Urbana: NCTE.

McDaniel, D., McKee, C., Smith Cairns, H. (1996) (Eds.), Methods for assessing children's syntax. Cambridge, MA: MIT Press.

McNeill, D. (1970). The acquisition of language: The study of developmental psycholinguistics. New York: Harper & Row.

Naigles, L., & Hoff- Ginsberg, E. (1998). Why are verbs learned before other verbs? Effects of input frequency and structure on children‟s early verb use. Journal of Child Language, 25, 95 – 120.

Radford, A. (1990). Syntactic theory and the acquisition of syntax. Oxford: Blackwell.

Rosenberg, S., & Abbeduto, L. (1993). Language and communication in mental retardation: Development, processes, and intervention. Hillsdale, NJ: Lawrence Erlbaum.

Rukmini, A.P. (1994) Malayalam Language Test (MLT). Unpublished master‟s dissertation submitted in part of fulfillment for the master‟s degree in speech and hearing. University of Mysore, Mysore.

Scarborough, H. S. (1990). The index of productive syntax. Applied Psycholinguistics, 11, 1-22.

Tager-Flusberg, H., & Calkins, S. (1990). Does imitation facilitate the acquisition of grammar? Evidence from a study of autistic, Down syndrome and normal children. Journal of Child Language, 17, 591-606.

Tamar Keren-Portnoy & Tamar Parush. (2005). The concept of reference in the cognitive sciences.

Templin, M. C. (1957). Certain language skills in children: Their development and interrelationships. Minneapolis: The University of Minnesota Press.


139

Tomasello, M. & Brooks, P. (1999) Early syntactic development. In M. Barrett (Eds.), The development of language. UCL Press.

Tunmer, W. E., & Bowey, J. A. (1984). Metalinguistic awareness and reading acquisition. Part III (3.2). In W. E, Tunmer, C., Pratt, & M. L. Herriman (Eds.),

Metalinguistic awareness in children (pp. 144-168). Berling: Springer-Verlag, US.

Quigley, S. P., Power, D. J., & Steinkamp, M. W. (1977). The language structure of deaf children. Volta Review, 79, 73-80.


140

Performance of PWS on Self Select Reaction Time Paradigm Using Speech and Non Speech Tasks

Mahesh B. V. M.1& R. Manjula 2

Abstract

The disorder of stuttering is viewed in terms of speech motor control perspective in the recent times. Most of the theories and models hint upon difficulty in initiating and controlling the speech movements as a common factor in persons with stuttering. Reaction time measures have been extended to study the common neuromotoric deficits across unrelated motor systems to generate an extensive data to verify that stuttering is a disorder extending beyond speech. Reaction time paradigms have been modified to tap the intended measure. In this study a modified reaction time paradigm called „Self-Select Reaction Time Paradigm‟ is used to delineate the motor programming deficits if any seen in Person with stuttering (PWS) across non speech and speech tasks. Fifteen PWS who had undergone speech therapy, 10 PWS without any treatment and 25 normal controls in the age range of 16-30 years participated in the study. The Self Select Reaction Time Paradigm was used to measure the reaction time for two motor programming processes namely INT and SEQ across speech and non speech tasks within four (1short, 1long, 4long & 4short) conditions. The results revealed that both the groups differed in various conditions across speech and non speech tasks. A modality independent deficit in the INT process was evident which supported a generalized motor programming deficit in the organization of spatiotemporal sequences in PWS. The potential utility of the paradigm and the implications of the results to the current understanding of the disorder of stuttering are discussed. Key words: stuttering, self select reaction time paradigm, motor programming

an1Riper (1982) defined stuttering as a disruption of the simultaneous and successive programming of muscular movements required

to produce a speech sound or its link to the next sound in a word. This definition suggests a possible scope of understanding the disorder from the speech motor control perspective. The speech motor control perspective of stuttering is more than just one single theory or model and all these theories share the common hypothesis that PWS have difficulties in initiating and controlling speech movements in one way or other. They suggest that, in stuttering the speech mechanisms responsible for a precise adjustment of the respiratory, laryngeal and articulatory movements are operating less efficiently. At certain moments, this inefficiency causes a breakdown of speech fluency and results in dysfluencies. How exactly this takes place has not been understood in a strict sense. The „discoordination hypothesis‟ states that stuttering is presumably the result of constitutional inability to temporally co-ordinate respiratory, phonatory and articulatory subsystems in speaking (Perkins, Rudas, Johnson & Bell, 1976; Caruso, 1991). Few studies supported the discoordination hypothesis

1e-mail: [email protected]; 2Professor of Speech Pathology, AIISH, Mysore, [email protected].

by using EMG measurements (Peters, Hulstijn & Starkweather, 1989). These studies reported a disruption of normal reciprocal action of abductor muscles in non fluent utterances which inturn suggested that stuttering might be due to the discoordinated activity between and within speech subsystems. Many other studies also supported the above hypothesis (Adams, 1974; Wingate, 1976; Zimmerman, 1980; Van Riper, 1982; Borden, 1983; Gracco, Caruso & Abbs, 1988; Harbison, Porter & Tobey, 1989). The hypothesis gradually lost its significance since few of the studies showed no differences in terms of discoordination between normals and PWS (Conture, Colton & Gleason, 1988). Also, it failed to account for the core behavioural features seen in PWS.

An alternative to the „discoordination hypothesis‟ is the „Speech planning hypothesis‟ (Postma & Kolk, 1993) where a central dysfunction is proposed which operates before the actual execution of speech occurs. The speech motor plan is an elaborate representation of all or most of the „intended utterance‟ constructed prior to the actual execution of the utterance itself (Sternberg, Monsell, Knoll & Wright, 1978). Many models were also proposed which explained stuttering as a motor planning/programming deficit (Mackay, 1982; Schmidt, 1988; Postma & Kolk, 1993; Van Leishout,

V

Performance of PWS on SRT

141

1995; Van Der Merwe, Mc Neil, Robin & Schmidt, 1997).

Reaction Time (RT) Paradigm is the most commonly used technique to investigate motor programming in speech production and many investigators have used RT paradigms to address the issue of speech motor control in general and particularly in stuttering (Kahneman, 1973; Peters et al., 1989; Van Leishout, Hulstijn & Peters, 1996; Aravind & Savithri, 1997). The underlying assumption of this paradigm is that differences in the latency of reaction time (dependent variable) consequent to manipulation of the elicited stimuli (the independent variable) are a result of alteration in motor programming and helps in studying the response preparation in the temporal domain. The majority of these studies have recorded slower reaction times for stutterers than for non stutterers (Adams & Hayden, 1976; Cross & Luper, 1979; Cross, Shadden & Luper, 1979; Starkweather, Franklin & Smigo, 1983) along with few studies which contraindicated the presence of slower reaction times in speech as well as in non speech tasks (McFarlane & Prins, 1978; Till, Reich, Dickey & Seiber, 1983).

A two-stage model of motor programming for

both speech and non speech movements was developed by Klapp (1995, 2003). Unlike the other models, this model distinguishes two separate processes in speech motor programming namely INT/SEQ and assumes that preparation of a sequential movement involves an organization of a series of motor programs. The first process (INT) refers to the internal spatiotemporal structure of an individual unit of movement and reads it into a motor buffer (Klapp, 2003). INT can be completed prior to initiation (preprogrammed) and is sensitive to unit complexity, with longer processing time for units that are more complex. The second process (SEQ) refers to the sequencing of units into their correct serial order after initiation. The SEQ process involves on-line retrieval of units from the motor buffer and therefore cannot be preprogrammed. SEQ is sensitive to the number of units in the buffer but not to the complexity of a unit.

Klapp (1995, 2003) validated the INT/SEQ

model using RT paradigms. In a simple RT paradigm, the response to be produced on a given trial is cued before the imperative signal that prompts response production; this allows pre-programming and reflects SEQ process. In a choice RT paradigm, the imperative signal specifies the response to be produced, and thus

preprogramming is not possible thereby reflecting the INT process. Klapp (1995) found an effect of button press duration (finger movements) on Choice Reaction Rime and an effect of sequence length on Simple Reaction Time.

Klapp‟s model (1995, 2003) was replicated

using a Self-Selection RT Paradigm which measured the INT and SEQ processes on each trial (Immink & Wright, 2001; Wright, Black, Immink, Brueckner & Magnuson, 2004). In these studies the participants prepare the upcoming responses and indicated the same by pressing a button when they are ready. This preparation duration was referred to as the ST (ST) which inturn reflected on the INT process. A go-signal will prompt the individuals to execute the response. The latency between the go-signal and the response is measured and this was called as RT which inturn reflected on the SEQ process.

Many of the studies in the past have reported a

programming deficit in Stuttering (Peters et al., 1989; Aravind & Savithri, 1997). All the studies viewed speech motor programming errors seen in stutterers as a unitary stage and a very few of these attempted to address the nature of speech and non speech motor programming deficit in stutterers. Studies based on Klapp‟s model (1995, 2003) have led to the observation that speech motor programming involves two distinct processes in a hierarchical sequence and it is not necessarily a unitary process (Immink & Wright, 2001; Wright et al., 2004). The two processes, INT and SEQ have been studied in subjects with Apraxia of speech (Immink & Wright, 2001; Wright et. al., 2004) using Self Select Reaction Time Paradigm. Such an attempt has not been made in persons with stuttering. This study is proposed to examine the performance of PWS on the Self Select Reaction Time paradigm for speech and non speech tasks.

The aim of the study was to compare the

performance of PWS and normal controls on speech and non speech tasks using Self Select Reaction Time Paradigm. The study investigated the difference if any between normal controls, PWS with treatment and PWS without treatment with respect to: (a) Motor programming for non speech and speech tasks, and thus its relation to INT or SEQ processes of programming (b) the modality independent or modality dependent factors with respect to INT or SEQ processes and (c) the effect of treatment in PWS with respect to INT or SEQ processes.


143

A pilot study was conducted on five participants to ensure the utility of the program developed using the DMDX software for the experimental tasks. Totally there were 10 blocks containing 40 trials in both speech and non speech tasks. Each block consisted of four trials which included 1 Short, 1 Long, 4 Short or 4 Long. Followed by the instructions a single block consisting of 4 trials was shown to the participants‟ in order to understand the task better. Also, it was noticed that the subjects committed more mistakes in the first block and the responses were slowed down at the last block. Hence, it was decided to exclude the first and last block in the experimental trials. Experiment of the study: The RT paradigm developed by the investigator which was tested and modified based on the outcome of the pilot study was included in the experiment of this study. The present study included two experiments: (a) Experiment 1: RT paradigm for non speech tasks and (b) Experiment 2: RT paradigm for speech tasks. Experiment 1: RT paradigm for non speech tasks Finger movement task as used in earlier studies (Klapp, 1995; Wright et al., 2004) and that was tested in the pilot study was included in this experiment. Appropriate instructions were given to each participant regarding different key presses and their sequences which are used in the experiment. The instruction was prepared using Power Point slides by the experimenter and this was presented to each subject before they participated in the experimental trials.

Initially, subjects were familiarized with the different key press responses which included a “Short press” („S‟ Key for 150 ms) and a “Long press” („L‟ Key for 450 ms). Each subject was provided with an auditory model regarding “short” and “long” press responses which consisted of two separate tones which included a short duration tone of „150 ms” and a long duration tone of “500 ms”. Later they were familiarized with the number of responses which they have to produce which included either a “single key press” or a “multiple key presses”. Accordingly, four different key press responses would elicit four targets namely; 1S (Single short press: 150 ms), 1L (Single long press: 450 ms), 4S (SLLS sequence: 150-450-450-150 ms), 4L (LSSL sequence: 450-150-150-450 ms).

Each experimental trial followed a particular

sequence in which each trial was initiated by presenting a visual symbol of “star” which was

followed by presenting the visual cue which could be either 1 Short, 1 Long, 4 Short or 4 Long. The subjects were asked to think about the required response which they have to produce mentally and press the space bar when they are ready to respond. This preparation interval is termed as “ST” and this would reflect the demands associated with the INT process. Followed by a variable delay, „go‟ signal of 300 ms was presented which prompted the individual to execute the required response. The time between the „go‟ signal and the response is called “Reaction Time” and this would reflect on demands associated with SEQ process. Experiment 2: RT paradigm for speech tasks Speech movements used in earlier studies (Klapp, 1995; Wright et al., 2004) and that was tested in the pilot study was included in this experiment. Appropriate instructions were given regarding different speech movements and their sequences which were used in the experiment. The instruction was prepared using PowerPoint slides by the experimenter and this was presented to each subject before they participated in the experiment.

Initially an auditory model of non sense syllable /pa/ which varied in duration and length was recorded by a male native speaker. The participants were familiarized with the nonsense syllable /pa/ which varied in terms of syllable duration and sequence length. This included a “Short syllable” (/pa/ of 150 ms duration) and a “Long syllable” (/pa/ of 450 ms duration). When the participants were familiarized with the different types of responses, they were asked to produce a “Short /pa/” for a duration equivalent to “150 ms” and a “Long /pa/” for a duration equivalent to “450 ms”. Later they were familiarized with the number of responses which they had to produce which included either a “single syllable” or a “multiple syllable sequence” responses. Accordingly, four different syllabic productions would elicit four targets namely; 1S (Single short syllable: 150 ms), 1L (Single long syllable: 450 ms), 4S (SLLS sequence: 150-450-450-150 ms), 4L (LSSL sequence: 450-150-150-450 ms).

The experiment began with the presentation of

the “READY” signal which was followed by a visual cue of 1S, 1L, 4S or 4L. The presentation of the visual cue prompted the required response from the subjects. The subjects were instructed to press the space bar when they were ready; this measures the ST and in turn reflects on the INT process. Pressing the „Space bar‟ produced a variable time delay followed by a „go‟


144

signal prompted for the execution of the response. The time delay between „go‟ signal and the response is called “Reaction Time (RT)” and reflected on the SEQ process.

Totally, both speech and non speech tasks consisted of 10 blocks in which each block consisted of 4 different types of responses i.e. 1S, 1L, 4S and 4L. The order of presentation of these was randomized across 10 blocks which totally constituted 40 trials. Analysis: In both experiments 1 & 2, raw scores were obtained for each condition (1S, 1L, 4S & 4L) in non speech tasks across ST and RT. Later mean scores were calculated for each condition across ST and RT. Mean scores of ST and RT for speech and non speech tasks of the participants were calculated and compared within the group and across the groups and also across four different conditions. While analyzing the speech motor programming the RT was measured from the burst of the syllable /pa/. The raw data was treated with suitable statistical procedures to make the inter and intra group comparisons.


A. Between group comparisons of ST & RT

a] Non speech Task: MANOVA was used to compare the differences across subjects in non speech task. Results in Table 1 point to a significant difference at less than 0.05 level between the three subject groups across Non Speech Study Time (NSST). The overall reaction time for the NSST was shortest for normal controls followed by PWS with treatment and then the PWS without treatment.

The longer ST reflecting INT process (Klapp, 1995, 2003) in the two experimental groups indicated that PWS took longer time in organizing the spatial temporal characteristics of an individual unit of utterance. The results obtained in the present study are in congruence with other studies which have used a different experimental paradigm (Webster, 1986; Rastatter & Dell, 1987; Webster & Ryan, 1991) which showed longer reaction times across various manual reaction time measures. Duncan‟s post hoc analysis revealed a significant difference between normals and PWS with no treatment and significant difference between PWS with treatment and without treatment. The PWS group with treatment performed similar to the normal controls, and this could be attributed to the treatment variable.

b] Speech Task: MANOVA was used to check the differences across groups statistically for the speech ST and RT tasks. As it is evident from Table 2, there was a significant difference at 0.05 level across the subject groups for Speech Study Time task (SST).

From the mean values, we can infer that the

overall ST for the SST was shortest for normal controls followed by PWS with treatment and finally by PWS without treatment. The results suggests that PWS with or without treatment takes longer time than normal controls while preparing the responses in advance before they execute the speech movements.

Table 1. Mean (in msec) and Standard Deviation (SD) for ST and RT across subjects in non speech tasks

Experimental condition

Group N Mean SD F

value

NSST

Normal 25 1049.79 331.66 5.138* PWS No

Tx 10 1472.71 324.71

PWS with Tx

15 1155.52 402.23

NSRT

Normal 25 815.90 277.86 3.872* PWS No

Tx 10 1121.61 360.06

PWS with Tx

15 941.91 283.70

*p< 0.05 NSST = Non Speech Study Time, NSRT = Non Speech Reaction Time, Tx = Treatment

Duncan‟s post hoc test revealed that there were no significant differences between normal controls and PWS who have undergone therapy. But, significant differences were found between normal controls and PWS with no treatment; PWS with treatment and PWS without treatment only for speech ST (SST) and not for Speech RT (SRT). It can probably be reasoned out that many of the fluency remediation therapies would provide sufficient time to plan the upcoming utterances by reducing the overall speech rate and this inturn gets reflected in the preparation of responses in advance. A similar notion has been supported by the study done by Savithri and Pooja (2000) wherein a reduced reaction time was observed after therapy in PWSA modality independent motor programming deficit in INT process irrespective of variables like treatment and severity of the condition is evident from the above findings.


145

B] Within group comparison across non speech and speech task a] Within group comparison of normals for non speech and speech tasks: Table 3 shows that the mean of the non speech ST (NSST) is longer than the non speech reaction time (NSRT) in normals. Paired t test revealed significant difference between the two tasks, [t(24) = 4.085, p < 0.05]. A longer ST in non speech could be due to an additional processing of the choices that were required which inturn required much more time since the choices carry more information load than an alerting signal which does not have sufficient information.

Table 2. Mean (in msec) and Standard Deviation (SD) for speech ST and RT across subjects in speech tasks


Group N Mean SD F

value

SST

Normal 25 821.02 397.60 5.853* PWS No

Tx 10 1273.92 338.72

PWS with Tx

15 974.70 277.88

SRT

Normal 25 894.17 250.97 2.646 PWS No

Tx 10 1046.67 276.46

PWS with Tx

15 1060.23 219.70

*p< 0.05 SST = Speech Study Time, SRT = Speech Reaction Time, Tx = Treatment

This created a significant difference between ST and RT of non speech task. The table also shows that there was no significant difference between ST and reaction time of speech task, [t (24) = 0.905, p = 0.375]. In the speech task subjects were made to artificially associate a visual cue with a verbal response. This association might have created an additional load on processing the visual stimuli and this could have inturn affected the online retrieval of the individual syllables. Hence, there was no significant difference in the ST and the reaction time.

b] Within group comparison in PWS with no treatment: From Table 4 it is evident that the ST and RT is different in non speech and speech tasks and they were significantly different from each other at 0.05 level of significance in the non speech task for PWS with no treatment, [t(9) = 3.181, p < 0.05]. An evident significant difference between the ST and the RT in the speech tasks for PWS with no treatment was seen and this was also found to be significant [t(9) = 2.635, p < 0.05]. It can thus be inferred that PWS with no treatment took longer time in programming the spatio

temporal parameters of the key pressing as well as in programming syllable /pa/ which inturn is reflected as a longer ST. A shorter RT shown by this group in the speech tasks compared to ST reveals that they had problem in retrieving the programmed sequences since normals did not show differences between ST and RT in speech task. c] Within group comparison of PWS with treatment: Table 5 shows no significant differences in ST and RT of both non speech [t(14) = 1.925, p > 0.05] and speech task [t(14) = 0.941, p > 0.05]. An increased arousal level after improved speech fluency can be attributed for the insignificant changes seen across tasks.

Table 3. Mean (in msec), Standard Deviation (SD) and

paired t test values of normals for non speech and speech tasks


N Mean SD t-value

(24) Non speech task

NSST 25 1049.79 331.66 4.085*

NSRT 25 815.90 227.86

Speech task

SST 25 821.03 397.60 0.905

SRT 25 894.17 250.87 *p< 0.05 NSST = Non Speech Study Time, NSRT = Non Speech Reaction Time, SST = Speech Study Time, SRT = Speech Reaction Time

Table 4. Mean (in msec), Standard Deviation (SD) and paired t test values of PWS with no treatment within

non speech and speech tasks

Experimental condition N Mean SD t-value (24)

Non speech task

NSST 10 1472.71 324.71 3.181*

NSRT 10 1121.61 360.06

Speech task

SST 10 1273.92 338.72 2.635*

SRT 10 1046.67 276.46 *p< 0.05 NSST = Non Speech Study Time, NSRT = Non Speech Reaction Time, SST = Speech Study Time SRT = Speech Reaction Time C] Within subject comparison for different conditions a) Normal controls: In the non speech tasks the results of Repeated measure ANOVA revealed that all the


146

conditions across the ST and RT were significantly different from one another respectively [F(3, 72) = 2.868, p < 0.05, F( 3, 72) = 11.882, p<0.05].

Paired t test run on the ST revealed that only the conditions 4S and 1S [t (24) =2.117, p < 0.05)], 4L and 1S [t(24) = 2.088, p < 0.05)] were significantly different with each other ST, and none of the other pairs were significant. Boneferroni‟s multiple group comparison checked the individual differences in RT and revealed 4S and 4L are significantly different than 1S and 1L. A sequence length effect of increased response latency with increase in response complexity was evident. No duration effect was seen i.e. there were no differences between 1S, 1L and 4S, 4L. Table 5. Mean (in msec),Standard Deviation (SD) and paired t test values of PWS with treatment within non

speech tasks


N Mean SD t-value (24)

Non speech task

NSST 15 1155.52 402.23 1.925

NSRT 15 941.91 283.70

Speech task

SST 15 974.70 277.88 0.941

SRT 15 1060.23 219.70

In speech ST task showed that there was an absence of sequence length effect. Also, there were no differences between the responses requiring same sequences i.e. 4S, 4L and 1S, 1L, revealing once again that there was no duration effect on the ST of speech task. Further repeated measure ANOVA showed no significant difference across different conditions in ST task, [F(3, 72) = 1.155, p > 0.05]. Absence of sequence length effect suggests that all the units, irrespective of the length seemed to be preprogrammed as a single chunk in the ST task and the absence of duration effect suggests that the addition of the syllable duration did not tax the speech motor system. In RT task only sequence length effect was evident with the absence of duration effect.

b) PWS with no treatment: Repeated measure ANOVA revealed no significant difference across conditions for ST measures [F (3, 27) = 2.620, p > 0.05] and RT measures [F(3, 27) = 1.177, p > 0.05] of non speech task in PWS without treatment there was no significant difference across conditions also in ST measures [F(3, 27) = 2.300, p > 0.05] of speech task. There was however a significant difference between the conditions in RT measures [F(3, 27) = 7.422, p <

0.01] of the speech task. There was also a lack of duration effect in all the tasks. Lack of sequence length effect in both ST of non speech and speech tasks in PWS with no treatment shows no differences in programming a shorter or a longer chunk. c) PWS with treatment: Repeated measure ANOVA revealed no significant differences between conditions in ST of non speech task, [F (42, 3) = 0.956, p > 0.05]. When compared to PWS with no treatment the reaction time of PWS with treatment was much shorter. This could be attributed to the treatment variable but, it is not statistically significant. There was no sequence length or duration effect on the responses indicating that treatment had a positive effect on motor programming, it could not rectify completely the deficit seen in PWS.

Repeated measure ANOVA across conditions revealed that there was no significant difference between conditions in the ST of the speech task [F (3, 42) = 2.056, p > 0.05]. There was no statistically significant difference between shorter and longer sequences suggesting that the motor programming of speech was aberrant even after PWS had undergone therapy. But the reaction time of PWS with treatment was shorter compared to the PWS with no treatment. Repeated measure ANOVA revealed a significant difference across the conditions in the RT of speech task [F(3, 42) = 4.121, p < 0.05].

There was a sequence length effect and an

absence of duration effect which again revealed that PWS have difficulty in integrating the individually programmed chunks into a single cohesive unit and the presence of duration as a complex factor did not tax their motor system.

D] Within condition comparison across groups To understand the within subject differences across the groups, Multivariate Analysis of Variance i.e. MANOVA was carried out. Since the subject size was less, the MANOVA results were cross verified with Kruskal-Wallis test. The MANOVA results revealed that both the ST and RT in all the four conditions (4S, 4L, 1S, 1L) of non speech task were significantly different across the groups. In the ST of speech task, conditions like 4L, 1S and 1L were significantly different across the groups and in the RT, only 1L was significantly different across the groups. This was also supported with the earlier findings that non speech ST and RT showed a significant difference between groups owing to the sequence length effects.


147

The ST of speech task produced significant differences between the groups in conditions like 4L, 1S and 1L. In the RT of speech task there was no significant difference across groups found. This could be due to the fact that all hearing adults are highly skilled in sequencing and controlling the syllable duration due to continuous practice in speaking over their lifetime. Also, it has been reported in the motor learning literature that a sequence of units may completely get reorganized as a single unit due to extensive random practice (Klapp, 1995; Sakai, Hikosaka & Nakamura, 2004; Wright et. al., 2004) Though, different stimulus varying in length and complexity was presented, it could have been reorganized as a single chunk, because of which there was no significant difference in various conditions across groups.

Duncan‟s Post hoc analysis test was carried out

to know the factors that accounted for the group difference. Few of the dependent variables of ST (4L, 1S, 1L) , RT (4S, 4L, 1S, 1L) of non speech task and RT (1L) of speech task were significantly different between normal controls and PWS with no treatment. These findings show that normals and PWS with no treatment behaved like two different groups across various conditions taken in the experiment but, PWS with treatment were falling between the normal controls and PWS with no treatment.

The results of Kruskal- Wallis test revealed that all the conditions in the ST of non speech task were significantly different across groups but, no significance was seen within all the conditions of RT of non speech task. Also, the findings of both MANOVA and Kruskal-Wallis test matched the ST of speech task where significant results across groups were revealed. The findings of MANOVA and Kruskal-Wallis are showing contradictory results with respect to the RT of the non speech task. But going by findings of within subject comparison of different conditions, it can be stated that there was no significant difference between RT of non speech across subject groups.

Also, the results should be explained by

keeping the findings of Kruskal-Wallis test, since it has more statistical power than MANOVA when the subject size varies significantly. Hence a deficit at the stage of INT leaving SEQ process intact best explains the non speech motor programming errors seen in PWS. Both MANOVA and Kruskal-Wallis revealed significant differences in ST of speech task, the findings can be interpreted that the PWS have deficit at the INT stage of speech motor programming. In the RT of speech

task, few of the conditions are found to be significantly different by MANOVA such as 4S, 4L and 1S and Kruskal-Wallis revealed significant difference across 4L and 1S. Hence, these two tasks were found to be significantly different from each other whereas the other two were not.

It is however difficult to conclude whether

PWS have deficits in the RT of speech task. This variable has to be studied further in greater detail to delineate the presence of programming errors in the speech task. Mann-Whitney U test was used to understand the difference between the groups across conditions.

Table 6. Chi-square and its significance of Kruskal-

Wallis test within conditions across groups

Conditions X2 (2) Sig. nspST4S 7.649 0 .022* nspST4L 6.819 0.033* nspST1S 9.222 0.010* nspST1L 7.344 0.025* nspRT4S 3.699 0.157 nspRT4L 3.234 0.198 nspRT1S 4.817 0.090 nspRT1L 4.872 0.088 spST4S 9.734 0.008* spST4L 14.889 0.001* spST1S 11.088 0.004* spST1L 15.164 0.001* spRT4S 6.487 0.039* spRT4L 2.066 0.356 spRT1S 4.067 0.131 spRT1L 6.011 0.040*

*p< 0.05 nspST= Non speech study time, nspRT= Non speech reaction time, spST= Speech study time, spRT= Speech reaction time

Normals vs. PWS with no treatment: Mann-Whitney U test revealed significant differences across all the conditions in the ST and RT of non speech tasks and ST of speech tasks across groups but, no significant difference was obtained in the RT of speech task across three conditions such as 4L, 1L, and 1S.Hence these findings support modality independent deficit in PWS in INT stage and modality dependent SEQ difficulty in the non speech task. Normals vs. PWS with treatment: No significant differences across all the conditions except for the conditions of 4L of ST, 1L and 4S of RT of speech task was found. No significant difference across most of the conditions between the two groups could be attributed to the treatment variable. The significant


148

differences seen in the above said conditions should be tested further to understand the group differences. PWS with treatment vs. PWS with no treatment: Mann-Whitney U test revealed no significant differences across the two groups in majority of the conditions but, there were few conditions which were significantly different across the groups which included 4S in ST of non speech task and 4L, 1S and 1L of ST of speech task. It can be concluded that PWS with and without treatment did not differ on the majority of the variables but, when it comes to the ST of speech task, there was a significant difference in three out of four conditions suggesting an INT deficit in PWS irrespective of the treatment variable.

Conclusions

There was a significant difference between normal controls and PWS in the motor programming stages outlined by Klapp (1995, 2003) in both non speech and speech tasks irrespective of the treatment variable in effect. PWS with treatment and without treatment were significantly different from the normal controls in the ST of both non speech and speech task. This suggests a modality independent deficit in ST which in turn points to a deficit in the INT process. There was no significant difference in the RT in the non speech task revealing an intact SEQ and aberrant INT process in the non speech task. But, the findings of the study in terms of Speech Reaction Time did not show any consistent trend across the groups hence it should be explored further. Also, treatment variable has a positive effect on both speech and non speech motor programming even though the therapy was addressed only with respect to speech remediation. Al l these findings suggest that there are few common neuromotor control strategies which subserves two unrelated motor systems.

In conclusion, though treatment showed

favorable effect on the speech motor programming as inferred through the Self Study Reaction Time paradigm used in this study, all the effects could not be attributed to treatment. Some of the similarities seen between normals and PWS with treatment group could be attributed to the motivation factors, arousal, practice effects along with some uncontrolled processing at the central level which cannot be addressed with the design used in this study.

References Adams, M. R. (1974). A physiologic and aerodynamic

interpretation of fluent and stuttered speech. Journal of Fluency Disorders, 1, 35-67.

Adams, M. R., & Hayden, P. (1976). The ability of stutterers and nonstutterers to initiate and terminate phonation during production of an isolated vowel. Journal of Speech and Hearing Research. Vol. 19, 290-296.

Aravind, N., & Savithri, S. R (1997). A comparative study of speech motor programming in stutterers and nonstutterers. Research at AIISH; Dissertation Abstracts. Vol. IV. M. Jayaram & S. R. Savithri (Eds.). pp 4-5. Mysore; AIISH publications.

Boersma, P., & Weenink, D. (2010). Praat: doing phonetics by computer [Computer program]. Version 5.1.45, retrieved 26 October 2010 from http://www.praat.org/.

Borden, G. J. (1983). Initiation versus execution time during manual and or counting by stutterers. Journal of Speech and Hearing Research, 26, 389-396.

Caruso, A. J. (1991). Neuromotor processes underlying stuttering. In H. F. M. Peters, W. Hulstijn, and C.W. Starkweather (Eds.), Speech motor control and stuttering. Elsevier/Excerpta Medica, Amsterdam.

Conture, E. G., Colton, R. H., & Gleason, J. R. (1988). Selected temporal aspects of coordination during fluent speech of young stutterers. Journal of Speech and Hearing Research, 31, 640-653.

Cross, D. E., & Luper, H. L. (1979). Voice reaction time of stuttering and nonstuttering children and adults. Journal of Fluency Disorders, 4, 59-77.

Cross, D. E., Shadden, B. B., & Luper, H, L. (1979). Effects of stimulus ear presentation on the voice reaction time of adult stutterers and nonstutterers. Journal of Fluency Disorders, 4, 45 -88.

Freeman, F. J., & Ushijima, T. (1978). Laryngeal muscle activity during stuttering. Journal of Speech and Hearing Research, 21, 538-562.

Gracco, V. L., Caruso, A. J., & Abbs, J. H. (1988). Kinematic analysis of multiple movement coordination during speech in stutterers. Brain, 111, 439-455.

Harbison, D. C., Porter, R. J., & Tobey, E. A. (1989). Shadowed and simple reaction I times in stutterers and non-stutterers. Journal of the Acoustical Society of America, 86, 1277-1284.

Immink, M. A., & Wright, D. L. (2001). Motor programming during practice conditions high and low in contextual interference. Journal of Experimental Psychology: Human Perception and Performance, 27, 423-437.

Kahneman, D. (1973). Attention and effort. Englewood cliffs, Prentice - Hall, New Jersey.

Kenneth, I. F., & Jonathan, C. F. (2003). DMDX: A window display program with millisecond accuracy. Behavior Research Methods, 35, 116-124.

Klapp, S. T. (1995). Motor response programming during simple and choice reaction time: The role of practice. Journal of Experimental Psychology: Human Perception and Performance, 21, 1015-1027.


149

Klapp, S.T. (2003). Reaction time analysis of two types of motor preparation for speech articulation: Action as a sequence of chunks. Journal of Motor Behavior 35, 135-150.

Levelt, W. J. M. (1989). Speaking: From intention to articulation. M.I.T. Press, Cambridge, M.A.

MacKay, D. G. (1982). The problems of flexibility, fluency, and speed -accuracy trade - off in skilled behaviour, Psychological Review, 89, 483 - 506.

McFarlane, S. C., & Prins, D. (1978). Neural response time of stutterers and nonstutterers in selected oral motor tasks. Journal of Speech and Hearing Research, 21,768-78.

Perkins, W., Rudas, J., Johnson, L., & Bell, J. (1976). Stuttering: Discoordination of phonation with articulation and respiration. Journal of Speech and Hearing Research, 19, 509 - 522.

Peters, H. F. M., Hulstijn, W., & Starkweather, C.W. (1989). Acoustic and physiological reaction times of stutterers and non-stutterers. Journal of Speech and Hearing Research, 32, 668-680.

Peters, H.F.M., & Hulstijn, W. (1987). Programming and initiation of speech utterances in stuttering. In H.F.M. Peters and W. Hulstijin (Eds.), Speech motor dynamics in stuttering. Springer Verlag, Wien.

Pooja, K., & Savithri, S. R. (2003). Speech motor programming in stutterers: a pre-post therapy comparison. Research at AIISH: Dissertation Abstracts. Vol. III. Geetha, Y. V., & Vijayalakshmi, B (Eds.). pp 82-94. Mysore: AIISH publications.

Postma, A., & Kolk, H. (1993). The cover repair hypothesis: Prearticulatory to repair processes in normal and stuttered disfluencies. Journal of Speech and Hearing Research, 36, 472-487.

Rastatter, M. P., & Dell, C.W. (1987). Reaction times of moderate and severe stutterers to monaural verbal stimuli: some implications for neurolinguistic organization. Journal of Speech and Hearing Research, 30, 21–27.

Riley, G. D. (1986). Stuttering severity instrument for children and adults. Texas: pro-ed, Inc.

Sakai, K., Hikosaka, O., & Nakamura, K. (2004). Emergence of rhythm during motor learning. Trends in cognitive sciences, 8, 547-53.

Schmidt, R. A. (1988). Motor control and learning, a behavioural emphasis, Ilionois: Human Kinetics Publishers.

Starkweather, C. W., Franklin, S., & Smigo, T.M. (1984). Vocal and finger reaction times in stutterers and nonstutterers: Differences and correlations. Journal of Speech and Hearing Research, 27, 193 - 196.

Starkweather, C. W. (1982). Stuttering and laryngeal behaviour: A review. ASHA Monographs, 21, 1-45.

Sternberg, S., Monsell, S., Knoll, R. L., & Wright, C. E. (1978). The latency and duration of rapid movement sequences: Comparisons of speech and type writing; In G. E. Stelmach, (Eds.), Information processing in motor control and learning (pp. 117-152). New York, Academic Press.

Till, J., Reich, A., Dickey, S., & Seiber, J. (1983). Phonatory and manual reaction times of stuttering and non-stuttering children. Journal of Speech and Hearing Research, 27, 171-180.

Van der Merwe, A., McNeil, M. R., Robin, D. A., & Schmidt, R. A. (1997) (Eds.), A theoretical framework for the characterization of pathological speech sensorimotor control. Clinical management of sensorimotor speech disorders (pp. 1-25). New York, Thieme.

Van Riper, C. (1982). The nature of Stuttering. Englewood Cliffs, Prentice hall.

VanLieshout, P. H. H. M. (1995). Motor planning and articulation in fluent speech of stutterers and non stutterers. Nijmegen: Nijmegen University Press.

VanLieshout, P. H. H. M., Hulstijn, W., & Peters, H. F. M. (1996). Speech production in people who stutter: Testing the motor plan assembly hypothesis. Journal of Speech and Hearing Research, 39, 76-92.

Webster, W. G. (1986). Neuropsychological models of stuttering- II. Interhemispheric interference. Neuropsychologia, 24, 737-41.

Webster, W. G., & Ryan, C. R. (1991). Task complexity and manual reaction times in people who stutter. Journal of Speech and Hearing Research, 34, 708-14.

Wijnen, F., & Boers, I. (1994). Phonological priming effects in stutterers. Journal of Fluency Disorders, 19, 1-20.

Wingate, M. E. (1976). Stuttering; theory and treatment. New York, Academic press.

Wright, D.L., Black, C.B., Immink, M.A., Brueckner, S., & Magnuson, C. (2004). Long term motor programming improvements occur via concatenating movement sequences during random but not blocked practice. Journal of Motor Behavior, 36, 39-50.

Zimmermann, G. (1980). Articulatory dynamics of fluent utterances of stutterers and nonstutterers. Journal of Speech and Hearing Research, 23, 122-136.


150

Verbal Perseveration in Malayalam - English Bilinguals Manju Mohan P.1 & Swapna N.2

Abstract

The present study intended to investigate the nature of perseverations in Malayalam-English bilingual typical elderly individuals and compare the perseveratory errors between their languages viz. Malayalam, (their mother tongue), a language spoken by the native people of the state of Kerala, in South India and English which was acquired later in life. The study included twenty typical bilingual {first language (L1)-Malayalam and second language (L2)-English} elderly individuals and twenty age and gender matched monolingual (Malayalam speaking) elderly individuals within the age group of 60-80 years. The subjects were tested in two language conditions using the following five tasks: confrontation naming, generative naming, picture description, word definition and, question-answering. The audio recorded responses were transcribed and the data was subjected to appropriate statistical analysis. The results revealed significant lesser perseveration in bilingual individuals which highlighted the presence of cognitive advantage in the bilingual group. There was no significant difference between the type and frequency of perseverations across both languages of a bilingual. Further, no gender and age related differences were seen in perseveration. The study supports bilingualism as a type of cognitive stimulation. Also it extends support for the disinhibition account of perseveration as well as for the inhibitory deficit hypothesis of language and aging. The study implies the striking need for a deeper understanding of perseveratory phenomenon so as to reflect on the potential of this particular cognitive linguistic behavior as a sensitive cognitive linguistic measure.

Key words: bilingualism, verbal perseveration, cognitive advantage

peech-Language 1Pathologists‟ (SLPs) in particular are concerned with the study of speech and language characteristics of elderly

individuals as it has direct implications on clinically aging population. Speech errors such as verbal perseveration in specific, whether from normal or impaired speakers, provide data that reveals the nature of linguistic representations and the cognitive mechanisms underlying the production of words and sentences. Verbal perseverations are speech errors in which the flow of speech is disturbed by the intruding material that comes from the preceding speech. As defined by Sandson and Albert (1984) perseverations are the inappropriate recurrence or continuation of an earlier response.

Perseveratory errors reflect the malfunction of fundamental mechanisms of the normal language processor that can be disturbed by brain damage, by circumstances that stress the unimpaired language system in some way (e.g., increasing rate of speech), or by aging. This malfunction can operate at different levels of neurocognition to produce distinctly different kinds of perseverative symptoms. The most widespread taxonomy of perseveration classifies perseveration into continuous, stuck-in-set and recurrent types and implicates disturbances of adrenergic, dopaminergic and cholinergic neurotransmitter systems respectively. The descriptions of the different types of perseveration are provided in Table 1.

1e-mail:[email protected]; 2 Lecturer in Speech Pathology, AIISH, Mysore, nsn112002 @ yahoo. com

These neurochemical systems exert its effect on fundamental cognitive mechanisms that are thought to influence verbal perseveration, such as working memory, planning, shifting of cognitive sets and attentional processes. Thus perseveratory errors are expected in normal aging also, apart from brain damaged patients, as a result of decline in cognitive functions, specifically the executive functions. Changes in inhibitory control, component of executive function is the first to decline during cognitive aging (Bedard, Nichols, Barbosa, Schachar, Logan & Tannock, 2002). There are two different theories of perseveration namely the „disinhibition theory‟ (Vitkovitch & Humphreys, 1991; Wheeldon & Monsell, 1994) and the „underlying language processing breakdown theory‟ (Dell, 1986; Dell, Burger & Svec, 1997; Cohen & Dehaene, 1998). According to the disinhibition theory or competing activation account the residual activation from the prior response interferes with the person‟s ability to retrieve a new response from long-term memory because its representations have been recently activated. This activation interferes with activation of the current target, resulting in the erroneous and perseverative selection of the prime. While according to the underlying language processing breakdown or reduced language-processing efficiency account two vital components lead to perseveration: 1) weakened activation of a target at any processing level (e.g., semantic, phonological), 2) normally existing persistent activation from previous responses. In this sense, the persistent activation from a previous

S

Verbal perseveration in bilinguals

151

Table 1. Types of perseveration along with their description* Type of

perseveration Motoric description

Areas of brain involved

Continuous

Abnormal repetition of a response token without cessation Stimulus: “Name this picture” (of a dog) Response: “Dog Dog Dog Dog”

Damage to thalamus, arcuate fasciculus and deep nuclei of subcortical structures Right hemisphere damage Norepinephrine depletion

Stuck-in-set

Inappropriate maintenance of a response type even though task demands have changed Stimulus: “Now point to the picture of the dog” Response: Continues to name, not point to, „dog‟

Left frontal lobe &/ mesolimbic frontal damage Dopamine depletion

Recurrent

Repetition of a previous response token to a subsequent stimulus within an established task set (Has pointed to dog and book) Stimulus: “Now point to the picture of the table” Response: Points to the „dog‟

Posterior left hemisphere damage, Left temporal/ parietal damage Acetylcholine depletion

*(Sandson & Albert, 1987)

response overcomes that of the target due to weakened activation of the target, rather than being due to interference from the previous response as earlier studies suggested.

There are two highly influential models proposed to explain these language changes during cognitive aging viz. the „transmission deficit hypothesis‟ (MacKay & Burke, 1990) and „inhibitory deficit hypothesis‟ (Hasher & Zacks, 1988; Hasher, Zacks & May, 1999). According to the transmission deficit hypothesis the weakened connection strength caused by disuse or aging produces transmission deficits that can impair activation, resulting in retrieval failure (Burke, MacKay, Worthley & Wade, 1991) while inhibitory deficit hypothesis states that older adults are less able to inhibit irrelevant information than young adults.

But it is not adequately explored whether these

cognitive linguistic changes with definite underlying neural mechanisms affect the linguistic representations of both languages of a bilingual similarly or not. There are reports which suggest that the second language is learned through explicit memory while first language is acquired through implicit memory in late bilinguals (Paradis, 2004a). Under these circumstances, it can be assumed that, as different cognitive processes are involved in first and second language acquisition, there may be some differences in the underlying neural mechanisms of both languages of a late bilingual.

Research into bilingual language processing

report bilingual advantages in nonverbal executive control in both children (Bialystok, 2001; Carlson & Meltzoff, 2008), and adults (Bialystok, Craik, Klein & Viswanathan, 2004; Bialystok, Craik & Ryan, 2006;

Costa, Hernandez & Sebastian- Galles, 2008). This advantage has been attributed to the enhancement of executive processes through their constant involvement in the resolution of conflict between the two competing language systems, in bilingual language production. The extended experience of bilingualism thus builds up cognitive reserve and protects against the onset of dementia (Bialystok, Craik & Freedman, 2007).

It is presumed that perseveration may serve as a

better behavioral tool to study second language changes during healthy aging as it has an implicated neurochemical substrata. The various neurochemical modulations have an effect on fundamental cognitive mechanisms particularly executive functions such as working memory, planning, shifting of cognitive sets and attentional processes which are thought to influence verbal perseverations. This offers the opportunity of considering perseveratory measures to assess bilingual cognitive advantages and to study whether these effects persists into old age. The current study also realizes the complete lack of any theoretical accounts or models proposed to study bilingual perseveratory behavior.

There is a lack of objective data concerning the

nature and occurrence of perseverative behavior in the aging population. The study of perseveratory characteristics in the normal aging population would help an SLP to screen/evaluate geriatric clinical population for their speech and language deficits, if any. In treating the brain damaged population, there is no definitive „cure‟ for perseverative errors till date. Thus the increased knowledge of the underlying nature of perseverative errors will assist SLPs in the treatment of this problematic symptom. Investigations to study perseveration have also raised diverse opinion on the


152

potential of the tasks that are used to elicit perseveration. Research in this direction may reveal the various underlying language mechanisms during aging and this may provide evidences to bridge the gap between the seemingly diverse theoretical accounts of perseveration as well as the cognitive models on language and aging by substantiating or contradicting the existing literature. Moreover, all the studies done so far, with respect to perseveration were in monolingual population. There is a pressing need to conduct similar studies in bilingual population as it may reveal interesting findings regarding the bilingual language representation, processing and bilingual cognitive advantages. It still remains unclear whether bilingual cognitive advantages persist into old age. Perseveratory errors may be used as a tool to examine this. Perseverations may also reflect language specific changes during healthy aging in bilingual elderly and research in this direction may also pave way into the less explored frontiers of second language loss. Keeping these in view, this study was planned. The specific objectives of the study were (1) to look for the type and frequency of perseverations if any, in bilingual normal elderly individuals and compare these with that of monolingual age and gender matched individuals to specifically examine the existence of bilingual cognitive advantage and (2) to compare the nature of perseveratory errors, in the first language, L1 (Malayalam) and second language, L2 (English) of Malayalam-English bilingual elderly speakers. In addition, the performance of the subjects with reference to age and gender was analyzed.

Method

The study included twenty normal bilingual elderly speakers (L1-Malayalam, L2-English) and twenty normal monolingual elderly speakers in the age group of 60-80 years. In each age group viz. 60-70 years and 70-80 years, ten monolingual and ten bilingual speakers were included with equal number of males and females in each group. The subjects had to satisfy a set of inclusionary criteria of which scoring in the “no cognitive impairment category” (severity score – 24 - 30) of Malayalam Mini-Mental State Examination, M - MMSE, (Mathuranath, Hodges, Mathew, Cherian, George & Bak, 2004) and obtaining a minimum score of „4‟ on each of the 4 macro skills in the International Second Language Proficiency Rating Scale (ISLPR) (Ingram & Wylie, 1997) on the second language proficiency for the Malayalam-English bilingual group were also considered. The subjects were tested individually in a quite environment in two language conditions (Malayalam and English) using five tasks viz. confrontation naming, generative naming, picture description, word definition and

question-answering, so as to elicit perseveration. The materials under each of the tasks were selected from already published test materials. The responses were audio recorded, transcribed and analyzed for the type and frequency of perseverations. If perseverations were absent, a score of „0‟ was given and if perseverations were present, a score of „1‟ was given for each perseveratory utterance. The data was then compared for the frequency of perseveration for each subject on each task as a ratio which was finally converted to percentage using the following formula.

Percentage of perseveration= Total number of perseveration X 100

Total number of utterances

The total percentage of perseveration for each task was computed for each subject in a similar manner. The data were subjected to statistical analysis using SPSS (version 16.0 package).


The mean and standard deviation for each subject belonging to each language status (bilingualism and monolingualism) and each language condition (Malayalam and English) along with the type of perseveration were calculated. These results have been presented and discussed below under separate sections.

I. Comparison of perseveration between the groups: a. Frequency of perseveration: The mean percentage of perseveratory errors and standard deviation across the groups of differing language status (bilingualism versus monolingualism), across two language conditions (Malayalam and English) and with respect to age and gender are depicted in Table 2.

The mean percentage of perseveration obtained for the bilingual speakers was lesser (M= 2.15, 2.25 in L1 and L2 respectively) compared to the monolingual speakers (M= 3.65). Two-way ANOVA revealed a significant difference in perseveration between both the bilingual and monolingual speakers (F (1, 36) = 32.93, p<0.05). The bilingual speakers had significantly lesser percentage of perseveration which could be because of the bilingual speaker‟s advantage in nonverbal executive control (Bialystok et al., 2004; Bialystok et al., 2006; Costa et al., 2008). Moreover, in individuals who are bi/multilingual, activation of lexicons are facilitated (Finkbeiner, Forster, Nicol & Nakamura, 2004), thereby gaining proficiency in both languages. Higher states of activation enhance accurate selection and thus diminish the chances of occurrence of perseveration or any other linguistic errors, thus supporting the results of the present study. The current finding also suggests that bilingual cognitive advantages persists even in the elderly individuals and


153

continue to influence changes in cognitive processing in bilingual older adults and supports that cognitive processing can be modulated by bilingualism.

Table 2. Mean percentage of perseveration (M) and Standard Deviation (SD) with respect to age, Gender

(G), language status (LS, Bilingualism Vs Monolingualism) and language conditions (LC, (Malayalam Vs English) for the bilingual and

monolingual group

*LS, LC & G 60- 70 yrs 70 - 80 yrs Total M SD M SD M SD

BL1M 2.05 0.99 2.25 0.68 2.15 0.81 BL1F 2.08 1.08 2.24 0.59 2.16 0.82 BL2M 2.12 0.77 2.36 0.97 2.24 0.84 BL2F 2.18 0.85 2.35 0.71 2.26 0.74 ML1M 3.55 0.80 3.77 0.95 3.66 0.84 ML1F 3.54 0.68 3.74 1.03 3.64 0.83

*BL1M: Bilingual male - percentage of perseveration in L1; BL1F: Bilingual female - percentage of perseveration in L1; BL2M: Bilingual male - percentage of perseveration in L2; BL2F: Bilingual female - percentage of perseveration in L2; ML1M: Monolingual mal - percentage of perseveration in L1; ML1F: Monolingual female - percentage of perseveration in L1

Thus bilingualism can be considered as a life

style factor that involves sustained complex mental activity which can add on to behavioral brain reserve/cognitive reserve and thus delay the rate of cognitive decline. This finding is in consonance with the studies by Bialystok et al., (2007), Rajsudhakar and Shyamala (2008) and Vijay Kumar and Prema (2010). Thus perseveration can be used as a behavioral measure to assess bilingual cognitive advantage.

The mean percentage of perseveration obtained

in the study is also very less for both the groups (3.65% and 2.15% in L1, Malayalam for monolingual and bilingual speakers) compared to the significantly greater percentage of perseveration reported in the brain damaged population (Mukunthan & Prema, 2003). Ramage, Bayles, Helm-Estabrooks and Cruz (1999) concluded that in normal aging individuals the frequency of perseverations was less (4%) and a significant difference existed between normal elderly individuals and individuals with brain damage in terms of perseveration which can be used to differentiate both the groups. Chandralekha and Prema (2003) reported 3.6% of perseveration in the higher age group of 75-80 year old normal elderly subjects included in their study. Likewise Bayles, Tomoeda, Mc Knight, Helm-Estabrooks and Hawley (2004) also reported significantly less amount of perseveration in normal

elderly speakers (8.5%) compared to persons with Alzheimer‟s Disease (30%). Preethi and Goswami (2010) in their study could elicit only 2.23% of perseveratory errors in normal elderly controls while in Alzheimer‟s Disease, the percentage of perseveration obtained was around 11.69%. Thus the perseveratory percentage obtained in the current study is comparable with the studies done previously in normal aging population.

b. Type of perseveration elicited: The type of perseveration seen in both languages of the bilingual and monolingual speakers in the current study was the recurrent type. There were no instances of continuous or stuck-in-set type of perseverations. But, there were reports of significant percentage of continuous type of perseveration compared to other types of perseveration in normal elderly Tamil speaking population in the study by Chandralekha and Prema (2003) and Mukunthan and Prema (2003). Preethi and Goswami (2010) reported of both continuous and recurrent types of errors in normal Malayalam speaking elderly. However, the results of the current study are in consonance with the study by Troster, Salmon, McCullough and Butters (1989) and Bayles, Tomoeda, McKnight, Helm-Estabrooks and Hawley (2004), in which only recurrent perseveration was seen during generative naming task. Thus it can be presumed that generative naming tasks may be more sensitive towards eliciting recurrent perseveration than other types of perseverations. The absence of other types of perseveration in the current study, could also be attributed to the strict inclusionary criteria used wherein the mental status of the subjects were screened using Malayalam version of Mini Mental State Examination (M-MMSE) before including them in the study. It is apparent from various studies that continuous and stuck-in-set types of perseveration are seen significantly in brain damaged population (Pekkala, Albert, Spiro & Erkinjuntti, 2008; Preethi & Goswami, 2010). Pekkala et al., (2008) studied Alzheimer‟s disease (AD) patients and reported that the different types of perseveration are likely to reflect the progressive deterioration of different brain regions. In their study with dementic patients, they could elicit recurrent and continuous perseverations in early stages of AD. As the disease progressed in severity into moderate stage, the number of recurrent and continuous perseverations increased and stuck-in-set perseverations emerged. According to Yamadori (1981) continuous perseverations are seen when there is a complete failure of post activation inhibition mechanism, and recurrent would reflect partial failure of that mechanism. It can thus be speculated that continuous and stuck-in-set perseverations may be reflecting more severe disruptions in post activation


154

mechanisms than the recurrent type. Thus the current study presumes that the different types of perseveration, recurrent, continuous and stuck-in-set falls into a continuum with recurrent appearing even with slight constraints over inhibitory mechanisms and stuck-in-set with more severe disruptions in inhibitory mechanisms.

Among the five different tasks employed, only

generative naming task was found to elicit perseverative errors in both monolingual and bilingual speakers. The other tasks namely, confrontation naming, picture description, defining words and answering questions failed to elicit perseverative errors. This finding is in consonance with the findings reported by Bayles et al., (2004). According to them, the task difficulty influenced the rate of occurrence of perseveration. According to Craik (1986) the cognitive and language processes may vary according to changes in tasks, materials and strategies. Generative naming is a more difficult task as it assesses verbal fluency both letter and category. Generative naming requires actively searching for the lexicon in the semantic buffer, retrieving the target item and finally stating the names of the items rapidly (Bayles et al., 2004). Thus generative naming is likely to recruit additional processing mechanisms than just lexical retrieval including executive functioning and short term memory. Preethi and Goswami (2010) also reported that generative naming yielded highest mean percentage of errors, which occurred due to increased cognitive demands.

Confrontation naming of pictures was another

task that was carried out. Here since the stimuli were presented visually, they provide a perceptual additional cue for the lexicon retrieval from the memory. According to Bayles et al., (2004) this cue offers increased activation in the semantic system and reduces the stress on the working memory during the retrieval. This plausibility explained as to why confrontation naming resulted in no perseverations. Similarly other tasks such as picture description and question and answer also required less effort as the subjects had sufficient time to recognize and generate ideas. According to Helmick and Berg (1976), the tasks that elicited the fewest number of perseverative responses were defining words and answering questions. These tasks were not bounded by speeded time conditions and thus may not have stressed the language system adequately to elicit perseverations. Rather enough time was given and the subjects were not constrained in any way to produce large number of ideas. Most often, it was observed that they enjoyed the freedom to limit their speech output. This would have affected their overall frequency scores (Bayles et al., 2004). In word

definition tasks too, the subjects had the freedom to give any relevant responses, there were no time constraints or constraints regarding the number of alternate ideas that had to be produced. Even though this task tests divergent thinking of cognitive flexibility, the system may have still enjoyed the flexibility as no constraints were imposed over cognitive linguistic processing.

However, the finding of perseveration as being

seen in only the generative naming task is in contrast to several other studies. In the Indian scenario, it contrasts the studies by Chandralekha and Prema (2003), Mukunthan and Prema (2003) and Preethi and Goswami (2010), wherein they could elicit perseveration in various tasks other than generative naming alone. This difference could be because of the subject selection criteria used in the present study; the subjects were screened for their mental status unlike the above mentioned studies. In such cases, it can be assumed that in order to elicit perseverations the language processing system needs to be highly constrained.

II. Comparison of perseveratory errors of L1 vs. L2 in bilingual speakers: In the bilingual group, the percentage of perseveratory errors with respect to the two different languages viz. first language (L1) and second language (L2) were analyzed. The mean values obtained were subjected to paired t-test. The mean and standard deviation values in both the languages along with the t-values obtained are depicted in Table 3.

Table 3. Mean (M), Standard Deviation (SD) and t-value of percentage of perseveration for L1 (Malayalam) and L2 (English) in bilinguals

Bilingual Group

M SD t-value(19)

First Language (L1)

2.15 0.80 0.424 (p>0.05)

Second Language (L2)

2.25 0.77

The results of the paired t-test indicated that

there was no significant difference in the percentage of perseverations between both languages for the bilingual speakers (t (19) = 0.424, p>0.05). Moreover, the only type of perseveration seen was recurrent perseveration in both languages of the bilingual. This finding supports the inhibitory deficit hypothesis of language changes with aging and the disinhibition account of perseveration since the perseverations in L1 were more or less similar to perseverations in L2. Inhibitory deficit hypothesis suggests deficient


155

inhibitory processes as contributing to cognitive linguistic changes during aging. According to the disinhibitory account of perseverations, the residual activation from the prior response interferes with the person‟s ability to retrieve a new response from long-term memory because its representations have been recently activated and thus cause perseverations. On the other hand, for the transmission deficit hypothesis and reduced language processing efficiency account of perseveration to have been true, there should have been significantly higher rates of perseverations in L2 compared to L1. This is assumed because there would have been reduced transmission to the linguistic representations of L2 due to less frequent use of second language in bilingual elderly which would then lead to weakened activation of the target words. Thus there are more chances for the persistent activation of previously uttered words to overcome the current target‟s activation levels, consistent with the reduced language processing efficiency account of perseveration, inducing greater perseveration in L2. But as the results did not reveal significant differences in the perseveratory patterns across both languages, the current findings extend support to the inhibitory deficit hypothesis of language and aging as well as to the disinhibition account of perseveration. The findings also support the conjectures put forth by Yamadori (1981) and Pietro and Rigordsky (1986) wherein failure of inhibitory mechanisms is implicated in causing perseverations. Thus from the current study it can be postulated that the changes in inhibitory functioning will affect the bilingual lexicons equally. That is the general cognitive decline associated with aging affects the two languages of a bilingual equally. In other words, the neural level mechanisms associated with general cognitive decline during aging affects the representation of both languages of a bilingual more or less similarly. However, in spite of the above findings, it is too premature to argue on lines of semantic degradation versus lexical access deficits of language representations in second language so as to conclusively postulate that there is no language specific loss/ attrition that are seen in bilingual elderly. III. Effect of age on perseveration: The mean perseveratory errors for the monolingual and bilingual group were analyzed to examine whether any significant difference existed between the age groups. The data was subjected to Mann Whitney U test and the mean, SD and the /z/ values are shown in Table 4.

The combined mean (M) and standard deviation (SD) of all the participants in L1 for the two age groups viz. 60-70 years and 70-80 years were M=2.80 (SD=1.12) and M=3.0 (SD=1.19) respectively.

Table 4. Mean (M), standard deviation (SD) with /z/ values for the two age groups for both groups

*LS, LC & G

60- 70 yrs 70 - 80 yrs /z/ value M SD M SD

BL1M 2.05 0.99 2.25 0.68 0.73 BL1F 2.08 0.08 2.24 0.59 0.10 BL2M 2.12 0.77 2.36 0.97 0.52 BL2F 2.18 0.85 2.35 0.71 0.31 ML1M 3.55 0.80 3.77 0.95 0.31 ML1F 3.54 0.68 3.74 0.03 0.21

*LS- Language status, LC-Language conditions, G-Gender

Similarly the combined mean and standard

deviation of all the participants for the two age groups in L2 were 2.14 (SD=0.76) and 2.35 (SD= 0.80) respectively. The results of the Mann Whitney test showed no significant difference between various age groups across any of the variables. This particular finding is not in consonance with the studies which have reported an age effect on perseveratory errors such as Troster et al. (1989), Daigneault, Braun and Whitaker (1992) and Chandralekha and Prema (2003). But support can be drawn for the current findings from the studies done by Ramage et al. (1999) and Foldi Helm-Estabrooks, Redfield and Nickel (2003), in which they report no age effect on verbal perseveration. As in the study by Ramage et al. (1999), the current study also assessed mental status before including the subjects for testing. Thus the lack of any age effect may be because of the confirmation of absence of pathological cognitive impairment by screening the participants using M-MMSE before including them in the study, which was not carried out in the above mentioned opposing studies. Moreover, the factors of personality, literacy, educational history etc. could be some other factors contributing to this finding. Schooling has also been reported to improve cognitive functioning (Garcia & Guerreiro, 1983; Roselli, Ardila & Rosas, 1990). Socioeconomic status and cultural factors also play a significant role in literacy and cognition (Reis & Castro-Caldas, 1997). Another factor that may play a role would be social engagement, which is defined as the maintenance of many social connections and a high level of participation in social activities and this has been thought to prevent cognitive decline in elderly persons (Bassuk, Glass & Berkman, 1999). It‟s apparent that today‟s older people are much likelier to have had more formal education, higher economic status, and better care for risk factors such as high blood pressure, high cholesterol and smoking that can jeopardize their brains. These results are significant especially in the Indian context for prevention of age related


156

communication disorders. The findings are in consonance with the study by Vijay Kumar and Prema (2010) wherein the authors attribute extraneous variables such as linguistic exposure i.e., bilingualism, life style, culture, profession, physical and communicational activity, physical and mental exercises, dietary habits etc. to the prevention of cognitive rigidity in elderly. Thus the absence of any age effect on the frequency of perseveration may be because of the above mentioned factors which might be contributing to lesser cognitive decline in elderly.

IV. Effect of gender on perseveration: The results of two-way ANOVA revealed no interaction effects between language status and gender [F (1, 36) = 0.002, p>0.05]. Moreover, there was no significant difference between both genders [F (1, 36) = 0.00, p> 0.05] in L1. An independent t-test was used to analyze gender effects in second language, L2 (English). The results revealed that there was no difference between males and females even in L2 (t (18) = 0.068, p>0.05). Thus, on the whole, there was no gender difference that could be found in the current study with respect to the percentage of perseveration. The absence of any gender effect in the current study is not in agreement with the study in Indian Tamil population by Chandralekha and Prema (2003), while it is in consonance with the study by Ramage et al., (1999).

Studies discussing gender differences in

cognitive functions report that women perform at a substantially higher level than men on verbal production, episodic memory, and face recognition tasks while men perform at a higher level on visuospatial tasks. The reasons speculated for these differences include variations in the sex hormones, socio-cultural factors, educational factors, training etc. (Weiss, Kemmler, Deisenhammer, Fleischhacker & Delazer, 2003; Herlitz & Lovén, 2009). A study by Mohan and Shyamala (2009) on the development of stroop effect in bilinguals also showed a substantial difference in the performance between males and females where females outperformed males. But the authors also report that there was an absence of gender effect after the age of 60 years due to the general cognitive decline nullifying the still debated female advantage in language processing. Thus the same explanation can be reasoned out for the absence of gender differences in the present study.

In sum, the results of the present study revealed

that bilingual speakers showed significantly less perseveratory errors compared to the monolingual speakers highlighting the bilingual cognitive advantage. The present study revealed only recurrent perseverations in healthy aging population which

indicates that tenably the various types of perseveration falls into a continuum wherein recurrent perseverations are elicited even with minimal disturbances to language processing system whereas stuck-in-set perseverations are obtained only if the system is severely disturbed. The findings also suggest similar nature of perseveratory errors in both languages of bilingual speakers which permits to foresee that both languages of a bilingual may be equally vulnerable to the general cognitive decline associated with aging. Further, there were no age or gender effects on the frequency of perseveration.

Conclusions

It can be concluded from the study that perseveration is a cognitive linguistic behavior with a neurophysiological basis and reflect deficits in executive functions. The current study extends its support to the disinhibition account of perseveration as well as to the inhibitory deficit hypothesis of language and aging since no significant difference was found between the two languages of bilingual speakers. Moreover, the conclusions drawn from the current study is corroborated with the previous studies, wherein the task difficulty is suggested as a factor which determines the nature of perseveratory errors. The study proposes generative naming, as the most useful task to elicit perseverations, particularly in individuals with adequate cognitive reserve.

The current study adds evidence to the literature

that supports bilingual cognitive advantage and its persistence to old age by using behavioral data on perseveration and highlights the use of perseveratory measures for assessing bilingual cognitive advantage. It also highlights the importance of cognitive stimulation which can delay the devastating effects of cognitive impairments.

The study assumes that the general

neurocognitive changes seen during aging affect the linguistic representations of a bilingual similarly giving some insight in to the less explored frontiers of second language loss in healthy elderly bilinguals, which is a fertile area where research is heavily warranted. The study also has implication with respect to intervention strategies for treating perseveration. Nevertheless, the results of the present study have to be interpreted with caution as the findings are concluded on the basis of the data obtained from a single task (generative naming) which was used for eliciting perseveration. Moreover, there was limited number of subjects within each age group.


157

More systematic and in depth analysis of perseveration especially in terms of its linguistic description is recommended especially in bi and multilingual population so as to elaborate on the currently accepted models on language processing.

Acknowledgements

The authors wish to express their gratitude to Dr. Vijayalakshmi Basavaraj, Director, AIISH for permitting to carry out this study. They also thank all the subjects for their cooperation for the study.

References

Bassuk, S. S., Glass, T. A., & Berkman, L. F. (1999). Social disengagement and incident cognitive decline in community-dwelling elderly persons. Annals of Internal Medicine, 131(3), 165-173.

Bayles, K. A., Tomoeda, K. C., McKnight, E. P., Helm Estabrooks, N., & Hawley, N. J. (2004). Verbal perseveration in individuals with Alzheimer‟s disease. Seminars in Speech and Language, 25(4), 335-347.

Bedard, A. C., Nichols, S., Barbosa, J. A., Schachar, R., Logan, G. D., & Tannock, R. (2002). The development of selective inhibitory control across the life span. Developmental Neuropsychology, 21(1), 93-111.

Bialystok, E. (2001). Bilingualism in development: Language, literacy, and cognition. New York: Cambridge University Press.

Bialystok, E., Craik, F. I. M., Klein, R., & Viswanathan, M. (2004). Bilingualism, aging, and cognitive control: Evidence from the Simon task. Psychology and Aging, 19, 290-303.

Bialystok, E., Craik, F. I. M., & Ryan, J. (2006). Executive control in a modified anti-saccade task: Effects of aging and bilingualism. Journal of Experimental Psychology: Learning, Memory, and Cognition, 32, 1341– 1354.

Bilaystok, E., Craik, F. I. M., & Ruocco, A. C. (2006). Dual modality monitoring in a classification task: The effects of bilingualism and aging. The Quarterly Journal of Experimental Psychology, 59(11), 1968-1983.

Bialystok, E., Craik, F. I. M., & Freedman, M. (2007). Bilingualism as a protection against the onset of symptoms of dementia. Neuropsychologia, 45, 459-464.

Burke, D. M., MacKay, D. G., Worthley, J. S., & Wade, E. (1991). On the tip of the tongue: What causes word finding failures in young and older adults. Journal of Memory and Language, 30, 542-579.

Carlson, S. M., & Meltzoff, A. N. (2008). Bilingual experience and executive functioning in young children. Developmental Science, 11, 282–298.

Chandralekha, C., & Prema, K. S. (2003). Verbal perseveration in normal geriatrics: A study on Tamil speakers. Students Research at A.I.I.S.H. Mysore (Articles based on dissertations done at AIISH), Vol IV, 132.

Cohen, L., & Dehaene, S. (1998). Competition between past and present: Assessment and interpretation of verbal perseverations. Brain, 121, 1641-1659.

Costa, A., Hernandez, M., & Sebastián-Gallés, N. (2008). Bilingualism aids conflict resolution: Evidence from the ANT task. Cognition, 106, 59-86.

Craik, F. I. M. (1986). A functional account of age differences in memory. In F. Klix & H. Hagendorf (Eds.), Human memory and cognitive capabilities, mechanisms, and performance (pp. 409–422). Amsterdam: North-Holland.

Daigneault, S., Braun, C. M. J., & Whitaker, H. A. (1992). Early effects of normal aging on perseverative and non-perseverative prefrontal measures. Developmental Neuropsychology, 8 (1), 99-114.

Dell, G. S. (1986). A spreading activation theory of retrieval in sentence production. Psychological Review, 93, 283-321.

Dell, G. S, Burger, L. K., & Svec, W. R. (1997). Language production and serial order: A functional analysis and a model. Psychology Review, 104, 123-147.

Finkbeiner, M., Forster, K. I., Nicol, J., & Nakamura, K. (2004). The role of polysemy in masked semantic and translation priming. Journal of Memory and Language, 51, 1-22.

Foldi, N. S., Helm-Estabrooks, N., Redfield, J., & Nickel, D. G. (2003). Perseveration in normal aging: A comparison of perseveration rates on design fluency and verbal generative tasks. Aging, Neuropsychology, and Cognition, 10(4), 268-280.

Garcia, C., & Guerreiro, M. (1983). Pseudo-dementia in illiterates. Paper presented at the 6th INS European Conference, Lisbon, 14–17 June.

Hasher, L., & Zacks, R. T. (1988). Cited in Taylor, K. J., & Burke, M. D. (2002). Asymmetric aging effects on semantic and phonological processes: naming in the picture-word interference task. Psychology and Aging, 17(4), 662-676.

Hasher, L., Zacks, R. T., & May, C. P. (1999). Inhibitory control, circadian arousal and age. In D. Gopher & A. koriat (Eds.), Attention & performance XVII: Cognitive regulation and performance: Interaction of theory and application (pp.653-675). Cambridge, MA: MIT Press.

Helmick, J., & Berg, E. (1976). Perseveration in brain-injured adults. Journal of Communication Disorders, 9, 571-582.

Herlitz, A., & Lovén, J. (2009). Sex differences in cognitive functions. Acta Psychologica Sinica, 41 (11), 1081-1090.

Ingram, D. E., & Wylie, E. (1997). International Second Language Proficiency Ratings (ISLPR). In Adult Migrant Education Program Teachers‟ Manual. Canberra: Department of Immigration and Ethnic Affairs.

Mathuranath, P. S., Hodges, J. R., Mathew, R., Cherian, P. J., George, A., & Bak, T. H. (2004). Adaptation of the ACE for a Malayalam speaking population in southern India. International Journal of Geriatric Psychiatry, 19, 1188 – 1194.

Mackay, D. G., & Burke, D. M. (1990). Cited in Taylor, K. J., & Burke, M. D. (2002). Asymmetric aging effects on semantic and phonological processes: naming in the picture-word interference task. Psychology and Aging, 17(4), 662-676.

Mohan, M. P., & Shyamala, K. C. (2009). Development of stroop effect in bilinguals. Language in India Strength for Today and Bright Hope for Tomorrow, 9, 1-11. Retrieved


158

from http:// www. languageinindia. com/ feb2009/ manjustroop.pdf.

Mukunthan, L., & Prema, K. S. (2003). Verbal perseveration in Broca‟s Aphasics: A study on Tamil speakers. Student Research at A.I.I.S.H. Mysore. (Articles based on dissertations done at AIISH), Vol IV, 181.

Paradis, M. (2004a). A neurolinguistic theory of bilingualism. Amsterdam, The Netherlands: John Benjamins Publishing Company.

Pekkala, S., Albert, M. L., Spiro, A., & Erkinjuntti, T. (2008). Perseveration in Alzheimer‟s disease. Dementia and Geriatric Cognitive Disorders, 25, 109-114.

Pietro, S. M., & Rigordsky, S. (1986). Patterns of oral-verbal perseveration in adult aphasics. Brain and Language, 29, 1-17.

Preethi, T. T., & Goswami, S. P. (2010). Verbal perseveration and anticipation in multilingual persons with Alzheimer‟s disease. Student Research at A.I.I.S.H. Mysore. (Articles based on dissertations done at AIISH), Vol. VI , Part B, 217 – 235.

Rajsudhakar, R., & Shyamala, K. C. (2008). Effects of age, gender and bilingualism on cognitive-linguistic performance. Student Research at A.I.I.S.H. Mysore. (Articles based on dissertations done at AIISH), Vol. III, , Part B, 127-146.

Ramage, A., Bayles, K. A., Helm-Estabrooks, N., & Cruz, R. (1999). Frequency of perseveration in normal subjects. Brain and Language, 66, 329-340.

Reis, A., & Castro-Caldas, A. (1997). Illiteracy: a cause for biased cognitive development. Journal of the International Neuropsychological Society, 3(5), 444-450.

Rosselli, M., Ardila, A., & Rosas, P. (1990). Neuropsychological assessment in illiterates II: Language and praxic abilities. Brain and Cognition, 12, 281-296.

Sandson, J., & Albert, M. L. (1984). Varieties of perseveration. Neuropsychologia, 22, 715 -732.

Sandson, J., & Albert, M. L. (1987). Perseveration in behavioural neurology. Neurology, 37, 1736-1741.

Troster, A. I., Salmon, D. P., McCullough, D., & Butters, N. (1989). A comparison of the category fluency deficits associated with Alzheimer‟s and Huntington ‟s disease. Brain and Language, 37, 500 – 513.

Vijay kumar & Prema, K. S. (2010). Cognitive flexibility in aging. Student Research at A.I.I.S.H. Mysore. (Articles based on dissertations done at AIISH), Vol. V, Part B, 246 – 258.

Vitkovitch, M., & Humphreys, G. W. (1991). Perseverant responding in speeded naming of pictures: It‟s in the links. Journal of Experimental Psychology: Learning, Memory and Cognition, 17, 664-680.

Weiss, E. M., Kemmler, G., Deisenhammer, E. A., Fleischhacker, W. W., & Delazer, M. (2003). Sex differences in cognitive functions. Personality and Individual Differences, 35(4), 863-875.

Wheeldon, L. R., & Monsell, S. (1994). Cited in Moses, M. S., Nickels, L, A., & Sheard, C. (2007). Chips, cheeks and carols: A review of recurrent perseveration in speech production. Aphasiology, 21(10/11), 960 – 974.

Yamadori, A. (1981). Verbal perseveration in aphasia. Neuropsychologia, 19, 591-594.

Some aspects of syntax: a computerized approach

159

Some Aspects of Syntax in 4-7 Year Old Typically Developing Malayalam Speaking Children: A Computerized Approach

Maria P. R1 & Swapna N.2

Abstract

The study was designed to investigate the developmental pattern of certain aspects of syntax of Malayalam speaking children in the age range of 4 to 7 years and to find the relationship between Mean Length of Utterance (MLU) and the syntactic development. Three groups of Malayalam speaking children in the age range of 4-7 years with equal age intervals were considered for the study. Using the SALT Software the speech samples elicited through story retell task and conversation task were analyzed for the MLU in words(w) and MLU in morphemes(m), Type Token Ratio (TTR) and other grammatical structures. The results revealed a developmental trend across age groups in MLU. Although MLU did not correlate with age, it correlated with the other grammatical structures studied such as plurals, causatives, Person-Noun-Gender (PNG) markers, tenses, case markers, conditionals, conjunctives, adjectives, adverbs, nouns, verbs, causatives, auxiliaries, interrogatives and negatives. Amongst these structures, adjectives, tense markers and PNG markers showed a significant developmental trend. In addition there was no gender difference in the acquisition of grammatical structures. These findings have implications in the assessment and intervention of children with developmental language disorders. Key words: syntax, mean length of utterance, SALT software

lthough 1we use it every day, and even if we all have strong opinions about its proper form and appropriate use, we rarely stop to think about

the wonder of language. Language is a psychological or cognitive property of humans. That is, there is some set of neurons in the head firing away that allows producing a set of letters, and there is some other set of neurons in the head firing away that allows translating these squiggles into coherent ideas and thoughts. There are many subsystems working simultaneously. Language consists of several components which are interrelated and linked with each other, amongst which the „syntax‟ studies the level of language that lies between words and the meaning of utterances: sentences. The study of syntax is the aspect of how we subconsciously get from sounds to meaning. Syntax is the central component of human language which is a rule system that governs the structure of sentences. It specifies the order that the words must take and the organization of different sentence types.

The learning of language by a child is not just the imitation of an adult model but is an insightful progressive discovery of grammatical structures by the child. This process of acquisition is dependent on the ability of the child to perceive and organize the environment and the language that is part of the environment; in addition the child must relate these two. The development of language and speech, use of rules appropriately occurs over the first few years of life of the child. Children acquire syntax and

1e-mail: [email protected]; 2Lecturer in Speech Pathology, AIISH, Mysore, [email protected].

morphology from its very beginning until the end of pre-school years. During these few years children develop an extremely rich and intricate linguistic system. They go from expressing just a few simple meanings in two words in a systematic manner to expressing abstract and complex ideas in multiword sentences. Brown (1973) introduced the Mean Length of Utterance (MLU) as measure of syntactic development. The addition of each morpheme reflects the acquisition of new linguistic knowledge.

Speech-language pathologists have been involved in the assessment of children‟s language since 1950s. The intervening years have brought diversity in the theories and practices of language assessment. Changing views of the nature of language have spawned new procedures for sampling and describing language and for categorizing deviations from normal language. MLU is one such procedure that has been used for language acquisition studies. Mean Length of Utterance in words-MLU (w) and Mean Length of Utterance in Morpheme-MLU (m) are being used for language analysis. MLU was reported by many authors as a measure that indicates the syntactic complexity and that it correlates with age (Bowerman, 1970; Devilliers & Devilliers, 1973; Miller & Chapman, 1981; Rondal, Ghiotti, Bredart & Bachelet, 1987; Blake, Quoataro & Onorati, 1993). Deepak, Karanth and Deepak (2009) reported that higher the MLU, higher was the percentage of usage of the grammatical forms.

But this viewpoint was contradicted by many who stated that MLU does not have a linear

A


160

relationship with age (Klee & Fitzgerald, 1985; Conant, 1987; Chan, McAllister & Wilson, 1998). Scarborough, Wyckoff and Davidson (1986) and Rollins, Snow and Willet (1996) cautioned in using MLU as a language match. Chabon, Kent-Udolf and Egolf (1982) stated that beyond the age of five years MLU does not correlate well with age. Though lot of research has gone into the field of MLU, studies still need to be conducted to confirm the reliability of MLU. Of late, language sample analysis using computer software is becoming popular in the field as they are time efficient.

Several studies have been carried out in different parts of the world on typically developing children to study the acquisition of various grammatical categories such as negation (Bellugi, 1967; Vijayalakshmi, 1981; Prema & Rangan, 2002; Sreedevi & Thirumalai, 2002), „wh‟ questions (Smith, 1933; Erwin-Tripp, 1970; Mc Grath & Kunze, 1973; Quigley, Wilbur & Montanelli, 1975; Sreedevi & Thirumalai, 2002), „tense‟ markers (Berko, 1958; Vijayalakshmi, 1981; Sreedevi & Thirumalai, 2002) etc. These studies throw light into the process of learning several aspects of language by a child. Further, other language acquisition studies have shown regularities in linguistic performance and acquisition in typically developing children. Popular belief and scholarly opinion have generally maintained that girls are more advanced in language development than boys (Templin, 1957). However studies have also reported of no gender differences in the acquisition of language (O‟Donnell, Griffin & Norris, 1967).

Although there are many studies catering to the needs of speech-language pathologists and language teachers in Western countries in terms of studying the various aspects of morphology and syntax, developing norms for aspects of language development etc., one cannot blindly follow the patterns and norms established for a set of population. Here, though the question of universality in language development presents itself for argument, there is necessity to test this question of universality. To test the above issue, studies have to be conducted on different sets of population of children speaking different languages.

Studies of language development in typically developing children are essential to understand the delays in development and to provide effective and efficient methods of rehabilitation. Moreover, due to the tremendous increase in literacy rate among the general public especially in the females and the early exposure of the children to a wide variety of environment and technology, they develop many language aspects at a much earlier age than previously reported. Hence to understand and study the

development of language, especially in the present day population is crucial.

Several studies have been conducted in the past to investigate a wide range of aspects in language acquisition viz. the development of grammatical structures and syntax, assess whether a linear relationship exists between age, MLU (m) and the grammatical development etc. The results of such studies are mixed as mentioned in the review, in that some studies report of a linear relationship while others deny the existence of such a relationship. Since the relationship between these aspects are inconsistent which is revealed through several studies, further research is required to study the association between these in different languages and age groups.

In the Indian context, in the recent past, although there have been studies conducted on typically developing children to explore the different developmental aspects of grammar, such studies are limited. Further, there is a dearth of research in India with respect to studying the pattern of language development in different Indian languages and analyzing the data through the use of computer software programs specifically developed for the purpose. India offers a highly challenging and interesting ground for studying language acquisition because of her multilingual environment.

Considering that language development may

vary across culture and the data with respect to language acquisition is limited in the Indian context, there is a dire need to carry out detailed studies examining language acquisition. Hence this study was planned which to certain extent will help in overcoming this dearth of information regarding the developmental pattern of syntax. Thus the specific objectives considered were (1) to examine the pattern of development of syntax of Malayalam speaking children in the age range of 4 to 7 years (2) to determine whether there is a relationship between MLU and the acquisition of aspects of syntax in Malayalam speaking children and (3) to analyze gender differences, if any, in the acquisition of syntax.

Method

Participants: Three groups of native Malayalam speakers in the age range of 4-7 years with 10 subjects each in the three age ranges of 4-5, 5-6, 6-7 were considered for the study. The subjects considered were matched for their socio-economic status. Equal number of males and females were selected in each age group. It was made sure that the subjects had no obvious medical history, no history of any sensory deficits or


161

any sort of speech and language disturbances, no oro-motor weakness or deficits and no evident emotional or behavioural disturbance. The WHO Ten-question disability screening checklist (Singhi, Kumar, Malhi & Kumar, 2007) was used to rule out any disability.

Ethical procedures were used to select the participants. The parents were explained about the purpose and the procedures of the study through the phone or in person and an informal verbal and written consent was taken. Test material: The speech samples were collected using two tasks-Story retell and conversation tasks. Task 1: Story retell task: The examiner showed the child a sequence of five pictures that illustrate the critical events in a single-episode story that the examiner narrated. The story that was selected here was “The Thirsty Crow.” Task 2: Conversation task: The subjects were asked simple questions about general information pertaining to themselves such as, “How do you come to school every day?”, “What all games do you play at school?” etc. The examiner also used toys to elicit speech sample from the subjects. Procedure: The children were tested at schools where they were studying. A classroom which was away from noisy environment of the school was selected. The data was collected after building rapport with the child and only when the child was comfortable with the investigator. Positive reinforcement was given whenever necessary. The child was made to sit on a chair and the investigator sat opposite to the child. A portable digital sound recorder was (Sony Digital Recorder ICD-320) was used to record the speech sample. Language samples with approx. 150 utterances were collected.

Analysis Transcription: The examiner listened to the audio recording and transcribed verbatim. The middle 100 utterances were considered for the analysis so as to avoid introductory or closing conversation effects that might bias results, such as shyness or problems focusing on the task at hand. Transcription reliability was conducted on a random 30% of the transcripts and was found to be 98% for individual morphemes. Segmentation: The language samples were formatted to comply with SALT (Systematic Analysis of Language Transcripts) (Miller & Chapman, 2008) transcription conventions and guidelines. Each utterance was given a speaker code, and inflectional morphemes within words were divided by using a slash. All the grammatical structures were coded

according to the conventions for obtaining the code summary option provided in the SALT Software. The definitions of utterance segmentation provided in the SALT manual were used in the present study. According to the SALT manual (Miller & Chapman, 2004), a P-unit represents documentation of a complete thought. Thought completion is generally characterized by a rise or fall in intonation and the presence of a pause. When conjoined and complex sentences do not contain pauses or changing intonation, then thought completion is determined by independent and dependent clauses. In these instances, P-units are separated after two conjoined independent clauses. In the present study the samples were segmented using the P-units.

Analysis using SALT Software: The segmented transcripts were fed into the computer using SALT conventions specified in the software. The SALT (Miller & Chapman, 1985) software (2008) was used to analyze the transcripts for Mean Length of Utterance (MLU) in words (w) and MLU in morphemes (m) and other various grammatical structures. During the analysis disfluencies such as repetitions, fillers, interjections were not considered. Compound words, yes/no responses to questions, abandoned and interrupted utterances, non-verbal utterances, and imitative utterances were not considered. The standard measures option was used to obtain the Type Token Ratio (TTR), MLU (w) and MLU (m) and code summary option was utilized to obtain the scores of other grammatical structures.

Statistical analysis: The raw data was tabulated and further suggested to quantitative analyses. SPSS version 16.0 was used for detailed analyses. The statistical mean scores were obtained and other statistical procedures were applied for further analysis.


The results have been presented and discussed under different sections.

I. Development of syntactic aspects: The mean scores of Type Token Ratio (TTR), MLU (w), and MLU (m) were computed and compared across the three age groups. The comparison revealed that these were higher for the 6-7 year age group compared to the 4-5 year and 5-6 year age group. The mean and standard deviation values for these parameters and other aspects of syntax have been depicted in Table 1.

The results of MANOVA revealed that there was no significant difference in the three parameters TTR, MLU (w) and MLU (m) considered across the three age groups. No significant difference in the


162

scores of TTR across age groups found in this study is in consonance with the study by Klee (1992) who also found similar results. Further Karl Pearson‟s correlation coefficient was carried to find the relationship between age and MLU. The results revealed that there was no significant correlation between the two (r = 0.22, p>0.05). These results are in consonance with the finding by Brown (1973), Scarborough et al. (1986) and Chabon et al. (1982) who found that the relationship between age and MLU was nonlinear beyond the age of 4 years. Further Miller and Chapman (1981) added that beyond the MLU of 4 years, the children would have control over diverse sentence structures and that MLU would probably depend more on the nature of the interaction rather than on the limits of children‟s grammatical knowledge.

Muma (1986) suggested that when MLU

exceeds 4.0, knowledge of formal grammatical mechanisms can no longer be indexed by increments in MLU and contends that the sequence of acquisition and use of grammatical structures is a more useful measure.

A detailed analysis of the language sample revealed the presence of the grammatical structures

such as nouns, verbs, adverbs, auxiliaries, adjectives, conjunctions, tenses, PNG markers, case markers, conditionals, causatives, interrogatives, and negatives and these were analyzed and subjected to statistical analysis.

In the first set, the grammatical structures such

as nouns, verbs, adverbs, auxiliaries, adjectives and conjunctions were studied to examine differences in these across age groups. On comparing the mean values of these structures across the age groups, depicted in Table 1, it was seen that 6-7 year old group had higher mean values for all the structures except auxiliaries. The results of MANOVA revealed that there was no significant difference in all the parameters except for the category of adjectives across age groups [F (2, 27) =5.89, p<0.05]. The Table 2 depicts the F-values across the grammatical structures. The Duncan Post Hoc analysis revealed that the 6-7 year old age group was significantly different from the 4-5 year old group and 5-6 year group at 5% level of significance. This shows that there was a significant increase in the usage of adjectives as a function of age.

Table 1. Mean and standard deviation (SD) scores obtained for Type Token Ratio (TTR), Mean Length of

Utterance in words and morphemes {MLU (w), and MLU (m)} and the different aspects of syntax across age groups

Aspects of syntax 4-5 years 5-6 years 6-7 years Total

Mean S.D Mean S.D Mean S.D Mean S.D

TTR 1.48 0.13 1.44 0.21 1.44 0.13 1.42 0.16 MLU (w) 7.38 1.55 7.00 0.83 8.142 1.25 7.46 1.27 MLU (m) 12.88 2.20 12.22 1.35 13.61 2.28 12.87 2.01

Nouns 49.80 18.34 47.50 11.53 56.70 12.83 51.33 14.58 Verbs 35.20 10.81 36.70 10.85 41.30 8.52 37.73 10.11

Adjectives 13.10 5.40 11.60 5.87 19.50 5.08 14.73 6.31 Adverb 11.50 5.56 10.80 4.80 16.30 5.57 12.86 5.68

Auxiliaries 12.30 6.44 10.80 6.59 10.70 3.71 11.26 5.58 Conjunctions 9.60 3.94 11.30 7.33 11.00 5.83 10.50 5.70

Past tense 28.10 13.0 17.50 5.87 22.40 5.83 22.66 9.65 Present tense 6.10 5.60 6.90 5.98 5.80 2.65 6.36 4.82 Future tense 3.30 1.70 7.60 4.22 11.50 2.87 7.56 4.53 PNG Person 17.90 6.31 9.90 5.04 14.80 7.06 14.20 6.85

PNG Number 4.50 2.50 3.70 3.97 4.80 2.44 4.33 2.95 PNG Gender 7.40 2.22 12.40 2.17 13.70 1.76 11.16 3.40 Case markers 16.00 9.53 17.30 8.19 21.80 8.38 18.36 8.78 Conditionals 1.50 0.83 3.50 4.56 3.30 1.33 2.50 2.80 Causatives 1.00 0.00 1.00 0.00 1.16 0.40 1.08 0.28 Negations 2.88 1.70 4.20 2.44 4.5 1.87 3.53 2.04

Interrogations 3.16 2.71 2.14 1.34 1.85 1.21 2.35 1.81


163

Table 2. F values for the six grammatical structures Grammatical structures F(2,27)

Nouns 1.08 Verbs 0.98 Adjectives 5.89* Adverbs 3.20 Auxiliaries 0.24 Conjunctions 0.21

*p<0.05

The mean values of the auxiliaries were higher in the younger group (Table 1) possibly because of the widely varying rates of language acquisition (Brown, 1973). Moreover, the lack of developmental trend with respect to auxiliaries in the current study could also be due to the task‟s sensitivity to the contextual variables such as the nature of the interaction, rapport between the examiner and the child, the person with whom the child is interacting and the child‟s intent to communicate (Miller & Chapman, 1981). Yet another factor that could have possibly influenced the responses was the task used in eliciting the speech sample. Since it was a story retell task it could be that it underestimated the grammatical capabilities of the older group. Similar finding was also reported in a study by Slobin and Welsh (1973).

Following this the mean values of past,

present and future tense were compared and it was seen that the mean values of the future tense showed a developmental trend with higher mean values for the higher age group, however the past tense usage in the younger group was higher. MANOVA revealed a significant difference in the usage of only past tense [F (2, 27) = 3.55, p<0.05] and future tense [F (2, 27) =17.40, p<0.05]. The Table 3 depicts the F values for all the three tense markers. The Duncan Post Hoc analysis revealed that for the usage of past tense markers there was a significant difference between 4-5 year old group and 5-6 year old group and for future tense markers, there was a significant difference among all the three age groups at 5% level of significance.

The poorer scores obtained by the older group

with respect to the usage of past tense could be because of the nature of the response given. For example, when questions regarding what they liked to do were asked to the older group, they provided with appropriate answers using future tense, however the younger groups often described what they did and what they saw. Hence the speech sample consisted mostly of past tense than the other tenses. According to George and Krantz (1981), young children bring in more utterances from their topic of interest and it has been seen that fewer than 20 percent of the preschooler‟s responses

may be relevant to the partner‟s previous utterance. DeThorne, Johnson and Loeb (2005) also stated that variables such as pragmatics influence the linguistic output.

Table 3. F values of the tense markers Tense markers F(2,29)

Past tense 3.55* Present tense 0.13 Future tense 17.40*

*p<0.05

The PNG markers were compared for their means across age groups and it was found that among the three markers, the person marker was used maximally by the 4-5 year age group. The number markers were used almost equally by all the three age groups and the gender marker usage increased with the increase in age. The application of MANOVA for PNG markers revealed that there was a significant difference between age groups for person marker [F (2, 27) = 4.23, p<0.05] and for gender marker [F (2, 27) = 25.9, p<0.05)]. The Table 4 depicts the F-values for the PNG markers. The Duncan Post Hoc analysis revealed that there was a significant difference in the usage of person marker between the 4-5 year old and 5-6 year old groups and a significant difference was seen for the gender marker between all the three age groups at 5% level of significance.

Table 4. The F values of the PNG markers

PNG markers F(2,29) PNG person 4.24* PNG number 0.36 PNG gender 25.99*

*p<0.05

A developmental trend was seen in the usage of gender marker while the same was not seen in the person marker across age groups. This could be attributed to the fact that the younger group used more person markers to describe an event within an utterance (marked on the basis of P-unit as per the SALT guidelines) compared to the older group as their MLU was higher.

Case markers were then considered for analysis.

For case markers the mean values showed a developmental trend, that is the usage of these increased with age. However, the results of MANOVA revealed that there was no significant difference between the three age groups.

A non-parametric test was applied for the

structures namely conditionals, causatives,


164

interrogatives and negatives due to their lesser occurrence in speech samples in all the three age groups. The mean value of all the four structures revealed that the usage of all the three categories increased with age except for the category of interrogatives. This could be attributed to the inquisitive nature of preschoolers (4-6 year group). Kruskal Wallis test revealed that there was no significant difference in all the four structures. Such findings could be attributed to the elicitation task used and the task‟s sensitivity to the nature of interaction and the material presented. This could also be attributed to the highly varying rates of acquisition. On the whole, it was seen that the causatives were used the least by children in all the age groups and the highest occurring grammatical structures were the nouns followed by the verbs.

II Relationship between MLU and the grammatical structures: Karl Pearson‟s correlation coefficient was applied to examine the relationship between the MLU and the usage of grammatical structures and the results revealed that there was a significant correlation between MLU(w) (r = 0.48) and MLU(m) (r = 0.49) and the usage of different grammatical structures at 5% level of significance. Figure 1 represents the correlation between MLU (w) and MLU (m) and grammatical structures through scatter plot.

Thus the results of the present study revealed that the acquisition of different aspects of syntax was directly proportional to the MLU measures. This is in consonance with the study by Devilliers and Devilliers (1973) and DeThorne et al., (2005) who documented that MLU (m) in the English language correlated with the development of morphological and syntactic skills in young children. Deepak et al., (2009) also found that acquisition of syntax was directly proportional to the MLU in Konkani speaking children between 3 and 5 years of age.

Karl Pearson‟s correlation coefficient was also applied to examine the relationship between MLU (w) and MLU (m). The results revealed that there was a significant correlation between MLU (w) and MLU (m) (r = 0.92 p<0.05). The findings are in consonance with similar studies carried out in other language too.

Researchers in Dutch, Irish, Icelandic and English languages have found correlations of 0.98–0.99 between MLU (m) and MLU (w) (Arlman-Rupp, de Haan & Van de Sandt-Koenderman, 1976; Hickey, 1991; Thordardottir & Weismer, 1998; Parker & Brorson, 2005).

Figure 1. Scatter plot of the correlation between MLU(w), MLU (m) and grammatical structures.

This finding suggested that MLU (w) can be used as effectively as MLU (m) as a measurement of a child's gross language development. Though the current study supports the finding that MLU (w) is also a good measure as MLU (m), to know the development in inflections it is important to analyze the MLU in terms of morphemes. Further MLU (m) would allow the comparison of development between languages of different types, thereby assisting in cross linguistic studies.

III . Gender differences in the acquisition of syntax: Mann Whitney test revealed that there was no significant difference between males and females in all the age groups. This in consonance with the study by O‟Donnell, Griffin and Norris (1967). Though Deepak et al., (2009) found differences in MLU in words as well as in morphemes and in the range and variety in the usage of grammatical structures between males and females, such findings were not observed in the present study. This study supports the view of Macaulay (1977) who stated that the female superiority of language might be more of an apparent nature than a


165

real one. If any difference exists it is only of transient nature in language acquisition.

Conclusions

The results of the present study revealed that there was a developmental trend across age groups in MLU, although not significant. A positive correlation was found between MLU and other grammatical structures. That is, with increase in MLU (w) as well as MLU (m) the usage of other grammatical structures also increased. The results also revealed that MLU did not correlate with age. Among the grammatical structures occurrence of structures like plurals, causatives, PNG markers, tenses, case markers, conditionals, conjunctives, adjectives, adverbs, nouns, verbs, causatives, auxiliaries, interrogatives, negatives were seen. Amongst these structures, adjectives, tense markers and PNG markers showed a significant developmental trend. The structures that were most commonly used by the children in the age range of 4-7 were nouns followed by verbs and the least frequently used were the causatives. In addition, that there was no gender difference in the acquisition of grammatical structures.

This study provides an insight into the

development of language especially the syntax in the Malayalam speaking children. However, caution must be taken while drawing inferences from this study given the small number of participants. Nevertheless, this study has important implications for early childhood assessment and intervention. The results suggest that it is crucial to examine several aspects of language particularly the morphosyntax since only then can the relationships can be revealed. The knowledge of the language acquisition in typically developing children speaking different languages belonging to different cultures would further help in the early identification and assessment of children with developmental language disorders and will help speech-language pathologists to arrive at an accurate and appropriate diagnosis. This would in turn contribute to the provision of early intervention services.

References

Arlman-Rupp, A., de Haan, D., & Van de Sandt-Koenderman, M. (1976). Brown‟s early stages: Some evidence from Dutch. Journal of Child Language, 3, 267-274.

Bellugi, U. (1967). The acquisition of negation. Unpublished Doctoral Dissertation. Harvard University.

Berko, J. (1958). The child‟s learning of English morphology. Word, 14, 150-177.

Blake, J., Quartaro, G., & Onorati, S. (1993). Evaluating quantitative measures of grammatical complexity in spontaneous speech samples. Journal of Child Language, 20, 139-152.

Bowerman, M. (1970). Reorganizational processes in lexical and syntactic development. In E. Wanner & L. Gleitman (Eds.), Language acquisition: The state of the art. New York: Cambridge University Press.

Brown, R. (1973). A first language: the early stages. Cambridge, MA: Harvard university press.

Chabon, S. S., Kent-Udolf, L., & Egolf, D. B. (1982). The temporal reliability of Brown‟s mean length of utterance (MLU-M) measure with post-Stage V children. Journal of Speech and Hearing Research, 25, 124-128.

Chan, A., McAllister, L., & Wilson, L. (1998). An investigation of the MLU-age relationship and predictors of MLU in 2- and 3-year-old Australian children. Asia Pacific Journal of Speech, Language and Hearing, 3, 97-108.

Conant, S. (1987). The relationship between age and MLU in young children: A second look at Klee and Fitzgerald‟s data. Journal of Child Language, 14, 169-173.

Deepak, D. R., Karanth, P., & Deepak, D. T. (2009). Mean length of utterance and syntax in Konkani. Retrieved from www. Language in India. Com.

DeThorne, L. S., Johnson, B. W., & Loeb, J. W. (2005). A closer look at MLU: What does it really measure? Clinical Linguistics & Phonetics, 19, 635-648.

DeVillers, P. A., & DeVillers, J. G. (1972). Early judgement of semantic and syntactic acceptability by children. Journal of Psycholinguistic Research, 1, 299-310.

DeVillers, J. G., & DeVillers, P. A. (1973). A cross-sectional study of the acquisition of grammatical morphemes in child speech. Journal of Psycholinguistic Research, 2, 267-278.

Erwin-Tripp, S. (1970). Discourse agreement: How children answer. In J. R. Hayes (1970). How children answer questions in cognition and the development of language. John Wiley & Sons: New York.

George, S. W., & Krantz, M. (1981). The effects of preferred partnerships on communication adequacy. Journal of Psychology, 109, 245-253.

Hickey, T. (1991). Mean length of utterance and the acquisition of Irish. Journal of Child Language, 18, 553-569.

Klee, T. (1992). Developmental and diagnostic characteristics of quantitative measures of children‟s language production. Topics in Language Disorders, 12, 28–41.

Klee, T., & Fitzgerald, M. D. (1985). The relation between grammatical development and mean length of utterance in morphemes. Journal of Child Language, 12, 251-269.

Macaulay, R. K. S. (1977). The myth of female superiority in language. Journal of Child Language, 4, 211- 224.

McGrath & Kunze, L. H. (1973). Development of phrase structure rules involved in tag questions elicited from children. Journal of Speech and Hearing Research, 16, 498-512.

Miller, J. F., & Chapman, R. S. (1981). The relationship between age and mean length of utterance in morphemes. Journal of Speech and Hearing Research, 154-161.

Dissertation, Vol. VIII, 2009-10, Part-B, SLP, AIISH, Mysore

166

Miller, J. F., & Chapman, R. S. (1985). Systematic analysis of language transcripts: User‟s guide. Madison: University of Wisconsin Language Analysis Laboratory.

Miller, J. F., & Chapman, R. S. (2004). Systematic analysis of language transcripts (SALT, v8.0) [Computer software]. Madison, WI: Language Analysis Laboratory, Waisman Center, University of Wisconsin-Madison.

Miller, J. F., & Chapman, R. S. (2008). Systematic analysis of language transcripts (SALT software) 8. Madison: University of Wisconsin Language Analysis Laboratory.

Muma, J. R. (1986). Language acquisition: A functionalist perspective. Austin, TX: Pro-Ed.

O‟ Donnel, R. C., Griffin, W. J., & Norris, R. C. (1967). Syntax of K.G and elementary school children: a transformational analysis, (Research report No: 8), Champaign, III: National Council of teachers of English.

Parker, M., & Brorson, K. (2005). A comparative study between mean length of utterance in morphemes (MLUm) and mean length of utterance in words (MLUw). First Language Acquisition, 25 (3), 1-12.

Prema, K. S., & Rangan, K. (2002). Some aspects of syntax of 5-6 years old children: A descriptive study in Kannada. Research at AIISH, Dissertation abstracts: vol. I, 133.

Quigley, S. P., Wilber, P. B., & Montanelli, D. S. (1974). Question formation in the language of deaf students. Journal of Speech and Hearing Research, 17, 699-713.

Rollins, P. R., Snow, C. E., & Willett, J. B. (1996). Predictors of MLU: Lexical versus morphological developments. First Language, 16, 243-259.

Rondal, J. A., Ghiotto, M., Brédart, S., & Bachelet, J. F. (1987). Age-relation, reliability and grammatical validity of measures of utterance length. Journal of Child Language, 14, 433-446.

Scarborough, H. S., Wyckoff, J., & Davidson, R. (1986). A reconsideration of the relation between age and mean length of utterance. Journal of Speech and Hearing Research, 29, 394–399.

Singhi, P., Kumar, M., Malhi, P., & Kumar, R. (2007). WHO ten questions screen for disability detection. Journal of Tropical Pediatrics, 53, 6, 383-387.

Slobin, D., & Welsh, P. (1973). In J. Harris (1973). Early language development: implications for clinical and educational practice, London: Routledge.

Smith, G. (1933). The influence of age, sex and situation on the frequency, form and function on questions asked by preschool children. Child Development, 4, 201-213.

Sreedevi, S. V., & Thirumalai, M. S. (2002). The aspects of acquisition of Kannada by 2+ year old children. Research at AIISH, Dissertation abstracts: vol. I, 77.

Templin, M. (1957). Certain language by skills in children: their development and interrelationships, Minneapolis, MN: The University of Minnesota Press.

Thordardottir, E.T., & Ellis Weismer, S. (1998) Mean length of utterance and other language sample measures in early Icelandic. First Language, 18, 1-32.

Vijayalakshmi (1981). Test For assessing syntax in Kannada. Unpublished Thesis Dissertation. University of Mysore, Mysore.

Computerized assessment of Phonological Processes-Malayalam

167

Computer Based Assessment of Phonological Processes in Malayalam (CAPP-M) - A Preliminary Attempt

Merin John 1 & N. Sreedevi 2

Abstract

In English, many computerized phonological assessment tools have been developed. In India too, attempts have been made to computerize the assessment of phonological processes. This paper introduces Computerized Assessment of Phonological Processes in Malayalam (CAPP-M), which is a user friendly software. Malayalam Articulation Test (MAT; Maya, 1990) was administered on 30 typically developing Malayalam speaking children, in the age range of 3-3.6 years. Out of the 72 test words administered, 40 target words were errored. Phonological processes for these 40 words were identified. Out of these 40 errored words, 20 most commonly erorred words were selected for software making. The software developed consists of the picture of the target word along with the IPA representation of the correct production of the word along with three other most occurring erroneous productions for the same word. The clinician has to listen to the child‟s production and select one of the four options provided. After the administration of 20 target words, on clicking the option “Report” the software provides the list of processes the child has produced with its frequency. Microsoft.net framework 2.0 is mandatory to run the software. The sensitivity of the software was tested on 6 children with hearing impairment, The administration revealed that CAPP-M was sensitive to the productions of children with hearing impairment. However, CAPP-M should be run on a larger group of normal children aged 3-3.6 years as well as other categories of communication disordered children to evaluate its sensitivity more comprehensively.

Key words: computerized assessment, phonological processes, sensitivity n1 the present days computers have extended its

veins to every field in the world. It spreads its usage in analysis of phonological processes with

less effort and time for its administration. In the recent past, in English many computerized assessment tools for phonological patterns have been developed. One such computerized version is Computer Analysis of Phonological Processes Version 1.0 (CAPP; Hodson, 1985), which yields data on percentage of occurrence for phonological processes described by Hodson (1985). CAPP includes a closed set of 50 words for which the user enters the transcription form produced by the client in a modified IPA format. Another such computerized tool is „Process Analyses -Version 2.0‟ (PAC; Weiner, 1986), which analyses phonetic responses to a closed set of 59 monosyllabic words. The program yields a phonetic inventory of initial and final sounds and a frequency count of several phonological processes. Software named „Programs to Examine Phonetic and Phonologic Evaluation Records -Version 4.0‟ (PEPPER; Shriberg, 1986), provides information about an individual‟s phonetic abilities and the phonological simplifications used. Entry lines can accommodate either single words or connected speech. Masterson and Bernhardt (2001) introduced „The Computerized Articulation and

1e-mail:[email protected]; 2Lecturer in Speech Sciences, AIISH, Mysore, [email protected].

Phonology Evaluation System‟ (CAPES); which evaluate the phonology. The stimuli are photographs of 46 words with various word lengths, structures and stress patterns in the single- word tasks that are displayed on the computer screen. The testee names the items, the computer audio records the responses (which can be played back later), and the clinician transcribes the client‟s words directly into the computer as the test is being administered. The tester can also choose among predicted word productions.

Another computerized version is known as the Hodson Computerized Analysis of Phonological Patterns (HCAPP; Hodson, 2003) is a computer software program that was developed to analyze the major phonological deviations appearing in young children‟s speech. The software is compatible with most popular types of microcomputers available. The programs differ in their scope, ranging from those designed to analyze the responses from a particular test of phonology focusing on the assessment of spontaneous samples of connected speech (Hodson, 1985; Shriberg, 1986; Long, Fey, & Channell, 2002). There is no question that computer application offer the clinician tremendous options for analysis (Masterson 1999; Louko & Edwards, 2001). Ingram and Ingram (2002) advocate using computer- assisted methods for sampling, transcription and storage. It is also possible to interface this sample with various speech analysis programs. One example of such a program is provided by www.sil.org and is called Speech Analyzer. This program is freeware and can

I


168

be downloaded for use in analyzing wave files and subjecting them to spectrographic analysis.

Masterson and Long (2004) introduced

Computer phonological analysis (CPA). This software involves in putting phonetic transcriptions from a keyboard and or by selecting from predetermined stimuli, displaying this data on the screen, and ultimately printing results of an analysis. Analysis often includes both relational and independent analysis of consonants and vowels, word position analysis, syllable shapes used patterns among errors and calculation of consonants correct. Each of the current programs has its own strengths and limitations, and undoubtedly future procedures will add new and helpful procedures for clinicians.

In India an attempt to use computerized

assessment of phonological process is still in the baseline. An attempt was made by Ramadevi (2006) by presenting the target words through Microsoft power point. However, the tedious task of keeping track of the data on a host of different worksheets, tallying up percentages and frequency counts, and cross checking a variety of relationships found in different portions of the client‟s transcript are still done manually. Hence the present study is an attempt to develop a software which reduces the laborious and repetitive manual work involved in traditional phonological analysis. Thus the aim of the study was to develop an indigenous computer based software to assess the phonological processes in native Malayalam speaking children. The specific objectives were (1) to obtain the most common phonological processes in native Malayalam speaking children in the age range of 3-3.6 years, (2) to develop a computer based software based on the normative data collected in collaboration with software engineers to assess (a) the common phonological processes in each child‟s utterance (b) to calculate the frequency of their occurrence (c) to rank the order of phonological processes and (3) to administer the developed software on a few children with communication impairment for sensitivity evaluation of the tool developed on a pilot basis.

Method

The study was carried out in three phases: Phase I: To obtain the most common phonological processes in native Malayalam speaking children in the age range of 3-3.6 years. Phase II: To develop a computer based assessment tool for phonological process analysis in Malayalam language. Phase III: To administer the developed software on 6 children with communication impairment for sensitivity evaluation on a pilot study basis.

Phase I- Obtaining norms Subjects: A total of 30 (15 males and 15 females) typically developing native Malayalam speaking children, in the age range of 3-3.6 years were enrolled in this study. All the subjects were selected from different localities of Alappuzha district, which is located in the southern part of the state of Kerala. Children included for the study had normal speech, language and hearing development, had no known reports of difficulties in behavioral and /or intellectual functioning and were free from any neurological illness or trauma. Material: Malayalam Articulation Test (MAT; Maya, 1990) was administered to all the 30 subjects. 72 test words which intended to test consonants were included excluding the 14 words which were intended to test the vowels out of 86 total target words. The picture stimuli of the 72 test words were recorded in Microsoft Power Point on a laptop computer for better appeal and attraction with the children. Care was taken to depict the target words in colorful pictures of convenient size which were unambiguous and elicited only a single target response. Procedure: In administering MAT, the examiner showed target pictures to one subject at a time and elicited a response. Subject was seated on one side of the examiner. The stimulus picture was presented through Microsoft Power Point mode using a laptop (Dell- Vostro 1400) computer. The responses obtained were audio recorded using a multimedia microphone. If any of the subjects failed to identify a target word, additional cues were presented by the examiner. In spite of the additional cues, if the child failed to name the target picture, „repeat after the examiner mode‟ was used for elicitation. In this way, approximately the test procedure was carried in a time span of 30-40 minutes for each subject. The test procedure involved the following steps. 1. Transcription: The speech sample of all the 30 subjects were transcribed using broad and narrow transcription (IPA, revised to 2005). 2. Identification of the phonological processes: The phonological process were identified by analyzing the whole target word sound by sound and not just the target phoneme in the word. Based on the sound changes in the word, the phonological processes operating were identified. 3. Calculating the percentage of subjects using the processes: As we know, calculating the percentage of occurrence of each process is tedious since determining total opportunities for occurrence of a particular process is very difficult. Also it is inappropriate to derive percentages for phonological processes that have only a few opportunities for occurrence. For example, there are only two

Computerized assessment of Phonological Processes-Malayalam

169

affricates in Malayalam, and then most speech samples would have limited opportunities for de-affrication to occur. According Hodson and Paden (1991), deriving percentages for phonological processes that have fewer than 10 opportunities for occurrence may yield rather skewed results which may give a false impression regarding the importance of the percentage score. Hence in the present study the percentage of children using a particular process was calculated instead of the percentage of occurrence of each process. The percentage of children using a process was calculated by the formula: Percentage of children using a process = Number of children using a process × 100 Total number of children tested 4. Statistical Analysis: Manual statistical procedure was employed to obtain significant difference across gender. Related graphs were generated using SPSS 16 package. Phase II - Preparation of Assessment Software: Preparation of the software involved the following steps 1. Ordering of the target words produced incorrectly : The number of subjects producing each target word erroneously among the test stimuli was calculated. Then the erroneous words were ordered from the maximally errored word to the least erroneously produced word by the subjects. For example, the word /d┗ktԥR/ was incorrectly produced by all the 30 subjects included in the study. The word /sImhԥm/ was erroneously produced by 29 out of 30 subjects. Hence while ordering; the target word /d┗ktԥr/ was followed by the target word /sImhԥm/ 2. Selection of words for the software: Out of the 72 test words administered, children errored only on 40 target words. From the descending ordered list of 40 erroneous words, the first 20 words were selected for including in the software. The details of the list of the erroneous words are provided in the result section. 3. Selection of possible utterances of each target word: For the 20 words selected for inclusion in the software, three varied utterances of the same target word obtained from the sample were required. For this, all the possible productions of the subjects for a particular target word were noted down along with the number of subjects using that particular pattern of production. Among the different possible productions, three patterns which outnumbered the other patterns were selected. For example, for the word /n殴k捲殴t索R殴m/, the different productions included /n殴t検殴t索R殴m/ produced by 3 children out of 30 subjects, /n殴検殴t索殴R殴m/ by 5 subjects, /n殴捲殴t索R殴m/ by 1

subject, /n殴kt検殴t索R殴m/ by 6 subjects, /n殴ktȒ殴t庚殴R殴m/

by 4 subjects, /n殴tȒ殴t庚殴R殴m/ by 6 subjects, /n殴kȑ殴t庚殴R殴m/ produced by 4 and the utterance /n殴kȑ殴t庚殴m/ produced by 1 subject out of 30 subjects. So from the different patterns of productions seen,

the 3 most common patterns i.e., /n殴tȒ殴t庚殴R殴m/, /n殴ktȒ殴t庚殴R殴m/ and /n殴kȑ殴t庚殴R殴m/ were selected. Thus the various patterns obtained for each target word are described in the results. So the 20 most erroneous words along with its 3 most possible commonly occurring patterns were selected for the software. 4. Collaboration with software professionals: Once the 20 words along with their most possible commonly occurring patterns were selected, this material was provided to a software professional in Bangalore (ENFIN Technologies India Pvt Ltd), for the preparation of the analysis software, which was the main objective of the present study. Phase III - Sensitivity Checking Verification of sensitivity on a pilot basis: The developed assessment software was administered on 6 children with hearing impairment, for carrying out the sensitivity checking on a pilot basis.


Phase I intended to obtain the normative data on the phonological processes present in 30 children within the age range of 3.0-3.6 years. Out of 72 test words administered, 40 words were erroneously produced by the subjects. Sound by sound analysis of these 40 words revealed a total of 11 phonological processes operating in the utterances of the 30 subjects tested. The distribution of the phonological processes in all the subjects including males and females are provided in Tables 1 and 2 respectively.

In this study the percentage of the subjects

using a particular phonological process was calculated and not the number of occurrences of each process. Number and percentage of subjects exhibiting different phonological processes in the age range of 3-3.6years old are provided in Table 3.

After obtaining the percentage of subjects exhibiting the various processes, the statistical technique “Equality of Proportion” was used to test the following hypothesis: “the percentage of children exhibiting phonological processes in males is greater than females at 0.05 level of significance”. The result of the statistical test run is shown in Table 4.

The statistical analysis revealed that only the process fronting and metathesis were significantly higher in males compared to females. There was no


170

Table 1. Distribution of different phonological processes in males (M)

Table 2. Distribution of different phonological processes in females(F)

Sl. No.

PRO F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14 F15

1 CR 6 7 5 8 6 6 10 3 3 3 5 1 6 7 6 2 Epn 5 4 5 5 6 4 3 5 1 4 2 6 1 5 1 3 St 1 - 1 - 1 5 3 3 - 2 1 1 4 3 1 4 Fr - 3 - - - - - 1 2 - 3 2 3 - - 5 FCD - - - - - - - - - - - 3 - - - 6 Pal 1 2 2 3 1 - 1 - 1 2 3 1 - 2 2 7 AF 1 3 1 5 - 6 1 9 - 2 4 1 4 1 - 8 DeA - - - - - - - - - - - - - - - 9 DAS - - - - - - - 1 - - - - - - - 10 PrV - - - - - - - - - - - - - - - 11 Mt - - - - - - - - - - - - - - -

Note: Symbols used are CR-Cluster reduction; Epn- Epenthesis, Fr-Fronting, St-Stopping, FCD-Final Consonant deletion, Pal-Palatalization, AF- affrication, DeA- Deaffrication, Das-Deaspiration, PrV-Prevocalic voicing, Mt-Metathesis

significant difference noticed in the percentage of the subjects exhibiting other processes across gender. However it is to be noted that metathesis occurred only in one male subject out of 30 subjects tested. Hence it could be placed under the idiosyncratic category. Phase II: Phase II of the present study was planned for the preparation of a software to assess the phonological processes in children. For this purpose, a word list was developed for including in the software. Out of the 72 test stimuli, 40 words were erroneously produced by children in the age range of 3 -3.6 years. Table 5 shows the number of subjects producing incorrect responses for every target word tested.

The 40 erroneously produced words (Table 5) were ordered in descending order, i.e., in the order of maximally erroneously produced test word to minimally erroneously produced test word by the subjects. From this list, first 20 words were selected for the software. For these 20 words selected, 3 varied utterances of each word obtained from the

sample were listed. For this, all the possible productions of the subjects for a particular target word were noted down along with the number of subjects using that particular pattern of production.

Among the different possible productions three productions which outnumbered the other productions were selected. Few of the words had only two or only one form of error production. The three varied utterances for the target words selected from the sample are listed in Table 6. Thus 20 most erroneous words along with its most possible commonly occurring patterns were ready for the software. This material was provided to a software professional at Bangalore (ENFIN Technologies India Pvt Ltd), for the development of the assessment software.

With the list of words provided, the software

was developed and is named as Computer based Assessment of Phonological Processes in Malayalam (CAPP-M). The 20 test stimuli are represented in colour picture form. The installation and working of the software (CAPP-M) is described below.

Sl. No.

PRO M1 M2 M3 M4 M5 M6 M7 M8 M9 M10 M11 M12 M13 M14 M15

1 CR 5 6 7 2 5 9 5 3 7 2 9 6 7 2 9 2 Epn 2 5 5 5 5 4 4 1 2 4 3 6 4 5 - 3 Fr 2 - - - 1 2 3 1 2 1 2 3 2 2 1 4 St 1 1 1 - 9 4 1 2 4 2 1 1 10 1 1 5 FCD - - - - - - - 1 - - - - 1 1 - 6 Pal - 2 5 - 1 - 1 2 - - _ 2 - 1 - 7 AF - - 1 - 4 8 - 6 7 1 2 3 1 - - 8 DeA - - - - 3 2 - 1 1 1 1 - 4 - - 9 DAS - - - - - - 1 1 - - 1 1 - 2 - 10 PrV - - 1 - - - - - - - _ 1 - - - 11 MT - - - - - - - - - - 1 - - - -


172

Table 4. Presence or absence of significant difference between male and

female subjects

Sl. No Phonological process Raw Scores

Percentage of subjects exhibiting the process

/Z/

Males Females Male Females 1 Epenthesis 14 15 93 100 1.42 2 Fronting 12 6 80 40 2.235* 3 Stopping 14 12 93 80 1.176 4 Final Consonant Deletion 3 1 20 6.6 1.08 5 Palatalization 7 12 46.6 80 0.176 6 Affrication 9 12 60 80 0.117 7 De-affrication 7 1 46.6 6.6 0.235 8 De-aspiration 5 1 33.3 6.6 1.33 9 Pre-vocalic voicing 2 0 13.3 0 1.18 10 Metathesis 1 0 6.6 0 3*

* p < 0.05

Table 5. Number of children producing the test words incorrectly

Sl.No. Words No of children with incorrect production

Sl.No Words No of children with incorrect production

1. d┗kt殴R 30 21 surj殴n 10 2. n殴kȑ殴tR殴m 30 22 s殴斡d識i 13

3. sImh殴m 29 23 Ȓ殴斡kȸ殴 7

4. tȒ殴d庚 r殴n 20 24. mǪȒa 7

5. bR殴球殴 22 25. m┗d庚Ir殴m 1

6. pust庚殴k殴m 20 26. tȒevI 3 7. skut殴R 9 27 al殴marI 4 8. bIsk殴t殴 23 28. rad識av殴 1 9. v殴st 庚 R殴m 28 29. l┗rI 3

10. pRダv殴 12 30. mukȸ殴m 3

11. tȒ殴kR殴m 19 31. ka d庚殴ka宜i 1

12. 球殴Rt殴 25 32. gaǰȹ 殴m 2

13. p┣tȒa 1 33. t 航Ȓȸaja 1

14. R┗d殴 5 34 ȁ殴d殴 5

15. b殴s殴 11 35. ȁ殴ȁul 2 16. RIb殴n 6 36. vIral 2 17 k殴sǪra 12 37 d識殴nil 4

18. RedIo 5 38. udup殴 1 19. pat庚R殴m 21 39 koȊi 1

20. ǪȒu 8 40. pȸal殴m 3

http://en.wikipedia.org/wiki/Retroflex_approximant

Computerized assessment of phonological processes-Malayalam

173

Table 6. Selected words with their most common forms of production

Framework: To run the software, it is mandatory to install Microsoft.net framework 2.0 in the system. So as the first step, Dotnet (.net) framework was downloaded from DotNet framework 2.0 from Google and installed. The software (CAPP-M) can be run in any computer provided the framework dotnet is installed in the system intended for assessment procedure. The program developed is in c#(c sharp) Programming Language.

Working: After the installation of the framework, the software program (CAPP-M) file can be opened. On the opening page there would be the first test stimuli in a color picture form as shown in Figure 2.

The screen contains the picture of the intended target word to be tested on the left corner. Below the picture its correct production is shown in IPA symbols. At the bottom of the page four options are provided towards the left side, i.e. the three most possible patterns of the intended target word along with an option called “any other”. This “any other” option is meant for any other production by the subject which does not fall under the common patterns of production. The right side top portion of the screen contains the analysis report. The report section will have the phonological processes listed one by one along with idiosyncratic productions and correct productions. The report generated will provide the various phonological processes in the child tested and also its frequency of occurrence. This software assesses eight most commonly occurring phonological processes in the age range of 3 – 3.6 years, i.e., cluster reduction, epenthesis, affrication, stopping, palatalization, fronting, metathesis and de-affrication. After the administration of the complete test consisting of the 20 words, the clinician needs to click on “report” to obtain a summary of the processes exhibited by the individual child along with its frequency in descending order.

Sl.No Correct Production

1 2 3

1 d┗kt殴R d┗t殴R d┗kt 庚殴R d┗t 庚殴R 2 n殴kȑ殴tR殴m n殴Ȓ殴t庚殴R殴m n殴ktȒ殴t庚R殴m n殴tȒ殴t庚殴R殴m 3 sImhǨm sImg殴m tȒIm殴m tȒImg殴m 4 bRǨ球Ǩ bǨtȒǨ bRǨȒǨ bǨȒǨ 5 pust庚ǨkǨm put庚ǨkǨm puȒt庚ǨkǨm putȒǨkǨm 6 bIskǨtǨ bIkǨtǨ bIȒkǨtǨ

7 v殴st 庚 R殴m v殴st 庚殴R殴m v殴t 庚 R殴m v殴t 庚殴R殴m 8 球ǨRtǨ tȒǨRtǨ ȒǨRtǨ t 庚ǨRtǨ 9 pat庚R殴m pat庚殴R殴m pat庚殴m - 10 tȒ 殴斡Ǩd 庚 rǨn tȒ 殴斡d庚 ǨrǨn tȒ殴斡d庚 Ǩn -

11 skut殴R kut殴R tȒut殴R -

12 pRダv殴 pダv殴 paRv殴 - 13 tȒ殴kR殴m tȒ殴k殴R殴m tȒ殴k殴m k殴k殴R殴m

14 b殴s殴 b殴Ȓa b殴tȒ殴 b殴t 庚殴

15 k殴sira k殴Ȓira k殴tȒira k殴t庚ira

16 s殴斡d識i t 庚殴斡d識i t 庚殴斡di - 17 Ȓ殴斡ǰȸ殴 tȒ殴斡ǰȸ殴 t 庚殴斡ǰȸa -

18 miȒa mitȒa mit庚a -

19 surjen Ȓurjen t 庚urjen tȒurjen 20 iȒu itȒu it庚u -


174

Phase III – Verification of sensitivity on a pilot basis: Verification of sensitivity on a pilot basis was carried out by administering CAPP-M on 6 children with hearing impairment. The administration reveled that CAPP-M was sensitive to the various patterns of productions of children with hearing impairment. That is to say that CAPP-M could identify the most commonly occurring processes in these children. The details of how many productions of these children matched with the patterns provided in the software are shown in Table 7. For example subject 1 produced 7 productions which matched with the templates (different patterns of production) in CAPP-M. The processes exhibited by these 6 subjects that were identified by CAPP-M are represented in Table 8.

Table 7. Number of productions matching with the templates in the software

Subjects Age (years) No of productions which matched with templates in CAPP-M

Subject 1 3.9 16 Subject 2 5.6 14 Subject 3 2.6 11 Subject 4 4 13 Subject 5 7.6 14 Subject 6 4 15

However the present sensitivity testing has to be interpreted very cautiously as it was administrated only on 6 hearing impaired subjects. From Table 8, it is evident that subjects 3 and 4 had 6 and 7

productions each which did not match with the templates in CAPP-M, hence categorized under the idiosyncratic processes. Therefore it is emphasized that CAPP-M should be run on a larger group of normal children aged 3-3.6 years as well as several other categories of communication disordered children to evaluate the sensitivity of the test more comprehensively.

Conclusions

The present study aimed to develop indigenous software for the assessment of phonological processes in Malayalam speaking children. Based on the normative data obtained from typically developing Malayalam speaking children in the age range of 3.0-3.6 years old, an assessment software tool was developed named Computer based Assessment of Phonological Processes in Malayalam (CAPP-M). The testing can be completed within duration of 8-10 minutes for each subject. This newly developed phonological assessment tool can be effectively used by clinicians for evaluation and post therapy assessment of children with communication disorders in Malayalam. It can thus be concluded that CAPP-M can be used as a quick screening tool for automatic and easy analysis of the phonological processes. However the present sensitivity testing has to be interpreted very cautiously as it was administrated only on 6 hearing impaired subjects. Therefore it is emphasized that CAPP-M should be run on a larger group of normal children aged 3-3.6 years as well as several other categories of communication disordered children to evaluate the sensitivity of the test more comprehensively.

Table 8. Phonological processes exhibited by 6 children with hearing impairment identified by CAPPM

Note: Symbols used are CR-Cluster reduction; Epn- Epenthesis, Fr-Fronting, St-Stopping, FCD-Final Consonant deletion, Pal-Palatalization, AF- affrication, DeA- Deaffrication, Das-Deaspiration, PrV-Prevocalic voicing, Mt-Metathesis

Acknowledgements

The authors extend their gratitude to Dr. Vijayalakshmi Basavaraj, Director, All India Institute of Speech and Hearing, Mysore, for granting permissin to conduct the study. They thank Mrs. Vasanthalakshmi, Lecturer in Biostatistics, for her statistical assistance. They also sincerely thank Mr. Shyam Kumar from ENFIN Technologies India Private Limited, Bangalore, for helping them with

software making. They wish to thank the subjects who participated in the study.

References Hodson, B. W. (1985). In J. Bernthal, & Bankson, N.

(1993). Articulation and phonological disorder. N.J Englewood Cliffs: Prentice Hall.

Hodson, B. W., & Paden, E. (1981). Phonological processes which characterize unintelligible and intelligible speech in early childhood. Journal of Speech and Hearing Disorders, 46, 369-373.

SLNo Subjects CR Epn AF St Pal Fr Mt DeA Correct production Idiosyncratic processes 1 Subject 1 7 - 6 3 1 - - - 1 3 2 Subject 2 4 - 2 6 2 - - - 2 4 3 Subject 3 6 - 3 - 2 - - - 3 6 4 Subject 4 6 - 6 1 2 - - - - 7 5 Subject 5 5 2 2 2 3 - - - 5 2 6 Subject 6 9 - 4 3 1 - - - 3 2

Computerized assessment of phonological processes-Malayalam

175

Hodson, B. W., & Paden, E. (1983). In J. R Lowe, (1994). Phonology: Assessment and intervention applications in speech pathology. Baltimore: Williams and Wilkins.

Hodson, B. W., & Paden, E.P. (1991). In Bernthal & Bankson. Articulation and phonological disorders. N.J Englewood Cliffs. Prentice Hall.

Hodson, B.W. (2003). In M.E. Gordon- Brannan & C.E. Weiss (2007). Clinical management of articulatory and Phonologic Disorders. Baltimore: Lippincott Williams & Wilkins.

Ingram, D., & Ingram, K.D. (2002). In W. O. Haynes & R. H. Pindzola (2008). Diagnosis and evaluation in Speech pathology (7th Eds). Boston: Pearson Education Inc.

Long, S., & Fey M. E., & Channell (2002). In W. O. Haynes & R. H. Pindzola (2008). Diagnosis and Evaluation in Speech pathology (7th Eds.). Boston: Pearson Education Inc.

Louko, R., & Edwards.(2001). In W. O. Haynes & R. H. Pindzola (2008). Diagnosis and evaluation in speech

pathology (7th Eds.). Boston: Pearson Education Inc.

Masterson, J. (1999). In W. O. Haynes & R. H. Pindzola (2008). Diagnosis and evaluation in speech pathology (7th Edn.). Boston: Pearson Education Inc.

Masterson, J., & Bernhardt, B. (2001). Computerized Articulation and Phonology Evaluations System. San Antonio. TX: The Psychological Corporation.

Masterson, J., & Long. (2004). In W. O. Haynes & R. H.Pindzola (2008). Diagnosis and evaluation in Speech pathology (7th Edn.). Boston: Pearson Education Inc.

Maya (1990) „An articulation test battery in Malayalam. Research at A.I.I.S.H, II, 179-180.

Ramadevi. S. (2006). Phonological Profile in Kannada, A study on hearing impaired. An unpublished doctoral dissertation submitted to the University of Mysore, Mysore.

Shriberg, L.D. (1986). In J. Bernthal & N. Bankson (1993). Articulation and phonological disorder. N.J Englewood Cliffs: Prentice Hall.

Weiner, F. F. (1986). In S. Ramadevi. (2006) Phonological profile in Kannada, A study on hearing impaired. An unpublished doctoral dissertation submitted to the University of Mysore, Mysore.


176

L

Spoken Language Proficiency (Discourse Level) in Bili ngual Children with Learning Disability

Mohana P.1 & Shyamala K.C. 2

Abstract

This study was mainly designed to investigate the language proficiency in bilingual children with Learning Disability (LD) and typically developing children. The study also aimed to identify and quantify the language impairment at discourse level in children with LD and typically developing children. It was compared across the group and language for two different tasks (picture description or story narration). The participants included a total of 60 school going children in the age range of 8 to 13 years. An attempt was made to investigate specifically, the aspects of discourse (propositional/non-propositional) that were affected in children with LD and typically developing children. The results indicated significant differences in the discourse of narrative productions in the LD compared with controls. The findings of this study support past literature, which calls for greater research in this area using stricter reliability measures.

Key words: learning disability, language proficiency, discourse aspects/propositional & non-propositional

anguage 1is like any other skill or aptitude; some people are proficient in languages, while others are better at math, science, or music. “Language

is presumed to have developed on a solid foundation establi shed during infancy and early childhood" (Nippold, 2006). Therefore, everyone has the potential to learn, but the fact is that some people are just more capable of learning language than others.

“Proficient” implies an advanced degree of

competence acquired through training. Language proficiency skills is the knowledge and abili ties which impact on the capacity of a given individual to communicate spontaneously, accurately, intelligibly, meaningfully and appropriately in a given language.

“Learning disability (LD) is a specific

language disabili ty that is neurobiological in origin. It is characterized by diff iculties with accurate and/or fluent word recognition and by poor spelling and decoding abili ties. These difficulties typically result from a deficit in the phonological component of language that is often unexpected in relation to other cognitive abili ties and the provision of effective classroom instruction” (Lyon, Shaywitz & Shaywitz, 2003). Children with LD have the greatest problems li ke organizing in an appropriate sequence of words (Clif ford, Reilly & Wulfeck, 1995; cited in Leonard & Kondrick, 1998). Paul and Smith (1993) reported diff iculty of narrative skil ls in LD children, they pointed that language difficulty may be due to deficit

1e-mail: [email protected]; 2Professor of Language Pathology, AIISH, Mysore, [email protected].

beyond language, difficulty in encoding, organizing, linking proposition and retrieval of words.

Chinn and Crossman (1995) described

dyslexics as having difficulty in “finding the right word͉ - They know that they have a good idea but cannot find a way to communicate it to others. They describe how young dyslexics tend to be disadvantaged because of misperceptions misjudgements and misreading within the social sphere, and that this can continue into adulthood. Investigators have found that pre-school impairment in language skills are associated with later problems in reading and spelling, and that children with LD have particular problems with complex language demands such as narratives or storytelling, lexical retrieval and recognition of melody patterns (Denckla & Rudel, 1976; Donahue, 1984).

Language can be viewed and analyzed on many levels, one of which is “Language in use͇ (Frattali & Grafman, in press) compared to production of sounds, words, or sentences in isolation, discourse production. Discourse production which is an integrative and context driven construct, is also a representative of the complex communication needed for daily life activities. School age children are expected to comprehend and produce a range of discourse types (Nelson, 1993; Scott, 1994). These different discourse types included during assessment practices present in children speaking different languages may reveal conditions under which language production problems arise.

There is considerable research effort towards organization of several languages. There are numerous reasons to believe that the use of language as a medium for acquiring knowledge is crucial to academic success.



Discourse proficiency in bilingual children with LD

177

Thus deficits in discourse may impact significantly on a child‟s academic achievement. Learning Disabled (LD) students, by definition, demonstrate primary problems in academic learning. Although the precise relationship between discourse and academic competencies is far from clear, a growing body of evidence suggests that LD students manifest problems in discourse. Conversation is one form of discourse in which deficits have been fairly well documented in LD children (e.g., Bryan, Donahue, Sturm & Pearl, 1981; Donahue, 1984). Another discourse form that poses problems for these students is the narrative.

In language competence, traditionally, the self assessment literature has used three distinct measures to index bil ingual language competence: (a) language proficiency, (b) language dominance, and (c) language preference. Adding to the confusion is that researchers have also at times used language preference, instead of language proficiency or dominance, as the domain of interest (e.g., Marian & Neisser, 2000).

Hence, the study attempts to identify and

quantify impairments in communication abilities of children with LD. The aim was to identify commonly shared discursive resources (shared patterns of talking). This study is formulated using Tamil which is a Dravidian language spoken mainly in the state of Tamil Nadu and English which is an Indo European language. It becomes interesting to study the differences in narrative skills in bil ingual Tamil-English children and to check whether such findings support or refute the earlier findings that bili nguals lack language proficiency at discourse level.

Despite the growing body of literature on LD,

there have been relatively few studies carried out in the Indian context. Recent research with regard to discourse in individuals with LD has confirmed deficits in the same. But, only few studies give information regarding the specific types of deficits in discourse of individuals with LD, especially with respect to differences in discourse deficits across monolingual and bilingual LD. The few studies reviewed revealed that impairment in discourse is one of the persisting, subtle and subclinical features in children with LD. Although there is much research in western countries in this area, limited literature is available in Indian context. An effort is made in these studies to analyze each feature of discourse and score them using perceptual rating scales. Thus in order to get a comprehensive picture of all the affected parameters at discourse level, this study was formulated.

The aim of the current study was to explore and

compare the language characteristics and deficits in

spoken language proficiency at discourse level in the children with LD whose mother tongue was Tamil and who had acquired English in formal education set up. Discourse skills were examined because the abili ty to initiate and maintain a conversation is essential for social communication and communication effectiveness. The study aimed to assess the discourse deficits in children with LD and to compare with narrative discourse of typically developing children across both the languages (Tamil-English) and various age groups (8 to 13 years) and across various tasks (picture description or story narration). This study was designed to specifically find which aspects of discourse (propositional/non-propositional) are affected in children with LD and typically developing children.

Method

Participants: The study was conducted on two groups of subjects, one experimental and one control group. The experimental group consisted of 30 dyslexics in the age range of 8 to 13 years. Age, gender and education matched 30 normal subjects served as the participants for control group. All the subjects included in the study were Tamil-English bil inguals. Apart from the mother tongue the other language to which the children were exposed to was English, which was learnt as a language as well as a medium of instruction in school.

All the children who participated in the study were right handed with normal hearing thresholds in both the ears and had average intelligence as verified on observation and parental/teacher interview and no formal tests however, were administered.

Those who had no known sensory deficits and neurological impairment were selected. The subjects used no other language in their daily communication other than Tamil and English. They had acquired English as their second language (L2-English for all the participants) both for academic and communicative purposes.

Table 1. Details of the participants of the study

Type of Population Age Range (in years)

Males

Females

Total

Normal participants 8-10 8 8 16 11-13 8 6 14

Persons with LD 8 – 13 19 11 30 Experimental / clinical group: A total of 30 children with LD were carefull y selected from specialized schools, or units within mainstream schools, that were organized to support children with the learning


178

disabilities. Children with b e h a v i o u r a l disorders, sensory deficits, cognitive and neurological impairment were excluded from the study.

Only children for whom an Educational

Psychologists‟ and Speech-Language Pathologists‟ (SLPs) assessment report was available were included in the LD group. In addition, LD children had, prior to this study received at least one year of special teaching support and had been attending their current school for at least two terms. Selection procedures targeted a relatively focused age range to ensure that similar curriculum levels had been met by all LD children. It was ensured that no undiagnosed dyslexics found their way into the clinical group.

Contr ol group: The child‟s age at the time of full assessment considered in this study varied considerably within these groups (from 8 to 13 years of age). Control group subjects consisted of 30 volunteers were selected from mainstream schools. They were screened for any speech and language, cognitive-linguistic and hearing impairment using assessment battery developed by the institution. Parental/guardian permission formed part of the criteria on which the control children were selected. Teachers were consulted to ensure that there was no reported evidence of learning dif ficulties or concerns about the acquisition of reading/writing skil ls. All control children were meeting school achievement levels based on their current and previous curriculum-based tests. Finally, control children were selected to mirror age levels and sex ratios of the LD groups.

Procedure: Testing environment was chosen with minimum distraction. Children were given the tasks individually in a quiet room. Testing took place on several different days over the period of 2-3 weeks. Each test session lasted no longer than 30 min. The narrative task and other tests were conducted solely by the examiner. Participants were selected by ethical procedures. The current study was carried out in different levels as discussed further.

Material: Initiall y both the experimental and control group was subjected to Language Proficiency Rating Scale (LPRS) (Vishnu, Deepa, Hema & Chengappa, 2010). The participants or the parents of the participants rated LPRS based on the proficiency scales, in both Tamil and English. Overall scores were calculated and converted into percentage score. Subjects were verified whether they score 90% to 100% which indicated higher language proficiency. Pursued by the judgment of the language proficiency of all the participants, further each participants were provided with a creative picture description (Western Aphasia Battery by Kertesz, 1982) task followed by

a story narration task. Duration of five minutes was given for picture description task where the children were asked to narrate gist of information about the picture. Warm up time of approximately 2-3 minutes was provided for each child before the commencement of the tasks. At the same time recording was done.

The testing was done first in Tamil and second

in English for both the tasks. While testing in Tamil, all conversations, instructions was given only in Tamil and similarly English was used exclusively for testing in English. All the narratives were videotaped for further transcription and analysis.

Discourse analysis procedure was used to assess the discourse abili ty in individuals with LD and normal speakers who were Tamil-English bilinguals. Various speech discourse parameters under the propositional and non-propositional aspects of conversation and picture description tasks were analyzed. Discourse Analysis Scale, (Hema, 2008) was used. Transcription procedure: The participants' spoken narratives were transcribed according to the conventions of the discourse analysis. From the recorded audio sample, transcription was done for picture description and story narration task individually using broad International Phonetic Alphabet, 2007. During transcription, initiation time, pause time, filled pauses, unfilled pauses and false starts etc were carefully noted for each episode. Scoring and statistical analysis: Statistical analysis was done using SPSS software 16 version. Test measures were selected to cover different aspects of language abili ty in Tamil-English bil ingual LD and normal individuals. Individual scores were calculated, Mann Whitney-U test and Wilcoxon‟s signed Rank test was applied for the sub-parameter of the following parameters to measure the significance of the value obtained. Nonparametric tests were used to find significant difference within LD group and control group across languages and tasks. Independent t-test was applied for the percentage scores of the following parameters to find the significant difference between the LD and normal speakers across two tasks (picture description/ story narration) and two domains (propositional/ non-propositional).

The investigator repeated the process of transcription of discourse sample i.e., the story narration and picture description sample of five children with LD and five normal subjects for verification of transcription, scoring, and reporting of the features. The findings were found to be correlating in both instances.


179

Parameters* Experimental group Control group Mean SD Mean SD

Epdprper 50.96 3.70 84.04 7.39 Epdnpper 31.67 8.34 70.33 9.28 Esprper 51.54 3.05 84.04 7.39 Esnprper 31.67 7.91 70.33 9.28 Tpdprper 49.23 4.62 88.91 2.97 Tpdnpper 30.00 12.03 74.33 10.73 Tsprper 48.97 4.56 90.00 2.69 Tsnprper 31.00 12.96 74.33 10.73


The performance of LD subjects as compared to typically developing children are presented and discussed in detail under various sections as follows.

Comparison between LD children and normal speakers for percentage scores: The raw scores obtained from each subsection of discourse analysis were converted into percentage score and was statistically analyzed to see the significant difference between the two groups.

Table 2. Mean and Standard deviation (SD) for percentage scores across subjects, languages & tasks

*(Epdprper- Engli sh picture description proportional percentage score, Epdnpper-Engli sh picture description non-proportional percentage score, Esprper- English story narration proportional percentage score, Esnprper- Engli sh story narration non-proportional percentage score, Tpdprper- Tamil picture description proportional percentage score, Tpdnpper- Tamil picture description non-proportional percentage score, Tsprper- Tamil story narration proportional percentage score, Tsnprper- story narration non-proportional percentage score).

The mean and standard deviation across children with LD and typically developing children across both the languages (Tamil and English) and across two aspects of discourse (propositional and non- propositional) using two different tasks are represented in Table 2. The values are shown as percentage scores and are graphically ill ustrated in the Figure 1.

The mean scores differed significantly from

each other and also with the LD group and typically developing children. However, there were variations in the scores obtained among the two tasks. The findings also indicated that the performance varied greatly across the two languages. Thus, the pattern of scoring across the tasks and languages is similar for children belonging to experimental group and control group.

The present study disclosed a trend in the performance of the LD children and typically developing children, in language proficiency. Although LD children did not perform on par with typically developing children, they were certainly better in propositional than non-propositional aspects. This result is in harmony with Laughton and Morris (1989) where the children with LD have lower values and discourse deficits compared to typically developing children.

Figure 1. The mean percentage score compared with languages, tasks and discourse aspects for LD and

normal. LD children and typically developing children

were compared using Wilcoxon signed rank test across the two languages and tasks based on their percentage scores. Significant difference was present between propositional and non propositional aspects of discourse across both the groups and languages.

In summary, the study discovered mixed

results where two groups demonstrated significant differences between languages and some for tasks which could be due to factors such as linguistic, social and environmental influences experienced by the child in their day to day life. However, the review of literature lacks studies examining whether the discourse skil ls of children with LD are similar to the discourse deficits of typically developing children and if these discourse skills differ from those of either picture description task or story narration task in Tamil or English. The same need to be studied and well documented in Indian languages.

Comparison within LD group and control group of normal children: In this section, propositional aspects li ke Failure to Structure Discourse, Communication


180

intent, Topic management, Information content, Information adequacy, Message Inaccuracy, Use of nonspecific vocabulary, Linguistic Non-fluency, Inappropriate Speech Style, Inappropriate Intonation Contour were compared within typically developing children as well as children with LD. Non- propositional aspects li ke delay in describing picture, Repair strategy, Revision behaviors and Gist of Information was also carried out within the groups.

Wilcoxon signed rank test was carried out for children with LD and typically developing children across language and task. Significant difference at p<0.05 level of significance in structuring discourse for both the tasks under same language, Tamil but not English was noted in all the parameters (propositional and non propositional). Parameters like Inappropriate Speech Style, Coherence, Inappropriate Intonation Contour and Delay in describing did not show significant difference among the subjects with LD. Comparison of propositional aspects and non- propositional aspects of discourse percentage scores across LD and typically developing children: LD Children and typically developing children were compared using Wilcoxon signed rank test across the two languages and tasks based on their percentage scores. Significant dif ference was present between propositional and non-propositional aspects of discourse across both the groups and languages. Children with identified learning difficulties are known to show impaired performance on oral narrative tasks compared to their peers with typically developing skills throughout their primary school years. More specifically, children with reading disabilit y use fewer words (Feagans, Garvey & Golinkoff, 1984) and include fewer of the original propositions when retelling a story compared to age-matched peers (Feagans, Garvey, & Golinkoff, 1984; Roth & Spekman, 1986). Moreover, in a cross-sectional study, Snyder (1978) found that oral narrative abilit y (as measured by the number of original propositions included in a story retell) accounted for more of the variance in reading comprehension in 11- to 14-year- old children with reading disabili ty than it did for the 8- to 11-year-old age group of children with reading disabili ty. These findings suggest that oral narrative diff iculties in children with reading disabili ty do not spontaneously improve over time, and stress the importance of investigating and remediating the oral narrative skills during the early school years.

Black and Logan (1995) found that these patterns in parent-child interaction, among others such as appropriate turn-taking, are linked to chil dren‟s conversational skills and peer acceptance. Chil dren‟s

conversational patterns were similar to those of their parents. Parents of rejected children were more li kely to make more demands and closed requests, take irrelevant turns, interrupt, and not leave time for a response after taking a turn.

Delgado, Guerrero, Goggin and Ellis, (1999) tested Spanish–English bil inguals and correlated self assessed proficiency in English and Spanish with performance on the Woodcock– Muñoz Language Survey (Woodcock & Muñoz-Sandoval, 1993). Delgado et al. (1999) found that participants assessed first- language (L1) skil ls more accurately than they did second-language (L2) skill s. Woodcock–Muñoz scores correlated with all self reported measures of L1 proficiency but with only self reported measures of L2 reading and writing (and not with L2 speaking and understanding). Similarly, Bahrick, Hall, Goggin, Bahrick and Berger, (1994) found that language dominance ratings correlated highly with performance on some tasks (e.g., category generation and vocabulary recognition) but correlated less with performance on other tasks (e.g., oral comprehension). Together, studies of domain-general self-assessment in bili nguals suggest that the relationship between self- reported and behavioural measures of language performance varies across languages and tasks (e.g., Bahrick et al., 1994; Delgado et al., 1999).

In summary, the study discovered mixed results where the two groups demonstrated significant differences between languages and some for tasks which could be due to factors such as linguistic, social and environmental influences experienced by the child in their day to day life. However, the review of literature markedly lacks studies examining whether the discourse skil ls of children with LD are similar to the discourse deficits of typically developing children and if these discourse skil ls differ from those of either picture description task or story narration task in Tamil or English. The same need to be studied and well documented in the Indian languages.

Conclusions

The current study was mainly designed to investigate the language proficiency in bil ingual children with LD and typically developing children. The study also aimed to identify and quantify the language impairment at discourse level in LD and typically developing children and also was compared across the group and language for two different tasks. The results revealed a significant difference between the experimental group (LD) and the control group (typically developing children), wherein the overall performance of the experimental group was poorer as compared to the control group.


181

Within the experimental group, significant difference was noticed between the languages with better scores in Tamil as compared to English. However, they did not show a significant difference among tasks. Within the control group, similar pattern was observed as that of the experimental group.

The results thus correlate with the language proficiency rating which was explored in Phase I. L1 proficiency was reported to be better than L2 proficiency. It was also found that self-reported reading proficiency was a more accurate predictor of first- language performance, and self-reported speaking proficiency that of second-language performance. Family based experiences, years spent in a L1 region may have contributed to L1 competence and proficiency while L2 was mainly school based and learnt later in a successive context, hence less proficient. It can be concluded that higher language proficiency reflects greater performance in spoken discourse among children with LD and typically developing children.

References

Bahrick, H. P., Hall, L. K., Goggin, J. P., Bahrick, L. E., & Berger, S. A. (1994). Fifty years of language maintenance and language dominance in bilingual Hispanic immigrants. Journal of Experimental Psychology: General, 123, 264–283.

Black, B., & Logan, A. (1995). Links between communication patterns in mother-child, father-child, and child-peer interactions and children's social status. Child Development, 66, 255-271.

Bryan, T., Donahue, M., Pearl, R., & Sturm, C. (1981). Learning disabled children's conversational skills: The 'TV Talk Show. Learning Disability Quarterly, 4, 250-259.

Chinn S. J., & Crossman M. (1995) Stress factors in the adolescent. In: T. R. Miles & V. P. Varma ( Eds.), Dyslexia and stress. London: Whurr.

Clifford, J., Reilly, J., & Wulfeck, B. (1995). Narratives from children with specific language impairment: An exploration in language and cognition. University of California, San Diego.

Delgado, P., Guerrero, G., Goggin, J. P., & Ellis, B. B. (1999). Self-assessment of linguistic skills by bilingual Hispanics. Hispanic Journal of Behavioral Sciences, 21, 31–46.

Denckla, M. B. & Rudel, R.G. (1976). Rapid automatized naming (R.A.N.): Dyslexia dif ferentiated from other learning disabilities. Neuropsychologia, 14, 471-479.

Donahue, M. L. (1984). Learning disabled children's

conversational competence: An attempt to activate the inactive listener. Applied Psycholinguistics, 5, 21-35.

Feagans, L., Garvey, G. J., & Golinkoff , R. (Eds.) (1984). The origins and growth of communication. Norwood, NJ: Ablex.

Frattali, C., & Grafman, J. (in press). Language and discourse deficits following prefrontal cortex damage. In L. L. LaPointe (Eds.), Aphasia and related neurogenic language disorders. New York: Thieme.

Hema, N. (2008). A study of discourse analysis in traumatic brain injury (TBI): Left hemisphere damage vs. Right hemisphere damage. Unpublished Masters Dissertation submitted to the University of Mysore, Mysore.

Kertesz, A. (1982). Western aphasia battery. New York: Grune and Stratton.

Laughton, J. & Morris, N. (1989). Story grammar knowledge of learning disability students. Learning Disabilities Research, 4(2), 87-95.

Leonard B. G & Kondrick, P. A. (1998). Psychological behaviorism's reading therapy program: Parents as reading therapists for their children's reading disability. Journal of Learning Disability, 31, 278-285.

Lyon, G. R., Shaywitz, S. E., & Shaywitz, B. A. (2003). A definition of dyslexia. Annals of Dyslexia, 53, 1-15.

Marian, V., & Neisser, U. (2000). Language-dependent recall of autobiographical memories. Journal of Experimental Psychology: General, 129, 361–368.

Nelson, N. (1993). Childhood Language Disorders in Context: Infancy through adolescence. MacMillan Publishing Co., U.S.A.

Nippold, M. (2006). Later language development: School-age children, adolescents, and young adults (Eds.), Austin, Texas: pro-ed.

Paul, R. , & Smith, R. (1993). Narrative skills in 4-year-olds with normal, impaired and late-developing language. Journal of Speech and Hearing Research, 36, 592-598.

Roth, F., & Spekman, N. (1986). Narrative discourse: spontaneously generate stories of learning –disabled and normally achieving students. Journal o f Speech and Hearing Disorders, 51, 8 –23.

Scott, C. M. (1994). A discourse continuum for school-age students. In G. P. Wallach & K. G. Butler (Eds.), Language learning disabilities in school-age children and adolescents: Some principles and applications (pp.219–252). New York: Macmillan.

Snyder, L. D. (1978). Communicative and cognitive abilities and disabilities in the sensorimotor period. Merril l Palmer Quarterly, 24, 161-180.

Vishnu, K., Deepa, M. S., Hema, N., & Chengappa, S. K. (2010). Language Proficiency Rating Scale. Department of Science and Technology. Unpublished Project. University of Mysore.

Woodcock, R.W., & Muñoz-Sandoval, F. (1993). Woodcock–Johnson language survey. Itasca, IL: Riverside.


182

English Language Test for Indian Children (ELTIC) Bhuvaneswari N.1 & Jayashree C. Shanbal 2

Abstract

Language development is a process that starts early in human life, when a person begins to acquire language by learning it as it is spoken and by imitations. Children with sensory impairments have delay in speech and language skills. Many assessment tools were developed in the past decades that help in providing the child with early intervention, due to which the language skills delaying for some extent can be improved. In particular these days English language has been used most of the time by the children. Therefore, this study was developed to assess the English language abilities in children with English as their second language. The aims of the study were, to develop English Language Test for the Indian children (ELTIC) and to study the developmental pattern for acquisition of language skills in children. Participants in the present study included 80 typically developing children in the age range of 4.0- 6.0 years, who were grouped into four as 4.0-4.6 years, 4.7-5.0 years, 5.1-5.6 years, and 5.7-6.0 years. The test was developed based on the Bankson Language Screening Test (BLST; Bankson, 1977) which was named as English Language test for Indian Children (ELTIC). It included three domains namely, Semantic knowledge, Morphological rules, and Syntactic rules. The results showed that there was a developmental trend observed in the performance of the children from younger to older groups i.e., 4.0-6.0 years across all domains of language mentioned.

Key words: ELTIC, semantic knowledge, morphological rules, syntactic rules

anguage is a 1socially shared code, or conventional system, that represents ideas through the use of arbitrary symbols and rules

that govern combinations of these symbols. Language development is a process that starts early in human life, when a person begins to acquire language by learning it as it is spoken and by imitations. Also the most important period of language development in children is the life between birth to three years, called as “critical period”. It was said to be important because the more the language exposure occurs during this period child will be able to learn faster.

Actually language is a complex combination of

several component rules systems and it is divided into: form, content and use. Form includes the linguistic elements that connect sounds and symbols with meaning i.e. phonology, rules that govern the internal organization of words i.e. morphology and rules that specify how words should be ordered to produce a variety of sentence types i.e. syntax. The content involves the language meaning, it maps knowledge about objects, events, and people, and relationship between them i.e. it included the rules governing semantics a subsystem of language. The use component of language encompasses rules that govern use of language in social contexts, so called pragmatics (Bloom & Lahey, 1978). The main functions of language relate to the speakers intention or goal, e.g., greeting, asking questions, giving information etc. When there is a problem in the integration of form,

1e-mail:[email protected]; 2Lecturer in Language Pathology, AIISH, Mysore, [email protected].

content and use; the language in the children will be disrupted i.e. often found in children with language disabilities. Children with some sensory impairment like hearing loss, delay in motor skills, poor cognitive abilities etc. cause delay in the development of speech and language skills of child compared to the typically developing children. Hence assessment plays an important role in finding the deficits in different disordered children.

Many assessment tools were developed in the

past decades that help in providing the child with early intervention, due to which the language skills delaying for some extent can be improved. Many investigators have developed and have standardized the tests like Comprehensive Receptive & Expressive Vocabulary Test (Wallace & Hammill, 1994); Clinical Evaluation of Language Fundamentals- third edition (Semel, Wiig & Secord, 1995) etc. were developed in western context. Also the language tests were developed in India like- A Screening Picture Vocabulary Test in Kannada- KPVT (Sreedevi, 1988); A Syntax Screening Test in Tamil, (Murthy, 1993); Linguistic Profile Test in Hindi Normative Data for children grade I to X (Monika Sharma, 1995). The available tests assess language skills of children in only a particular age range and were more restricted to few locations only. In the recent years scenario of language usage like English is quite high by children from various language background as the exposure to English is greater for children at home and in schools. For this reason there is a need for the development of test for assessment of English language skills in children. This would further

L

English Language Test for Indian Children (ELTIC)

183

serve as norm for assessment and identification of children with language difficulties.

Language codes human interaction within the

social environment. During the first two years of life, children learn many behaviors that they integrate and use to communicate. Communication gradually changes from use of personal and idiosyncratic forms within the family to use of conventional forms within society (Bates, 1993). The term language can be defined as a system that codes the social process in which communicative forms of exchange are learned; i.e., it is a system that is lend in the social context. It may be thought of either mental system of rules or as a form of social behavior. Those who believe that language is a mental system of rules recognize that communication is a social act (Hegde & Maul, 2006).

Researchers have attempted to explain language

and its components in different ways. This was also explained by various other disciplines who dominated the view point of language for the past two decades such as the behavioral, the psycholinguistic/syntactic, the semantic/cognitive, and the pragmatic (McCormick & Schiefelbusch, 1990; Bohannon & Bonvillian, 2001; Owens, 2005).

The behavioral approach to language

development was first presented by Skinner (1957). Language learning, according to this approach depends on environmental variables, which are mastered by imitation, practice, and selective reinforcement. In this process of language acquisition, parents and significant others are crucial because they model the appropriate utterances that children imitate and practice. By rewarding children‟s correct productions, parents shape the children‟s utterances until they are grammatically acceptable. In early 1960‟s criticism was leveled at the behavioral school for overemphasizing parental input, recent researchers have revised their thinking to acknowledge its importance in language development. Another contributions made by SLP‟S in the field of behaviorists is systematic training designs and their applications to non speaking individuals (McCormick & Schiefelbusch, 1984). These approaches were commonly used in providing therapy for children with language disorders.

In late 1960‟s, linguists, mainly Noam

Chomsky, hypothesized that human brain contains a mental plan to understand and generate sentences (Chomsky, 1957, 1965). According to Chomsky, a baby is born with an innate linguistic mechanism (called the language acquisition device (LAD) that is activated by exposure to linguistic input. The LAD has two parts: a set of rules or general principles for

forming sentences, and procedures for discovering how these principles are to be applied to the child‟s particular language. The development of language rests upon several major variables that interact with one another. In the process of language development firstly the innate characteristics of the child is important along with the child‟s I.Q, attention and others for the situations. Secondly the environment plays a major role in language development i.e. the more stimulating the environment, the better and more thoroughly the child develops the language skills (Roseberry-Mckibbin, 2007).

The term language impairment applies to a

heterogeneous group of developmental and/ or acquired disorders and/ or delays that are principally characterized by deficits in use of spoken or written language for comprehension and / or production purposes that may involve the form, content and/or function of language in any combination. The individuals with language impairments include disorders like specific language impairment (SLI), Mental Retardation (MR), Autism etc. Due to these reasons the language delay may take place in morphology, semantics or syntax. Eadie, Fey, Doughlas, and Parsons (2002) reported that children with SLI have considerable difficulty with morphological aspects of language. Difficulty with morphology is a hallmark of SLI. Children with SLI will frequently omit or make errors on bound morphemes such as regular and irregular past tense inflections, possessive morpheme‟s (That‟s the girl bicycle), articles a, an, the (e.g.: boy fell into pond), third person singular forms (e.g., she run into the classroom. He pet the kitty), contractible copula forms (e.g., Daddy not watching TV).

Assessment refers to the ongoing procedures

used to identify the child‟s unique strengths and needs as well as the family concerns, priorities, and resources regarding the child‟s development in order to plan intervention services (Crais, 1995). Assessment of children‟s language disorders consists of clinical activities that precede treatment and result in an accurate, thorough description of the child‟s existing and non existing communicative behaviors, the communication demands of a child‟s environment, associated factors that may affect language skills, the communication patterns of a child‟s family, and the strengths and limitations of the child and the family.

The assessment for children with the language disorders was performed by speech-language pathologists (SLP). SLP‟s are professionals who provide an assortment of services that relate to communicative disorders. The distinguishing role of


184

the SLP is to identify, assess, treat and prevent communication disorders in all modalities (including spoken, written, pictorial and manual) both receptively and expressively given by American Speech- Language- Hearing Association (ASHA, 2000).

Diagnostic screening tests: Diagnostic or screening tests allow the professional to identify the levels of competence of child for any language disorder. Table 1 shows list of diagnostic/screening tests to assess language abilities in children. Indian tests for screening language skill: These tests are mainly developed for Indian population to test their language abilities (morphology, semantics, syntax and pragmatics). They were developed to help out diagnosing the children in India with language disorders through which early intervention is possible. Table2 shows list of Indian tests useful in assessing language abilities in children.

There are various tests that were developed to

assess language and found useful to identify children with language impairments. Among which there are many tests developed earlier to screen children with language impairments i.e. to identify the language impairment cross the areas of language. For this reason many tests were developed for western population and for Indian population as well, that assess mainly the language components during their critical age period that is further helpful for early intervention. Therefore, the main need of the present study was to develop an English Language Test for Indian Children (ELTIC). But the present need of the study was not only to know their language deficits alone but to find out their English language skills in children who have exposure

to English at home and other environmental situations i.e. in the present scenario; English is being used majorly apart from cultures, places for communication. Children in India are exposed to English as the second language at school. Also this screening tool takes a short span of time to assess children with language problems who use only English language for communication.

The main aim of this study was to develop an English Language Test for the Indian children (ELTIC) and to study the developmental pattern for acquisition of language skills in children.

Method

Participants: The participant group for the present study included typically developing children in the age range 4.0- 6.0 years. These children were grouped as 4.0-4.6 years, 4.7-5.0 years, 5.1-5.6 years, and 5.7-6.0 years.

Each group consisted of twenty children. All the children studied in schools with English as the medium of instruction with Kannada as native language background. All the children had an exposure to English language at home at least 50% of the time. Test Development and administration Phase I: Development of test material: The test was developed based on the Bankson Language Screening Test (BLST; Bankson, 1977). The original BLST contains five domains which included Semantic knowledge, Morphological rules, Syntactic rules, Visual perception and Auditory perception.

Table 1. List of diagnostic/screening tests to assess language abilities in children Sl. No.

Name of the screening test Author and year Age range Test items and material

1. The Northwestern Syntax Screening Test

Lee (1971) 3 years – 7year 11 months

Receptive and expressive use of syntax

2. The Fluharty Preschool Speech and Language

Screening Test

Fluharty (1974) and Simmons (1988)

3years -5years Vocabulary, articulation, receptive language, expressive language

Table 2. List of Indian tests useful in assessing language abilities in children Sl. No. Name of the test Author and year Age range Test items

1. Linguistic Profile Test (LPT) Karanth (1980) > 6years Assess phonology, semantics and syntax

2. Linguistic Profile Test in Malayalam, normative data for children in grades

I to X

Asha (1997) 6years-15years

Assess phonology, semantics, and syntax

3. Linguistic Profile Test in Telugu, normative data for children in grades

I to X

Suhasini (1997) 6years-15 years

Assess phonology, semantics, and syntax


185

For the present study, only three domains from BLST were considered in order to specifically develop a language screening tool in young children. The domains included Semantic knowledge, Morphological rules, and Syntactic rules. These three domains are considered for the study because they are the earlier components of language development in children. The literature shows that active learning begins early in children development (Sokolov & Snow, 1994). Domain I: Semantic knowledge

Semantics is the component of language concerned with meaning. Meaning can be conveyed through language at the word, sentence, and discourse levels. The familiar sense of meaning is lexical meaning. The preschool period is one of the rapid lexical growth (GolinKoff, Mervis, & Hirsh-Passek, 1994). The first domain of the test material is semantic knowledge; there are eight sub domains in it namely: body parts, nouns, verbs, categories, functions, prepositions, colors/quantity, and opposites. In each of the sub domains there are nine items.

Among the eight sub domains, seven sub

domains have both reception and expression skills of the items in it and the last sub domain have only expression skill. The scores for each sub domains are mentioned in the test material. Among the eight sub domains, seven sub domains have both reception and expression skills of the items in it and the last sub domain have only expression skill. The scores for each sub domains are mentioned in the test material. The maximum score for reception was 63 and maximum score for expression was 72; and the total for this domain was 135. Domain II: Morphological rules :

Several morphological structures emerge during the early school –age years. Gerunds, which are verbs to which –ing has been added to produce a form that fulfills a noun function (e.g.: to fish becomes fishing) (Owens, 2005).

Derivational morphemes first appear in the late preschool years with the adjectival comparative –er (bigger) emerging between 4 and 5 years of age and –est (biggest) emerging between 5 and 6 years (Norris, 1998). The second domain of the test material is morphological rules, it has three sub domains namely: pronouns, verb tenses, & plurals/ comparatives/ superlatives; in which each sub domain has nine items as above. In this sub domain only expression skill was only performed. The maximum score for this domain was 21.

Domain III: Syntactic rules The studies reported that overall sequence of

development of syntax is similar for children with mild mental retardation and those developing typically; however the rate of development is also slower among those with mental retardation. Both sentence and complexity increase with development. The third domain of the test material is syntactic rules; it has two sub domains namely: subject-verb agreement/negation and sentence repetition/ judgment of correctness. These sub domains have nine items where the expression skill was only performed. The maximum score of this domain is 18. Phase II: Test administration: Before the final test administration a pilot study was done. It was done for a group of eight participants, two participants in the age range of 4.0 to 4.6 years, 4.7 to 5.0 years, 5.1 to 5.6 years, and 5.7 to 6.0 years. During this phase of testing few items were answered wrongly by children because of the confusion for pictures presented in the stimulus book. Therefore the items of pictures said wrongly were replaced with the new pictures in the stimulus book.

Therefore, the final test developed after the

changes made, which was called as English language test for Indian children (ELTIC) included three domains and thirteen sub domains as described above in phase I. The examiner provided appropriate instructions to the child as mentioned in the manual, along with the picture stimuli for each sub domains. The approximate time needed to complete the testing was 30-35 minutes. Adequate time was provided for each child to give their response. Also a scoring sheet was provided for scoring the responses of the child. Phase III: Scoring and analysis: During the testing the correct responses of the child were marked on a scoring sheet under each sub domains of the test material that was developed. The responses were given scoring „1‟ for the correct response which was said in English and „0.5‟ for the partially correct response in which the response was told in English but nearly to the target response and „0‟ was given as a score when child gives the incorrect response instead of the target response. The scores were calculated and analyzed using appropriate statistical analysis i.e. Multivariate Analysis of Variance (MANOVA) has been administered for the data to get the mean and standard deviations of all the domains across all the age groups. To explain the significance between each domain and between the age groups post Hoc Duncan‟s test was performed.


186

0

20

40

60

80

100

120

Me

an

sco

res

Age range (in years)


The results have been presented and discussed under the several sections. I. Comparison of overall performance across all domains and across age groups: The data was analyzed across age groups for all the domains. Mean and standard deviation (SD) scores were computed across age groups (see Table3). Analysis of results revealed that the mean scores across age groups increased from group 1 i.e. from 4.0-4.6 years (Mean=79.10, SD=16.54) compared to group 4 i.e. from 5.7 - 6.0 years (Mean=113.12, SD=11.86). The

Figure 1. Mean scores across age groups for all domains.

group 1 i.e. 4.0- 4.6years performed poorer compared to group 4 i.e. 5.7-6.0 years. This indicates that the groups followed a developmental sequence in the acquisition of language skills from 4.0 - 4.6 years, 4.7-5.0 years, 5.1-5.6 years and 5.7-6.0 years (Figure1).

The results showed that there was a developmental pattern observed in the performance of children from younger to older age groups i.e., from 4.0-6.0 years across all domains of language which included semantics, morphological and syntactic rules. Owens (2005) reported that the children‟s rate of language development shows variation due to the differences in intellect, learning style and socioeconomic factors and also reported that the later stages of language development is based on the earlier ones; language growth is a gradual process. II. Comparison of performance of semantic knowledge domain across age groups: Analysis of results for semantic knowledge reception revealed that there was an increase in the mean scores from group 1 (Mean=45.07, SD=5.04) compared to group 4 (Mean=54.27, SD=4.28) (Table 4). The group 1 performance was poorer compared to group 4. The results also revealed that there was a developmental

trend in the performance of the skill as a function of increasing age (Figure 2).

In the present study results revealed that there was a developmental pattern in the acquisition of semantics, morphology and syntactic rules in children from 4.0 years to 6.0 years. The findings of the study support Rukmini (1994) who developed the Malayalam Language Test (MLT) and assessed children from 4-7 years age. She studied the performance of children on phonology, semantics and syntactic skills. She also reported that there was a pattern of acquisition of semantic and syntactic skills in children across age groups and semantic domain was better than the syntactic domain which indicated that semantics in children are acquired before syntactic skills. The analysis of the results for all the sub domains in semantic knowledge- reception and expression skills revealed that the mean scores was found to be increasing from group 1 to group 4.

This has indicated a developmental sequence of

skills as a function of increase in age groups. Among all the sub domains the group 1 has performed poorer comparatively with other age groups. This indicates that the development of skills increased with increase in age.

Lahey and Edwards (1999) studied typically

developing children and children with SLI in the age range of 4.3 to 9.7 years. They studied the pattern of errors in naming pictures of common objects. They explained that typically, as language skills expand, children learn new words with abstract meanings. This could be the reason attributed to lexical development in typically developing children. The findings of the present study revealed that for semantics, a developmental trend was observed from 4.0-6.0 years.

III. Comparison of performance of morphological rules across age groups: The analysis of the results of for the overall performance of the morphological rules across age groups showed a developmental pattern the skills from group 1 to group 4 (see Figure 2). In the group 1 the morphological skills performance has showed a lower scores, but there is a sudden rise in the morphological skills in the group 2, which was followed by other age groups.

Table 3. Overall Mean and Standard Deviation (SD)

scores across age groups (N=20)

Age range Mean SD 4.0-4.6years (Group1) 79.10 16.54 4.7-5.0years (Group2) 92.62 22.39 5.1-5.6years (Group3) 109.97 16.82 5.7-6.0years (Group4) 113.12 11.86


187

Table 4. Mean and SD for all the domains across age groups Domain Max.

score Age groups

4.0-4.6years (Group 1)




Mean SD Mean SD Mean SD Mean SD Semantic

knowledge reception

63 45.07 5.04 50.17 7.19 53.95 5.43 54.27 4.28

Semantic knowledge expression

72 29.65 11.72 35.25 12.80 43.20 8.99 45.42 6.70

Morphological rules

27 1.52 1.48 2.92 2.45 6.77 3.16 6.07 2.32

Syntactic rules 18 2.85 1.70 4.27 1.98 6.05 2.01 7.35 1.94

Age Range (yrs.)

5.7-6.05.1-5.64.7-5.04.0-4.6

100

80

60

40

20

0

Domains

SK-C

SK-E

MR

SR

Figure 2. Average scores (%) across age groups for all domains.

Note: SK-C= semantic knowledge reception, SK-E= semantic knowledge expression, MR= morphological rules, SR=syntactic rules

Among the sub domains pronouns expression (PRNE) has showed poor scores compared to other sub domains scores. But there is an increase in score pattern of mean even in the sub domains pronouns expression (PRNE). And the other sub domains verb tenses (VTE) and Plurals/comparatives/superlatives expression (PCSE) performed better even in group1 compared to pronouns (PRNE).

Rescorla and Roberts (2002) reported that in typically developing children morphological forms are evident during the second and third years of life in English speaking children. The present study findings supported above study that development of morphological rules start developing in the early years of life, except for the development of pronouns

(PRNE) that has started developing from age of 4.6-5.0 years. IV. Comparison of performance of Syntactic Rules across age groups: The analysis of the results for the overall performance of the syntactic rules across age groups showed a developmental pattern of the skills from group 1 to group 4. Among the 2 sub domains sentence repetition/ judgment of correctness (SRJCE) showed better scores across age groups compared to the sub domains subject-verb agreement/ negation (SVANE). The present study findings supported the findings of Rescorla and Roberts (2002), that the most common measures of syntactic development is the increase in mean length of utterance (MLU) in morphemes.


188

Table 5. Mean and standard deviation scores for all the sub domains across age groups

Note: BPR= body parts reception, BPE= body parts expression, NR= nouns reception, NE= nouns expression, VR= verbs reception, VE= verbs expression, CR= categories reception, CE=categories expression, FR= functions reception, FE= functions expression, PR= prepositions reception, PE= prepositions expression, CQR= colors/quantity reception, CQE= colors/quantity expression, OPE= opposites expression, PRNE= pronouns expression, VTE= verb tenses expressions, PCSE= plurals/comparatives/superlatives expression, SVANE= subject-verb agreement/ negation expression, and SRJCE= sentence repetition/judgment of correctness.

Conclusions

In sum, the analysis of results mean and standard (SD) scores revealed a significant difference across age groups and also cross domains. The overall performance of the participants across age groups increased from group 1(4.0-4.6years) to group 4 (5.7- 6.0 years) across all the domains. Thus ELTIC can be used as an assessment tool to assess the English language abilities of children across the domains semantic knowledge, morphological rules and syntactic rules. The normative data obtained reveals that the performance of the children increases as a function of age.

The present test ELTIC aids in assessing and identifying the language deficits in the areas of semantics, morphology and syntax in children between 4 - 6 years. It can be used as a quick screening tool in assessing language deficits. It can be easily used with children who know English language with any native language background. This further helps to make a

base line for child‟s language abilities, and helps in planning therapy for the child.

References

American Speech- Language- Hearing Association (ASHA) (2000a). Council on Professional Standards. Background information for the standards and implementation for the certificate of clinical competence in speech-language pathology. (Effective date: January 1, 2005).

Asha, M. M. (1997). Linguistic Profile Test in Malayalam, normative data for children in grades I to X. Unpublished Master‟s dissertation submitted to the University of Mysore, Mysore.

Bankson, N. (1977). Bankson Language Screening Test. Baltimore: University Park Press.

Bates, E. (1993). Commentary: Comprehension and production in early language development. Monographs of the society for research in child development, 58, 222-242.

Bloom, L., & Lahey, M. (1978). Language development and language disorders. New York: Macmillan.

Domain Sub domains 4.0-4.6year (Group 1)




Mean S.D Mean S.D Mean S.D Mean S.D Semantic knowledge

BPR 3.95 1.79 5.70 1.89 6.15 1.56 5.65 1.34 BPE 3.52 1.75 4.30 2.19 4.60 1.38 4.45 1.02 NR 6.95 1.43 7.30 1.30 8.40 0.88 8.55 0.68 NE 3.55 1.74 4.22 2.05 5.42 1.57 5.60 1.34 VR 7.45 1.23 8.15 0.81 7.95 1.14 7.90 1.16 VE 4.07 7.46 3.35 2.32 4.85 1.34 5.35 1.13 CR 7.15 1.30 7.70 1.08 7.90 0.71 7.97 1.10 CE 5.45 1.17 6.67 0.74 6.87 1.35 7.00 1.03 FR 6.87 0.91 6.97 1.26 7.45 1.27 7.15 0.93 FE 4.25 1.42 4.87 1.66 5.85 1.53 6.30 0.86 PR 6.50 2.16 7.25 0.91 7.90 1.11 8.50 0.68 PE 2.27 1.31 3.45 2.16 5.00 2.14 5.30 1.83

CQR 6.20 2.64 7.10 2.77 8.20 1.90 8.55 1.14

CQE 4.62 2.63 5.52 2.45 6.47 2.03 7.27 1.68 OPE 1.90 1.20 2.85 1.18 4.12 1.53 4.15 1.59

Morphological rules PRNE 0.00 0.00 0.10 0.30 0.55 1.05 0.30 0.57 VTE 0.47 0.71 0.97 1.63 3.27 1.83 2.67 1.98 PCSE 1.05 1.05 1.85 1.22 2.95 1.14 3.10 0.77

Syntactic Rules SVANE 1.32 0.74 1.97 1.09 2.70 1.01 3.20 1.17 SRJCE 1.52 1.31 2.30 1.15 3.35 1.27 4.15 1.31


189

Bohannon, J., & Bonvillian, J. (2001). Theoretical approaches to language acquisition. In J. Berko Gleason (Eds.), The development of language (5th ed.). Boston: Allyn & Bacon.

Chomsky, N. (1957). Syntactic structures. The Hague: Mouton.

Chomsky, N. (1965). Aspects of the theory of syntax. Cambridge, MA: MIT press.

Crais, E. R. (1995). Expanding the repertoire of tools and techniques for assessing the communication skills of infants and toddlers. American Journal of Speech- language Pathology, 4, 47- 59.

Eadie, P. A., Fey, M. E., Doughlas, J. M., & Parsons, C. L. (2002). Profiles of grammatical morphology and sentence imitation in children with specific language impairment and Down syndrome. Journal of Speech, Language, and Hearing Research, 45, 720-732.

Fluharty, N. B. (1974). The design and standardization of a speech and language screening test for use with preschool children. Journal of Speech and Hearing Disorders, 39, 75-84.

GolinKoff, R. M., Mervis, C. B., & Hirsh-Pasek, K (1994). Early object labels: the case for a developmental principles framework. Journal of Child Language, 21, 125-155.

Hegde, M. N., & Maul, C. A. (2006). Language disorders in children: An evidence based approach to assessment and treatment. USA.

Karanth, P. (1980). Linguistic profile test in Kannada. Journal of All India Institute of Speech and Hearing, 14, 121-126.

Lahey, M., & Edwards, J. (1999). Naming errors of children with specific language impairment. Journal of Speech, Language, and Hearing Research, 42, 195- 205.

Lee, L. (1971). Northwestern syntax screening test. Northwestern University Press, Evanston.

McCormick, L., & Schiefelbusch, R.L. (1984). Early language intervention. Columbus, OH: Merrill/Macmillan.

McCormick, L., & Schiefelbusch, R.L. (1990). Early language intervention (2nd ed.). Columbus, OH: Merrill/Macmillan.

Monika Sharma (1995). Linguistic profile test (LPT) (Hindi) Normative data for children grade I to X. Unpublished Master‟s dissertation submitted to the University of Mysore, Mysore.

Murthy, S. K. (1993). Syntax screening test in Tamil. Unpublished Master‟s dissertation submitted to the University of Mysore, Mysore.

Norris, J. A. (1998). Early sentence transformations and the development of complex syntactic structures. In W.O. Haynes and B. B. Shulman (Eds.), Communication development: Foundations, processes, and clinical applications (pp. 263-310). Baltimore: Williams & Wilkins.

Owens, R. E. (2005). Language development: An introduction (6 th ed.). Boston: Allyn & Bacon.

Owens, R. E., Jr. (2005). Language development: An introduction (6th ed.). Columbus, OH: Merrill/ Macmillan.

Rescorla, L., & Roberts, J. (2002). Nominal versus verbal morpheme use in late talkers at ages 3 and 4. Journal of Speech, Language, and Hearing Research, 45, 1219- 1231.

Roseberry-Mckibbin., C., (2007). Language disorders in children: A multicultural and case perspective. Pearson Education, Inc. USA.

Rukmini, A. D., (1994), Malayalam Language Test (MLT), Unpublished Master‟s dissertation submitted to the University of Mysore, Mysore.

Semel, E., Wiig, E. H., & Secord, W. A. (1995). Clinical evaluation of language fundamentals (3rd ed.). San Antinio, TX: The Psychological Corporation.

Simmons, J. O. (1988). Fluharty Preschool Speech and Language Screening Test: Analysis of construct validity. Journal of speech and Hearing Disorders, 53, 168-174.

Skinner, B. F. (1957). Verbal behavior. New York: Appleton-Century-Crofts.

Sokolov, J. L., & Snow, C. E. (1994). The changing role of negative evidence in theories of language development. In C. Gallaway and B. J. Richards (Eds.), Input and interaction in language acquisition (pp. 38-55). Cambridge, UK: Cambridge University Press.

Sreedevi, N. (1988). KPVT- A screening picture vocabulary test in kannada. Unpublished Master‟s dissertation submitted to the University of Mysore, Mysore.

Sudha, K. M. (1993). A syntax screening test in Tamil. Unpublished Master‟s dissertation submitted to the University of Mysore, Mysore.

Suhasini, G. (1997). Linguistic profile test (LPT) Telugu- Normative data for children in grades I to X. Unpublished Master‟s dissertation submitted to the University of Mysore, Mysore.

Wallace, G., & Hammill, D. D. (1994). Comprehensive receptive and expressive vocabulary test. Austin, TX: Pro-Ed.


190

Gender Difference in Nature of Disfluencies in Children with Stuttering Nisha Sudhi 1 & Y. V. Geetha 2

Abstract

Gender difference in stuttering is a much debated issue. A lot of studies have been done on this, mostly in the western countries. Stuttering is not only reported to be less common in female compared to male children, with a 1: 4 ratio, but its onset, development and recovery characteristics are more in favor of females. The present study is aimed to explore the similarities and differences in nature of disfluencies, if any, in male and female children with stuttering (CWS), across the age groups in the Indian context. 24 male and 24 female CWS registered at the All India Institute of Speech and Hearing were followed up. An analysis of the nature of disfluencies in terms of age of onset, nature and type of onset, time since onset, type of stuttering, nature and duration of treatment, associated problems, etiology of stuttering, and recovery patterns was done. Across gender significant differences have been obtained in most of the parameters under study. The study confirms many of the earlier findings by other authors with regard to gender differences in stuttering. Female CWS are different in the nature of disfluencies and recovery characteristics of stuttering compared to male CWS. Key words: stuttering, nature of disfluencies, gender difference

tuttering, 1in its most basic sense is a disruption/break in the forward flow of speech and a term that is most commonly used/

recognized by most people. This disorder manifests itself predominantly in childhood, most often within 2-6 years of age. Hence, it has often been described as a disorder of childhood. Literature also depicts stuttering to be a disorder of males. There have been investigations carried out in various aspects of stuttering to arrive at such a conclusion. However, they reveal divided opinions and findings. The earliest investigation into the gender difference in stuttering was from the early 1890‟s. Investigations into this ratio are being carried out since then. Van Borsel, Moeyart, Mostaert, Rosseel, VanLoo and Van Renterghem (2006), in agreement with past studies conducted, support that stuttering prevalence is higher in males than females and also reported the tendency for stuttering prevalence to decrease with increasing age. The difference in age of onset of stuttering (if any) across male and female children with stuttering (CWS) is an aspect of prime interest. Yairi and Ambrose (1992) reported the onset of stuttering in males to be 40.56 months and in females 34.21 months, with a 5 month difference in mean age at onset between males and females. However, Kent (1983) had discussed the fact that increased occurrence of stuttering in males is one of the few consistencies about the disorder and it appears that stuttering behavior begins with approximately equal frequency in young boys and girls. However, females are more likely to recover from stuttering during pre-school years than are males.


The research into the nature of stuttering in boys and girls reveals significant findings too. According to Yairi and Ambrose (1992), out of 87 children (59 boys and 28 girls), 44% (26 boys and 12 girls) had a sudden onset and 56% (33 boys and 16 girls) had gradual onset. Also, out of the 87 participants, 60 children were reported as having mild stuttering, 14 as moderate stuttering and 10 children (6 males and 4 females) were rated as having severe stuttering at onset. Another significant finding was that all the 6 male and 2 out of the 4 females had a sudden onset of stuttering. This study dictates a positive relationship between severity of stuttering and sudden onset. However, Yairi and Ambrose (2005), in a recent study reported that patterns of onset age are similar for both males and females.

Regarding the type of stuttering, clinically, the

number of disfluencies, especially of certain types has been regarded as the most important index of stuttering severity. They include SLD (Stuttering like Disfluencies) and OD (Other Disfluencies) as stated by Young (1984) and Yairi and Ambrose (1992). Accordingly, sound repetitions, single syllable word repetitions, syllable repetition, prolongations and blocks are considered as SLDs. Multi syllabic word repetitions, phrase repetitions, interjections and revisions are classified as Other Disfluencies (ODs). This classification helps to determine whether a child can be classified as a CWS or not. In a very recent study, exploring the difference between the gender in relation to type of stuttering, Anjana and Savithri (2007) analyzed the speech sample of 10 boys and 10 girls in the age range of 5.1-6 years and found that majority of the children had almost all the disfluency types. The most prominent disfluency type was sound repetitions. Also, significant gender differences were obtained, with boys showing

S


Gender difference in CWS

191

significantly higher percentage of SLDs than the girls. This also shows that boys are at a greater risk for stuttering than girls.

One consistent finding in the literature on

stuttering is that a small but significant percentage of children who stutter exhibit concomitant speech/language disorders in addition to their stuttering. Making a comparison across gender, there have been reports that male CWS have more associated problems than female CWS. Kolk and Postma (1997) suggested that CWS are more prone to phonological encoding errors because they are continuously slower at this task. At young ages, male children are more prone to develop phonological deficits than female children. According to Yairi and Ambrose (1992), the later age of stuttering onset for boys may reflect a slower language/phonological development. There has been contradictory evidence too. Ryan (1992) examined the potential performance difference between 20 male and female preschool children with and without stuttering. There were no differences between the two groups on articulation proficiency, although 25% of the CWS group (all boys) later required articulation treatment.

The etiology of stuttering is one of the most

investigated aspects. Various causes have been proposed to account for stuttering in boys and girls and more stuttering in boys. A difference in the reaction and responsiveness with favoritism exhibited on male children has been implicated widely in earlier literature. Silverman and Van Opens (1980) reported that boys are exposed to greater pressure to produce fluent speech than are girls. Current research considers females to be less susceptible to stuttering than male CWS or that they have equal chances of inheriting the disorder, but that female CWS recover faster, while male CWS tend to persist (Kidd, Kidd & Records, 1978; Ambrose, Yairi & Cox, 1993. Differences in language ability, reaction time, as well as the more recent differences in neuro-anatomical and functional changes of the brain have been implicated. Females have been considered to have an increased bilateral speech and language representation compared to males. The males are more likely to have a strong left hemisphere lateralization for speech and language.

There is a lot of debate about the onset, nature,

development, type, associated problems, and cause of stuttering in male and female CWS. Valuable opinion is available in scattered texts but these available information need to be compiled and a comprehensive comparison is necessary to give a better picture of the difference in the nature of disfluencies between the male and female CWS. This in turn will help in understanding their problems in a better way, finding out the prognosis, help in early

intervention and serve better in treating the male and female individuals with stuttering. The outlook towards the female CWS, their characteristics and needs can be understood better. In addition, most of the studies regarding gender and stuttering have been conducted in the western countries and such intensive studies have not been conducted in India. Further investigation into the cause of the condition will also be possible. With this need the present study was planned.

This study aimed to explore the difference in

nature of disfluencies if any, in male and female CWS with regard to the age of onset, nature, development of the problem, type of disfluencies, time since onset of stuttering, etiology, associated problems, nature and duration of treatment. Also, to study the pattern of recovery/relapses and severity levels with regard to male and female CWS.

Method

Subjects: Participants for the study included Kannada and Telugu speaking, male and female CWS in the age range of 3-6 years, registered at the All India Institute of Speech and Hearing, Mysore. 27 male and 24 female CWS were considered as participants in the present study. Materials: The following materials were used in the study: Stuttering severity instrument -SSI-3 (Riley, 1994), Language Assessment Checklist for children (Swapna, Prema & Geetha, 2010), Pictures for description from the fluency test (Nagapoornima, 1990; Yamini, 1990 & Rajendraswamy, 1991), Questionnaire, Set of questions for conversation sample, topic for monologue, A sample for writing (spontaneous and copying); Words and numbers for dictation, Video recorder (Digital video camera recorder-Sony model : HDR-TG-1E) and SPSS-16.

Procedure: The data was collected in two phases. Phase 1: In Phase I case histories of all children registered at the All India Institute of Speech and Hearing (AIISH) with the complaint of stuttering were reviewed. A total of 84 case files of children diagnosed by a speech-language pathologist as children with stuttering, and who were within 3-6 years of age were selected. Of these, 51 were of male and 33 of female CWS. Phase 2: In Phase 2, the selected CWS with complete address and/or telephone numbers were followed up through telephone/correspondence. A total of 80 CWS were followed up of which 51 children (27 male CWS and 24 female CWS) reported for assessment. Of these, 24 male and 24 female CWS who reported for the follow up were considered for the study. 3 male CWS were not considered for the study due to lack of adequate samples. Follow up included 10 CWS in mild


192

severity level, 10 in moderate and 4 in severe stuttering for both male and female CWS. In male CWS, available case file details of the remaining 6 children are also presented. The period of follow up ranged from 10 days to 2.7 years. On follow up, the questionnaire was administered to the parents/caregiver through an interview. A minimum of 150-200 word speech samples was video recorded during 30-45 minutes of interaction between the child, the parent and the investigator. Care was taken to ensure that the sample was no less than 5-minutes duration of the child‟s tal�ing.

Data analysis: The recorded samples were transcribed verbatim. The samples obtained from each participant were analyzed for the frequency of total disfluencies per 100 words. Within this, the mean frequency of Stuttering Like Disfluencies (SLDs) (i.e, sound repetitions, single syllable word repetitions, syllable repetition, prolongations and blocks) and Other Disfluencies (ODs) (multisyllabic word repetitions, phrase repetitions, interjections and revisions) per 100 words were calculated. SSI-3 was administered to arrive at an overall score and severity of stuttering. The following percentages were computed: Percentage of total disfluencies/total words, Percentage of SLDs/total disfluencies, Percentage of ODs/total disfluencies. The data was tabulated and statistically analyzed using SPSS package to answer the research objectives.


The purpose of the present study was to explore the gender difference, if any, in male and female CWS with respect to parameters such as age of onset, nature and development of stuttering, type of stuttering, causative factors, associated problems, time since onset (TSO) and nature and duration of treatment taken. It also aimed to compare the recovery across male and female CWS (as a group and within each severity level) using SSI-3 severity ratings. Attempts were made to look into the percentage of SLDs vs. ODs. The results have been described under each category.

Age and onset across gender and severity: The data collected were grouped into 3 age groups, < 2; 2 - 4 and 4 - 6 years for the analysis of age of onset information. The results are shown in Table 1.

The age of onset of stuttering has been reported in literature to be within 2-6 years of age. The present study also obtained similar results. Most of the studies on onset of stuttering across gender suggest that female CWS have an earlier age of onset than male CWS (Yairi & Ambrose, 1992; Mansson, 2000). However, opposing results have been reported by Andrews, Craig, Feyer, Hoddinott, Howie and Neilson (1983) wherein, the authors reported the age

of onset to be same for both genders. Data obtained from the present study supports their study results. 8.3% of female CWS had onset before 2 years compared to 6.6% of male CWS. Almost equal numbers of male and female CWS had onsets of stuttering between <2, 2-4 and 4-6 years. However, smaller intervals of age could have given a better insight. Nature and development of stuttering across gender and severity: Nature of onset: Figure 1 shows the nature of onset across the gender and different severity groups. In the present study 45% of CWS had a sudden onset of stuttering. With regard to gender, around 50% of both male and female CWS exhibited a gradual onset. Earlier literature reported that stuttering in children had a gradual onset 70% of the time and a sudden onset only 30% of the time. The present data obtained is in agreement with Yairi and Ambrose, (2003) that a significant number of preschool children exhibit sudden onset of stuttering. Analyzing the data in the present study, a correlation is also observed between sudden onsets and greater severity levels of stuttering.

Analyzing the data in the present study, a correlation is also observed between sudden onsets and greater severity levels of stuttering. For both male and female CWS, sudden onsets were associated with moderate and severe stuttering. This result supports the findings of Yairi, Ambrose and Nierman (1993) who reported that a significant number of preschool children exhibit a sudden onset of moderate to severe stuttering. Comparing the results across gender, the present data revealed that for male CWS, severe stuttering was associated with a more sudden onset (60%) than moderate stuttering (50%) whereas for female CWS, there was no such difference. Current status: The current status of the condition was classified as progressive, static, regressive or fluctuating. Table 2 provides data regarding the current status data for different severity groups in male and female CWS. The present data indicate that among both male and female CWS more number of children were found to have a decline/regression in their disfluencies (Male CWS: 41.6%, female CWS: 54.16%). Several reports in the literature too indicate that in childhood the remission rates of stuttering are high (Mansson, 2000; Yairi & Ambrose, 2005).

The current status of the condition is a good indication of the recovery from stuttering. When a regressive/fluctuating condition of stuttering is noted, it is a better predictor of recovery than a static/progressive condition. This can be accounted to the fact that stuttering in children, in the initial stages, is highly variable and the child often fluctuates


196

Table 5. Types of stuttering across gender

(Bloodstein & Smith, 1954; Goldman, 1967) indicated that male CWS were subjected to more environmental pressure. However, the present study does not support these findings. The current data indicated that environmental factors were the least implicated in male CWS (10%) and in female CWS, they were the second most reported causative factor (16.6%). Type of stuttering: As shown in Table 5, the type of stuttering in male and female CWS indicate that there was not much difference between male and female CWS. Both the groups had similar types and frequency of disfluencies. Comparing SLDs in both male and female CWS, sound repetitions were found to be the most frequent disfluency type. In the category of ODs, multisyllabic word repetition was found to be the most frequent disfluency. This supports the results of the study by Anjana and Savithri (2007) who reported similar results. Recovery between male and female CWS across severity levels and as a group Estimation of recovery from stuttering was done in 2 ways. Comparing the previous and current SS1-3 scores: To examine the recovery pattern in male versus females CWS, previous SSI-3 scores were compared with the current scores on SSI-3 (Riley, 1994). The results were compared between gender and across severity of stuttering. Mann-Whitney U test was carried out to obtain the significance, if any between male and female CWS, with respect to previous and current SSI-3 scores as a whole group (i.e, male CWS and female CWS) and within each severity level of stuttering (i. e, previous and present scores of male children with mild stuttering versus previous and present scores of female children with mild stuttering). This nonparametric test was used for the statistical analysis of these scores because the data were not normally distributed.

Table 6 depicts the previous and current SSI

scores for children with mild (1-10), moderate (11-20) and severe stuttering (21-24), based on previous scores. For children with mild stuttering male subjects 3, 4 and 5 and female subjects 2, 3, 5, 7, 8 and 10 presently obtained a score of less than 10 on the SSI-3 (Riley, 1994). This indicated that they can be classified as children with no stuttering (CWNS) as suggested by Coulter, Anderson & Conture, (2009). Thus, better recovery was seen in female (6

out of 10) than male CWS (3 out of 10). However, statistical comparison of previous and present scores across gender within the category of mild stuttering using Mann- Whitney U-test indicated that there was no significant difference between gender on previous and current scores of SS1-3, in mild stuttering group (z = -1.77, p > 0.05), which could be due to the fact that both male and female CWS recover better when they have mild stuttering. For children initially classified as having moderate stuttering Table 8 clearly depicts the difference in scores between the male and female CWS. It can be seen that no male CWS obtained a score of 10 and lesser, to classify them as CWNS, whereas in female CWS, subjects 1, 2, 4, 5 and 7 obtained a score of 10 and less, indicating complete recovery in them. Male subjects 4, 5 and 7 previously diagnosed with moderate stuttering showed a steady decline in their scores, falling into the category of mild stuttering. Female subjects 3 and 10 also were found to have a regression in their scores, coming under the category of mild stuttering.

Therefore, summarizing the obtained results, it was found that more female CWS recovered than male CWS. Statistical analysis using Mann-Whitney revealed a significant difference between male and female children with stuttering with respect to the SSI-3 scores (z = -2.58, p < 0.05), further supporting that female CWS have better chances of recovery than male CWS. For children initially classified as severe stuttering it can be observed from the table that all female CWS had partial recovery. 3 female CWS were found to have moderate stuttering presently (subjects 1, 3 and 4) and one female was found to have only mild stuttering (subject 2). In male CWS, 3 of the subjects were found to have partially recovery, 2 recovered to a moderate severity level (1, 4 subject) of stuttering and 1 subject recovered better, having only mild stuttering currently (subject 2). 1 male subject was found to have no recovery (subject 3). Therefore, in the category of severe stuttering, it can be seen that there was partial recovery for both male and female CWS. Also, there was 100% recovery in females whereas only 75% in male CWS. Results of Mann-Whitney U test indicated that there was no significant difference (z = 0.00, p > 0.05) between male and female CWS on the basis of SSI-3 scores in the severe category. However, the results need to be interpreted with caution due to the less number of subjects considered in both groups.

Gender Percentage Mean (%) SR SSWR SYR PR BL MSWR PHR INT REV

Male 26.46 8.30 17.84 7.38 11.38 11.07 3.69 6.46 7.38 Female 25.12 9.35 12.80 8.37 6.89 11.82 5.41 9.35 10.83

Gender difference in CWS

197

Table 6. Previous and current SSI scores across gender and severity

In addition, a Mann-Whitney U test was

carried out between gender on the previous and current SSI scores, as a whole, without dividing the subjects into different severity levels. The results indicated that there was a significant difference (z = -3.01, p < 0.05) between gender with respect to the previous and current SSI-3 scores. This implies that better recovery was observed in female CWS than male CWS. Therefore, the present study is in support of majority of the literature findings that female CWS recover better than male CWS (Felsenfeld, 1997; Yairi & Ambrose, 2004).

Type/frequency of disfluencies across gender and severity: The data was also analyzed to obtain the type and frequency of disfluencies across severity level (classified according to the previous SSI-3 scores) and gender. Frequency of SLD and OD in male and female CWS within the category of mild stuttering are shown in table 7. The results indicated that 3 male subjects (subjects 3, 4, and 5) had less than 2 SLDs per 100 words. Within female CWS, 6 subjects (subjects 2, 3, 5, 7, 8 and 10) had less than 2 disfluencies per 100 words of conversational speech. Therefore, these 3 male and 6 female CWS can be classified as CWNS (Pellowski & Conture, 2002). Frequency of SLD and OD in male and female CWS within the category of moderate stuttering are shown

in Table 8. Within the moderate stuttering group, no male CWS were found to have less than 2 SLD per 100 words, whereas 5 female CWS were found to have so (subjects 1, 2, 4, 5 and 7). No male subjects with moderate stuttering was found to have recovered completely whereas 5 female subjects with moderate stuttering were found to have recovered completely, which supports the findings by Pellowski and Conture, (2002).

Table 7. Frequency of SLD and OD in male and

female CWS (mild stuttering)

MIL

D S

TU

TT

ER

ING

Male Disfluency Female Disfluency SLD OD SLD OD

1 7 4 1 9 2 2 5 3 2 2 3 3 2 3 3 1 2 4 1 4 4 6 2 5 2 3 5 2 4 6 6 5 6 4 2 7 8 4 7 2 3 8 5 2 8 1 2 9 6 2 9 6 3 10 6 4 10 1 2

Table 8. Frequency of SLD and OD in male and

female CWS (moderate stuttering)

MO

DE

RA

TE

S

TU

TT

ER

ING

Male Disfluency Female Disfluency SLD OD SLD OD

1 20 8 1 2 5 2 16 4 2 2 6 3 14 3 3 9 2 4 9 3 4 2 3 5 4 1 5 2 4 6 22 5 6 11 2 7 6 5 7 2 5 8 14 4 8 7 2 9 21 7 9 7 2 10 7 3 10 4 2

Frequency of SLD and OD in male and female CWS within the category of severe stuttering are shown in Frequency of SLD and OD within the category of severe stuttering are shown in table 9 It was found that none of the male or female CWS had 2 or lesser disfluencies per 100 words of conversational speech and therefore, none of them can be classified as completely recovered.

Male SSI scores Female

SSI scores

Previous Present Previous Present 1 16 20 1 16 12 2 15 16 2 14 10 3 13 10 3 12 6 4 12 6 4 14 12 5 16 10 5 14 8 6 16 12 6 16 12 7 16 14 7 14 8 8 16 12 8 14 10 9 14 12 9 16 12 10 16 15 10 16 6 11 25 25 1 18 10 12 20 18 2 23 10 13 23 21 3 18 12 14 24 15 4 23 8 15 23 14 5 22 10 16 26 23 6 20 17 17 22 17 7 17 8 18 19 17 8 19 14 19 25 24 9 21 17 20 20 18 10 22 14 21 27 22 1 28 25 22 28 20 2 27 18 23 30 28 3 29 25 24 30 25 4 27 22

SE

VE

RE

Male Disfluency Female Disfluency

SLD OD SLD OD 1 13 4 1 13 7 2 9 2 2 11 3 3 15 6 3 13 3 4 14 4 4 8 5

Table 9. Frequency of SLD and OD in male and female CWS (severe stuttering)


198

Conclusions

The aim of the present investigation was to explore the gender differences, if any, in male and female CWS with regard to parameters such as age of onset, nature and type of stuttering, causative factors, associated problems, time since onset (TSO) and nature and duration of treatment taken. It was also aimed to compare the recovery rates across male and female CWS (as a group and within each severity level) using SSI-3 severity ratings and the percentage of SLDs vs. ODs. Analysis of the results revealed significant gender difference across all the parameters under study excluding the nature of onset.

References Ambrose, N., Yairi, E., & Cox, N. (1993). Genetic aspects of

early childhood stuttering. Journal of Speech and Hearing Research, 36, 701-706.

Andrews, G., Craig, A, Feyer, A-M., Hoddinott, S., Howie, P. M., & Neilson, M. D. (1983). Stuttering: A review of research findings and theories circa 1982. Journal of Speech and Hearing Disorders, 48, 226-246.

Andrews, G., & Harris, M. (1964). The syndrome of stuttering, Clinics in developmental medicine, No. 17. London: Spastics society Medical education and information unit in association with William Heineman Medical Books Ltd.

Anjana. B. R. (2004). Genetics in stuttering. Unpublished Master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing. University of Mysore.

Anjana B. R., & Savithri S. R. (2007). Disfluencies in 5.1 to 6 year old Kannada speaking children. Journal of All India Institute of Speech and Hearing, 26, 3-8.

Blood, G. W., Ridenour, C. Jr., Qualls, C. D., & Hammer, C. S. (2003). Co-occuring disorders in children who stutter. Journal of Communication Disorders, 36, 427-488.

Bloodstein, O., & Smith, S. M. (1954). A study of the diagnosis of stuttering with special reference to the sex ratio. Journal of Speech and Hearing Disorders, 19, 459-466.

Coulter, C. E., Anderson, G. D., & Conture, E. G. (2009). Childhood stuttering and dissociations across linguistic domains: A replication and extension. Journal of Fluency Disorders, 34, 257-278.

Felsenfield, S. (1997). Epidemiology and genetics of stuttering. In R. F. Curlee & C. M. Seigel (Eds). The nature and Treatment of Stuttering: New Directions (2nd ed.). Boston: Allyn & Bacon.

Goldman, R. (1967). Cultural influences on the sex ratio in the incidence of stuttering. American Anthropologist, 69, 78-81.

Kidd, K. K., Kidd, J. R., & Records, M. (1978). The possible cause of sex ratio in stuttering and its implications. Journal of Fluency Disorders, 3, 13-23.

Kent, R. D. (1983). Facts about stuttering: Neurologic perspectives. Journal of Speech and Hearing Disorders, 48, 249-255 Kolk, H., & Postma, A. (1997). Stuttering as a covert repair

phenomenon. In R.Curlee & G. Seigel (Eds.), Nature

and treatment of Stuttering: New Directions (2nd ed.) (pp. 182-203). Boston: Allyn & Bacon

Mansson, H. (2000). Childhood stuttering: Incidence and development. Journal of Fluency Disorders, 25, 47-57.

Nagapoornima, M. (1990). Disfluencies in children (3-4 years). In M. Jayaram & S. R. Savithri (Eds.), Research at AIISH, Dissertation Abstracts: Volume II, 171-173.

Pellowski, M. W., & Conture, E. G. (2002). Characteristics of speech disfluency and stuttering behaviors in 3- and 4- year old children. Journal of Speech, Language and Hearing Research, 45, 20-34.

Rajendraswamy (1991). Some aspects of fluency in children (6-7 years). In M. Jayaram & S.R.Savithri (Eds.), Research at AIISH, Dissertation abstracts: Volume III, 6-7.

Riley G. (1994). Stuttering severity instrument for children and adults (3rd Ed.). Austin, TX: Pro-Ed Publishing.

Ryan, B. P. (1992). Articulation, language, rate, and fluency characteristics of stuttering and non-stuttering preschool children. Journal of Speech and Hearing Research, 35, 333–342.

Silverman, E. M., & Van Opens, K. (1980). An investigation of sex bias in classroom teachers‟ speech and language referrals. Language, Speech and Hearing Services in the Schools,11,169-174.

Swapna, N., Jayaram, M., Prema, K. S., & Geetha, Y. V. (2010). Development of intervention modules for preschool children with communication disorders. An unpublished ARF project report undertaken at AIISH, Mysore.

Van Borsel, J., Moeyart, J., Mostaert, C., Rosseel, R., Van Loo, E., & Van Renterghem, T. (2006). Prevalence of stuttering in regular and special school populations in Belgium based on teacher perceptions. Folia Phoniatrica et Logopaedica, 58, 289–302.

Van Riper, C. (1982). The nature of stuttering (2nd ed.). Englewood Cliffs, NJ: Prentice Hall.

Wingate, M. E. (1964). Recovery from stuttering. Journal of Speech and Hearing Disorders, 29, 312-321.

Yairi, E., & Ambrose, N. (1992). Onset of stuttering in preschool children: Selected factors. Journal of Speech and Hearing Research, 35, 782-788.

Yairi, E., & Ambrose, N. (2003). Childhood stuttering: advances in knowledge and their clinical implications. American Speech-Language-Hearing Association annual convention, Chicago, IL. Abstract published in ASHA Leader, 8.15, 151.

Yairi, E., & Ambrose, N. (2004). Early childhood stuttering. Austin TX: Pro-Ed.

Yairi, E., & Ambrose, N. (2005). Early childhood stuttering: For clinicians by clinicians. Austin, TX: Pro-Ed, Inc.

Yairi, E., Ambrose, N., & Niermann, R. (1993). The early months of stuttering: A developmental study. Journal of Speech and Hearing Research 36, 521-528.

Yamini, B. K. (1990). Disfluencies in children (5-6 years). In M. Jayaram & S. R.Savithri (Eds.). Research at AIISH, Dissertation abstracts: Volume III, pp 171-173.

Young, M. A. (1984). Identification of stutterers and stuttering. In R. F. Curlee & W. Perkins (Eds.), Nature and treatment of stuttering. San Diego: College-Hill.

Resource manual for multisensory skills

199

Resource Manual for Children with Communication Disorders (3 – 8 Years) on Multisensory Skills

Priyashri S. 1 & R. Manjula 2

Abstract

Multi-sensory instruction refers to any learning activity that includes the use of two or more modalit ies to receive or express information. Children with communication disorders often demonstrate deficits in one or more sensory modalities. Multisensory approach capitalizes on their strengths to stimulate learning by engaging students on multiple levels. The sensory modali ties often focused in multisensory training include visual, auditory, tactile, kinesthetic and olfactory modalities. In India, where there is a large population of persons with communication disorders, such a training package is not available in any of the languages. Speech therapists and parents involved in treating children with communication disorders do not have access to any ready-made material that incorporates a multi sensory approach. Hence, there is a need to develop a resource manual that acts as a guide and enables the therapist and parents / caregivers to improve or train multisensory skills in children with communication disorders. The manual is one of the first attempts at helping parents and caregivers help their children develop multisensory skills. The structure of the manual has been so arranged so as to facilitate the parents/caregivers to fi rst assess their child‟s sensory strengths and weakness in various senses. The assessment has been made easy through a simple questionnaire that assesses the various sensory skills. This will give the parent/caregivers a good idea about which sensory skill is lacking/deficient and needs to be worked upon. They can move onto the appropriate sections in the subsequent part to develop/improvise the respective sensory skill using activities and suggestions in the section. Key words: multisensory approach, communication disorder, manual, training package, multisensory skills

ultisensory 1training approach is a model of language instruction which holds that learning or acquiring knowledge and skil ls

in given subject areas is best facili tated by involving more than one of learners‟ senses (McIntyre & Pickering, 1995). The sensory modalities which are often focused in multisensory training include visual, auditory, tactile, kinaesthetic and olfactory. Multisensory approach is used to facili tate a person‟s abili ty to learn and recall information by combining explicit information and multisensory strategies. Multisensory approach is the most effective teaching method for some groups of children with different disabili ties. Using a multisensory approach with children helps in channelizing more appropriately the information to be learnt by them to the brain and this in turn facilitates learning. Studies in the domain of cognitive psychology show that there are both cooperative functions of the senses and disruptive and inhibitory effects of one sensory system upon another. It is important, therefore, to use all of the child‟s senses while teaching him/her a new concept. However, empirical studies about sensory and motor development in children with autism are

1 e-mail:[email protected]; 2Professor of Speech Pathology, AIISH, Mysore, [email protected].

limited compared to studies of other aspects of development. Empirical evidence converges to confirm the existence of sensory and motor diff iculties in many children with autism at some point in their early development (Ornitz, Guthrie & Farley, 1977; Ohta, Nagai, Hara & Sasaki, 1987; Dahlgren & Gilberg,1989; Scharre & Creedon, 1992; Baranek, 1999). Reviewing sensory and motor impairments for children with autism, Baranek (2002) maintained that current evidence supports a correlation between sensory symptoms and Autism Spectrum Disorders (ASD), specifically a tendency for individuals with ASD to exhibit under- and over arousal in response to sensory stimuli. Sensory integration theory emerged at the same time that therapies aimed at individuals with learning disabili ties were increasing. The definition of the term learning disabilities, introduced by Ki rk in 1962, gave rise to assessments and treatments based on suspected sensory-processing, such as the Illinois Test of Psycholinguistic Abili ties (Kirk, McCarthy & Kirk, 1968), psycholinguistic approaches, and a myriad of perceptual motor interventions (Barsch, 1967; Delacato, 1963; Doman, Spitz, Zucman, Delecato & Doman, 1960; Frostig & Horne, 1964; Getman, Kane, Halgren & McKee, 1964).

M

Dissertation Vol. VIII. 2009-10, Part-B, SLP, AIISH, Mysore

200

Like sensory integration theory, the work of Doman et al., (1960) was based on the theory that individual development follows an evolutionary process. The National Research Council (2001) recognized that sensory integration therapy contains components of scaffolding, child-centred approach,

meaningful and appropriate play. The increasing number of rooms for multisensory training across American educational settings (Hogg, Cavet, Lambe & Smeddle, 2001) stands in favour of sensory integration. Such rooms also called Snoezelen rooms have been used extensively in individuals with profound and severe physical and mental disabili ties, as well as with individuals with autism, dementia, and traumatic brain injury. Here gentle stimulation of the primary senses of touch, taste, sight, sound and smell are provided to the individual to gain experience of autonomous discovery and to react and respond with no declared aims or purpose. Many comprehensive educational models have also been developed for children with different communication impairments. For children with autism, the TEACCH program that includes a sensory processing or motor development component has been proposed. For children with developmental apraxia of speech there have been several multi sensory approaches to treatment such as the association method, sensory motor approach and so on. For children with learning disabilit y, several multisensory training programs have been proposed and studies have been conducted to examine their efficacy and advantages over other forms of treatment. Thorpe and Bordon (1985) discussed the visual, auditory and kinesthetic importance when learning the sound to symbol relationship for reading and spelling. Anselmo and Kulp (1997) conducted a study to see if a multi-instructional program would increase the pre-reading skil ls of kindergarten children who were at risk. Their findings suggested that there was an impact on the student‟s ability to identify initial consonant sounds using multi-sensory approach. The students did not show a growth in their abili ties to identify or match corresponding upper and lower case letters when following the multisensory instructional intervention.

The investigators recommended that a multi-sensory instructional program be used with kindergarten children so that their reading readiness is better achieved by first grade. The Orton Gillingham Method and the Visual-auditory-kinesthetic-tactile method also utilize multi-sensory stimulation to teach children with learning disabili ty. As mentioned earlier, the Association method developed by McGinnis and Pribram (1980) believes that the essential processes of learning attention, retention and recall has to be

integrated to acquire automaticity of language. The investigator reports that such a multi-sensory form of instruction is effective not only for children with developmental apraxia of speech but also children with profound mental retardation, autism, cerebral palsy, traumatic brain injury and dyslexia. With the limited review of li terature that is available, it follows that multi-sensory instruction is fast becoming a popular method of intervention. For any child to develop communication through the multi-sensory form of instruction, she/he must have better functioning sensory abilit ies. Visual, auditory, kinesthetic, tactile, olfactory and gustatory modes are used as the input senses to teach language. Hence, it is obvious that these senses need to function at their optimum best both physically and cognitively. There are very limited multisensory training packages, comprehensive, ready-to-use manuals to guide the parents / caregivers in the development of these multisensory abili ties in their children who are diagnosed as having various communication disorders available in published forms. There have been many articles on integrated classroom and the advantages of such a set up. However, the child will never be able to utilize them all unless she /he has the readiness in her/him to receive the instruction. Thus there is a need for a manual that incorporates all these features to be introduced in the Indian context.

Method

The purpose of this study was to develop a resource manual to develop multisensory abili ties in children with developmental disabilities between the age group of 3 to 8 years. The manual is intended to be used by clinician and caregivers of children with developmental disabil ities. The manual was prepared in stages and there were five major stages as described below: Stage 1

A. A compilation of available information was done based on web search and other available actual resource materials. Various centres that offer training in specific sensory modalities were also researched and a note of the modalities addressed along with how each of them was addressed was also carried out. This yielded a fairly good idea of what already exists in different Indian languages. B. Based on the literature review, the drawbacks and lacunae in training children in sensory modalities were noted. This helped in identifying the constituents of the manual and additional requirements to be met. The outline and structure of the resource manual was


201

developed. The manual was planned to include five sensory domains of visual, auditory, tactile- kinaesthetic- proprioceptive, olfactory and gustatory. Stage 2

The manual was prepared with the resource materials for all the five sensory domains. Table 1 and Figure 1 and 2 show the overall structure of the manual developed in English language. The manual has been organized in a very simple and user-friendly way so that any individual who wishes to use the manual is able to follow it. It has been proposed in English. The manual focuses on developing skil ls in five major modalities in children between the age range of 3 – 8 years. Each of the modalities comprise of several skill s, each of which are addressed in the manual. The manual is therefore broadly divided into five main sections as follows:

A)Vi sual B) Auditory C)Tactile, kinesthetic and proprioceptive D)Olfactory E)Gustatory

Within each of the modalities a number of

skills are included (see appendix). Each modality consisted of a simple and easy-to-administer questionnaire that tests the abili ties of the child for various skil ls that fall under the modality. Questions that target each of the skills individually and help identify the child‟s level of functioning in a particular skill have been included in the questionnaire. For e.g., in part I of visual modality, questions that assess the child‟s abilit y for visual discrimination, form constancy etc. are included under the respective skil l-headings.

The second part in each modality consists of activities to improve the skil ls that the child does not have or has not mastered completely. Here again, each skill has been individually addressed through activities. This part further consists of definition of what the skill is and how it is important for the child, instructions to carry out the activity, type of activities to establi sh/improvise the skill and suggestions for home training.

Each of the sections has been organized in a similar fashion. However, for certain skil ls where the activity cannot be represented in the book/ manual, suggestions to carry out the same has been given through elaborate descriptions.

Stage 3

A. The material developed in stage 1 was subjected to familiarity rating by typically developing children in the age group of 3 to 8 years. The material was tested for famil iarity, ambiguity of pictures, clarity of audio recording, comprehension of instructions, scoring procedures and ease of use of material. B. The material was subjected to item/ content validity by giving the same to three experienced speech-language pathologists with a minimum experience of three years. Each activity includes appropriate instructions, resource material and scoring procedure wherever required. Stage 4

Based on the results obtained in stage 2, suitable modifications were incorporated in the manual for each domain. These modifications were in terms of increasing/reducing the size of the picture and incorporating pictures that related more to the Indian context and substitution of two pictures which were felt to be ambiguous.

Manual

Visual modality

Auditory modality

Tactile,Kines-thetic,propri-

oceptive

Olfactory Functions

Gustatory functions

Figure 1. Overall structure.

Figure 2. Structure of each module.

Questionnaire to assess the skills.

Activities to train in specific skills.


202

Stage 5

The manual was administered on three children with Autism, Expressive language disorder and Learning disability respectively. All these children were between the age range of 3 – 8 years. The parents (mothers) of these children were literate (minimum education of 10th standard) and could read and understand English. Feedback was obtained from these mothers which was rated on a three point rating scale for familiarity of pictures and photos used in the manual, comprehension of instructions, ease of administering the questionnaire to understand the specific deficit in the respective sensory domain and ease of carrying out the activities suggested in each of the subskil ls of the respective sensory domain. The items / skil ls which were rated low by the mothers were further probed into to obtain a description of the actual error/opinion. Stage 6

Based on the feedback obtained from the parents (mothers) suitable modifications were incorporated in the manual and the manual was finally printed. Who can use the manual?

The manual has been designed for clinicians and parents/caretakers alike. Clinicians working with children having communication disorders and parents/caretakers who are bringing up children with communication disorders can avail the best out of this manual by following instructions given at the beginning of every section. Suggestions and Guidelines for users

Both the parts of the manual, namely questionnaire and the section on therapy activities have an instruction set preceding them for each modality. This means that at the beginning of every questionnaire and activity section there is an instruction set for the user (clinician/ parent/ caretaker). All that the user has to do is to follow the instructions given prior to each section. Caretaker/ parent/ clinician are highly encouraged to be creative when administering the manual and the activities can be shaped in a way to suit the child‟s interests by using appropriate items.

Implication of the manual

The manual is one the few attempts at helping parents and caregivers help their children develop multisensory skill s. The structure of the manual has been so arranged so as to facil itate the parents/ caregivers to first assess their child‟s sensory strengths and weakness in various senses. The assessment has been made easy through a simple questionnaire that

assesses the various sensory skill s.

The manual has a provision to evaluate as well as develop sensory skills in various modalities. It plays in preparing the child to learn language both receptive and expressive. Sensory readiness as a prerequisite to learning language has often been emphasized in the literature and is also being touted as essential to schooling. Typically developing children who seem to acquire such a form of readiness naturally without any specific intervention of any sort demonstrate the usage of such sensory skills in their classroom learning as well as other communicative situations.

Certain developmental disorders such as autism, profound mental retardation etc. are characterized by very severe sensory impairments in specific sensory modalities that come in way of learning to communicate. These sensory skil ls when developed and when unwanted sensory behaviours are eliminated can facilitate language therapy and language learning on part of the child. The child will be in a better position to explore her/his environment and learn things better and retain them too.

Many a time olfactory and gustatory skills are ignored and grouped as the child‟s likes and dislikes. However, there may be more serious sensory issues that need to be addressed. This manual takes into account little consistent behaviours that indicate sensory issues in the child.

The manual will also help the child to retain what she/he goes onto learn in a better manner as indicated by some research studies that support multi-sensory instruction. The manual when administered also indirectly helps achieve the concept of an integrated classroom in its truest sense by enabling children with developmental disabilities to be equal to their counterparts in sensory abili ties thereby reducing a huge burden on them in terms of coping with peers. The manual developed keeping the Indian population in mind addresses the need for a multisensory training manual for children with developmental disabili ties in India.

Li mitat ions

The questionnaires that assess the visual and auditory skil ls have been developed keeping the developmental milestone of particular skil ls in mind. However, the questionnaires of the tactile, kinesthetic, proprioceptive, olfactory and gustatory modalities have been developed keeping the unwanted


203

/abnormal behaviours in mind. Thus the caregiver/clinician must be cautious when using them. The activities under each modality are not proportionate. More number of activities has been proposed under the visual, auditory and tactile modalities as compared to the gustatory and olfactory modalities.

Future recommendations

The manual needs to be validated on more number of subjects with various developmental disorders and tested for content/ item validity. Based on this, suitable changes/modifications need to be

incorporated. Once the manual is run on different types of developmental disabili ties and on more number of subjects it needs to be seen whether such a form of multisensory stimulation has differential effects on different disabilities and if so, what the differences would be. More number of activities to develop the olfactory and gustatory modalities need to be added to overcome the disparity in the number of activities suggested. The manual needs to be translated into other Indian languages so that it is accessible to more caregivers who are from different language backgrounds.

References

Anselmo, M. G., & Kulp, P. M. (1997). Phonemic awareness/ Multisensory instruction: An intervention for kindergarten children at risk in pre-reading. Eric Document reproduction Service.

Baranek, G. T. (2002). Eff icacy of sensory and motor interventions for children with autism. Journal of Autism and Developmental Disorders, 32, 397-422.

Barsch, R. H. (1967). Achieving perceptual motor efficiency: A space oriented approach to learning. Seattle, WA: Special Child.

Dahlgren, S. O., & Gillberg, C. (1989). Symptoms in the first two years of life. European Archives of Psychiatry and Neurological Science, 238, 169–174.

Delecato, C. H. (1963). Neurological organization and reading. Springfield, IL: Charles C Thomas.

Doman, R. J., Spitz, E. B., Zucman, E., Delecato, C. H., & Doman, G. (1960). Children with severe brain injuries: Neurological organization in terms of mobility. Journal of the American Medical Association, 174, 257-262.

Frostig, M., & Horne, D. (1964). The Frostig program for the development of visual perception: Teacher‟s guide. Chicago: Follett.

Getman, G. N., Kane, E.R., Halgren, M. R., & McKee, G. W. (1964). The physiology of readiness: An active program for the development of perception for children. Minneapolis, MN: Programs to Accelerate School Success.

Hogg, J., Cavet, J., Lambe, L., & Smeddle, M. (2001). The use of Snoezelen as a multisensory stimulation with people with intellectual disabilities: A Review of the research. Research in Developmental Disabilities, 22, 353-372.

Kirk, S. (1962). Educating exceptional children. Boston. Houghton Miff lin.

Kirk, S. A., Mc Carthy, J.J., & Kirk, W. D. (1968). Il linois Test of Psycholinguistic Abilities. Urbana: University of Il linois Press.

McGinnis, D., & Pribram, K. (1980). The neuropsychology of attention emotional and motivational controls. In E. Wittrack (Eds.), The Brain and Psychology: Academic Press.

McIntyre, S. E., & Pickering, D. (1995). Autism multisensory approaches: Retrieved July 23 2009 http://autism.lovetoknow.com/Autism_Multisensory_Approaches

National Research Council (2001). Sensory integration: Retrieved August 16 2009 from http:// webcache. googleusercontent. com/ search? q= cache: b2046s9ZorwJ: www. autismnetwork. org/ modules/ sensory/ reference. html+ National+ research+ council+ 2001+ sensory+ integration+ therapy& cd= 1& hl=en&ct=clnk&gl=in

Ohta, M., Nagai, Y., Hara, H., & Sasaki, M. (1987). Parental perception of behavioural symptoms in Japanese autistic children. Journal of Autism and Developmental Disorders, 17, 549–563.

Scharre, J. E., & Creedon, M. P. (1992). Assessment of visual function in autistic children. Optometry and Vision Science, 69, 433–439.

Thorpe, H. W., & Borden, K. S. (1985). The effect of multisensory instruction upon the on-task behaviours and word reading accuracy of learning disabled children. Journal of Learning Disabilit ies, 18, 279- 286.


204

APPENDIX

Details of subskills of each sensory domain for which activities are suggested

Number of sub skills

SENSORY MODALITIES

Visual Auditory Tactile-

Kinesthetic – Proprioceptive

Olfactory Gustatory

Visual Awareness

Awareness of sound

Tactile Awareness

Hypersensitivity to Smells (Over-Responsive)

Hypersensitivity to Oral Input (Oral Defensiveness)

Visual discrimination

Early listening ---- talking loop

Tactile Hypersensitivity

Hyposensitivity to Smells (Under-Responsive)

Hyposensitivity to oral and gustatory input (Under-Registers)

Visual memory and recall

Association of sound with meaning

Tactile Hyposensitivity

Olfactory identification

Gustatory identification

Visual closure

Localization Skills

Tactile Perception and Discrimination

Visual figure ground discrimination

Auditory Discrimination

Hypersensitivity to Movement (Over-Responsive)

Form constancy

Distance and Directional Listening

Hyposensitivity to Movement (Under-Responsive)

Position in space

Auditory – foreground background discrimination

Poor Muscle Tone and/or Coordination

Depth perception

Auditory Memory and Sequencing

Proprioceptive dysfunction- Sensory Seeking Behaviours

Visual Sequencing

Auditory Closure

Proprioceptive dysfunction-Grading Of Movement

10. Visual Seriation

11. Part/Whole and Whole /Part Relationships

12. Colour perception and colour constancy

13. Visual conceptualizing

14. Visual pattern-following

15. Visual analysis and synthesis

Effect of training on voice of carnatic singers

205

The Effect of Training on the Voice of Carnatic Classical Singers Ranjini Mohan 1 & S.R. Savithri 2

Abstract

There is a need to know the effect of training on the various parameters of voice in Carnatic singers. In this context, the present study was undertaken. The objective of the study was to compare the voice of trained Carnatic and beginning Carnatic singers. Two groups of subjects participated in the study. Group I consisted of 20 female trained singers and group II consisted of 20 female beginning singers. Subjects in group I had at least 10 years of training in Carnatic classical singing and subjects in group II had less than 2 years of training in Carnatic classical singing. The age range of subjects in group I was 12-40 years (mean = 16.46 years and SD = 7.93) and those in group II was 16-40 years (mean = 27.5 years and SD = 8.56). Specifically, the acoustic parameters maximum phonation duration, habitual frequency, frequency of the base note, frequency range in singing, speaking, and phonation, vibratos, presence / absence of singer‟s formant and voice projection (through measures of skewness and kurtosis) were measured. The results revealed a significant difference between the two groups for maximum phonation duration, difference between the habitual frequency and base note, range of frequencies in phonation, number of vibrato and Skewness and Kurtosis in singing and speech. This indicates that the trained singers did have an edge over the beginning singers over some acoustic features as a result of practice.

Key words: carnatic music, trained singers, beginning singers, skewness, kurtosis

oice is 1a marvelous device, and the human voice that produces it is compared to a musical instrument. Because singing is a common

musical outlet, the complexities of the vocal mechanism, the high degree of coordination and the amount of energy required for artistic performance are often overlooked, taken for granted, or not understood.

Singing is defined as a sensory motor

phenomenon that requires particular balanced physical skills (Bunch, 1982). Sundberg (1994) stated, “the singer must gain control over all perceptually relevant voice parameters, so that they do not change by accident and signal an unintended boundary”. Technical control of all voice parameters is a pre-requisite for artistic expression.

The primary difference between singers and non

singers is not greater total lung capacity, but rather it is the use of greater proportion of air in his or her lungs, thus decreasing residual volume and increasing respiratory efficiency (Boominathan, 2008). Hirano (1981) reported MPD of 34.6 secs in normal adult males and 25.7 in normal adult females. Venugopal, Rajasudhakar and Savithri (2005) measured MPD in 80 normal children between 5 and 13 years. They found that the mean MPD for /a/ among girls between 5-5.11 years was 5.33 seconds, which increased to 14.8 seconds in the girls between 12-12.11 years. However, Sheela (1974) reported no significant difference in


phonation time between the trained and untrained singers.

Chandra (2001) compared base pitch for singing

of trained Carnatic singers with their habitual frequency for speaking/phonation. She found that subjects with more than six years of training had a lower base pitch than those who had training for less than six years.

The actual definition of singing range is the span

from the lowest to the highest note a person can create with their voice. In speech, pitch slides up and down a restricted range of semitone covering a range of 6 or 8 notes, i.e. three or four notes above and below the middle note. In singing, however, pitch reaches one and a half octaves in untrained voices and two to two and half octaves in trained voices, i.e. on the 8th or 10th note above and below the middle note (Greene, 1972). Sheela (1974) studied frequency range in 30 (24 female and 6 male) trained and untrained singers between 19 and 57 years. They were asked to phonate their lowest and highest possible notes. She found the frequency range in trained singers to be 2.0 to 2.75 octaves, which was significantly higher than in untrained singers, which varied between 1.0 to 1.5 octaves. The average voice can master a variation from one and one and a half octaves to two octaves during speech (Karr, 1953).

Vibrato is a pulsation of pitch, usually

accompanied by a synchronous pulsation of loudness and timbre. It is important as it gives pleasing

V

http://howtosing.org/


206

flexibility, tenderness and richness of tone to good singing. Average rate of vibrato of a good singer is 6 to 7 c/s. Vibrato develops with training.

The term “singer‟s formant”, was introduced by Sundberg (1974) who showed that the singer‟s formant could be interpreted articulatorily as a clustering of third, fourth and fifth formants, which may be an acoustic mismatch between the pharynx and the entrance of the larynx tube. Sengupta (1990) studied singer‟s formant in North Indian classical singing and singer‟s formant was observed. Chaya Devie (2003) compared Hindustani and Carnatic singers for presence of singer‟s formant. The results showed that singer‟s formant was present in Indian classical singers. Boominathan (2004) used LTAS module (FFT analysis) to determine if Carnatic singers demonstrate this peak energy contour. Contrary to the above studies, he did not find a singers‟ formant in Carnatic singers as such music is essentially homophonic and accompaniments usually shadow the singer and hence the singer would not require to project his/her voice over an orchestra.

Voice projection is the strength

of speaking or singing whereby the voice is used loudly and clearly. Skewness and Kurtosis are used to infer projection.

The review of literature throws light on some

aspects of acoustics of singing. However, it is insufficient to draw any conclusions. Further, it is known that training improves any skill and singing is not an exception. However, there is no empirical evidence to state that the acoustic measures in trained singers are different from untrained singers. In this context, the present study compared the voice of trained Carnatic with those of beginning Carnatic singers. The results of the study would throw light on how vocal training would affect the voice of these singers during singing and speech.

Method

Participants: Two groups of subjects participated in the study. Group I consisted of 20 female trained singers and group II consisted of 20 female beginning singers. Subjects in group I had at least 10 years of training in Carnatic classical singing and subjects in group II had less than 2 years of training in Carnatic classical singing. The age range of subjects in group I was 12-40 years (mean = 16.46 years and SD = 7.93) and those in group II was 16-40 years (mean = 27.5 years and SD = 8.56). Those with any other form of voice training (like Hindustani music) prior to or during the study period were excluded from the study.

Those with a history of vocal pathology and ongoing pubertal or maturational voice changes, auditory problems and other respiratory or cardiac illness were also excluded. Procedure: Phonation of vowel /a/, gliding from lowest to highest pitch, singing of gamaka, and singing and speaking of the song /kereja ni:ranu/ formed the material. All the voices were recorded in a quiet room on a Sony ICD-UX71 digital voice recorder at a sampling frequency of 441000 Hz. The following parameters were considered for the study: maximum phonation duration, habitual frequency, frequency of base note, singing frequency range, speaking frequency range, maximum frequency range in phonation, number of vibratos, singer‟s formant, skewness and kurtosis.

The researcher demonstrated each task before the participants‟ voice was recorded and three recordings were made for each parameter. The recorded phonation/ speech/ singing samples were transferred to the computer and saved on to the computer memory. PRAAT (Boersma & Weenink, 2006) software and LTAS of CSL 4500 (Kay Pentax, New Jersey) were used to extract the parameters. Number of vibratos/ gamaka was calculated.

The data was compiled and statistical analysis

was done using commercially available SPSS 16.0 software. Mean and SD of the parameters were computed and MANOVA was done to find out the significant difference between groups on the 9 parameters investigated. Independent t-test was also used to confirm the results of MANOVA.

Results

MPD: Results of MANOVA showed a significant difference {F (1.38) = 22.63, p<0.001} between groups. Trained singers had longer MPD than beginning singers. Habitual frequency: Results of independent t-test indicated no significant difference {t (38) = 1.720, P > 0.05} between groups. However, the mean habitual frequency and SD in group II was higher than those in group I. Frequency of base note: Results of independent t-test indicated no significant difference {t (38) = -0.351, P>0.05} between groups on frequency of base note. However, subjects in group I had higher frequency of base note compared to those in group II.

http://en.wikipedia.org/wiki/Speech_communication

http://en.wikipedia.org/wiki/Singing

http://en.wikipedia.org/wiki/Voice

http://en.wikipedia.org/wiki/Enunciation


207

Difference between habitual frequency and frequency of base note: Results of MANOVA indicated significant difference {F (1, 38) = 1.012, P<0.05) between groups. Subjects in group I had less difference compared to those in group II. Singing frequency range: Results of MANOVA showed no significant difference {F (1, 38) = 2.224, P>0.05} between groups. However, group I had higher singing frequency range and SD compared to group II. Speaking frequency range: Results of MANOVA showed no significant difference {F (1, 38) = 0.197, P>0.05} between groups. However, group I had higher speaking frequency range compared to group II. Frequency range in phonation: Results of MANOVA showed significant difference {F (1, 38) = 32.001, P<0.01} between groups. Group I had higher frequency range and SD compared to group II. Number of vibratos / gamakas: Most of the beginning singers were not able to produce any gamakas, but for two participants who produced two gamakas each. On the other hand, the trained singers were able to produce a maximum of 5 gamakas (6 participants); others were able to produce 4 (4 participants), 3 (9 participants) and 2 (1 participant). Figure 2 shows the number of gamakas produced by the two groups of subjects. Singers Formant (Fs): No singer‟s formant was noticed in any of the subjects in either group gamakas within one second. Skewness and Kurtosis: Results indicated significant difference between groups on Skewness {F (1, 38) = 11.978, P<0.01} and Kurtosis {F (1, 38) = 25.075, P<0.01} in singing and speaking [skewness – {F (1, 38) = 61.383, P<0.01; kurtosis – {F (1, 38) = 162.299,

Figure 2. Number of gamakas produced by the two groups of singers.

P<0.01}]. Both groups had negative skewness indicating no projections in the voice. Figures 1 and 3 show an example from the sample of LTAS in singing and speaking, respectively. Table 1 shows the results of the present study.

Discussion

The results revealed several points of interest. First, a significant difference between groups on MPD was observed. Subjects in group I (singers with > 10 years of experience) had longer MPD compared to those in group II (singers with < 2 years of experience). First of all we shall observe whether the MPD in subjects of the present study was normal or not. Hirano (1981) reported MPD of 34.6 secs in normal adult males and 25.7 in normal adult females. Venugopal, Rajasudhakar and Savithri (2005) reported MPD of 5.33 secs in children between 5 and 5.11 years and 14.8 secs in girls between 12 and 12.11 years. The results of the present study are not in consonance with that of Hirano. There are differences in MPD between countries owing to the build of people. Therefore, it cannot be expected that Indian subjects will have the same MPD as reported by Hirano (1981). The MPDs obtained in the present study was 14.10 in subjects of group I and 10.30 in group II. However, there were 10 subjects in group II who had MPD below 10 seconds.

Figure 1. LTAS in singing in group I (left) and group II (right) subjects in singing task..

Figure 3. LTAS of speaking in group I (right) and group II (left).


208

Table 1. Results of the present study Parameters I (trained) II (beginning)

MPD (sec) 14.10 (1.87) 10.30 (3.04) Habitual frequency (Hz) 205.59 (17.40) 222.94 (41.63) Frequency of base note (Hz) 205.59 (17.40) 202.84 (30.45) Habitual frequency ~ frequency of base note (Hz) 10.47 (15.75) 20.11 (39.82) Singing frequency range (Hz) 226.62 (54.96) 205.25 (28.83) Speaking frequency range (Hz) 105.89 (19.41) 103.40 (15.69) Frequency range in phonation 301.40 (70.47) 191.39 (50.95) Singer‟s formant Absent Absent Skewness (singing) -1.93 -1.22 Skewness (speaking) -1.23 -0.47 Kurtosis (singing) 3.84 2.03 Kurtosis (speaking) 1.75 -1.15

These results are in contrast to those of Sheela (1974), who found no significant difference in phonation time between trained and untrained singers. Also, Carnatic vocal practice consists of singing many notes or long verses in a single breath and in various tempos. Such training starts early in the syllabus of Carnatic music. The experienced singers might have greater practice using their respiratory apparatus to maintain singing longer utterances. Hence, their phonation time also increases with practice. Therefore, the results can be interpreted that the trained singers had greater respiratory efficiency i.e. their laryngeal mechanism uses the air supply effectively.

Second, the results showed no significant

difference between groups on habitual frequency, frequency of base note, frequency range of singing and speaking. The first result that there was no significant difference in habitual frequency is expected, as habitual frequencies need not be different just because a group of subjects are trained singers. Many a time the music teachers do not select an appropriate base pitch as relevant to the physique of the student. Therefore, it cannot be expected that the frequency of base note will be different in two groups of subjects. This result does not correlate with the study by Chandra (2001) who found that subjects with more than six years of training had a lower base pitch than those who had trained for less than six years. However, many of the beginning singers‟ production of a:dha:ra sr.ti did not match that of sr.ti box, and hence their production of the pitch was sharp. But the trained singers were successful in matching their pitch with the sr.ti box. The lack of significance between the trained and beginning singers on frequency range in singing could be because of the song selected. Both the groups of singers were asked to sing the same song, which restricted their singing frequency range as they had to stick to the notes of the song. Hence, this task does not elicit the wide singing

range in group I subjects. However, it was noticed that subjects in group I had wider range of frequency compared to those in group II. This result reflects the fact that in spite of the same song, the subjects in group I was able to use ornamentation in singing which reflected in wider range.

Third, the difference between habitual frequency and that of base note was significantly different in the two groups with group I subjects‟ habitual frequency closer to the frequency of base note. Though there was no significant difference between groups on habitual frequency and frequency of base note, the difference between the two was significant. While subjects in group I had lower habitual frequency than frequency of base note, those in group II had higher habitual frequency than frequency of base note.

Fourth, subjects in group I had significantly wider range of frequency in phonation than those in group II. The frequency range in group I was 301.4 Hz, while that in group II was 191.39 Hz. This study is in agreement with Sheela (1974) who found a greater range of pitch in trained singers than in untrained singers. She found the frequency range in trained singers to be 2.0 to 2.75 octaves, which was significantly higher than in untrained singers, which varied between 1.0 to 1.5 octaves. The results could be interpreted as the wider range in group I resulting from practice.

Fifth, subjects, in group II (except two of them)

were not able to produce gamaka/vibrato. This may be because of the curriculum in Carnatic music. The initial part of the curriculum focuses more on ta:la, tempo and rhythm and gamakas are not taught. Therefore, subjects in group II, in spite of a demonstration might not be able to produce gamaka. Also, production of Gamakas requires great control of respiratory and laryngeal


209

mechanism. The physiologic tremors in the cricoarytenoid and thyroarytenoid muscles are a result of practice and training which might not be achieved by singers with singing experience of < 2 years.

Sixth, none of the subjects in the present study exhibited singer‟s formant. This is in contrast with the results of Sundberg (1974), Sengupta (1990) and Chaya Devie (2003). However, it is in consonance with that of Boominathan (2004) who stated that in Carnatic singing, music is essentially homophonic and accompaniments usually shadow the singer and hence the singer would not require projecting his/her voice over an orchestra.

Seventh, a significant difference between groups

on Skewness and Kurtosis in singing and speaking was noticed. Subjects in group I had lower skewness and higher kurtosis compared to those in group II. The data on skewness and kurtosis indicated absence of voice projection. This is in consonance with the results of Devvuru & Savithri (2006) who reported no one to one correspondence between acoustic and perceptual correlates of projection. Voice rated as having good projection did not have low skewness and high kurtosis. The LTAS in the present study needs some description. They were unusual with the spectra spread over all frequencies and most of the time tilting upwards in group I. The reason for this is not known. Initially it was thought that the result was related to the recording procedure. Hence, a singer sang and spoke the song in to the microphone of CSL 4500. However, the results were the same with LTAS tilting upwards in high frequency. The results are interesting as it is known that LTAS is predominant in low frequencies and has lower energy at high frequencies. A software scientist was consulted and he also was of the opinion that the recordings were good (did not clip or was not recorded on multiple channels and was free from any other distortions that could affect the analysis). Hence the unusual LTAS in the present study still remains unsolved.

Conclusions

The results of the present study have thrown light on the acoustic parameters in beginning and trained singers. The results are encouraging in that some of the acoustic parameters were significantly different in two groups of subjects. However, the results of the present study are restricted to Carnatic singing. In the present study the same song was used in two groups of subjects for comparison. This was not able to reflect the frequency range used by two groups of subjects. Future studies on various glottal

parameters and extension of the present study for frequency range of singing using various songs are warranted.

Acknowledgements

The authors would like to thank Dr. Vijayalakshmi Basavaraj, Director, All India Institute of Speech and Hearing, Mysore for granting permission to carry out this study. Also, the authors acknowledge the subjects who participated in the study.

References Boominathan, P. (2004). Do carnatic singers have singer‟s

formant? Proceedings of Frontiers of Research in Speech and Music.

Boominathan, P. (2008). Assessment of professional voice. Proceedings of Seminar on Voice: Assessment and Management, All India Institute of Speech & Hearing, Mysore.

Boersma, P., & Weenink, D. (2006). PRAAT v 4.5.01. Institute of Phonetic Science, University of Amsterdam.

Bunch, M. (1982). Dynamics of the singing voice. New York: Springer, Verlag.

Chandra, A. (2001). Relationship between fundamental frequency of voice and „a:dhara shruthi‟ in Carnatic vocal music. Students research at AIISH: Articles based on Dissertation, Vol IV.

Chaya Devie, N. (2003). Singer‟s formant in Indian classical singers – Carnatic vs. Hindustani. Students research at AIISH: Articles based on Dissertation, Vol I.

Devvuru, S., & Savithri, S. R. (2006). Acoustic and perceptual attributes of vocal projection in Carnatic classical singers. Proceedings of PAS3 06, London.

Greene, M. C. L. (1972). The voice and its disorders. Pittman Medical Publishing Co., III Edn., London.

Hirano, M. (1981). Clinical disorders of voice, Disorders of Human Communication, 5, Springer Weain.

Karr, J. M. (1953). Developing your speaking voice. Harper and Row Publishers, New York.

Sengupta, R. (1990). Study on some aspects of the “Singer‟s Formant” in North Indian classical singing. Journal of Voice, 4(2), 129-134.

Sheela, E. V. (1974). A comparative study of vocal parameters of trained and untrained singers. Students research at AIISH: Articles based on Dissertation, Vol I.

Sundberg, J. (1974). Articulatory interpretation of the “singing formant”. Journal of the Acoustic Society of America, 55, 838-844.

Sundberg, J. (1994). Perceptual aspects of singing. Journal of Voice, 8(2), 106-122.

Venugopal, M. B., Rajasudhakar, R., & Savithri, S. R. (2005). Maximum phonation duration and s/z ratio in Dravidian children. Journal of the Indian Speech and Hearing Association, 19, 47-51.


210

Treatment Manual in English for Indian Children with Dyslexia Ranjini R. 1 & Jayashree C. Shanbal 2

Abstract

Dyslexia refers to difficulty in the acquisition of literacy skills that may be caused by a combination of phonological processing, visual and auditory processing difficulties. Management issues of these children with dyslexia have always been a challenge for the speech and language therapist. Though many treatment strategies have been discussed and their efficacy reported in literature, not many attempts have been made at training the Indian children. The present study thus aimed at developing a manual in English of Indian Children with Dyslexia (CWD) and also field testing the same. Such a focus of intervention would help the children tackle the difficulties they face in their second language at school level. The study initially involved the development of the treatment manual. The manual consisted of four sections namely listening comprehension, phonological awareness, reading skills and reading comprehension. This therapeutic intervention was carried out for four children with dyslexia for 10 sessions. Pre-therapy scores of these children on the skills worked upon were compared to post-therapy scores. Results revealed improvement in all the domains which were worked upon though not statistically significant. Participant 1 (D1) and D2 showed greater improvement across all domains worked upon. This study thus brings light into the domains which need to be incorporated when dealing with the therapeutic intervention of children with dyslexia. However generalization issues still needs to be addressed. Key words: treatment manual, dyslexia, phonological awareness, intervention

yslexia is a 1term that has been loosely applied to reading disabilities. Specific definitions for dyslexia vary with disciplines. Dyslexia refers

to difficulty in the acquisition of literacy skills that may be caused by combination of phonological processing, visual and auditory processing difficulties. One of the most daunting and clearly defined current challenges for both researchers and Speech-Language Pathologists is to develop, disseminate and implement methods for training children on reading skills. Existing research related to the treatment trends available till now is briefly explained in the following sections.

While considering remediation for children with dyslexia, phonological awareness training is found to be the most discussed in literature. Phonological awareness is a metalinguistic ability which enables the child to analyze the sound structure of language. Many studies have been conducted in the western setting indicating significant improvement in decoding and word reading accuracy, though generalization remained a question. In the Indian context, Shilpashree (2004) found that training of meta-phonological skills in Kannada speaking children also showed improved reading abilities in that language. Auditory training has also been applied in treating children with dyslexia as they are thought to have difficulty processing auditory information. Perfetti (2007) reported improvement in reading abilities of six children with dyslexia with a

1e-mail: [email protected]; 2Lecturer in Language Pathology, AIISH, Mysore, [email protected].

combined auditory and articulatory training. Auditory temporal processing could be trained effectively at the sound and phoneme levels. However, no significant stable transfer of these improved abilities on reading and spelling exceeding the effect of the school-based standard training was demonstrated.

Reading fluency is another aspect which has received immense attention recently. Recent researchers opine this skill to be important in determining the overall reading efficiency. For reading, fluency speed and accuracy have been traditionally considered the hallmark or most essential features used to describe the measure and interpret fluency (LaBerg & Samuels, 1974). Researchers also suggest that accuracy in itself is not sufficient because for complete understanding of any given text, reading should be rapid enough to understand the connections between the printed ideas (Nathan & Stanovich, 1991). These research reports thus emphasize the importance of rate for fluent reading and also for better comprehension. Reading rate in turn depends upon the purpose of reading as stated in the theory of reading (Carver, 1992). He mentioned 5 important gears of reading which makes explicit predictions about the skill which needs to be emphasized depending upon the purpose of reading. 5 gears include scanning, skimming, raiding, reading to learn and reading to memorize. Reading fluency involves various strategies which have been reported in the literature. Many studies have been carried out to find the efficacy of these strategies.

D

Treatment manual in English for Indian CWD

211

Repeated reading is one of the oldest adapted strategies by most clinician while working at improving reading fluency (Rose, 1984; Rashotte & Torgesen, 1985; Cohen, 1988; Daly & Martens, 1994; Marston, Deno, Dongil, Diment & Rogers, 1995; Swain & Allinder, 1996). Repeated reading with modeling (adult, proficient peer or computer) was also adapted a little later, found to be more efficient and effective than reading repeatedly alone (Gilbert, Williams & McLaughlin, 1986; Moseley, 1993; Daly & Martens, 1994; Vaughn, Chard, Bryant, Coleman, Tyler, Thompson & Kouzekanani, 2000).

Reading comprehension is often limited in children with reading disabilities because of difficulties with accurate and fluent word recognition, and because they have missed opportunities to acquire reading comprehension strategies (Brown, Palincsar & Purcell, 1986). Hence, strategies like making use of contexts to comprehend, making inferences, visualization of the material being read etc. which have proven to be effective with children with dyslexia needs to be focused to attain a complete and efficient management outcome. For any remediation program for reading disability to attain completion reading fluency and automaticity needs to be achieved at the end of the program. Various approaches are available to treating the reading difficulties. As discussed earlier one approach would focus only on intense training on word reading and related phonological processing skills (Vellutino, Scanlon, Sipay, Small, Pratt, Chen & Denkla, 1996; Torgesen, Rashotte & Alexander 1997; Foorman, Francis, Fletcher, Schatschneider & Mehta, 1998; Vadasy, Jenkins & Pool, 2000) other being the one involving primary focus on training reading comprehension strategies (Englert, Tarrant, Mariage & Oxer, 1994). Yet another approach involves a combination of word reading and comprehension instruction (Iverson, Tunmer & Chapman, 2005). None of these approaches are mutually exclusive though. When working with reading comprehension, inferential thinking is one skill that is a major predictor of one‟s reading ability because it is known to play an important role in understanding and integrating texts (McKoon & Ratcliff, 1992; Graesser, Singer & Trabasso, 1994; Kucan & Beck, 1997). Most children who have relatively normal reading abilities are found to have more difficulty at answering inferential questions when compared to literal questions.

Though such wide range of research review is available discussing the various methods of treating children with dyslexia, most of the above mentioned studies holds good only for those children being

rehabilitated in L1 which is relatively implicit. The dyslexia instruction becomes much more challenged when the treatment needs to be given in L2 which is explicitly learnt by child in school settings. As mentioned by Gersten et al., (2001) many additional issues such as cultural acceptance and so on needs to be addressed when treating children who are English learners as in Indian setting. Studies reporting the efficacy of treating second language are very scanty and needs to be investigated in depth. Studies need to be carried out in children with English as L2 in determining how efficient the above mentioned treatment strategies and approaches are when dealing with a child acquiring the language. Children with English as second language may require intense instruction similar to low achievers owing to the differences in the orthography of Indian languages in comparison to English. And with intense training in English language Indian children may be benefitted when culturally appropriate materials are made use and also when native language is strategically used to teach the overcome the complexities with English. Such an approach in turn would not require exemption of English language for these CWD in India which is now being done to tackle the difficulty with the language. Henceforth an attempt to provide an intense instruction and thereby helping those children with the potential to learn to overcome the language difficulties needs to be taken up rather than blindly exempting them from this opportunity.

The aim of this study was to develop a treatment manual in English for Indian children with dyslexia which is culturally appropriate and also incorporate the rules of the language which may not be explicitly taught to these children in school settings.

Method

Participants: Participants included children with Learning Disability (in the age range of 11-13 years) from Grades V and VI. This clinical group consisted of 3 males and one female. None of the children enrolled for treatment had any known or reported history of hearing, neurological, developmental or emotional problems. Subject selection criteria: All the children with learning disability (CWLD) were diagnosed on the Test of Early Reading Skills (Loomba, 1995). Native language of all the participants was Kannada with English as the medium of instruction in school. All the participants were assessed by a clinical psychologist for their intelligence reported average or above average


212

intelligence in the children. All these children were enrolled for speech-language intervention. Development of the treatment manual: The manual was prepared incorporating four domains of treatment explained below. Materials and activities were taken from various textbooks, internet resources and journals. Each of the above mentioned domains had several other sub sections arranged in the hierarchical order of difficulty. Criteria: On attaining a 90% criteria (in phonological awareness, phonics) and 85% (in reading comprehension, and listening comprehension) the subject could advance to the next level of the corresponding section. Section 1: Listening comprehension skills: This domain consists of the following sub-sections. Literal comprehension- This includes questions seeking the information directly stated in the passage played/read. Missing information- This includes listening material with few of the content words missing, which the child is required to arrive at. Inferential information - The child was made to inference new information which is unstated in the listening sample played/read. Scoring: Each section consisted of listening passage accompanied with questions. Each question was awarded a score of „1‟ if answered correctly and „0‟ if not correct. Section 2: Phonological awareness skills: Word recognition is considered to be important process in reading and usually emerges in the initial stages of reading acquisition. For proper decoding and access to the appropriate semantic component word recognition requires to be adequate. This domain consisted of the following sub-skills: Pre-phonemic skills- This consisted of activities like counting the number of words in a given sentence, counting the number of syllables in a given word, judging a pair of words for similarity, choosing the odd one from a set of rhyming words, etc. Phonemic awareness skills- This is considered to be deep phonological awareness requiring manipulation tasks such as consists manipulation activities like phoneme segmentation, onset-rime blending etc. This

skill consisted of the following activities, onset-rime blending, onset-rime segmenting, phoneme blending, phoneme segmenting, phoneme addition, phoneme deletion and phoneme substitution. Section 3: Reading skills: This section consisted of skills beginning from recognition of letter names and sounds (phonics), sight word recognition and reading fluency. Phonics: Phonics involved activities aimed at establishing letter-sound knowledge. These skills involved learning about the letter sounds and their relationship to words. These instructions were included once the child familiarizes with the alphabets. Sight word recognition: Sight word reading in the present manual involved the game Bingo where fast identification of words are facilitated by imposing a time limit for recognizing the word and also for producing it. This section involved activities promoting sight-word reading. Accuracy and the speed of processing the word were focused by providing feedback to the child. Reading fluency: Reading fluency consisted of strategies for improving the fluency of reading. The activities included teaching sight word reading and reading using strategies like repeated reading with feedback/modeling, rereading with timing and listening preview. In the present study reading fluency consisted of strategies namely repeated reading, listening preview and re-reading with time. This section consisted of a graph for plotting the reading rate in each session. Accuracy and rate measures were obtained in this section. Section 4- Reading comprehension skills: This domain involved various strategies which need to be incorporated in order for better comprehension of the material read. The strategies mentioned below were to be worked out in parallel most of the times when reading any material; hence there was no specific order of teaching them. The following were the strategies incorporated in the manual for facilitating better comprehension.

Fix it up strategy- This strategy helps to teach the child chunking any unknown word into parts and arriving at its pronunciation by chunking into parts similar to already known words and trying to sound them out the same way or just by chunking and trying to sound it out using the known phonetic rules of English.


213

Retell and summarize- Retell and summary involved various steps to help the child retell any given story retaining three important aspects required namely, the facts of the given story, the sequence and the meaning using the retelling cards and graphic organizers. This strategy has been reported to be effective instruction strategy and its long term influence on reading has also been emphasized. Coding: Coding involved teaching the child to code the important portions of the text, the main ideas, the words which were not understood, events in the text which the child can connect to his earlier experiences, and also concepts new to him/her again using coding cards. Connect: Connect involved training the child to associate any read material to his own real life experiences wherever applicable. Referencing and inference strategy: Inferencing is often considered to be more complex to literal interpretation and hence also helps in greater thought process following reading. This skill teaches the child to infer information which is not given in the text and generate new ideas. It also helps in improving referencing the text which leads to such an inference which ultimately results in better understanding of the material read. Graphic Organizers: Story mapping, KWL strategy (K- what we know, W- what we want to know, L- what you have learned): is a strategy used when reading expository texts. This teaches the child to link the already known information with the newly learnt one. Graphic organizers are considered strategies of content enhancement which have been proven to be efficient. Administration of the manual: A 10 session treatment was planned and carried out for four Dyslexic children using the above prepared manual and pre and post test measures of various reading skills of these children was also obtained. Procedure: Before enrolling the subjects for therapy a Pre-therapy evaluation was done using Test of Early Reading Skill (Loomba, 1995). The subjects were given an intense reading training for 10 sessions of 60 minute duration. In each of the sessions the skills taken up were, listening comprehension, phonological awareness and word recognition, reading comprehension, and reading fluency. The progression from one activity to the next was done only after attaining 90% accuracy.

Treatment manual included training listening comprehension, phonological awareness and word recognition, reading fluency, reading comprehension. This therapeutic intervention was carried out on a daily basis for 10 days for all the four participants. Before beginning each session a baseline measure of all the skills to be worked upon was obtained, and similar measures were also obtained post session on a daily basis. After 10 sessions of therapy a post therapeutic assessment was carried out by administering Early Reading Skill. The obtained measure was subjected to statistical analysis and the results discussed in the following sections. Scoring and analysis: Baseline and post therapy scores for pre-phonemic, phonemic and fluency measures were obtained. For reading comprehension and listening comprehension scores were obtained however the data was more qualitative with the emphasis on strategies used for each child than the scores. The raw scores were averaged and percentage calculated for all averaged scores. Baseline and post therapy comparison after ten sessions was made for pre-phonemic and phonemic level skills using Wilcoxon Signed Rank test to determine the significance. Graphs were also plotted to show the trend in other domains (fluency, comprehension) with each session. Qualitative analysis was done for all other domains owing to the small sample size.


The present study primarily aimed to develop a treatment manual for children with dyslexia (CWD). Non-parametric test was employed along with Wilcoxon Signed Rank test as the sample size (N=4) was small in the present study. Wilcoxon Signed Rank test was used to compare the baseline and post-therapy scores and to check for its significance. For all the domains baseline and post-therapy scores were obtained and average scores were converted into percentage scores. The skills taken up for speech-language therapy included Listening Comprehension (LC) skills, Phonological Awareness (PA) skills, Reading Skills (RS) & Reading Comprehension (RC) skills. The results are presented and discussed under different sections.

Listening comprehension (LC) skill To work on LC, activities related to literal

comprehension, inferential comprehension and sequencing of main ideas were considered. Baseline and post-therapy scores for the entire domain were calculated and compared across participants. Scores shown under LC represent the scores of literal comprehension sub-skill as these were the skills


214

worked upon for all participants under this domain. Table 1 shows the average scores (in %) for baseline and post-therapy conditions for all the participants across all the domains.

From the Table 1 it is evident that the post therapy scores of all the participants for the LC domain were higher than the baseline scores. For LC domain only qualitative analysis and descriptive statistics were carried out. Inclusion of LC in the treatment for CWD has long been debated. Though no conclusive research reports are available, when considering the simple view of reading (Gough & Tunmer, 1986) verbal comprehension is considered to play a role in facilitating reading comprehension in addition to decoding and word recognition skills, however only general listening comprehension inclusion is supported by this simple view model. Simple view model (Gough et al., 1986) supported the inclusion of listening skills and also emphasized on the interaction between the components of reading namely comprehension and decoding in leading to overall reading skill.

This model has provided evidence for the

multiplicative interaction of decoding and comprehension, thus suggesting that only when both skills are attained reading can occur effectively (Juel, Griffith & Gough, 1986; Stanovich, Nathan & Vala-Rossi, 1986).

Though many research reports support the involvement of speech perception deficit in CWD, (Bradley & Bryant, 1983) unequivocal reports on this notion are still lacking. Strategy used for literal comprehension, namely graphic organizer has been used by many researchers to facilitate reading comprehension by a means of text enhancement (Griffin, Simmons & Kame‟enui, 1991; DiCecco & Gleason, 2002).

However owing to the similarity in reading

comprehension and listening comprehension similar strategies like graphic organizers can be used in LC to facilitate content enhancement and thereby better comprehension. The same strategy was also used for RC domain for all the four participants of the study. Though graphic organizers have been reported to be efficient in facilitating semantic mapping, the extent to which it can be generalized to other contexts and with different materials are questionable (Kim, Vaughn, Wanzek & Wei, 2004).

Phonological awareness (PA) skill

To work on Phonological Awareness (PA), activities related to pre- phonemic awareness skills (PPAS) and phonemic awareness skills (PAS) were considered. Wilcoxon Signed Rank test was administered to compare baseline and post therapy

Table 1. Average scores (in %) of all participants in baseline and post therapy conditions across all the domains

Note: Participant 1 (D1), participant 2 (D2), participant 3 (D3), participant 4 (D4), pre-phonemic awareness skills (PPAS), phonemic awareness skills (PAS),yes/no (Y/N), word matching (WM), odd one out (ODD), rhyme generation (RHY-G), onset-rime blending (ORB), initial phoneme deletion (IPI), phonics (PH), reading fluency (RF), rule „c‟ (Rlc), rule „e‟ (Rle), letter sound (LS), rate (RT), accuracy (ACC), sentence puzzle (SP), story sequencing (SS), summarization (SUM).

Participants

Domains

Sub-sections

Sub-skills

D1 D2 D3 D4

Pre Post Pre Post Pre Post Pre Post

Listening Comprehension (LC)

76.00 84.00 85.00 95.00 77.00 82.00 69.00 79.00

Phonological Awareness

(PA)

PPAS

Y/N 76.00 90.00 91.00 95.00 85.00 91.00 31.00 47.00 WM 80.00 93.00 89.00 96.00 63.00 76.00 26.00 41.00 ODD 71.00 85.00 61.00 71.00 61.00 73.00 30.00 45.00

RHYG 51.50 66.50 56.00 67.00 51.00 66.50 19.50 43.50

PAS ORB 51.00 81.00 59.00 77.00 56.00 70.00 42.00 58.00 IPI 72.18 81.15 77.5 85.62 74.37 84.06 70.00 79.06

Reading Skills (RS)

PH

Rle 62.00 84.00 55.00 77.00 53.00 71.00 49.00 67.00 Rlc 58.00 75.00 67.00 82.00 52.00 70.00 51.00 63.00 LS 72.18 81.15 77.5 85.62 74.37 84.06 70.00 79.06

RF RT 32.00 42.00 49.00 69.00 25.00 33.00 17.00 23.00

ACC 66.00 79.00 54.00 74.00 58.00 68.00 42.00 58.00 Reading comprehension

(RC) SP 66.00 85.00 72.00 87.00 49.00 54.00 44.00 57.00 SS 74.00 89.00 78.00 87.00 63.00 82.00 69.00 74.00


215

scores on PPAS and PAS skills and to check for significance. Table 1 shows results on phonological awareness baseline established before therapy and post therapy scores for PA of all the participants.

As evident from Table 1 scores showed a

general trend of improvement in the post-therapy scores in comparison to baseline scores across all participants. Wilcoxon Signed Rank test results revealed that there was no significant difference in the performance of subjects in baseline and post-therapy scores. There were variations in few of the activities across participants, though overall trend showed higher post-therapy scores.

To summarize the results of this domain,

analysis of results revealed an improvement in post therapy scores following treatment. Baseline and post therapy scores were compared for each participant for all ten sessions on these four tasks under pre-phonological skills using Wilcoxon Signed Rank test which showed no significant difference. In addition to this qualitative analysis was also done which revealed some differences between baseline to post-therapy conditions. Results of the performance of individual participant suggested a similar trend to be evident in all of them across the tasks. The results in this section also revealed an easy acquisition of Y/N by three participants (D1, D2, & D3). WM was also found to be relatively easy for participants in comparison to other PAS tasks i.e. ODD and RHY-G. D4 also showed a similar trend in the scores irrespective of its poor baseline scores. In all the subjects, irrespective of the baseline levels an improvement in the post-therapy scores was observed.

The reason for no significant difference across

baseline and post-therapy conditions in the tasks PPAS and PAS may be due to the lesser number of sessions taken up. Though no conclusive reports on the total duration and dosage of PA therapy is available in literature owing to the fewer studies on efficacy of PA, National Institute of Child Health and Human Development (NICHD, 2000a) has suggested 5 to 18 hours of instruction to bring about considerable improvement in phonological awareness skill, however this cannot be generalized because this criterion was not established taking the heterogeneity of the reading performances of children into consideration. Bailet, Repper, Piasta and Murphy (2009) reported an intense instruction on PA of 30 minute long session for 9 weeks to show some improvement on overall literacy skills. Considering the durations specified in the literature the duration of PA therapy provided in the

current study (15 minutes approx for 10 sessions) is suggestive to be insufficient (Torgesen, Wagner & Rashotte, 1999).

Inclusion of PA in CWD has been emphasized

by researchers since ages together and is not a new finding; however with respect to older children deep phonemic awareness or PAS in the present study are considered to be of more significance in bringing change in the reading skills. PPAS only facilitates the development of PA and has no direct contribution to the acquisition of reading skills. It is PA which helps children to develop the alphabetic principle, the basic concept of phonemes of the language etc. Hence with respect to older children PAS is more concentrated in therapy than PPAS. Effectiveness of phonemic awareness intervention in older children with dyslexia and specifically for bilingual children having English as their second language has been emphasized by the findings of Swanson, Hodson and Aikins (2005), in their study on Spanish-English older bilingual children (7th grade). This would support the inclusion of these skills in the intervention of the participants of the present study (5th and 6th grade). Though not significantly deviant, the improvement in IPI of PAS was found to be lesser in comparison to ORB, this was because the scores of children went down in IPI task as most of them produced initial syllables instead of phonemes which was scored as incorrect and hence the overall improvement came down. Participants of the study tended to produce syllables in IPI task owing to some amount of cross-linguistic influence which needs to be addressed in the intervention of these children. Indian languages are syllabic and hence children find syllabification much easier. English in turn is considered to be a phonemic language, with phonemes forming the minimal units. Thus when dealing with ELL as in the case of Indian children, intervention should incorporate all the basic principles of the language since the language for ELL would be learnt explicitly unlike native speakers.

Reading skills

To work on Reading Skills (RS), activities for improving phonics (PH), word recognition (WR) and reading fluency (RF) were considered. As shown in Table 1 phonics (PH) showed improvement in general following therapy in all the participants. From the table it is evident that the changes in scores were noticeable enough in all the participants again irrespective of their baseline scores. The general trend was thus the same across the participants for this sub-section. A similar trend was also observed in RF (in RT and ACC). However the improvement in ACC was found to be less in all participants in comparison to RT which


216

showed better changes post treatment for participants. Word recognition was also worked upon however the scores were not taken up for analysis owing to the redundancy of the material as the same words were scored for accuracy under RF. The words used in this section were from the passage taken up for reading fluency, hence word recognition scores and ACC were almost very similar and hence the data was not taken up for analysis.

Comparison of the performance of all the

participants on each task of PH revealed a similar pattern of improvement on all tasks (Rle, Rlc and LS) by all participants except for participant D1, where a slight disparity in the improvement of Rlc and Rle was observed following therapy. Above results suggest that with respect to phonics the improvement was minimal in the rules. All participants experienced difficulty understanding the phonic rules which can be established only over a longer period of time. The performance in letter sound identification showed improvements across all subjects owing to its less complexity in learning. Also the letters taken up initially were easy ones such as „p‟, „t‟ etc. than „c‟ or „g‟ which have more than one sound depending upon the context.

RF measures as mentioned earlier had RT and

ACC measures which were scores of which were compared in baseline versus post therapy condition. In general there was improvement in RT though not in all participants and improvement in ACC was common for all participants following therapy. The general trend showed an increase in RT though not consistent with all. Another observation with ACC scores revealed a similar baseline scores in ACC across participants.

Comparison of performance of all participants

across each task of RF reveals an improvement in RT in participants D1 and D2 from baseline to post therapy condition and a not so significant change in RT of participants D3 and D4 from baseline to post therapy condition as evident from Table 1. In RT maximal improvement was seen in participant D2 as mentioned earlier in individual performance. With respect to ACC improvement was observed in all participants from baseline to post therapy scores. When comparing RT and ACC scores for each participant disparity was observed in the scores of participant D2, with higher RT in comparison to ACC baseline scores, with higher RT scores to ACC baseline scores. Similarly the scores of participant D4 showed a relatively lower baseline RT measures in comparison to ACC baseline.

Thus in this domain a general improvement in phonics was observed across all participants though acquiring the rules was observed to be little difficult. LS task was relatively easier for the participants as only consonant sounds were only taken up and vowels and letters having more than one sound such as „c‟, „g‟ were not taken up. Though the concept of letter sounds was easily acquired by children, the scores were relatively good in this section. In RF, RT and ACC measures also followed the general trend with higher post scores following therapy. However RT did not show significant improvement in participants D3 and D4. ACC showed to have improved for all participants. In general performance of participant D2 was higher in comparison to other participants and especially noticeable in progress in RT. Inclusion of phonics in reading instruction is supported by many studies incorporating emergent literacy skills. Phonics has been suggested to be useful for kindergarten and first grade students to associate the symbol to sound correspondence (Schneider, Kuspert, Roth, Vise & Marx, 1997). Phonics was used with these children since they were learners of English and providing explicit instruction on phonics would be beneficial to them in developing overall better performance in word recognition.

Strategies used for each participant was

different with respect to their baseline reading rate and accuracy measures. Repeated reading with modeling and feedback was used uniformly across all participants as this has been the most effective strategy as reported in literature till date (National Reading Panel, 2000).

To work on reading comprehension (RC),

activities such as sentence puzzle (SP), story sequencing (SS) and summarization (SUM) were taken up. General trend was similar to other domains with an improvement in post therapy scores in comparison to baseline scores across all tasks in all participants. SUM was worked upon only for participants D1 and D2 as scores of D3 and D4 were below criterion for earlier tasks. When considering comparison of scores across tasks for all participants, SS was found to be relatively easier for all participants with a better score. This may be because the SS scores (in %) were obtained by combining the scores on sequencing of story cards and also by asking child to sequence story using graphic organizers. Scores on SP were slightly lower as children found difficulty putting the chunked sentences in order especially when the sentences were complex. SUM on the other hand was taken up for only participants D1 and D2 and results do not show much


217

improvement owing to lesser number of sessions for which this task was taken up.

Overall in this section on RC, improvement was observed in the tasks on SP and SS for all subjects. And SUM was taken up for participants D1 and D2 but only for few sessions. The scores on SS were higher as it was a combined score of sequencing cards and saying out the sequence where sequencing cards was performed well by all participants. Reading comprehension began with sentence puzzle (Scott, 2009). As sentence level comprehension was found to have a positive influence on reading comprehension. Thus for all the participants this skill was worked upon initially as there are existing models and research reports supporting the breakdown at the sentence level resulting in reading comprehension deficits. All the participants were found to have problems at sentence level processing especially with complex sentences, passive forms and for those featuring distance dependencies. However there also exist reports in literature reporting no significant positive outcomes from explicit teaching of sentence complexity. SUM was taught for participants D1 and D2 by familiarizing with „self cue‟ cards which also served as self monitoring strategy. Summarization is one strategy which has reported to be beneficial for CWD in literature, especially when supported with cueing cards assisting self monitoring. Though comprehensive reports on the improvement in reading comprehension is available, with such limited duration as in the present study no comments can be made on the efficacy of the approaches used. Curriculum based text was used as the material for both the participants for better understanding and to assure relevancy of the material and also as supported by narrow view of reading. Summarization was initially carried out in native language as the comprehension of the material is the focus in this strategy. Such an approach making use of native language strategically can be used with English Language Learners (ELL) as in Indian setting.

Conclusions

In the present study a treatment manual in English for Indian children with dyslexia was developed, incorporating all the important domains crucial for the acquisition of fluent reading. This would serve as an important tool for the clinicians in clinical setting when working with children with learning disability. The sensitivity of the present manual needs to be tested across many subjects.

The treatment strategy followed in the present study was intense with explicit instructions. Such an approach to reading instruction has been highlighted by

Response to Intervention (RTI) which is advocated by explicit and intense instruction. RTI has been the focus of interest for assessment and intervention in the recent years. Recently this RTI model has gained immense attention in determining the eligibility of having a reading disability apart from the IQ discrepancy criteria which was earlier considered a main criterion for enrolling any child for special reading instruction. As suggested in the literature effective reading instruction incorporates use of the following evidence-based skills, known to promote successful beginning literacy: (a) awareness of and ability to manipulate phonemes in segmenting and blending strategies (b) awareness and understanding of letter-sound correspondence (Byrne & Fielding-Barnsley, 1989; Foorman et al., 1998; Abbott, Walton & Greenwood, 2001) (c) the ability to translate the speech stream sound structures of oral language (phonological processing) into written language (Adams, 1990) and (d) fluency in decoding words and understanding word meaning (Adams, 1990).

References

Abbot, M., Walton, C., & Greenwood, C. R. (2001). Phonemic awareness in kindergarten and first grade. Teaching Exceptional Children, 34(4), 20-26.

Adams, M. J. (1990). Beginning to read: Thinking and learning about print. Cambridge, MA: MIT Press.

Bailet, L. L., Repper, K. K., Piasta, S. B., & Murphy, S. P. (2009). Emergent literacy intervention for prekindergarteners at risk for reading failure. Journal of Learning Disability, 42, 336-355.

Bradley, L., & Bryant, P. E. (1983). Categorizing sounds and learning to read: A causal connection. Nature, 301, 419-421.

Brown, A. L., Palincsar, A. S., & Purcell, L. (1986). Poor readers: Teach, don't label. In U. Neisser (Eds.), The school achievement of minority children: New perspectives (pp. 105-143). Hillsdale, NJ: Erlbaum.

Byrne, B., & Fielding-Barnsley, R. (1989). Phonemic awareness and letter knowledge in the child's acquisitions of the alphabetic principle. Journal of Educational Psychology, 80(1), 313-321.

Carver, R. P. (1992). Reading rate: Theory, research, and practical implications. Journal of Reading, 32, 84-95.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd Ed). Hillsdale, NJ: Eribaum.

Cunningham, P. (1999). What should we do about phonics? In L. Gambrell, L. Morrow, S. Neuman, & M. Pressley (Eds.), Best practices in literacy instruction (pp. 68–89). New York: Guilford.

Daly, E. J., & Martens, B. K. (1994). A comparison of three interventions for increasing oral reading performance: Application of the instructional hierarchy. Journal of Applied Behavior Analysis, 27, 459–469.


218

DiCecco, V. M., & Gleason, M. M. (2002). Using graphic organizers to attain relational knowledge from expository text. Journal of Learning Disabilities, 34(4), 306-321.

Englert, C. S., Tarrant, K. L., Mariage, T. V., & Oxer, T. (1994). Lesson talk as the work of reading groups. The effectiveness of two interventions. Journal of Learning Disabilities, 27, 165–185.

Foorman, B. R., Francis, D. J., Fletcher, J. M., Schatschneider, C., & Mehta, P. (1998). The role of instruction in learning to read: Preventing reading failure in at-risk children. Journal of Educational Psychology, 90, 37-55.

Gersten, R., & Baker, S. (2001). Teaching expressive writing to students with learning disabilities. Elementary School Journal, 101(3), 251-272.

Gilbert, L. M., Williams, R. L., & McLaughlin, T. F. (1996). Use of assisted reading to increase correct reading rates and decrease error rates of students with learning disabilities. Journal of Applied Behavior Analysis, 29, 255–257.

Gough, P., & Tunmer, W. (1986). Decoding, reading, and reading disability. Remedial and Special Education, 7, 6-10.

Graesser, A. C., Singer, M., & Trabasso, T. (1994). Constructing inference during narrative text comprehension. Psychological Review, 101 (3), 371-395.

Griffin, C. C., Simmons, D. C., & Kame‟enui, E. J. (1991). Investigating the effectiveness of graphic organizer instruction on the comprehension and recall of science content by students with learning disabilities. Journal of Reading, Writing, and Learning Disabilities International, 7, 355–376.

Iverson, S., Tunmer, W. E., & Chapman, J. W. (2005). The effects of varying group size on the reading recovery approach to preventative early intervention. Journal of Learning Disabilities, 38, 456–472.

Juel, C. (1988). Learning to read and write: A longitudinal study of 54 children from first through fourth grades. Journal of Educational Psychology, 80 (4), 443–47.

Kim, A., Vaughn, S., Wanzek, J., & Wei, S. (2004). Graphic organizers and their effects on reading comprehension of students with LD: A synthesis of research. Journal of Learning Disabilities, 37, 105-118.

Kucan, L., & Beck, I. L. (1997). Thinking aloud and reading comprehension research: Inquiry, instruction, and social interaction. Review of Educational Research, 67(3), 271-299.

LaBerge, D., & Samuels, S. J. (1974). Toward a theory of automatic information processing in reading. Cognitive Psychology, 6, 293–323.

Loomba, M. (1995). Descriptive analysis of the sequential progression of English reading skills among Indian children. University of Mysore Retrieved May 22, 2010, from http:// aiishdigilib. in: 8080/ digitallibrary/ HomeAdvanveResult. do

Marston, D., Deno, S. L., Dongil, K., Diment, K., & Rogers, D. (1995). Comparison of reading intervention approaches for students with mild disabilities. Exceptional Children, 62, 20-38.

Mattingly, I. G. (1972). Reading, the linguistic process, and linguistic awareness. In J. F. Kavanagh & I. G. Mattingly

(Eds.), Language by ear and by eye: The relationships between speech and reading. Cambridge, MA: MIT Press.

Mattingly, I. G. (1984). Reading, linguistic awareness, and language acquisition. In J. Downing & R. Valtin (Eds.), Language awareness and learning to read. New York: Springer-Verlag.

McKoon, G., & Ratcliff, R. (1992). Inference during reading. Psychological Review, 99 (3), 440-466.

Moseley, F. (Edn.) (1993). Marx‟s method in “Capital”: a reexamination. Atlantic highlands, N. J., Humanities Press.

Nathan, R. G., &. Stanovich, K. E. (1991). The causes and consequences of differences in reading fluency. Theory into Practice, 30, 176–184.

National Reading Panel. (2000, April). Teaching children to read: An evidence-based assessment of the scientific research literature on reading and its implications for reading instruction. (NIH Pub. No. 00-4769). Retrieved July 27, 2001, from http:// www. nichd. nih. gov/ publications/nrp/smallbook.htm.

National Institute of Child Health and Human Development. (2000a). Report of the National Reading Panel. Teaching children to read: An evidence-based assessment of the scientific research literature on reading and its implications for reading instruction (NIH Publication No. 00-4769). Washington, DC: Government Printing Office.

Perfetti, C. A. (1999). Comprehending written language: A blueprint of the reader. In C. Brown & P. Hagoort (Eds.), The neurocognition of language (pp. 167-208). New York: Oxford University Press.

Perfetti, C. A. (2007). Reading ability: Lexical quality to comprehension. Scientific Studies of Reading, 11(4), 357-383.

Rashotte, C. A., & Torgesen, J. K. (1985). Repeated reading and reading fluency in learning disabled children. Reading Research Quarterly, 20, 2, 180-188.

Rose, T. L. (1984). The effects of two prepractice procedures on oral reading. Journal of Learning Disabilities, 17, 544–548.

Schneider, W., Kuspert, P., Roth, E., Vise, M., & Marx, H. (1997). Short- and long-term effects of training phonological awareness in kindergarten: Evidence from two German studies. Journal of Experimental Child Psychology, 66, 311 – 340.

Scott, C. M. (2009). A case for the sentence in reading comprehension. Language, Speech and Hearing Services in School, 40, 184-191.

Shilpashree, H. N. (2004). Children with reading disability: A remedial manual on meta phonological skills. Unpublished Master‟s dissertation, University of Mysore.

Swain, K. D., & Allinder, R. M. (1996). The effects of repeated reading on two types of CBM: Computer maze and oral reading with second-grade Students with learning disabilities. Diagnostique, 21(2), 51-66.

Swanson, T, J., Hodson, B, W., & Aikins, M. S. (2005). An examination of phonological awareness treatment outcomes for seventh-grade poor readers from a bilingual community. Language Speech and Hearing Services in Schools, 36, 336-345.


219

Torgesen, J. K. (1998). Individual differences in response to reading intervention. Paper presented at the Pacific Coast Research Conference, LaJolla, CA.

Torgesen, J. K., Rashotte, C. A., & Alexander, A. W. (1997). Principles of fluency instruction in reading: Relationships with established empirical outcomes. In M. Wolf (Eds.), Dyslexia, fluency and the brain (pp. 332–355). Timonium, MD: York Press.

Vadasy, P. F., Jenkins, J. R., & Pool, K. (2000). Effects of tutoring in phonological and early reading skills on students at risk for reading disabilities. Journal of Learning Disabilities, 33(6), 579-590.

Vaughn, S., Chard, D. J., Bryant, P., Coleman, M., Tyler, B., Thompson, S. L., & Kouzekanani K. (2000). Fluency and comprehension interventions for third-grade students: Two paths to improved fluency. Remedial and Special Education, 2, 325-335.

Vellutino, F. R., Scanlon, D. M., Sipay, E. R., Small, S. G., Chen, R., Pratt, A., & Denckla, M. B. (1996). Cognitive profiles of difficult to remediate and readily remediated poor readers: Early intervention as a vehicle for distinguishing between cognitive and experiential deficits as basic causes of specific reading disability. Journal of Educational Psychology, 88, 601-638.


220

Developing a Protocol for Measuring Participation of Persons with Aphasia Rupali Mathur 1 & S. P. Goswami 2

Abstract

For persons with aphasia communication (activity & participation) and context (personal & environmental factors) are very important. Hence, the current investigation was aimed to develop a protocol for measuring participation in life situations in person with aphasia and administrater of this protocol on persons with aphasia and their caregivers. Development of a protocol for participation of person with aphasia was carried out in two stages. First the protocol was developed. which consisted of 4 domains i.e. understanding and expression, general tasks and demands, interpersonal interactions and relationships and facilitator and barriers. Further, it was administered on a total of twenty persons with aphasia and their family member/ caregiver. The results indicated that participation in terms of various contexts was greatest for persons with aphasia demonstrating mild communicative deficits and less associated problems. Across each domain, there was significant correlation among the responses of persons with aphasia and their caregivers. Duration of therapy played a significant role in enhancing participation of persons with aphasia across various contexts. Comprehension and expression ability was reported to be average among persons with aphasia. 40-50 % of the time persons with aphasia required help to understand and express the conversation. As result of stroke persons with aphasia loose almost 50-60% of the ability to carry out activities of daily living depending on the type of aphasia. At work place only 40% of the persons with aphasia could return. Majority of caregivers who supported persons with aphasia in various conditions were women. Thus, the results of the study do provide a conclusive base that as in persons with aphasia the activity and participation is affected in various situations which varies from the activities of daily living to the vocation. Hence, it is very imperative for speech-language pathologists to measure participation of persons with aphasia which in turn, will help in planning treatment. Key words: protocol, participation of persons with Aphasia

phasia is 1characterized by a wide array of symptoms and characteristics which varies depending on the site of lesion, extent of

lesion and associated problems. There have been numerous attempts to explain aphasia. However, no one definition has succeeded in explaining the exact nature, characteristics and consequences of this condition. Aphasia can influence one or more elements of communication such as speech, language or gesture. These deficits impact upon the expression and/or comprehension of language and cause remarkable alterations in day to day functioning of persons with aphasia. Aphasia can diminish participation of persons with aphasia across several real life contexts such as activities of daily living, at home, in social situations, in academics, and at work place.

The major emphasis in the literature is towards linguistic ability and associated aspects of persons with aphasia. However, in real life situations various other issues also play a vital role. For persons with aphasia and their family members participation across various situations is more important than knowing names of few lexical categories or repetition etc. (Le Dorze, Julian, Brassard, Durocher & Boivin, 1994; Le Dorze & Brassard, 1995). Following are few aspects which are noteworthy and needs to be kept in mind while working with persons with aphasia.

Accomplishment of several activities of daily

living for instance bathing, eating, cleaning, shopping, travelling, watching TV, listening to

1e-mail:[email protected]; 2Reader in Speech Pathology, AIISH, Mysore, [email protected]

music, reading newspaper, writing letters, using computers, calculation, managing finance are more imperative for persons with aphasia. Meanwhile, most of these activities also involve some or the other aspect of communication. Aphasia also affects domestic life of persons with aphasia (Hilari, Smith, 1985; Wade, Hewer, David & Enderby, 1986; Wiggins, Byng & Smith, 2003).

Interpersonal interactions and social life

include relationship with family members, relatives, friends, participating in various social events like festivals, religious activities, parties etc. Interpersonal interactions and relations are reported to be altered in persons with aphasia (Smith, 1985).

Work integration or returning to work place is

a very essential issue in the life of persons with aphasia. Stroke hampers ability of an individual to return back to work (Le Dorze & Brassard, 1995; Salonen, 1995; Parr, 2001). There are more chances of change of job or reduction in timing etc. This can severely affect life of persons with aphasia. This reduces their confidence and can lead to depression.

The factors which help persons with aphasia

to participate in various contexts act as a facilitator. Knowledge about them can help enhance communication of persons with aphasia apart from therapy. Another important issue is who provides support to persons with aphasia either family members or professionals or friends etc. Knowledge about barriers experienced by persons with aphasia is also necessary. If we know what are hurdles are faced by them, various steps can be taken to remove or reduce them to further enhance participation of persons with aphasia.

A

Protocol for measuring Participation of Persons with Aphasia

221

Even though above described elements are more relevant to life of persons with aphasia, very few attempts have been made to assess these aspects. In spite of the abundance of available measures, existing appraisal of aphasia may be inadequate. To focus socially relevant and important treatment, a clinician must also evaluate the impact of residual deficits on a person‟s participation in life situations. Available functional communication measures, for example: Communicative Abilities in Daily Living (CADL Holland, 1980), Functional Communication Profile-(FCP Sarno, 1969), American Speech Language hearing Association Functional Assessment of Communication Skills for Adults(ASHA FACS,Frattali, Thompson, Holland, Wohl & Ferketic, 1995) may or may not measure aspects of actual life participation (Simmons-Mackei, Threats & Kagan, 2005).

Further research is required to fill the empty

space in the literature regarding participation. Such research should be guided by a clear conceptual framework and well-defined concepts. These could possibly be derived from existing frameworks like:

The International Classification of Functioning, Disability and Health-ICF (World Health Organization, 2001): This framework conceptualizes health and the consequences of disease (World Health Organization, 2001). It is a valuable tool in research into disability, in all its dimensions that is impairments at the body and body part level, person level activity limitations, and societal level restrictions of participation. It is a biopsychosocial model. Disability Creation Process model-DCP (Fougeyrollas, St-Michel, Bergeron, Cloutier & Cote, 1999): DCP is a theoretical model. It explains and focuses specifically on the social construction of disability and impact on participation or „„life habits‟‟ (Noreau, Fougeyrollas & Vincent, 2002). It comprises of four components, which are: risk factors (cause), personal factors (organic system and capabilities), environmental factors and life habits. There is dynamic interaction among factors of this model. Living with Aphasia: Framework for Outcome Measurement-AFROM (Kagan, Simmons-Mackie, Rowland, Huijbregts, Shumway, McEwen, Threats, & Sharp, 2008) ICF classification is very broad while A-FROM tries to simplify it by considering the elements which are more relevant for aphasia. It includes following domains: Aphasia severity (correlate of ICF body function/impairment), participation/life habits, personal factors including identity and emotions and environment.

In Indian context very less work has been done in this area. A study by Jeslma (2009) showed that only six studies have been published related to ICF from developing countries, with Germany and the USA responsible for almost 50% of the total. Thus, there is a need, to strengthen the existing research and to study from Indian per se.

Communication (functioning, activity, participation) and context (personal and environmental factors) are equally important, as well as interdependent, for the person with aphasia. There is a necessity to develop measures that appraise ability of persons with aphasia to participate in desired life situation (Ross & Wertz, 2003). Very few measures exist for the purpose of appraising ability of person with aphasia to become or remain involved in desired life situations.

The study was planned to develop a protocol

for measuring participation in life situation in person with aphasia and to administer this assessment protocol on person with aphasia and on their family member/ caretaker and thereby gain an understanding towards ability of persons with aphasia to participate in life situations.

Method

The main aim of the current investigation was to develop a protocol for measuring participation in real life situation in person with aphasia and to administer this on person with aphasia and on their family member/ caretaker. Development of the protocol was done in two stages.

Stage 1: Development of the protocol Protocol for measuring activity and participation was developed based on the principles of ICF-WHO, DCP and AFROM. Protocol consisted of following four sections: A. Understanding and Expression: This section

has questions regarding their comprehension and expression of verbal, nonverbal and figurative language.

B. General Tasks and Demands: It comprised of questions related to ability of persons with aphasia to perform different daily activities.

C. Interpersonal Interactions and Relationships (participation in society): This included their performance at their work place, academic, social situations, and home.

D. Facilitator and Barriers: It incorporated questions concerning the factors which acts as a facilitator and barriers to participate in various contexts. Section A, B, and C also had questions

regarding whether they require help, their ability is same as before stroke and their satisfaction level. Each section had questions which were supposed to be rated on a five point rating scale. Following rating


222

scale was used: 1- Able to do less than 25% of the time, 2- Able to do 25-50% of the time, 3- Able to do 50-75% of the time, 4- Able to do 75-90% of the time and 5- Able to do more then 90 % of the time

These questions were given to ten judges. Speech- Language pathologists (SLPs) served as the judges. They were asked to rate the questions on a Feedback questionnaire for aphasia management manual (Field testing of Manual for Adult Non-fluent Aphasia therapy manual- MANAT-K, Goswami, Shanbal, Samasthitha & Navitha, 2010). Few parameters from the feedback questionnaire were removed as they were not relevant such as size of the picture, color and appearance, arrangement iconicity, trainability. Judges were also asked to give suggestions regarding the rating scale used and any other changes in the protocol.

Further, pictures were incorporated to capture

all the concepts and to make it easier to understand for persons with aphasia. A professional artist made all the pictures. Further, these pictures were subjected to their ambiguity and familiarity validation by five judges. Judges were experienced SLPs. They were asked to rate pictures on their familiarity and ambiguity. Items with 90% agreement were included in the assessment tool. Based on suggestions of judges modifications were made. Stage 2: Administration of this protocol on persons with aphasia and their family members Participants: A total of 20 persons with aphasia along with their caregivers (20) participated in the study. Persons with aphasia were identified through hospitals, neurological clinics and/ or speech and hearing centers. They were diagnosed using Western Aphasia battery (WAB, Kertesz, 1982) by experienced speech language pathologist. The age range of persons with aphasia was from 28 to 80 years. Among the aphasia group various types of aphasia were represented (six global aphasia, four Broca‟s aphasia, two transcortical motor aphasia, one Wernicke aphasia, one transcortical sensory, three anomic aphasia, two subcortical aphasia, and one progressive non fluent aphasia).

Participants were selected by adhering to the

appropriate ethical procedures. Participants and caregivers were explained the aim and procedures of the study, and an informal verbal and/ or written consent were obtained. Participants were randomly selected based on the inclusionary criteria. Inclusion Criteria for person with aphasia: Persons with aphasia following stroke or other left hemisphere damage were included. Those with no known history of pre-morbid neurological illness, psychiatric disorders and/or cognitive decline, and no

other significant sensory and/or cognitive deficits that could interfere with the individual‟s performance in the investigation were also selected. Procedure: The protocol was administered on persons with aphasia and their family members. They were instructed to rate it as per their ability to perform different tasks on a rating scale ranging from 1 to 5. The responses included were either pointing or verbal responses. Their responses were recorded on a scoring sheet. Testing time varied from 20-30 minutes. Further these responses were subjected to analysis using SPSS software (version 16.0).


The results of the current investigation are being presented and discussed under the following sections:

A. Understanding and expression: This domain evaluated the comprehension and expression (verbal, nonverbal and figurative language) in persons with aphasia. Furthermore, it assessed whether they require help, performance as compared to previous and the satisfaction level.

The results showed the following trends. Comprehension was reported to be good among almost all types of aphasia by persons with aphasia and their caregivers. The reason behind this can be as it was self rated so there were more chances of bias and other reason could be that caregivers tend to overestimate the ability of person with aphasia as they were more interested in participation ability of person with aphasia rather than their linguistic capacity. Expression ability was reported to be average among persons with aphasia. 40-50% of the time persons with aphasia required help to understand and express the conversation. These results indicate that aphasia can have a profound impact on ability of person with aphasia to participate in conversations.

Comprehension and expression was greatest

for persons with aphasia demonstrating mild communicative deficits and less associated problems.

The results are being strengthened by the findings reported by Peuser and Schriefers (1980). Various other investigators have also reported similar findings (Milberg & Blumstein, 1981; Semenza & Goodglass, 1985; Goswami, 2004). Hence, results of current investigation are in accordance with the literature. The findings indicated that linguistic ability, language recovery, pre morbid linguistic ability, associated problem, duration of therapy attended, age of onset of problem were the few factors contributing to change in their ability to comprehend and express with respect to before stroke. These factors can affect ability of person with aphasia to comprehend and express to a large extent.


223

Therefore, these factors should be kept in mind while considering the participation of a person with aphasia.

There was significant correlation noticed

among the responses of persons with aphasia and their caregivers. The results are in accordance with the view of Davis and Mehan (1988), where they have reported better results in understanding as reported by the caregivers compared to the scores shown by the rehabilitation workers. Indeed, many researchers advocate that families should be included in aphasia therapy and that they should also be given counseling and support (Wahrborg & Borenstein, 1989; Denman, 1998; Le Dorze, Croteau, Brassard & Michallet, 1999; Chapey, Duchan, Elman, Garcia, Kagan & Lyon, 2001; Pound, Parr & Duchan, 2001). Hence, findings of current investigation are in harmony with above mentioned studies.

Duration of therapy played a significant role

in enhancing participation of persons with aphasia across various contexts. Various studies have reported that if therapy duration was more than it was found to be more effective (Hagen, 1973; Mazzoni, Vista, Geri, Avila, Bianchi & Moretti, 1995). B. General Tasks and Demands: This domain assessed questions concerning ability of person with aphasia to perform a variety of daily activities. Like previous section it also assessed whether they require help, performance as compared to previous and their satisfaction.

As a result of stroke persons with aphasia

loose almost 50-60% of the ability to carry out activities of daily living depending on the type of aphasia. Reduced linguistic, sensory and motor ability following stroke can be the reason behind reduction in performance across tasks. Various researchers have reported that aphasia can have a profound impact on person with aphasia everyday activities and communication (Bullinger, Anderson, Cella & Aaronson, 1993).

Almost half of the time persons with aphasia

required help to execute activities of daily living. Ability of person with aphasia to fully participate in activities of daily living and their satisfaction level could be affected by interaction of many factors together such as severity of deficits, associated problems, support from family and professionals etc. Language deficits and other associated problems (paralysis, sensory deficits) can be the major reason which hampers their ability to participate in various daily life contexts. Due to which support from others is important factor while performing everyday life activities (Bullinger, Anderson, Cella & Aaronson, 1993; Worrall & Hickson, 2003; Hilari & Northcott, 2006).

Almost similar responses were obtained for persons with aphasia and their caregivers with respect to their performance to carry out everyday activities, need for help, performance as compared to previous and their satisfaction. These findings indicate a greater understanding of caregivers towards the ability of persons with aphasia to execute various general task and demands (Lubinski, Duchan & Weitzner-Lin, 1980; Ferguson, 1992, 1994; Milroy & Perkins, 1992; Goodwin, 1995; Laasko, 1997, 2002; Laasko & Klippi, 1999; Lindsay & Wilkinson, 1999; Oelschlaeger, 1999; Oelshlaeger & Damico, 2002; Perkins, 2002). Thus, results are being supported by above mentioned studies.

Results also revealed that along with

caregivers‟ assistance, speech and language therapy helps persons with aphasia to fully participate and perform daily activities almost similar to before stroke. This depicts role of therapy in recovery from the aphasia and can lead to better participation in social context (Hagen, 1973; Marshal, Pound, White-Thomson & Pring, 1990; Davis & Pring, 1991; Byng, 1993; Mazzoni, et al., 1995; Nickels & Best, 1996). Thus, therapy is very important factor while dealing with persons with aphasia which enhances their activity and participation in various social domains. C. Interpersonal Interactions and Relationships (participation in society): This domain assessed their performance at their work place, academics, social situations, and home. Like previous domain it also assesses whether they need help, performance in contrast to previous and their satisfaction in various situations.

At work place only 40% of the persons with

aphasia could return. However, they could not regain full employment stage same as before stroke. Ability of persons with aphasia to perform well at work place is affected to a large extent following factors like working requirements and support from other employee, linguistic and motor ability. These findings are also supported by various other researchers i.e. return to work is often characterized by reduced hours, return to another job, or return to the same job with modifications (Carriero, Faglia & Vignolo, 1987; Black-Schaffer & Osberg, 1990; Sarno, 1992; Hinckley, 2002; Parr, 2001). Often there was no return at all (Le Dorze & Brassard, 1995; Garcia, Barrette & Chantal, 2000; Parr, 2001; Hinckley, 2002; Hilari, Wiggins, Byng & Smith, 2003). Persons with aphasia are less satisfied with their performance at work place. This could be attributed to their reduced ability to perform the task assigned to them at work place, reduction in working hour or change of job/vocation due to occurrence of stroke. Here, also severity of aphasia played a significant role to predict the performance of persons with aphasia in vocation/ employment.


224

Ability of persons with aphasia to participate in social situations and communication at home was reduced by 50% after the onset of the stroke. There are reports of change in interaction of persons with aphasia with friends and relatives. However, many factors tend to influence participation ability of persons with aphasia such as communicative ability, support from family and professionals. Communication difficulty may result in social isolation, challenges with interpersonal relationships, mental and emotional changes, and corresponding lack of independence (Cruice, Worrall, Hickson & Murison, 2003; Ross & Wertz, 2003; Worrall & Holland, 2003). This can further restrict participation of persons with aphasia social contexts.

Persons with severe communication deficits

were less satisfied with their communication at various contexts. These findings could be attributed to their inability to communicate in various situations as reported by Cruice, Worrall, Hickson and Murison (2003); Hilari and Byng (2009).

There was significant correlation among the

responses of persons with aphasia and their caregivers. However, only for one question i.e. whether performance at work place is same as before stroke (C2), there was less correlation. This indicates that caregivers need to understand persons with aphasia in various aspects of employment. Therapy had a positive effect on ability of persons with aphasia to participate in society. Research by various other authors reported the similar findings (Hinckley & Packard, 2001; Hinckley, 2002).

D. Facilitators and Barriers

This included questions which inform us about the factors which acts as a facilitator and barriers for persons with aphasia to participate in various daily situation. It also included questions which inform us about who help persons with aphasia to participate in various daily situations such as at home, in social situations, at work place, in academics etc.

Majority of caregivers who supported persons

with aphasia in various conditions were women. These findings receivie support from studies conducted by Hodson, Wood and Langton, (1996) and Swati (2008), who reported majority of women and spouse are caregivers of persons with aphasia.

Almost 60% of the factors helped persons

with aphasia to participate in various contexts. However, for person with aphasia having severe communication deficits, fewer factors facilitated their participation in turn leading to participation restriction. Speech and language therapy enhanced their participation to some extent, however the trend obtained was not much conclusive.

40% of the problems are faced by persons with aphasia. This can hamper their participation in daily situations. A study conducted by Davidson, Worrall and Hickson (2003) also reported that persons with aphasia generally engage in communicative activities less frequently than others. Thus results of the present study obtain support from above mentioned studies. Therapy tends to show a positive impact on their participation ability. However, the trend obtained is not much conclusive.

Conclusions

Hence, to conclude this protocol can be administered on persons with aphasia to measure their activity and participation in real life situation. Poor scores i.e. rating of 1 or 2 signifies restricted participation of person with aphasia, on the other hand score of 4 or 5 signifies good participation in various questions. Domains of this protocol are more relevant for persons with aphasia.The areas of participation where participation is restricted can be taken as intervention goals. This can also help in planning the intervention goals which are more relevant for persons with aphasia.This protocol can be used in assessment to determine severity of participation restriction. It can aid in research regarding the effectiveness of intervention. As it is aphasia friendly version it provides a format for communicating with person with aphasia about goals and expected outcomes. Person with aphasia can have an opportunity to determine and choose what outcomes are relevant from his/ her perspective.

It can also be helpful in deciding which

factors are facilitating participation in persons with aphasia and which factors create barriers which hinder participation in persons with aphasia. Knowledge regarding facilitators and barriers can in turn help to enhance participation of persons with aphasia. This can also help in generating a barrier free environment for person with aphasia. Application need not be limited to the assessment of person with aphasia. With slight adaptation this protocol would be appropriate for measuring participation restrictions associated with other communication disabilities.

References

Black-Schaffer, R. M., & Osberg, J. S. (1990). Return-to-work after stroke: Development of a predictive model. Archives of Physical Medicine and Rehabilitation, 71, 285-290.

Bullinger, M., Anderson, R., Cella, D., & Aaronson, N. K. (1993). Developing and evaluating cross cultural instruments: From minimum requirements to optimal models. Quality of Life Research, 2, 451-459.

Byng, S. (1993). Hypothesis testing and aphasia therapy. In A. Holland & M. Forbes (Eds.) Aphasia treatment:


225

World perspectives (pp. 115-130). San Diego: Singular Publishing Group.

Carriero, M., Fagila, Z., & Vignolo, L. A. (1987). Resumption of gainful employment in aphasics: Preliminary findings. Cortex, 26, 667-672.

Chapey, R., Duchan, J. F., Elman, R. J., Garcia, L. J., Kagan, A., & Lyon, J. G. (2001). Life participation approach to aphasia: A statement of values for the future. In R. Chapey (Eds.), Language intervention strategies in aphasia and related neurogenic communication disorders (pp. 235-245). Baltimore: Lippincott Williams & Wilkins.

Cruice, M., Worrall, L., Hickson, L., & Murison, R. (2003). Finding a focus for quality of life with aphasia: Social and emotional health, and psychological well-being. Aphasiology, 17, 333-353.

Davis, A., & Pring, T. (1991). Therapy for word finding deficits: More on the effects of semantic and phonological approaches to treatment with anomic patients. Neuropsychological Rehabilitation,1,135-145.

Davis, P., & Mehan, H. (1988). Professional and family understanding of impaired communication. British Journal of Communication, 23, 141-151.

Davidson, B., Worrall, L., & Hickson, I. (2003). Identifying the communication activities of older people with aphasia: evidence from naturalistic observation. Aphasiology, 17, 243-264.

Denman, A. (1998). Determining the needs of spouses caring for aphasic partners. Disability and Rehabilitation, 20, 441-423.

Ferguson, A. (1992). Conversational repair of word finding difficulty. In M. Lemme (Eds.), Clinical aphasiology conference proceedings, 21, 299-307. Austin, TX: Pro-Ed.

Ferguson, A. (1994). The influence of aphasia, familiarity and activity on conversational repair. Aphasiology, 8, 143-157.

Fougeyrollas, P., St-Michel, G., Bergeron, H., Cloutier, R., & Cote, J. (1999). Quebec Classification: Disability Creation Process; International Network on Disability Creation Process, Quebec-City. Retrieved december 21, 2009, from http://www.ripph.qc.ca.

Frattali, C., Thompson, C. K., Holland, A. L., Wohl, C. B., & Ferketic, M. K. (1995). American Speech Language-Hearing Association assessment of functional communication skills for adults. Rockville, MD: American Speech-Language-Hearing Association.

Garcia, L. J., Barrette, J., & Chantal, L. (2000). Perceptions of the obstacles to work reintegration for persons with aphasia. Aphasiology, 14(3), 269-290.

Goodwin, C. (1995). Co-constructing meaning in conversations with an aphasic man. Research on Language and Social Interaction, 28, 233-260.


Goswami, S. P., Shanbal, J. C., Samasthitha, S., Navitha,U. (2010). Field testing of Manual for Adult Non-fluent Aphasia therapy- Kannada (MANAT-K). Research Project, All India Institute of Speech and Hearing, Research Fund, Mysore, India.

Hagen, C. (1973). Communication abilities in hemiplegia: Effect of speech therapy. Archives of Physical Medicine and Rehabilitation, 54, 454-463.

Hilari, K., & Byng, S. (2009). Health-related quality of life in people with severe aphasia. International Journal of Communication Disorders, 44 (2), 193-205.

Hilari, K., & Northcott, S. (2006). Social support in people with chronic aphasia. Aphasiology, 20, 17-36.

Hilari, K., Wiggins, R., Byng, S., & Smith, S. (2003). Predictors of health-related quality of life (hrql) in people with chronic aphasia. Aphasiology, 17(4), 365-381.

Hinckley, J. J. (2002). Vocational and social outcomes of adults with chronic aphasia. Journal of Communication Disorders, 35(6), 543-560.

Hinckley, J. J., & Packard, M. E. (2001). Family education seminars and social functioning of adults with chronic aphasia. Journal of Communication Disorders, 34(3), 241-254.

Hodson, S. P., Wood, V. A., & Langton, H. R. (1996). Identification of stroke carers at risk: a preliminary study of predictors of psychological well being at one year post-stroke. Clinical Rehabilitation, 10, 337-346.

Holland, A. (1980). Communicative abilities in daily living: A test of functional communication for aphasic adults. Baltimore: University Park Press.

Jelsma, J. (2009). Use of the international classification of functioning, disability and health: a literature survey. Journal of Rehabilitation Medicine, 41, 1-12.

Kagan, A., Simmons-Mackie, N., Rowland, A., Huijbregts, M., Shumway, E., McEwen, S., Threats, T. & Sharp, S. (2008). Counting what counts: A framework for capturing real-life outcomes of aphasia intervention. Aphasiology, 22 (3), 258- 280.

Kertesz, A. (1982). Western Aphasia Battery. NewYork: Grune & Stratton.

Laasko, M. (1997). Self-initiated repair by fluent aphasic speakers in conversation. Studia Fennica Linguistica, 8, Helsinki, Finland: Finnish Literature Society.

Laasko, M. (2002). Collaborative construction of repair in aphasic conversation: An interactive view on the extended speaking turns of persons with Wernicke‟s aphasia. In C. Goodwin (Eds.), Conversation and brain damage, (pp. 163-188). New York: Oxford University Press.

Laakso, M., & Klippi, A. (1999). A closer look at the “hint and guess” sequences in aphasic conversation. Aphasiology, 13, 345-363.

Le Dorze, G., & Brassard, C. (1995). A description of the consequences of aphasia on aphasic persons and their relatives and friends, based on the WHO model of chronic diseases. Aphasiology, 9, 239-255.

Le Dorze, G., Croteau, C., Brassard, C., & Michallet, B. (1999). Research considerations guiding interventions for families affected by aphasia. Aphasiology, 13, 922-927.

Le Dorze, G., Julien, M., Brassard, C., Durocher, J., & Boivin, G. (1994). An analysis of the communication of adult residents of a long-term care hospital as perceived by their caregivers. European Journal of Disorders of Communication, 29, 241-267.

Lindsay, J., & Wilkinson, R. (1999). Repair sequences in aphasic talk: A comparison of aphasic-speech and language therapist and aphasic-spouse conversations. Aphasiology, 13, 305-325.

Lubinski, R., Duchan, J., & Weitzner-Lin, B. (1980). Analysis of breakdowns and repairs in aphasic adult communication. In R. Brookshire (Eds.), Clinical aphasiology conference proceedings (pp. 111-116). Minneapolis: BRK.

Marshall, J. C., Pound, C., White-Thompson, M., & Pring, T. (1990). The use of picture matching tasks to assist in


226

word retrieval in aphasic patients. Aphasiology,4, 167-184.

Mazzoni, M., Vista, M., Geri, E., Avila, L., Bianchi, F., & Moretti, P. (1995). Comparison of language recovery in rehabilitated and matched, non-rehabilitated aphasic patients. Aphasiology, 9, 553-563.

Milberg, W., & Blumstein, S. (1981). Lexical decision and aphasia: Evidence for semantic processing. Brain and Language, 14, 371-385.

Milroy, L., & Perkins, L. (1992). Repair strategies in aphasic discourse: Towards a collaborative model. Clinical Linguistics & Phonetics, 6(1), 27-40.

Nickels, L., & Best, W. (1996). Therapy for naming disorders (Part I): Principles, puzzles and progresses. Aphasiology, 10, 21-47.

Noreau, L., Fougeyrollas, P., & Vincent, C., (2002). The Life-H: Assessment of the quality of social participation. Technology and Disability, 14, 113-118.

Oelschlaeger, M. (1999). Participation of a conversation partner in the word searches of a person with aphasia. American Journal of Speech-Language Pathology, 8, 62-71.

Oelschlaeger, M., & Damico, J. (2002). Word searches in aphasia: A study of the collaborative responses of communicative partners. In C. Goodwin (Eds.), Conversation and brain damage (pp. 211-230). New York: Oxford University Press.

Parr, S. (2001). Psychosocial aspects of aphasia: Whose perspectives? Folia Phoniatrica et Logopaedica: Official Organ of the International Association of Logopedics and Phoniatrics (IALP), 53(5), 266-288.

Perkins, L. (2002). Negotiating repair in aphasic conversation: Interactional issues. In C. Goodwin (Eds.), Conversation and brain damage (pp. 147-162). New York: Oxford University Press.

Peuser, G., & Schriefers, H. (1980). Sentence comprehension in aphasics: Result of administration of the „three-figure test‟. British Journal of Disorders in Communication, 115-127.

Pound, C., Parr, S., & Duchan, J. (2001). Using partners‟

autobiographical reports to develop, deliver and evaluate services in aphasia. Aphasiology, 15, 477-493.

Ross, K. B., & Wertz, R. T. (2003). Quality of life with and without aphasia. Aphasiology, 17, 355-364.

Salonen, T. (1995). Report of a questionnaire survey of poststroke patients with aphasia and their families. Topics in Stroke Rehabilitation, 2(3), 72-75.

Sarno, M. T. (1969). The functional communication profile manual of directions. Rehabilitation Monograph, 42, New York University Medical Center.

Sarno, M. T. (1992). Preliminary findings in a study of age, linguistic evolution and quality of life in recovery from aphasia. Scandinavian Journal of Rehabilitation Medicine, Supplement, 26, 43-59.

Semenza, C., & Goodglass, H., (1985). Identification of body parts in brain-injured subjects. Neuropsychologia, 23, 161-176.

Simmons-Mackie, N., Threats, T., & Kagan, A. (2005). Outcome assessment in aphasia: a survey. Journal of Communication Disorders, 38, 1-27.

Smith, L. (1985). Communicative activities of dysphasic adults: A survey. British Journal of Disorders of Communication, 20(1), 31-44.

Swati, B. (2008). The burden of persons with aphasia on the caregivers. Unpublished Masters Dissertation, University of Mysore, Mysore, India.

Wade, D. T., Hewer, R. L., David, R. M., & Enderby, P. M. (1986). Aphasia after stroke: Natural history and associated deficits. Journal of Neurology, Neurosurgery, and Psychiatry, 49(1), 11-16.

Wahrborg, P., & Borenstein, P. (1989). Family therapy in families with an aphasic member. Aphasiology, 3, 93-97.

World Health Organization (2001). International Classification of Functioning, Disability and Health (ICF). Geneva: World Health Organization.

Worrall, L., & Hickson, L. (2003). Communication disability in aging: from prevention to intervention. Albany, NY: Delmar Publishers.

Worrall, L., & Holland, A. L. (2003). Editorial: Quality of life in aphasia. Aphasiology, 17, 329-332.

��A�B��CDEF��B��E��

�

��

��AB�C�D�C��EF��A��C��A��

��EF��E��A��A��E�D��C��

�!��

��E D�E�EA� ��B�� E� �� A�E��A� ��E ��A��F��F �E��B�E� �A ��E �DEEF� �� F��EA �� B�� CDEF��B�

��E�� C�� A F��D��A �� D�F�� E�E��D�A� F��EA� �� BD� �� B��EF�� E�E ��EA �� E

��B�� F�A��BD �!" ��D�F��E�E��D�A�F��EA��A�F��A�F��BD �!"F��EA��C��CB��EF��E�E

��E��A��#��E��BD�$%&"�� "&'�� A�'&(��E�B��EF��A��ED�E�EA��B��E�E��E��)��

*�A�B��E �E�� )*� +B��A� !,,%��A��D�F�B�E�E�F��D��A ��E�E F��E��B� �� E �B��EF�� A ��

F�A�� A� F��A�F�� BD� ��E �DEEF� ��D�E� �EF��E� �E�E ��A�F��E� �A -.� �A� �A��/E� �� B�

��F��E��E�E�B��E��F��EA��C��)*��F��E�D��E��AA��A��E��E

��BD� ��EFED��A�A�E0D�E��A ��B��EF��EA��C��E��E��BD�DE��E��E��E��A

��E)��D��D��AE��FC��BF�B�E�)C��B��E��-A��ED�F�B�E�E�F��D��A�� EF��EA��C��F��E�D��E�

��A ��E ��D�F�� E�E��D�A� F��EA �A �� E ��E ��BD� �A� �A �� E D��E�E�� E0FED� ��.E�FEA��E ��

1��F��2��EF�C��D�ECEA�EAFE��E�E�B��ED�E�EA��B��A��F��E��EAEE��A��E��E��EA��A�

��E��D��AF��EA��C��

��A�BC�BDDBEFD��B�EFD��E��D��F��DB�B�B�BD��B�C�BC��E�E��F�E��F�EF��EB��

��ABCDEF� E��C�BF�� A�E�� A�� ABCDEF�

E��C�BF� �DE��E� �� ABCDEF� E��C�BF�

��DCD��E� �� AED�� F��CA��

�DE�� DC�� C�� ABCDECD�� E�F�CDF�E�

��A�D�� ABCDEF�� DE� A� E��C�BF� �� E ��D�A��

��DCD��E� ��A�A�C��D!�� " � �D��E��A�� A"��FA�DCD�E�

�� E��DA�� D�C��A�CD��E� A�� FFB�D�ACD�� AE� �� AE�

E�� EC�D�C�� D�C��ECE� A�� D�� CDCD��

"��A�D��#��$�A�C��%��A�D!ACD��&&'�(�

�BCDEF� DE� A� �D�� A�DA"�� "�AD�� F��C�

�DE��C�AC��D�EC�A��A�E��B�D��D��A�� D��

A�� A�� E� A� EC�A� � ��B�E�� DC��BC�

��FDEED�� #�� $�A�C�� %��A�D!ACD�� &&�)�

*�E��D��&&+�(�� F�C�FE�C��C��CD�B��C��B��

A�B�C,��A�C��B��C��D��F��FBC��F��-A�D��

.�/B��FA��&&+�(�

� �CA0� ��E� C�� C�� A � D�� D�� A��A��

��E� �D1�� 2"BD��D�� "��1E3� C�� EC�B�C� ��AE�E� A��

E��C��E� C�AC� �0��EE� �B�� D��AE(� /�BE� E �CA0� ��A�E�

�DC�� C�� E4� C�� A � D�� D�� E�

��F"D�� D�C�� EB�� B�DCE� AE� A� ��AE�� A� ��ABE�� A�� A�

E��C��(��E��C��EDECE��F��ABE�E(��

��ABE�� EDECE� �� F�� AE�E(� �� AE��

��EDECE��A��B��E(��EDECE��

F�� F��F�E(� � �CA�CD�A�� C� E��C��E��

E��B��"��"�C��AFFACD�A�� C� A��F�A�D��B�(�

5�� 0AF�� A� E��C�� D1�� 2��B��EE� �� D��AE�

E�� B�D�BE� 3� DE� ��AFFACD�A�� C� "BC� DC� DE� ��C�

F�A�D��B�(� /�BE�� B��E� �� E �CA0� E��B�� "�� F��,

�� ,FAD�4E�A��A6A�A�7�FAD�(��F)� ��EE�� 8A��BA��

�AC�� 99�$��: E��E� AFA�A1�7 A��(��F�

��ED�� B�� C�� 0��AD�� C�� FACD�� E��C��E�

��D��A��"�C��AFFACD�A��A��F�A�D��B�(�

� �CA0� DE� C�� ECB� � �� E� 2��BF��3�

C��C�� D�� AE�E� A�� ABE�E(� 9C� DE� A�E�� C��

�D��BDECD�� A�D!ACD�� C�AC� F��DAC�E� "�C�� E�B��

A��F�A�D��E� A�� A�D!�� D�C��AE�E�

A�� E��C��E(� 9�� C �D�A�� D�� D�� C��

E �CA�CD��F��C��F��C�E�" �� A�E(�9��C��DE�

A� � �A�DA�� D�� F��CA�� AE��CE� ��

��D�� DC� ��A�E� C�� E �CA�CD�� D�DCE� A��DC��C��

�A��BA��D�DCE(�

� �CA0� DE� ��C�� A� ;��FA��

D�C��AC��;� E��E�� D�C��A�CD�� C��B�� A��BA�� A��

�B��0DE�DE�F��D��C��C��C��E��D�D��D��AE�C�AC��

�A��E��A�� D��B��FD��E��C�D�1�EFA�� CA�1�(�:A� ��

ABCDECD�� D��D�D�BA�E� � FA � �A�� D��D�B�C � D�� BCCD��

E��C��E� C��C�� A�� B��ECA��D�� AFFACD�A��

��C�E��C��EC�B�CB��(�

�CB� � �� AFFACD�A�� F��C� D�� D��

�DC� �� DE� �EE��CDA�� C�� B��ECA�� C�� A E� A��

��DA��E� A�� A�E�� C�� D�� CD�� A�� D�D��C�

F�C��E� �� -��A"D�DCACD��(� *�AFFACD�A�� E� �A��

"�� C�� C�� "�� C�� ED��D�D�A�C� �A�DA"�� D��

C�� A��BA�� ABCDECD�� D��(� 9C� �AE� "�� D��

ECB�D��D��C��<��DE��E��A1D��ABCDECD��D��(�9��DA��

ECB�D�E� �� AFFACD�A�� E� �� C�� DE��

��B�ACD��AE��A�� (�$��C�DE�ECB� ��AE�

CA1�� B�� C�� D��ECD�AC�� C�� AFFACD�A�� E� � D��

:A�A A�AF� E��A1D�� D�� DC�� (� /�DE� ECB� �

��B�� AB�F��C� C�� B��C� B��ECA��D�� E �CA�CD��

��E�� DA��E� A�� A��BA�� A"D�DCD�E� D�� D��

�DC��(�/��E��C� ECB� � ADF�� C�� D��ECD�AC�� C��

/

��E��A3��3--- 455,&!5 .��&6 C*. �--C7 )��E

��+��

E �CA�CD��ACB��E�D��C��E��D��DC��BCDEF�

E��C�BF� �DE��E� D�� F�A�DE�� DC�� C �D�A��

��D��D��(�

B��

E�!"��#/��B�E��EB"6��CE��ED��

C�� E��C� ECB� (� *��B�� 9� ��EDEC�� =� ��D��

�DA��E��AE��" �>BA�D�D��E ��DEC�?�E��,�

�A��BA�� AC��DECE� BED�� :,��AC� �-�"D�E�� &&��

A�� D��CDA�� DA��EDE� ��1�DEC,�BCDEF� ��C�BF�

�DE��E� ��,�� @�� AFA�A�� B�DC��A�� .�

AD6A AE��&&+B��DC�D��C��A��B��C,�� E(��

�

*��B�� 99� ��EDEC�� =� A��FAC�� C��E�

�DC�D�� C�� A�� B�� C,�� A�E(� �B"6��CE� ��

�D�D��D�C��C��A��B�E)�C,=� �E��=,'� �E��A��',��

�E(��C��EB"6��CE�D��C��E��C�ECB� ��FA��E(�

E�C��A ��# �B"6��CE� �DC�� ACD�� A��BA�� AE�

:A�A A�AF� �DC�� FD�� :-�� "��D�� ?� A��A��

D�C��D�� AE� �DA��E�� " � �E ��DEC�� A�� A�D��

�� F��AD�CE� �� A�D�� EE�� DEBA�� DF�AD�F��C�� A��

F��D�A�� F��D�ACD��E� �� A��CA�� D�C��D��

E��C��(�� B�� C ��E�� D��B�� 0��C�

-�CCDE��DE��(��B"6��CE��DC�� FD��,F��AC��

�� DF�AD�F��C�� E��C��(�/�� A��BA��A�� C��

EB"6��CE� �A�� ,�� A�E� "��D�� C��D�� D�A��

A�� A�� B�ACD�� F�� C�A�� '�F��C�E� C�� A��

E��C��A� ��AE�ACC��" �A��C��EB"6��CE(�

B��C�$��#

B�C�D�C��%�A��&��'B%&()��A�(�**+��,

/�DE� C�EC� AEE�EE�E� D��D�D�BA�E� ��F� E,�� A�E� ��D��

D��B��E� "�C�� CD�� A�� 0��EED�� E1D��E(� 9C� ��E�

D�� B��ECA��D�� "�C�� CD�� A��BA�� A�� A��

�0��EED�� A��BA�� A�� C�� D�� D�� E� D��

��A��D�� A�E� �� C��A� (� 9C� �A�� "�� BE�� C�� C�EC� C��

DF��CA�C� ��FAD�E� �� A��BA�� D(�(�� E�FA�CD�E� A��

E �CA�CD��A"D�DCD�E(�/��E �CA�CD��A�AF�C��E��:8/�DE�

BE��C��E��C�ECB� (�9C��EDECE��EB"E��CD��E�

�AF�� :��FD�� EC�B�CB��E� �:�� B�A��

��FE� ��5�� /��E�E� �/�� F*� FA�1��E� ��F*�� AE��

FA�1��E� ��:�� /�A�EDCD�� 9�C�A�EDCD�� A��ABEACD��E�

�/9�� C�� C ��E� ��/�� 6B��CD��E� A��

GB�CACD��E� ��G�� F�A�ACD��E� �� DCD��A��

��ABE�E� �� A�� A�CD�D�A�� EC�B�CD�� (� 9C�

D��B��E� "�C�� F��ED�� A�� 0��EED�� CAE1E�

��D��CAD�E�=�>B�ECD��E�D��A��EB"E��CD��

-�� F��A �� ,� /�� D�CB��E� ��D�CD��

E��D��F��C�A��A�1��D��F��C��AE�BE��

�D�CB�� E��D�CD�� CAE1(� �AF�� D�CB��E� �� BE��

A�� C�� EB"6��CE(� /�� A�AF�C��E� D��B�� / �� /�1��

-ACD�� //-�� /�CA�� FBF"�� C��E�� CA��

��*�AFFACD�A�� C��C��E��*��CA,�

�� *�AFFACD�A�� C� �DF�� C��E�

��*��D�� CA�� *�AFFACD�A�� C�

��F��B�� C��E� ��*�� A�� CA��

*�AFFACD�A�� C��F��0��C��E��*��0��(

-��#<A�� EB"6��C� �AE� C�EC�� D��D�D�BA�� D�� A�

>BD�C� ��F(� /�� EB"6��C� �AE� AE1�� C�� EDC� ��F��CA"� �

�� A� ��AD�� A��EE� C�� CA"�� DC�� C�� 0��DF��C��

EDCCD�� A� ��AD�� EDC�� C�� C�� EB"6��C(� H�� C��

A�FD�DEC�ACD�� C�� C�EC�� A��C� �DC�� C�� D�� AE�

"BD�C(� /�� E �CA�CD�� CD�� :8/� �AE� C�EC�� " �

D�EC�B�CD�� C�� EB"6��C� �� A�� CAE1� AE� �� C��

D�EC�B�CD�� D�� D�� C�� C�EC(� <A�� EB"6��C� �AE�

D�EC�B�C��AE��E4�29��D��AE1� �B��D��>B�ECD��E�D��

�A��EB"E��CD��A�� B��A��C��A�E��A��D�� (�9�

�D��D�� B�C��0AF��E�D��A��EB"E��CD��(3�

�

�B"E�>B��C� � �D�� D�CB�� A��E� ��D�CD��

E�� D��F��C� A�� D�� D�CB�� A��E� ��D�CD��

�A�1� ��D��F��C� �AE� BE�� C�� C� C�� E��

EAF��(�<A��EB"6��C��AE�D�EC�B�C��AE��E4�29��D��

E��C��E�CE��D�CB��E��A�� B�E��B��E��D"��C��

�D�CB��E�D��E��C��E(�9�� C��EB"6��CE��AD�� C�� D�DCDAC��C��

��EACD�� 0��DF��C�� B�� D�� B�E?>B�EC,

D��E� ��A�C� C�� C�� D�CB�� "�C�� C �D�A��

��D��D��AE��AE�ABCDECD��D��3(�

�

�� EAF��E� ��D�DC�� BED�� "�C�� C�� CAE1E�

��BED��D�DCA��D��/�A�E��,

'&�(�� C��EB"6��CE��D��C�1��E�AC� C��

��C��E�EED��AE��D��F��C�

�A�CD��#

��A� EDE� �� :8/4 /�� EAF��E� ��

C�A�E��D"�� D�� 9�� A�� A�A� !�� A�D�BE�

��AFFACD�A��E�D��C��D��C�E �CA�CD��AC��D�E�

��C��EB"6��CE��F�A��"AE��C��D��E��E�D��

�A��AC�� (�/��C�CA��E��E��A��AC��

�"CAD��" ��A��B�ACD��C��BF"��DC�FE��0��EE��

��C� � �� A�� AC�� (� �� C�� AC��D�E� ��

C��F�BC��C��A�E��=�A��A��C��EB"6��CE�

E�� C�� AC��D�E� �AE� �A��B�AC��(� ��CA��

�I�� E�� AE� �A��B�AC�� AE � ��E��CACD�� D��

��A��D��FE(�

�

��A� EDE� �� D�CB�� E��D�CD�� CAE14 /�� E��

EAF��E� �� C�A�E��D"�� D�� 9�� A��

A�A� !�� A�D�BE� ��AFFACD�A�� E� �� "�C�� C��

��B�E(� ��CA��BE� E�� EAF�� AE� �"CAD��

�A�D�BE� CAE1E� �D1�4� C ��E� �� E��C��E� A�� C �� C�1��

�ACD�� A��AE� A�A� !�� C�� ECA"�DE�� C��CA��

��B��C��A"��F��CD��AC��D�E��E �CA0�

D��A�E��CA��BE�E��EAF��(��CA��I��E��

�AE� �A��B�AC�� AE � ��E��CACD�� D�� A��D��

��FE(�

�


�

��J��

/�� D��C� C ��E� �� E��C��E� ��

A�A� !��,� EDF�� E��C��E�� F��B�� E��C��E� A��

��F��0�E��C��E(��EDF��E��C��EDECE��

�� FAD�� ABE�(� �� F��B�� E��C�� AE� C��

F��FAD��ABE�E��D�1��" �C��D�AC��E�2A��3��

2"BC3(��F��0�E��C��EDECE��FAD��ABE��

�DC�D��D��C�� A��EB"��D�AC��ABE�E(�/ ��,C�1��

�ACD��AE��A��B�AC��BED��C��FB�A4

/ ��,C�1��ACD��K��FBF"��C ��E?FBF"��

C�1��E��L�&&

/ ��E� �� C�� C�� C�CA�� BF"�� D��C�

��E�A�� C�1��E�F�A�E� C�CA��BF"�� >B��

�A��BCC��D��E��(�

)��C��A��A

/�� ACA� �"CAD�� F� �DC�D�� A�� A��EE� C��

��B�E� �� EB"6��CE� �AE� ECACDECD�A�� A�A� !��(� 5��

:A�A A�AF�8A��BA��/�EC��:8/��E��D�CD��ECACDECD�E�

�AE� �F�� "��ABE�� B�ACD�� ECB�D�� A��

�� D�CB��E��D�CD�� CAE1E�� ,�A�AF�C�D�� C�ECE��AE�

�A��D�� BC� �� "�C�� C�� B�E� A�� AE� ��F�A��(�

��F�A�DE��F�A��:8/�E��A��//-�"�C��C��

C �D�A�� D��D��A��D��DC��

C��EAF��A��B�E��AE�ECB�D��BED��:A��#�DC��

M,C�EC(�/�� EB�CE� �A��"��E��C�� A��DE�BEE��

B��D��C�EB"E��CD��E(��

B%&# /�� F�A�� A�� ECA��A�� DACD�� :8/�

EB"C�ECE� ��CD�� A�� 0��EED�� E �CA0� �� C��

C �D�A�� D��D��A��D��DC��

�D��C� A�� B�E� A�� E�� D�� /A"�� C,=� �E��

/A"��=,'� �E��A��/A"��E��',�� E�(��

/�� C �D�A�� D�� D�� D�� C�� A��

��B�� C,=� �A�E� �*��B�� A�� E�� A� F�A��

��CD��E��=(&��C��:��A��EB"C�ECE��

"�D��C��D��EC��A��A�F�A��E��&(E&��C��:��

EB"C�EC��"�D�� C�� EC�AF�� C��EB"C�ECE� D�� C��

:8/�ECB�D��(�/��C �D�A�� D��D��C��

EAF�� A�� B�� *��B�� A�� E�� A� F�A��

�0��EED��E��C(+&��C��:�EB"C�EC��"�D��C��

�D��EC�� A�� A�F�A�� E�� (&� �� C�� EB"C�EC��

"�D��C��EC(��

�

/�� D��DC�� C�� A�� B��C,=�

�A�E��*��B��H��A��E��A�F�A��CD��E��

�(+��C��:��A��/9��EB"C�EC��"�D��C��D��EC��A��

A�F�A��E��(C��C��5�A��EB"C�EC��"�D��

C�� EC�AF�� C��EB"C�ECE� D�� C��:8/�ECB�D��(�

5�� 0��EED�� CAE1�� C�� EAF�� A�� B�� *��B�� H��

�A��E��A�F�A��E��(J��:��EB"C�EC��"�D��

C�� D��EC�� A�� A� F�A�� E�� &(�� C��

EB"C�EC��"�D��C��EC(�

/A"��(�)*��+EFED��A�A�80D�E��A��A��0�A

��D�F��E�E��D�A�2��EA��2��A�F��EA

��C��%&"��E��BD�

.� �CA�CD��

�AC��D�E�

-��CD�� <0��EED��

/��

�*��

��

�*��H��

/��

�*��

��

�*��H��

:��

:�A�� &(E� �(+� C(C� (J��

�� (� &(C�� (&� &(==�

�5�

:�A�� E(�� (C�� (C� &('C�

�� (�� &(=E� (J� &(=E�

�

/��

:�A�� C(E� ('�� (+� (&E�

�� (E� &(=�� (E� &(&��

�

�F*�

:�A�� C(C� �(CE� C(E� ('��

�� &(J� &(=�� &(J� &(=C�

�:�

:�A�� =(&� �(CE� C(+� ('�

�� &(&� &(=�� &(=� &(==�

/9��

:�A�� C(J� �(+� E(�� ('��

�� &(E� &(C�� (C� &(=C�

�/�

:�A�� C(J� �('� (+� (+E�

�� &(�� &(=E� &('� &(CE�

�G�

:�A�� C(J� �(CE� �(�� ('��

�� &(E� &(=�� (� &(=C�

��

:�A�� =(&� �(�� E(� (&��

�� &(&� &(CE� &(�� &(�&�

��

:�A�� C(+� �(�&� C(&� (&�

�� &(+� &(CC� (� &(�&�

��

:�A�� =(&� (C�� (� &(��

�� &(&� &(+J� (E� &(CC�.:��N�:��FD��EC�B�CB��E)��5�N��B�A��FE)�/��N�

/��E�E)� �F*� N� �F*�FA�1��E)� �:� N� �AE��FA�1��E)� /9�� N�

/�A�EDCD��9�C�A�EDCD��A��ABEACD��E)��/�N��C��C ��E)�

�G�N��6B��CD��E�A��GB�CACD��E)��N��F�A�ACD��E)��

N��DCD��A��ABE�E)�A��N��A�CD�D�A��EC�B�CD��(�

/�� C �D�A�� D�� D�� D�� C�� A��

��B�� =� N� '� �A�E� �*��B�� A�� E�� A� F�A��

��CD��E��=(&��C��:��A��EB"C�EC��

"�D��C��D��EC��A��A�F�A��E��&(J&��C��:��

EB"C�EC��"�D�� C�� EC�AF�� C��EB"C�ECE� D�� C��

:8/� ECB�D��(� /�� D�� C�� EAF�� A�� B��

�*��B�� A�� E�� A� F�A�� 0��EED�� E��

C(J&� �� :� EB"C�EC�� "�D�� C�� D��EC�� A�� A� F�A��

E��(&��C��EB"C�EC��"�D��C��EC(�

/�� D��DC�� C�� A��B��=�N�'�

�A�E��*��B��H��A��E��A�F�A��CD��E��

�(+��C��:��EB"C�EC��"�D��C��D��EC��A��A�F�A��

E�� (C�� C�� 5� A�� EB"C�EC�� "�D�� C��

��EC�AF��C��EB"C�ECE�D��C��:8/�ECB�D��(�5��

�0��EED�� CAE1�� C�� EAF�� A�� B�� *��B�� H�� A��

E��A�F�A��E��(��:��EB"C�EC��"�D��C��

�D��EC��A��A�F�A��E��&(+&��C��/9��EB"C�EC��

"�D��C��EC(�

��E��A3��3--- 455,&!5 .��&6 C*. �--C7 )��E

�E&��

/A"��(�)*��+EFED��A�A�80D�E��A��

��A��0�A��D�F��E�E��D�A�2��EA��2��A�

F��EA��C��"9'��E��BD

.�� CA�CD��

�AC��D�E�

-��CD�� <0��EED��

/��

�*��

��

�*��H��

/��

�*��

��

�*��H��

:��:�A�� &(J� �(+�� C(+� �(��

�� (�� &(C=� &(E� &(+�

�5�

:�A�� C(&� (C�� C(�� (&��

�� (�� &(=�� &(�� &(&��

/��

:�A�� C(+� (CE� E(E� &(+C�

�� &(E� &(=�� &(+� &(CC�

�F*�

:�A�� C(�� ('�� C('� (C&�

�� &(+� &(=�� &(J� &(+J�

�:�

:�A�� =(&� �(CE� C(J� (��

�� &(&� &(J&� &(�� &(+C�

/9��

:�A�� C(J� �('� E(+� &(+&�

�� &(E� &(=E� &(�� &(CC�

�/�

:�A�� C(�� ('� �(=� �(&�

�� &(J� &(=E� (� &(�&�

�G�

:�A�� C(J� �(CE� �(�� ('&�

�� &(E� &(=�� &(J� (C�

��

:�A�� =(&� �(�� (J� (C&�

�� &(&� &(CE� (E� &(+J�

��

:�A�� =(&� �(�&� C(=� �(&�

�� &(&� &(CC� &(J� &(J+�

��

:�A�� C(J� (C�� (&� &(+�

�� &(�� &(+J� (J� &(C=�.:��N�:��FD��EC�B�CB��E)��5�N��B�A��FE)�/��N�





�

/�� C �D�A�� D�� D�� C�� A��

��B�� ',�� A�E� �*��B�� E�� A�� E�� A� F�A��

��CD��E��=(&��C��:��/��G��.�

��EB"C�EC��"�D��C��D��EC��A��A�F�A��E��C(&�

�� C�� :�� EB"C�EC�� "�D�� C�� EC� AF�� C��

EB"C�ECE� D��C��:8/�ECB�D��(�/��D��C��EAF��

A�� B�� *��B��E�� A�� E�� A�F�A�� 0��EED��

E�� C(J&� �� :�� F*�� :� EB"C�EC�� "�D�� C��

�D��EC�� A��A�F�A�� E��E(&� �� C��G�EB"C�EC��

"�D��C��EC(�

�

/��D��DC�� C��A��B��'�N��

�A�E��*��B��HE��A��E��A�F�A��CD��E��

E('��C��G�EB"C�EC��"�D��C��D��EC��A��A�F�A��

E�� ('E� �� C�� 5� EB"C�EC�� "�D�� C�� EC�

AF�� C�� EB"C�ECE� D�� C�� :8/� ECB�D��(� /��

��D��DC��C��A��B��'�N�� A�E��*��B��

HE��A��E��A�F�A��0��EED��E��E(&��C��

:��EB"C�EC��"�D��C��D��EC��A��A�F�A��E��(&�

��C��EB"C�EC��"�D��C��EC(�

/A"��E(�)*��+EFED��A�A�80D�E��A��A��0�A

��D�F��E�E��D�A�2��EA��2��A�F��EA

��C��'9(��E��BD

.� �CA�CD��

�AC��D�E�

-��CD�� <0��EED��

/��

�*��E��

��

�*��HE��

/��

�*��E��

��

�*��HE��

:��

:�A�� C(&� E(�E� C(J� E(&��

�� (�� &(CE� &(E� &(&��

�5�

:�A�� C(+� ('E� C('� (&E�

�� &(C� &(==� &(�� &('+�

/��

:�A�� C(J� �(&� C(&� (&C�

�� &(� &(&�� &('� &(&��

�F*�

:�A�� C(J� E(&E� C(J� �(C�

�� &(C� &(&�� &(=� &(J&�

�:�

:�A�� =(&� E(&E� C(J� �('��

�� &(&� &(&�� &(E� &(J&�

/9��

:�A�� C(J� E(C� C(�� (��

�� &(E� &(=E� &(+� (�

�/�

:�A�� =(&� E(&C� E(J� �(+�

�� &(&� &(&�� (E� &(C=�

�G�

:�A�� =(&� E('� E(� (C&�

�� &(&� &(=E� &(+� &(=C�

��

:�A�� =(&� E(C�� E(=� (C&�

�� &(&� &(=�� &(+� &(=C�

��

:�A�� =(&� E(&&� C(+� �('�

�� &(&� (�� &(C� &(J&�

�

��

:�A�� =(&� �('�� E(�� (&�

�� &(&� &(J&� (E� &(&��O:��N�:��FD��EC�B�CB��E)��5�N��B�A��FE)�/��N�





�

��F�A�DE��F�A��:8/�E��"�C��C��

��D��DC��D��C�A��B�E��AE�ECB�D��

BED�� :A�� #�DC�� M,C�EC(� /�� AE� �A�DACD��

AF��C��EB"6��CE��*��B��9��*��B��99�A��*��B��999(�

9C��AE��"E��C�AC��D��CD��A��0��EED��C��

�AE� ED��D�D�A�C� �D�� D�� :8/� E�� AE� A��

��EE�� F�*��B�� 9� C��*��B�� 99� �� A�� EB"C�ECE(�

�D��D�D�A�C��D��AE��B��"�C��C��*��B��99�

A��*��B��999��A�F�EC�A��EB"C�ECE� D��"�C��CD��

A�� 0��EED��(� /�� B�E� �A�� D�� A�B�E�

��F�A��C�� B��B�(�

�

PA��JC'��-D�1E�A��#D��J�=��HA�CA1��

-BCC�� A�� 0� �J�=�� A�C�� HA1�� A�� -BCC��

�J�+�� D�� A�� HA�C��B��D� �J�� C�� C�AC�

ABCDECD�� D�� A�� D��D�B�C � D�� D�� C��

E �CA�CD�� B��E(� P6��BA�� A�� /A��,5�BE"�� &&��

A�� P�� &&�� C�� C�AC� ��D�� DC�� ABCDEF�

�D��A��A��AFFACD��E��(�

-�� F��A# /A"�� C� E��E� C�� F�A�� A��

ECA��A�� DACD�� D��C� �A�AF�C��E� �� D�CB��

�

�

��E��D�CD�� CAE1� �� C�� C �D�A��

A�� D��DC�� C�� A�� B

�E�� A��',�� E(�5D�B�� E��E� C��F�

��//-�D��D��DC��A��C �D�A

��D��A��EE�A�� C��A��B�E(�5D�B�

��A�� FBF"�� C��E� D�� D��

A�� C �D�A�� D�� D�� A�

��B�E(� 5D�B�� E� E��E� C�� F�A��

*�AFFACD�A�� C� ��C��E� D��

��A�� C �D�A�� D��D��

��B�E(� 5D�B�� C� E��E� C�� F�A��

*�AFFACD�A�� C� �DF�� C��

�DC��A��C �D�A�� D��D��

A�� B�E(� 5D�B�� =� E��E� C�� F�A��

*�AFFACD�A�� C� ��F��B�� C

A��FA�E�A��EE�C��A��B�E(�5D�B

F�A�� CA�� *�AFFACD�A��

��C��E� D�� D�� DC��

��D��D��A��EE�C��A��B�

�

�

5D�B��(�)E�ADE�FEA��E��+�AF

�C��A��D�F��E�E��D�A�2��EA

��E��BD��

&DF� &��A )�� '&&)/# /�� C �D�A

��D�� C��A��B��C�N�=� �A�E� �*

E��A�F�A��//-�E��C(+�(��D��

�� C��EAF��A��B��*��B��H��A��

//-� E�� E'(��(� �� F�A�DE��

"�C��C �D�A�� D��D��A�

A��B��AE��A��D��BC�BED��:A��,

/�� EB�CE� E�� C�AC� C�� AE�

�D�� ECACDECD�A�� "�C�� C �D�A�

��D��A��D��DC��C��EA

/�� C �D�A�� D�� D��

��B��=,�'� �A�E� �*��B��A�� E��

E�� ='('�(� ��D�� DC��

��B�� *��B�� H�� A�� E�� A� F�A�

C'(JJ(�/��F�A�DE��C �D�A��

A��C�DE�A��B��D��D�AC��C�A

ED��D�D�A�C� �D�� D�� //-� "�C�

��D�� D�� A�� D�� DC

&(&�'��Q&(&=�(�

��A�B��

��D�� D��

�� B�� C,=� �E�� =,'�

C��F�A��CA��

C �D�A�� D��

5D�B��E��E� C��

� ��D�� DC��

�� A��EE� C�� A��

�A�� CA��

E� D�� D�� DC��

��A��EE� C��A��

�A�� CA��

��C��E� D�� D��

��D��A��EE�C��

F�A�� CA��

� ��C��E� D��

5D�B��'�E��E�C��

� � ��C� ��F��0�

�� A�� C �D�A��

��B�E(�

�AF��EA��

��EA��2��F��

C �D�A�� D��

A�E� �*��B��A��

(��D��DC��

��A��E��A�F�A��

DE�� F�A�� //-�

��A��C�DE�

,#�DC�� M,C�EC(�

�AE� �� ED��D�D�A�C�

�D�A�� D��

�C��EAF��A��B�(

��D�� C�� A��

E��A�F�A��//-�

�� C�� EAF�� A��

F�A�� //-� E��

��D��D��

��C�AC�C��AE��

"�C�� C �D�A��

�� DC�� ?!?K�

/A"��C(��.�F�B�E�E�F��D��A

�C��A��D�F��E�E��D�A�2��

��E�EA��E��BD

�

�A�AF�C��E

O�

��

��D��

�E�

/��

:�A��

//-� C�N�=�

=�N�'�

'�N��

C(+��

='('��

'�(�J�

�(C'

+('&

+(+E

/F�� C�N�=�

=�N�'�

'�N��

�(C&�

J(&&�

�E('&�

(C

�(&&

(C

�*�� C�N�=�

=�N�'�

'�N��

J+(C'�

&&(&�

&&(&�

E(CC

&(&&

&(&&

�*��D�� C�N�=�

=�N�'�

'�N��

��(+C�

=+(J&�

C=(E��

'(J'

C(&

�(�=

�*�� C�N�=�

=�N�'�

'�N��

J(�&�

�C(=�

EE(J��

�(J&

E(�E

=(&

�*��0�� C�N�=�

=�N�'�

'�N��

�(++�

'(JC�

��(&=�

=(

�(C�

(JE

O//-,/ ��/�1��-ACD��/F�,/�CA��FB

�*��,� ��CA�� *�AFFACD�A��

�*��D�, ��CA�� *�AFFACD�A�

��C��E�� *��, ��CA�� *�

��F��B�� C��E�� *��0�

*�AFFACD�A�� C��F��0��C��

��

/�� C �D�A�� D��

��B��'�N�� A�E��*��B��E��A��

E�� '�(�J(� ��D�� DC��

��B�� *��B�� HE�� A�� E�� A� F�

=E(=(� #�� C�� C �D�A��

��D��DC��B�E��F

C�� AE� �� ED��D�D�A�C� �D��

��B�E��?!?�K&(&�+��Q&(&=�(�

�

��F�A�DE��F�A��//-�

��D�� DC�� A�� B�E� �

:A��,#�DC�� M,C�EC(�/��AE��A

A��B�E�AF��C��EB"6��CE��*�

A�� *��B�� HE(� 9C� �AE� �"E��

ED��D�D�A�C� �D�� D��//-� E��

��F�*��B��H� C��*��B��H��?!?K�

ED��D�D�A�C��D��AE��B��"�C�

A��*��B��HE� �?!?K�&(EC�� Q&(&=�(

�A��D��A�B�E��F�A��C�� B�

��CDEF��B��E��

�E�

D��A�AF��EA��

2��EA��2��F��

��BD��

��

� :�A��

�(C'�

+('&�

+(+E�

E'(��

C'(JJ�

=E(=�

=(C��

C(&+�

J(++�

(C�

�(&&�

(C�

C(�&�

�(C&�

(C&�

('C�

(=�

(+�

E(CC�

&(&&�

&(&&�

CE(�E�

=J(E+�

'=(��

&(=

C�

J(�C�

+(&'�

'(J'�

C(&�

�(�=�

JE(EE

+=(J�

��(='�

J(��

J(�J�

�(+E�

�(J&�

E(�E�

=(&�

'(''�

&(&��

�(=�

J(��

=(�C�

=(�E�

=(�

�(C��

(JE�

&(&�

C(��

C(J�

&(&�

'(C+�

C(=E�

�CA��FBF"��C��E��

D�A�� C� ��C��E��

ACD�A�� C� �DF��

�� *�AFFACD�A�� C�

�0�,� ��CA��

��C��E

�� D�� C�� A��

�E��A�F�A��//-�

� �� C�� EAF�� A��

� A� F�A�� //-� E��

��D�� D�� A��

��F�A��ECACDECD�A��

�� "�C�� C�� C��

�//-�E��"�C�� C��

B�E� �AE� ECB�D�� BED��

AE��A�DACD��D��D��C�

��*��B��H��*��B��H��

E�� C�AC�� C�� AE�

E�� AE� A�� EE��

?!?K�&(&'��Q&(&=�(�F��

��"�C��C��*��B��H��

&(&=�(�/�� B�E�

�� B��B�(�

��E��A3��3--- 455,&!5 .��&6 C*.

�E��

&��C0��!��E�A��A��'&0E/#

5D�B��(��E��E:B��E��CEA�EAFE

��C��A��D�F��E�E��D�A�2�

�F��E��BD��

/�� C �D�A�� D�� D��

��B��C�N�=� �A�E��*��B��A��E��

�BF"��E��C��E��(C&(�/��D��

�� C��EAF��A��B��*��B��H��A��

C�CA�� BF"�� E��C��E� �� C(�&(� ��

F�A�� C�CA�� BF"�� E��C��E� "�C�

��D�� D�� A�� C�DE� A

�A��D�� BC� BED�� :A��,#�DC�� M,C�E

E�� C�AC� C�� AE� ED��D�D�A�

ECACDECD�A�� "�C�� C �D�A�� D�

��D��DC��C��EAF��A��B

�Q&(&=�(�

��

/�� C �D�A�� D�� D��

��B��=�N�'� �A�E��*��B��A��E��

�BF"��E��C��E��J(&(�/��D��

C�� EAF�� A�� B�� *��B�� H�� A�� E

C�CA�� BF"�� E��C��E�� (C&(�/��

C �D�A�� D�� D�� A��

��B��D��D�AC��C�AC�C��AE�ED��D�D�A�

C�CA��BF"��E��C��E�"�C��C��C�

&(&&J��Q&(&=�(�

�

/�� C �D�A�� D�� D��

��B��'�N�� A�E��*��B��E��A��E��

�BF"��E��C��E��E('&(�/��D��

�� C��EAF��A��B��*��B��HE��A��

C�CA��BF"��E��C��E��(C&(�#��

��D�� D�� A��B�E��

ECACDECD�A�� C�� AE� ED��D�D�A�C� �D��

C��C��B�E��?!?K�&(&&J��Q&(&=�(�

�

�� F�A�DE�� F�A�� C�CA

E��C��E� "�C�� C�� D��DC��

�AE� ECB�D�� BED�� :A��,#�DC�� M,C�

�A�DACD��D��D��C�A��B�E�AF��

*��B��H��*��B��H�� A��*��B��HE(� 9C

C�AC� C�� AE� ED��D�D�A�C� �D�� D��

C�CA�� BF"�� E��C��E� AE� A��

*��B�� H� C�� *��B�� H�� ?!?K� &(&�=��

�

��

��

��

�� A

��A�B�

C�DE�DF��

��

C*. �--C7 )��E

EAFE��AF��EA

�2��EA��2�

��D�� C�� A��

�E��A�F�A��C�CA��

��D��DC��

��A��E��A�F�A��

&(� �� F�A�DE��

E� "�C�� C �D�A��

� C�DE� A�� B�� AE�

C�EC(� /�� EB�CE�

�D�D�A�C� �D��

��D��D��A��

��B��?!?K�&(&&+��

��D�� C�� A��


�D��DC��

�A�� E�� A� F�A��

(�/�� F�A�DE��

�� C�DE� A��

�D�D�A�C��D��D��

�C��C��B�E��?!?K�

��D�� C�� A��


��D��DC��

��A��E��A�F�A��

(�#��C��C �D�A��

B�E�� F�A��

� �D�� "�C��

�� C�CA�� BF"��

DC�� A�� B�E�

C�EC(� /�� AE�

F��C��EB"6��CE��

HE(� 9C��AE� �"E��

�� D�� CA��

�� EE�� F�

�=�� Q&(&=�� A�� C��

*��B�� HE� �?!?K� &(&�� Q&(&=�(� /�

"�CC��E��E��F�A��C��C�� B��

-��A�� 1��CCD �

'-1�E/

5D�B��E(�)E�A.E�FEA��E��1��

CEA�EAFE��.12C��AF��EA��

�E�E��D�A�2��EA��2��F��

/�� C �D�A�� D��

��B��C�N�=� �A�E��*��B��A��E��

��J+(C'(�/��D��DC��C

�*��B�� H�� A�� E�� A� F�A�� *

��F�A�DE�� F�A�� *�� "

��D�� D�� A�� C�

�A��D�� BC� BED�� :A��,#�DC�� M

E�� C�AC� C�� AE� ED��D

ECACDECD�A�� "�C�� C �D�A��

��D��DC��C��EAF��A��

�Q&(&=�(�

/�� C �D�A�� D��

��B��=�N�'� �A�E��*��B��A��E��

��&&(&(�/��D��DC��C

�*��B�� H�� A�� E�� A� F�A�� *

��F�A�DE�� C �D�A�� D��

�� C�DE� A�� B�� D��D�AC�� C�AC� C��

�D�� D�� *�� "�C�� C �

��D��A��D��DC��?!?K

�

/�� C �D�A�� D��

��B��'�N�� A�E��*��B��E��A��E��

��&&(&(�/��D��DC��C

�*��B��HE��A��E��A�F�A��*�

C��C �D�A�� D��D��A��

��F�A��ECACDECD�A�� C��AE�ED�

"�C��C��C��B�E��?!?K�&(&&=��

�

�� F�A�DE�� F�A�� *

��A��B�E��AE�ECB�D��BED��

C�EC(�/��AE��A�DACD��D��D��C�

C��EB"6��CE��*��B��H��*��B��H�

�AE��"E��C�AC�C��AE��ED��D�

�*��AE�A��EE��F�*��B

�?!?K�&('��Q&(&=�(��D��D�D�A�C��D�

"�C�� C�� *��B�� H�� A�� *��B�

BCD

EDF

�

��

��

��

��F�DE��B�

��C

��

=�(� /�� B�� AE�

�B��B�E(�

CCD �� E�A��A��

��F��2��EF�

��C��A��D�F��

�F��E��BD��

�� D�� C�� A��

A��E��A�F�A��*��

��C��EAF��A��B��

�A�� *�� =�(�E(� ��

�� "�C�� C �D�A��

� �� C�DE� A�� B�� AE�

�� M,C�EC(� /�� EB�CE�

ED��D�D�A�C� �D��

��D��D��A��

��A��B��?!?K�&(&&��

D�� D�� C�� A��

A��E��A�F�A��*��

��C��EAF��A��B��

A�� *�� 'J(E+(� /��

��D��D��A��

AC� C�� AE� ED��D�D�A�C�

�� C �D�A�� D��

?!?K�&(&&=��Q&(&=�(�

D�� D�� C�� A��

A��E��A�F�A��*��

��C��EAF��A��B��

��*��+=(�(�#��

��A��B�E��

AE�ED��D�D�A�C��D��

(&&=��Q&(&=�(�

�A�� *�� "�C�� C��

BED��:A��,#�DC�� M,

��C�A��B�E�AF��

B��H��A��*��B��HE(�9C�

�ED��D�D�A�C��D��D��

�*��B��H� C��*��B��H��

�C��D��AE� ��B��

*��B�� HE� �?!?K� &(&�+��

�� A

��

BCD

EDF

�

�

�Q&(&=�(�/��B��AE�"�CC��E��

C�� B��B�E(�

-��A�� 1��CCD ��

E�A��A��'-1�E�E/

5D�B��C(�)E�A.E�FEA��E��1��

C��D�ECEA�EAFE��.12C�C��AF��EA

��D�F��E�E��D�A�2��EA��2�

��BD��

/�� C �D�A�� D�� D��

��B�� C,=� �A�E� �*��B�� A�� E�

�*��D��(+C(�/��D��DC��

A�� B�� *��B��H��A�� E��A�F�

JE(EE(� �� F�A�DE�� F�A�� *�

C �D�A�� D�� D�� A��

��B��AE��A��D��BC�BED��:A��,#�DC�

��EB�CE� E�� C�AC� C�� AE� ED��D�D�



�Q&(&=�(�

�

/�� C �D�A�� D�� D��

��B�� =,'� �A�E� �*��B�� A�� E�

�*��D��=+(J&(�/��D��DC��

A�� B�� *��B��H��A�� E��A�F�

+=(J(�/��F�A�DE��C �D�A��

A�� C�DE� A�� B�� D��D�AC�� C

ED��D�D�A�C� �D�� D�� *��D�� "�C�

��D�� D�� A�� D�� DC

&(&&J��Q&(&=�(�

�

/�� C �D�A�� D�� D��

��B�� '� N� �� A�E� �*��B�� E�� A�� E

�*��D��C=(E�C(�/��D��DC��

A�� B�� *��B��HE��A�� E��A�F�

��(='J(� #�� C�� C �D�A�� D��

�� B�E� �� F�A�� ECACDECD�A

ED��D�D�A�C� �D�� "�C�� C�� C��

&(&&J��Q&(&=�(�

�

/�� D�� DC�� A��

C�A�� C�� C �D�A�� D�� D�� D�

��B�E�� ED�� F�EC� �� C�� E��C��E� ��

��D��DC��EDF��E��C��E

��A�B��

��E��E��F�A��C��

�� E��FC�

��F��2��EF�

�EA��C��A�

�2��F��E

��D�� C�� A��

A�� E�� A� F�A��

DC��C��EAF��

� A�F�A��*��D��

� �*��D�� "�C��

�� C�DE� A��

#�DC�� M,C�EC(�/��

D��D�D�A�C� �D��

��D��D��A��

��B��?!?K�&(&&+��

��D�� C�� A��

A�� E�� A� F�A��


� A�F�A��*��D��

��D��D��

AC�� C�AC� C�� AE�

� "�C�� C �D�A��

�� DC�� ?!?K�

� ��D�� C�� A��

�A�� E�� A� F�A��


� A�F�A��*��D��

��D�� D�� A��

CDECD�A�� C�� AE�

�� C�� B�E� �?!?K�

� �A�� "�CC�� E��E�

�� D�� A�� C�� A��

�E� �� BCC�� " � C��

C��E(�

�� F�A�DE�� F�A�� *

��D�� DC�� A�� B�E� �

:A��,#�DC�� M,C�EC(�/��AE��

�D��C�A��B�E�AF�� C��EB"6

*��B�� H�� A��*��B�� HE(� 9C��AE� �"

�AE� �� ED��D�D�A�C� �D�� D��

��EE�� F� *��B�� H� C�� *��B

�Q&(&=�(�F�� ED��D�D�A�C� �D��

C��*��B��H��A��*��B��HE�A�E��

/�� B�� B�� AE� "�CC�� E��

��B�E(

-��A�� 1��CCD ��

E�A��A��'-1��E/

5D�B��=(�)E�A.E�FEA��E��1��

2��D�BA�CEA�EAFE��.122�C��AF

�A��D�F��E�E��D�A�2��EA

��E��BD��

�

/�� C �D�A�� D��

��B�� C� N� =� �A�E� �*��B��

�*��J(�&(�/��D��DC�

A��B��*��B��H��A��E��A�

'(''(� �� F�A�DE�� F�A��

C �D�A�� D�� D�� A��

��B��AE��A��D��BC�BED��:A��,#

��EB�CE� E�� C�AC� C�� AE� ED��

ECACDECD�A�� "�C�� C �D�A��

��D��DC��C��EAF��A��

�Q&(&=�(�

�

/�� C �D�A�� D��

��B�� =� N� '� �A�E� �*��B��

�*��C(=(�/��D��DC�

A��B��*��B��H��A��E��A�

&(&'(�/��F�A�DE��C �D�A��

A�� C�DE� A�� B�� D��D�AC

ED��D�D�A�C� �D�� D�� *��

��D�� D�� A�� D��

&(&&J��Q&(&=�(�

�

/�� C �D�A�� D��

��B�� '� N� �� A�E� �*��B�� E�� A

�*�� EE(J�(� /�� C�

�*��B�� HE�� A�� E�� A� F�A��

#�� C�� C �D�A�� D�� D�

�DC�� B�E� �� F�A��

�

��

��

��

��


��C

��

��CDEF��B��E��

�EE�

A�� *��D�� "�C�� C��


AE��FB��A�DACD��D��

EB"6��CE��*��B��H��

AE� �"E�� C�AC�� C��

�� D�� *��D�� AE� A��

� *��B�� H�� ?!?K� &(�+&��

��AE� ��B��"�C��

�E��?!?K�&(��Q&(&=�(�

� E��E� ��F�A�� C�� C��

CD �� F��A�

��F��2��EF�

��AF��EA��C�

�EA��2��F��E

�� D�� C�� A��

�� A�� E�� A� F�A��

��DC��C��EAF��

��A�F�A��*��

�A�� *�� "�C��

� A�� C�DE� A��

#�DC�� M,C�EC(�/��

E� ED��D�D�A�C� �D��

��D��D��A��

��A��B��?!?K�&(&��

�� D�� C�� A��

�� A�� E�� A� F�A��


��A�F�A��*��

A�� D��D��

D��D�AC�� C�AC� C�� AE�

�� "�C�� C �D�A��

D�� DC�� ?!?K�

�� D�� C�� A��

E�� A�� E�� A� F�A��

� �� C�� EAF�� A�� B��

�A�� *�� (=(�

�� D�� A�� D��

�A�� ECACDECD�A�� C��

�� A

��

BCD

EDF

��E��A3��3--- 455,&!5 .��&6 C*.

�EC�

�

�AE�ED��D�D�A�C��D��"�C��C��C�

&(&&J��Q&(&=�(�

�� F�A�DE�� F�A�� *��

��D�� DC�� A�� B�E� �AE�

:A��,#�DC�� M,C�EC(�/��AE��FB

�D��C�A��B�E�AF�� C��EB"6��CE

*��B�� H�� A�� *��B�� HE(� 9C� �AE� �"E��

�AE� �� ED��D�D�A�C� �D�� D�� *�

��EE�� F� *��B�� H� C�� *��B�� H

�Q&(&=�(� F�� ED��D�D�A�C� �D�� AE�

C��*��B��H��A��*��B��HE�A�E��?!?K�

/�� B�� AE� "�CC�� E��E� ��F

�B��B�E(�

-��A�� 1��CCD ��

E�A��A��'-1��2E/�

5D�B��'(�)E�A.E�FEA��E��1��

2��D�E0CEA�EAFE��.1220C��AF��

�A��D�F��E�E��D�A�2��EA��

��E��BD��

/�� C �D�A�� D�� D��

��B�� C� N� =� �A�E� �*��B�� A�� E

�*��0��(++(�/��D��DC��

A��B��*��B��H��A��E��A�F�A

&(&(��F�A�DE��F�A��*��0��"�C

��D�� D�� A�� C�DE� A

�A��D�� BC� BED�� :A��,#�DC�� M,C�E

E�� C�AC� C�� AE� �"E��A"



�Q&(&=�(�

�

/�� C �D�A�� D�� D��

��B�� =� N� '� �A�E� �*��B�� A�� E

�*��0��'(JC(�/��D��DC��

A��B��*��B��H��A��E��A�F�A

C(��(�/��F�A�DE��C �D�A��

A�� C�DE� A�� B�� D��D�AC�� C

ED��D�D�A�C� �D�� D�� *��0�� "�C�

��D�� D�� A�� D�� DC

&(&&+��Q&(&=�(�

�

/�� C �D�A�� D�� D��

��B�� '� N� �� A�E� �*��B�� E�� A�� E

�

��

��

��

��

��


��C

��

C*. �--C7 )��E

�C��C��B�E��?!?K�

�� "�C�� C��

�AE� ECB�D�� BED��

��FB��A�DACD��D��

"6��CE��*��B��H��

� �"E�� C�AC� C��

�� *�� AE� A��

�B�� H�� ?!?K� &(+�+��

� �AE� C�� "�C��

?!?K�&(&�=��Q&(&=�(�

�E� ��F�A�� C�� C��

�� FC�2

��F��2��EF�

��EA��C�

��2��F��E

��D�� C�� A��

�A�� E�� A� F�A��

C�� C��EAF��

�A�F�A��*��0��

0��"�C��C �D�A��

� C�DE� A�� B�� AE�

C�EC(� /�� EB�CE�

��A"�� D��

��D��D��A��

��B��?!?K�&(&J��

��D�� C�� A��

A�� E�� A� F�A��


�A�F�A��*��0��

��D��D��

AC�� C�AC� C�� AE�

�� "�C�� C �D�A��

�� DC�� ?!?K�

��D�� C�� A��

A�� E�� A� F�A��

�*��0��(&=(�/��D��DC�

A��B��*��B��HE��A��E��A�

C(J(�#��C��C �D�A�� D��

��B�E� �� F�A�� ECACDECD�

ED��D�D�A�C� �D�� "�C�� C��

&(&&J��Q&(&=�(�

�

�� F�A�DE�� F�A�� *�

��D�� DC�� A�� B�E� �

:A��,#�DC�� M,C�EC(� /�� AE

�D��C�A��B�E�AF�� C��EB"6

*��B�� H�� A�� *��B�� HE(� 9C� �AE� �

�AE� �� ED��D�D�A�C� �D�� D��

��EE�� F� *��B�� H� C�� *��B

�Q&(&=�(�F�� ED��D�D�A�C� �D��

C��*��B��H��A��*��B��HE�A�E��

/�� B�� AE� "�CC�� E��E�

�B��B�E(�

�

��D��DC��B��

��F��0� BCC��A��E� C�A�� C�� C �

��D��(� �DFD�A�� EB�CE� ��

A�� %R�� J'�� .� �B��AF

��D�� ABCDECD�� D�� C� A"

EC�B�CB�� F�A�� C�� C �D�A��

HA�� J�+�� C�� C�AC� ��D��

�D��D�B�C �D��B�D��E �CA�CD�A��

E��C��E(� M��ECA��D�� A�C

E��C��E� D�� D�� DC�� DE

� 1��E��8��1FA��#ACE��H��A��

��A�C��A�

/�� D�� BED��E� �

ECB� 4� ��D�� DC�� DE� �A�� :

C�A��FA�E� D�� A�� C�� A�� B�E��

A�� 0��EED��F��E �CA0(�/�

�� C�� D�� DC�� F

:��FD�� C�B�CB�� EB"C�EC��

��B��E��D��D��C��6B��CD�

EB"C�EC��:8/�CAE1(��D��DC��

��B�E� ��F�� "�CC�� D�� C��

�C�B�CB��:��EB"C�EC��" �/

�:�� /�� A�� 5(� /��

�D��E��E�D��F�EC��C��:8/�CAE

�B�� B�� D�� E� ��

"BC� >BA�DCACD�� C�� F� �D��

��FB��D��ECACDECD�A�� "�C�

�� A�� C �D�A�� D�� D��

��B�� D�� C�� //-� �A�AF�C��

CAE1(� �� "E��A"�� D�� A

E��AE�A��EE��F�*��B�

/��AE� ED��D�D�A�C� �D�� "�C

�DC��A��C �D�A�� D��

F�(��E��C��E��CA��*�AF

A

BCD

EDF


��A�F�A��*��0��

��D��D��A��

CACDECD�A�� C�� AE�

�� C�� C�� B�E� �?!?K�

� �*��0�� "�C�� C��


� �AE� �� A�DACD�� D��

��EB"6��CE��*��B��H��

�AE� �"E�� C�AC� C��

�� D�� *��0�� AE� A��

� *��B�� H�� !?K� &(E'��

��AE� ��B��"�C��

�E��?!?K�&(J��Q&(&=�(�

E��E� ��F�A�� C�� C��

�B�� E �CA�CD�A�� EE�

�� C �D�A�� D��

� ��C�� " � $��F��D��

F��,��DC�� J'J�� D��

�C� A"�� C�� BE�� E��C��

�� D�� D��(�

��DC�� ABCDEF� �A��

D�A�� C��A�ACD��

�� A�CD�� A�� AEED��

DE� ��C�� " � �AB��

��A��J+��(�

��A�

��E� �� A�� C��

�A�� :8/� E��E� ��

�B�E�� "�C�� CD��

/�� B�� B�E�

��F�� "�CC�� D�� C��

"C�EC��AE� C��

6B��CD��E�A��GB�CACD��E�

��DC��C��A��A��

� C�� :��FD��

��" �/9��G��/��F*��

/�� B�E� �A��

8/�CAE1��F�A��C��C��

E� ��FA�� F��C��

� �D��C� (� /�� AE�

�"�C��D��DC��

� ��D�� A�� C�� A��

� �� D�CB�� E��D�CD��

�� AE� ��B�� D�� //-�

*��B��H�C��*��B��H�(�

�� "�C�� C�� D��

�D��D��D��C��C�CA��

��*�AFFACD�A�� C�


�

�E=��

��C��E�� CA�� *�AFFACD�A�� C�

�DF�� C��E�� CA�� *�AFFACD�A��

��C� ��F��B�� C��E�� A�� CA��

*�AFFACD�A�� C� ��F��0� ��C��E� �� A�� C��

A��B�E(�/��D��DC��E��C�A��

C�� C �D�A�� D��D�� D��A�� C��A��B�E�

A�� A�� C�� A�AF�C��E� �0��C� �� CA��

*�AFFACD�A�� C� �DF�� C��E(� 9�� C��

��CA��*�AFFACD�A�� C��DF��C��E�

�A�AF�C�� D�� DC�� A�� "�CC�� E��E� C�A��

C�� C �D�A�� D�� D�� D�� A�� C�� A��

��B�E�� ED�� F�EC� �� C�� E��C��E� BCC�� " � C��

��D��DC��EDF��E��C��E(�

�

9�� A�� A�AF�C��E�� C�� B�� A�� "�CC��

E��E� ��F�A�� C�� C�� B�� A�� B�E(� /��

��F��0DC � �� E��C��E� ��B�� " � ��D�� DC��

��D��AE��A��DF��A�BA�� DC��A��(��C��

��B�� D��DC��A��A��AFFACD�� E��

D��F�A�DE��DC��C �D�A�� D��D��(��

�

)��A��

�B��AF��,��DC��M(��J'J�(�<F��AEDE�A��F�A�D��D��A��

D�� FA�� A�� ABCDECD�� D��(� *�A�B��E �A� CDEEF� �

��J,E+(�

HA�� S(� �J�+�(� �� A�FACD�� A�A� EDE� �� ABCDECD�� D��RE�

�A��BA�� DC�� E��C� C�� A��AED�� A�� FA�� A��BA��

��F��C(� M��B"�DE�� C��A�� DEE��CACD��

:��"�B��M�D��EDC ��:��"�B��BEC�A�DA(�

HA�CA1�� 8(�� -BCC�� :(�� .� ��0�� (� �J��(� �� F�A�ACD��

ECB� � �� D��A�CD�� ABCDEF� A�� E��D�D�� F��CA��

��CD�� A��BA�� DE��E(� 999(� �DE��DFD�A�C� 5B��CD��

��A� EDE(� ;�B�A�� B�� A� 2��

CF��/�D��EA��(��E+E,EJ'(�

�A�C��(��HA1��8(��.�-BCC��:(��J�+�(��F�A�ACD��

ECB� � �� D��A�CD�� ABCDEF� A�� E��D�D�� F��CA��

��CD�� A��BA�� DE��(� 9A(��A� EDE� �� E �CA0� A��

�A��BA�� B��CD��(� ;�B�A�� 2�� .��F�� A�

.��F��!,��E=,E'�(�

*�E��D�� (� $(� ��&&+�(� �BCDEF4� FA� � ��E�� FF��

�AC��A ET�2E��E=��E��EJNEJ=(�

$��F��D��.�%R��J'��(��DCD��D�DCE�D��D��(��

6��)E��F��6B��E��A��,�E(�

PA�� 8(� �JC'�(� 9��A�C� A��F�CA��D�A�� A��BA�� D��

�A�� D��A�CD�� ABCDEF(� ��E��F�A ;�B�A�� .��F��

!5# ��C�,�C'(�

P�� :(��(�� &&��(� 9��*(�*��EC�D��:(�$��E��(��.��(�-(�

H��E(��<�E(�(�2��D�E�EA��E��A��D��F��F��

��E��EA�(�F��S��E� 4�S��#D�� .��E(�

P6��AA�� :(�� .� /A��,5�BE"�� $(� ��&&�(� ��

D��ECD�ACD�� A��BA�� DF�AD�F��C� D�� ABCDEF4�

9F��D�ACD��E� �� CD�� EB"��B�E� *�A�B��E �A�

2��A��E.��FE��E��!'��+�NE&+(�

�AB��-(�� 1��E��(��8��1FA��S(��#ACE��:(��H��#(��.�

�� (� �J+��(� �� F�A�DE�� A��BA��

��A�A�C��DECD�E��F��CA�� CA��A�B�CE��DC��A�D��L�

E ��F��A�� C��E��DC��E��D�D��F��CA�� CA��ACD��

A�� ABCDEF(� ;�B�A�� B�� A� �E�E��D�EA��

��E��C=�,C'+(�

�D��(��.�HA�C��B��D��*(� �J��(��E �CA�CD�� D��ECD�ACD��

��"A��ABCDECD��F��CA�� CA��A��FA��D��(�

;�B�A��B��A�2��CF��/�D��EA��(��,

EC(�

-A�D�� 9(�� .� /B��FA�� -(� 5(� ��&&+�(� �BCDEF4� ��D�DCD��

��B��"D�� E��D�� DA��EDE(� .E��F 2��A�F� ��

:��E��F��==��=��JNC'(�

-D�1E��(�:��.�#D��8(� �J�=�(�8A��BA��FFB�D�ACD��

A�� C�� BE�� E F"��E� D�� FA�� A�� ABCDECD�� D��(�

;�B�A��B��A�2��CF��/�D��EA��=��J,

��(�

-�"D�E��(�8(��&&�(�/��:��D�D��,��1�DEC��BCDEF�D��

/��E��:,�$�/�4��D�DCDA��D��ECD�ACD��D��C��A��

��C��CD�� ABCDEF� A�� AED�� F��CA��

�DE��E� ;�B�A�� B�� A� �E�E��D�EA��

��E��E��E,CC(�

-B1FD�D�� (� �(� �JJC�(� :A�A A�AF� �A��BA�� C�EC(� �A

B��B"�DE�� :AEC��DE� �DEE��CACD�� EB"FDC�� C�� C��

M�D��EDC ��: E��: E��(�

�� AFA�A�� P(� �(�� B��DC��A�� (�� .� AD6A AE�� &&+�(�

�D��CDA�� DA��EDE� ��1�DEC� ��,��(�

M��B"�DE��6��C��C��99�$��E�A��B��(�

#D��8(��<�E(�(��J�'�(�<A�� D��ABCDEF(�F��U��14�

��AF��EE(�

#�� $�A�C�� %��A�DEACD�� &&'�(� 9�C��ACD��A�� ECACDECD�A��

��AEED�D�ACD�� DE�AE�E� A��-��AC��$�A�C�� "��FE(�

&C�� <�� 9��,&�(� @�DC�� &&�,&',�=B5+C(� ��AED��

��F��CA��DE��E(�

#�� $�A�C�� %��A�D!ACD��(� 5+C(� ��&&��(� ��AED��

��F��CA��DE��E(�@�DC��&&J,&,&B(�

�


236

Written Language Skills in Children with Dyslexia Sheetal R.1 & Sangeetha Mahesh2

Abstract

The aim of the study was to provide preliminary data for establishing grade level guidelines of performance on several writing variables. In addition it also provides comparison and differences found between children with dyslexia and typically developing children (TDC) from first to fifth grade. An efficient system for evaluating written language behaviors in children with dyslexia has been provided. The study included two groups of participants. The first group included TDC from grade 1 to grade 5 and second group included 8 children with dyslexia. The study was carried out in two phases. The first phase included developing norms for TDC across grades for the written language variables. The second phase included comparing within the group of children with dyslexia and between the groups of children with dyslexia and TDC. Eight variables considered for the study included Total number of Words, Total number of T-units, Mean Length of T-units, Number of Clauses, Clause Density, Percentage of Grammatical T-units, Percentage of spelling errors and Errors in Writing Conventions. Children with dyslexia showed poorer scores compared to TDC in almost all the aspects of productivity, complexity and accuracy except for MLT-UNIT. However, the high variability within children of the same grade could be attributed to the severity and type of dyslexia and to the treatment effects. A multidimensional database of variables that contribute to overall writing proficiency across grades is needed to provide assessment and remediation in school age children.

Key words: written language, dyslexia, typically developing children

riting 1is an essential means of communication and helps children to organize their thoughts in a structured way.

Writing is one of the most complex of all the language behaviours and students of all ages have difficulties becoming proficient in writing. Some children with writing difficulties have language or learning disabilities, whereas others do not.

Writing is a cognitive activity. It involves use of

executive process and self regulation all throughout the writing activity which includes four stages: planning, organizing, generating and revising. Planning and organizing stages are internal processes which cannot be measured explicitly. Generating stage is where text is generated where the writer translates his or her planned ideas into meaningful chunks of sentences, phrases, words into written symbols and includes spelling, handwriting and punctuation (Scott, 2005).

In the early primary grades text generation and

writing quality are most constrained by a child‟s handwriting fluency (Berninger & Swanson, 1994). Because children who have not yet mastered handwriting must direct attention to letter formation, they do not generate much text. By the intermediate grades, when handwriting is automatised for most children, its constraint on text generation is minimized and texts become longer. With age, text length and

1e-mail:[email protected]; 2Clinical Lecturer, AIISH, Mysore, [email protected]

quality become increasingly related (Berninger & Swanson, 1994; Shanbal & Prema, 2003).

Writing is a multifaceted activity ranging from

the production of legible handwriting to the production of organized discourse. It is important to know the components of written language in order to plan appropriate instructional interventions for children with writing difficulties. The components of written language are handwriting, spelling, usage, vocabulary and text structure. These skills are interlinked to one another and thus help in forming a structured and organized written text (Mather, Wendling & Roberts, 2009).

International Dyslexia Association (2002)

defined dyslexia as a specific learning disability that is neurological in origin which is characterized by difficulties with accurate and/or fluent word recognition and by poor spelling and decoding abilities. These difficulties typically result from a deficit in the phonological component of language that is often unexpected in relation to other cognitive abilities and the provision of effective classroom instruction. Secondary consequences may include problems in reading comprehension and reduced reading experience that can impede growth of vocabulary and background knowledge.

A disorder of written language is termed as

dysgraphia by Johnson and Myklebust (1967). As we know that writing is a multifaceted activity, it calls for a classification of written language disorders. Written

W

Written language skills in children with dyslexia

237

language deficits can be due to inappropriate visuo motor integration, due to deficits in revisualization and can be due to formulation difficulties in syntax.

From past few decades research is gradually

gearing up on reading and writing aspects of language particularly research on writing is a recent research interest for most of the Speech-Language Pathologists (SLP), psychologists and educators. Written language can be assessed using standardized instruments which test the basic skills like spelling, handwriting, vocabulary, syntax quantitatively. These standardized measures do not follow the classroom curriculum. Hence, these tests should be used in conjunction with an informal assessment by doing curriculum based measurement which allows the SLP to quantify the writing samples and measure in the way the children are taught in school by testing their basic skills.

Testing a wide variety of written language skills

at various levels of language development has important implications for assessment, diagnosis and therapy. Research indicates that children with language difficulties have problems in several aspects of written language. But, we do not have readily available data on typically developing primary children that can be used to compare the clinical samples of children‟s writing.

Most of the standardized tests check narrative

style of writing. Although this is appropriate in the primary grades, children are expected to create expositions as they move on to higher classes. Expository writing requires children to employ skills that can be particularly challenging for language disorders. Commercial test seldom cover these areas neither in any depth nor with resources adequate to scaffold them.

Also, studies on written language and

difficulties associated with written language have been very few in the Indian context. India is a multilingual country and majority of the children are exposed to their mother tongue and have English as their medium of instruction in school. Thus, there is a need to study the language aspects of writing in typically developing children (TDC) and also to derive normative for different grades in TDC in English. The development of norms can be used as an index for comparing writing samples of TDC and the clinical groups particularly in children with dyslexia. Clinically, this study will help SLPs to screen and to identify their difficulties and design appropriate management procedures relevant to linguistic aspects of writing. Further, the results can be compared and differentially diagnosed with other types of dyslexia. Hence, the present study has been planned to meet the following

objectives: 1) To derive a normative data for the written language skills using expository style of writing, 2) To study the pattern of written language skills in children with dyslexia, 3) To perform quantitative analysis of written language skills in both the groups and 4) To compare and analyze the written language skills in TDC and children with dyslexia.

Method

The current study was taken up to investigate the written language skills in children with dyslexia. The study was conducted in two phases. First phase aimed at developing a normative data. In second phase, a comparison of written language skills in TDC and in children with dyslexia was taken up. It is essential as norms with reference to the topic „My School‟ have not been established in Indian children. Participants: The study included two groups of participants. Group 1 included TDC and group 2 included children with dyslexia. These children were students from first grade to fifth grade having English as their medium of instruction. All children selected had Kannada as their native language. All children were right handed. Group 1: Inclusionary criteria for TDC: They were selected from different schools in Mysore City. All the children in the group were screened for having any history of in (a) spoken language (b) sensory, motor or other neurological functions and (c) behavior including attentional deficits. This group comprised of 27 children from first grade, 30 from second grade, 28 from third grade, 30 from fourth grade and 29 from fifth grade. Group 2: Inclusionary criteria for children with Dyslexia: For the second phase of the study, participants were children who had reported to AIISH and were diagnosed as having dyslexia according to evaluation reports. Participants were diagnosed as dyslexia considering the definition given by International Dyslexia Association (2002) which defines dyslexia as a specific learning disability which is characterized by difficulties with accurate and/or fluent word recognition and by poor spelling and decoding abilities. Eight children with dyslexia participated for the study. It comprised of 1 child from first grade, 2 from third grade, 2 from fourth grade and 3 from fifth grade.

Task and procedure: Participants were instructed to produce a text on „My school‟ in English. This topic was selected as it is familiar and easy for all children across grades. Initially a picture prompt was used. The participants were given around twenty minutes to produce the text. The written sample was collected.


238

The author read the participant‟s sample to clarify any words that are unclear due to poor penmanship or incorrect spelling.

All the written samples were transcribed by an experienced SLP and coded according to the computerized language analyzer, the Systematic Analysis of Language Transcript conventions (SALT; Miller & Chapman, 2001). After practicing and establishing coding guidelines, the written samples were coded.

Scoring: The written sample of TDC and children with dyslexia were analyzed quantitatively using the following 8 variables: 1. Total number of words (TNW): It is the total number

of words written in the text which will be automatically calculated by SALT. It was used to measure the productivity and written fluency.

2. Number of T-units (T-UNIT): A T-unit is a sentence. It is considered to be one main clause with all subordinate clauses embedded. This was used to measure syntactic complexity.

3. Mean length of T-unit (MLT-UNIT): This is a measure of syntactic complexity. This was calculated by dividing TNW by T-UNIT.

MLT-UNIT= TNW T-UNIT 4. Number of clauses (CLAUSES): A clause is a group

of words containing a subject and a predicate. This is a measure of productivity.

5. Clause density (C-DENSITY): This was calculated by the ratio of CLAUSES to T-UNITS. This is another measure of syntactic complexity.

C-DENSITY= CLAUSES T-UNIT 6. Percentage of grammatical T-units (GRAM T-UNIT):

This was calculated by using the ratio of number of T-units without errors divided by the total number of T-units in the sample.

Gram T- unit= Num T-units without errors Total number of T-units 7. Percentage of spelling errors (SPELL): SPELL is

calculated by dividing the number of spelling errors by TNW. A word was counted as a spelling error only once if the same spelling is used every time. However, if the word is spelt differently every time then each incorrect spelling was considered as an error. This is a measure of accuracy.

SPELL= Number of spelling errors TNW 8. Errors in writing conventions (CONVEN): This was

a measure to check the appropriate use of punctuation marks. The total number of these errors was calculated. This is a measure of accuracy.

Statistical analysis: The numerical values obtained after quantitative analysis done using SALT was subjected to statistical analysis using SPSS software (version 10). The tabulated scores were used to obtain the mean, standard deviation (SD) and confidence interval (CI) using MANOVA on TDC with grade as the between subject factor. Post hoc analysis was also carried out to get pair wise comparison across the grades to obtain significant difference measures using Duncan‟s Test.


In the first phase normatives were derived for TDC. In the second phase the same norms were compared with that of the scores of children with dyslexia.

A) Written language skills in TDC across grades 1. Total Number of Words (TNW): Examination of MANOVA for TNW revealed a significant main effect for grade [F (4, 139) = 25.928, p < 0.0001]. The results revealed that there is a clear-cut increasing trend in the TNW from grade 1 through grade 5. They are in agreement with previous studies of written language (Puranik, Lombardino & Altman, 2008) which suggests that there is an increase in productivity aspect of written language. The possible reason could be that the oral vocabulary shows a marked increment with increase in age which is reflected even in the written language abilities.

Table 1. Mean and Standard Deviation (SD) for TNW

across grades for TDC Grade Mean SD

I 42.33 24.76 II 61.51 37.28 III 78.25 50.66 IV 115.07 47.10 V 161.52 73.90

2. Total Number of T-units (T-UNIT): MANOVA for T-UNIT revealed a significant main effect for grade [F (4, 139) = 12.051, p < 0.0001]. There was an increasing trend in the T-UNIT from grade 1 through grade 5 (Table 2). These results are consistent with the findings of previous studies (Puranik et al., 2008) indicating that there is an increase in the productivity of written language with increase in grade. From Table 2, it is evident that there is a marked increment from 3rd grade to 4th grade because a marked shift in literacy skills occurs at grade 3 and children would have mastered the basic reading skills and begin to read for meaning (Chall, 1983).


239

Table 2. Mean and Standard Deviation (SD) for T-UNIT across grades for TDC

Grade Mean SD I 8.78 5.50 II 10.97 6.83 III 11.50 5.15 IV 15.33 4.51 V 17.79 5.58

3. Mean Length of T-unit (MLT-UNIT): MANOVA for MLT-UNIT revealed a significant main effect for grade [F (4, 139) = 23.684, p < 0.0001]. The Duncun‟s post hoc analysis was done at p <0.05 level of significance which gave the pair wise comparison between the grades. There was a clear cut increasing trend in the MLT-UNIT from grade 1 through grade 5 (Table 3). Pair wise comparison showed that children in grade 5 differed significantly from that of children in all the other grades. 4th graders showed significantly higher scores from 2nd and 1st graders. Children in 3rd grade showed significantly higher scores than children in 1st grade.

Table 3. Mean and Standard Deviation (SD) for MLT-

UNIT across grades for TDC Grade Mean SD

I 4.82 1.76 II 5.68 0.90 III 6.60 1.39 IV 7.45 1.77 V 9.09 2.72

The results are not in agreement with the

previous studies (Puranik et al., 2008). However, the results revealed that there was no significant difference between adjacent grades till 4th grade which is in partial agreement with the previous studies. This suggests that mean length of T-unit is a measure of complexity and children acquire competence in syntactic complexity slowly and steadily as a function of age and grade. 4. Total Number of Clauses (CLAUSES): MANOVA for CLAUSES revealed a significant main effect for grades [F (4, 139) = 23.472, p < 0.0001]. There was a clear cut increasing trend in the total number of clauses from grade 1 through grade 5 (Table 4). These results are in agreement with previous studies of written language (Puranik et al., 2008). There was an increase in number of clauses used with an increase in the grade indicating that there was an increase in productivity aspect of written language. The possible reason could be that with the increase in age, there was an increment in the spoken language structure in children. Children tend to use more complex and long

sentences as they grow older. A similar developmental process takes place also in written language skills.

Table 4. Mean and Standard Deviation (SD) for

CLAUSES across grades for TDC Grade Mean SD

I 7.70 5.83 II 11.53 7.31 III 13.93 7.25 IV 20.17 7.97 V 25.14 9.39

5. Clause Density (C-DENSITY): Examination of MANOVA for C-DENSITY revealed a significant main effect for grades [F (4, 139) = 24.528, p < 0.0001]. There was a clear cut increasing trend in the C-DENSITY from grade 1 through grade 5 (Table 5). However, the results are not in agreement with the previous studies. The reason could be because the other studies used retelling or narrative texts and hence the scope for them to use complex sentences would have been limited. This study employed expository style of writing where the child had to give information regarding the topic „My School‟ which is so familiar and easy for them to formulate complex sentence structure.

Table 5. Mean and Standard Deviation (SD) for C-

DENSITY across grades for TDC Grade Mean SD

I 0.81 0.33 II 1.04 0.10 III 1.17 0.23 IV 1.30 0.27 V 1.41 0.27

6. Percentage of grammatical T-units (GRAM T-UNIT): There was no significant difference seen in percentage of grammatically correct T- units across the grades [F (4, 139) = 0.526, p = 0.717]. Although, the Table 6 shows an increasing trend from 1st grade to 4th grade, it was not statistically significant. But, the variability was greatest in 1st graders compared to all the other grades. With the increase in the grade, variability decreased. The results are in agreement with the previous study done by Puranik et al., 2008. One possible explanation could be that children in higher grades attempted to write more complex sentences which was very well evident from the results on C-DENSITY. This attempt would have led to produce more errors than might be expected. Also, this study did not count the number of errors per T-unit; a T-unit was scored as error only once irrespective of whether it had one or more errors. It is highly possible that children in lower grades would have had more errors


240

per T-unit than children in higher grades. This should be considered in the future research. Table 6. Mean and Standard Deviation (SD) for GRAM

T-UNIT across grades for TDC Grade Mean SD

I 0.44 0.39 II 0.47 0.30 III 0.49 0.29 IV 0.54 0.21 V 0.52 0.23

7. Percentage of Spelling Errors (SPELL): Examination of MANOVA for SPELL revealed a significant main effect for grades, [F (4, 139) = 6.115, p < 0.0001]. There was a clear cut decreasing trend in the percentage of spelling errors from grade 1 through grade 5 (Table 7). The results indicate an improved spelling accuracy with an increase in age which is in consonance with the previous studies done by Puranik et al., 2008. Although there was a decrease in spelling errors, the errors still persisted in higher grades. The possible reason could be because spelling is the most difficult aspect of basic skills because the child had to recall and produce the entire word correctly. Even as their reading skill improves the spelling difficulties persisted, particularly in English, as it has a deep orthography. The variability in the performance of spelling decreased with increase in grade. This is an indication that with the increase in grade, spelling errors are seen mostly on „exception‟ words only which do not have a regular pattern and these have to be memorized (Mather et al., 2009).

Table 7. Mean and Standard Deviation (SD) for SPELL

across grades for TDC Grade Mean SD

I 0.10 0.10 II 0.07 0.08 III 0.06 0.06 IV 0.03 0.03 V 0.03 0.01

8. Errors in Writing Conventions (CONVEN): There was no significant difference seen in writing grades across the grades [F (4, 139) = 1.081, p = 0.398]. There was no trend seen across the grades and has high variability in all the grades (Table 8). The children at different age levels did not differ significantly on writing conventions which is in agreement with the previous study done by Puranik et al., 2008. Researchers have suggested that children at this grade level are less sensitive to the use of punctuations and they typically confine their writing to the most basic writing conventions. The variability was

higher in all the grades as reflected from the SD values from Table 8. The variability in higher grades was unexpected but, it is likely that children in higher grades would have concentrated more on the productivity and complexity of language than on the accuracy. Also, children in an attempt to write complex sentences tend to write run- on sentences without using any punctuations and connecting the clauses by conjunctions like and, or etc.

Table 8. Mean and Standard Deviation (SD) for

CONVEN across grades for TDC Grades Mean SD

I 9.74 7.77 II 7.03 5.40 III 11.18 9.72 IV 10.90 10.10 V 10.38 9.72

The findings from this study indicate that

children demonstrate knowledge about the writing from Grade 1 itself but to lesser extent compared to other grades. The probable reason for knowledge at such an young age in TDC is that, they start receiving formal instructions before beginning school. As age increases, there was improvement noticed in terms of productivity, complexity and accuracy.

B) Written language skills in children with Dyslexia Here we discuss about the second phase of the method where quantitative scores of children with dyslexia were compared within the group and between the groups with that of TDC. 1. Total Number of Words (TNW):

Table 9. Comparison of TDC and children with

dyslexia across grades for TNW Grade Mean of

TDC Group

95% Confidence interval for mean

of TDC

Quantitative data of children with

dyslexia Lower bound

Upper bound

Sub. no. Score

I 42.33 32.54 52.13 1 5 II 61.51 47.65 75.59 - - III 78.25

58.61 97.89 2 2

3 3

IV 115.07 97.48 132.65 4 62

5 62

V 161.52 133.51 189.63 6 47 7 86 8 33


241

Table 9 gives a clear picture that children with dyslexia scored lesser than their TD peers across the grades. It was also evident that higher grade children with dyslexia had higher scores compared to children in lower grades. This was in agreement with the study done by Houck and Bellingsley (1989). The possible reason could be because of word retrieval problems which was present in children with dyslexia (Mather et al., 2009). 2. Total Number of T-units (T-UNIT): Table 10 gives a clear picture that children with dyslexia scored lesser than TDC across the grades. It was also evident that higher grade children with dyslexia had higher scores compared to children in lower grades. This was in agreement with the study done by Houck and Bellingsley (1989). The possible reason could be that as there can be a word retrieval problem thereby a delay in producing words which limits them to produce lesser number of T-units thereby reducing the overall productivity.


dyslexia across grades for T-UNIT Grade Mean

of TDC Group

95% CI for mean of TDC

Quantitative data of children with dyslexia

Lower bound

Upper bound

Sub no. Score

I 8.78 6.60 40.95 1 1 II 10.97 8.42 13.52 - - III 11.50

9.50

13.50

2 1

3 0

IV 15.33 13.65 17.02 4 8

5 8

V 17.79 15.67 19.92 6 7 7 13 8 5

3. Mean length of T-units (MLT-UNIT): From Table 11, it was evident that there was an increasing trend in the MLT-UNIT in children with dyslexia and most of them had a score similar to that of their TDC. This could be because they would be lacking in their ability to use age appropriate vocabulary and tend to explain the meaning more elaborately. 4. Total number of clauses (CLAUSES): The results from Table 12 showed that children with dyslexia had lesser number of clauses compared to their typically developing peers across the grades. Higher grade children had higher scores compared to lower grade children. The total number of clauses is also a measure of productivity. There was an overall decrease in the

measure of productivity as seen earlier in TNW and T-UNIT which is reflected in CLAUSES also.


dyslexia across grades for MLT-UNIT

Table 12. Comparison of TDC and children with dyslexia across grades for CLAUSES

Grade Mean of TDC group

95% CI for mean of TDC group


Lower bound

Upper bound

Sub no. Score

I 7.70 5.40 10.01 1 1 II 11.53 8.80 14.26 - - III 13.93

11.12

16.74

2 0

3 0

IV 20.17 17.19 23.14 4 10

5 10

V 25.14 21.57 28.71 6 8 7 16 8 4

5. Clause Density (C-DENSITY): The results revealed that children with dyslexia were not that deviant from that of TDC in C-DENSITY particularly in higher grades. Taking support from the results obtained for MLT-UNIT, children with dyslexia may tend to write longer sentences involving more number of clauses per sentence compromising for their inability in using an appropriate vocabulary accordingly. Another reason could be that children lack the use of appropriate punctuations and write run on sentences with more and more clauses embedded in the same T-unit.




Lower bound

Upper bound

Sub no. Score

I 4.82 4.1252 5.5178 1 5 II 5.67 5.3429 6.0158 - - III 6.60

6.0633

7.1424

2 2

3 0

IV 7.45 6.7952 8.1148 4 7.75

5 7.75

V 9.09 8.0583 10.125 6 6.71 7 6.62 8 6.60


242

Table 13. Comparison of TDC and children with dyslexia across grades for C-DENSITY

6. Percentage of grammatical T-units (GRAM T-UNIT) : From Table 14 it was evident that higher grade children had higher number of grammatical T-units compared to children in lower grades in children with dyslexia. The results suggested that children with dyslexia produced less correct grammatically correct sentences when compared to TDC. But, there was an increasing trend seen from grade 1 through 5 in children with dyslexia. Higher graders have less erred sentences compared to low graders. This could be because of limited productivity seen in children in lower grades.

Table 14. Comparison of TDC and children with dyslexia across grades for GRAM T-UNIT




Lower bound

Upper bound

Sub no.

Score

I 0.44 0.29 0.60 1 0 II 0.47 0.35 0.58 - - III 0.49

0.38

0.61

2 0

3 0

IV 0.54 0.46 0.62 4 0.38

5 0.25

V 0.52 0.43 0.61 6 0 7 0.61 8 0.50

Another reason could be that, children in higher grades could have made more grammatical errors in an attempt to produce longer sentences. Third reason could be because grammaticality includes morphological and syntactical abilities which involved

use of suffixes that typically undergo significant development between fourth grade and high school in TDC. 7. Percentage of spelling errors (SPELL): Spelling errors had a lot of variability in children with dyslexia. There was no increasing or decreasing trend seen across the grades. Spelling is the most basic and most difficult skill of all. Spelling is a sensitive index of phonological processing. To spell the words correctly, children should have good knowledge in phonology, orthography, morphology and semantics of a language. Children with dyslexia have impaired phonological abilities and thus can have poor spelling abilities. A low quality phonological representations can result in less clearly defined, less explicit word representations and can result in pervasive effects on many aspects of language use (Perfetti & Hart, 2001).


dyslexia across grades for SPELL Grade Mean of

TDC group




Upper bound

Sub no. Score

I 0.101 0.06 0.14 1 0.20 II 0.067 0.04 0.10 - - III 0.055

0.03

0.08

2 0.50

3 1.00

IV 0.034 0.02 0.04 4 0.23

5 0.23

V 0.027 0.02 0.03 6 0.06 7 0.14 8 0.27

8. Errors of Writing Conventions (CONVEN): CONVEN has a lot of variability in TDC across the grades and there was no significant difference across groups. The same trend was also seen in children with dyslexia. The result of the present study is not in agreement with the previous studies done by Houck and Bellingsley (1989). Conventional errors were almost equal to that of the TDC. But, this result is not sensitive as there is a lot of variability within each grade in all the grades. A review of research reports that struggling writers have problems in using correct capitalization and punctuation rules. Children who struggle to write tend to write run-on sentences and sentences with too many clauses that are joined using conjunctions like „and‟. The same could be the reason in children with dyslexia as well.




Lower bound

Upper bound

Sub no. Score

I 0.81 0.68 0.94 1 1 II 1.04 0.10 1.07 - - III 1.17 1.09

1.26

2 0

3 0

IV 1.30 1.19 1.40 4 1.25

5 1.25

V 1.41 1.31 1.52 6 1.14 7 1.23 8 0.80


243

Thus, children with dyslexia showed poorer scores compared to TDC in almost all the aspects of productivity, complexity and accuracy except for MLT-UNIT . These children exhibited lesser TNW, less number of T-units but the number of words per T-unit was more though, these T-units had incorrect grammar and spelling errors. However, the high variability within children of the same grade could be attributed to the severity and type of dyslexia and to the treatment effects.


dyslexia across grades for CONVEN

Conclusions

The present study was conducted to determine the written language skills in TDC and in children with dyslexia. Children with dyslexia showed poorer scores compared to TDC in almost all the aspects of productivity, complexity and accuracy except for MLT-UNIT. These children exhibited lesser TNW, less number of T-units but the number of words per T-unit was more, though, these T-units had incorrect grammar and spelling errors. However, the high variability within children of the same grade could be attributed to the severity and type of dyslexia and to the treatment effects. Writing is a complex process that requires generation and integration of many levels of linguistic material. There might be difficulty at one or more levels of processing in children with dyslexia.

The norms developed can be used for the assessment of written language skills in children with dyslexia across grade 1 to 5. The topic „My School‟ used in the present study had increasing trend on productivity, complexity and accuracy starting from first grade to fifth grade. Hence, this topic can be used

as standard stimuli as it is easy, familiar and no ceiling effect noticed across the grades. Writing is a language based activity. So, SLPs can make valuable contributions to improving the writing of the children. This can be used for preparing appropriate management procedures to improve the linguistic aspects of writing. The measures discussed in the study can be evaluated at multiple points in time during the course of written language intervention and to assess the progress during and after intervention.

References

Berninger, V. W., & Swanson, H. L. (1994). Modifying Hayes and Flower‟s model of skilled writing to explain beginning and developing writing. In E. C. Butterfield (Eds.), Children‟s writing: Toward a process theory of the development of skilled writing (pp. 57-81), Greenwich, CT: JAI Press.

Chall, J. S. (1983). Stages of reading development. New York: McGraw Hill.

Houck, C., & Billigsley, B. (1989). Written expression in students with and without learning disability: differences across the grades. Journal of Learning Disabilities, 22, 561-565.

International Dyslexia Association. (2002). What is dyslexia? Definition of dyslexia adopted by IDA board of directors. Retrieved April 25, 2010, from http://www.interdys.org/FAQ WhatIs.htm.

Johnson, D., & Myklebust, H. (1967). Learning Disabilities: Educational Principles and Practices. New York: Grune and Stratton.

Mather, N., Wendling, B., & Roberts, R. (2009). Writing Assessment and Instructions for Students with Learning Disabilities (2nd Edn.). San Francisco: Jossey Bass.

Miller, J., & Chapman, R. (2001). Systematic Analysis of Language Transcripts (Versions 7.0) [computer Software]. Madison: University of Wisconsin-Madison, Waisman Center, Language Analysis Laboratory.

Prefetti, C. A., & Hart, L. (2001). The lexical bases of comprehension skill. In D. Gorfien (Eds.), On the consequences of meaning selection (pp. 67- 86). Washington, DC: American Psychological Association.

Puranik, C., Lombardino, L., & Altmann, L. (2008). Assessing the microstructure of written language using a retelling paradigm. American Journal of Speech- Language Pathology, 17, 107-120.

Scott, C. (2005). Learning to Write. In H. Catts & A. Kahmi (Eds.), Language and reading disabilities, 2nd Edn., (pp. 233-273). Needham Heights, MA: Allyn & Bacon.

Shanbal, J. C., & Prema, K. S. (2003). Tool for Screening Children with Writing Difficulties- ToSC-WD. Unpublished Master‟s Dissertation submitted to University of Mysore, Mysore.

Singer, B., & Bashir, A. (2004). Developmental variations in writing composition skills. In R. Silliman, J. Ehren, Apel (Eds.), Handbook of language and literacy development and disorders.

Grade Mean of TDC

95% Confidence interval for mean of

TDC



Upper bound

Sub no. Score

I 9.74 6.67 12.81 1 1 II 7.03 5.02 9.05 - - III 11.18

7.41

14.95

2 1

3 3

IV 10.90 7.13 14.67 4 15

5 17

V 10.38 6.68 14.08 6 8 7 15 8 9

Dissertation Vol. VIII, 2009-10, SLP, AIISH, Mysore

244

Effects of Aging on Naming in Bilingual Older Adults Sharon Susan Sam1 & Jayashree C. Shanbal 2

Abstract

The act of naming is an immensely complex process which is basic to our understanding of the world. Successful naming requires effective semantic and phonologic processing and normal visuo-perceptual function. Any disruption in these functions will lead to a difficulty in naming called anomia. Aging is one such factor which affects the naming ability. Bilingualism is yet another factor which can influence naming. The current study used an experimental method to examine naming performances in participants with ages beginning from 40 years till 80 years of age and these participants consisted of both monolinguals and bilinguals. The study used two experiments. Experiment 1 was done to compare the performance of naming on measures of accuracy and latency. This experiment measured accuracy scores and naming latencies obtained for a 24 item stimuli. Accuracy scores and reaction times were measured. The second experiment used a priming paradigm in order to carry out an in-depth analysis of the processing mechanism in the participant groups. Six priming paradigms were used. The findings support the Transmission Deficit Hypotheses (TDH, Mackay & Burke, 1990) which was put forth in an attempt to explain naming deficits while aging, that older individuals do have a reduction in priming transmission. It supported the suggestion that the naming failures in older adults were due to transmission deficit and not due to an inhibitory deficit. Further it also suggested that the presence of an extra lexicon in bilinguals does not cause an increased interference in bilingual participants.

Key words: naming, aging, bilingualism, cognition, transmission deficit hypothesis

he 1act of naming is important for our understanding of objects in the world (Howes, 1979). Naming is considered a relatively

straightforward cognitive operation, whose outlines are well understood. The various factors which might influence these naming latencies as well as naming accuracy include visual qualities of the stimuli, imageability, concreteness, prototypicality of the stimuli with the original object, semantic category, characteristics of the name like the phonemes in it, or the length of syllables, and frequency of occurrence or familiarity to the words. However, it is not yet clear which of these factors are more basic than the others. For naming in adults, the latencies of correct responses (from picture onset to the onset of spoken responses) for the more frequent items on visual confrontation naming tests were found to; typically range from 400 to 1500 ms (Goodglass, Theurkauf & Wingfeild, 1984).

Any disruption in the function described above which has a role in the process of naming can lead to naming difficulty or anomia. Although word-finding difficulties are found among individuals of all ages, this type of error are found to occur more frequently with ageing. Though, not all studies show a significant age-related word finding deficit on confrontation naming tasks there are enough studies which report word-finding difficulty in older adults (LaBarge,

1e-mail:[email protected]; 2Lecturer in Language Pathology, AIISH, Mysore, [email protected].

Edwards & Knesevich, 1986; Le-Dorze & Durocher, 1992). However, Goulet, Ska and Kahn (1994) pointed out that most research conducted in this area focused mainly on naming accuracy scores rather than on retrieval latency. They also reported that most of the available norms of picture naming test were based on naming accuracy and not naming latencies. Hence they suggested that in order to conclusively comment on the decline in naming skills in older adults, researchers also need to focus on measuring the naming latencies along with naming accuracy.

The decline in naming ability with ageing may

be multifactorial. Few researchers suggest that (Albert, Heller & Milberg, 1988; Goulet, Ska & Kahn, 1994), age decline in picture-naming abilities could be due to nonlinear modifications in cognitive function related to selective changes in the brain evolving at differential rates across the life span. Considering the relationship between cognition and language, a decline may be expected in the case of language performance which is associated with the decline in cognitive performance. However, variability was seen in the results of studies that focused on language skills. Two main areas found problematic for older individuals were lexical retrieval during word production (Goulet, Ska & Kahn, 1994) and comprehension of complex material (Wingfield & Stine-Morrow, 2000). Two types of paradigms were commonly used to measure lexical retrieval abilities- picture-naming tasks and analysis of tip-of-the-tongue (TOT) states.

T

Naming in bilingual older adults

245

An early study of confrontation naming in healthy subjects using Boston Naming Test (BNT; Kaplan, Goodglass & Weintraub, 1983) indicated a decline in naming ability with increasing age, i.e., older adults named fewer pictures than younger adults. (Borod, Goodglass & Kaplan, 1980). Another study by Nicholas, Obler, Albert and Goodglass (1985) supported the previous finding of decreased naming ability with age. Additionally, the analyses of response cues showed that, although older groups received significantly more phonemic cues, there was no difference between the age groups in the ability to utilize such cues. They also found that the most common errors for all subjects were semantically related to the target word. Comments about the target words and circumlocutions were also common errors. Younger subjects produced more semantically related errors, more errors that were semantically and phonologically related to the target than the older adults. Older subjects produced proportionately more circumlocutions and augmented correct responses than younger subjects. Age related decline of naming abilities were found as early as at age 50 years (Au, Joung, Nicholas, Obler, Kass & Albert, 1995) or even less than 40 years (Connor, Spiro, Obler & Albert, 2004) but significant reductions in performance were mostly evident at age 70 or over.

Evidence in this direction also comes from

investigating the tip-of-the tongue (TOT) phenomenon. TOT is said to occur when a familiar person or object cannot be named. Studies have commonly reported a larger number of TOTs in older adults (Burke, MacKay, Worthley & Wade, 1991). In addition most studies have found that older subjects have less access to information about the TOT target word than younger subjects do (Burke et al., 1991). It was also found that both older (mean age 71.0 years) and mid age (mean age 38.7 years) adults reported significantly more TOTs than younger (mean age 19.4 years) adults. Burke et al., (1991) also found that TOTs for proper nouns were more frequent than those for object names or abstract words for older and mid-age adults. Furthermore, consistent with the findings of their first experiment, Burke et al. (2002) took their results as support for their hypothesis that older adults have difficulty accessing at least parts of the phonological information needed for lexical production, i.e., the transmission deficit hypothesis.

Various theories were put forth in an attempt to

explain the above reported decline in naming abilities in older adults. The Transmission Deficit Hypothesis (TDH, MacKay & Burke, 1990) accounts for several findings related to naming difficulties and may explain

why a reduction in priming transmission co-exist with intact semantic priming. The TDH (Burke et al., 1991; James & Burke, 2000) suggests that this breakdown in priming transmission occurs because of weakening of lexical- phonological connections in memory with age. TDH is based on the framework of Node Structure Theory (Mackay, 1982, 1987) supported by others (White & Abrams, 2002, Abrams, White & Eitel, 2003; Cross & Burke, 2004). The TDH was also found to be consistent with neurobiological characteristics of aging.

With respect to naming, bilingualism is yet

another aspect widely studied. Bilingualism is defined as the regular use of two (or more) languages (Grosjean, 1992). Researches till date have shown that bilinguals may have many cognitive advantages. However there are few studies which have compared monolinguals and bilinguals. These studies have reported that bilinguals are slower in naming pictures but do not differ in the time required to access their meaning in order to classify them as human-made or natural (Gollan, Montoya, Fennema-Notestine & Morris, 2005). This finding probably reflects the fact that bilinguals are in fact less practiced in naming words in either of their languages. In the Indian context very few studies have been conducted in this area of research on naming. Vijaya Kumar (2007) found that there was no significant deterioration in cognitive linguistic flexibility as age advances. He also found that there was a significant increase in mean reaction times for target word retrieval. This indicated that there could be a slower cognitive processing in older individuals. He suggested that this could be due to the effects of slowing of cognitive linguistic processes. The critical age identified in the study for onset of cognitive aging was 60 years onwards. Based on extensive research in bilingual language, it is now clear that bilingual language processing is fundamentally language non-selective in nature. Effects of cross language activation have also been observed for words that do not have the same degree of lexical and semantic overlap as cognates such as interlingual homographs, which do not share meaning (“fin” which means “end” in Spanish) (Dijkstra, Van Jaarsveld & Ten Brinke, 1998; Dijkstra, DeBrujin, Schriefers & Ten Brike, 2000), and cross-language orthographic neighbors (words that share all but one letter across languages such as dine/cine in English and Spanish) (Jared & Kroll, 2001). However, Gollan, Fennema-Notestine, Montoya and Jernigan (2007) found that naming difficulties for bilinguals persist into older age; however, unlike the research on the executive control advantages for bilinguals, it does not appear that the magnitude of the bilingual disadvantage on these


246

linguistic skills changes with aging. The joint activation of both languages contributes to how lexical items are selected, leading to both facilitation and interference depending on the relation between the words (Costa, 2005). Further studies showed that linguistic deficit for bilinguals appears to be confined to situations requiring rapid retrieval of specific lexical items and is not found in linguistic or conceptual processing more generally.

Dijkstra, van Jaarsveld and Ten Brinke (1998)

extended the monolingual word recognition model called the Interactive Activation (McClelland & Rumelhart, 1981) to the bilingual domain, naming it Bilingual Interactive Activation or the BIA model. This was later extended to the BIA + model. Dijsktra et al., (1998) explains that this type of view has two consequences. One that the mental lexicon of bilinguals is integrated; that is there is no qualitative distinction between words of one language or the other, and the other that when a particular input is processed, the model does not make any distinction between words from the first or second language. This is termed language non selective lexical access and is because word recognition is an automatic process in which language membership information becomes active only after the word candidates have been activated.

This BIA + model differs from BIA model also

in terms of its representations. It incorporates orthographic, phonologic and semantic representations that are assumed to interact within the language processing system.

Several theories have been proposed to explain

bilingual language attrition as a function of age. An early hypothesis of language attrition was the regression hypothesis (Jakobson, 1941), which suggested that language components might be lost in attrition in the reverse order in which they were acquired. Following this the Activation Threshold Hypothesis (ATH) was proposed by Paradis (1985, 1993). Relative activation thresholds were used to explain for the language-selection patterns in polyglot aphasia and in bilingualism in general. Kopke (2004) employed the ATH to study language attrition and found that the results did not support ATH. Olshtain and Barzilay (1991) proposed a more specific hypothesis of lexical-retrieval decline in bilingualism. They proposed that language attrition degraded the speakers‟ lexicon and reduced the degrees of specification available. The Olshtain–Barzilay study (1991) assumes that words of basic meanings are retained but specific lexical items of highly specific concepts are less accessible.

The above literature points to the large amount of research evidence which suggest that cognitive abilities show a decline with aging and subsequently a decline in cognitive processing as well as naming abilities. However, there are also reports which contradict those findings. Hence, there is a need to look at the effects bilingualism has on age related decline in naming. The study proposes to analyze the pattern of naming difficulties seen in the older adults, both monolinguals and bilinguals. The study uses both accuracy and reaction time measurements to gather data on the naming failures related to aging.

Method

The study used an experimental design in order to quantify the effects on age related changes on naming. The naming latencies were considered across age groups under seven conditions, which included one confrontation naming and six priming conditions. The comparison of naming latencies in primed and unprimed conditions will help in drawing conclusions regarding the possible mechanisms in age related naming difficulties. The latencies were also be compared across ages. Among the age groups the younger participants acted as a control group while older participants were be considered the experimental group.

A total of eighty participants were included in the study. They included forty adult monolinguals in the age range 40 -80 years, 10 adults in each age group (40-50 years, 50-60 years, 60-70 years and 70-80 years) and forty adult Malayalam- English bilinguals in the age range 40 - 80 years, 10 adults in each age group (40-50 years, 50-60 years, 60-70 years and 70-80 years).

All the participants were informed about the purpose and procedure of the study and an informed consent was obtained from them. Participants selected used Malayalam as their first language (L1) and had learnt English as a part of their formal education. They were proficient speakers of Malayalam and English but were dominant speakers of Malayalam.

The experiment was carried out in a quiet room using a Lenovo Y410 model laptop and headphones. The stimulus presentation and responses were obtained using the DMDX software (Forster & Forster, 2003). The study included two experiments: Experiment 1: The first experiment involved a confrontation naming task. The test stimuli were prepared by selecting words from the Hundred Pictures naming test (Fisher & Glenister, 1992).The monolingual participants named only one set of


247

pictures, 24 in number, in their L1 and the bilinguals named two sets of pictures (24 pictures each) in their L1 and L2. The items were subjected to familiarity rating by 10 SLPs. Only the items rated as “most familiar” or “familiar” were selected. Those rated as “unfamiliar were omitted. 48 such pictures were selected. These picture stimuli were presented using a laptop using DMDX software. The DMDX software was programmed in order to present these pictures for 1.5 s and the patients were instructed to name the pictures as fast as possible. Cues were provided if the participant fails to name. The cues initially were a semantic cue and then a phonemic cue. Performance of the participants on measures of accuracy scores and naming latencies were measured. The DMDX software recorded the responses and further analysis using check vocal determined the reaction times as well as the errors exhibited by the participants. Experiment 2: The second experiment attempted to obtain naming latencies from both monolingual and bilingual subjects under six priming conditions. The experiment used two types of word lists- one set of homophone words and a set of non homophone words. The non-homophone stimuli to be used in the experiment were selected from the language training manual “With a Little Bit of Help” (Karanth, Manjula, Geetha & Prema, 1999). In addition to this, two homophone word lists were also prepared. These included separate lists for English and Malayalam homophone words. These entire word lists were subjected to familiarity rating by 15 SLPs who were dominant speakers of Malayalam and were bilinguals. Only the most familiar and familiar ones were chosen. The ones rated as unfamiliar were omitted. Bilingual and monolingual participants named these pictures in two different conditions. For bilingual participants these pictures were presented in 6 priming conditions. The tasks involved naming in both L1 and L2 languages with primes in the other language. Monolingual participants were also asked to name these pictures under six priming conditions. However, for monolingual participants, both the primes and the targets were in the same language. The six priming conditions included: (1) Appropriate-meaning distracters which were competitors of the depicted meaning of a picture (e.g., tree, picture -/log/; /puli/, picture-bear) (2) Inappropriate-meaning distracters which were competitors of the nondepicted meaning of the name of the homophone picture (e.g., water, picture -/honey/; answer, picture-ceiling) (3) In addition to the homophone pictures, pictures with non-homophone

names were paired with phonologically related (carrot for /kap偉al/or //su:tram/for sun), semantically related (pond for /araja偶am/ or /mul偉ap偉u/ for hibiscus), semantically- phonologically related distracters [S-P](aligator, picture-tortoise or /pa:tram/ for pot) and unrelated distracters were included (Car for book). The error responses were omitted and only accurate responses were considered for measuring the naming latencies. Appropriate statistical analysis was used to compare the effects of priming on naming latencies. The data collected from the 80 participants was subjected to quantitative analysis using the SPSS (16.0) version. Qualitative analysis of the responses for the Experiment 1 was done to determine the nature of errors and type of cueing which benefitted the older participant. Quantitative analysis was done to analyze the effect of age on naming accuracy and latency.


The results of the study are presented discussed under separate sections below.

Comparison of performance on naming across age groups: Table 1 shows the mean and standard deviations (SD) for accuracy scores and naming latencies across age groups for monolingual and bilingual participants. Experiment I demonstrated that for both monolinguals and bilinguals, there was significant effect of age on the naming accuracy in adults at p< 0.05 level of significance. In contrast, the effect of age on naming latencies was significant at p< 0.05 level of significance. i.e., as age increases there was significant increase in naming latencies with age although the accuracy is not compromised. There was no significant difference between the two languages (L1 and L2) of a bilingual at p< 0.05 level of significance. The analysis of monolingual and bilinguag naming performance on the basis of naming accuracy revealed that there was a progressive decline in the naming accuracy as the age advanced. The younger age groups 40-50 years and 50-60 years scored better than the older age groups, 60-70 years and 70-80 years. However, the decline was not statistically significant. Only for L2 there was a significant difference seen in the age group 70-80 years. However, qualitative analysis of data revealed that these groups showed difference in the type of errors exhibited during naming, i.e., the younger groups made greater semantically related errors (e.g.,‟lizard‟ for „snail‟) and the older groups exhibited circumlocutions (e.g., “it flies” for „Kite‟).


248

Table 1. Mean and standard deviation (SD) for accuracy and naming latencies across age groups

The latencies also showed a pattern similar to the accuracy scores. There was a decline in naming latencies as age advanced. The participants in the younger age groups (40-50 and 50-60 years) had faster latencies than the participants in the age group (60-70 and 70-80 years). The age group 70-80 years had the slowest latencies. The findings support the previous research on naming abilities in older adults that increasing age presents with a decline in the naming abilities. Nicholas et al., (1985) measured overall correctness levels, responsiveness to cues, and error types. They found a significant main effect of age for overall correctness. They found that older adults made more errors than younger adults and specifically, the 70-year-old group had significantly higher error rates than all the other groups, i.e., the 30-39 year old participants in both naming tasks. There are other longitudinal studies which replicated the same finding as that of the above researches. Au , Joung, Nicholas, Obler, Kass and Albert (1995) demonstrated that the initial signs of decline appeared by the age of 50 years, whereas study by Conner et al., (2004) showed that the decline may start as early as 40 years. The deficits in naming performance can be accounted by the TDH (Burke et al., 1991). TDH states that naming difficulties in older individuals can be due to a deficit in transmission of information across the mental nodes. The impairment could be at the level of the semantic nodes, at the level of the phonological nodes or at the connections between them. These impairments would cause an under activation of the target node thereby causing a word finding difficulty.

Bilingual participants also showed a pattern

similar to monolinguals. The accuracy scores were the better for younger age groups where as it was poorer for older age groups. This pattern was observed in both L1 and L2. There was a decline seen in the naming latencies also with aging. i.e., the younger age groups had faster latencies and older age groups had poorer latencies. The participants in the age group 70-80 years had the slowest latencies. The decline in naming ability

with aging in bilinguals could be due to two reasons; one could be that there is impairment in transmission of information across the lexical nodes as indicated in the case of monolinguals. The second factor could be impairment in inhibition of incorrect responses, which could lead to incorrect naming responses and longer latencies. The possibility of an inhibition deficit is further implicated in the bilingual lexicon since bilingual lexical selection involves the inhibition of the translation equivalents in the other language.

However, the present experiment also

demonstrated that the naming latencies followed similar patterns across languages i.e., the latencies did not differ significantly across languages. Support for this finding can be drawn from the BIA+model (Dijkstra et al., 1998) which stated that the mental lexicon of a bilingual is an integrated system. The model suggested non- selectivity in bilingual language processing. The similarity in accuracy scores and naming latencies as seen in the bilinguals naming performance could be due to the fact that these lexicons are retrieved from a single integrated system rather than from two separate stores.

Analysis of errors made by participants

indicated that younger participants made more of semantic errors and older participants exhibited circumlocutions. This was the pattern seen in both the languages and in both monolingual and bilingual participants. Participants above 60 years of age exhibited circumlocutions when confronted with a word whose name they were unable to retrieve.

Comparison of performance on naming across age groups in different priming conditions: Figure 1 and 2 illustrates the pattern of naming latencies across age groups in both monolinguals and bilinguals. A pattern

Age Group

(in years)

Monolinguals Bilinguals

L1 L2 Accuracy Latency (ms) Accuracy Latency (ms) Accuracy Latency (ms)

Mean (ms)

SD Mean (ms)

SD Mean (ms)

SD Mean (ms)

SD Mean SD Mean (ms)

SD

40-50 23.40 1.07 475.94 116.72 22.90 0.64 117.10 68.57 23.50 0.60 151.02 57.85 50-60 22.90 1.10 661.70 26.09 23.00 0.64 688.11 210.3 22.90 0.58 413.73 34.94 60-70 21.00 1.60 1205.59 203.42 22.00 0.66 709.35 160.19 22.20 0.60 732.03 46.9 70-80 20.00 2.90 1738.32 564.80 21.20 0.63 1560.16 362.27 21.10 0.59 1551.83 531.73


249

similar to that was seen in accuracy scores was observed in latencies also. However, the effect was more pronounced in latencies. When naming latencies were compared, it was seen that the Semantic- Phonologic (SP) and phonologic priming produced fastest priming latencies and unrelated and homophone priming produced slowest latencies in all the age groups. There is also evidence for the slowing latencies for semantic distracters. Homophone priming did not yield significant priming effects in monolinguals as their latencies looked much like that of much like unrelated primes. The Figure 2 also showed increased naming latencies as the age advanced. This effect was seen both in monolinguals and bilinguals and for the two languages of a bilingual. In the age group 60-70 years, a different pattern was observed wherein the latencies did not show much difference across the prime types. However S-P priming and Phonologic

priming produced longer latencies compared to other priming types as well as semantic priming produced interference. A similar pattern was seen in the age group 70-80 years. The variation in latencies reduced and the effects of various primes on the latencies was found to have reduced. Unlike in monolingual, in bilinguals, a homophone priming effect was observed. Homophone priming effect was observed when the targets were to be named in L2. There was a facilitation effect seen for homophone words primed with inappropriate meaning words (reduction of latencies) in all the age groups except in age groups 70-80 years.

In conclusion, the accuracy scores and naming

latencies both showed a decline with age, i.e., accuracy decreased and latencies increased as age advanced. However, the effects were significant only for the participants in the age group 70-80 years. The introduction of primes did influence the naming latencies and in some cases the accuracy scores. The most profound effect was observed for semantic phonologic priming which improved accuracy scores and reduced latencies with all age groups. Homophone priming was not evidenced in monolinguals. In bilinguals, there were no effects of homophone priming observed in L1, but there was an increase in accuracy and reduction in latency observed with homophone priming using inappropriate meaning words in L2. This effect was seen in all age groups except 70-80 years. Unrelated priming produced the slowest latencies and poorest accuracies. There were also differences in the pattern of priming accuracies and latencies within age groups.

For monolinguals, the following pattern

emerged during the analysis. First, there was an increase in the naming latencies as the age advanced even in the presence of primes. This finding supports the previous research findings which reported that there was an increase in the naming latencies with increasing age. In an early study by Borod et al., (1980) reported a decline in naming efficiency with age. They also reported that the older individuals received more number of phonemic cues compared to younger participants. The deficits in naming performance can be accounted by the TDH (Mackay & Burke, 1990). Based on the TDH, encoding of new memories and retrieval of existing memories depends on the rate of transmission of priming across the connections linking representational units in memory. Burke et al., (1991) suggested that connections become stronger with use (activation), and especially recent use, and weaken over time as a result of disuse. Ageing is also considered to weaken connection strength. These

0

500

1000

1500

2000

2500

Ph

on

Se

m

Se

m-P

ho

n

Un

rel

Ho

m-

ap

Ho

m-i

na

p

Mea

n na

min

g la

tenc

ies

Prime types

40-50

50-60

60-70

70-80

Age groups

0

200

400

600

800

1000

1200

1400

1600

1800

Ph

on

Se

m

Se

m-P

ho

n

Un

rel

Ho

m-a

p

Ho

m-i

na

p

Mea

n na

min

g la

teni

ces

Prime types

40-50(L1)

50-60(L1)

60-70(L1)

70-80(L1)

40-50(L2)

50-60(L2)

60-70(L2)

70-80(L2)

Age groups

Figure 1. Mean naming latencies across age groups in monolinguals.

Figure 2. Mean naming latencies across age groups in bilinguals.


250

impairments would cause an under activation of the target node thereby causing a word finding difficulty.

Thus the TDH accounts for the reduction in

accuracy and the increase in latencies due to aging in monolingual participants. It was found that younger participants benefitted most from semantic phonologic primes and phonologic primes and least from unrelated and homophone primes. Taylor and Burke (2002) reasoned out why phonologic and semantic phonologic primes could be of greater benefit while naming. Based on the node structure theory (NST) proposed by MacKay (1987) phonologic or semantic phonologic prime‟s offer bottom-up priming which facilitates faster access of the intended word. Burke and Taylor and Burke (2002) stated that in addition to the bottom-up facilitation effect, the S-P prime also offers a top-down priming effect. This is because of the fact that in addition to being phonologically related to the target word, the prime also belongs to the same semantic category, giving additional facilitatory effect. Taylor and Burke (2002) stated that semantically related distracters at-150ms SOA produced interference effect and caused slower naming latencies. They suggested that this is because a semantically related prime activates all the semantic nodes related to the prime and thus increase the number of nodes which compete for selection. This may be the reason why semantic priming produced longer latencies in all age groups. Taylor and Burke (2002) reported significant differences between homophone priming using appropriate and inappropriate meaning primes. However, this was not found in the current study. This could be due to the fact that the inappropriate meanings of homophone words are not in frequently used in daily life and are used only in formal educational and literary settings. However the participants in older age groups, 60-70 years and 70-80 years seemed to differ from the participants in younger age groups. Apart from having higher naming latencies, they did not show clear evidence for any of the priming types. These age groups were also benefitted by semantic phonologic and phonologic cues, but consistent with previous reports, they benefitted less from the primes. Thus the results of the current study supports the previous findings in the area that there is a reduction in transmission of priming from semantic to phonologic nodes in older individuals thus leading to increased latencies and reduced accuracy. It also suggests that the semantic system is not impaired by age but rather the transmission of information to lower nodes is what is getting affected.

The pattern of naming performance seen across ages in monolinguals was observed in bilingual data

also for both the languages i.e., as age advanced there was an increase in naming latencies in both the languages. There has been evidence suggesting that bilingual exhibit a cognitive advantage. Vijay Kumar (2007) found that bilingualism was a strong factor in delaying cognitive aging. However, as the data suggested, this advantage was not seen for naming in the present study. The results from the current study also pointed to the same fact that there is an age related decline in naming latencies that causes reduction in the latencies and accuracy scores for naming in bilinguals also in spite of the advantage that has been suggested to be present in them. i.e., there may not be a bilingual advantage when it comes to lexical access.

In all the age groups, semantic-phonologic and

phonologic primes produced faster latencies and better accuracy scores, but this effect was not significant in the older groups, 60 years above. The TDH (Burke et al., 1991) can be used to account for the absence of this facilitation effect for older adults. Based on the TDH (Burke et al., 1991), older adults might have deficits in the transmission of priming along single connections as those found in lexical to phonological representations. This transmission deficit would cause a reduction the priming that reaches the target or in other words delay the activation of the target in response to the stimuli. In case of homophone priming, the effects were similar to that seen in monolinguals for naming target in L1. This could be because the inappropriate meanings of homophone words were not in frequent use and were used only in formal situations. This may have led to weaker activation of the related nodes and hence failed to provide any facilitation.

However for L2, evidence of homophone

priming effect was seen. The homophone words primed with inappropriate meaning words produced better latencies in all the age groups except 70-80 years. The presence of homophone priming effects could be due to the fact that homophones with both their meanings in L2 (English) was in regular use compared to the homophone words in L1 (Malayalam) which was not in regular use as explained in the section above. Hence both the meanings are active for the bilingual in case of an English homophone, but only the dominant meaning was active in case of a Malayalam homophone.

Statistical analysis revealed that there was a

significant difference across age groups at p<0.05 level of significance when priming was compared with age groups. However there was no significant difference across languages, i.e., the pattern of decline was similar across languages. This adds further support to the


251

BIA+ model which states that bilinguals have an integrated mental lexicon. The presence of an extra lexicon was not found to be influencing the latencies of naming in bilinguals. There have been few recent studies which puts forward the same proposal. In general, the naming latencies were found to increase in both monolinguals and bilinguals. There was no significant language effect in bilinguals suggesting that the bilingual lexicon could be one integrated system. The decline was seen in both languages for the bilingual and was comparable to that in monolinguals.

Hence the present study supports the previous literature which shows that bilinguals do not show an advantage in lexical access. The priming experiment suggested that the naming deficits in older adults could be attributed to transmission deficits. Additional support is provided by the homophone priming effect seen in bilinguals for L2. The results of homophone priming also seemed to contradict the inhibitory deficit hypothesis. Also it was found that inhibitory deficit was not significant in older adults as demonstrated by the less significant semantic interference effect. Also it was found that Phonologic and S-P primes consistently offered facilitatory effect across age groups, the better of the two being Semantic- Phonologic. All age groups benefitted from the phonological and semantic phonological primes due to the bottom-up priming in phonological primes and top-down bottom- up priming in S/P priming condition. Moreover, the onset of deficits in language processes can be thought of as beginning by the age of 60-70 years since there was a pattern change observed across all priming types in this age group. There is the need for further studies to support this claim.

Conclusions

The focus of the current study was to investigate the effects increasing age has on naming abilities. It intended to verify the literature which establishes that increasing age is accompanied by a decline in naming performance. The performance was examined naming decline with aging in both monolingual and bilingual context. This was done to find out if the presence of an extra lexicon influences the naming performance in bilingual older adults. Thus in turn, the study seeks to verify the literature which suggests that there is a cognitive advantage for bilingualism, but, causes a disadvantage in lexical access.

The findings support the TDH which was put

forth in an attempt to explain naming deficits in aging, that older individuals do have a reduction in priming transmission. This finding also supports the suggestion that the naming failures in older adults were indeed due

to transmission deficit and not due to an inhibitory deficit. Further it also suggests that the presence of an extra lexicon in bilinguals does not cause an increased interference in bilingual participants. The study also offers evidence for the BIA+ model put forth to explain bilingual lexicon organization. Since there were no differences in performance in both languages of the bilingual, it can be considered that they form part of the same integrated system of a bilingual mental lexicon which comprises of words of both languages.

References

Abrams, L., White, K. K., & Eitel, S. L. (2003). Isolating phonological components that increase tip-of-the-tongueresolution. Memory and Cognition, 31, 1153–1162.

Albert, M. S., Heller, H. S., & Milberg, W. (1988). Changes in naming ability with age. Psychology and Aging, 3, 173–178.

Au, R., Joung, P., Nicholas, M., Obler, L. K., Kass, R., & Albert, M. L. (1995). Naming ability across the adult life span. Aging and Cognition, 2, 300–311.

Borod, J. C., Goodglass, H., & Kaplan, E. (1980). Normative data on the Boston Diagnostic Aphasia Examination, Parietal Lobe Battery, and the Boston Naming Test. Journal of Clinical Neuropsychology, 2, 209–215.

Burke, D. M., MacKay, D. G., Worthley, J. S., & Wade, E. (1991). On the tip of the tongue: What causes word finding failures in young and older adults? Journal of Memory and Language, 30, 542–579.

Connor, L. T., Spiro, A., Obler, L. K., & Albert, M. L. (2004). Change in object naming ability during adulthood. Journal of Gerontology: Psychological Sciences, 59, 203–209.

Costa, A. (2005). Lexical access in bilingual production. In J. F. Kroll & A. M. B. D. Groot (Eds.), Handbook of bilingualism: Psycholinguistic approaches (pp. 308–325). New York: Oxford University Press.

Cross, E. S., & Burke, D. M. (2004). Do alternative names block young and older adults‟ retrieval of proper names? Brain and Language, 89,174–181.

Dijkstra, A., de Bruijn, E., Schriefers, H., & Ten Brinke, S. (2000). More on interlingual homograph recognition: language intermixing versus explicitness of instruction. Bilingualism: Language and Cognition, 3, 69–78.

Dijkstra, A., van Jaarsveld, H., & Ten Brinke, S. (1998). Interlingual homograph recognition: effects of task demands and language intermixing. Bilingualism: Language and Cognition, 1, 51–66.

Fisher, J., & Glenister, J. M. (1992). The Hundred Pictures Naming Test. Hawthorn, Vic: Australian Council for Educational Research.

Forster, K. I., & Forster, J. C. (2003). DMDX: A Windows display program with millisecond accuracy. Behavior Research Methods, Instruments & Computers, 35, 116–124.

Gollan, T. H., Montoya, R. I., Fennema-Notestine, C., & Morris, S. K. (2005). Bilingualism affects picture naming


252

but not picture classification. Memory and Cognition, 33, 1220–1234.

Gollan, T. H., Fennema-Notestine, C., Montoya, R. I., & Jernigan, T. L. (2007). The bilingual effect on Boston Naming Test performance. The Journal of the International Neuropsychological society, 13, 197–208.

Goodglass, H., Theurkauf, J. C., & Wingfeild, A. (1984). Naming latencies as evidence for two modes of lexical retrieval. Applied Psycholingusitics, 5, 135-146.

Goulet, P., Ska, B., & Kahn, H. J. (1994). Is there a decline in picture naming with advancing age? Journal of Speech and Hearing Research, 37, 629–644.

Grosjean, F. (1992). The bilingual and bicultural person in the hearing and deaf world. Sign Language Studies, 77, 307-320.

Howes, D. (1979). The naming act and its disruption in aphasia. In D. Aaronson & R. W. Rieber (Eds.), Psycholinguistic research: Implications and applications (pp. 435–470). Hillsdale, NJ: Erlbaum.

Jakobson, R. (1941). Child language, aphasia, and phonological universals. The Hague: Mouton, 1968.

James, L. E., & Burke, D. M. (2000). Phonological priming effects on word retrieval and tip-of-the-tongue experiences in young and older adults. Journal of Experimental Psychology: Learning, Memory, and Cognition, 26, 1378–1391.

Jared, D., & Kroll, J. (2001). Do bilinguals activate phonological representations in one or both of their languages when naming words? Journal of Memory and Language, 44, 2–31.

Kaplan, E. F., Goodglass, H., & Weintraub, S. (1983). The Boston Naming Test. 2. Philadelphia, PA: Lea & Febiger.

Karanth, P., Manjula, R., Geetha, Y.V., & Prema, K, S. (1999). Production of Language Training Materials in Major Indian Languages. A UNICEF funded project submitted to All India Institute of Speech and Hearing, Mysore.

Köpke, B. (2004). Neurolinguistic aspects of attrition. Journal of Neurolinguistics, 17, 3-30.

La Barge, E., Edwards, E., & Knesevich, J. W. (1986). Performance of normal elderly on Boston Naming Test. Brain and Language, 27, 380-384.

Le Dorze, G., & Durocher, J. (1992). The effects of age, educational level, and stimulus length on naming in normal subjects. Journal of Speech- Language Pathology and Audiology, 16, 21–29.

MacKay, D. G. (1982). The problems of flexibility. fluency, and speed-accuracy trade-off in skilled behavior. Psychological Review, 89, 483-506.

MacKay, D. G. (1987). The organization of perception and action: A theory for language and other cognitive skills. New York: Springer-Verlag.

MacKay, D. G., & Burke, D. M. (1990). Cognition and aging: A theory of new learning and the use of connections. In T. Hess (Eds.), Aging and cognition: Knowledge organization and utilization (pp.13–263). Amsterdam: North Holland.

McClelland, J. L., & Rumelhart, D. E. (1981). An interactive activation model of context effects in letter perception: Part 1. An account of basic findings. Psychological Review, 88, 375-407.

Nicholas, M., Barth, C., Obler, L. K., Au, R., & Albert, M. L. (1997). Naming in normal aging and dementia of the Alzheimer‟s type. In H. W. Seliger, & R. M. Vago (Eds.), First language attrition (pp. 139–150). Cambridge: Cambridge University Press.

Nicholas, M., Obler, L. K., Albert, M. L., & Goodglass, H. (1985). Lexical retrieval in healthy aging. Cortex, 21, 595–606.

Olshtain, E., & Barzilay, M. (1991). Lexical retrieval difficulties in adult language attrition. In H. W. Seliger, & R. M. Vago (Eds.), First language attrition (pp. 139–150). Cambridge: Cambridge University Press.

Paradis, M. (1985). On the representation of two languages in the brain. Language Sciences, 7, 1–39.

Paradis, M. (1993). Linguistic, psycholinguistic, and neurolinguistic aspects of interference in bilingual speakers: The activation threshold hypothesis. International Journal of Psycholinguistics, 9, 133–145.

Taylor, J. K., & Burke, D. M. (2002). Assymmetric aging effects on semantic and phonological processes: Naming in the picture words intereference task. Psychology and Aging, 17, 662-676.

Vijay Kumar. (2007). Cognitive-linguistic flexibility and aging. Unpublished Master‟s Dissertation. University of Mysore, Mysore.

White, K. K., & Abrams, L. (2002). Does priming specific syllables during tip-of-the-tongue states facilitate word retrieval in older adults? Psychology and Aging, 17, 226–235.

Wingfield, A., & Stine-Morrow, E. A. L. (2000). Language and speech. In F. I. M. Craik & T. A. Salthouse (Eds.), The handbook of aging and cognition (2nd edn., pp. 359–416). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.

http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6VDJ-47RB6MF-9&_user=10&_coverDate=01%2F31%2F2003&_rdoc=1&_fmt=high&_orig=search&_origin=search&_sort=d&_docanchor=&view=c&_searchStrId=1514185975&_rerunOrigin=scholar.google&_acct=C000050221&_version=1&_urlVersion=0&_userid=10&md5=a12ca122b9c9d91bfefc222c025cb0fe&searchtype=a#bbib24

Vocal load in students of speech-language pathology

253

Vocal Load in Undergraduate Students of Speech-Language Pathology

Shwetha C. 1 & S. R. Savithri 2

Abstract

Vocal load is the demand placed on the voice organ by voice use needs. Vocal load is measured in terms of time, voicing and intensity: the longer and louder a person has to talk, greater the strain on the voice will be and more voice capacity is needed. The objectives of the present study included (a) investigation of prevalence of vocal fatigue symptoms in students of Speech-Language Pathology (SLP), (b) quantification of voicing periods (vocal load), and (c) determination of short-term vocal recovery index in SLP students. The term „amount of voice use‟ or „time dose‟ is measured as „voicing periods‟ in this study. The study was conducted in two phases: phase-1 investigating the prevalence rate of vocal fatigue symptoms in seventy four B. Sc. (Speech & Hearing) students and phase-2 investigating the percent of voicing duration, mean F0, maximum F0, minimum F0, and SD of F0 before and after therapy sessions and short-term vocal recovery index during speech-language therapy in five students from phase-1. Out of 74, 60 students scored less than 17 and 14 students scored greater than 17, and the total prevalence rate of vocal fatigue symptoms was18.91% (risk of developing voice disorder). The results indicated that the average percentage of voicing was 58.54%. There was no difference in F0 measures of the pre- and post therapy, and within pre- and post therapy sessions. However, the mean and highest F0 were higher in post-therapy compared to pre-therapy sessions. The short-term vocal recovery index was found to be 0.42.

Key words: vocal load, speech-language therapist, voicing period, short-term recovery index

rofessional voice users can be defined as 1individuals who use their voice to earn their daily living. Stemple (1993) defined

professional voice users as those individuals who are directly dependent on vocal communication for their livelihood. Professional voice users are more prone to laryngeal pathologies than the general population due to their nature of work and life style (Stemple, Glaze & Gerdeman, 1995). Koufmann (1998) evolved a classification of vocal professionals based on their voice use and risk. The four levels are as follows: The elite vocal performers (level I) includes singers and actors, where a slight vocal difficulty cause serious consequences to them and their careers. The professional voice users (level II), is a person for whom even moderate vocal difficulty would prevent adequate job performance. Clergymen, lecturers/teachers, politicians, public speakers and telephone operators would fall into this group. The non-vocal professional (level III), is a person for whom a severe vocal problem would prevent adequate job performance. This group includes lawyers, businessmen and physicians. The non-vocal non-professional (level IV), is a person for whom vocal quality is not a prerequisite for adequate job performance. Laborers and clerks fall into this group.

Vocal load is defined as the demands placed

on the voice organ by voice use needs. Vocal loading is a combination of prolonged voice use and “additional loading factors (e.g. background noise, acoustics, air quality) affecting the fundamental


frequency, type and loudness of phonation or the vibratory characteristics of the vocal folds as well as the external frame of the larynx” (Vilkman, 2004). Vocal load is measured in terms of time, voicing and intensity: the longer and louder a person has to talk, the greater the strain on the voice will be and the more voice capacity is needed (Buekers, 1998).

There are on-going studies of a large number

of teachers who are recorded with the dosimeter for several days and weeks both at work and leisure time. Parameters used to measure different vocal doses are: Time dose (total phonation time, measured in seconds), Cycle dose (the total number of vocal fold vibrations) and Distance dose (the total distance the vocal folds are moving, measured in meters). Common values from cycle dose measurements are 1 ½ - 2 million vocal fold vibrations during a working day for a female teacher (Titze, Svec & Popolo, 2003). The simplest vocal dose is time dose, often called the voicing time, which accumulates the total time the vocal folds vibrate during speech.

Russel, Oates and Greenwood (1998)

investigated the prevalence of voice disorders in 1168 state school teachers using mail survey and reported that 22% of teachers experienced regular voice problems that interfered with their ability to use their voices as they wish, and 38.7% of teachers reported missing at-least 1 day of work in the past year as a result of their voice disorder. This result is also supported by Smith, Kirchner, Taylor, Hoffman and Lemke (1998). Kowalska, Bogusz, Fiszer, Spychaska, Kotylo, Przygocka and Modrzewska (2006) investigated the prevalence and the risk

P



254

factors for occupational voice disorders in two groups (425 female primary and secondary school teachers and 83 non teaching women of age range 23 to 61yrs). Vocal symptoms were more frequent in the teachers than in the non teachers (the mean number of voice symptoms were 3.21 and 1.98 respectively). Based on 400 voice professionals (100 singers, 100 teachers, 100 politicians & 100 vendors) in India, Boominathan, Rajendran, Nagarajan, Jayashree and Muthukumaran (2008) reported the most common symptoms in professional voice users as change in voice quality, voice fatigue, discomfort in throat, hoarseness, loss of voice, loss of intelligibility/clarity of speech/song, dry throat, shortness of breath, frequent throat clearing, itchy throat, voice tightness, loss of voice control, and inability to maintain shruthi/range.

In the past, several studies have been

conducted on vocal loading in professional voice users. Ohlsson, Brink and Lofqvist (1989) recorded the fundamental frequency and phonation time using the voice accumulator. In the application measuring the phonation time of speech pathologists (age range, 27-46) and hospital nurses (age range, 27-42), the phonation time of the former was higher (6.6%) than that of nurses (5.4%). The fundamental frequency of speech was higher among nurses (233.1 Hz) than among speech pathologists (216.1 Hz). In a group of nine different occupation (bookkeepers, teachers, sports instructors, swimming instructors, nurses, telephone operators, receptionists, speech therapists, and day care center teachers), the vocal loading of sports instructors was the highest, followed by that of day care center teachers (Buekers, Bierens, Kingma & Marres, 1995). Airo, Olkinurora and Sala (2000) studied speaking time and speech SPL in eleven female subjects using a pair of modern noise exposure analyzers. The speech detection was estimated in quiet as well as background noise at 50, 60, 70, 80, 85 and 90 dB in ascending order and found that a 6 dB detection threshold was better in separating speech from background noise and speech detection was consistent up to 70 dB background noise. A pilot study was done by Wingate and Sapienza (2007) on four teachers of singing and/or choral directors with regard to vocal loading and they were employed full-time in public school and university settings. Evening lessons and rehearsals were included in the data collection for five days. The KayPentax Ambulatory Phonation Monitor, Model 3200, was used to record total phonation time, average F0, and SPL values. The results showed that there was high vocal load which was moderately correlated to their perceived vocal effort.

Lohscheller, Doellinger, McWhorter and

Kunduk (2008) evaluated the voice effects of vocal loading and also determined whether high-speed digital imaging with phonovibrogram (PVG) analysis

would identify changes in vocal fold vibratory characteristics following prolonged reading (vocal fatigue) in subjects with normal voice. In all subjects, the effect of vocal loading was reflected by alterations of PVG parameters representing the posterior opening and closing dynamics. Evaluation within subjects revealed slight asymmetric vibratory behavior between the left and right vocal folds. Doellinger, Lohscheller, McWhorter and Kunduk (2009) investigated the potential of high-speed digital imaging technique (HSI) and the phonovibrogram (PVG) analysis in a normal vocal fold dynamics by studying the effects of continuous voice use (vocal loading) during the workday before and in the afternoon after vocal loading, respectively. The result revealed significant changes in vibration behavior, indicating impact of occurring vocal load. Constant voice usage stresses the vocal folds even in healthy subjects and can be detected by applying the PVG technique. Hence, HSI in combination with PVG analysis seems to be a promising tool for investigation of vocal fold fatigue and pathologies resulting in small forms of dynamical changes.

Morrow and Connor (2009) measured total phonation time, fundamental frequency (F0), and vocal intensity (dB SPL [sound pressure level]) in elementary class (n = 7) and music teachers (n = 5), using a KayPentax Ambulatory Phonation Monitor (KayPENTAX, NJ) during five full teaching days of one workweek. The results suggest that typical vocal loads for music teachers are substantially higher than those experienced by classroom teachers. Rajasudhakar and Savithri (2009) measured the voicing periods and vocal short-term recovery index using digital audio tape (DAT) recorder on two female primary school teachers of age 32 and 42 years, respectively. The average voicing percentage was found to be around 26.91% (1hour 34mins 28secs) and 0.72 was the average short-term vocal recovery index for teachers.

Short-term vocal recovery index

Titze (2002) reported the effects of short (ls)-and long (l l) -term vocal recoveries. The long-term vocal recoveries are repair of extra-cellular matrix, blood vessels, basement membrane and growth of epithelial cells. On the other hand, the short-term vocal recoveries are restoration of circulation, water to tissue, calcium and removal of lactic acid.

Recovery indexes were calculated as follows:


255

Risk factors in vocalization included (a) vibration overdose, (b) lack of recovery time, (c) poor tissue environment, (d) genetically weak tissue structure, and (e) poor vocal efficiency (economy). Genetic factors include thickness of mucosa, toughness of skin (epithelial cells), basement membrane and extracellular matrix, Proteoglycan (liquid composition), and vocal fold geometry (length, thickness, left-right asymmetry etc.). Poor tissue environment includes dehydration, exposure to smoke, chemicals, pollens, drugs, and inadequate nutrition. Hunter and Titze (2009) quantified the recovery of voice following a 2-hour vocal loading exercise (oral reading). Vocal recovery trajectory was tracked in eighty-six adult participants using short vocal tasks and perceptual ratings after an initial vocal loading exercise and for the following two days. The results showed that the short-term recovery was apparent, with 90% recovery within 4 to 6 hours and full recovery at 12 to 18 hours. They compared the vocal fatigue with dermal wound healing, and suggested that the vocal fatigue from the daily use of the voice could be treated as a chronic wound with the healing and repair mechanisms in a state of constant repair. Determination of this short-term vocal recovery index is useful in the management of professional voice users with voice disorders. Speech induced voiceless parallels noise-induced hearing losses. Ear and larynx are the two unique organs where tissue is vibrated at high frequencies. Ear tissue vibrates at low amplitude (0.001mm) but the larynx-tissue vibrates at high amplitudes (1mm). Therefore, the protection of the sounding mechanism (the voice) is as vital as protection of the hearing mechanism (the ear) (Titze, 2002). However, the occupational health care of professional voice users is surprisingly undeveloped compared to the attention given to occupational hearing disorders. This is due to lack of reliable evidence of higher risk of voice disorders in this profession and what specific risk that can cause voice problems. Even though the voice is essential for speech therapist in daily living, they are not much studied. Prevalence of voice problems in professional voice users is valuable in prevention of the same in this population. Also, there are no studies on Speech-Language Pathologist (SLP) who are nevertheless professional voice users and thus, are more prone to voice disorders. Further, determination of vocal recoveries is useful in the management of professional voice users with voice disorders. In this context, the present study was planned. The

objectives of the study were three fold and included (a) investigation of the prevalence of vocal fatigue symptoms in SLP students, (b) quantification of the voicing periods (vocal load), and (c) determination of the short-term vocal recovery index in SLP students.

Method

The study was carried out in two phases. Phase 1 investigated the prevalence of vocal fatigue symptoms in students doing B.Sc (Speech and Hearing) and phase 2 investigated the voicing periods and short-term vocal recovery index in them.

Phase 1: Prevalence of vocal fatigue symptoms in B. Sc. (Speech & Hearing) student. Subjects: Seventy four B. Sc (Speech and Hearing) students in the age range 18- 22 years were randomly selected from the All India Institute of Speech & Hearing, Mysore, India. The criteria for selecting them were, (1) they should have normal speech, language and hearing, (2) they should not have any voice problems, (3) they should have normal structure and function of the articulatory and phonatory structures and (4) they should provide speech-language therapy for persons with developmental language disorder. Procedure: Vocal fatigue symptom questionnaire was used to collect the information on vocal fatigue features along with standard demographic data. The vocal fatigue questionnaire of Rantala, Vilkman and Bloign (2002) adopted by Rajasudhakar and Savithri (2009) was used.

The questionnaire contains nine vocal fatigue statements. The students were instructed to answer the questionnaire without exaggerating/under estimating the vocal fatigue features during the therapy hours (see Appendix). The B.Sc (Speech and Hearing) students were asked to answer on a 4-point rating scale where „1‟ indicates less than once a year, „2‟ indicates a few times per year or relatively seldom, „3‟ indicates once a month or relatively often, and „4‟ indicates almost every wee or very often.

Analyses: The scores for each student were summed up as a sum variable called „voice fatigue complaints‟. If complaint score exceed 17, the subject is at a higher risk to develop voice disorders. A score of less than 17 indicates no risk to develop voice disorders (Rantala, Vilkman & Bloign, 2002). The total prevalence rate of vocal fatigue symptoms was calculated by using the following formula:


256

Phase 2: Voicing periods and short-term vocal recovery index (Is) in B.Sc (Speech and Hearing) students: Phase 2 was carried out in the natural therapy settings. Subjects: Five of the students (1 male, 4 female) who participated in phase 1 were selected on the basis that they should be providing at least 4 sessions of language therapy for children with developmental disorder. Instruments used: A portable, light-weight digital audio tape (DAT) recorder (Olympus digital voice recorder WS- 100, Japan) was used. The recorder had in-built condenser microphone (ME 15) and the weight of the device was about 54 grams (including battery). The overall frequency response of the microphone was 100 Hz to 5000 Hz and size of the DAT recorder was about 94(L) x 40(W) x 15.1(T) mm. The sampling frequency was 12 kHz and the maximum power consumption of the DAT recorder was 100mW. PRAAT (Boersma & Weenink, 2009) software was used to extract voicing periods. Figure 1 shows the subject wearing the DAT recorder.

Figure 1. Subject wearing the digital audio tape recorder.

Procedure: The objectives of the study were explained to the students and also regarding the confidentiality of the data. The voice/speech sample was recorded in therapy settings, the natural therapy environment of students. They were asked to wear a tiny, light weight digital voice recorder (54 grams, Olympus WS-100) around the neck throughout the therapy hours (4 hours) and speech recording was done during 5 working days, from Monday to Friday (4 hours × 5 days = 20 hours) . The distance between the mouth and microphone (10 cm) was kept constant for all students. Apart from the speech sample, they were instructed to phonate vowel /a/ for 3-4 seconds before and after therapy of 4 hours every day. Analyses: The recorded five-day voice/speech samples were transferred onto the computer memory. The entire day speech sample was truncated into ten minutes token and a total of 24 ten minute tokens were made for each participant using Adobe Audition software. Each of the ten minute samples was viewed as waveform on PRAAT (Boersma & Weenink, 2009) software and the voicing periods and

F0 measures were extracted. In each sample, the speech of others, like speech of a child who is receiving the therapy, parent‟s speech, external noise, and other therapist‟s speech was removed, retaining only the intended therapist speech. i) Quantification of voicing duration: The participant‟s voicing (pitch) periods was estimated using PRAAT software by removing the unwanted speech/voice in the samples. The degrees of voice breaks were extracted from the software and it was converted into the voicing percentage by subtracting the value from 100. The average of voicing percentage for 24 ten minute tokens for each day was calculated. Figure 2 shows an illustration of quantification of voicing periods.

Figure 2. Illustration of voicing periods for a sample of 17 seconds.

ii) F0 measures: PRAAT software was used to analyze the frequency related parameters for the speech sample. Using the first and the last ten minute sample, the mean fundamental frequency (F0), minimum F0, maximum F0, and standard deviation in frequency was extracted pre (before) and post (after) therapy and comparison was made between pre and post therapy context as well as between participants for each day. iii) Short-term vocal recovery index (Is): The short-term vocal recovery index (Is) is defined as the ratio of the total resting time (rest periods) to the actual performance (voicing) time. This can be calculated using the formula, performance time – phonation time (voicing duration) / performance time (Titze, 2002). The performance time represents the total recorded duration of speech sample on a workday and the phonation time represents the total voicing periods/voicing time. Lower Is indicates risk for voice disorders. The data on voicing periods and short term vocal recovery index were tabulated and the mean and SD of percent voicing was calculated. The percent voicing was compared across five students.

Results

Phase 1: The questionnaire was collected from the seventy four B.Sc. (Speech and Hearing) students to investigate the prevalence of vocal fatigue symptoms. Out of 74, sixty students scored less than 17, having fewer complaints of vocal fatigue symptoms (no risk to develop voice disorder), and


259

Table 4. Is of each student

and telephone marketers, 0.3. Rajasudhakar and Savithri (2009) reported an average short-term vocal recovery index of 0.72 in teachers. The Is in the present study was much lower than that reported by Titze (2002) in teachers and Rajasudhakar and Savithri (2009) in primary school teachers. Shorter the Is value, greater the vocal injury. The results indicated that students of speech pathology were likely to have injuries to vocal folds. The short-term vocal recoveries are restoration of circulation, water to tissue, calcium and removal of lactic acid. Lactic acid (IUPAC systematic name: 2-hydroxypropanoic acid), also known as milk acid, is a chemical compound that plays a role in several biochemical processes. It is thought that the muscle cells of our body are converting this glucose to lactic acid and then the lactic is absorbed and converted by mitochondria to fuel for muscle cells. So by training at a high intensity, we are not only getting in a good work out, but our body is creating additional proteins that help in absorbing and converting lactic acid to energy. Lactic acid is primarily found in sour milk products, such as koumiss, leban, yogurt, kefir, and some cottage cheeses. The casein in fermented milk is coagulated (curdled) by lactic acid (en.wikipedia.org/wiki/Lactic_acid).

The ability of a muscle to sustain contraction

over a prolonged period is in part related to the distribution of motor unit fiber types within the muscle body. Type I (slow-twitch) motor units are more fatigue resistant than type IIa and IIb (fast-twitch) motor units. Thus, muscles containing a high proportion of type I fibers are considered fatigue-resistant. Artificial stimulation studies using feline and canine thyroarytenoid muscles have demonstrated high levels of resistance to fatigue (Edstrom, Lidquist & Martensson, 1984; Cooper & Rice, 1990).

Titze (1984, 1994) has hypothesized a

relationship between vocal fatigue and altered vocal fold viscosity. Prolonged periods of phonation may lead to changes in the composition of fluids within the vocal folds, the result being an elevation in the viscosity and stiffness of the folds. According to Titze, increased tissue viscosity should result in proportionally greater friction and heat dissipation

during vocal fold vibration. This reduction in phonatory efficiency would suppose greater energy input in order to initiate and sustain oscillation of the folds.

It has been noted that blood circulation to the

vocal folds decreases during phonation (Hiroto, Toyozumi, Tomita, Miyagi, Kuroki & Koike, 1969; Matsuo, Oda, Tomita, Machara, Umuzaki & Shin, 1987). This phenomenon is most likely due to a constriction of blood vessels associated with increased intramuscular pressure during contraction (Enoka, 1994).

Titze (1984, 1994) has hypothesized regarding

the impact of excessive or prolonged phonation on the nonmuscular tissues of the larynx. Titze suggests that the repeated application of mechanical stress to epithelia and the lamina propria during vocal fold elongation may contribute to vocal fatigue. Similar effects may be significant for ligamental tissue and the cartilaginous framework of the larynx. The nature of this proposed fatigue mechanism (vocal fold elongation) indicates that it is most likely applicable to prolonged phonation at high pitch. While Titze has suggested tissue strain as a potential factor in vocal fatigue, no data exist demonstrating the relative contribution of this mechanism.

Research questions and experimental designs

should accommodate probable individual differences in response to occupational and other voice demands. By further understanding the factors that promote individual susceptibility to vocal fatigue, clinicians can better advise in issues relating to voice care. Variables of interest include age, gender, nature and degree of voice training, nature and degree of occupational voice demand, vocal hygiene, and history of voice problems. Issues related to gender and voice training has begun to be explored (Scherer, Titze, Raphael Wood, Ramig & Blager, 1986; Gelfer, Andrews & Schmidt, 1991; Scherer, Titze, Raphael Wood, Ramig & Blager, 1991; Solomon, Glaze, Arnold & van Mersbergen, 2001; Chang & Karnell, 2001). Improved understanding of the psychological constructs and processes involved in vocal fatigue may help to partially explain individual differences in fatigue susceptibility.

Besides examining the underlying

mechanisms and salient features of vocal fatigue, research must address treatment issues. No empirical research exists specific to the treatment of vocal fatigue. Solomon and DiMattia's (2000) work provides some support for hydration as a useful management strategy. However, the forms of voice therapy or vocal endurance training to delay, attenuate, or prevent fatigue are unknown. How intervention should be tailored for different individuals is also unknown. Research in this area is

B. Sc. (Sp & Hg) students Is

1 0.47

2 0.37

3 0.44

4 0.33

5 0.49

Mean Is 0.42

http://en.wikipedia.org/wiki/Chemical_compound



http://en.wikipedia.org/wiki/Biochemistry

http://en.wikipedia.org/wiki/Milk

http://en.wikipedia.org/wiki/Koumiss

http://en.wikipedia.org/wiki/Yogurt

http://en.wikipedia.org/wiki/Kefir

http://en.wikipedia.org/wiki/Cottage_cheese

http://en.wikipedia.org/wiki/Casein


260

warranted to the appropriate management of all voice users experiencing vocal fatigue difficulties (Retrieved from http:// www.jvoice.org/article/ PIIS089219970300033X/fulltext).

Conclusions

The objectives of the present study were three-fold and included (a) investigation of the prevalence of vocal fatigue symptoms in SLP students, (b) quantification of the voicing periods (vocal load), and (c) determination of the short-term vocal recovery index in SLP students. Seventy four B.Sc. (Speech and Hearing) students were included in the phase 1. Out of 74, 60 students scored less than 17 and 14 students scored greater than 17, and the total prevalence rate of vocal fatigue symptoms was 18.91% (risk of developing voice disorder). Five students from phase 1 participated in the phase 2. The results indicated that the average percentage of voicing was 58.54%. There was no difference found in f0 measures between the pre- and post therapy, and within pre- and post therapy sessions. However, the mean and highest F0 were higher in post-therapy compared to pre-therapy sessions. The short-term vocal recovery index was found to be 0.42. Further vocal fatigue was reported by 19 % of subjects. The percent voicing duration was 52 to 69. There were no changes in mean, maximum, and minimum F0 across days and between pre- and post-therapy sessions. However, mean and maximum F0 increased from pre-to post therapy sessions in all subjects. The Is

varied between 0.33 to 0.47.

In the vocal mechanism, fatigue of the intrinsic and/or extrinsic laryngeal musculature could result in a reduced capacity to maintain tension in the vocal folds and stability in laryngeal posture. Thus, while no evidence exists from in vivo studies in humans, it may be that the human thyroarytenoid muscle is highly resistant to neuromuscular fatigue.

As the results of the present study were

obtained from lesser sample size, it cannot be generalized to all contexts in SLP students. Further, the percent voicing in normal speech depends upon the frequency of occurrence of voiced speech sounds in a language. Future research including males and various languages in India is warranted.

Acknowledgements

The authors would like to thank Dr. Vijayalakshmi Basavaraj, Director, All India Institute of Speech & Hearing, Mysore for granting permission to carry out this study. Also the authors acknowledge the subjects who participated in the study.

References Airo, E., Olkinuora, P., & Sala, E. (2000). A method to

measure speaking time and speech sound pressure level. Folia Phoniatrica Logopedica, 52, 275-288.

Boersma, P., & Weenink, D. (2009). PRAAT. Institute of phonetic sciences, University of Amsterdam. Retrieved July 20, 2009, from http://www.praat.org/.

Boominathan, P., Rajendra, A., Nagarajan, R., Jayashree, S., & Muthukumaran, G. (2008). Vocal abuse and vocal hygiene practices among different level professional voice users in India-a survey. Asia Pacific Journal of Speech Language and Hearing, 11(1), 47-53.

Buekers, R. (1998). Are voice endurance tests able to assess vocal fatigue? Clinical Otolaryngology, 23, 533-538.

Buekers, R. & Koten, J. W. (1991). De verzekeringsgeneeskundigebeoordeling van stem stoornlssen.Tijdschr Verzekeringsgeneesk, 29, 74-78.

Buekers, R., Bierens, E., Kingma, H., & Marres, E. (1995). Vocal load as measured by the voice accumulator. Folia Phoniatrica Logopedica, 47, 252-261.

Chang, A., & Karnell, M. P. (2001).The relationship among vocal fatigue, perceived phonatory effort and phonation threshold pressure (PTP). Paper presented at: The Voice Foundation's 30th Annual Symposium: Care of the Professional Voice, Philadelphia, PA.

Cooper, D. S., & Rice, D. H. (1990). Fatigue resistance of canine vocal fold muscle. Annals of Otology, Rhinology and Laryngology, 99, 228–233.

Doellinger, M. J., Lohscheller, A., Mc Whorter, M.,& Kunduk, M. (2009). Variability of normal vocal fold dynamics for different vocal loading in one healthy Subject investigated by phonovibrograms. Journal of Voice, 23(2), 175-181.

Edstrom, L., Lindquist, C., & Martensson, A. (1984). Correlation between functional and histochemical properties of the intrinsic laryngeal muscles in the cat. In B. Wyke (Eds.), Ventilatory and phonatory control systems. (pp. 392–404) London: Oxford University Press.

Enoka, R. M. (1994). Neurochemical Basis of Kinesiology. 2nd ed. Champaign, Ill: Human Kinetics.

Gelfer, M. P., Andrews, M. L., & Schmidt, C. P. (1991). Effects of prolonged loud reading on selected measures of vocal function in trained and untrained singers. Journal of Voice, 5, 158–167.

Hiroto, I., Toyozumi, Y., Tomita, H., Miyagi, T., Kuroki, K., & Koike, Y., (1969). An experimental study on the hemodynamics of the vocal fold during vibration. Japan Journal of Otolaryngology, 72, 884–888.

Hocevar-Boltezar, I. (2009). Prevalence and risk factors for voice problems in priests. Wiener Klinische Wochenschrift, 121, 7-8, 278-291.

Hunter, E, J., & Titze, I. R. (2009). Quantifying vocal fatigue recovery: dynamic vocal recovery trajectories after a vocal loading exercise. Annals of Otology, Rhinology and Laryngology, 118(6), 449-60.

Koufmann, J. A. (1998). What are voice disorders and who gets them? Retrived from www. Bgsm.edu/voice/voice disorders.html

Kowalska, M. S., Bogusz, E. N., Fiszer, M., Spychaska, T. L., Kotylo, P., Przygocka, B. S., & Modrzewska, M. (2006). The prevalence and risk factors for


261

occupational voice disorders in teachers. Folia Phoniatrica, 58, 85-101.

Lohscheller, J., Doellinger, M., McWhorter, A. J. & Kunduk, M. (2008). Preliminary study on the quantitative analysis of vocal loading effects on vocal fold dynamics using phonovibrograms. Annals of Otology, Rhinology and Laryngology, 117(7), 484-493.

Matsuo, K., Oda, M., Tomita, M., Machara, N., Umuzaki, T., & Shin, T. (1987). An experimental study of the circulation of the vocal fold on phonation. Archives of Otolaryngology and Head Neck Surgery, 113, 414–417.

Morrow, S. L., & Connor, N. P. (2009). Comparison of voice-use profiles between elementary classroom and music teachers. Retrieved from http://www.jvoice.org/article/S0892-1997.

Ohlsson, A. C., Brink, O., & Lofqvist, A. (1989). A voice accumulator-validation and application. Journal of Speech and Hearing Research, 32, 451–457.

Rajasudhakar, R., & Savithri, S. R. (2009). A method to measure voicing periods and vocal short-term recovery index. Journal of All India Institute of Speech & Hearing, 23, 1-9.

Rantala, L., Vilkman, E., & Bloign, R. (2002). Voice changes during work: Subjective complaints and objective measurements for female primary and secondary schoolteachers. Journal of Voice, 16(3), 344-355.

Russel, A., Oates, J., & Greenwood, K. (1998). Prevalence of voice problems in teachers. Journal of Voice, 12(4), 467-479.

Scherer, R. C., Titze, I. R., Raphael, B. N., Wood, R. P., Ramig, L. A., & Blager, R. F. (1986). Vocal fatigue in a professional voice user. In V. L. Lawrence (Eds.), (pp. 124–130). Transcripts of the Fourteenth Symposium: Care of the Professional Voice. New York: The Voice Foundation.

Scherer, R. C., Titze, I. R., Raphael, B. N., Wood, R. P., Ramig, L. A. & Blager, R. F. (1991). Vocal fatigue in a trained and an untrained voice user. In T. Baer, C. Sasaki, K. Harris (Eds.), Laryngeal function in phonation and respiration (pp. 533–555). San Diego, Calif: Singular Publishing Group.

Smith, E., Kirchner, H., Taylor, M., Hoffman, H., & Lemke, J. (1998a). Voice problems among teachers:

Differences by gender and teaching characteristics. Journal of Voice, 12(3), 328-334.

Solomon, N. P., Glaze, L., Arnold, R. & Mersbergen, M. (2001). Vocal fatigue and systemic hydration in men. Paper presented at: The Voice Foundation's 30th Annual Symposium: Care of the Professional Voice, Philadelphia, PA.

Solomon, N. P., & DiMattia, M. S. (2000). Effects of a vocally fatiguing task and systemic hydration on phonation threshold pressure. Journal of Voice, 14, 341–362.

Stemple, J. C. (1993). Voice therapy: Clinical studies, San Diego: Singular publishing group.

Stemple, J. C., Glaze, L., & Gerdeman, B. K. (1995). Clinical voice pathology: Theory and management, San Diego: Singular publishing group.

Titze, I. R. (1984). Vocal fatigue: some biomechanical considerations. In V. L. Lawrence (Eds.),. Transcripts of the Twelfth Symposium: Care of the Professional Voice (Part One: Scientific Papers). (pp. 97–104) New York: The Voice Foundation.

Titze, I. R. (1994). Principles of Voice Production. Englewood Cliffs, NJ: Prentice Hall.

Titze, I. R. (2002). How far can vocal folds travel? Presentation at ASHA conference. Retrieved from http://www.ncvs.org/museum-archive/ asha 2002 – irrt /frame.htm.

Titze, I. R., Svec, J., & Popolo, P. (2003). Vocal dose measures-Quantifying accumulated vibration exposure in vocal fold tissue. Journal of Speech and Hearing Research, 46, 922-935.

Vilkman, E. (2004). Occupational safety and health aspects of voice and speech professions. Folia Phoniatrica Logopedica, 56, 220-253.

Wingate, J. M., & Sapienza, C. (2007). Objective measurement of vocal load in teachers of singing and choral directors: A Pilot Study. Occupational voice conference, University of Texas at San Antonio.

http://projects.dlc.utsa.edu/ovc/Schedule.html.Retrieved from http:www.google.com

http:// en.wikipedia. org/ wiki/ Lactic_acid on 21st may. http://www.jvoice.org/article/PIIS089219970300033X/fullt

ext).

Appendix – Vocal Fatigue Questionnaire

Name: Age/Gender: Education: Voice use:

a) How many hours a day do you give therapy? b) How many hours a week do you give therapy? c) How many hours do you speak in a day? d) Do you have presentations during class hours?

If yes, specify no. of hours a day? e) Other types of work requiring voice use? (e.g. Singing)

If yes, specify no. of hours a day? Please underline appropriate response:

f) Do you take part in sports where you are required to shout? Often Occasionally Seldom Never

g) Do you watch or go to sports competition where you are required to shout? Often Occasionally Seldom Never

h) Do you smoke?

http://projects.dlc.utsa.edu/ovc/Schedule.html.Retrieved

http://www.jvoice.org/article/


262

i 20 or more/day ii 10 or 20/ day iii <10/day iv Occasionally v Given up smoking. How long ago? vi No

i) How many cups of coffee do you drink each day, approximately? j) Do you suffer from allergies?

Often Occasionally Seldom Never k) Do you suffer from asthma?

Often Occasionally Seldom Never l) Do you have hearing loss?

Often Occasionally Seldom Never m) Do you have gastro esophageal reflux disorder (GERD)?

Often Occasionally Seldom Never n) Have you consulted an ENT specialist because of voice problem?

How often in you carrier you visited an ENT specialist? Did you take medical treatment?

Vocal fatigue: Instructions: Please read the statements presented below. Circle one number to correspond with its statement, indicating how frequently you experience the situation described. Keys:

1. Less than once a year or never 2. A couple of times a year or occasionally 3. About once a month or quiet often 4. Almost every week or very often.

1. My voice gets tired after a long period (2 hours) of talking

1 2 3 4 2. My voice is hoarse or rough without infection/ disease/ sickness

1 2 3 4 3. I feel lump and/or mucous in the throat

1 2 3 4 4. When I talk a lot, my throat hurts/ pains

1 2 3 4 5. My voice does not have good quality in situations where much talking is needed

1 2 3 4 6. My voice does not penetrate noise/ audible in presence of noise

1 2 3 4 7. I have voice breaks when talking

1 2 3 4

8. I had aphonia (complete loss of voice) without infection/ disease/ sickness 1 2 3 4

9. I take sick leaves due to voice problems 1 2 3 4

10. When do you suffer from these symptoms? (Questions from 1 to 9) a) In the evening b) In the morning c) In the afternoon d) During the week ends e) While giving therapy f) During the summer

Nonword Repetition in Children with Language Impairment

263

Nonword Repetition in Children with Language Impairment: An Exploratory Study

Shylaja K. 1 & Swapna N. 2

Abstract

Children with Specific Language Impairment (SLI) are significantly delayed in acquiring multiple aspects of language even in the absence of mental retardation, sensory deficits, frank neurological damage and serious emotional problems. It is plausible that these language difficulties could arise due to limitations in Phonological Working Memory (PWM). Therefore this study was planned to investigate the PWM in children with language impairment (CWLI) and the children with normal language (CWNL) using a nonword repetition task (NWR). Three to seven year old Kannada speaking CWLI (n=9) and typically developing children (CWNL, n=10) with a language age of three to four years were administered the NWR task through the DMDX software. To examine the relationship between NWR accuracy and receptive vocabulary skills, a vocabulary test was administered. The responses of the children were recorded, transcribed, scored, and subjected to statistical analysis. The results indicated that the CWLI scored poorer on the NWR task compared to the CWNL especially on nonwords of 4 and 5 syllable length. In addition it was found that the CWLI had a higher percentage of vowel and consonant errors, compared to CWNL. No significant correlation was found between the NWR accuracy and receptive vocabulary scores. These findings have implications with respect the assessment and management of children with language impairment. Key words: language impairment, phonological working memory, nonword repetition

pecific 1Language Impairment (SLI, Leonard, 1998) is a condition in which the acquisition of language in children is neither rapid nor

effortless as is generally seen in typically developing children. These are the children who show significant language limitation in the absence of mental retardation, sensory deficits, frank neurological damage and serious emotional problems. Deficits in grammatical morphology, phonology, syntax, lexicon and pragmatic skills are observed in children with SLI (Joanisse & Seidenberg, 2003). Children with SLI also exhibit other types of deficits that extend beyond language including problems with working memory (Johnston & Weismer, 1983) and speech perception (Tallal & Piercy, 1974).

There is a general debate among researchers on the underlying cause for language impairment in children with SLI. A few researchers claim that language impairment in children with SLI arises as a result of deficits in linguistic knowledge (van der Lely & Stollwerck, 1997; van der Lely, 2005), while others explain SLI in terms of domain-specific cognitive-linguistic processes.

One of the language specific cognitive

processing areas that have received much attention in the past is the Phonological Working Memory (PWM)

1e-mail: [email protected]; 2Lecturer in Speech Pathology, AIISH, Mysore, [email protected]

(Gathercole, 2006). A deficit in PWM means a decreased capacity to store phonological information which affects both the acquisition of new words (which demand the retention of new phonological sequences) and broader levels of language processing such as sentence comprehension that require the manipulation of phonological information (Briscoe, Bishop, & Norbury, 2001). In the recent past, the assessment procedures of SLI have gained momentum and specifically, the phonological working memory has been studied by a number of researchers using different methods and stimuli. One such diagnostic tool that has been recently researched upon is the use of nonword repetition as a clinical marker in children with SLI.

Nonword repetition task (NWR) requires

perception, storage and retrieval of its phonological constituents in a sequence and it is proposed as a potential screening tool to identify the deficits related to phonological working memory in children with SLI (e.g., Botting & Conti-Ramsden, 2001). Several investigations using NWR in English speaking children with SLI have documented that these children are less proficient at nonword repetition than normal language peers there by indicating that they have an apparent deficit in their phonological working memory (PWM) (e.g., Dollaghan & Campbell, 1998; Marton & Schwartz, 2003; Gathercole, 2006). An association has also been found between nonword repetition and language skills in school-age children with both typical and atypical language development (Montgomery, 2002). Studies pertaining to NWR task have been also

S


264

attempted in languages other than English, such as Spanish, Dutch and Swedish and NWR task has been found to be an identifier of SLI in these languages (Sahlen, Wagner, Nettelbladt & Radeborg, 1999; de Bree, Rispens & Gerrits, 2007; Girbau & Schwartz, 2008).

However, a study carried out in Cantonese by

Stokes, Wong, Fletcher and Leonard (2006), wherein they examined fourteen children with SLI, thirty typically developing age matched and younger peers using NWR and sentence repetition tasks and found that there was no limitation in phonological working memory in these children with SLI. They attributed the poorer NWR for English-speaking children with SLI to the weaker use of redintegration strategy in nonword repetition. They recommended further cross-linguistic investigations of processing strategies.

In the Indian context very limited studies have

attempted to investigate the nonword repetition performance in children with SLI. One such study by Prema, Prasitha, Savitha, Purushotham, Chitra and Balaji, (2010) was conducted on a 14 year old male child with SLI who was a native speaker of Kannada, a South-Indian Dravidian language. They examined his performance on a NWR task with 15 nonwords of 3-syllable length each and found that he had poor nonword repetition accuracy compared to the typically developing child, matched on age, gender, language, socio-economic status and grade. They also recommended further research with large sample size to corroborate the results.

A look into the review of literature suggests

that NWR has been studied extensively in English speaking children with SLI. Moreover, some research has also been carried out in other languages such as Swedish, Spanish, Dutch and Cantonese. The results of some of these studies are contradictory especially the study carried out in Cantonese SLI population. These studies suggest that NWR tests may depend on the linguistic and the paralinguistic aspects of the language such as prosodic properties, phonotactic structure and the phonetic inventory of the language. This calls for the question whether NWR is language dependent or processing dependent. Since the linguistic and paralinguistic aspects of the languages have an influence on the PWM, it is essential that such studies be replicated in other languages to examine the same. Moreover in the Indian context, such studies are limited. Hence this study was planned with the objectives of examining the performance of Kannada speaking children with and without language impairment on a nonword repetition task and to determine the relationship between the nonword

repetition accuracy and receptive vocabulary in these children.

Method

Participants: A total of nineteen Kannada speaking children with chronological age ranging between 3 to 7 years served as subjects for the study. The clinical group consisted of nine children in the age range of 4.5 to 7 years diagnosed as Delayed speech and Language (language impairment) by a qualified team of professionals including speech-language pathologist and psychologist. They were matched for language to the control group by determining their language age with the help of a standardized test of language viz. Kannada Language Test (Karanth, 1995) a diagnostic language tool. The language age of all the children ranged from 3-4 years and all the children had average intelligence quotient. Subjects were also evaluated for their articulation abilities using the Kannada Articulation Test (Babu, Rathna & Bettageri, 1972) and only those children with known minimal consistent articulatory errors and whose speech could be easily understood were included in the group. The control group consisted of ten typically developing children with normal hearing and normal receptive and expressive language skills, matched for language age and socioeconomic status. They were mainly recruited through nursery and kindergartens.

The children included in both the groups had no history of sensory, intellectual, neurological, medical, oro-motor, emotional, or behavioral disturbances. In addition the WHO Ten-question disability screening checklist (Singhi, Kumar, Malhi & Kumar, 2007) was used to rule out any disability for the children in the control group. All the children in the clinical group had attended speech-language therapy for an average duration of one week (demonstration therapy) and were advised to continue to train the child at home. Ethical procedures were used to select the participants. The parents were explained the purpose and the procedures of the study and an informed verbal consent were taken.

Material : The nonwords used in the study were constructed by selecting meaningful words from Computerized Linguistic Protocol (in Kannada) for Screening Children (CLIPS) (Anitha & Prema, 2008) and words from „With a little bit of help-Early Language Training Manual‟ (Karanth, Manjula, Geetha & Prema, 1999). A total of 52 meaningful words were selected and these were converted to nonwords by transposing the vowel/syllables, depending on the word length, e.g., mane (word) to mena (nonword), ki職aki (word) to 職akiki (nonword). The nonwords included


265

followed the phonotactic rules of the Kannada language. Within a given nonword, no consonant occurred more than once and the consonants of the original word were maintained. The list of the 20 nonwords was such that none of their individual syllable combinations (CV or CVC) corresponded to a meaningful Kannada word. This was done to ensure that the nonwords included were not affected by a subject‟s vocabulary nowledge. In addition, the nonwords did not include consonant clusters and contained sounds that were within the phonetic inventory of the children selected. These 52 nonwords prepared were subjected to a judgement on word-likeliness on a 4-point rating scale by five adult native speakers of Kannada and the words which were rated with a point of „0‟ (not at all similar to any meaningful Kannada word) or „1‟ (slightly or 25% similar to a meaningful Kannada word) were included in the final list of nonwords.

The final list contained a total of 25 nonwords, with 20 nonwords as the test items (5 under each of the syllable lengths used) and 5 nonwords as the practice items (Included in Appendix). The prepared list of nonwords were then audio-recorded by a female native speaer of Kannada, using the “PRAAT” software (downloadable software for speech recording and analysis) loaded in the Compaq Presario C700 laptop system and then loaded into DMDX software to maintain a constant inter-stimulus interval of 4sec. A pilot study was carried out to ensure that the nonwords can be repeated by 3-4 year old group of normal children with ease and to check whether the interstimulus interval was adequate or not to repeat the nonwords. Procedure: The list of recorded nonwords along with five practice items were presented using DMDX software, through headphones auditorily at the comfortable listening level to the individual participants in a quiet listening environment. Each participant was given the recorded instructions in Kannada through headphones as following: “You are going to listen to some funny made-up words. Your job is to say them back to me, exactly the way you hear them. Some of the words will be short, and others will be longer. Listen carefully, because the words will be said only once. Here comes the first word.” The list of five practice items was presented first followed by the test items. The responses were audio recorded directly into the DMDX software. No feedback was given on the test items, but encouragement was given as required. The nonwords were presented in the order of increasing difficulty (all two-syllable non words, followed by three-syllable nonwords etc.). In addition, the subjects were evaluated for their receptive

vocabulary using KPVT- A Screening Picture Vocabulary Test in Kannada (Sreedevi & Karanth, 2002). The test consisted of thirty pictures. The subject was instructed to point to the appropriate picture named by the experimenter from a set of four pictures. The total time taken to complete the nonword repetition task and vocabulary assessment task was approximately 20 minutes. Analysis: The participants‟ productions were audio recorded and transcribed verbatim using broad phonetic transcription by the experimenter. The audio recorded responses were analyzed for the accuracy of the repetition and the type of errors. Accuracy of the response: This was calculated as the whole word correct or incorrect out of total of 20 nonwords and also at each syllable length. Exact repetition of all the syllables in a nonword was assigned a score of „1‟. Any syllable substitutions, omissions, and additions in a nonword were considered as incorrect and scored „0‟. The distortions of a syllable and segmental substitutions based on normal phonological processes were counted as correct. Error analysis: The total number of vowels and consonants repeated correctly and the total number of different types of errors such as substitutions, omissions, and additions were calculated across the different syllable lengths. The raw scores were converted into percentage of vowels/consonants correct by dividing the number of vowels/consonants correct by the total number of vowels/consonants and multiplying it by 100. The total percentage of different errors was also computed in a similar manner for each subject for the entire set of nonwords and also at each of the different syllable length.

The subject‟s vocabulary score was computed

from the KPVT. A correct response was scored as „1‟and an incorrect response was scored „0‟. The total number of pictures identified correctly was determined and the total score was computed.

The obtained data were appropriately tabulated

and subjected to statistical measures. SPSS software (version 16.0) package was used for statistical analysis. Mean and standard deviation values were computed. Independent t-test, Repeated Measure ANOVA, Boneferroni‟s pairwise comparison test, Karl Pearson‟s correlation coefficient, MANOVA and Paired t-test were used to answer the research questions of the present study.


The results of the statistical analysis for both groups on both the tasks have been presented and discussed under various sections:


266

I. Overall accuracy of responses: The mean and standard deviation values were computed and independent t-test was administered to determine whether any significant difference existed in the performance between the two groups. The values for both the groups are depicted in Table 1. The overall mean score for accuracy for the nonword repetition task for the CWLI group was 9.78 (SD=4.41), which was lower than the mean of the CWNL group which was 15.20 (SD=2.20). This indicated that the CWLI had lower accuracy than the CWNL for the nonword repetition task and the same has been depicted in the Figure 1.

These results are in consonance with the studies by Dollaghan and Campbell (1998); Sahlen, Wagner, Nettelbladt and Radeborg (1999); Marton and Schwartz (2003); de Bree, Rispens and Gerrits (2007); Girbau and Schwartz (2008). According to Edwards and Lahey (1998) the cognitive complexity of a nonword repetition task overtaxes the general processing resources of children with SLI, thereby hindering their ability to create and thus store accurate phonological representations of unfamiliar input and hence poor performance on nonword repetition task. Thus the results replicates the results of earlier studies which indicate that the deficit in phonological working memory capacity plays a role in the language impairment in children with SLI leading to their poorer performance on nonword repetition task (Gathercole & Baddeley, 1990a, 1990b). On specific examination of the performance of the two groups across different syllable lengths, it was seen that there was no statistically significant difference in the performance on 2- and 3-syllable length nonwords between the groups (p>0.05). However there was a significant difference in the performance between groups on the repetition of 4- and 5-syllable length nonwords, which has been depicted in Table 1.

This is in consonance with the study done in Dutch where they reported that there was a sharp decrease in the accuracy of the responses at a target length of 4-syllable nonwords in children with SLI (de Bree, Rispens & Gerrits, 2007). However, studies done in English and Spanish languages suggest that the children with SLI have difficulty in repeating nonwords of 3-syllable length and above (Dollaghan & Campbell, 1998; Girbau & Schwartz, 2007). This could be attributed to the differences in the complexity of the syllable structure between the Kannada and English languages; Kannada is a syllable-timed language considered to be having relatively simple syllabic structure, e.g., CVCVCV pattern than English which is a stress-timed language (Abercrombie, 1967).

Table 1. Mean, standard deviation (SD), and t-values for accuracy of responses on the nonword repetition

task in both the groups Accuracy of responses

Group

Mean

SD

t-values(17)

A2s CWLI 3.56 1.42 1.41 CWNL 4.30 0.82

A3s CWLI 3.22 1.30 1.75 CWNL 4.20 1.14

A4s CWLI 2.00 1.00 4.58* CWNL 4.10 0.99

A5s CWLI 1.00 1.32 2.65* CWNL 2.40 0.96

Overall accuracy

CWLI 9.78 4.41 3.45* CWNL 15.20 2.20

Note: A2s- accuracy at 2-syllable length nonwords; A3s-accuracy at 3-syllable length nonwords; A4s- accuracy at 4-syllable length nonwords; A5s-accuracy at 5- syllable length nonwords, *p<0.05.

Figure 1. Mean accuracy of responses on the nonword repetition task in both the groups.

However, these results are in contrast to the study by Stokes, Wong, Fletcher and Leonard (2006) who found that Cantonese speaking children with SLI performed at par with the typically developing age matched children. They concluded that the NWR task in Cantonese does not tax the working memory in the same way that nonwords do in other languages such as English and Swedish which has complex phonotactic

structures, variable stress patterns, prosodic (temporal and sequential properties) and difficult-to-articulate

consonants compared to the Cantonese language. They also suggested that the other possible factors contributing to the better performance of English speaking typically developing age-matched children on nonword repetition task was that the target nonwords in their test stimuli had items which were more similar to the real words. Hence the typically developing children could relate the target nonwords to the long-


267

term memory language store and used lexical and phonotactic information to "fill in the blanks" of the skeletal score (the CVC pattern), creating either an accurate response or a close approximation to the target nonword. This process of redintegration was lacking in the children with SLI lead to their poorer performance on NWR compared to the typically developing children. Hence the results of the present study with respect to the accuracy of responses in NWR task seemed to follow the same pattern as in English, Spanish, Dutch and Swedish.

II. Accuracy across syllable length The mean values of accuracy for both the

groups across the syllable length showed a decreasing trend with the increase in the syllable length. The mean values depicted in Table 1 represent the decrease in the accuracy of the responses with the syllable length in both the groups. The same has been depicted in Figure 2.

Repeated Measure ANOVA was used to

examine if any significant difference existed across each syllable length in both CWNL and CWLI group. The results indicated that there was a significant difference in the performance of CWNL [F (3, 27) =10.50, p<0.001)] and in CWLI [F (3, 24) =23.03, p<0.001)] across syllable length. In order to determine, which two pairs were significantly different from each other, Bonefferoni‟s pairwise comparison test was used. The results indicated that the performance of the CWNL was significantly different only on the 5-syllable length nonwords compared to 2-, 3- and 4-syllable length nonwords (p<0.05) whereas the performance of the CWLI was significantly different on 4-syllable length nonwords compared to 2- and 3-syllable length and also on 5-syllable length nonwords compared to 2- & 3-syllable length nonwords (p<0.05). These values are depicted in Table 2.

These results are in agreement to earlier studies

in which it was reported that because of the limited capacity nature of phonological short-term memory the typically developing children perform better on shorter syllable length nonwords than longer syllable length nonwords (e.g., Gathercole, 2006). Further the greater repetition decrement for lengthier nonwords in children with SLI could arise either from accelerated rates of decay before output or from inadequate encoding in the short-term store. Moreover the unfamiliarity of phonological structure of nonwords forces participants to rely heavily on temporary phonological representations to support their repetition attempt, preventing the reliance on activated lexical representations that arises in memory tasks using

familiar verbal stimuli (e.g., Hulme, Maughan & Brown, 1991).

Figure 2. Mean accuracy of the nonword repetition task at different syllable lengths in both

the groups.

Table 2. Pairwise comparison of the syllable lengths in both the groups

Syllable length

CWNL CWLI

MD Level of

significance MD

Level of significance

2s

3s 0.10 1.00 0.33 1.00

4s 0.20 1.00 1.56 0.02*

5s 1.90 0.01* 2.56 0.001*

3s 4s 0.10 1.00 1.22 0.01*

5s 1.80 0.05* 2.22 0.00*

4s 4s 1.70 0.001* 1.00 0.10

Note: 2s- accuracy at 2-syllable length nonwords; 3s-accuracy at 3-syllable length nonwords; 4s- accuracy at 4-syllable length nonwords; 5s-accuracy at syllable length nonwords; MD- Mean difference, *p<0.05.

The CWNL had poorer repetition accuracy in nonword repetition at only 5-syllable length nonwords, where as the CWLI had poorer repetition accuracy at both 4- and 5-syllable length nonwords. This indicates a more limited capacity of the phonological loop function in children with SLI than in children with typical language development, similar to the results of the study by Marton and Schwartz (2003).


268

III. Error analysis a. Percentage of Vowels Correct (PVC) and Percentage of Consonants Corrects (PCC) across both the groups

The mean and standard deviation values for overall PVC and PCC for the nonword repetition task and the PVC and PCC at each syllable length for both the groups are shown in Table 3. The overall mean scores of PVC and the mean values at all the syllable lengths were higher for the CWNL when compared to the CWLI. The PVC also decreased from 2-syllable nonwords to 5-syllable length nonwords, that is, the errors increased from shorter syllable length to the longer syllable length nonwords in the CWLI group. MANOVA indicated a significant group difference in the total PVC for the overall nonword repetition task, PVC at the 3-syllable length and 5-syllable length nonwords ( p<0.05).

The mean values for overall PCC and the mean of PCC at different syllable lengths revealed similar pattern as that of PVC, i.e. the PCC was higher in CWNL than in CWLI which is depicted in Table 3. In addition, the PCC was least at the 5-syllable nonwords for the CWLI group. Analysis done using MANOVA indicated that there was a significant difference between both the groups in the total PCC and also for the PCC at the 5-syllable length nonwords (p<0.05). Table 3. Mean, standard deviation (SD) and F-values

indicating the PVC and PCC at different syllable lengths for both the groups

Syllable length

Group PVC PCC

Mean SD F(1,17) Mean SD F(1,17)

2s CWLI 92.22 10.93

0.16 73.33 26.46

1.65 CWNL 94.00 8.43 85.00 10.80

3s CWLI 88.89 12.01

4.55* 86.67 21.60

1.45 CWNL 97.34 3.44 95.33 7.06

4s CWLI 78.34 25.86

3.30 73.33 22.36

4.32 CWNL 94.10 8.43 89.50 9.85

5s CWLI 63.56 25.49

6.53* 58.22 27.21

9.94* CWNL 85.20 8.01 87.60 10.91

Overall CWLI 77.80 18.20

5.78* 70.00 22.45

7.06* CWNL 91.57 4.54 89.43 5.48

Note: PVC - Percentage of Vowels Correct; PCC - Percentage of Consonants Correct; 2s- 2-syllable length nonwords, 3s- 3-syllable length nonwords; 4s-4-syllable length nonwords; 5s- 5-syllable length nonwords; *p<0.05.

The results of the present study are in consonance with the study done by Girbau and Schwartz (2008) who found that the children with SLI aged 8.10years, though scored higher on vowels compared to consonants, did not perform on par with the children with typical language development who had almost 100% score on the percentage of vowels

correct. They also reported that the children with SLI made more consonant errors overall and in the 3-, 4-, and 5 -syllable nonwords. b. Comparison between total PVC and PCC in the

CWNL and CWLI group The CWNL group had higher PVC (Mean =

91.57, SD = 4.54) than PCC (Mean = 89.43, SD = 5.48). Paired t-test used indicated that there was significant difference between total PVC and PCC in the CWNL group at [t (9) = 2.764, p<0.05]. The CWLI group also had higher PVC (Mean = 77.30, SD = 18.20) than total PCC (Mean = 70.00, SD = 22.4) and the paired t-test indicated a significant difference in the total PVC and PCC at [t (8) =3.68, p<0.05]. This suggests that both the groups had more difficulty in repeating consonants than vowels. The results of the present study are in consonance with the study done by Girbau and Schwartz (2008) who concluded that vowels are preferentially preserved in the phonological working memory task in children with SLI and children with typical language development. c. Percent Syllable Substitution (PSS), Percentage of Syllable Addition (PSA) and Percentage of Syllable Omission (PSO)

The mean and standard deviation scores depicted in Table 4 indicated that the CWLI had higher PSS than CWNL overall and at each syllable length. MANOVA indicated that there was a significant difference in the PSS only at the 5-syllable length [F (1, 17) = 5.84, p<0.05]. The results of the present study are similar to the studies by Marton and Schwartz (2003) and Girbau and Schwartz (2008). They found that the children with SLI produced more percentage of substitutions than children with typical language development overall and in the 3-4- and 5-syllable nonwords.

The mean and standard deviation scores for percentage of syllables addition (PSA) and percentage of syllable omission (PSO) indicated that the CWLI had higher PSA and PSO than CWNL (Table 4). The results of MANOVA revealed that there was no significant difference in the PSA and PSO between the groups (p>0.05). The results by Girbau and Schwartz (2008) also reported of lesser addition errors in the SLI group and hence the results with respect to PSA is in agreement with this study.

d. Predominantly occurring errors

The mean and standard deviations values depicted in Table 4 revealed that both the groups had highest percent of syllable substitution (PSS) [CWNL: mean=10.57, S.D=4.48; CWLI: mean= 20.64, S.D= 14.64] followed by PSO errors [CWNL: mean=2.29,


269

S.D=2.63; CWLI: mean= 12.86, S.D= 19.74] and PSA [CWNL: mean=0.43, S.D=0.69; CWLI: mean= 1.75, S.D= 3.18] which were the least.

Table 4. Mean and standard deviation (SD) values indicating the PSS, PSA and PSO across different

syllable lengths in both the groups Syllable length

Group PSS PSA PSO


2s CWLI 20.00 11.18 2.22 4.41 8.89 20.28 CWNL 14.00 11.74 1.00 3.16 0.00 0.00

3s CWLI 12.59 16.81 0.74 2.22 5.19 13.24 CWNL 4.00 6.44 1.33 2.81 0.00 0.00

4s CWLI 21.67 14.14 3.33 8.30 12.22 25.99 CWNL 10.50 12.79 0.00 0.00 2.00 4.22

5s CWLI 31.11 22.78 0.89 2.67 19.56 27.67 CWNL 12.80 7.25 0.00 0.00 4.80 5.59

Overall CWLI 20.64 14.64 1.75 3.18 12.86 19.74 CWNL 10.57 4.48 0.43 0.69 2.29 2.63

Note: PSS - Percentage of Syllable Substitutions; PSA - Percentage of Syllable Additions; PSO - Percentage of Syllable Omission; 2s- 2-syllable length nonwords, 3s- 3-syllable length nonwords; 4s-4-syllable length nonwords; 5s- 5-syllable length nonwords

Repeated measure ANOVA indicated a

significant difference in the type of error in both CWNL group [F (2, 18) = 53.29, p<0.05] and CWLI group [F (2, 16) = 4.366, p<0.05]. Boneferroni‟s pairwise comparison test in the CWNL group indicated a significant difference between the PSS and the other two types of errors (p<0.05). In the CWLI group Boneferroni‟s pairwise comparison test, revealed a significant difference only between the PSS and the PSA (p<0.05). These results are depicted in Table 5.

The result of the present study is in congruence

with the results of the earlier studies done by Marton and Schwartz (2003) and Girbau and Schwartz (2008) who found that consonant substitutions were the most frequent type of error found in both the children with typical language development and children with SLI. These errors in children with SLI were attributed to the some underlying weakness in phonological knowledge or memory for phonological information, including the nature of phonological representations in working memory (Edwards & Lahey, 1998; Ellis Weismer & Edwards, 2006).

Table 5. Results of Boneferroni‟s pairwise comparison test for the different types of errors in both the groups

Note: PSS - Percentage of Syllable Substitutions; PSA - Percentage of Syllable Additions; PSO - Percentage of Syllable Omission; MD- Mean Difference; LS-Level of Significance, *p<0.05. IV. Relationship between accuracy of the nonword repetition and receptive vocabulary

The mean scores on the KPVT- A Screening Picture Vocabulary Test in Kannada (Sreedevi & Karanth, 2002) indicated better performance of the CWNL group (Mean= 17.40, SD=2.41) compared to the CWLI group (Mean= 14.33, SD=2.12). Further analysis using independent samples t-test revealed a significant difference in the performance between the groups on the vocabulary test [t (17) = 2.93, p<0.05].

Karl Pearson‟s correlation coefficient was used

to determine the relationship between nonword repetition accuracy and receptive vocabulary scores. The results indicated that there was no significant correlation between these two measures in both CWLI (r= 0.109, p>0.05) and CWNL (r=0.249, p>0.05) group. The same is depicted in the Figure 3.

Figure 3. Scatter plot between the nonword repetition

and receptive vocabulary scores.

Pairwise comparisons of error types

CWNL CWLI

MD LS MD LS

PSS PSA 10.14 0.00* 18.89 0.007* PSO 8.29 0.00* 7.78 1.00

PSA PSO 1.86 0.12 11.11 0.36


270

The above results are in consonance with the results obtained by Metsala (1999) and Sahlen, Wagner, Nettelbladt and Radeborg (1999). Sahlen et al. (1999) found a weaker correlation between vocabulary and fable comprehension and nonword repetition than between non-word repetition and comprehension of grammar. They concluded that tests assessing the comprehension of grammar strained language processing and storage more than the other tests. However, these results are in contrast to the earlier studies where significant correlation was found between nonword repetition accuracy and receptive vocabulary in typically developing children (Roy & Chiat, 2004). The possible factors contributing to the lack of correlation between the nonword repetition and receptive vocabulary scores especially in the CWNL in the present study was the inappropriate and lack of clear representation of the target words in the picture form in the KPVT. Further, KPVT is a relatively old test and a few of the target word stimuli used in the test were found to be inappropriate and unfamiliar to the present day children as these words are rarely used today due to modernization. Hence the children with age appropriate language skills also obtained poorer scores in the receptive vocabulary test.

Conclusions

It can be concluded from this study that Kannada speaking children with language impairment exhibit a deficit in the phonological working memory which is revealed through their poor performance on the nonword repetition task. The longer syllable length nonwords (4- and 5-syllable length nonwords) were found to be better indicators of phonological working memory deficit in CWLI. Hence it is recommended that the nonword repetition test be included in the routine clinical assessment procedures. Assessing phonological working memory skills using nonwords in children with language impairment may help us to predict whether the children might be at risk for specific language impairment and further have greater language and literacy deficits. This study has implications in intervention as along with teaching language, clinicians can also incorporate tasks which require mental manipulation of language and nonword/sentence material which would in turn improve the overall processing abilities in terms of phonologically encoding and representing novel material. Further research work can be undertaken to estimate the developmental norms for the nonword repetition and evaluate the clinical utility of the same in the Indian scenario.

References Abercrombie, D. (1967). Elements of General Phonetics.

Chicago: Aldine Pub. Co. Anitha, R., & Prema, K.. S. (2008). Computerized Linguistic

Protocol (in Kannada) for Screening (CLiPS). Student research at AIISH (Articles based on dissertation done at AIISH), Vol II: 2003-2004, 110-120.

Babu, P. R. M., Ratna, N., & Bettagiri, R. (1972). Test of Articulation in Kannada. The Journal of All India Institute of Speech and Hearing, 111, 7-20.

Botting, N., & ContiRamsden, G. (2001). Non-word repetition and language development in children with specific language impairment (SLI). International Journal of Language and Communication Disorders, 36 (4), 421- 432.

Briscoe, J., Bishop, D. V. M., & Norbury, C. F. (2001). Phonological processing, language, and literacy: A comparison of children with mild-to-moderate sensorineural hearing loss and those with specific language impairment. Journal of Child psychology and Psychiatry, 42, 329- 340.

de Bree, E., Rispens, J., & Gerrits, E. (2007). Non-word repetition in Dutch children with (a risk of) dyslexia and SLI. Clinical Linguistics and Phonetics, 21(11-12), 935-44.

Dollaghan, C., & Campbell, T. (1998). Nonword repetition and child language impairment. Journal of Speech, Language, and Hearing Research, 41, 1136-1146.

Edwards, J., & Lahey, M. (1998). Nonword repetitions of children with specific language impairment: Exploration of some explanations for their inaccuracies. Applied Psycholinguistics, 19, 279- 309.

Ellis Weismer, S., & Edwards, J. (2006). The role of phonological storage deficits in specific language impairment: reconsideration. Invited commentary on S.E. Gathercole, Nonword repetition and word learning: The nature of the relationship. Applied Psycholinguistics, 27, 556- 562.

Gathercole, S. E., & Baddeley, A. (1990a). Phonological memory deficits in language impaired children: Is there a causal connection? Journal of Memory and Language, 29, 336- 360.

Gathercole, S. E., & Baddeley, A. (1990b). The role of phonological memory in vocabulary acquisition: A study of young children learning new words. British Journal of Psychology, 81, 439–454.

Gathercole, S. E. (2006). Nonword repetition and word learning: The nature of the relationship. Applied Psycholinguistics, 27(4), 530- 540.

Girbau, D., & Schwartz, R. (2007). Non-word repetition in Spanish-speaking children with specific language impairment (SLI). International Journal of Language and Communication Disorders, 42, 59– 75.

Girbau, D., & Schwartz, G. R. (2008). Phonological working memory in Spanish-English bilingual children with and without specific language impairment. Journal of Communication Disorders, 41, 124-145.

Hulme, C., Maughan, S., & Brown, G. D. (1991). Memory for familiar and unfamiliar words: Evidence for a long-term memory contribution to short-term memory span. Journal of Memory and Language, 30 (6), 685- 701.


271

Joanisse, M. F., & Seidenberg, M. S. (2003). Phonology and syntax in specific language impairment: evidence from a connectionist model. Brain and Language, 86 (1), 40-56.

Johnston, J. R., & Weismer, V. (1983). Cognitive development in pre-adolescent language impaired children. British Journal of Communication Disorders, 18 (1), 49-55.

Karanth, P., Manjula, R., Geetha, Y.V., & Prema, K.. S. (1999). With a little bit of help- Early language training manual. Books for change, Bangalore.

Karanth, P. (1995). Development and standardization of language and articulation test in Indian languages. An inter-institutional project: AIISH, Mysore in collaboration with AYJNIHH, Mumbai and RRTC, Madras. Funding agency- Unicef.

Leonard, L. (1998). Children with specific language impairment. Cambridge, MA: MIT Press.

Marton, K., & Schwartz, G. R. (2003). Working memory capacity and language processes in children with specific language impairment. Journal of Speech, Language and Hearing Research, 46, 1138-1153.

Metsala, J. L. (1999). Young children‟s phonological awareness and nonword repetition as a function of vocabulary development. Journal of Experimental Child Psychology, 91 (1), 3-19.

Montgomery, J. (2002). Understanding the language difficulties of children with specific language impairments: Does working memory matter? American Journal of Speech-Language Pathology, 11, 77– 91.

Prema, K. S., Prasitha, P., Savitha, S., Purushotham, A. N., Chitra, P., & Balaji, R. R. (2010). Clinical markers for identification of children with Specific Language

Impairment (SLI). Indian Journal of Applied Linguistics, 36 (1-2), 181-198.

Roy, P., & Chiat, S. (2004). A prosodically controlled word and nonword repetition task for 2- to 4-year-olds: Evidence from typically developing children. Journal of Speech, Language, and Hearing Research, 47, 223-234.

Sahlen, B., Wagner, C. R., Nettelbladt, U., & Radeborg, K. (1999). Language comprehension and non-word repetition in children with language impairment. Clinical Linguistics and Phonetics, 13 (5), 369 – 380.

Singhi, P., Kumar, M., Malhi, P., & Kumar, R. (2007). Utility of the WHO ten questions screen for disability detection in rural community-The North Indian experience. Journal of Tropical Pediatrics, 53(6), 383-387.

Sreedevi, N., & Karanth, P. (2002). KPVT-A Screening Picture Vocabulary Test in Kannada. Research at AIISH (Dissertation abstracts), Vol II, 148.

Stokes, S. F., Wong, A. M-Y., Fletcher, P., & Leonard, L. B. (2006). Nonword repetition and sentence repetition as clinical markers of specific language impairment: the case of Cantonese. Journal of Speech and Hearing Research, 49 (2), 219 - 236.

Tallal, P., & Piercy, M. (1974). Developmental aphasia: Rate of auditory processing and selective impairment of consonant perception. Neuropsychologia, 12, 83-93.

van der Lely, H. K. J., & Stollwerck, L. (1997). Binding theory and grammatical specific language impairment in children. Cognition, 62 (3), 245-290.

van der Lely, H. K. J. (2005). Domain-specific cognitive systems: Insight from grammatical specific language impairment. Trends in Cognitive Sciences, 9(2), 53-59.


272

Appendix: Nonword List

Practice Items

Test Items

Syllable lengths

Nonwords

2-syllable

bussa

3-syllable

lana:gi

4-syllable

ka召annaka

4-syllable

d索 uno:vu召u

5-syllable

t索 in�a:ginaji

Syllable lengths Nonwords

2-syllable

nija:

mina

nuga

bija:

t索 ila

3-syllable

職akiki

t索 ipa:t検a

lit索 t索 a�a

ku:lud索 a

lippat検a

4-syllable

nu召召ad索 ova

vakkat検inu

d索 unnuvut索 i

ginna:d索 it検i

d索 uvu召uko

5-syllable

gid索 ujunivu

jut索 avud索 ugi

nimagid索 d索 a:la

�od索 d索 a:t索 t索 i召uni

d索 ujuvu召iku

��AB�CB�DB��E�C��F��

�

��

��A�BAC�D�E�AF��D��A��E�A��A��A�CF��A

A�E�A��A��A�A A��A!�A"�A

#AA

AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA

F$DE�%A

FA�� B�� B�� AB�� AB�� B��C�� B��B�� C��C�A� BAB�CB�� A��

�B��B��C�C��C��C�A�BAB�CB��B�C��CB��B��B��A��E��B��B��

��C��AB�CB�DB��E�� !�"�#��$%&'(�C��F��FA��C��$(�)��B��A�

�*(��C��E�+��A��C��,(�-��C�C��'(�.B�C��/(�-�B�C��0(��C�C��&(�#�B1C��B��2(�+��C��B��

�C��3�B�CB��B��+B��B�C��B��FA��B��B��C�C��B��$44��B��B��F��B��C��

�A�� B�� B�� *4� 5� &4� E�B�� A�� B��C!�� C�� C�� B�� C�A� *4� ��6�� $4� �B�� B�� $4�

��B��(�C��B�A��A��6��B�C��B��F��B��*4�B��BAB�C��B�C��A�

��"��3��(��C��C�AC��A��B��B��,2�5�&4�E��FA��AB�CB�DB��E�C��

F��B��B��C�C��B��B��AB��AC��B��E��B��B��B��C�E��B�C��BAB�C��E��

B��A��E��7��A��B1��C��B��C�C�B�C��

�

��A�BCDBEBF�B��F��BCDBEB��B��F�BCDBEB��

��

��ABC�DE� �FB� E�C�� B� ��B� ��E��

B��A�F��B�� E��CE�C��B�� BFE��C�B� �� FB��C�B�

�B��E� �� C��C�� C�� CE� ��B�

��EECA�B� A�� BCF� ��FB� ��B�� B�F��EC��DC��

�B��CE�� D��DB� CE� ��EC�BFB�� E� �� FC��F��

�B��E� �� C��C�� C�� CE� �� F�� E��C��

AB��C�F��FB��BF� ��D��DB� CE� �� EB�� E��A��E� ��

��B��C��FB�ABC�D�B��B��A��ABC�DE��

�FB� ��A�B� �� C�D� �EE��C��C�� AB��BB�� BEEB��C��

�FAC�F�F�� FB�FBEB��C��E� �� B�B��E� �� B��FBEE� ��BCF�

��D��E�� BCF��CE�BE��BCF��BB�C�DE��C��FBE��E�

C�� FB� B��B��C�B� ��F�� C��C�� !�CE� CE�

��EC�BFB��E��B��B�"��C#�B$��B��FBE��C��

�BF��C��E��

�

%��EC�� CE��B�C�B��E� &��B� ��EE��F��B�BFC�F��C��

�� BFA�� C��C��B� �� #�CFB�� BEC��

��B� �BF��E� E�E�B�� C��C�D� ��B� �F� ��FB� �E�B��E� ��

��B� ��FB�B��C�D� �� F��C�D� �BFA�� BEE�DBE$�

'(�EE��)�*�AB��C��+++,��

�

%��EC��CE��B�C�B��E��B��EE��F�C��CF�B��

��D��DB� ��EB�� A�� AF�C�� DB�� %��D�� EC��

�CE�AC�C�� CE� ��B�� A�� EC�� CB��E� �FB�

��C�C�� EC�C��F� �� C�� FB��FFC�D� C�B��C�C�A�B�

DF��E�� -�BF� ��B� �B�FE�� AB�C��BFC�D� ��

��B��FB� ��E� ABB�� EB�� BE�FCAB� �� EEC��

��B� ��FC��E� ��EC�� E��F��BE�� .BD�F�C�D� ��BEB�

��EEC�C��C��E�� /BF�BE0� '+�+,� E��BE� ��

FBEB�F��BFE� �BE�FCAB� ��B� E��B� ��B��B�� F��

�C��BFB��CB�E�A�EB�� BCF� E��CBE�� C�� B��

��C�B�� F��BF� �� F��C�� B�� BF�� %�

��ABF� �� B��E� ��B� ABB�� EB�� EEC��

�� B1��C�2� �E��E��345D��C��6� �7F��BEE�F� �� D��DB�

7��D��%889:��E�FB��E�� 5��

��D��DB� �B�C�C�E� �� D��DB� C��CFB�� DF��E��

;��D��EE� �� /�� '+��,� ��C�B�� B� ��<�F�

��F��BFCE�C�E��C��BFB��BE��EC��

�

=B�C�C�C�� F�A�B�E� �FCEB� AB��EB� ��B� �BF��

>��EC�?� ��E� ABB�� EB�� BE�FCAB� �� FCB��

C��CF�B��E� C�� D��DB�� C�� FB� ��EB�� A�� AF�C��

��DB��F��FB�E�E�B��B��AB��EB��A��AF�C��DB��

9�� D��DB� C��CF�B�� B�BF�� CE� ��B� �FC��F��

��F��BFCE�C�� EC�� C�B� C�� C� ��B�B��

�CE�F�BFE� �C B� �E��ECE� �F� �B�B��C�� C�� CD�� AB� ��

EB��F�� F��BFCE�C�� 8�� EC��D�� B� �BF��

>�C�D��ECE?��B��EEBE��BFE�FB��C�D��E�B��E��

��D��DB� C��CF�B�� BF��C��E� C�� AF�C�� C��

�� B� �B�BF�C��C�� B� EC�BE� �� B� �BEC��

�F��C�D��B�C��CF�B��

�

%EEBEE�B�� CE� �B�C�B�� E� �� FD��C0B�� D��1

�CFB��B�� B��C�� B� ��FCB�� D�C�C�B��

�C�D�CE�C��F�D��C��B��E��D��DB��9��

��EEBEE�B��CE��FFCB��B�BF�C�B�B��CB��@E�

��D��DB� E�FB�D��E� �� B� �BEEBE� �� B� �BDFBB� ��

��C�� D��DB� �B� �BEEBE� �� AB� ��C�CB�� '��B��

+446�*��B�� ++A,�� 8�B�� C�� B��FBE� &��B� ��FB�

�� D��DB� C��CF�B�� C��C��BE�� E�B��E� ��

��D��DB� �BF��F��B� �FB� ��E�� F��FC��B� ��F�

�FB��B��&�'(��D��/��B��E��)�9��++3,��

�

%EEBEE�B�� EC�� C��BE� ��FBB�

C��BFFB��B�� B��E2� �� B��C�� BECE�

��F��C��BECE��BE�C�D�'*��B��++A,��!�B�

��D��DB� �EEBEE�B�� CE� �CD�� E�F��FB�� AEBF��C��

A�EB�� B� �EB� �� AB�EC�B� �� E�FBB�C�D�

�EEBEE�B�� E�� FB�B�EC�B� ��EC�� A��BF��

��C�F��BE�E��E�B�C�C��D��DB��C��E��

�

:

8C��B�C��9��9��*44%3$4��#B��3D��):#��);��<E��

��A�

�

��DBE�BF�� %��EC�� (��BF�� 'D%(6� /BF�BE0��

+4�,��E��EBB E��EECD��CB��E��EEC��EC��

E��F��B� ��BE� �� C�E� �FC��FC�� EB�� C�� B�

�C�D��ECE�� 8�� F��C�BE� B��C�C�� B�CEC�� FC�BFC�� F�

�EEBEEC�D� �� F�C��F� ��EEC�C��C�� FB��BF�� C��

��C�E� �� B� C��C�C�� CB��E� �� AB� ��B��

��B#�C�� C�� B� �� B� BCD�� A�EC�� BE�

��F�C�D� �� B� E��B�� E��FB� �A��C�B�� B�

�C�D��E�C��E�A�BE�E��

�

:��B�BF��B�D%(�CE��EB�FB��C�B��(�E��

=C�D��E�C��%��EC��B��C��C��'(=%B6�;��D��EE�)�

/�� +��,� �� C�� F��C�BE� ��B� �C�D��E�C�� D��

��EEC��C�D��EC�� E�A��BE� �� F��C�D� ��B� EB�BFC��

��EC�� C��CF�B�� !�CE� �BE�� CE� �BECD�B�� F� A��

��C�C�� FBEB�F�� EB� �� C�� FCEBE� �� F�

��D��DB� �� F� �BF��F��B� ��C�E�� 9��F��B�

��EEC�C��C�� CE� �B�BF�C�B�� A�� B� ��BF��

�BF��F��BE� �� B� ��F� �F��C��D��DB1��C��

E�A�BE�E�� C�� EEBEE� E��B��E� E�BB��

��FB�B�EC��FB�B�C�C��C�D��

!�B�D%(�CE��BECD�B�� EEBEE��C�C��E�B��E�

�� D��DB� ��C��E� C�� EC�� CB��E� ��

�F��C�B� ��B� �� BB�B�� BE��A�CE�� F�D��ECE� ��F�

��BF�� !�B� �F��B��FB� CE� A�EB�� B� �FC��C��B� ��

��BF�� B�F��C�D�CE�C�E� �� B� �B�F�1��C��

��B�� !�B� D%(� ��FCEBE� BCD�� E�A�BE�E� �� BCF�

E��FBE��FB�DC�B��C��!�A�B��

!�A�B��)��D�B��A�C��B1C��

9��E�� =��C�E� ��

E��FBE�

� 9��B��E�E�BB�� 3�

�� %��C��F��BFA��FB�B�EC�� 33�

�� .B�B�C�C�� 33�

A� E��C�D� 33�

F� .B��C�D� 33�

G� DFC�C�D� 33�

�� %�F��C�� G3�

4� *��E�F��C��HCE��E��C��

��C��E E�

33�

�

!�B�%��EC��I��CB�� '%I,��C��FCEBE�

��B� 9��B��E� 9�BB�� '9,�� %��C��F�� HBFA��

*��FB�B�EC�� '*,�� .B�B�C�C�� '.,� �� E��C�D� 'E,�

�EBE� ��B� �F�� F�C�� B� ��D��DB� �EEBEE�B��6� ��B�

*�F�C��I��CB�� '*I,� C��BE� ��B��BFA��9��FBE��

.B��C�D��FC�C�D�� F��C�� E�F��C�� E E� �FB�

��B��E�7BF��F��B�I��CB��'7I,��

��

��*I�J%I�K�7I�

�

%I� L� +��4� C��C��BE� %��EC�� C�� CE� �EB�� C��

FBEB�F�� E��CBE� '/BF�BE0�� +�+,�� 8�� F��E�� %I� CE�

��EC�BFB�� E� +4�A� '�F,� ++�G� '�B�� %I,�� (�EB��

��BEB� ��F� ��F��B�BFE� �C0�� 9��B��E� E�BB��

��FB�B�EC�� FB�B�C�C�� C�D�� B� ��BE� ��

��EC�� FB� FB��D�C0B�� !�B�� AB� ��EEC�CB�� BF�

(F��@E�� DBF�C� B@E�� !F��E��F�C�� 9B�E�F�� '!*9,��

!F��E��F�C�� F� '!*�,�� *��C�� %��C��

8E��C��;��A��EC��

�

%��F�C�D� �� 9�B�� /BF�BE0� '+43,�� "��B�

��EC�� #��CB�� '%I,� CE� �� C�� B�E�FB� ��

EB�BFC�� B� E�� B�� D��DB� �B�C�C�� C�� EC��$�

B��C��C�C��E�A�BE��FCA��BE��C��BFB��BF�B��DB�

��C�� B�%I��8��F��C��B��6� ��B��

�� FB�B�C�C�� B�� FCA��B� �3M�� -A<B�� C�D�

��FCA��BE��M��9B#�B��C��E��FCA��B�4M��

NBE1E��I�BE�C��E��%��C��F��F��FB��D�C�C��B��

��FCA��BE�GM��D�F�1��B��FCA��BE�AM��OC��

9B��B��B� ��B�C�� .BE��EC�B� E�BB�� B��

��FCA��B��M��!�BEB��BF�B��DBE��B��E�F��B��B�

D%(� ��EC�� #��CB�� CE� �BCD��B�� B��C�� F�E�

B��FBEEC�B� ��E E� '43M�� B�%I,��(B��EB� ��B�%I� CE�

�BCD��B�� B��C�� A�� E��FBE� �F�� B��FBEEC�B� ��E E�� C��

�CD�� FB��C�� FB�FBEB�� CB��@E� B��FBEEC�B�

��D��DB� �AC�C�� !�CE� �BCD��DB� #�BE�C��E� �A�� B�

FB��C�B��FCA��C��E��B��FC��E�B��FBEEC�B��E E��

��B� �FB�C��C�� B� %I�� ;C�B�� C��F��C��

��B�� B�� FB�B�C�C�� E��FBE� ��FCA��B��E��

�� B� %I�� B�� CD�� AB� B��B��B�� AB� ��B� ABE��

�FB�C��FE��B�E�FB��A��B�%I��!�B�E�A�E��FBE��F�

C�B�E� �� B� �BE�E� �FB� '�,� E��B��E� E�BB�� 'A,�

��FB�B�EC�� '�,� FB�B�C�C�� '�,��C�D��

��EEC�C��C��B��CB��F�C�D��B��C��

�FC��C��B�C��B��4�E�A��BE��EC��

!�CE� ��EEC�C��C�� CE� ��EC�BFB�� E� ��C�C��

��C�� A�EB�C�B� ��F� FBEB�F�� C�D��ECE� �� F�D��ECE��

/BF�BE0� '+�+,� E��B�� B� E��FB� ��F� C��F��C��

��B�� E� ��B� �CD�BE�� FFB��C�� C�� B� %I6�

��B�BF�� B� �FBEB��B�� E�AE��C��B� ��CE�

��C��!��E��D��B�%I�CE��FBEB��B��E��C��B��

�� B� EB�BFC��EC�� C��CF�B�� B� FB��C��E�C��

AB��BB��C��B��B��C��C�C��E�A�BE�E��B�D%(�

��B��ABB��C��BE�CD��B��

D�C�B� ��E�F��C�D� �� E��F�C0B�� BE�� B�

E�� E�� B� C�� ',� ��B� �FB#�B��

��FFB��B� �� F�E� '�,� �� C�E� FB��C�B� B�EB� �F�

�C��C�� F��C�D� ��F�E� '�,� ��B� DF��C��

��F��C��C��B��FB��EC�C��CE�E��B��C��F�C��F�

��D��DB��FB��BF�� B� E�� DC�B� ��B��C�� B�

��DB�BC��B�E�C��C��BE��C�B�E��!�BFB�E��

AB��F��DB��C��C��C��B��BE�C�D�C�B�E��!�BFB��FB��

��B� FB�CB�� C�BF��FB� �� F�A�B�E� C�� B�B��C�D� ��

�BE�� C��C��BE� �� F�� F�� C��BFB��F�A�B�E��C��


�

��F��

�FB� �B��C�� C�� FB�� B� �� E�� B��BF�

�F�A�B�E� �� EC�D� ��F��FC��B� ��B�� F��B�

��F ��

!�A�B��)��C��D�

� B��FBEEC�B�9��E��

!��BE��

%��EC��

O��B�� *��FB�B�1�

EC��

.B�B�C�C�� E��C�D�

� ;��A�� 31A� 31��+� 31�A�+� 31G�

�� (F��?E� 31A� A13� 31��+� 314�

�� 8E��C�� 31A� 31��+� F13� 31G�

A� !F��E��F�C��

��F�

31A� A13� 413� 314�

� .B�B��C�B�

F� DBF�C� B?E� F13� 31�G�+� 31��+� 31+�

G� !F��E��F�C��

EB�E�F��

F13� 31�G�+� 413� 31+�

�� *��C�� F13� �13� 31�G�+� 31+�

4� %��C�� F13� �13� �13� 31+�

�

!�B�D%(� CE� �� FB�B�EC�B� �EEBEE�B��

��F� �EEBEEC�D� ��D��DB� ��C��C�D� C�� EC�E�� !�B�

�FC��F�� D��E� �� B� D%(� �FB� �� EEC�� FC��E�

��EC�� E��F��BE� �� B��B� ��B� EB�BFC�� B�

��EC�� C��CF�B�� !�B� B��C��C�� CE� �BECD�B�� F�

A�� C�C�� FBEB�F�� EB� �� C�� CE� ��EB��

��F� ��D��DB� E�A�BE�E� �� FBB� �BF��F��B� �BE�E��

*��EEC�C��C�� E��F��BE� CE� �B�BF�C�B�� A�� B�

��BF�� BF��F��B� �� B� ��F� ��D��DB� E�A�BE�E6�

��B� E��B�� BF�� BF��F��B� �� BEB� E�A�BE�E�

�CB��E��F��C�D��BF��F��B�EB�BFC��'%I,��

!�B� E��F�C0��C�� C��F��C�� CE� �F��C�B�� A��

/BF�BE0��7��B�'+�A,��E��B��A��/BF�BE0�

'+�+,�� !�BEB� FB�BFB��BE� C��B� �FC�BFC�� F�

��EEC�C��C�� EC�� E��F��BE� A�EB�� B�

��D��DB� E�A�BE�� BF��F��B� �� E��B� �� F3�

��EC��CB��E��C��FC��E�B�C��DCBE��.B�C�AC�C��

��C�C�� FB� ��E�� F��C�B�� !�B� D%(� ��C�BE�E�

�CD�� C��BF1F��BF� FB�C�AC�C�� D�� C��BF�� ECE�B��

�� CD�� BE�1FB�BE�� FB�C�AC�C�� 9��BEE�� FC�BFC��

��C�C�� E� ��E�� ABB�� BE�FCAB�� %��EC�E� �BFB�

�C��BFB��C��B�� F�� AF�C�� DB�� E� �� B�

D%(� �BF��F��B�� !�B� �EBE� �� %I� �CE�C�D�CE��

��EC�E��F��AF�C��DB��F��E��/BF�BE0�

E��BE� �� D��DB� E�A�BE�E� �� AB� ��C�CE�BFB�� C��

��F��C��B�� B� ��F� A�� B� �� D%(� FB#�CFB� ��

�B�E��BE��EBEEC��E��B�B��

8�� E��F�� B� D%(� CE� �� FB�B�EC�B�

�C�D��E�C�� 8�� CE� #�C�B� �EB�� C�� C�D��E�C��

��EEC�C��C�� EC�� C�� FC��E� E��F��BE�� !�B�

�BE�� B�FE� �� BF��B� �C�� B� �EE��C��

��EBE� �� B�F�� AB� ��EEC�CB�� E� ��B� �� B� 4� A�EC��

��BE�� 9�� B�F� �� EEC�C��C�� CE� �� FEB��

C��C�C�� C�� E�BF� ��C�� FBEB�F�� A�� E� �C��B�

�B��C�D� ��F� ��B� "�C�B�$� ��EC�E� �� F� ��

��FB� ��B�� C�� B� ��C�C�� F��C�B� �� B�

��EEC�C��C��E�E�B��E�DDBE�E��!�B��EB��BEE��D%(�

�E��FBEB�F��CE��B��BFB��A��/BF�BE0�'+�+,��

%��D�� EC�BF�A�B� ��ABF� �� BE�E� ��F�

��EC�� CE� ��C��A�B� C�� A�� B�D�CE�� BF� 8��C��

��D��DBE�� BFB� �FB� �� EC�� BE�E� ��C��A�B� C��

!B��D�� B� �� B� ��F� ��<�F� =F��C�C�� D��DBE�

�C�B�� E�� B�� C�� %��F�� 7F��BE�� E��BF�� F��

8��C��!�BFB��FB� ��B��FBEB�� E��B��B�� BECD��

�� EC�� BE�� C�� !B��D�� A�EB�� B� �FC��C��BE� ��

D%(�'/BF�BE0��+�+,��

9�� BE�� B�� C�� C�B��C��C�D� ��B�

��EC�� BE�FCAC�D� ��B� ��EC�� EEC��C�D� C�� C��

��FC��E�E�ADF��E��F��B��F��EB��C�D��ECE��BF��

�� F�D��ECE�� 9�� BE�� AB� �BF�� EB�� F�

E�BB�� D��DB� ��BF��CE�E� �B��C�D� �C�� !B��D��

��EC��C��EC��B��B��B��FC��F��BF��

E�BB��1��D��DB��DCE�E� CE� ��EEBEE�� C��F��B�

��B� ��C��C�� E C��E� C�� C��C�C��E� �C�� EC��

!�B� �C�� B� �FBEB�� E�� E� �� B�B��

E��F�C0B�DBE�BF��%��EC��(��BF��C��!B��D��'D%(1

!,�A�EB��B��C��DBE�BF��%��EC��(��BF��

'/BF�BE0�� +�+,� �� EEBEE� ��B� ��D��DB� �AC�C�CBE� ��

��EC��E�A<B��E��

��&A

��$'�%D(�!��DF��E��E�A<B��E��BFB��EC�BFB��F�

��B� �FBEB�� E�� C�� C�� B� DF�� ECE�B��

��F��E�C��C��BFB��DB�DF��E��B��BF��B�

B��BFC�B��DF��ECE�B��EC��E�A<B��E�A

"��E��A �E��(� 33� ��F�� !B��D�� E�B� BFE� C��

��B� �DB� F��DB� �� 3� P� �3� �B�FE� �BFB� �� B�� E� ��F��

DF�� !�BEB� 33� E�A<B��E� �BFB� ��BD�FC0B�� C�� C�B�

�DB�DF��E��C B��31�3��1A3��A1F3��F1G3��G1�3�

�B�FE��C��3� E�A<B��E� C�� B��DF��%�� B� E�A<B��E�

�BFB� ��B�� F� ��BCF� �DB�� EB�� D��DB� �B�B��

��B��BEE��B��C��B�B��%��B��F�C�C��E�C��

��B� E�� BFB� ��C�B� E�B� BFE� �� !B��D�� C��

�CE��F��B�F��DC��F��E��C��FC�� C��BEE��F��

��CE��F��F�D��A�EB��

�

)*��E��A �E��(� %� �� 3� �� EC��

��CB��E� C�� B��DB�F��DB��A3�P��3��B�FE��F�C�C��B��

C�� B� E�� %�� B� C��C�C��E� �C�� EC��

�CE��F�� BFBAF��E��F� ��C�B��E� '*H%,�� !�B� �DB�

F��DB�� B��F�C�C��E� F��DB�� F��4�P��3��FE��

��B� B��C�� E��F�E� �� BEB� C��C�C��E� F��DB��

�F�� 3�� E��F�� F��B� �B�B�� %�� B� ��EC��

E�A<B��E� �BFB� FCD�� B�� !�B� ��F�C�C��E� �BFB�

�C�D��EB�� A�EB�� B� �B�F��DC�� C��C�DE� ��C��

�BFB��A��C�B��F��B�F��DCE��

A

8C��B�C��9��9��*44%3$4��#B��3D��):#��);��<E��

��G�

�

�E�%�&�E�(� !�B� �B�B��B�� B� �BE�� E� ��B� C��

��EBE�� 8�� B� �CFE��EB��B�B��B�� B� �BE��

��BFC��C��!B��D��E��B��!�B��BE��A��BF��C��B��

��B� ��D��DB� ��F��B�BFE� ��B��2� �BE�FC��C��

E��B��E� E�BB�� F� ��BFE��C�� E�BB��

�B�E�FB� �� C��F��C�� B�� B�E�FB� �� B��

��C��F��FB�B�EC�� FB�B�C�C��C�D�� 8�� B�

EB�� EB�� B� �BE�� A��BF�� 'D%(1!,� ��E�

��C�CE�BFB�� F�� B�� B� �BFE��E� �C��

��EC��A

AAAAAAAAAAAAAAA

+�D��DA��&A��D%�DD��A

!�B��E��0B��EC�D�9799�G�3��BFEC��

��BE�C��B��B��B��E��FBE�'�,��E��F��B�C��C��

'9=,��F�A��DF��E�A<B��E��A

A

��$�D(A��DA��%�A

!�CE� E�A�BE�� E� �BECD�B�� B�C�C��

��BFE��C�� E�BB�� F�� B� ��CB�� C�� FB��

#�BE�C��E� �E B�� C�� B� ��B�� C��BF�CB��

�C��FB� �BE�FC��C�� !�B� �� C��F�� E�B��E� ��

E��B��E� E�BB�� B��C�B�� BFB� ��B� C��F��C��

��B�� B�� !�B� ��C�� E��FB� ��F� B��

�E�B�� E� 3�� !�B� �B�� 9��F�� B�C��C��

��F��DF��E��FB�DC�B�� C�� B�!�A�B��B��

9=��EC��E�A<B��E��FB�DC�B��C��!�A�B�A��A

�

!�A�B��<�B��<(�B��)�B��B��CB�C��)8(��

��B��6��C��B�C��B��E�

�

!�B� �B�� 9��F�� B�C��C�� E��FBE� ��

��F�� E�A<B��E� C�� B� �� DB� F��DBE� ��F� C��F��C��

��B�� BFB� 3�33� �� 3�33�� 8�� B�� E � �� B�

C��C�C��E� ��F�EE� �� B� �DB� DF��E� �BF��F�B��

EC�C��F�� !�BFB� ��E� �� ECD�C�C�� B��C�B� �� E�

�AEBF�B��F�EE��B��DB�DF��E��

-��B��E ��C��F��C��B��B��C��

��EC�E� �� CD�BF� �B�� '+�GG,� �C�� 9�=� ��

'3�F��,��B��A��F��E��F�C��EB�E�F��E�A<B��E��C��

�� B�� '��F3,� �� 9=� '3��3�,�� !�B� C��C�C��E� �C��

��C�� EC�� B�� E��FB� �� 'G�33,�

�(F��?E��DBF�C� B?E��F��E��F�C��F��EC�E��

D��A�� B� �BF��F�B�� F�� FB�� C�E��

�F��E��F�C�� EB�E�F�� C�� EC�E� C�� CE�

��E �� !�BEB� FBE��E� E��F�� B� �C��C�DE� �� (��

9�C��F�� =B�CE�� /F��BF� �� /�� '�33,� ��

FB��F�B�� B� ��CB��E� �C�� B�� F�� BEC��E�

�BF��F��F��B��CB��E��C��FCD��F��BEC��E�

��B��BFA��B��E ��

FB��'��<�B��B��)�B��B��CB�C��)8(��BAB�C��

��6��C��B�C��B��E�

Q(%1� (F��?E� %��EC�6� D%1� DBF�C� B?E� %��EC�6� %%1�

%��C��%��EC��!*�1�!F��E��F�C��F�%��EC�6�!*91�

!F��E��F�C�� 9B�E�F�� %��EC�6� *%1� *��C�� %��EC��

;%1�;��A��%��EC��

*F�F�� .��C� '+4+,� FB��F�B��

C��F��C�� B�� E� ��B� ABE�� FB�C��F� �� B�

EB�BFC��B��EC��C��CF�B��E��B�E�FB��A��B�

%I�� !�B� C��F��C�� B�� E��FB� FB��B��E� EB�BF��

�C�B�EC��E� �� CB��?E� ��C��C�B� �AC�C�CBE� ��

��FCA��BE�� CD��BF�B��DB� �� B��C�� B�

%��EC��I��CB��!C�B��E��EB��C��B��B��

��B�FB��C��E�C�E��D��B�E�A�BE�E��F�AB��BB��B�3�

E�A�BE�E� �� %I�� /BF�BE0� '+�+,� E�DDBE�B�� B�

C��F��C�� B�� E��FB� FB�FBEB��E� �� B�E�FB� ��

��C�� C��C��B��E�� CB�� E��

��EEBEE� E��B� �BDFBB� �� A�� FB�B�EC��

B��FBEEC�� AC�C�CBE� �� FBE�� F��FC��B�� C�� B�

��E ��

�

��$�D(AF�&��E�A��E$��A%��E��D��A

!�B� � ��CB�� BF��F��B� CE� ��B�� C��B��

A�� C��C��CBE� �� BFA�� B��FBEEC�� F��C��

C��B��B�� C��E��*��FB�B�EC�� E ��B��E� ��

��BF��FC��E��E�B��E��CE��B��FB��A��EC�D�'�,�NBE1

E��#�BE�C��E�'A,��C��C�D��E ��C��F��FB��D�C�C��

�� '�,� �� EBFCBE� �� EB#�B��C�� E�� !�B�

��C��E��FB��F�B��E�A�BE��CE�G3��G3��43��!�B�

�B��9=��F��DF��E��FB�DC�B��C��B�!�A�B�F�

�� B�� 9=� �� EC�� E�A<B��E� �FB� DC�B�� C��

!�A�B�G��

�

!�B� ��F�� E�A<B��E� E��FB�� B� �B�� DB�

'GG�GG,� �� E��F�� B�C��C�� '+�F3,�� E� ��B� �B��

�DB�� B�E�A<B��E��F�EE�� B��DB�DF��E��1�3�

�B�FE��

A

;F��

'NFE,�E�

8��F��C��

��B��O��B�� !��

�� 9=� �� 9=� �� 9=�

�31�3� �3� 3� 3�33� 3� �33� 3� �33�

�1A3� �3� 3� 3�33� 3� �33� 3� �33�

A1F3� �3� 3� 3�33� 3� �33� 3� �33�

F1G3� �3� 3� 3�33� 3� �33� 3� �33�

G1�3� �3� 3� 3�33� 3� �33� 3� �33�

;F��Q� E�8��F��C��B�� O��B�� !��

�� 9=� �� 9=� �� 9=�

(%� F� F��3� ��G� ��3� �3+� A��3� �+��

D%� F� F�43� �4�G� F�33� �33� F�A3� �+GG�

%%� �� +�GG� �F�� 33� �� 33� �F�

!*�� A�33� �A� A�33� �A� 4�� 4G�

!*9� �� F3� ��3�� F�F3� ��3�� A�33� �F�

*%� � G�33� 1� ��33� 1� G�F3� 1�

;%� �� 33� �33� �33� �33� A�F3� ��


�

��

FB��/��<�B��B��)�B��B��CB�C��)8(��B��C��E�

��B��A��C��

�

!�A�B�G��<�B��B��)�B��B��CB�C��)8(��BAB�C��

��6��B��C��E��B��A��C��

�

�

;F�

��Q�

�

�

E�

NBE�CE��

%��C��F��F��

FB��D�C�C��

�

9B#�B��C��

��E�

�

!�� 9=�

�� 9=� �� 9=� �� 9=�

(%� F� A��43�FA A��3� ��+A� �4�33� �3�+�A�GG4�43

D%� F� �3��3 G�AF� �F�G3� ��3� ��3� ��A��+�GG ��G��

%%� �� G3�33 �333� F4�GG� �F� ��33� 4�4�+�GG��F4

!*

��

�� F��33 A��A� F��33� ��3�� G3�33� 4�A4�G��44 4�33�

!*9� �� AG�F3��

A4�33� 4�A4� 4�33� ��4��FG�GG G�F�

*%� � AF�33 1111� A��33� 11� AG�33� 11� ��F3F�GG

;%� �� 33 A��A� ��33� �33� �333� �333�AA��34

Q(%1� (F��?E� %��EC�6� D%1� DBF�C� B?E� %��EC�6� %%1�

%��C��%��EC��!*�1�!F��E��F�C��F�%��EC�6�!*91�

!F��E��F�C�� 9B�E�F�� %��EC�6� *%1� *��C�� %��EC��

;%1�;��A��%��EC��

!�B� C��C�C��E� �C�� EC�� BF��F�B�� F��

��B�� FB�� B� ��F�� E�A<B��E�� !�B� NBE� CE��

#�BE�C��E� �FB� ��EC�BFB�� AB� ��B� EC��BE�� E � ��

�EEBEEBE� ��C��F�� BFA�� FB�B�EC�� !�B� ��C�E�

�� F��E��F�C�� F� ��EC�E� �A��C�B�� B� �CD�BE��

E��FB� C��B�� 'G3�33,� ��9=� '�333,6� 'F��33,� ��9=�

'A��A,6� ��B�� A�� F��E��F�C�� EB�E�F�� EC�E�

E��FB�� B�� 'AG�F3,� �� 9=� '��,�� B��

��C��EC�E��B��E��FB��'AF�33,��

(F��?E��EC�E� ��B��C��C��B��'A��43,��C��

9=�� '�FA,�� 8�� CE� B�C�B�� B� E�A<B��E��

AB��BF� ��FB�B�EC�� NBE� CE�� #�BE�C��E� ��B��

��FB�� C��F��F�� FB��D�C�C�� EB#�B��C��

��E�� =�FC�D� ��C�CE�F��C�� B� �BE�� E��

��B�� FB#�CFB�� B� FB�B�C�C�� #�BE�C��E� �� BFB�

�E B��A��B��C�C�C��

8��C�C��E� �C�� DBF�C� B?E� ��EC��

DFB��BF��C��C��C��FB�B��C�D��B�#�BE�C��E��!�B�

��C�C�C��EC��C�CB�� B�FB��B��#�BE�C��E��E� ��B�C�C��

��B� FBE��EB� �F�� B� ��CB��E�� E� ��BEB� E�A<B��E� ��

��F��B��E��FB�'�3��3,��C��9=��'G�AF,��!�B�D��A��

��EC�E�� BF��F�B�� BF�� F� C�� BF�E� ��

��FB�B�EC��!�B��E��FB��B��'��33,��8��E�C�B�

�� FB�B�C�C��E��F� B�B��B��F��C�B��C�� BE� ��BEB�

E�A<B��E��BF��F��BF��F��B��

8�� B� E�A� EB��C�� C��F�� BFA��

��FB�B�EC�� BEB� E�A<B��E� �BFB� �E B�� C��

��B� A�� F�E�� FB�� A<B��E�� F�� A<B��E�� FE��

��F�E��B��BFE��ABFE��F�C��FB��FB��E B��A��

��B� ��C�C�C��:BFB� ��B� C��C�C��E� �C�� C��

�B�� 'F4�GG,� �� 9=� �� '�F,�� BEB� C��C�C��

�BF��F��BE� �BFB� �B�F� �� B� E��FBE� �� F��

E�A<B��E� �F��E��F�C�� EC�� B� �� B�� E��FB�

'F��33,��B��FB��C��E�A<B��E��

8��C�C��E� �C�� F��E��F�C�� EB�E�F�� EC��

��B��'A4�33,��B��B��E��FB��(F��?E��EC�E�

'A��3,�D�B�� B� FBE��EBE��BFB� ��EC�BFB�� C�� BF�E�

�� BD�FCBE� ��B� �BF��F��B� �� BEB� ��EC�� DF��

�BFB� AB��BF� ��F� �A<B��E�� F�� A<B��E�� FE�� A��

��F�E�� B��BFE� �� F�C��FB� ��FB�� C�DBFE��

��ABFE� �� B� DB��B�FC�� E��BE�� B��1FCD��

�CFB��C��E�� *��EC�BFC�D� ��B� DBF�C� B?E� ��EC�E�

�A��C�B�� B�� '�F�G3,� �� 9=� �� '��3,�� BEB�

C��C�C��E� �BFB� �� A�B� �� FB��D�C0B� ��B� �B��BFE��

��ABFE�� C�DBFE�� DB��B�FC�� E��BE� �� B��1FCD��

�CFB��C��E�� -�� B� ��BF� �� B�� BFB� �A�B� ��

FB��D�C0B� ��B� FB�� A<B��E�� A�� F�E� �� F�C��FB�

��FFB�� *��EC�BFC�D� ��B� FBE��EBE�� BFB� �BFB� ��

��B�F� �� F��B�� !�B�� FB#�CFB�� FB�B��C�D� ��B�

E�C��E��E��FBEB��B��B��

;��A�� EC�E� �A��C�B�� B�� '��33,� ��B�

�BF��F�B��F��B��FB��BF��EC�E��

!�B�� BFB� �A�B� �� FB��D�C0B� �B�� FB�� A<B��E�� !�B��

��C�B�� C�B��C�� B� �B��BFE�� ABFE�� C�DBFE��

DB��B�FC��E��BE��B��1FCD��CFB��C��B�B��B��

��B��FBEB��C��B�E�C��C��E�FB�B��B��E��B��

��B��C�B��FB��D�C0B��BF��BD�FCBE�C��B��

C��EB��C��

!�B� EB��C�� C�� C��BE� EB#�B��C��

��E� CE� ��EC�BFB�� AB� ��B� ��B�� E E�

�EEBEEC�D� ��B� ��C��F�� BFA�� FB�B�EC�� :BFB� C��

B�C�B�� B� ��C�� EC�E� �� B� �CD��B�� B�

�� '��33,� ��B�� A�� F��E��F�C�� F� ��EC�E�

�A��C�B��B��'G3�33,��!�BEB��BE��EC��

E�A<B��E� �� CD�� E��FB��B�� FB�� BF�

��B��EC�E��8��E��C�B��BEB�E�A<B��E��B�

�B��E�F��B�� AB��BF� �BF��F��B� ��B�� B�

��E� �BFB� FB�B��B�� B�� B� F��B� ��

;F��

'NFE,�

E� NBE�CE�� %��C��F��

��F��

FB��D�C�C��

9B#�B��C��

��E�

!��

�� 9=�

�� 9=� �� 9=� �� 9=�

�31�3� �3� G3� �33� G3� �33� 43� �33� GG�G� +�F3�

�1A3� �3� G3� �33� G3� �33� 43� �33� GG�G� +�F3�

A1F3� �3� G3� �33� G3� �33� 43� �33� GG�G� +�F3�

F1G3� �3� G3� �33� G3� �33� 43� �33� GG�G� +�F3�

G1�3� �3� G3� �33� G3� �33� 43� �33� GG�G� +�F3�

8C��B�C��9��9��*44%3$4��#B��3D��):#��);��<E��

��4�

�

�FBEB��C�� B� E�C��C� ��E� E��B��

*��EC�BFC�D� ��B� FBE��EBE� �� BFB� B�C�C�B�� A�� BEB�

C��C�C��E��BFB��F��B��

!�B� ��C�� EC�� E�A<B��E� �A��C�B��

�B��'AG�33,��AB��EB��B�C��C�C��E�� B��C��

��CE� ��B� �� EC�� (F��?E� ��EC�� E�A<B��E� ��

�B�� '�4�33,�� *��EC�BFC�D� ��BCF� FBE��EB�� BEB�

E�A<B��E��BFB��A�B��B��B�E�B��E�B��

��E��C�BE��B��B�B��FB#�CFB��FB�B�C�C��B�

��E��B��B�E�A<B��E��EB��FFB��C��

�BFB��C�B��FC�D��B��BE�C�D��

!�B��B��B��DBF�C� BE?E��EC�E�'��3,�

��9=��E� '��A�,��!�BFB� ��FB�B�EC�� B� E�B��

��E� �BFB� D�� C�� E� B�C�B�� F�� BCF�

FBE��EBE��!�B��C�B��BF��F��E�B��E��

*��EC�BFC�D� ��B� �F��E��F�C�� EB�E�F�� EC�E� ��B��

�A��C�B�� B�� E��FB� �� '4�33,�� BFB� A�� BEB�

E�A<B��E� �BFB� �A�B� �� FB�B�� B� E�B��

��E�� !�BFB� ��E� �� FB��B�� BF��F��B� �E� ��B�

��B�C�� E�C��C� ��E� C��FB�EC�D�� B�B�� D��

��B�� B� FB�B�C�C��E� �BFB� DC�B�� B�B�� B�

�FBEB��C�� F��B� ��E� E��B�� B�� C�B��

��FB�B�� B� DC�B�� E�C��C�� %��F�C�D� �� :BC��

��9��BE�'+�G,�E��CB��FB��FB�B�EC��

BFF�FE�C��(F��@E��*��C��DBF�C� B?E��EC�E��

!�B��1EC�� EC�� CB��E� '�C�B� (F��@E�� BCD��

��C�� C�B�DBF�C� B@E,� �� BCD�� F��E� �BFB�

DC�B�� BE�� C�� B��B�� C��BFB��C��B�

��FB�B�EC�� BFF�FE� ��EB�� A�� E��C��

C��B�B��B� �F�� EB� ��EB�� A�� B�C��

C��B�B��B��!�B�� FB��F�B�� DBF�C� B@E� ��EC�E�

��B�ECD�C�C��FB��B�C��BFF�FE��B��BF�

DF��E� �� ECD�C�C�� C��BFB�� AB��BB�� B�

�B�C��BFF�FE��B�A��(F��@E��C��F��

DF��E�� !�B�� E�� ECD�C�C�� C��BFB��BE�

AB��BB��(F��@E��C��EC�E��B�BF�A��

��BEB� DF��E� ��B� ��FB� E��C�� BFF�FE� �� B�

��F��E��

��$�D(A+��AA

.B�B�C�C�� CE� �BE�B��A��CD��FB#�B��F�E��

C��FB�EC�D��B�D��EC�B��F�E��ABFE��ABF1

��F��AC��C��E��CD��F�A�AC�C��EB��B��BE�

�� EB��B��BE� �� C��FB�EC�D� �B�D�� DF��C��

��B�C�� 8�� C��BE� �BE�E� �� F�� DC�C�� BE��

EB��B��B��C�E�� B� �B��BFE��BE��EB��B��BE�

��C��ECE�E�E�B�C�C��E��F��DF��C��F�E��

!�B� ��C�� E��FB� ��F� FB�B�C�C�� CE� 33�� !�B� �B��

��9=��F��DF��E��FB�DC�B��C��B�!�A�B��

�B��9=��EC��E�A<B��E��FB�DC�B��C��!�A�B�4��

OF��B��A��B��A�B�� C�� CE�#�C�B�B�C�B�� B�

��F�� C��C�C��E� �F�� B� �DB� F��DB� �3� 1�3� �FE��

FB�B�C�C�� E � ��B� �BF��F�B�� F��B�� %�F�EE� ��B�

�DB�DF��E� ��BFB��E��B��C�B��F� ��B�BFF�FE�EBB��(��

�B��E�A<B��E�� EB��1��FFB��C��E��E� ��B��BFB�B�C�B��

��FC�D��B��BE�C�D��BFC��

!�A�B��<�B��<(�B��)�B��B��CB�C��)8(��

��B��6��-��C�C��B��

;F��'NFE,� E� .B�B�C�C��

�� 9=�

�31�3� �3� �33� �333�

�1A3� �3� �33� �333�

A1F3� �3� �33� �333�

F1G3� �3� �33� �333�

G1�3� �3� �33� �333�

�

!�A�B�4��<�B��<(�B��)�B��B��CB�C��)8(��

BAB�C��6��C�C��B��

Q(%1� (F��?E� %��EC�6� D%1� DBF�C� B?E� %��EC�6� %%1�

%��C��%��EC��!*�1�!F��E��F�C��F�%��EC�6�!*91�

!F��E��F�C�� 9B�E�F�� %��EC�6� *%1� *��C�� %��EC��

;%1�;��A��%��EC��

!�B� ��C�� EC�E� �� A��C�B�� B��

'44��,�� !�BEB� C��C�C��E� �BFB� �A�B� �� FB�B�� B�

��E� ��C�� BFB� �FBEB��B�� B�� !��E� ��E�

��B�� A�� (F��?E� ��EC�E� �� B��

'�+�G,��BFB�A�� BEB� C��C�C��E� FB#�CFB�� FB�B�C�C��

��B��E��B�C��BFF�F��BFB��AEBF�B��

!�B� �B�� E��FB� ��F� *��C�� EC�E� ��E�

'4�33,�� BEB� E�A<B��E� ��E�� FB#�CFB�� FB�B�C�C��E� A��

��FB� �� B�C�� BFF�FE� �BFB� EBB�� C�B� �BE�C�D��

��BFB�E�DBF�C� B?E��EC�E��BF��F�B��BF��F�

C�� B�FB�B�C�C�� E �� B��B��E��FB� '�33,�!�B�E��

��FFCB�� A�� *�C�� C� ��DC��C��E� '++3,�� BFB�

��B�� FB��F�B�� (F��?E� ��EC�E� B��CAC�B�� FB�

��ABF� �� B�C�� BFF�FE�� BFB�E�� C��

��EC�E��DFB��BF��ABF��B�C��B��E��F�

��F�E��F�EBE��DBF�C� B?E��EC�E�FB�B��B��FB�

��ABF��FB��B��F�E�� B�� <�FD��E��BFB��E��

��C�B��C��BCF�E�BB��

*�C�� C� ��DC��C��E� '++3,� C��BE�CD��B�� B�

FB�B�C�C�� B�C�C�E� C�� +F� E�A<B��E� C�� FBB� ��EC��

E��F��BE� '�� C�� 4� (F��@E�� F�

DBF�C� B@E� ��EC�E,�� 9�A<B��E� FB�B��B�� F�EBE� ��

EB��B��BE� �F�� B� .B�B��C�D� 7�F�EBE� 9�A�BE�� B�

;F��Q� E�.B�B�C�C�� !��

�� 9=� �� 9=�

(%� F� �+�G� �+�3�� +�G3� �+�3��

D%� F� � �+�3�� G�G3� �G��

%%� �� 44�� 3�F4� �� A�

!*�� A��3� 44�� 3�F4�

!*9� �� A��A� �� A�3�

*%� � 4� 111� +�� A��A�

;%� �� A��A� 4� 111�


�

��+��

(�E�� =C�D��E�C�� %��EC�� B��C��C�� !�B��

FB��F�B�� C�� EC�E� B��CAC�B�� DFB��BF�

��ABF� �� B�C�� B��E�� F�� FB�CEC��E��

��F�� F�EB� FB�B�C�C��E�� (F��@E� ��EC�E� �EBE�

��FB� ��B�C�� BFF�FE� �� CEEC��E� �� DBF�C� B@E�

��EC�E�E��FB��B��F�E��<�FD��

��$�D(A,��AAA

E��C�D� �� A<B��E� �� CE�� F��C��

��E�C��BE�G+M��B��C�D�E��FB��!�B��

�F��C��A<B��E��FB�B�EC��C��A�B��FB�E��

C��C�C�� !�B� E��B� ��C�E� ��FC��E� ��BD�FCBE��

E��BE��EC0B��!�B��CB�� CFE�� CE��E B�� B� ��B�

�A<B�� CE�� FBEB��C�� D�F�� O��B�� CE� CE�

�3M��B��C�D�E��FB��8��CE��B�E�FB��A��C�D��E�

�� C��E� �E� ��B� ��CB�� C�� B� �C��B��

9B��B��B� ��B�C�� !�CE� 3M� �� B� ��C�D� E��FB��

:BFB� ��B� ��CB�� E� �E B�� B�B� �� B�

B��C�BF�E��E��.BE��EC�B�E�BB��1��CE��E�3M��B�

��C�D� E��FB��:BFB� ��B� ��F�� C��C�D� CE� ��C�C��B�� A��

��B� ��B�� B� �FB�B�C�D� EB��B��B�� !�B� ��C��

E��FB��F�B��E�B��CE�G3��3��3��3��

A

8��C�D��E ��F��E�A<B��E�E��FB��B�B�

E��FBE�C��A<B��C�D��E ��%�F�EE��DB�DF��E��

�� B� E�A<B��E� ��BFB� ��E� �� B��C�B� C�� BF��F��B�

�B�B�E��

8��BF�E��A<B��C�D��B��E��FB��A��C�B��

��F� �DB� F��DB� �31�3� �FE� CE� '4�AF,� �� 9=� '��G�,��

��B��EC�BFC�D��B��DB�DF��E��F��1A��A1F3�FE�

��B�� B�� '��+F,� �� '��A3,� FBE�B��C�B��

:BFB��B�E�A<B��E��BFB��A�B��B��B��C��BE��

A�� B�� B�� FB�B�C�C�� E��B� ��C�� B�

�� EB�� FFB��C��E� �BFB� ��E�� EBB�� !�B� ��BF� DF��

F1G3� �FE�� B�� B�� 'A��3,� �� 9=� 'F�A�,��

�BFB�C��E�#�C�B�B�C�B��BFB��E��B��C�B�C��B�

��F��B�� E ��!�B��BF��DB�DF��G1�3�FE� ��B��

�� B�� '��A3,�� BFB� ��B� �BF��F��B� �� BEB�

E�A<B��E��E��F��B��FB�� BF��DB�DF��

8��CE��E ��BFB��E��B��C�B��BF��F��B��F��DB�

DF��E��31F3��FE��A��BFB��E��F�E�C��FB��B��E��FB�

C��B��DB�F��DBE��F1G3��G1�3�FE�FBE�B��C�B��

!�B� ��BF� EB��C��E� �� C�D� C��B� EB��B��B�

��B�C��E ��FBE��EC�B��C�D��BFB�C��B�

E�A<B��E� �A��C�B�� B�B� E��FB�� 8�� A<B�� C�D�

��E � ��C�� EC�E� ��A��C�B�� E��FB�� 'A+��,��

�F��E��F�C�� EB�E�F�� E��FB� �� 'A��F3,�� B��

A��C��EC�E��B��'��33,��BEB�

C��C�C��E� �BFB� �A�B� �� B� ��B� �� A<B��E��

(F��?E��EC�E��BFB��A�B��B��B��C�B�E�

�� B�� B�� '��33,��DBF�C� B?E� �� D��A��

C��C�C��E��B��BF��F�B��BF��F��A��C�B��

�� B�� E��FB� �� '��A3,� )� '��F3,� FBE�B��C�B�� !�B�

�FBEB�� E�� E��F�E� ��B� �C��C�DE� �� /��

;��D��EE� '+4F,� FB��F�B�� C�C�� F�� P

�F��C��C��C�� C��C��EC��FB�EBB��B��

��FB��BF��B��EC�E��

/�� ;��D��EE� '+4F,� B��C�B�� B�

�CE�FCA��C��BFF�F��BE�C��C��FB��C�D��+�(F��@E�

��EC�E��+�DBF�C� B@E��EC�E��F��C�E��

+� ��E�BFC�F� ��C�E�� !�B� FB��C�B� �CE�FCA��C��E� �� B�

��FBB��E��F��C�B��C�D�BFF�FE�P��B�C��BFF�FE��

EB��C��BFF�FE��C��F��CF��C��E��B��B��

��CE�C�D�CE��B��C��E�ADF��E��F��B��BF�

��EC��E�ADF��E��!�B��FB��F�B��C��EC�E�

�F��B�� B� �B�BE�� B�C�� BFF�FE�� B� ��E��

��C��F��CF��C��E��C��E�DDBE��C�C��

��F�1�F��C�� C��C�� C�� C�� EC�� FB��C�B�

��BF��EC��E��F��BE��

F��D��A -��(A !�B� E��F�� E��FBE� �� B�

�A��B� �B��C��B�� E�A�BE�E� DC�B� ��B� %��EC�� I��CB��

!�B��B�� 9=� �� F�� DF��E� �FB� DC�B�� C�� B�

!�A�B�+��B��9=��EC��E�A<B��E��FB�DC�B��

C��!�A�B�3��

!�A�B�+��<�B��<(�B��)�B��B��CB�C��)8(��

��B��6��7��

�

D�B�� EC�BFC�D� ��B� %I� �� F�� E�A<B��E��

�� B��DB�DF��E�E��FB��A��B��B�%I�E��FB��+��4��

��B� �� E��FB��C��E� DC�B�� A��/BF�BE0� '+�+,��

%��D�� B� E�A<B��E� E��FB�� A��B� ��B� ��

��BE� ��B�%I�E��FBE�A�� C�� CE�#�C�B�B�C�B�� B��E� ��B�

�DB�DF��C��FB�EB��B�%I��BE�C��BEB�C��C�C��E�

��E��B�FB�EB��A

!�B� ��C�� EC�E� E��FB�� B� �CD�BE�� B��

��3��E��BEB�C��C�C��E��BF��F��B��E�AB��BF�C��

��C�E� B��B�� C�� C�D� ��E �� O��B�� A��

�F��E��F�C��EB�E�F��EC�E��B��G+�AF��

�F��E��F�C��F��EC�E��E��FB��FG�33��!�B�

��C�� EC�E� �� B�� A�3� ��B��

A��C��(F��?E��EC�E��BF��%I��B��A��3��

��BFB� �E� ��B� DBF�C� B?E� ��EC�E� ��B� E��FB�� B��

�� D��A�� EC�E� �� B�� 3��3�

��C�� E� �BF�� FB�� BF� ��B� ��

��EC�E��

;F��'NFE,� E�%I� !��

�� 9=� �� 9=�

�31�3� �3� ++�G+� 3�F�F� ++�G+� �F�F�

�1A3� �3� ++�� 3�F++� ++�� F++�

A1F3� �3� ++�AG� 3�4AG� ++�AG� �4AG�

F1G3� �3� +4�+A� �3+� +4�+A� �3+�

G1�3� �3� +4�A4� �F� +4�A4� �F�

8C��B�C��9��9��*44%3$4��#B��3D��):#��);��<E��

�43�

�

!�A�B�3��<�B��B��)�B��B��CB�C��)8(��

BAB�C��6��7��

;F��Q� E� %I� !��

�� 9=� �� 9=�

(%� F� A��3�� FA� A��3�� FA�

D%� F� ��

%%� �� 3� A�+� 3��3� ��A3�

!*�� FG�33� 3��A� ��3� A�+�

!*9� �� G+�AF� ��FG� FG�33� 3��A�

*%� � A�3� 11� G+�AF� ��FGG�

;%� �� 3��3� ��A3� A�3� 11111�

Q(%1� (F��?E� %��EC�6� D%1� DBF�C� B?E� %��EC�6� %%1�

%��C��%��EC��!*�1�!F��E��F�C��F�%��EC�6�!*91�

!F��E��F�C�� 9B�E�F�� %��EC�6� *%1� *��C�� %��EC��

;%1�;��A��%��EC��

%��F�C�D� ��/BF�BE0� '+�+,� ��B�%I� E��FBE� ��F�

��B� ��EC�E� �FB� �BEE� ��B�� FB�� B� ��

E��FB� '+��4,�� BFBA�� C�� CE� E�� E��F�C0��C��

��B�%I� E��FBE� ��F� ��EC�E��BFB� �E� ��E�P� ��C��

E��FB� %I� '4��,�� C�� EC�E� 'G3�F,��

�F��E��F�C��EB�E�F��'F+�G,�� F��E��F�C��F�'FA�A,��

DBF�C� B?E� '�+�3,�� (F��?E� '��,� �� D��A�� '3�F,�

FBE�B��C�B�� !�B� �FBEB�� E�� %I� E��FBE� �FB�

��C�D� ��B� ��C�� EEC�C��C�� DC�B�� A��

'/BF�BE0��+�+,��

AAAAAAAAAAAAAAAAAAAAA

"��%��D��DA

8�� B� �FBEB�� C��BE�CD��C�� B�� E�

��B� �� B�B�� BE�BF�� EC�� A��BF�� C�� !B��D��

A�EB�� B� �FC��C��BE� �� DBE�BF�� %��EC�� (��BF��

'/BF�BE0��+�+,��E��EEBEE��B��D��DB��AC�C�CBE�

C��E��C��C��D��DB��D��

!�B�E��E��B�C��7��EBE��8��7��EB�1��

�B�B��B��B��BE��E��B��C��EB�1��B�

�BE�� E� ��C�CE�BFB�� 33� ��F�� 3� ��EC��

E�A<B��E�� !�B� �FBEB�� BE�� A��BF�� E�A�BE�E� ��C��

�FB� A�EB�� EC�C��F� �C�BE� �E� �� '/BF�BE0�� +�+,��

R��BF� B�� E�A�BE�E�� B� �BE�� BFC��E� �BFB�

�F��E��C��E��D%(1B�D�CE��'/BF�BE0��+�+,��A��E��B�

��B��BFC��E��C�CB��F�C�D��B��FB#�B��

��FFB��B� �� F�E� �� C�D�CE�C�� FC��C��BE� ��

!B��D�� !�CE� �BE�� F��C�B� �� A<B��C�B�

�EEBEE�B�� F� �EEBEEC�D� C��C�C��E��C�� EC��

C�� !B��D�� D��DB�� !�CE� �BE�� A��BF�� AB� �EB��

��EEC�� FC��E� ��EC�� E��F��BE�� 8�� AB� �EB��

��BF��F�DF��E�C��C��C�C��E��C��EC��

+�B�E��%�D�

(�EE��%��)�*�AB��C��.��'+++,��*�C�C��E�B��E��EC��

8�� ;�� =B�BE� �� 7C00��CD�C�� 'B�E�,�� ;B��

+�C�C�B�� B�� =1��C��B�� .��E�A��E�� B�E��

9�EEB�2�7E��D��7FBEE�A(�� S�� H�� 9�C��F�� %�7�� =B�CE�� =�*�� /F��BF�� S�� )�

/��B��'�33,��HBFA��BECD��B��C��CB��E�

�C�� F�� AB� �BEC��E�� >��B�� B�C��B��

.��E�A��C�B��)��C��E��&��F4G1+G��

(��D�� 9�� /�� S�� B��E�� %�� )� 9�� *�� '++3,��

%��EC��BE�E�FB��EC�BFB��AB�C��E��'��G�1+��

*F�F��%��)�.��C��S��;��'+4+,��7FB�C��C�D��B��BE�BF��

��EC�� A��BF�� EC�� #��CB��>��B�� )��A�B��

;�B�C��8C��/'��G�1GG��

*F�F�� %�� :�� E�S�� )� ��C�E �� %�� B�� '+4+,��

7FB�C�C��F��7E��B�FC��B��C��B��EC��

E�FBB�C�D��F��AB�C��E��,��G1G4��

*�C��C��B��)�DC��C��E��9��B�� '++3,��.B�B�C�C��B�C�C�E� C��

��FBB� ��EC�� E��F��BE�� >��B�� +��C�B�C��

8C��*,��144��

*��B��.��'++A,��:B��B��C��C��B��C��C��B��

BAB�CB��(��C��FB2�DC��C��E�)�DC� C�E��

*��B�� .�� '�33,�� :B��B�� C��C�� B��C�� C��

BAB�CB�B��B��C��C�B�C��C��

'�F��B��,��(��C��FB2��C��C��DC��C��E�)�DC� C�E��

;��D��EE�� :�� )� /�� B�� '+��(�� D�� CB��C��

�1B�C�B�C��BAB�CB��D�E�C�D��2��B��OBACDBF��

:BC�� /�� )� 9��BE�� .�� S�� '+�G,�� !�B� ��FB� ��

��FB�B�EC�� BFF�FE� C�� (F��?E�� *��C��

DBF�C� B?E��EC�E��+��1��$*��F41�GF��

:�� %�� )� (�F��B�� *�� '+4F,�� 9��B� �C��BFB��C��

B��B��E��DB��E�F� B1FB��B��EC��8��:��R��E ��

'B�E�,��FA��B�C��BC��9��=CBD�2�*��BDB�:C��7FBEE��

/BF�BE0��%��)� 7��B�� B�� '+�A,�� !�B� ��EC�� #��CB��2� !�B�

��C�� F�� B�E�FB�B�� EC��

�CE�AC�C��+B�B�CB��>��B��.��C�B��)�C��

�1G��

/BF�BE0�� %�� '+�+,�� AB�CB� B�� B��CB�� C��?�

FB1��E��B�C!B�C��B��E��EB��N�F 2�;F��B�

)�9�F��8��

/BF�BE0�� %�� '+4�,�� BAB�CB� �B��E�� EB�� N�F 2�

;F��B��9�F��

/�� 9�� B��)� ;��D��EE�� :�� '+4F,�� 7C��FB1��C�D� C��

��EC��D�BC��B��:B��B��*'��GG14��

9�B��*��)�/BF�BE0��%��'+43,��.B�C�AC�C��C�C��

��F��BFCE�C�E��B��BAB�CB��B��E��>��B��

)��A�B��;�B�C��8C��'/��341��A��

9�FBB�� -�� )� .CEEBF�� %�:�� '�33�,�� BAB�CB��

EB��N�F 2�-��F��R�C�BFEC��7FBEE��

�

Cognitive cueing: Vocal variability and naturalness

281

Usefulness of Cognitive Cueing in Eliciting Vocal Variability and Vocal Naturalness

Tiffy George Roy1& Yeshoda K.2

Abstract

Voice is important as an emotional medium as it is the carrier for spoken language. The commonly measured acoustic correlates of emotions are pitch, duration, intensity and voice quality. The use of cognitive cues as a task elicitation method stems from cognitive-behavioral therapy and focuses on prompting the individual to think about and feel a task prior to completing the task. There are limited studies on the naturalness of speech with respect to normals and its influence on emotions. In the present study an attempt was made to understand the usefulness of cognitive cueing in eliciting vocal variability employing different emotions and also assess the naturalness of the same. Sentences were constructed for three emotions: neutral, happy and sad. Twenty participants who were competent in English language were considered for this study. Each participant was asked to read the sentences in the uncued and cued conditions. The results in general, revealed increased means values for most of the vocal parameters in cued conditions across all emotions. The qualitative analysis also revealed that the judges perceived the sentences in cued conditions to be most variable and more natural than the uncued conditions. This indicates variability in acoustic parameters both quantitatively as well as qualitatively across conditions (uncued and cued). Hence, it could be concluded that cognitive cueing brought about changes in vocal attributes which could be quantified and perceived, strengthening the view that cognitive cues could help in achieving greater variations in voice and speech thereby, making the speech more natural.

Key words: cognitive cues, emotions, mental imagery, vocal variability, vocal naturalness

veryday 1experience suggests that voice carries clues to the underlying emotional state of the speaker and this plays an important aspect in the

study of human communication. An emotion is a mental and physiological state associated with a wide variety of feelings, thoughts, and behavior. Emotions are subjective experiences, often associated with mood, temperament, personality, and disposition. Emotions color the language, and can make meaning more complex. Generally materials from three categories are used in investigating emotional speech: spontaneous speech, acted speech and elicited speech. Spontaneous speech contains the most direct and authentic emotions, but the difficulties in collecting this kind of speech are also extensive. In the ideal condition, speakers should be recorded without knowing about it, so that they behave completely naturally. The acted speech is more stereotypical, and that the expression of emotions is more extreme than in spontaneous speech. In elicited speech the idea is that certain emotions are induced. The induction method has the positive feature that it gives control over the stimulus; on the other hand, different subjects may react differently on the same stimulus. The validity of such elicited, or induced, speech depends to a large extent on how successful the induction process was (Stibbard, 2001).

1e-mail: tiffy 1986@ gmail.com; 2 Lecturer in Speech Sciences, AIISH, Mysore, [email protected].

Emotions have a physiological effect on the larynx which is reflected in the acoustic parameters of speech. Risberg (1986) reported that happiness gave an increase in pitch and pitch range while Öster and Risberg (1986) noted a slow tempo. Fónagy and Magdics (1963) described it as “lively”. However, findings described sadness as exhibiting normal or lower pitch, narrow pitch range and slow tempo (Skinner, 1935; Davitz, 1964). Comparing both these emotions, neutral emotion has a lower pitch, with pitch tending to be normally distributed about the average pitch level (Cowan, 1936). The commonly measured acoustic correlates of emotions are pitch (fundamental frequency, both average and range), duration, intensity and voice quality (Murray & Arnott, 1993). Scherer (1981) noted various vocal indicators of emotional expression principally pitch, intensity and vocal quality with pitch which is often used to differentiate between emotions.

The efficacy of cognitive therapy in the

modification of certain cognitive processes resulting in the amelioration of symptoms is well documented in psychological literature. Cognitive cueing is an approach to voice treatment that stimulates thought patterns changing the speakers‟ voice. Cues of this type have been traditionally used by drama and singing teachers to elicit vocal behaviors (Andrews, Shrivastav & Yamaguchi, 2000; Bohnenkamp, Andrews, Shrivastav & Summers, 2002). Mental images and cues have traditionally been employed by voice teachers and voice clinicians to elicit and shape vocal behavior. A

E


282

key aspect of cognitive cueing is an individual‟s level of cognitive development. In order for cognitive cueing to be effective, an individual must be able to form mental representations of objects and realize that objects exist even when they are not in sight (Piaget, 1952).

Studies on the effects of emotions on the

acoustic characteristics have shown that average values and ranges of F0 differ from one emotion to another. There are several reasons why changes in F0 with time are potentially capable of providing information concerning the emotional state of the speaker. First, considerable latitude of change is possible in the variation of F0 since only certain aspects of F0 contour carry information with regard to the linguistic content of a message. The principle linguistic functions of F0 changes are to indicate stress and to mark boundaries of different types of sentence length or phrase length units. Subject to these constrains, a speaker is relatively free to use changes in F0 to convey nonlinguistic information, such as, emotions. The fundamental frequency can undergo variations that may not be intended or be under overt control of the speaker and hence may provide an indication of the speaker‟s emotional state. Thus emotional states are said to influence vocal quality as a result of changes in the muscle tonus. These changes are primarily brought about by the functioning of the sympathetic division of the autonomic nervous system.

Cowan (1936) stated that unemotional speech

has a narrow pitch range compared to emotional speech, with pitch tending to be normally distributed about the average pitch level. Williams and Steven (1972) also reported that in neutral situations, the sentences were generated with shorter duration than for the emotional situations.

Cowan (1936) and Öster and Risberg (1986)

reported that happiness gives an increase in pitch and pitch range. Öster and Risberg (1986) also noted a slow tempo, while Fónagy and Magdics (1963) described it as “lively”. Davitz (1964) reported an increase in speech rate along with an increase in intensity.

Williams and Steven (1972) observed the

average fundamental frequency for speaking in sorrow situations and found it to be considerably lower than that for neutral situations and the range of F0 was usually quite narrow. This change in F0 was accompanied by a marked decrease in the rate of articulation and an increase in the duration of an utterance. The increased duration resulted from longer vowels and consonants and from pauses that were often inserted in a sentence. In the review of Johnstone and Scherer (2000); Scherer and Ceshi (2000), sadness was

proved to decrease the mean fundamental frequency, F0 range and variability, speed and articulation rate and intensity. The patterns were contrary to the acoustic patterns in happiness in which a rise in the mentioned patterns was seen.

In comparison with the neutral emotion, Murray

and Arnott (1993) classified happy and sad emotion on some acoustic characteristics. Happy emotion elicited a faster or slower rate, had a higher pitch and a wider pitch range, a higher intensity, with smooth upward inflections than neutral emotion. Comparing sad emotion with neutral emotion, it was reported that the average pitch and pitch range was slightly lower with reduced intensity and slower speech rate for sad emotion. The voice quality in this emotion was reported to be resonant and had a downward pitch inflection with articulation reporting to be slurred.

Scherer (1986) concluded that F0, energy and

rate may be the most indicative of arousal. Arousal is defined as a subjective state of feeling activated or deactivated (Sanchez, Kirschning, Palacio, & Ostrovskaya, 2005). Changes in F0 and rate of speech can be attributed to many factors such as general temperamental and personality characteristics of the individual. Individuals vary greatly in the rapidity of their motor activity such as walking, as well as with alacrity with which they think and come to a conclusion and even express themselves. Individuals who think slowly, move ponderously also speak with a deliberate; even tempo while for individuals who are active and quick, their speech is likely to be hurried and lively. Rate can also be a characteristic of the mental state. Slow rate can be a characteristic of mental state such as wonder, doubt, deep thought and sorrow while rapid tempo is associated with joy, excitement, humor and anger.

Voerman, Langeveld and Van Rossum (2009)

conducted a study on the short- and long-term results of two techniques (mental imagery and manual shaking of the larynx) in 116 patients with non-organic dysphonia or aphonia. It was found that the group of patients who retrieved their voice was subjected to mental imagery tasks but relapses were more common in this group. It was thus concluded that the cure rate for mental imagery was much higher than that for laryngeal shaking.

It has been reported that expression of emotions

can be mediated by personality characteristics (Gross, John & Richards, 2000; Barrett & Niedenthal, 2004). This entails that some people are more expressive than others and may be more emotional than others. Gender differences have also been reported stating that women are more expressive than men.


283

The dictionary defines naturalness as ease and ease as freedom. The adjective „natural‟ derived from the Latin word „naturalis‟ means of nature meaning true to nature and freedom from artificiality, affectation or constraint. Till today, there is no clear definition of naturalness. There are numerous studies that speak of the measurement of naturalness. But these studies pertain to its evaluation in stuttering especially to assess naturalness of speech in recovery from stuttering. There are limited studies on the naturalness of speech with respect to normals and its influence on emotions. Hence, in the present study an attempt was made to understand the usefulness of cognitive cueing in eliciting vocal variability employing different emotions and also assess the naturalness of the same. The aims of the study were to quantitatively analyze the vocal variability of speech in uncued and cued conditions across emotions and compare the same, to assess the naturalness of speech qualitatively in uncued and cued conditions across emotions and to correlate the obtained quantitative and qualitative measures.

Method

Participants (Speakers): Twenty participants (10 males and 10 females) in the age group of 20 to 25 years with mean age of 22.5 years were considered for this study. Participants who were fluent and competent in English language use were only chosen. The participants selected in this study had Kannada as their first language and English as their second language. Participants were excluded if they had a velopharyngeal disorder, or were perceived by a speech language pathologist to have a voice disorder, abnormal oral-peripheral structures, or hearing loss, neurological or psychological problems.

Participants (Judges for qualitative analysis): Five qualified speech-language pathologists (three males and two females) with a minimum of two years of clinical experience formed the judges for the perceptual experiment in qualitative analysis.

Procedure Materials (Quantitative analysis): Three primary emotions were considered in this study: happiness, sadness and neutral emotion. Sentences that evoked neutral, happy and sad emotion were considered. For each emotion two sentences (a total of 6) were constructed and written on A2 size cards. For every sentence cognitive cues were constructed and all cues were audio recorded. These cues were targeted to create specific mental images and to elicit the chosen emotions. Instructions: All the participants were instructed as follows: „A few written sentences will be shown. You are required to read the sentences aloud. After reading

all the sentences once, you will have to listen to a few audio recordings and then read the same sentences again. All your speech samples will be recorded‟. Recording: All the read samples were recorded in a quiet environment on to Olympus Digital Audio recorder (WS-100, Japan) with Verbatim Headset with Collapsible Microphone, (Recorder No: 41691) with the microphone-mouth distance of 5”- 6”. Each participant was asked to read the sentence three times. Trial 1: The cards with sentences were presented individually and instructed to read all the sentences aloud. This was the no cue condition. Trials 2 and 3: The subjects were asked to read the sentences after listening to the recorded cognitive cues. This was the cognitive cued condition. The sentences and cues used are shown in the Appendix. Instrument : Quantitative assessment was carried out using the software Real Time Pitch software (Model 5121) of Computerized Speech Lab 4500 (KAY Elemetrics).

Acoustic analysis involved two phases. The samples of trials 1 and 3 were considered for acoustic analysis.

Phase I: Quantitative analysis The quantitative analysis involved extraction of the following parameters using the software Real Time Pitch. 1. Mean Speaking Fundamental frequency (MSF0):

It is the average pitch that is used during speaking and is expressed in Hertz (Hz).

2. Standard Deviation of Speaking Fundamental frequency (SDSF0): The standard deviation reflects the frequency variability for a reasonably large time segment or passage.

3. Lowest speaking fundamental frequency (LSF0): The lowest pitch the individual can produce during speaking and is expressed in Hertz (Hz).

4. Highest speaking fundamental frequency (HSF0): The highest pitch the individual is able to produce during speaking and is expressed in Hertz (Hz).

5. Variation in Fundamental frequency (vF0): It is the relative standard deviation of fundamental frequency which reflects the variation of F0 within the analyzed voice sample. It is expressed in terms of %.

6. Sentence Duration: It is the period of time during which the sentence is spoken, expressed in seconds.

Phase II: Qualitative analysis Materials: The samples of subjects of the uncued trial (trial 1) and cued trial (3) were transferred onto an audio CDS. For each subject six tracks were created


284

and in total, 120 tracks formed the material for qualitative analysis. Instructions: Each of the judges were instructed to listen to the audio samples individually and rate each of the sentences (uncued and cued) on a three point rating scale where 1 represented least variable and 3 represented most variable. They were given a free hand and instructed to listen to the samples as many times required and rate the sentences. Qualitative parameters: Naturalness of speech is ambiguous and most difficult to quantify. Review of literature indicated no single attribute or any established attribute to signify naturalness. Parameters of speech as shown below were chosen for description of speech naturalness. 1. Voice - Pitch, loudness and quality 2. Articulation - Pronunciation and intelligibility 3. Fluency - Rate, effort and continuity 4. Prosody - Intonation and stress Scoring sheets incorporating these parameters were prepared and presented to the judges. Statistical analysis: All the extracted vocal parameters of the quantitative analysis of the two trials (uncued and cued) were subjected to statistical analysis. Descriptive measures, repeated measure ANOVA and paired t test were obtained to check for significance, if any, across the trials, emotions and gender. Qualitative analysis: The individual scores of the judges were pooled emotion wise and overall percentages were calculated. The results were tabulated and presented.


The results are discussed under different sections:

Section I: Quantitative analysis a) Mean Speaking Fundamental Frequency

(MSF0) The MSF0 for the uncued trial and cued trial is

shown in Table 1. In the uncued neutral emotion, the female group obtained a mean value of 228.56 Hz (SD=21.54) which was not significantly different from the cued neutral emotion with a mean value of 229.51 Hz (SD=31.48) Also, there was no significant difference between the mean scores obtained amongst the uncued and cued condition in neutral emotion for males with values of 139.61 Hz (SD=11.93) and 140.84 Hz (SD=26.85) respectively. The emotion of happiness yielded a significant difference between the trials and across gender. Similarly, males obtained a value of 125.83 Hz (SD=8.54) in the uncued condition and 169.12 Hz (SD=41.18) in the cued condition.

However, there was no difference in the means obtained across the trials for sad emotion for males and females. The females obtained a score of 219.33 Hz (SD=20.95) and 220.19 Hz (SD=22.64) in the uncued and cued condition respectively and males secured a mean value of 125.99 Hz (SD=7.99) and 125.99 Hz (SD=14.23) for the uncued and cued condition respectively. The overall level of significance calculated using the repeated measure ANOVA is displayed in Table 2. There was a significant difference observed across the three emotions [F (2, 114) = 21.20, p<0.05] and no difference across gender [F(1,114) = 3.567. p>0.05]. An overall change between the uncued and cued trials was also seen [F(1,114) = 33.30, p<0.05].

Table 1. MSF0 in uncued and cued condition across gender and emotions

Emotions* Gender Mean(Hz) SD

UCN Female 228.56 21.54 Male 130.61 11.93

UCH Female 222.61 20.51 Male 125.83 8.54

UCS Female 219.33 20.95 Male 125.99 7.99

UC Total

Female 223.50 21.00 Male 127.48 9.74

CN Female 229.51 31.48 Male 140.84 26.85

CH Female 247.49 26.41 Male 169.12 41.18

CS Female 220.19 22.64 Male 125.12 14.23

C Total Female 232.40 28.97 Male 145.03 34.36

*UCN=Uncued neutral emotion; UCH=Uncued happy emotion; UCS=Uncued sad emotion; UC= Uncued; CN= Cued neutral emotion; CH= Cued happy emotion; CS= Cued sad emotion; C= Cued

In general, the MSF0 varied across conditions

and emotions in both males and females as expected. However, a significant difference was observed only for the emotion happy. The results conform to the findings of the acoustic patterns of the emotions reported by various authors. Experiments have proved that in unemotional speech such as that of neutral emotion, there will not be much changed present in the mean F0 and its parameters and the above results are in line with the studies stated by Cowan (1936) and Williams and Steven (1972). The results are in parallel to the study done by Murray and Arnott (1993) as happy emotion had shown an elevation in mean F0 while sad emotion had not changed considerably from the uncued and the cued trial. This difference was noticed irrespective of gender. It can be thus assumed


285

that the presentation of cognitive cues elicited vocal variability in the direction of emotions. It is evident that the subjects were influenced by the cues given to them and this influence is manifested in the overt, vocal expression. Table 2. F value and significance of the overall effect

of uncued and cued conditions across gender and emotions for MSF0

MSF0 F value Change across emotion 21.205* Change across gender 3.567 Overall change between uncued and cued conditions

33.303*

*p<0.05 b) Standard deviation of Speaking Fundamental

frequency (SDSF0)

Table 3. SDSF0 in uncued and cued condition across gender and emotion

Emotions* Gender Mean S D

UCN Female 25.01 8.61 Male 19.21 12.29

UCH Female 32.57 13.95 Male 21.57 10.09

UCS Female 33.90 16.69 Male 19.46 10.88

UC Total

Female 30.49 13.85 Male 20.08 10.99

CN Female 42.27 12.65 Male 24.67 13.20

CH Female 45.37 11.66 Male 33.89 12.95

CS Female 30.16 10.28 Male 19.52 9.11

C Total

Female 39.26 13.16 Male 26.03 13.14

*UCN= Uncued neutral emotion; UCH=Uncued happy emotion; UCS=Uncued sad emotion; UC= Uncued; CN= Cued neutral emotion; CH= Cued happy emotion; CS= Cued sad emotion; C= Cued

Table 3 depicts the mean SDSF0. SDSF0

reflected the frequency variability across the two conditions [F(1,114)=22.29, p<0.05] and across emotions [F (2,114) =8.75, p<0.05] but there was no significance across gender [F (1,114)=0.82, p> 0.05] (Table 4).

It was seen that in neutral emotion, females were observed to have a significant variability in frequency from the uncued trial to the cued trial while males did not have a significant difference. Happy emotion, showed a significant difference between the trials and across gender. There was a significant rise from 32.57(SD=13.95) in uncued condition to 45.37 (SD=11.66) for females and a rise of values from 21.57 (SD=10.09) to 33.89 (SD= 12.95) in uncued and cued trials respectively in males. The values did not differ to a large extent for sad emotion in females and males. Females obtained a SDSF0 of 33.90 (SD=16.69) in uncued trial and 30.16 (SD=10.28) in the cued trial. Males also obtained a score of 19.46 (SD=10.88) in the uncued trial and it remained the same for the cued trial with a value of 19.52 (SD=9.11).

Table 4. F value and significance of the overall effect of uncued and cued conditions across gender and

emotions for SDSF0 SDSF0 F value Change across emotion 8.75* Change across gender 0.82 Overall change between uncued and cued conditions

22.29*

*p<0.05

As seen in Table 4, there was a significant difference in the reading of the subjects before the cues were presented and after the cues were presented. There was an obvious difference across the emotions as the acoustics between the emotions varied. It can be observed that variation of F0 was more for females than for males in the neutral emotion. This is quite contrary to the characteristics pattern for an unemotional speech. This could be because of the cognitive cues used to elicit the sentence. For the neutral emotion, the sentence, „this is a pen‟ was considered. The cue given for this sentence was to imagine the concept of a pen being taught to a child. It is well known that in sociolinguistics, a sentence when used with children will be modified drastically. In this study, although males had variations in F0, females were observed to have had a significant variation. This is consistent with the studies with respect to gender that females are more expressive compared to the male group. Happy emotion elicited a significant upward inflection after the cognitive trial which is in accordance to the studies by authors, Cowan (1936), Öster and Risberg (1986), Murray and Arnott (1993). But for the emotion of sadness, no significant change could be observed between the conditions. Johnston and Scherer (2000) found sadness to decrease the mean fundamental frequency and have a lesser variations. In the present study, the F0 did not vary significantly for the neutral emotion and the variations in F0 were minimum.


286

c) Lowest Speaking fundamental frequency (LSF0)

Table 5 shows the mean scores of LSF0. Repeated measure ANOVA showed that there was no significant difference between the trials, emotions and gender [F (1,114) = 2.37, p>0.05], [F(2,114)= 2.02, p>0.05] and [F (1,114) = 1.83, p>0.05] (Table 6). A significant difference was observed only for females in the neutral emotion. There was decrease in low F0 from 173.98 (31.48) to 144.66 (49.62) from uncued to cued condition. As for males, there was no statistical difference in the mean low F0 as the value obtained were 103.28 (13.65) in uncued condition and 97.09 (20.79) in cued condition. In happy emotion, there was a slight decrease in LSFO from 137.91 (39.06) to 133.27 (40.89) in females and in males, there was a slight increase from 95.83 (11.28) to 101.04 (27.28). As for sad emotion, there was only a difference of approximately 1Hz between the uncued and cued condition.

Table 5. LSF0 in uncued and cued condition across gender and emotions

Emotions* Gender Mean SD

UCN Female 173.98 31.48 Male 103.28 13.65

UCH Female 137.91 39.06 Male 95.83 11.28

UCS Female 145.91 38.94 Male 95.36 11.85

UC Total

Female 152.60 39.27 Male 98.16 12.63

CN Female 144.66 49.62 Male 97.09 20.79

CH Female 133.27 40.89 Male 101.04 27.28

CS Female 144.65 42.83 Male 94.05 11.88

C Total

Female 140.86 44.17 Male 97.39 20.80


Females scored a value of 145.91 (38.94) in uncued condition and it decreased to 144.65 (42.83) in cued condition. Males obtained a value of 95.36 (11.85) in uncued condition and it decreased to 94.05 (11.88) in cued condition.


emotions for lowest speaking fundamental frequency

LSF0 F value Change across emotion 2.018 Change across gender 1.83 Overall change between uncued and cued conditions

2.37

*p<0.05 d) Highest speaking fundamental frequency

(HSF0) The mean values of the change in HSF0 in uncued

and cued condition across gender and emotions are grouped in Table 7.

Table 7. HSF0 in uncued and cued condition across gender and emotions


UCN Female 287.56 34.52

Male 176.41 49.48

UCH Female 307.17 41.19 Male 208.40 55.81

UCS Female 300.50 44.97 Male 185.53 35.70

UC Total

Female 298.41 40.60 Male 189.80 48.66

CN Female 324.14 43.28 Male 189.03 53.69

CH Female 345.23 57.66 Male 269.72 75.82

CS Female 298.52 35.19 Male 180.61 37.63

C Total

Female 322.63 49.43 Male 212.16 69.32


Table 8 portrays that there is a statistically significant difference between the trials and emotions [F(1,114)= 17.642, p<0.05] and [F(2,114)= 7.428, p<0.05] respectively. From Tables 7and 8, it can be viewed that happy emotion evoked a rise significant difference from uncued to cued condition in both females and males. In the neutral emotion, only females showed a rise in HSF0 from 287.56(SD=34.52) to 324.14(43.28).


287

There was barely a change in males in neutral emotion across the trials ranging from 176.41(SD=49.48) in uncued condition to 189.03 (SD=53.69) in cued condition. In sad emotion, there was no significant difference between the uncued and cued trial across the genders. Females scored a value of 300.50 (SD=44.97) in uncued trial and it slightly decreased to 298.52 (SD=35.19) in cued trial which is not a significant change while males scored 185.53 (SD=35.70) in uncued trial which decreased to 180.61 (SD=37.63) in the cued trial. As observed, the lowest speaking fundamental frequency did not change much across the emotions for the cognitive cued trial but the highest speaking fundamental frequency did have a significant change and this change was different with respect to emotions.

Table 8. F values and significance of the overall effect of uncued and cued conditions across gender and

emotions for HSF0

*p<0.05

The difference between lowest and highest F0 (Range) was wider in females than males whose range was lesser. This shows the expressive nature of females. It also raises a significant issue of the cues that is used for eliciting the vocal changes. The sentence „this is a pen‟ with the situation of talking to a child elicited more pitch variations in the speech of the individuals while the following sentence „the tests are in the cupboard‟ with a situation of helping a friend find the test did not yield much variation. By manipulating the cognitive cues presented to the subjects, the vocal parameters could be varied to suit the variations desired by the experimenter. To further support this assumption, the study also proves a widened F0 range for happy emotion and a narrow range with no significant difference from the unemotional reading for sad emotion. The statement can be thus made that the subjects were influenced by the cognitive cues that were presented in the direction of the type of cues that were used.

e) Variation in fundamental frequency (vF0) Table 9 presents the mean values of the vF0

across the emotions. As seen in Table 10, there was an observable difference between the overall mean for uncued and cued condition but there is only a slight difference between the emotions and no difference in gender as per the repeated measure ANOVA. From Table 9, it can be observed that there is only a significant difference in the values obtained for females

in neutral and happy emotion and not in sad emotion. In neutral emotion, there was an increase in variability from 0.11 (SD=.037) to 0.18(SD=.060). In happy emotion, there was a slight raise from 0.14 (SD= .036) to 0.18 (SD=.045). But there was no significant difference in VF0 of males in all the three emotions.

Table 9. vF0 in uncued and cued condition across gender and emotion

Emotions* Gender Mean S D

UCN Female 0.11 0.037 Male 0.14 0.078

UCH Female 0.14 0.036 Male 0.17 0.076

UCS Female 0.15 0.080 Male 0.15 0.078

UC Total

Female 0.13 0.057 Male 0.15 0.079

CN Female 0.18 0.060 Male 0.16 0.052

CH Female 0.18 0.045 Male 0.19 0.064

CS Female 0.14 0.054 Male 0.15 0.066

C Total

Female 0.17 0.055 Male 0.17 0.063

*UCN= Uncued neutral emotion; UCH=Uncued happy emotion; UCS=Uncued sad emotion; UC= Uncued; CN= Cued neutral emotion; CH= Cued happy emotion; CS= Cued sad emotion; C= Cued Table 10. F value and significance of the overall effect

of uncued and cued conditions across gender and emotions for vF0 vF0 F Value

Change across emotion 3.64* Change across gender 1.00 Overall change between uncued and cued conditions

10.31*

*p<0.05 Exploring this particular parameter, it can be

noted that registers have played a role in the significant difference in variation that is observed in females as compared to males in the neutral emotion and sad emotion is known to have a very narrow range of change. However, there was a significant variation of F0 for females and a slight change in males for the emotion of happiness. This reason can be due to the individual differences in which the cognitive cues were perceived and to an extent, there was also an influence of the life experiences and the personality of the individual. The variability in voice was scattered in the study and when they are averaged, there was no significant difference across the trials especially in the case of males. Individual difference can be present on decoding of the cues that were presented.

HSF0 F value Change across emotion 7.428* Change across gender 0.011 Overall change between uncued and cued conditions

17.642*


288

f) Sentence duration Table 11 depicts the mean sentence duration in

uncued and cued condition across gender and emotions. It portrays that there was no significant difference between the sentence duration in neutral and happy condition for both males and females but there was a significant increase in duration for sad emotion for both females and males. Females had a mean duration of 1.87 (SD=0.359) in uncued condition which increased to 2.10 (SD=0.589) after the sentence was read after the presentation of the cognitive cue. Males, at the same time, had a mean duration of 1.62 (SD=0.354) in the uncued trial and the duration of the sentence read increased to 1.89 (SD=0.0455) after the cognitive cue was presented.

Table 11. Sentence duration in uncued and cued condition across gender and emotions


UCN Female 1.50 0.444 Male 1.35 0.495

UCH Female 2.30 0.230 Male 2.09 0.255

UCS Female 1.87 0.359 Male 1.62 0.354

UC Total

Female 1.89 0.480 Male 1.69 0.485

CN Female 1.46 0.321 Male 1.53 0.424

CH Female 2.26 0.260 Male 2.21 0.347

CS Female 2.10 0.589 Male 1.89 0.455

C Total

Female 1.94 0.539 Male 1.88 0.492


Table 12 shows that there was a significant change in the overall scores across trials [F(1, 114) = 10.85, p<0.05] and across emotions [F(2, 114) = 3.51, p<0.05]. However, there was no significant change across gender [F(1, 114) = 3.916, p>0.05].

As stated in literature, it can be noted in the

present study that the duration did not vary significantly for neutral and happy emotion while there was a longer duration for the emotion of sadness. Murray and Arnott (1993) reported that the duration can be faster or even slower in happy emotion in comparison to the neutral emotion while it is slower for the sad emotion. Davitz (1964) and Fonagy (1981) reported a downward inflection, slurred enunciation and rhythm with irregular pauses which contributes to

the longer time taken for this emotion. Williams and Steven (1972) also reported the increased duration resulted from longer vowels and consonants and from pauses that were often inserted within the sentence for sad emotion.


emotions for sentence duration Sentence duration F value

Change across emotion 3.517 Change across gender 3.916 Overall change between uncued and cued conditions

10.85

*p<0.05 Section II: Qualitative analysis

Table 13 gives an overall view of the overall percentage given by the five judges for the samples on the 3 point rating scale.

Table 13. Percent scores of the judges‟ ratings for uncued and cued trials across emotions

Emotion Uncued

condition Cued

condition

Neutral 1 = 8.10% 2 = 34.30% 3 = 57.60%

1 = 12.85% 2 = 31.95% 3 = 55.20%

Happy 1 = 13.65% 2 = 33.05% 3 = 53.25%

1 = 5.30% 2 = 26.00% 3 = 68.70%

Sad 1 = 7.40% 2 = 34.85% 3 =57.75%

1= 7.40% 2 = 31.8% 3 = 60.8%

It can be seen that for neutral emotion, the judges perceived the sentence read in the uncued trial a little more natural 57.6% of the times than the cued condition 55.2% of the times. This difference in the perception could be attributed to the cognitive cues that were provided. For the first sentence, „This is a pen‟, the cue given was that of „how would you teach a child the concept of a pen‟. This cue elicited exaggerated pattern making the listeners perceive the sentence as unnatural. Happy emotion exhibited the greatest change in the perception from the uncued to the cued trial. 68.7% of the times the judges perceived the cues condition more natural than the uncued condition while only 53.25% of the times, the listeners perceived the uncued happy sentence as natural. There was also a rise in the percentage of perception of cued condition for sad emotion than in the uncued condition. The results obtained from the perceptual evaluation can be thus attributed to the fact that the provision of cognitive cues, indeed did improve the naturalness of the sentences that were read. The overall results indicated a


289

variation across the conditions, in that, cued conditions were perceived to be most variable when compared to uncued conditions for the emotions happy and sad. The results of the quantitative analysis revealed increased mean values for most of the vocal parameters in cued conditions across all emotions. Further, the qualitative analysis also revealed that the judges perceived the sentences in cued conditions to be most variable and more natural than the uncued conditions. This indicates that variability in acoustic parameters was noticed both quantitatively as well as qualitatively across conditions (uncued and cued).

Conclusions

The observed outcome of the study was that there was a definite effect of cognitive cueing on the vocal variability. Gender difference was also observed in the subjects owing to the known fact that females are more expressive than males. The quantitative analysis showed changes with the emotions considered in the study. Happy emotion elicited a faster rate, had a higher F0 and a wider F0 range with upward inflections than neutral emotion. Comparing sad emotion with neutral emotion, the average F0 and F0 range was slightly lower with slower speech rate. The qualitative analysis also revealed that the judges perceived the sentences in cued conditions to be most variable and more natural than the uncued conditions. This indicates that variability in acoustic parameters was noticed both quantitatively as well as qualitatively across conditions (uncued and cued). Hence, it could be concluded that cognitive cueing brought about changes in vocal attributes which could be quantified and perceived. This strengthens the view that cognitive cues could help in achieving greater variations in voice and speech thereby, making the speech more natural. This study thus implies the usefulness of cognitive cues in obtaining desired changes in vocal parameters and thereby enhancing performance in professional voice users. The usefulness can further be extended in improving naturalness of speech (post therapy) in clients with speech disorder can also be studied.

References

Andrews, M., Shrivastav, R., & Yamaguchi, H. (2000). The role of cognitive cueing in eliciting vocal variability. Journal of Voice, 14(4), 494-501.

Barrett, L. F., & Niedenthal, P. M. (2004). Valence focus and the perception of facial affect. Emotion, 4(3), 266–274.

Bohnenkamp, T., Andrews, M., Shrivastav, R., & Summers, A. (2002). Changes in children‟s voices: the effect of cognitive cues. Journal of Voice, 16 (4), 430-443.

Cowan, M. (1936). Pitch and intensity characteristics of stage speech. Arch. Speech, Suppl. To December issue.

Davitz, J. R. (Eds.) (1964). A review of research concerned with facial and vocal expressions of emotion. The communication of emotional meaning. New York: McGraw- Hill.

Fonagy, I. (1981). Emotions, voice and music. In J. Sundberg, (Eds.), Research aspects on singing. Royal Swedish Academy of Music. 33, 51-79.

Fónagy, I., & Magdics, K. (1963). Emotional patterns in Intonation and Music. Z. Phonet. Sprachwiss. Kommunikationsforsch, 16, 293-326.

Gross, J. J., John, O. P., & Richards, J. M. (2000). The dissociation of emotion expression from emotion experience: A personality perspective. Personality and Social Psychology Bulletin, 26, 712-726.

Johnstone, T., & Scherer, K. R. (2000). Vocal communication of emotion. In M. Lewis, Haviland-Jones J. M. (Eds.), Handbook of emotions (pp. 220–235). New York: Guilford Press.

Murray, I. R., & Arnott, J. L. (1993). Towards the simulation of emotion in synthetic speech: A review of literature on human vocal emotion. Journal of Acoustical Society of America, 93 (2), 1097-1108.

Piaget, J. (1952). The child‟s conception of number. New York: Basic Books.

Öster, A., & Risberg, A. (1986). The identification of the mood of a speaker by hearing impaired listeners. Speech transmission Lab, Quarterly Progress. Stat. Rep. 4/1986, 79- 90.

Risberg, A. (1986). The identification of the mood of a speaker by hearing impaired listeners. Speech transmission Lab, Quarterly Progress. Stat. Rep. 4/1986, 79- 90.

Sanchez, J. A., Kirschning, I., Palacio, J. C., & Ostrovskaya, Y. (2005) Toward mood-oriented interfaces for synchronous interaction. CLIHC‟05, 23-26, Cuernavaca, Mexico.

Scherer, K. R. (1981). Speech and emotional states. In Darby, J. K. (Ed.) Speech evaluation in psychiatry. New York: Grune and Stratton.

Scherer, K. R. (1986). Vocal affect expression: A review and model for future research. Psychological Bulletin, 99, 143 -165.

Scherer, K., & Ceshi, G. (2000). Studying affective communication in the airport: The case of lost baggage claims. Personality and Social Psychology Bulletin, 26 (3).

Skinner, E. R. (1935). A calibrated recording and analysis of pitch, force and quality of vocal tones expressing happiness and sadness. Speech Monograph, 2, 81-137.

Stibbard, R. M. (2001). Vocal expression of emotions in non-laboratory speech: An investigation of the reading/leeds emotion in speech Project Annotation Data. Unpublished PhD thesis. University of Reading, UK.

Voerman, M. S., Langeveld, A. P. M., & Van Rossum, M. A. (2009). The Journal of Laryngology & Otology, 123, 528-534.

Williams, C. E., & Steven, K. N. (1972). Emotions and speech. Journal of Acoustical Society of America, 52 (4), 1238-1250.


290

APPENDIX

Sentence 1: This is a pen (Neutral emotion).

Cue: Imagine you are teaching a child diagnosed with mental retardation. You are teaching him the concept of things we write with. You begin to teach him the concept of the pen. How would you teach him?

Sentence 2: The tests are in the cupboard (Neutral emotion).

Cue: Imagine you performed a test on a case which is not very familiar and had still to finish. Your friend takes up the case the next day but she too does not know much about the test and asks you where they are kept. What would you tell her?

Sentence 3: I got the highest IA in class (Happy emotion).

Cue: Imagine you could not study for well for a difficult paper. But to your amazement you stand first in scores you obtained for IA. You never expected it. How would you feel?

Sentence 4: The director announced a holiday today (Happy emotion).

Cue: Imagine you are very tired after the annual day as you have been working very hard. But you also know that you haven‟t studied for a test the next day. But to your surprise you wake up to hear that classes are suspended. How would you feel?

Sentence 5: He cannot walk anymore (Sad emotion).

Cue: Imagine you have a cousin you like so much. One fine day, you receive a call saying he met with an accident. You are unable to meet him but a few months later you see him in your institute unable to walk and run as he used to. How would you feel?

Sentence 6: The results are out and I failed (Sad emotion)

Cue: Imagine you worked very hard for your exams. But despite the effort you lose a paper. All your friends passed and are joyful. How would you feel?

Articulatory acquisition in Telugu speaking children

291

Articulatory Acquisition in Typically Developing Telugu Speaking Children: 2-3 Years

Usha Rani K. & N. Sreedevi

Abstract

The aim of the investigation was to obtain articulatory norms for Telugu Test of Articulation and Discrimination (TTAD; Padmaja, 1988) in the age range of 2-3 years and to establish the ages at which 75% and 90% of the children produce the phonemes correctly. TTAD was administered to 120 (60 males and 60 females) children in the age range of 2-3 years, subdivided into (2 - 2.3, 2.3 - 2.6, 2.6 - 2.9 & 2.9 - 3 years). The responses were audio recorded and transcribed using IPA transcription and statistical analysis indicated an overall difference across age groups and gender. The age group 2-2.3 years was significantly different from the older age groups

for their articulatory scores. All the vowels, nasals, lateral /l/ and voiceless plosives (/p/, /k/, /tʒ /) were acquired by 90% of the children by the age of 2-2.3 years. The dipthong /ai/, voiced plosives (/b/, /g/, /dӁ/) and fricative continuants (/j/, /v/) acquired 90% criteria by the age of 2.3-2.6 years. By the age of 2.6-2.9 years the retroflex

stops (/Y/, /ૃ/), aspirated /bh/ the glottal /h/, affricate /tȉ/ reached 90% criteria. The fricative /s/ acquired 90%

criteria by 2.9-3 years, whereas /ȉ/, rest of the aspirated stops (/ph/, /dh/, /kh/ and /gh/), affricate /dȘ/ and retroflex and alveolar flap (/r/, /┛/) were acquired with only 75% accuracy and did not reach 90% criteria even by the age of 2.9-3 years. None of the clusters tested reached even 75% criteria by 3 years of age. Key words: articulatory acquisition, Telugu, place, manner, initial/ medial/ final position

he 1order of acquisition of sounds and age norms pertains to the area of phonology. The availability of normative data is essential to

clinical assessment of child articulation and phonology, without these norms misdiagnosis and unnecessary or inappropriate treatment is likely. Many investigators have studied the acquisition of phonology and gave the pattern that takes place in children at different age levels. Over the years the acquisition of phonemes in the speech of young English-speaking children has received considerable attention by speech-language pathologists (Wellman, Case, Mengert & Bradbury, 1931; Poole, 1934; Menyuk, 1966; Sander, 1972; Arlt & Goodban, 1976; Prather, Hedrick & Kern, 1975; Fudala & Reynolds, 1986; Smith, 1990; Dodd, Holm, Hua & Crosbie, 2003). In investigating speech sound acquisition two basic approaches have emerged. The classical approach provides developmental ages at which specific sounds are mastered by a significant percentage of children selected from a normal population. The other is distinctive feature approach proposed by Menyuk, (1966) which attempts to analyze the available data on correct usage of consonants during early articulatory development in terms of phonology or feature acquisition. Several studies in 1960s, 70s and 80s described normal phonological development in terms of distinctive features (Menyuk, 1966; Prather et al., 1975; Irwin & Wong, 1983) and according to them features which dominate children‟s correct

1 e-mail: [email protected]; 2 Lecturer in Speech Sciences, AIISH, Mysore, [email protected].

usage of consonant at the beginning stage of morpheme construction are + nasal, + grave, + voicing, +diffuse, + continuant and + stridency features. Amayreh and Dyson (1998) defined three types of age of acquisition: „Age of customary production‟ (i.e., at least 50% of children in an age group produce the sound correctly in at least two positions); „Age of acquisition‟ (i.e., at least 75% of children in an age group produce the sound correctly in all positions); and „Age of mastery‟ (i.e., at least 90% of children in an age group produce the sound correctly in all positions). Fudala and Reynolds (1986) in their normative data reported on a sample of 5,122 children that by the age of 1.6 to 1.11 years all the vowels and

diphthongs (i.e., /殴/, /】/, /i/, /æ/, /ǚ/, /Ȑ/, /a/, /I/, /i/, /u/, /ou/, /ai/, /ei/, and /au/) were produced by 97.7 to 100% of children whereas Stoel-Gammon and Otomo (1992) reported that the unrounded vowels in English /i/ and /a/ are mastered early and /I/ and // were less accurate by 2-2½ years.

Dodd, Holm, Hua and Crosbie (2003) reported phonological development of 684 monolingual British English-speaking children aged between 3 and 6.11 years with 6 months time interval. Each child was tested individually using two subtests from the Diagnostic Evaluation of Articulation and Phonology: DEAP (Dodd, Zhu, Crosbie, Holm & Ozanne, 2002) and child‟s responses were elicited using picture naming task and repetition where needed. Results indicated that

T


292

the sequence of sound acquisition reported in this study was consistent with previous studies: /m, n, p, b, d, w/ were among the first sounds acquired while/r, h, ð/ were the last sounds acquired considering the 90% criteria. Age levels for the development of speech sounds in English according to different authors are presented in Table 1.

A number of studies have been done to obtain

the norms in many of the Indian languages also (Kumudavalli, 1973; Tasneem Banu, 1977 and Prathima & Sreedevi, 2009 in Kannada, Usha, 1986 in Tamil; Padmaja, 1988 in Telugu; Arun Banik, 1988 in Bengali and Maya, 1990 in Malayalam). They concluded that the acquisition followed same pattern as in English but generally it was found that most of the sounds were acquired earlier in the Indian studies compared to western context.

Padmaja (1988) has developed a Test of

Articulation and Discrimination in Telugu and studied the acquisition of sounds in 160 Telugu speaking children in the age range of 2½-4½ years and found that all vowels and most of the consonants except aspirated stops, retroflex stops (/r/, /┑/, /居/, /危/) and fricatives (/s/, /Ȓ/) are acquired by 2½ years. The aspirated consonants and phonemes /s/, /r/ are

acquired by 3.3years and /居/, /Ȓ/ and clusters are acquired by 3½ years.

Maya (1990) studied 240 Malayalam speaking

children in the age range of 3-7 years. She reported that they acquire /s/, /r/, /l/, /f/, /tȒ/, /j/ and unaspirated stops at an earlier age of 3-3.6 years, while aspirated stops were acquired as late as 6-6.6 years. The criterion of acquisition considered was 75%.

Prathima and Sreedevi (2009) studied 120

Kannada speaking children in the age range of 3-4 years and found that all the vowels, diphthongs and most of the consonants were acquired by 90% of the children by 3-3.6 years except /d/, /r/ and /h/ in initial and medial positions. By 4 years of age /d/ was acquired in all the positions acquired in all the positions, /r/ was acquired in medial but not in initial positions by 90% of children and /h/ was not acquired even by 75% of children. Among the clusters only /ski/ was acquired by 90% of children by 4 years.

Table 1. Age levels for the speech sound development according to different authors in English

* Criteria for the sound to be considered as acquired. IP: Initial position, FP: final position Empty space indicates speech sounds not tested. F: Female, M: Male

Speech sounds

Wellman et.al.,

(1931) 75%*

Poole (1934) 100%*

Templin (1957) 75%*

Sander (1972) 75%*

Prather et.al., (1975) 75%*

Arlt & Goodban (1976) 75%*

Fudala & Reynolds (1986) 90%*

IP FP

Smith et.al., (1990) 90%*

M F

Dodd et al., (2003)

90%*

m 3 3½ 3 Below 2 2 3 2 2 3 3 3.0–3.5 n 3 4½ 3 Below 2 2 3 2 2 3 3 3.0–3.5 h 3 3½ 3 Below 2 2 3 1½ 3 3½ 3.0–3.5 p 4 3½ 3 Below 2 2 3 2 3 3 3 3.0–3.5 f 3 5½ 3 3 2-4 3 2½ 3 3½-5½ 3½-5½ 3.0–3.5

w 3 3½ 3 Below 2 2-8 3 1½ 3 3 3.6–3.11

b 3 3½ 4 Below 2 2-8 3 2 3 3 3 3.0–3.5 ┕ 4½ 3 2 2 3 3 7- 9 7-9 3.0–3.5 j 4 4½ 3½ 3 2-4 3 5 4 3.6–3.11 k 4 4½ 4 2 2-4 3 2½ 3 3½ 3½ g 4 4½ 4 2 2-4 3 2½ 3 4 3½ l 4 6½ 6 3 3-4 4 5 5 6-7 5-6 3.6–3.11 d 5 4½ 4 2 2-4 3 2½ 2½ 3½ 3 t 5 4½ 6 2 2-8 3 3 4 3½ 4 s 5 7½ 4½ 3 3 4 11 11 7-9 7-9 ð 6 6 ½ 7 4 7 4½ r 5 7½ 4 3 3-4 5 5½ 8 8 6.0–6.5

tȒ 5 4½ 4 3-8 4 5½ 5½ 7 6 4.0–4.5

v 5 6½ 6 4 4 3 ½ 5½ 5½ 5½ 5½ 3.0–3.5 z 5 7½ 7 4 4 4 11 11 7-9 7-9 4.0–4.5

Ȫ 6 6½ 7 6 4 4 7 6 4.0–4.5

7½ 6 5 4 5 6 6 8 6

Ȓ 6½ 4½ 4 3.8 4½ 5½ 5½ 7 6 5.0–5.5


293

More recently it has been the observation of speech-language pathologists that the children of this generation are acquiring sounds at an earlier age than their earlier counter parts and also the results of exploratory studies on articulation showed that children were acquiring proficiency in articulatory skills at an earlier age than would be expected from previously established norms.

In view of the fact that the previous norms were

obtained more than 20 years back in Telugu (Padmaja, 1988) and as in the recent years “Early intervention” for children (0-3 years old population) is becoming an increasingly common therapeutic option, there is an immediate need to test much younger children for determining the exact age of acquisition consistent with the performance of children of present trend and to update and revalidate the previously obtained norms for both clinical and research purposes. Thus this study was planned to obtain norms for the mastery of articulatory skills in typically developing 2-3 years old Telugu speaking urban children.

The objectives of the present study were 1) to administer the Test of articulation in Telugu (Padmaja, 1988) and to establish the ages at which 75% and 90% of the children produce the phonemes correctly for revising the norms established earlier, 2) to compare the articulatory skills across age and gender and 3) to compare the difference in the articulatory acquisition of phonemes in the initial and medial positions of the words and to compare the data obtained with that of the earlier reported studies in both English and Telugu.

Method

Participants: 120 typically developing 2-3 years Telugu speaking children were selected randomly from play schools and day care centres of different localities in Hyderabad city. The children were sub divided into four groups with an age-interval of three months (2-2.3, 2.4-2.6, 2.7-2.9 and 2.10-3 years) with each group comprising of 30 subjects including equal number of boys and girls.

Subjects were selected based on the following criteria that they should be native speakers of Telugu brought up in urban ambient environment and belonging to middle socio economic status. The subjects were also exposed to some amount of English language in the kindergarten set up and should not have any speech, language, hearing, cognitive or any other motor difficulties. All the children were informally screened for language ability and the oro-facial speech mechanism during data collection and any potential subject who did not appear to be within normal limits were excluded.

Test material: Telugu test of Articulation and Discrimination by Padmaja (1988) was used and the test comprises of two parts; Part I includes 80 items to test 10 vowels in all the positions and 34 consonants in the initial and medial positions as a consonant never occurs in final position in Telugu and few commonly used clusters in medial and initial positions and clusters occur frequently only in medial position in Telugu according to phonotactics rules of the language. A total of 4 clusters were tested, 2 clusters in medial position and 2 in initial position. Part II includes 20 minimal pairs for testing discrimination. In the present study part I was used to obtain the articulation norms. Test consists of 80 picturized target stimulus represented in line drawing. But in the present study coloured photographs of the same items collected on net and few photographs of real objects were used which were highly representative of the target stimuli. Each picture was designed to elicit the given target sound as a single phoneme or cluster at each position.

Data collection procedure: Each subject was tested individually in a quiet room for about 45 minutes to 1 hour. Once the rapport was established, the examiner presented the target pictures one after another. The target picture was presented one at a time on the DELL Inspiron (1525) laptop screen to the subjects and they were encouraged to name the picture spontaneously. If any subject failed to identify a target picture additional cues were presented. If the subject still did not name the picture examiner gave a forced-choice type question (“Is this a tomato or Apple”). In spite of the additional cues if subject didn‟t respond appropriately, the subject was asked to repeat the target word after the examiner.

The response obtained was audio recorded using CENIX (VR-P2170) digital recorder at High quality mode with an external mini microphone placed approximately 5-10 inches away from the subject‟s mouth.

Data scoring: The samples from all the 120 subjects were transcribed using broad and narrow standard IPA symbols to represent each sound with addition of diacritic marks. All the responses of each subject were analyzed sound-by-sound on a response sheet. Correct responses (CR), substitutions (S), omissions (O), distortions (D), additions (A) or any other type of articulatory deviation (Ao) was recorded on the response sheet. A score of „1‟ was assigned for consistent articulation of target sound, „0.75‟ for distortion, „0.5‟ for substitution error and „0‟ for omission error. The maximum score was 80. Finally the total score for each subject was calculated. The scoring sheet used for the study is presented in Appendix.


294

Inter-judge reliability: Inter-judge reliability for phoneme transcription was assessed by comparing the percentage of agreement between the transcriptions of the investigator and two experienced speech language pathologists on 10% of the samples randomly selected across the total 120 samples. The inter-judge agreement for phoneme transcription was 95% between them.

Data analysis: The data obtained was subjected to suitable analysis. The phonemes uttered correctly by each subject in both the positions was identified and a score of „one‟ was assigned to each correct item, the data was tabulated by age level and sex, with an entry for each phoneme in the recording sheet and scored. From the scores obtained for the correct articulatory production, mean, standard deviation and range of scores were calculated for each age group and independently for boys and girls in initial and medial positions. The phonemes which were produced correctly by 75% and 90% of the subjects were identified in each age group in the initial and medial positions separately were identified and the percentage tables were used to determine the ages of mastery and acquisition.


The results are presented and discussed under different main sections.

Age vs. Articulatory acquisition

Two-way ANOVA was carried out to find the significant difference in articulatory scores between different age groups (2-2.3 years, 2.3-2.6 years, 2.6-2.9 years and 2.9-3 years). The results revealed that there was an overall difference across age groups and gender. Hence further, one-way ANOVA was carried out to note the significant difference in articulatory scores between the four age groups. Table 2 and Figure 1 shows the mean articulation scores in the four age groups considered in the study.

Results indicated that in both boys and girls, the youngest age group considered i.e., 2-2.3 years group was found to be significantly different from the three older age groups for their articulatory scores. The significant difference was at 0.05 level of significance (p=0.00). The results also imply that in both genders the patterns of articulatory acquisition are similar that is the younger age group 2-2.3 years scored low on the articulation test compared to older age groups. The above results can be attributed to findings that the aspirated and retroflex stops, fricatives and affricates were not acquired by even 75% of the children by this age.

Diphthongs /ai/ and /ou/ and the semi vowels /j/, /v/ were acquired only by 75 % of the children by 2-2.3 years of age which were acquired with 90% accuracy by the older age groups (i.e., 2.3-2.6 years).

Comparing the scores across the three older age groups no statistically significant difference was evident. It was also observed that the standard deviation values decreased with age. This finding can be attributed to the wider variability in the production patterns of younger children due to ongoing neuromuscular maturation reflected in their articulatory skills which is very much a motor skill.

The above findings indicate the effects of articulatory complexity on sound acquisition. Also greater experience with speech production could be a factor contributing to reduced variability in the older groups.

Table 2. Mean articulation scores & standard deviation (SD) in different age group

Gender Age N Mean (SD)

Boys 2 -2.3 yrs 15 67.19 (5.87) 2.3-2.6 yrs 15 73.21 (3.64) 2.6-2.9 yrs 15 74.41 (2.82) 2.9-3 yrs 15 75.20 (2.97)

Total 60 72.50 (5.04)

Girls 2 -2.3 yrs 15 69.76 (4.85) 2.3-2.6 yrs 15 74.26 (2.47) 2.6-2.9 yrs 15 76.15 (2.33) 2.9-3 yrs 15 76.68 (2.24)

Total 60 74.21 (4.12)

Combined Scores

2 -2.3 yrs 30 68.47 (5.45) 2.3-2.6 yrs 30 73.74 (3.10) 2.6-2.9 yrs 30 75.28 (2.69) 2.9-3 yrs 30 75.94 2.69)

Gender vs. Articulatory articulation

As two-way ANOVA indicated overall gender differences in articulatory scores, independent T-test was carried out to check for any the significant difference across gender in each of the four age groups. Figure 1 shows the gender wise comparison of articulatory scores.

Results indicated that there was no significant difference across gender in each group separately. In the present study though females exhibited superior articulatory skills compared to males in all the age groups on overall comparison, the difference was not statistically significant. Relatively recent normative study of mastery of speech sound acquisition showed similar results with girls tending to acquire sounds earlier upto 6 years of age (Smith, Hand, Freilinger, Bernthal & Bird, 1990). Mc Cormack and Knighton (1996) reported that 2.5-year-old girls had more accurate phonological output than boys. Gender differences revealed in Templin‟s study which were also noted in the present study suggest that speech sound acquisition is an area in which gender differences favouring girls persist, atleast in the early years. Order of acquisition of sounds Vowel acquisition: All the vowels /a/, /a: /, /i/, /i: /, /e/, /e: /, /o/ and /o:/ were acquired by 100% of the


295

children by the age of 2.3 years except /u:/ and the dipthongs /ai/ met 90% criteria by the age of 2.3-2.6 years whereas /au/ met 90% criteria by the age of 2.6-2.9 years.

The finding of the present study that the long rounded vowel /u:/ is not acquired with 90% accuracy by 2-2.3 years and the vowels /i/, /e/ were produced with only 93.3% and 86.6% accuracy is in consonance with the study by Stoel-Gammon and Otomo (1992) where they concluded that the vowels /i:/ and /e/ tended to evidence high error rates compared to /a/ and /i/.

Figure 1. Mean articulation scores in different age groups for boys and girls.

Consonant acquisition: All the nasals were acquired by 90% of children by 2-2.3 years. The bilabial (/p/), velar (/k/), the dental (/tӁ/) voiceless plosives, lateral /l/ also met 90% criteria by the age of 2-2.3 years. The diphthong (/ou/), retroflex stops (/居/, /d/), aspirated stop (/bh/), fricative (/h/) and affricate (/tȒ/) reached only 75% criteria by the age of 2.3-2.6 years. The bilabial voiced plosives (/b/, /g/, /Ȫ/), fricative continuants (/j/, /v/) and retroflex geminated /ll/ met 90% criteria by the age of 2.6-2.9 years of age whereas rest of the aspirated stops (/ph/, /dh/, /kh/ and /gh/) were acquired by only 75% of the children by the age of 2.6-2.9 years. Table 3 presents the comparison of present results with other Indian studies which reached 90% accuracy by the age of 2.3-2.6years which were not acquired in the previous age group (aspirated stop /bh /, affricate /tȒ/ and glottal stop /h/).

The fricative /s/, affricate /tȒ/ in medial

position met 90% criteria by the age of 2.9-3 years whereas the fricative /Ȓ/, flap /r/ and retroflex /┛/ in medial position, the aspirated stops and affricate /Ȫ/ in both initial and medial positions were not acquired by 90% of the subjects by 2.9-3 years but they had reached 75% criteria earlier by 2.6-2.9 years. The results of the present study indicated that the aspirated stops were acquired 6-9 months earlier but only after modelling and repetition. Also the results

that aspirated stops are acquired after their unaspirated counterpart is in agreement with Dodd and So‟s (1995) findings on articulatory acquisition in Cantonese speaking children. There are significant discrepancies noted on comparison of the results of present study with that of Arun Banik (1988) in Bengali and Padmaja (1988) in Telugu.

The acquisition of the fricatives /s/, /Ȓ/ is 6 months earlier in the present study compared to Padmaja (1988) in Telugu and one year earlier compared to Arun Banik (1988) in Bengali, whereas retroflex /┛/ is one year earlier. Surprisingly in the present study the acquisition of /s/, /Ȓ/, /r/ is earlier compared to the recent report in Kannada (Prathima & Sreedevi, 2009) where in the mastery of these phonemes are seen above 3.6 years.

Consonant cluster acquisition: Another salient observation of the study is that none of the clusters tested reached 75% criteria by 3 years of age. Findings revealed that all the clusters tested were acquired with approximately 50% accuracy in the present investigation. In Padmaja (1988) study clusters were reported to be achieved by 3.6 years. However the ages of acquisition of different speech sounds in Telugu language are relatively earlier compared to other Indian studies and this may be attributed to the cultural differences. Acquisition based on place, voicing and manner features Place feature: In general the bilabials, dentals, labiodentals, velars except the aspirated plosives met 90% criteria by the age of 2-2.6 years whereas the glottal (/h/) reached 90% criteria by the age of 2.6-2.9 years in the present study. Palatal, retroflex and aspirated stops met only 75% criteria but not 90% by the age of 2.9-3 years. Voicing feature: In the present study, voiced plosives such as /b/ and /g/ were acquired later by the age of 2.3 - 2.6 years considering the 90% criteria while the voiceless cognates /p/ and /k/ were acquired earlier by 2 -2.3 years itself. The voiceless fricatives /s/, /h/ and affricate /tȒ/ met the 90% criteria by 2.7-2.9 years where as the voiceless fricative /Ȓ/, and voiced affricate /Ȫ/ reached only 75% criteria by 2.9-3 years. The voiced velar /g/ showed a gradual acquisition pattern. Velar /g/ was acquired with 75% of accuracy at 2-2.3 years and reached 90% accuracy at 2.3-2.6 years only in the medial position, later by 2.6-2.9 years /g/ met 90% criteria in initial position.

The findings indicate that the bilabials (/p/, /b/) and velar (/k/) were acquired 3 months earlier compared to Padmaja, (1988) and acquisition of plosives (/tӁ/, /dӁ/, /g/) indicate the same age of acquisition.


296

Table 3. Comparison of the present study with that of other Indian studies

„-„ indicates speech sounds not acquired. Empty space indicates speech sound not tested Manner feature Plosives: All the plosives in the present study (/p/, /b/, /k/, /g/, /tӁ/, /dӁ/) were acquired with 90% accuracy by the age of 2-2.6 years. The retroflex stops (/居/, /危/) met 75% criteria at 2.3-2.6 years and reached 90% criteria by 2.6-2.9 years in both boys and girls. The aspirated plosive / bh / was acquired with 90% accuracy by 2.6-2.9 years whereas the rest of aspirated stops /ph/, /dh/, /kh/ and /gh/ were acquired with only 75% accuracy by 2.6-2.9 years but they did not reach 90% criteria even by the age of 2.9-3 years. Fricatives: The alveolar fricative /s/ was acquired by 2.6-2.9 years of age by 75% of the children and reached 90% criteria by 2.9-3 years. The palatal voiceless fricative /Ȓ/ was acquired only by 75% of the children by 2.6- 2.9 years and did not reach 90% criteria even by 2.9-3 years. The glottal fricative /h/ met 75% criteria by 2.3-2.6 years and reached 90% criteria by 2.6-2.9 years. Figure 2 indicates the manner of speech sound acquisition by 90% of boys and girls.

Affricates: The voiceless palatal affricate /tȒ/ met 75% criteria by 2.6-2.9 years and reached 90% criteria by 2.9-3 years in boys, whereas in girls it met 75% criteria at 2.3-2.6 years and reached 90% criteria by 2.6-2.9 years. The voiced palatal affricate /Ȫ/ was acquired with 75% accuracy at 2.6-2.9 years in both boys and girls and did not reach 90% accuracy even by 3 years of age.

Nasals: The nasals /m/ and /n/ reached 90% criteria by 2-2.3 years in both boys and girls which was in consonance with Fudala and Reynolds (1986) and was earlier by one year when compared to Smit et al., (1990) & Dodd et al., (2003). Their acquisition is three months earlier compared to Padmaja (1988).

Laterals: Both the alveolar /l/ and voiced retroflex laterals /┑/ in medial position were acquired with 90% accuracy at 2-2.3 years in boys and girls. The alveolar /l/ in initial position reached 75% criteria at 2-2.3 years and later met 90% of accuracy by 2.3-2.6 years. On comparison with Padmaja (1988), the retroflex /┑/ is acquired almost one year earlier in the present study.

Flaps: The voiced alveolar flap /r/ and the retroflex flap /┛/ was acquired with 75% accuracy only in the medial position by 2.6-2.9 years in both boys and girls but were not acquired with 90% accuracy even by three years of age, however it was nearly one year earlier compared to Padmaja (1988). It was not acquired by 75% accuracy in initial position even by 3 years.

Fricative continuants: /v/ and /j/ reached 75% criteria by 2-2.3 years and reached 90% of accuracy by the age of 2.3-2.6 years in both boys and girls that is three months earlier when compared to Padmaja (1988).

Speech sounds

Tasneem Banu, 1977 (Kannada)

75%

Usha, 1986

(Tamil) 75%

Padmaja, 1988 (Telugu)

75%

Arun Banik, 1988

(Bengali) 90%

Maya, 1990 (Malayalam)

75%

Prathima & Sreedevi, 2009

(Kannada) 90%

Present study, 2010 (Telugu)

75% 90%

M 3 3 2.6 2.5 3-3.6 3-3.6 2-2.3 2-2.3 N 3 3 2.6 2.5 3-3.6 3-3.6 2-2.3 2-2.3 └ 2.5 3-3.6 P 3 3 2.6 2.5 3-3.6 3-3.6 2-2.3 2-2.3 F 2.9 3-3.6 2.6-2.9 - H 2.6 3 3-3.6 - 2.3-2.6 2.6-2.9 K 3 3 2.6 2.7 3-3.6 3-3.6 2-2.3 2-2.3 B 3 3 2.6 2.5 3-3.6 3-3.6 2-2.3 2.3-2.6 D 3.6 3 2.6 3 3-3.6 3-3.6 2.3-2.6 2.6-2.9 g 3 3 2.6 3 3-3.6 3-3.6 2.3-2.6 2.6-2.9 R 4.6 3.9 4 3.7-4 - 2.6-2.9 - S 3 3 3.3 3.6-4 3-3.6 2.6-2.9 2.9-3

Ȓ 5.1 6 3.6 3 5-5.6 3.6-4 2.6-2.9 -

tȒ 3.7 3 2.6 3 3-3.6 3-3.6 2.3-2.6 2.6-2.9 t 3 2.6 3 3-3.6 3-3.6 2.3-2.6 2.7-2.9 v - 3 2.6 3-3.6 3-3.6 2-2.3 2.3-2.6 l 3 3 2.6 3 3-3.6 3-3.6 2-2.3 2.3-2.6 Ȫ 2.6 3 6-6.6 2.6-2.9 -

Z 4

j 3 3 2.6 3 3-3.6 3-3.6 2-2.3 2.3-2.6


297

Figure 3. Age of mastery of clusters by both boys and girls.

Indicates sounds acquired by 90% of the children. Indicates sounds not acquired by 75% of the children.

Clusters: In the present study, 4 clusters were tested, 2 clusters (/kȒa/, /Ȓra/) were tested in medial position and 2 clusters (/bl/, /sk/) in initial position. None of the clusters were acquired by 75% accuracy even by three years of age both in boys and girls. All the clusters crossed approximately 50% of criteria. Figure 3 indicates the age of mastery of clusters in both boys and girls.

Speech sound acquisition vs. word position All the consonants reached 90% criteria in

both initial and medial positions simultaneously. However there were few discrepancies noted in one vowel and few consonant acquisition patterns. The long vowel /u:/ reached 90% criteria earlier in the medial position by 2.3-2.6 years and later in initial position in boys. Similarly the dipthong /ai/ reached 90% criteria earlier in medial position by 2.3-2.6 years whereas in initial position by 2.6-2.9 years. The plosive /g/ was also noted to be acquired earlier by 90% accuracy in medial position in boys and aspirated /dh/ reached 75% criteria earlier in medial position by 2.3-2.6 years in girls. The aspirated /kh/, fricative /Ȓ/ met 75% criteria earlier in medial position by 2.7-2.9 years and were not acquired by 75% of children in the initial position even by the age of 3 years in both boys and girls. The flap sound /r/ was acquired first in the medial position with 90% of accuracy in both boys and girls by the age of 2.7-2.9 years and did not reach 90% criteria in initial position even by the age of 3 years.

Considering the results there is a clear trend seen that few of the difficult consonants met 90% criteria earlier in medial position when compared to initial position. Such positional variations in acquisition are reported in other studies also which are in agreement with the present study. These positional wise cues may be used in the therapeutic process for targeting the specific consonants. Based on the results obtained, mean articulation scores expected for typically developing children in the age range of 2-3 years in boys and girls are depicted in the Table 4. Maximum score is 80.

Table 4. Articulation scores expected for typically developing children in the age range of 2-3 years

Conclusions

The aim of the present study was to obtain norms for the mastery of articulatory skills in typically developing 2-3 years old Telugu speaking urban children using Telugu test of Articulation and Discrimination (TTAD). The results revealed that all the nasal and vowels tested were acquired by the age of 2.3 years itself except /u:/ and dipthongs /ai/, /au/. The children acquired bilabials, labiodentals, dentals and velars earlier by 2- 2.6 years when compared to alveolars, palatal, retroflex and glottal sounds which were acquired by 2.6-3 years of age and more. The unaspirated and/or voiceless sounds were acquired earlier compared to the aspirated and/or voiced sounds. None of the clusters tested were mastered by 3 years of age. The prominent observation of the present investigation was that, present day children acquire most of the sounds at a younger age itself compared to the earlier reports in Telugu. This early acquisition may be attributed to a change in norms, over years due to effects of globalization such as Television and computers, greater exposure to speech environment or the effects of bilingualism and increased educational opportunities at younger age itself through improved preschool programmes, the technological and modified methods of teaching focusing more on multilingual education, differences

Position 2-2.3 years

2.3-2.6 years

2.6-2.9 years

2.9-3 years

Initial clusters /ble/ /ske/

Medial clusters

/kȒa/ /aȒr /

Age Gender Mean scores expected for typically developing

children 2 -2.3 years Boys 67.19 + 5.8 2 -2.3 years Girls 69.76 + 4.8 2.3-2.6 years Boys 73.21 + 3.6 2.3-2.6 years Girls 74.26 + 2.4 2.6-2.9 years Boys 74.41+ 2.8 2.6-2.9 years Girls 76.15 + 2.3 2.9-3 years Boys 75.20 + 2.9 2.9-3 years Girls 76.68 + 2.2


298

Figure 2. Manner of speech sound acquisition by 75% / 90% of both boys and girls.

Note: The affricate /tȒ/ was acquired by 75% of children by 2.6-2.9 years and reached 90% by 2.9-3 years in boys, whereas in girls it was acquired by 75% of them at 2.3- 2.6 years and reached 90% by 2.6-2.9 years. The long vowel /u:/ was acquired later in boys and met 90% criteria by 2.3-2.6 years

Indicates sounds acquired by 90% of the children. Indicates sounds acquired by 75% of the children.

Indicates sounds acquired in medial position (75%) not in initial position

Manner of articulation 2-2.3 years

2.3-2.6 years

2.6-2.9 years

2.9-3 years

Vowels /a/ /a:/ /i/ /i:/ /u/ /u:/ /e/ /e:/ /o/ /o:/

Diphthongs /ai/ /ou/

Plosives /p/ /b/ /居/ /危/ /tӁ/ /dӁ/ /k/ /g/

/ph/ /bh/ /dh/ /gh/ /kh/

Affricates /tȒ/ /Ȫ/

Fricatives /s/ /Ȓ/ /h/

Nasals /m/ /n/

Fricative continuants /j/ /v/

Flap/laterals /r/ /l/ /┑/


299

in child rearing practices, and increased awareness of speech defects among parents. Hence it is recommended to use these revised norms while administering Telugu Test of Articulation and Discrimination by Padmaja (1988).

Acknowledgements

The authors would like to sincerely thank Dr. Vijayalakshmi Basavaraj, Director, All India Institute of Speech and Hearing Mysore, for granting permission to carry out this study. They thank Ms. Vasanthalakshmi for helping out with the statistics and to all the little angles who participated in the study. They also thank all those who have directly or indirectly contributed for the successful completion of the study.

References Arlt, P. B., & Goodban, M. T. (1976). A comparative study

of articulation acquisition as based on a study of 240 normal‟s, aged three to six. Language, Speech, and Hearing Services in Schools, 7, 173-180.

Arun Banik (1988). Articulation Test in Bengali. An unpublished master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing, University of Mysore, Mysore.

Amayreh, M., & Dyson, A. (1998). The acquisition of Arabic consonants. Journal of Speech, Language and Hearing Research, 41, 642-53.

Dodd, B., & So, K. H. (1995). The acquisition of phonology by Cantonese-speaking children. Journal of Child Language, 22, 473-95.

Dodd, B., Holm, A., Hua, Z., & Crosbie, S. (2003). Phonological development: a normative study of British English-speaking children. Clinical linguistics and Phonetics, 17, 617-643.

Fudala, J. B., & Reynolds, W. M. (1986). Arizona articulation proficiency scales (2nd Ed.). Los Angeles: Western psychological Press.


Maya (1990). An articulation test battery in Malayalam. An unpublished master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing, University of Mysore, Mysore.

Mc Cormack, P., & Knighton, T. (1996), Gender differences in the speech development of 2.5-year-old children. In P. McCormack and A. Russell (Eds.), Proceedings of the sixth Australian International Conference on Speech Science and Technology (Adelaide: The Australian Speech Sciences and Technology Association), 217– 222.

Menyuk, P. (1966). A role of distinctive features in children‟s acquisition of phonology. Journal of Speech, Language and Hearing Research, 11, 138-146.

Padmaja, B. (1988). Telugu Test of articulation and Discrimination (TTAD). Unpublished Master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing, University of Mysore, Mysore.

Poole, E. (1934). Genetic development of articulation of consonant sounds in speech. Elementary English Review, 11, 159-161.


Prathima, S., & Sreedevi, N (2009). Articulation Acquisition in Kannada Speaking Urban Children: 3-4 Years. An unpublished master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing, University of Mysore, Mysore.


Smith, A. B., Hand, L., Freilinger, J. J., Bernthal, J. E., & Bird, A. (1990). The Iowa articulation norms project and its Nebraska replication. Journal of Speech and Hearing disorders, 55, 779-796.

Smit, A. B. (1993). Phonologic error distribution in the Iowa-Nebraska articulation norms project: Consonant singletons. Journal of Speech, Language and Hearing Research, 36, 533-547.

Stoel-Gammon, C., & Otomo, K. (1992). The acquisition of unrounded vowels in English. Journal of Speech and Hearing Research, 35, 604-616.

Thirumalai (1972). Acquisition of phonology of a child. Unpublished master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing, University of Mysore, Mysore.

Usha, D. (1986). Tamil articulation test. An unpublished master‟s dissertation submitted in part-fulfillment for the master‟s degree in Speech and Hearing, University of Mysore, Mysore.


Zhu, H. (2002). Phonological development in specific contexts. Clevedon, UK. Multilingual matter.


300

Appendix (Sample of the scoring sheet)

Name of the child: Age/Gender: School: Sl.No.

Position

Phoneme

Check word

Errors noted

Score

CR S O D A Ao

1 Initial a annamu 2 Medial a palaka 3 Final a kadava 4 Initial a : a : ku 5 Medial a : ka : lu 6 Initial I illu 7 Medial i mirapaka : ya 8 Final i go : li 9 Initial i : i : ka 10 Medial i : mi : sam 11 Initial u ullipa : ya 12 Medial u pustakamu 13 Final u tȒepu

14 Initial u : u : yala 15 Medial u : su : di 16 Initial r ruȒhi

17 Initial e eluka 18 Medial e meda 19 Initial e : e:nugu 20 Medial e : me : ka 21 Initial ai aisu 22 Medial ai railu 23 Final ai enabai 24 Initial o okati 25 Medial o kobbarika : ja 26 Initial o : o : da 27 Medial o : no : ru 28 Medial ou gounu 29 Initial k ka : lu 30 Medial k me : ka 31 Initial kh khadgamu 32 Medial kh mu khamu 33 Initial g gaȪelu

34 Medial g muggu 35 Initial gh ghanta 36 Initial tȒ tȒa : ku 37 Medial tȒ mantȒamu 38 Initial Ȫ Ȫada

39 Medial Ȫ gaȪelu


301

40 Initial Y Yama :Ya 41 Medial Y gariYa 42 Initial 危危abbu 43 Medial 危 ga危危amu 44 Medial n ra : ni 45 Initial tӁ tӁa : du 46 Medial tӁ ka tӁi 47 Initial dӁ dӁuvvena 48 Final dӁ ge :dӁe 49 Initial dh dhanasu 50 Medial dh ra dham 51 Initial n no : ru 52 Medial n kannu 53 Initial p palaka 54 Medial p kappu 55 Initial ph phalam 56 Initial b bendakaji 57 Medial b kobbarika : ji 58 Initial bh bhu : mi 59 Initial m mantȒmu 60 Medial m tȒi : ma 61 Initial j jamudu 62 Medial j pojji 63 Initial r ro : lu 64 Medial r ka : ru 65 Initial l langa 66 Medial l palaka 67 Initial v vankaja 68 Medial v kovvathi 69 Final r karra 70 Medial kȒ rickȒa 71 Initial s su : di 72 Final s glasu 73 Initial h hamsa 74 Initial Ȓ Ȓankam 75 Medial Ȓ a:Ȓramam 76 Final l kallu 77 Medial kȒa rickȒa 78 Initial bl ble : du 79 Medial Ȓra aȒramam 80 Initial sk ske : lu

Correct responses (CR-score 1), substitutions (S-score 1/2) indicating the substituted phoneme in the respective block, omissions (O- score 0), distortions (D-score 3/4), additions (A- score 0) any other type of articulatory deviation (Ao)


302

Linguistic and Metalinguistic Skil ls in Children with Stutteri ng Yashaswini R.1 & Y. V. Geetha2

Abstract

Since the age of onset of stuttering in most children is at around three years, speculations regarding the cause of stuttering are many. There has been limited attention in stuttering research on metalinguistic awareness. The present study aimed at comparing linguistic and metalinguistic abilities in Children with Stuttering (CWS) and Children with no stuttering (CWNS). The study included 30 CWS and a matched group of 28 CWNS. Each child was evaluated for linguistic and metalinguistic abilities using LPT, metaphonological skills task of RAP-K. These children were again divided into 4 age groups; 8-9year, 9-10years, 10- 11years and 11-12years. In Phonology section CWS in the age group of 10-11 years performed significantly poorer than CWNS. On semantics section there was no statistically significant difference between the performance of CWS and CWNS. On syntactic judgment task, the performance of 8-9 years age group CWS was significantly poorer. On metaphonological skills 8-9 years age group CWS performed poorer on phoneme stripping; 11-12 years CWS performed poorer on syllable oddity. However, when the age effects were eliminated statistically and subjected to further analysis, it revealed that CWS performed significantly poorer on all the metaphonological tasks, except for rhyme recognition and phoneme oddity. Thus it can be concluded that CWS have significantly poorer performance in higher language abili ties li ke syntactic judgment and metaphonological skills. However, these results must be cautiously considered due to the limited sample size considered. Key words: children with stuttering, children with no stuttering, linguistic skil ls, metalinguistic skil ls

tuttering 1is probably the best known and most researched speech disorder, but perhaps the most difficult to define, to explain and to

treat. Research studies indicate that the children develop stuttering at around the age of three years and most recover as they grow. Since the age of onset of stuttering in most children is at around three years, researchers have speculated and enumerated observations explaining that there could be a possible link between stuttering and early difficulty in language formulation (Bloodstein, 2006). One more important speculation about the cause of stuttering reported is that of motoric deficit (Peters & Starkweather, 1990). A combination of linguistic and motoric deficits due to the imbalance between linguistic and motoric development is also speculated (Peters & Starkweather, 1990). In addition, research focused on motor performance in the face of increased linguistic demand has suggested strong bi-directional language/motor interactions in Individuals with Stuttering (IWS) (Kleinow & Smith, 2000).

Interest in language and linguistic variabilit y in Children with Stuttering (CWS) compared to Children With No Stuttering (CWNS) has also stemmed from the observations that most of the onset of stuttering is during the language development period. Nearly 30-40% of CWS exhibit delay in their phonological and language development; there are many instances of children enrolled for speech therapy for delayed speech and language developing stuttering. In addition, the fact


that linguistic variability plays a role in the moments of stuttering has attracted wider research.

These reports attracted researchers to

investigate the language skills in younger children. Research findings from descriptive studies on speech and language abilities of CWS have been less consistent. On the one hand, some literature reviews have suggested that CWS may have less developed phonology, vocabulary, or overall language abilities than their normally fluent peers (Anderson & Conture, 2000). On the other hand, some empirical studies have found no evidence to suggest that speech or language abilities of CWS are less robust than those of CWNS. Howell, Davis and Au-Yeung, (2003) reported that CWS and CWNS (aged 2-10 years) performed similarly on Reception of Syntax Test, a measure of syntactic development. Silverman and Ratner (2002) failed to find significant group differences between children who do and do not stutter on measures of receptive vocabulary. In addition, some studies have reported that CWS may have above average expressive language abilities relative to their developmental expectations (Watkins, Yairi, & Ambrose, 1999).

However, to date there has been limited

attention in stuttering research on one area of language competence: metalinguistic awareness. Metalinguistic awareness has been defined as the “ability to reflect upon and manipulate the structural features of spoken language, treating language itself as an object of thought, as opposed to simply using language system to comprehend and produce sentences” (Tunmer & Herriman, 1984). In an attempt to answer the question whether the

S

Linguistic and metalinguistic skills in children with stuttering

303

performance of CWS and CWNS on phonological and grammatical awareness and metalinguistic tasks differ significantly or not, Bajaj, Hodson and Aikins (2004) conducted a study. The results of the study indicated no signif icant differences between CWS and CWNS on metaphonological skill tasks. In the grammaticali ty judgment task, CWS scored significantly lower than that of CWNS. Research in this area is scanty and needs to be probed further.

The rationale for examining metalinguistic

skills is drawn from aspects of relationship between stuttering, language abili ties, and phonological skil ls of CWS. The fact that performance on global measures of language, have not clearly distinguished CWS from CWNS (Ratner, 1995), examination of group performance in specific domains of language competence, such as metalinguistic awareness, may extend the scope of inquiry into the stuttering- language connection into new directions. Second, results of several studies have indicated that many CWS have concomitant articulation/ phonological disorders (Blood & Seider, 1981). Paralleling such findings, it has been documented that children who have phonological disorders demonstrate lower levels of phonological awareness, it is reasonable to think whether CWS, at least a subgroup of them, have atypical levels of phonological awareness compared to their fluent peers or not. The research in this area is limited and hence there is a need to look into this dimension in CWS compared to their fluent counterparts. Considering this, the current study was taken up with the following objectives: 1) To compare linguistic abili ties in CWS and CWNS; 2) To compare the performance of CWS and CWNS on metalinguistic tasks.

Method

Participants: The s t u d y included 30 CWS and a matched group of 28 CWNS. The CWS group had to satisfy the following inclusionary criteria to be considered for the study: 1) Age range of 8-12 years; 2) Kannada as their native language; 3) No complaints of any peripheral sensory impairment (hearing or visual); 4) No problems in the general intellectual and motoric abili ties. These two groups of participants were again classified into four different groups each based on the age range, which included: i) 8-9 years; ii) 9-10 years; iii) 10-11 years and iv) 11 -12 years. Materials: The test materials included in the study were: A checklist developed to gather information with respect to socio-demographic data, native language, medium of instruction at school, number

of languages to which the child was exposed, family, birth and developmental history, general intellectual and peripheral sensory abili ties, onset, development, severity of stuttering and academic performance; Stuttering Severity Instrument for Children and Adults- III edition (SSI-3) (Riley, 1994); The reading passages in Kannada were taken from Kannada Articulation Test (KAT) (Babu, Ratna & Betegeri, 1972). This included the passages on “maIsu:rina kh．dda” and “bakka tale manushja mattu nona”; Linguistic Profile Test (LPT) (Karanth, 1980). This language test taps the subject‟s mastery of the phonological, syntactic and semantic features of the Kannada language. The phonology and semantic sections were considered for evaluating the language skills. The section on syntax, which taps the subject‟s knowledge of the core syntactic features of Kannada through a grammaticality judgment task, was considered for metalinguistic skills; For the picture description task, three pictures from LPT were taken; Test of Metaphonological skil ls from Reading Acquisition Profile in Kannada (RAP-K), Prema, (1997) (grade III to grade VII ). This section included in RAP-K have Rhyme Recognition (R.R), Syllable Stripping(S.S), Syllable Oddity (words) (S.O.W), Syllable Oddity (non-words) (S.O.N.W), Phoneme Stripping (PH.S), Phoneme Oddity (PH.O). SPSS software version 10 was used for data entry and analysis.


As mentioned, with the aim of investigating whether there is any significant difference between the performance of CWS and CWNS on linguistic and metalinguistic tasks, the CWS and CWNS were again divided into four age groups: 8-9 years, 9-10 years, 10-11 years and 11-12 years. The gender and mean age for these four age groups is as given in the Table 1. According to the data obtained through administration of the checkli st in the CWS group the following were observed. Table 2 summarizes the data obtained through the questionnaire for CWS.

The results are presented and discussed under

different sections and subsections: 1) Linguistic skills a. Phonology: This section of the linguistic skills included two subsections: i. Phonemic discrimination test: The children were supposed to discriminate between two words e.g.,/huvu-havu/ and point out to the correct pair of picture among four choices in the same sequence as the order of presentation of the items auditorily.


304

Table 1. Number, gender, mean age of CWS and CWNS in four age groups

Age Group (years)

CWS CWNS Total

Mean age (years)

Male Female Mean age (years)

Male Female

8-9 8.61 7 2 8.69 6 1 16 9-10 9.44 8 0 9.53 8 0 16

10-11 10.5 6 0 10.51 6 0 12 11-12 11.27 6 1 11.52 6 1 14 Total - 27 3 - 26 2 58

Table 2. Summary of the data obtained through the questionnaire for CWS

Particulars Age range (years) Responses from the questionnaire 8-9

(years) 9-10

(years) 10-11 (years)

11-12 (years)

Total subjects

Number of subjects (N) 9 8 6 7 30 Positi ve Family history of stuttering 3 (33.3%) 2 (25.0%) 2 (33.3%) 1 (14.3%) 8 (26.7%)

Delayed speech & language development

3 (33.3%) 2 (25.0%) 1 (16.7%) 2 (28.6%) 8 (26.7%)

Kannada as medium of instruction 2 (22.2%)

2 (25%)

2 (33.3%)

1 (14.3%)

7 (23.3%)

Below average academic performance 0(0%) 2 (25.0%) 2 (33.3%) 0 (0%) 4 (13.3%) Exposure to >2 languages 7 (77.8%) 6 (75.0%) 4 (66.7%) 6 (85.7%) 23 (76.7%) Abnormal articulation 0 (0%) 0 (0%) 1 (16.7%) 0 (0%) 1 (3.3%)

Table 3. Performance of CWS & CWNS on phonology section

Age range

(years)

CWS CWNS p

N Mean Standard deviation

N Mean Standard deviation

8-9 9 95.88 1.83 7 97.14 2.03 .180 9-10 8 94.62 4.17 8 97.25 2.12 .141 10-11 6 95.83 2.22 6 99.00 2.44 .030* 11-12 7 96.71 4.23 7 96.71 2.21 .518

N= No. of subjects, SD= standard deviation; *p<0.05 ii. Phonetic expression: The children had to repeat back the words orally presented. Table 3 presents the number of subjects (N), mean scores, standard deviation (SD) and p values on phonology section of CWS and CWNS. The mean scores show differences between CWS and CWNS, where the CWNS have better mean than CWS; but statistically significant difference (*p < 0.05) between CWS and CWNS was found only in 10-11 years age group, where CWS performed significantly poorer than CWNS. It was observed that in 10-11 years age range, out of six children in CWS group one child (16.7%) had delayed speech and language development and the other child (16.7%) had misarticulation of /r/ and /l/. These observations of CWS having history of delayed speech language development is in agreement with the previous findings by Andrews and Harris (1964) who found that the eighty children who stuttered tended to be

late in starting to talk compared to controls. Five of the children in their survey of seventy- six (6.6%) were reported to be late talkers (two turned out later to be persistent and three recovered). Their study also had a significantly higher frequency of history of abnormal articulation compared to the controls. So the possible reason behind the poorer performance of CWS in 10-11 years age group may be because of the persistence of delayed speech and language development (16.7%) and speech sound disorder (16.7%) in some children in that age group. b. Semantics: This section consisted of two subsections. i) Semantic discrimination: Here the children were asked to point out to the named colour, furniture and body part. ii) Semantic expression: In semantic expression again there are items on a) Naming, b) Lexical category, c) Synonym, d) Antonymy, e) Homonymy, f) Polar questions, g) Semantic anomaly, h)


305

Age range

CWS CWNS N Mean SD N Mean SD

8-9 9 82.56 8.22 7 89.64 6.14 0.056 9-10 8 87.31 9.69 8 82.56 29.88 0.401 10-11 6 90.17 6.84 6 84.08 7.58 0.147 11-12 7 91.29 6.06 7 92.86 3.98 0.749

Dependent Variable F (1,56) Sig. Phonology 5.614 0.021*

Syntaxax

5.507 0.023* Semantics 0.002 0.965

Paradigmatic relations, i) Syntagmatic relations, j) Semantic contiguity and k) Semantic similarity.

There was no statistically significant difference between the performance of CWS and CWNS on semantics section. Results of this section have been summarized in Table 4. This finding support several other studies, which compared the receptive vocabulary of CWS and their age matched peers and found no significant dif ferences. Will iams, Melrose and Woods (1969) studied school age children with stuttering, and found differences in receptive vocabulary between the two talker groups, whereas another study by Perozzi and Kunze (1969) of the same talker groups of similar age found no significant differences.

Since 26.7% of CWS group had history of

delayed speech and language development, it was expected that they score lower on the semantics section of the language test also, based on the review. But in spite of dif ferences in the mean values which show CWS performing poorer than CWNS on semantics, the statistical analysis did not show any significant differences. This is in disagreement with the past research findings (Reed, 1977; Byrd & Cooper, 1989; Murray & Ryan, 1992), which have suggested that children who stutter (CWS) may have less well- developed language skills than fluent children, and that such relative linguistic deficiencies may play a role in precipitating their disfluencies.

Table 4. Performance of CWS & CWNS on semantics section; Number of subjects (N), mean,

Standard Deviation (SD) & p values

The underlying reason for this could be either, as noted (Ratner, 1997), standardized language tests are primarily designed to identify frank language disabili ty for diagnostic and therapeutic purposes. Hence they are unlikely to provide the more precise discrimination between groups that may be required when either subtle depression of skills or weakness in a very specific domain of language exists. Trends reported in studies that have not found significant differences between groups on a variety of measures suggest that this may be the case (Ratner, 1995). Alternatively, it could also indicate that the group of children considered in the present study in spite of having delayed speech and language development 26.7% might have recovered from the language disadvantage. Thus, to conclude

the results on the language skills in CWS, it is noted that CWS in 10-11 years age group performed significantly poorer on the phonology section of the language task. There were no significant differences in the performance of CWS and CWNS on this task in any other age groups, even in the younger age groups. On semantics section, both CWS and CWNS performed equally well, showing no statistically significant differences. These results also indicate that there were no marked developmental trends observed in the language tasks in CWS in the sample considered. To further confirm this observation, further statistical analysis was undertaken. Table 5. Results of Kruskal Wallis test on language

measures

Language skills

CWNS CWS Chi-

Square df Asymp.

Sig. Chi-

Square df Asymp.

Sig. Phonology 4.53 3 0.21 2.79 3 0.426

Syntax 0.78 3 0.85 5.71 3 0.126 Semantics 7.85 3 >0.05 5.68 3 0.128 LPT total 2.66 3 0.45 6.26 3 0.100

Table 6. Summary of 1-way MANOVA on LPT

subsections

* p < 0.05

Kruskal wallis test was administered on the CWS and CWNS groups to see whether there exists significant differences in the performance across the age groups. Results indicated no statistically significant differences in both CWS group and CWNS across the four age groups. Thus the age effects on the two groups were statistically eliminated. Summary of this is presented in Table 5. After the elimination of the age effects, the CWS (N=30) and CWNS (N=28) were considered as two individual groups. Table 6 summarizes results obtained from one-way MANOVA. On these two groups of CWS and CWNS one-way MANOVA was applied. The results revealed that on phonology and syntax section, CWS had poorer performance which was statistically significant. On Semantics section there was no such differences noted. 2. Metalinguistic skills a. Syntactic judgment task: This task again had subsections on i) Morphophonemic structures ii) plural forms iii) tenses iv) PNG markers v) case markers vi) transitives, intransitives and causatives vii) sentence types viii) conjunctives, comparitives and quotatives ix) conditional clauses x) participal


306

Age range (years)

CWS CWNS p

N Mean SD N Mean SD

8-9 9 72.28 9.05 7 82.36 6.35 0.049*

9-10 8 79.31 12.25 8 84.63 4.02 0.429

10-11 6 81.75 8.52 6 84.75 7.44 0.629

11-12 7 83.64 13.37 7 85.21 4.08 0.898

constructions. The results on syntactic judgment are presented in Table 7. It showed that the performance of CWS on syntactic judgment task was significantly poorer than (* p < 0.05) than that of CWNS in 8-9 years age group. This result supports the previous study results by Bajaj, Hodson and Aitkins (2004). The age groups studied for grammaticality judgment skil ls were 5-7.6 years and 7.8- 8.10 years. In both these age groups CWS had significantly poorer performance than CWS.

However, these results cannot be compared with the results of other studies which have used other measures for evaluating the expressive syntactic abil i ties of CWS, due to the methodological variations. Because of the fact that the syntax section employed syntactic judgments, this section of the LPT was considered on metalinguistic measure. The age effects were eliminated and one way MANOVA was applied on CWS and CWNS. This revealed significant differences between the performance of CWS and CWNS. b. Metaphonological skills. This section comprised of the following subsections. i. Rhyme recognition: Here the children were supposed to say whether the pair of words presented to them auditoril y sounded similar or not. Both CWS and CWNS performed equally well on this task. This could be because as reported in the previous literature, rhyming skills is a simple skill, which is acquired at a very young age of 3years by the children (Bryant, Bradley, MacLean & Crossland, 1989). Since the CWS considered in the present study are from 8 years, this skill might have been acquired much earlier. The results are as tabulated in Table 8.

Table 7. Performance of CWS & CWNS on syntax section; Number of subjects (N), mean, Standard

Deviation (SD) & p values

* p < 0.05 ii. Syllable stripping: The child was supposed to remove the given syllable from the word. The statistical analysis showed no signif icant difference between the performance of CWS and CWNS of all age groups considered. Table 9 presents the relevant results.

iii. Syllable oddity (words): In testing for syllable oddity the child was instructed to concentrate on the sound aspect of the words; among the four words which does not match the group, had to be identified. Table 10 presents the results on this section of metaphonological skills. Statistically significant

Table 8. Performance of CWS & CWNS on rhyme

recognition; Number of subjects (N), mean, Standard Deviation (SD) & p values

Age Range (years)

CWS CWNS P

N

Mean

SD

N

Mean

SD 8-9 9 11.56 1.014 7 11.86 0.38 0.64 9-10 8 12.00 0.00 8 12.00 0.00 1.00 10-11 6 11.33 1.211 6 11.83 0.41 0.46 11-12 7 11.86 0.38 7 12.00 0.00 0.32 Total 30 11.70 0.79 28 11.93 0.26 -

Table 9. Performance of CWS & CWNS on Syllable stripping; Number of subjects (N), mean, Standard


Age range

(years)

CWS CWNS N Mean SD N Mean SD p

8-9 9 10.22 2.167 7 11.14 0.69 0.658 9-10 8 10.13 3.56 8 11.38 0.92 0.809 10-11 6 10.17 4.49 6 12.00 0.00 0.317 11-12 7 10.86 1.86 7 11.71 0.49 0.410 Total 30 10.33 2.940 28 11.54 0.69 -

dif ferences between CWS and CWNS were found in the age group of 11-12 years. The CWS performed poorer than the CWNS. iv. Syllable oddity (non-words): The child was instructed to listen to the sound aspect of the non-word among a set of four words and identify one which does not match the group. The results on this subsection are presented in Table11. v. Phoneme stripping: Here the child was supposed to remove phoneme indicated by the clinician from the given word. Only in the younger age group 8-9 years, the CWS performed significantly poorer than CWNS. Table 12 presents the results on this section. vi. Phoneme oddity: A set of four words were presented auditorily. The children were instructed to concentrate on the sound aspect of the words and were supposed to identify the word that did not belong to the set. Table 13 presents the results on this subsection of metaphonological skills. Only CWS in 9-10 age group showed better performance on this task than their fluent peers. The CWS and CWNS were divided into four age groups because, in normal children it has been reported that the acquisition of metaphonological skill occur in a


307

Age range

CWS CWNS p

N Mean SD

N Mean SD 8-9 9 7.00 1.12 7 8.43 1.90 0.085 9-10 8 8.63 2.77 8 9.00 1.41 0.707 10-11 6 8.17 2.99 6 9.00 1.09 0.616 11-12 7 7.00 2.45 7 9.57 1.27 0.044* Total 30 7.67 2.35 28 9.00 1.44 -

hierarchy, and syllable manipulation skills are acquired first in the developmental sequence than the phoneme manipulation skil ls. But the present results demonstrated no such patterns. This is evident in the results where CWS in 8-9 years age group there was significantly poorer performance on phoneme oddity task. Whereas the performance of CWS in 11-12 years age group was significantly poorer than the CWNS only on the syllable oddity task. In CWS in 8-9 age group, since they have poorer performance on phoneme oddity, phoneme stripping was also expected to be equally poorer, but no such thing was revealed in the results. In CWS of 11-12 years age group, since poorer performance on an easier task like syllable oddity was seen, poorer performance on other more diff icult tasks like the phoneme manipulation tasks were expected. Again, results showed no such differences.

To confirm whether CWS showed any developmental trends in acquisition of metaphonological skills or not, the data was subjected to further statistical analysis. As mentioned earlier, the same statistical procedures were employed. The age effects on the two groups were statistically eliminated using Kruskal Wallis test. After the el iminat ion of the age effects, the CWS (N=30) and CWNS (N=28) were considered as two individual groups.

Table 10. Performance of groups on Syllable oddity (words); Number of subjects (N), mean, Standard


* p< 0.05 Table 11. Performance of Groups on Syllable oddity

(NW Number of subjects (N), mean, Standard Deviation (SD) & p values.

Age range

(years)

CWS CWNS p N Mean SD N Mean SD

8-9 9 6.11 2.62 7 8.57 1.27 0.051 9-10 8 7.12 2.03 8 8.13 1.89 0.356 10-11 6 7.50 3.15 6 7.67 1.37 1.000 11-12 7 7.43 1.90 7 9.29 1.11 0.058 Total 30 6.97 2.39 28 8.43 1.50 -

On these two groups of CWS and CWNS

one- way MANOVA was applied. The results of the data after applying this statistical procedure revealed that, there were significant differences in the performances of CWS and CWNS on all the metaphonological skill tasks, except for the rhyme recognition and phoneme oddity task. The results are tabulated in Table 14 and Table 15.

Table 12. Performance of CWS & CWNS on Phoneme stripping; Number of subjects (N), mean,

Standard Deviation (SD) & p values

Age range

(years)

CWS CWNS p N Mean SD N Mean SD

8-9 9 4.56 2.83 7 8.29 2.43 0.016* 9-10 8 6.88 4.36 8 8.00 3.78 0.524 10-11 6 8.83 4.67 6 9.83 2.04 0.681 11-12 7 7.14 3.89 7 9.14 2.12 0.367 Total 30 6.63 4.01 28 8.75 2.70 -

* p < 0.05

Table 13. Performance of CWS & CWNS on Phoneme oddity; Number of subjects (N), mean,

Standard deviation (SD) & p values

Age range

(years)

CWS CWNS p

N Mean SD N Mean SD 8-9 9 5.11 1.62 7 6.85 2.61 0.217 9-10 8 6.25 2.49 8 6.86 1.46 0.831 10-11 6 7.00 2.76 6 6.50 3.02 0.936 11-12 7 6.43 2.15 7 8.14 2.12 0.173 Total 30 6.100 2.23 28 7.10 2.27 -

Table 14. Kruskal Wallis test results across age

groups on metaphonological tasks

Metalinguistic skil ls

CWNS CWS Chi- Square

df

Asymp . Sig.

Chi- Square

df

Asymp. Sig.

Syntactic judgment

0.782

3

0.854

5.713

3

0.126

Rhyme recognition

2.423

3

0.489

3.035

3

0.386

Syllable stripping

6.628

3

0.085

1.297

3

0.730

Syllable oddity (words)

2.115

3

0.549

2.763

3

0.430 Phoneme oddity

1.226

3

0.747

3.382

3

0.336

Phoneme stripping

2.656

3

0.448

5.106

3

0.164

Metaphonolo-gical skills total

3.011

3

0.390

1.565

3

0.667

Thus to conclude the results on the

metaphonological section; when CWS and CWNS in different age groups were compared, statistically


308

significant differences were found in only 8-9 years age group, where CWS performed poorer than CWNS. On Syllable oddity section, statistically significant differences were observed between CWS.

Table 15. Results of one-way MANOVA for metaphonological skills section

Tests of Between-Subjects Effects Dependent Variable F Sig. Rhyme recognition 2.102 0.153 Syllable stripping 4.448 0.039*

Syllable oddity (words) 6.655 0.013* Syllable oddity (non-words) 7.676 0.008*

Phoneme oddity 5.471 0.023* Phoneme stripping 2.903 0.094

* p< 0.05 and CWNS in 11-12 years age group. No significant differences were shown on any other subsections. These results suggested that no marked developmental trends were followed by both CWS on metaphonological skills. However, the overall MANOVA results revealed significantly poor performance in CWS on syntax and all the metaphonological tasks except for rhyme recognition and phoneme oddity. Thus these results show that CWS have problems in higher level language skil ls like the metalinguistic abil ities. The possible reasons why the CWS showed irregular results on metaphonological skil ls could be as follows. In literature it has been reported that children‟s ability to segment words into syllables was easier than into phonemes and showed a clear hierarchy in the performance of these language analysis tasks. At the age 4, nearly half of the children could segment words into syllables but none could segment it into phonemes. By the end of I grade, 90% could perform the syllable segmentation task while 70% succeeded in phoneme segmentation task. They postulated that regardless of instruction results indicated that a greater level of maturity was necessary to analyze words into phonemes than into syllables (Lieberman, Lieberman, Mattingly & Shankweiler, 1980). However, since children in 8-9 years age group showed significantly poorer performance on syllable oddity section, which has been found to be easier than the phoneme stripping and phoneme oddity section, it could be assumed that these CWS do not follow a particular developmental trend in the acquisition of the metaphonological skills and have deficits in this domain. These reasons hold good when we consider the overall performance of CWS, with eliminated age effects. Thus from the results obtained it can be concluded that CWS considered in the present study have significantly poorer performance on metalinguistic skills. However, these results must be considered with

caution because the number of subjects considered in the present study in limited.

Conclusions

The role of language has long been considered important in theories and research concerning stuttering (Ratner, 1997). One focus of inquiry has been whether those who stutter score more poorly on tests of linguistic abili ty than controls. However, there is no clear stand about the point (Nippold, 1990; Ratner, 1997). Though research has focused on evaluating various language domains in children with stuttering, this direction of research is mainly done on younger children very near to the onset of stuttering and not on older CWS. Thus in an attempt to throw light on the possible factors influencing the CWS to persist in stuttering during their school age the current study was designed. The aim of the study was to compare CWS and CWNS on linguistic tasks and to see whether there exists a difference in the performance on CWS on metalinguistic tasks.

In order to get a picture of the

developmental pattern of language and metalinguistic abil ities in CWS and CWNS the 8-12 age range was further divided into four groups with one-year interval between each group. Language and metalinguistic measures were subjected to Mann Whittney test in order to find if statistically significant differences existed between CWS and CWNS on the language and metalinguistic skills measures. The results revealed that in Phonology section 10-11 years age group CWS performed significantly poorer than CWNS. It was observed that totally 8 children (26.7%) in CWS group exhibited delayed speech and language development and one child (3.3%) had misarticulation of /r/ and /l/. On semantics section there was no statisticall y significant difference between the performance of CWS and CWNS on semantics section. On metalinguistic skil ls test, the results indicated that the performance of in 8-9 years age group CWS on syntactic judgment task was signif icantly poorer. On metaphonological skills, 8-9 years age group CWS performed poorer on phoneme stripping; 11-12 years CWS performed poorer on syllable oddity. When the age effects were eliminated statistically further analysis was done, CWS performed significantly poorer on all the metaphonological tasks, except for rhyme recognition and phoneme oddity. Thus it can be concluded that CWS have significantly poorer performance in higher language abil i ties like syntactic judgment and metaphonological skil ls. However, the results must be cautiously considered, due to the limited sample considered. Future studies needed with larger number of subjects in each age group. Further the linguistic and metalinguistic


309

aspects could be studied on different subgroups of CWS, with regard to age, gender, associated disorders, severity of stuttering, genetic factors etc. More sophisticated language measures like reaction time measures could be used to study the group differences with respect to language processing.

References Anderson, J. D., & Conture, E. G. (2000). Language

abilities of children who stutter: A preliminary study. Journal of Fluency Disorders, 25, 283-304.

Andrews, G., & Harris, M. (1964). The syndrome of stuttering, Clinics in Developmental medicine (No 17). Heinemann; London.

Babu, R. M. P., Rathna, N., & Bettageri. (1972). Tests of Articulation in Kannada. Journal of All India Institute of Speech and Hearing, 3, 7-19.

Bajaj, A., Hodson, B., & Aikins, M. S. (2004). Performance on phonological and grammatical awareness metalinguistic tasks by children who stutter and their fluent peers. Journal of Fluency Disorders, 29, 63-67.

Blood, G., & Seider, R. (1981). The concommitant problemsof young stutterers. Journal of Speech and Hearing Disorders, 46, 31-33.

Bloodstein, O. (2006). Some empirical observations about early stuttering: A possible link to language development. Journal of Communication Disorders, 39, 185-191.

Byrd, K., & Cooper, E. (1989). Expressive and receptive language skills in stuttering children. Journal of Fluency Disorders, 14, 121–126.

Howell, P., Davis, S., & Au- Yeung, J. (2003). Syntactic development in fluent children, children who stutter, and children who have English as an additional language. Child language teaching and therapy, 19, 311-337.

Karanth, P. (1980). L i n gu i s t i c Profile Test in Kannada. Journal of All India Institute of Speech and Hearing, 14,121–126.

Kleinow, J. & Smith, A. (2000). Influence of length and syntactic complexity on speech motor stability of fluent speech of adults who stutter. Journal of Speech Language and Hearing Research, 43, 548-559.

Lieberman, I. Y., Lieberman, A. M., Mattingly, I. G. & Shakweiler, D. (1980). Orthography and the beginning reader. Cited in Frost, R. Prelexical and postlexical strategies in reading. Evidence from a deep and shallow orthography, Status report on speech research, SR-113, Haskins Lab, 153-170.

Murray, H., & Reed, C. (1977). Language abilities of preschool stuttering children. Journal of Fluency Disorders, 2, 171–176.

Nippold, M. (1990). Concomitant speech and language disorders in stuttering children: A critique of the literature. Journal of Speech and Hearing Disorders, 55, 51–60.

Perozzi, J. A., & Kunze, L. H. (1969). Language abilities of stuttering children. Folia Phoniatrica, 55, 51-60.

Peters, H. F. M., & Starkweather, W. C. (1990). The interaction between speech motor coordination and language processes in the development of stuttering: Hypothesis and suggestions for research. Journal o f Fluency Disorders, 15, 115-125.

Prema, K. S. (1997). Reading Acquisition Profile in Kannada (RAP-K). Unpublished Doctoral Thesis, Submitted to the University of Mysore, Mysore.

Ratner, N. (1995). Language complexity and stuttering in children. Topics in Language Disorders, 15, 32-47.

Ratner, N. (1997). Stuttering: A psycholinguistic perspective. In R. Curlee & G. Siegel (Eds.), Nature and treatment of stuttering: New directions (2nd ed.) (pp. 99–127). Boston: Allyn & Bacon.

Riley, G. D. (1994). Stuttering Severity Instrument for Children and Adults –III edition. Austin T X: Pro-ed. Cited in Manning, W. H. (2001). Clinical Decision Making in Fluency Disorders. II edition. New Zealand: Singular Thomson Learning.

Ryan, B. P. (1992). Articulation, language, rate, and fluency characteristics of stuttering and nonstuttering children. Journal of Speech and Hearing Research, 35, 333–342.

Silverman, S., & Ratner, N. B. (2002). Measuring lexical diversity in children who stutter: Application of vocd. Journal of Fluency Disorders, 27, 289–304.

Tunmer, W., & Herriman, M. (1984). The development of metalinguistic awareness: A conceptual overview. Ci ted in A. Bajaj., B. Hodson., & M.S.Aikins. (2004). Performance on phonological and grammatical awareness metalinguistic tasks by children who stutter and their fluent peers. Journal of Fluency Disorders, 29,63-67.

Watkins, R. V., Yairi, E., & Ambrose, N. G. (1999). Early childhood stuttering III: initial status of expressive language abilities. Journal of Speech, Language and Hearing Research, 421, 1125-1135.

Will iams, D. E., Melrose, B. M., & Woods, C. L. (1969). The relationship between stuttering and academic achievement in children. Journal of Communication Disorders, 2, 87–98.


310

Comparison of Signed Expression of Indian Sign Language (ISL) Users and Markers of Signed English (SE)

Yashomathi 1 & R. Manjula 2

Abstract

In sign language, syntax is primarily conveyed through a combination of word order and non manual features. Less attention is given to the area of word order and grammatical markers such as plurals, tense, agent marker, possessives, adverbs, adjectives, superlatives, comparatives, participle opposites (Un-, im-, Dis-) which are not studied in Indian Sign Language. This study aimed to compare the sequence in which Subject, Object, Verb (SOV) signs occur in Indian Sign Language (ISL) with that of Signed English (SE) and to investigate whether there are any stand alone signs for sign markers in ISL like in SE. The features which distinguish sign markers in ISL were investigated and compared with the sign markers of SE. Fifteen hearing impaired adults, who were users of ISL aged 15-30 years participated in the study. The participants were presented with the sentences one at a time embedded with the sign markers and instructed to express signs for each sentence through ISL signs. The word order was found to be predominantly SVO followed by SO in ISL. Separate manual signs representing “grammatical markers” were found in ISL for grammatical categories such as opposites „Un‟, agent marker, possessives, and superlatives characterized by definite features such as handshape, orientation of the hand, location, movement and non manual markers or facial expressions for which either any feature or combination of features matched with the SE. Key words: word order, non manual marker, sign markers/ grammatical markers

n 1sign languages, syntax is primarily conveyed through a combination of manual and non manual features. There are different viewpoints

pertaining to the word order in sign language. Major research is carried out to investigate the word order in American Sign Language, British Sign Language, and other sign languages like Israeli Sign Language, Sign Languages of Netherlands etc., but very less is being researched in Indian Sign Language (ISL). Even though the word order is being addressed in ISL there is not enough research carried out regarding the signed word order in ISL except for the study by Pallavi and Manjula (2010) which partially addressed this issue.

The number of sign languages in India and Pakistan is unclear. Zeshan (2001) has argued that India and Pakistan have only one sign language, which she terms Indopakistani Sign Language (IPSL). ISL is used by approximately 2,680,000 hearing impaired in India (Gordon Jr., 2005). Vasistha, Woodward and Wilson (1978) quoted an estimate of 1,000,000 deaf adults and 500,000 deaf children using ISL or “Urban Indian Sign Language” as suggested by Jepson (1991b). Gordon Jr. (2005) observed that there are three main sign language families in ISL; Mumbai-Delhi sign language, Kolkata sign language and Bangalore-Madras sign language. ISL has its own morphology, phonology, syntax, and grammar (Vasishta et. al, 1978; Zeshan,

1e-mail: [email protected]; 2Professor of Speech Pathology, AIISH, Mysore, [email protected].

Vasishta & Sethna, 2004) and is not based on any spoken language, and can be described by means of linguistic analysis. As in other sign languages, ISL uses non manual signs (primarily facial) in parallel with manual signs (hand/arm) to indicate negations, questions, and suggestive phrases. ISL has also been found to share grammatical features with many other sign languages, including the use of space, and simultaneity and the five meaningful parameters, that is, Handshape Orientation, Location, Movement, and non manual features such as body position, head movement and facial Expression (HOLME).

In a significant finding, Vasishta et al., (1978) found that ISL does not share the features of European sign language families which includes French and Spanish but shares few features of American Sign Language (ASL). They found that in ISL, whenever there is a sentence containing a subject and a verb, the subject always preceded the verb. No other grammatical marking was evident. For sentences containing a subject, verb and object, 95% of the sentences had a subject-object-verb word order. Some of these sentences as well as the other 5% of the sentences relied on directionality in three-dimensional space (movement from subject to object) for expressing subject-object relationships.

An Indian study by Pallavi and Manjula (2010) investigating the word order seen in signed expression of ISL users, the influence of the verbal language order on the signed expression by ISL users were found to be majorly SOV type. The expression of subject, verb, object and indirect object relations gave different results for different cities. Although

I


Word order, Grammatical markers: ISL vs. SE

311

subject occupied the initial positions and verb occupied the final positions, there were considerable variations in the sign order beyond the general rule. The subject-verb-object-indirect object relation in the signing patterns of these different cities emphasized the role of incorporation of ISL.

Table 1. Word order seen in ISL across various cities

Negatives in Bombay, Bangalore, and

Calcutta were always placed post verbally at the end of sentences. This was also the preferred order in Delhi. 77% of the sentences had negatives post verbally at the end of the sentences while 23% of the sentences placed the negative directly before the verb. E.g. English: The man does not cry.

ISL: MAN CRY NOT.

Past tense in ISL is expressed by a past tense sign that is placed at the final position of the sentence. Vasishta et al., (1978) found that like in ASL, even in ISL, a time frame is set by the use of the past tense marker and no further reference is made until a shift in tense made.E.g. English: The man cried.

ISL: MAN CRY PAST.

Vasishta et al., (1978) states: “Adjectival modification provided the most striking example of a consistent ISL grammatical process observed in each of the four cities and yet which did not bear a close relation with the indigenous spoken language.” In over 97% of the sentences, the adjective occurred after the noun. However when the adjective, especially one that modified an object noun, an equally high percentage of adjectives preceded the noun. E.g. English: The good man looked at the woman.

ISL: MAN GOOD WOMAN LOOK.

Pronominal forms are used when a context has been established earlier. But no specific classifiers were observed to designate pronouns in ISL unlike the ASL. Instead, effective use of space was seen which eliminated the use of classifiers.

In ISL, Wh-questions are strictly clause-final (Sinha, 2007). ISL is SVO with asymmetry between matrix and embedded clauses (Sinha, 2007). ISL is a verb/predicate final language (Zeshan, 2003).

In ISL syntax, less research is carried out in the area of morphological inflection like regular and irregular plurals, regular and irregular past verbs, possessives, comparatives, adverb, adjective etc. An Indian study by Sinha (2007) reported that the morphology is complex; it exhibits both sequential as well as simultaneous affixation of its manual as well as nonmanual components (Sinha, 2007). In this regard, the only sign language system reported as having separate signs for grammatical markers (sign markers) is Signed English (SE) system, which includes 14 sign markers: Regular plurals, Irregular plurals, Agent marker, Possessives, Regular past verb, Irregular past verb, 3rd person singular, Verb from –ing, Participle, Adjective, Adverb, Comparatives, Superlatives, and Opposite „Un‟.

As indicated by Bornstein (1973) and Bornstein, Saulnier and Hamilton (1980), no separate sets of grammatical markers are available in any of the sign languages. Thus, this study aimed to address the issue of word order and grammatical markers in ISL in the context of the known sign marker of SE. This study is a preliminary attempt to understand the predominant type of word order in sentences embedded with 14 markers with the signed expressions incorporating the grammatical markers in ISL (which may or may not be equivalent to sign markers of SE). Hence there is a need to investigate the presence of grammatical markers in the signed expression of ISL users. The present study aimed at comparing emerging word order with respect to combination of Subject, Verb, Object for a predetermined set of sentences expressed in ISL which were embedded with 14 sign markers of SE with ISL. Another aim of the study was to investigate whether there were any separate manual signs in ISL for grammatical markers like the ones seen in „sign markers‟ of SE. If these signs were present, the study aimed to find the similarities and differences of these in comparison to the „sign markers‟ used in SE and if separate manual signs were absent in ISL to express grammatical markers, then the other signs used to express these were investigated.

Method

Subjects: 15 hearing impaired adults, 8 males and 7 females aged 15-30 years participated in the study. The selection of these subjects was based on the following criteria: subjects were pre-lingually deaf and they were proficient in using ISL in day to day conversation as assessed on a checklist prepared for the purpose by the investigator. The parents of these subjects were reported to be having normal hearing

Cities Word order

Delhi S-O-IO-V (Incorporation of O and IO in V)

Bangalore S-IO-O-V (Incorporation of O in V)

Calcutta S-IO-O-V IO-S-O-V (Incorporation of O and IO in V)

Bombay S-IO-O-V IO-S-O-V (Incorporation of O and IO in V)


312

sensitivity. The minimum educational qualification of the subjects was Secondary education in a special school for the persons with Hearing Impairment. All these subjects had learnt ISL as part of the education in the special education. The medium of instruction of writing and reading during their academic skills considered here was English. Material: Preparation of master video tape: A prerecorded CD of SE was considered as a master video tape which consisted of signed expressions in the sentence form for all the 14 markers of SE embedded in few simple short sentences. There were three sentences under each sign marker and hence totally the video consisted of signed expressions for total of fourty two sentences (14x3= 42). The signed expression for these sentences thus served as the model/ referent against which the signs expressed for the same sentences by the ISL users were compared. Task and Procedure: The audio-video tape of the selected set of sentences (42) as signed by the “Model Signer” who was proficient in SE was the referent. Each sentence as expressed by the SE user in the tape was written on separate cards. The video/ audio recording of the participants of the study, who were ISL users were collected individually. The ISL participants were presented the cards containing the 42 sentences, one at a time and instructed to sign the sentences which were written on the card. This was carried out in individual setting, in a room with minimal distractions. Appropriate instructions were provided to the subjects before recording. Instruction given to the subjects was as follows: “You will be presented with simple and short sentences written on the card. A single sentence will be presented one at a time. You have to read carefully and sign for the sentence slowly and clearly”. Each participant was given two trials for each sentence. The signed expressions of each subject for each of the sentences were audio-video recorded using a video camera which was placed at a distance of 4-5feet away from the participants. All the participants were kept blind to the purpose of the study. The order in which the 42 sentences were signed by the ISL signers was randomized across the participants.

Analysis: Analysis of Signed Expressions of the ISL users by the judges was carried out in three steps.

Step 1: Three judges were selected. The judges were qualified sign interpreters of ISL. The purpose of the study was not revealed to the judges.

Training session: A training session was provided to these judges using a video sample of two subjects who were not a part of this study. At first the signed expression for all the 14 markers of SE embedded in few simple short sentences as recorded by a proficient SE signer was played to the judges.

Following this, the video sample of the ISL signer for the same sentences was presented to the judges. The judges were instructed to judge and compare the signed expressions of the ISL user with that of the model SE sample in terms of the sequence in which the Subject, Object, Verb occurred in ISL signed expression, to identify the “grammatical markers” if any in the signed expressions of ISL user that were a match with the “signed markers” of SE as recorded in the model speaker of SE and also to assess the “grammatical markers” equivalent to the 14 “signed markers” of SE if present, in terms of HOLME (Handshape, Orientation, Location, Movement and Expressions). The response for each of the test sentences or each marker was entered by the judges in the recording form which was specifically prepared by the investigator for the purpose.

Step 2: The three judges carried out the analysis of the recorded samples of the signed expression of ISL signers by comparing same with that of the model SE sign for each of the sentence. A total of 630 (42 x 15n = 630) video samples of ISL signed expression by ISL users was analyzed by all the three judges in terms of word order, presence of “grammatical markers” if any similar to “sign markers ” of SE, the Handshape, Orientation, Location, Movement and Expressions (HOLME) features of those markers if any identified in the samples. Each judge analyzed the samples independently without mutual consultation.

Step 3: Reliability check: Item by item inter judge and intra judge reliability was checked for both the judgments of word order type and for “grammatical markers”. 10% of the total video samples (63) of ISL signed expression by ISL users was considered for intrajudge reliability and this was carried out by all the three judges after a gap of one week for intrajudge reliability. The raw data obtained by the judges was tabulated and subjected to statistical analysis.


After the analysis by the judges, the frequency of occurrence of word order types, the frequency of occurrence of grammatical markers, and the HOLME features of the grammatical markers were compiled from the responses given by all the three judges. Item by item intra and inter judge reliability was checked using the same data. These results are presented and discussed under seven sections: Section 1: Comparison of percentage occurrence of word order types of target sentences in SE and ISL.

The target sentences selected from the SE referent video had SVO word order for all the sentences incorporating 14 markers and the word

Word order, Grammatical markers: ISL vs. SE

317

Intra judge reliability for word order: The item by item intra judge reliability for all the three judges was calculated and the results revealed 92%, 90% and 94% intra judge reliability for Judge1, Judge2 and Judge3 respectively.

Intra judge reliability for HOLME: The intra judge reliability for all the three judges for the judgment of HOLME features was calculated and it was found to be more than 80% for all the three judges.

Inter judge reliability for word order: The inter judge reliability for word order was found to be more than 90% for all the three judges. The inter judge reliability was found to be 94%, for Judge1 and Judge2, 95% for Judge2 and Judge3 and 98% for Judge3 and Judge1 respectively.

Inter judge reliability for HOLME: The item by item inter judge reliability for all the three judges for the judgment of HOLME features for sign markers was more than 80%.

Conclusions

The results revealed that the predominant word order seen for the set of sentences with 14 sign markers of SE as expressed by the signed expression was SVO followed by SO. The results also revealed the presence of independent signs for „grammatical markers‟ in the signed expression of ISL for grammatical categories such as opposite „Un‟, agent marker possessives and superlatives. The analysis for these grammatical categories based on the HOLME features was made and it was found that the any one of the feature or combination HOLME features in ISL signed expression matched with those of the SE. Excluding these grammatical categories, there was no independent sign for grammatical categories such as: regular plurals, irregular plurals, and comparatives, but variation in terms of Location, Movement, Orientation associated with non manual markers were found which served the function of sign markers. Hence the analysis of location, movement of the hand, orientation of palm and fingers and non manual markers for regular plural, irregular plurals and comparatives revealed that there were similarities as well as differences expressed in the signed expression of ISL users for the sentences embedded with these categories.

References Aarons, D. (1994). Aspects of the syntax of American sign

language. Boston University Graduate School. Bornstein, H. (1973). A description of some current sign

systems designed to represent English. American Annals of the Deaf, 118, 454-463.

Bornstein, H., Saulnier, K., & Hamilton, L. (1980). Signed English: A first evaluation. American Annals of the Deaf, 125, 467-481.

Fischer, S. D. (1973). In R. B. Wilbur. (Eds.), American sign language: linguistic and applied dimensions. San Diego: University Park Press.

Gordon. R., Jr. (2005). Retrieved from Ethnologue: Languages of the World, Fifteenth edition. Dallas, Tex.: SIL international. Online version: http://www.ethnologue.com/

Jepson, J. (1991b). Urban and rural sign language in India. Language in Society, 20(1), 37–57.

Jones, N., & Mohr. K. (1975). In R. B. Wilbur. (2nd Ed.), American Sign Language: Linguistic and Applied Dimensions. San Diego: University Park Press.

Lidell, S. (1977, 1980). In R. B. Wilbur. (2nd Ed.), American sign language: linguistic and applied dimensions. San Diego: University Park Press.

Pallavi, M., & Manjula, R. (2010). Comparison of subject, verb, object structure in sign language users with hearing impairment: Influence of the verbal native language? Journal of Indian Speech, Language and Hearing Association, 24 (1), 48-58.

Pizzuto, E., & Corazza, S. (1996). Noun morphology in Italian sign Language (LIS). Lingua, 98, 169-196.

Sinha, S. (2007). A skeletal grammar of Indian sign language. PhD thesis. University of Mysore.

Stokoe, W. (1960). In R. B. Wilbur (2nd Ed), American sign language: Linguistic and applied dimensions. San Diego: University Park Press.

Vasishta, M., Woodward, J., & Wilson, K. L. (1978). Sign languages of India: Regional variations within the deaf populations. Indian Journal of Applied linguistics, 4(2), 66–74.

Woodward, J. (1978b & 1979). Handshapes in Indian sign language. In M. M. Vasishta, J. Woodward, & S. De Santis. An introduction to Indian Sign Language. (pp. 3-6). All India Federation of the Deaf, New Delhi.

Zeshan, U. (2001). Mouthing in Indo-Pakistani Sign Language (IPSL): Regularities and variations. In P. Boyes Braem & R. Sutton-Spence, (Eds.), The hands are the head of the moth: the mouth as articulator in sign language. (pp. 247-71). Hamburg: Signum.

Zeshan, U., Vasishta, M., & Sethna. M. (2004). Implementation of Indian sign language in educational settings. Asia Pacific Disability Rehabilitation Journal, 15 (2), 15-35.

Zeshan, U. (2003). Indo-Pakistani sign language grammar: A typological outline. Sign Language Studies, 3(2), 157–212.

http://www.ethnologue.com/

part b speech-language pathologyaiish.ac.in/images/pdf/09-10slp.pdf · 1e-mail:...

Documents