ED 064 697

AUTHORTITLE

SPONS AGENCY

BUREAU NOPUB DATEGRANTNOTE

EDRS PRICEDESCRIPTORS

DOCUMENT RESUME

24 CS 000 072

Deutsch, Cynthia P.The Development of Auditory Discrimination:Relationship to Reading Proficiency and to SocialClass. Final Report.Office of Educaticn (DREW), Washington, D.C. Bureauof Pesearch.BR-6-3034Jun 72OEG-6-10-21571p.

MF-$0.65 HC-$3.29*Auditory Discrimination; Auditory Tests; BeginningReading; Cultural Differences; DisadvantagedEnvironment; Elementary Grades; *LearningDisabilities; *Reading Difficulty; Reading Readiness;Social Class; *Social Disadvantagement;*Socioeconomic Influences; Socioeconomic Status;Visual Perception

ABSTRACTThe determination of the relationships among auditory

discrimination ability, social class and age group differences,reading skill ability, and visual perceptual skills was the objectiveof this study. Fifteen New York City public schools provided 180first, third, and fifth grade white and black males from lower andmiddle socioeconomic status (SES). A variety of auditory tests wereadministered as well as a visual discrimination measure, an attentionmeasure, reading tests and an intelligence measure. The resultslargely support the hypothesis that poor auditory discrimination is amajor intervening variable between social conditions and readingretardation. The relationship is stronger for blacks than for whitesand decreases with age indicating that teaching and remedial trainingshould be oriented differently for varicus SES, racial and agegroups. (TO)

Final Report

Project No. 6-3034Grant No 6-10-215

,

SCOPE OF INTERESI NOTICE

The ERIC Faohty hes kssignocithis docu t lar ocesengto:

In our judgement. this deaurrerntts also of interest ta the eteertni-hottses noted RO the right. Indent-tog should relhect then seutoteIpoints oI view.

The Development of Auditory Discrimination:Relationship to Reading Proficiency and to Social Class

Cynthia P. Deutsch, Ph.D.Institute for Developmental Studies

New York UniversityNew York, New York

June, 1972

U.S. DEPARTMENT OFHEALTH, EDUCATION, AND WELFARE

Office of EducationBureau of Research

U F LiEPARTMENT OF HEALTH.tDUCATION & WELFAREOFFICE OF EOLICATiON

THIS DOCUMENT HAS BEEN REPRO-DUCED EXACTLY AS RECEIVED FROMTHE PERSON OR ORGANIZATION ORIG-INATING IT POINTS OF VIEW OR OPIN-IONS STATED DO NOT NECESSARILYREPRESENT OFFICIAL. OFFICE OF EDUCATION POSITION OR POLICY

Final Report

Projuct No. 6-iniqGrant. No. G-l0-215

The Development or Auditory Discrimination:Relationship to Reading Proficiency and to Social Class

Cynthia P. Deutsch, Ph.D.

Institute for Developmental Studies

New York University

New York, New York

June, l97,.

The research reported herein was performed pursuant to a grantwith the Office of Education, U.S. Department of Health, Educa-tion, and Welfare. Contractors undertaking such projects underGovernment sponsorship are encouraged to express freely theirprofessional judgement in the conduct of the project. Points ofview or opinions stated do not, therefore, necessarily representofficial Office of Education position or policy.

U.S. DEPARTMENT orHEALTH, EDUCATION, AND WEIXARE

Office ol EducationBureau of Research

TABLE OF CONTENTS

PaveAcknowledgements... . iv

Introduction and Review ol the Literature 1

Method13

Results19

Discussion and Conclusions45

Rei'erenecs55

Appendix I59

Appendix II63

3

iv

Acknowledgements

A long-term and complex research project such as thatreported herein owes whatever success it achieves to the jointefforts of many people: staff, consultants, administrators,colleagues, and, most of all, subjects. lt is patently impos-sible tu acknowledge all contributions, or to name all contrib-utors. However, it is appropriate to make some attempt to singleout those who helped to formulate the instruments and to gatherthe data for the study, and those who assisted in statisticalanalyses and in the final write up. In the former category Ishould like to thank (alphabetically) Ann Bishop, Richard Coleman,Judy Gorrell, Dorothy Schmidt, Dr. Sol Schwartz, Margaret Thomas,and Ann Weissmann. In the latter category, thanks are dueSandra Dalton, Dr. Judith Goldston, and Dr. Carol Silfen. MaureenMeArdle Kaley was involved in supervising the study during thedata collection and initial analyses and write-up, and Jo Rend'eFine has supervised and edited the final organization and writing.The final form of the study owes much to their efforts. LyndaHurwitz and Carol Kerr performed the necessary secretarial ser-vices with*cheerfulness and a high level of efficiency; thefinal draft was typed by Jane Cohen.

It is impossible to acknowledge sufficiently the contrib-utions of my colleagues at the Institute for Developmental Studiesduring these past 12 years: association with an active andproductive research and developmental group contributes immeasur-ably to the endeavors of each of its members. I would Jike atleast here to say "thank you" to my colleagues and to my husbandand colleague, Dr. Martin Deutsch, for the fostering atmospherewhica has obtained at the Institute. I should also like toacknowledge the cooperation of the New York City Board ofEducation in permitting the project to be conducted within theschools, and to express aPpreciation to the many district super-intendents, principals, assistant principals, and teachers inthe various schools who were so helpful. As is always the case,the greatest debt goes to th.q,subjects, each of whom submittedto several hours of testing procedures and who must of necessityremain anonymous.

With the cooperation and assistance of so many, I mustyet make clear that responsibility for opinions, conclusions --and errors -- is mine.

4

Chapter .1

Introductionand

Review u1 the Literature

ProhHm and ohjeetivesThu fligh ineidenee 01 reading and other learning disabil-

iLjes in childppn lpom lower socio-eyonomic eireumstances is pri-marily an educational problem. Not only is reading centPat LOthe c'dlleatiODUI 1,W0Ce5 in the tew years, but later subjectteaehing depends heavily on the student's ability to read. How-ever, readiirf; and other learning disabilities are also potentiallysevere social problems. The child with a reading disability islikely to experience more Jaii1 i1'0 in school than is the able read-CP. His failures will more than likely lead him to develop nega-tive attitudes toward school itselr and prohahly a more negativeand removed attitude toward other social institutions as well.In addition, on increasingly industrialized and automated so-ciety, there are Jewel' and Jewel., jobs fur unskilled workers. Anyperson who doc:s not deveLop adequate reading skills will find italmost impossibte to enter the skilled labor market-

With this awareness of.- both the y;reateP incidence and ser-ious implications oL reading disabilities In the lower socio-economic groups, the task remains to determine Lhe mediating Fac-tors between social conditions and learning periormunee. Sinceinstitute res2arch has shown such striking dirLerences in audi-

S.., child.cen between good and poorreaders (Deutsch, 1964; Katz and Deutsch; 1963b) it is possiblethat at least a portion or the class discrepancy in reading re-tardation is attributable to auditory problems, and thus develop-ment of auditory discrimination may be one of the mediating fac-tors (of. argument in C. Deutsch, 1964).

Reading research ih general has been primarily directedtoward analysis of underlying visual and more recently, linguis-tic, factors. Invest-jgation of auditory factors in leading has oc-cupied third place. One reason for the relative scai2city of suchresearch prior to initiation of the present study may be that in-terest mainly Focused on normal middle-elass populations, inwhich auditory discrimination problems are neither expected norfound in significant numbers. It was hypothesized by the Princi-pal Investigator that perhaps middle-class children, from theirquieter, more speech-direeted environments, more frequently at-tain minimal auditory discrimination levels by the time they en-ter first grade than do children from lower socioeconomic strata(Deutsch, 19(11).

1

With the high incidence of reading retardation among chil-dren from lower-class backgrounds, it is important to learn if a .

major intervening variable between social conditions and readingretardation is poor auditory discrimination. Thus, a main objec-tiv.2 of the present project was to determine if there are socialclass differences in auditory liscrimination ability.

A second objective was to determine the prevalence of aud-itory discrimination difficulties coincident with different lev-els of reading skill within and between class groupings. The re-lationship might differ in different age groups, and this alsowas investigated.

A third objective was concerned with exploring the possi-ble relationship between levels of auditory discriminatic., skillsand the visual-perceptual skills involved in reading.

Related Research

Socioeconomic status and auditory discrimination abilit .It has been suggested (Deutsch, 1964, 1968) that the environmentof the lower-class child, who is typically the poorer reader, isan important factor in his ability to develop auditory discrimin-ation skills. Not only does the environment affect the amountand variety of stimuli to which the child is exposed, but it alsoinfluences the nature and amount of practice he gets in learningto discriminate stimuli from each other. Higher socioeconomic(SES) children are exposed to a different and richer stimulus en-vironment than are children from less privileged homes. Thereis likely to be less verbal interaction, and less emphasis on dis-criminating differences between objects and on attaching appro-priate labels to objects in lower-class homes. The environmentof the lower-class child might also be characterized as a muchnoisier one than that of.the middle-class child.

Research on the reticular activating system lends cogencyto a hypothesized relationship between a noisy environment andauditory discrimination difficulties. The reticular system,which seems to be responsible for a genera/, over-all activatingfunction in the nervous system without which no stimulus is ef-fective, is able to influence transmission of sensory messageswithin the central nervous system. Animal experiments show thatwhen the reticular system is directly activated, sensory trans-mission is inhibited or facilitated deperding on where the acti-

.vating stimulus/is applied (11ern6ndez-Pe6n, 1961). Experimentsby Hern/andez-Peon and others indicate that activation of thebrain-stem reticular formation inhibits auditory transmissionvery early in the path of that transmission. In addition it hasbeen found in cats that auditory evoked potentials are reducedwhen the animal is attentive to stimuli in other modalities(Bach-y-Rita et al., 1961, reported in Hern6ndez-Pe6n, 1961). Sim-

ilarly, the potential in other modalities is reduced when the an-

2

imal's attention is on an auditory stimulus. It was also toundthat arrerent7 messa:-ws elicited hv stimuli which UPC attended toare facilitated.

While dipeet iurcronees from these animal experiments tuhuman behavior are obviously inappropriate, these animal oexperments do provide hypoLhoses about functioning in the human 'WV-vous system, and these hypotheses can be viewed in the light ofavailable behavioraj data. The animal findings on the reticularsystem are consistent with the inFormation-theory findings thatthe signal-to-noise ratio is influential in the stimuli per-ceived and in the responses evoked. The higher the ratio, i.e.,the greater the amount or signal as compared with noise, theMore likely will be the accurate perception of the signal. Onecould thus hypothesize that with a low ratio, i.e., with a lotof noise in the system, the excess activation of the reticularsystem is interfering with travel of the signal up the neuralpaths.

Thus a child raised in a noisy environment with littledirected and sustained speech stimulation might well be deficientin his discrimination and recognition of speech sounds. He couldalso be expected to be relatively inattentive Lo auditory stimuli,and iurther, to have difficulty with another skill, such as read-ing, which is partially dependent on auditory discrimination skill.

Only a few studies have focused specifically on the audi-L. 1ekioni.tig ui disadvantaged children as com-pared with advantaged children. Clark and Richards (1966) ad-ministered the Wepman Test of Auditory Discriminat:ion to twogroups: 29 paying and 29 non-paying Head Start Children in Mad-ison, Wisconsin. The results showed a significant difference0)(.001) between disadvantaged and advantaged groups. MeArdle(1965) investigated the auditory discrimination ability of agroup of 20 day-care and 24 laboratory nursery-school children inKnoxville, Tennessee. There was a significant difference(p.001) between the groups as measured by a modified form of theBoston University Speech Sound Discrimination Test. The investi-gators of both studies ccncluded that the auditory discrimina-tion abilities of disadvantaged preschool children were poorerthan the auditory discrimination abilities of those from advan-taged environments.

In a study comparing 27 "culturally deprived" 5 year oldsfrom a Project Headstart with 27 middle class children from anursery school, Giebink and Marden (1968) also found that scoreson the Auditory Vocal Automatic Subtest of the Illinois Test ofPsycholinguistie Abilities were significantly lower for the so-called "culturally deprived" group.

Auditory discrimination and readinr Several investiga-tors have examined the relationship between auditory perceptualfunctioning and reading achievement within groups other than the

disadvantaged. In order to standardize the Reading Aptitude Test,Monroe (1935) administered a number of readiness tests to childrenin the primary grades. Correlation coefficients between each ofthe major types of readiness tests (auditory, language, visual,etc.) and end of first-grade reading achievement were computedfor 85 children in four first-grade classrooms. There was ahigher correlation between reading and a composite auditoryscore, including measures of blending, auditory memory, and cor-rect pronunciations (.66) than for any other readiness var-iable.

Wepman (1960) administered tests of auditory discrimina-tion, articulation, intelligence, and reading to 156 first andsecond grade children at the end of the school year in Chicago.The Chicago Reading Tests and the Wepman Auditory DiscriminationTest were among the tests given. There was a significant re-lationship between low reading attainment and poor auditory dis-crimination, but a comparison of the data for the two gradesshowed that the number of children with poor auditory discrimin-ation decreased in the second grade.

The relationship between prereading measures of auditorydiscrimination and reading achievement at the end of the firstgrade was examined by Dykstra (1964) in eight Minneapolis schools.Seven auditory discrimination subtests selected from readingreadiness tests were administered at the beginning of the schoolyear and two subtests of the Gates Primary Keading Test were givenat the end. Complete data were gathered on 632 pupils. Resultsindicated that: (a) intercorrelations among auditory discrimina-tion measures and subsequent reading achievement were low; how-ever five of the seven auditory discrimination measures made asignificant contribution to a multiple regression equation whichwas designed to predict reading achievement; and (b) variationsin performance on the auditory discrimination and intelligencemeasures accounted for less than half of the variation in per-formance on the reading measures.

Harrington and Durrell (1965) tested 500 parochial school,second-grade pupils in Boston. The investigators devised an audi-tory discrimination instrument to measure the subject's abilityto perceive initial consonant sounds, rhyming at the ends ofwords, final consonants, and a combination of initial and finalconsonants in words spoken by the examiner. The subjects werematched for mental age, visual discrimination ability, and pho-nics ability. For each matched pair, a comparison of readingability between the high- and low-discrimination scorer was made.It was found that the pupils with superior auditory discriminationwere also significantly superior in reading ability.

A longitudinal study of the 105 first-grade children whoentered the two elementary schools.in Oxford, Ohio, in September,1958, was conducted by Thompson (1963). In the month preceding

the iirst grade and in the eighth month of the second grade, threeauditory discrimination tests were administered: the Wopman Audi-tory Discrimination T0:1L, the Boston University Speech Sound Dis-eriminaLion Test, and the Auditory Discrimination and OrientationTest. In the ninth month uf the second grade, the word recogni-tion and the paragraph reading subtests ui Lhe Gates AdvancedPrimary Reading Tests were administered. Subjects WcrY alsogiven the Wechsler intelligence Scale Lbr Children (MSC) and anaudiomeLrie Lest, Two criteria were used to select good andpoor ...caders: (a) those 24 who were at the upper and those 24who were at the lower end ot the reading distribution; and (b)those 24 whose reading ability greatly exceeded their mental abil-ity, and those 24 whose mental ability greatly exceeded theirreading ability. The findings indicated that: (a) of the 24best readers, 16 had possessed adequate auditory discriminationupon entering the first grade. ExaminUtion of the 20 poorest read-ers, however, indicated that only one had demonstrated adequateskill in making auditory discriminations at the beginning of thefirst grade; (b) high auditory discrimination test scores weremore likeiy to be characteristic of the good readers at the upperend of the men:al ability distribution than the good readerswhose reading ability greatly exceeded their mental ability; and(c) approximately 24 percent of the subjects had inaccurate audi-tory discriminative ability at the end of the second grade, andone-half of these were classified as poor readers.

Christine and Christine (1964) used the Wepman AuditorycvaluaLe subjects From a midwestern school

system. They tested 27 subjects in the second and third gradesreading at or above grade level; 15 primary-grade subjects readingbelow grade level; and 11 primary-grade subjects with speech de-fects. The first group was significantly superior to the secondin auditory discrimination ability as measured by the WADT. Theyconcluded that their data supported the hypothesis that poor audi-tory discrimination is one causal factor in reading retardation.

The auditory abilities or 188 fourth-grade childrel wereinvestigated by Reynolds (1953) in four schools in Minnesota. Theschools were randomly selected, and no attempt was made to sep-arate good and poor readers or hearers. Included among the 14tests administered were the Gates Basic Reading Test, the BondSilent Reading Diagnostic Test, the Seashore Pitch DiscriminationTest, and the Word Discrimination Test.

Reynolds found that auditory measures used for the pre-diction of performance on two reading tests - word recognitionability and knowledge of sound values fur common word elements -were inconsistent in analyses made separately for each school.In three or the schools, the auditory measure provided a predic-tive value that was not significantly better than mental age. Ina fourth school, it was found that measures of auditory memoryspan, word discrimination ability, and pitch discrimination, when

6

Combined in multiple regression equatiops, providt-J significantly

better predicti.ons of performance of the two reading tests than

mental age alone.

Wheeler and Wheeler (1954) described a study in whicli 629

childreP in the fourth, fifth, and sixth grades of the Coral

Gables Elementary School were given the Metropolitan Achievement

Test, the Seashore Pitch Test, and an auditory disc,.imination

test designed by the investigators to measure a subject's abllity

to: (a) discriminate typical word-pairs; (b) discriminatebe-

tween paired sound elements and determine whether each pair was

the same or different; (c) select the one word from four which

did not rhymo; and (d) select from a list of three sounds the

one sound which he had heard in a stimulus word previously pro-

nounced by the examiner. The authors reported the following re-

sults: (a) a significant but very low correlation existedbetwr-en

the Seashore Pitch Test and the Wheeler Auditory Discrimination

Test. The correlation for the fourth grade was significantat

the .05 level and for the fifth and sixth grades at the .01 level;

(b) a significant but low correlation (.01 level) existedbetween

pitch discrimination and reading for children in the fifth and

sixth grades, while a negligible relationship existed between

pitch and reading for children in the fourth grade; and (c) all

correlations of the auditory discrimination factor with reading

abilities were statistically significant (.01 level); however,

these correlations were low.

Templin (1954) measured the ability of the fourth-grade

children in five Minneapolis schools to discriminate between

consonant sounds. The 26 subjects having the highest scores on

the Durrell-Sullivan Reading Test were classified as the upper

reading group, and the 26 having the lowest scores were classified

as the lower group. The author reported no significant differene

between the scores of the upper and lower reading groups on this

group test of sound discrimination. She conclPded that the test

was comparatively easy for fourth-grade pupils evidenced by

the number of subjects who achieved the maximum possible score.

S_ocioecononiesta-tt_g_tualfunctioninandpeadin& While some investigators have examined the relationship

of auditory perceptual furctioning to reading in disadvantaged

subjects, these are few in number.

Katz and Deutsch (1963b) studied the relationship of audi-

tor', discrimination to reading among disadvantaged boys. Sub-

ects were black males in the second and fourth grades in two

Harlem schools. All students had a Lorge-Thorndike IQ above 70

and evidenced no severe emoional or physical disabilities. The

Gates Advanced Primary Ruading Test was administered at the be-

ginning of the school year. Two word discrimination tests were

administered to 72 subjects: one with 24 English word pairs and

the other with 24 Hebrew word pairs. The inc3usion of the Hebrew

word pairs was to assess whether or not antj.cipated differences

10

liehiWOH ChildVoji IO dilferelW0:4 In discrimination abil-ities meoeJy io Jamillarity with the material.Thy HtE (o) ;:pod tild poorreaders (.t,e !mu.: lo 6.fliee si:,;nifiodotAy.on the discriminationtosts, thy !ood roJdyis makjw, levier errors on both tests;(b) all children riponded byrter to moaningrul stimali. Mean-ingfulnes:;, however, aid nf,t interact significantly with readinglevel, S,vceStiui. Ldat the underlyin!.., discrimination skillsrather than dill-eoonces in familiarity with the stimuli differen-tiated the ,..woup:,.; and (c) the interaotion of reading level wi.thage was siniFicant, indieoting that discrimination skills Oil-Lt2rentiuted the r.,00d and poor readers best at the second gradelevel. IQ scores were sinificantly related to auditory discrim-ination at all age lovels.

Robinson and Hanson (TOGS) compared disadvantaged first,second, and third grade children on reliability of instrumentsused to moasure factors related to reading success. They foundthat, among other tests, the Metropolitan Readiness Tests andAudiometer test wore reliable in all groups; the Word Discrimina-tion Test was reliable with dlsadvantaged children in the firstand third grades and the Weprdan was reliable with disadvantaged

seeond, and third grode children, and with middle-elasschildren in the first grade.

Socioeconmie status auditory and visual perceptual _lune-__tionino. und readifw. The ability to make shiits from one sense

! to 1)2 crucial for the learningof reading. The child must, for example, be able to develop cor-respondences between his previous auditory perceptions and wordson a printed page>

Katz and Deutsch (1063a) studied the ease with which atten-tion is shifted between auditory and visual stimuli in 48 first,third, and fifth grade block children grouped according to readingability. A bimodal reaction time app,-,ratus was employed. Itautomatically presented four separate stimuli: a red light, agreed lignt, a high tone of 1200 cps, and a low tone of 400 cps.The subject's task was to lift his finger from a button every timea stimulus was pereeived. The results indicated that older chil-dren hod faster overall reaction times (p Z.001), and that re-tarded readers exhibited ,,reater difficulty than normal readersin shilling from one modality to another (p.01).

Raab, Dontsch, aad Freedman (19(i0) had earlier employedthe same apparal-us and procedure with 111 average readers and 10poor readers in the Yonrill and fifth grades from lower-middIcand lower-elass Jamilies. As in the above exl-ierlilent, the subjectwas i:isfrueted to raiSO his index finger as quickly as possiblewhenever any of Elie stimuli were presented. Reaction time wasmeasured fvom stimulus onset to the lilting of the finger. Re-

117

8

sults indicated that poor readers showed significantly greaterdifficelty in making a erossmodal shift than normal readers andsignificantly less visual responsiveness and efliciency. Nosignificant difierenees in reaction time to sound were found.Both good and poor readers Eormed sets to light but formed nosets to souod.

In a later investigation, Katz and Deutsch (1964) comparedauditory, visual, and auditory-visual learning on a serial learn-ing task. The subjects were good and poor readers, male, black,and in the first, thjrd, and fifth grades; An analysis of vari-ance of the number of correct results indicated that: (a) oldersubjects learned more efficiently under all the stimulus condi-tions; (b) the differences between good and poor readers, whichwere evidenced at all age levels, were.greatest in younger chil-dren; (c) auditory presentation was most difficult for almost allsubjects; and (d) an ieteraction between reading level and modeof presentation suggested that retarded readers had particulardifficulty with auditory tasks.

Weiner, Wepman and Morency (1965), studying a sample of 28good and 28 poor readers in the fourth grade, found that low audi-tory discrimination and low visual discrimination were both sig-nificantly correlated with low reading attainment, although notwith each other. The correlation between the two perceptual tests(Wepman ALiditory Discrimination Test and the Chicago Test of Vis-ual Discrimination) was extremely low (.05). The authors con-cluded that there are two different learning types, related toability to use vision and audition in gaining information: chil-dren with auditory problems who have reading difficulties andchildren with visual problems who have reading dificulties.

Lovell and Gorton (1965) investigated the differences be-tween 50 backward readers, 9 to 10 years of age, and 50 normalreaders, both of average intelligence, on a number of selectedtests including the ITPA, the Birch and Belmont test of auditory-visual integration, Gorton's test of auditory discrimination,the Bender Visual-Motor Gestalt, Shapiro's test of rotation, aspatial orientation test and a motor impairment test. Backwardreaders differed from normal in the expected direction on alltests although not significantly on the test of auditory discrim-ination, the Bender Visual-Motor Gestalt and the Shapiro. Whentest scores were factor analyzed, the facto), patterns of the twogroups were different, such that in the backward group readingage was linked most strongly with auditory-visual integrtion andmotor performance. This suggested to the authors that some kindof neurological impairment, not necessarily specific, may belinked with reading skills among backward readers.

Butenica (1968) studied advantaged and disadvantagedfirst grade youngsters to see if auditory and visual tests wererelated to socioeconomic background and also to reading achieve-ment and spelling. He found that auditory and visual tests ac-counted For much more variance in reading than did a group 112

12

test; that auditory and visual perception were virtually inde-

pendent ol one another; and that middle-elass.children perFormedbetter than _lower-class ehildven on ail tasks. Thy tests he used

iLe[uded: the Reversals and Picture Squares, the Visual-Motor In-

tegration Test, the Wepman Test ot Auditory Discrimination, und

the Non-Verbal Auditory DiscrimivaLion Test.

In 1969 Bruininks studied the relationship oF auditory dis-

crimination to read.inl!, among black disadvantaged boys. His sample

of 105 subjeets had a mean age of 8 years, 7 months. Subjects

were screened to insure that they had adeguate auditory and visual

acuity. He used a number of perceptual tests, among them the

Wepman. There was a significant association between audition and

reading, bat he did not Lind a curvilinear relationship betweenauditory discrimination and reading us he had predictLd. He also

found that partial correlations between auditory tests and reading

decreased markedly when verbal intelligence was controlled. Ile

did find that it was auditory tests rather than visual, which

wore significantly, though not highly, correlated with reading

over and above intelligence.

Weaver and Weaver (1967) reported that disadvantaged black

children earn scores-on the ITPA which are significantly lower

than their mental ages, and that they earn scores on the subtests

involving auditory and vocal channels that are significantly

lower than scores obtained through visual-motor channels.

Results from studies investigating the relationship be-

tween auditory discrimination and reading with primary-grade,

middle-class subjects indicate a significant difference in auditory

discrimination abilities between retarded readers and subjects

reading at, or above grade level: (with first grade subjects -

Monroe, 1935; Dykstra, 1964; with second grade subjects - Harring-

ton & Durrell, 1965; with.first and second grade subjects - Wep-

man, 1960; Thompson, 1963; and with second and third grade sub-

jects - Christine & Christine, 1964). With intermediate grade

subjects, however, findings were less consistent. With Fourth

grade populations, Reynolds (1953), in a predictive study using anon-selected sample, and Templin (1954), in a comparison study of

good and poor readers, found no significant relationship between

auditory discrimination skills and reading ability; however,

Wheeler and Wheeler (1954) reported that with a non-selected sam-

ple of fourth, fifth, and sixth graders, there was a low but sig-

nificant correlation between pitch discrimination skill and read-

ing ability. They found slightly higher correlations with fifth

and sixth graders than with fourth graders. Also, Reynolds

(1953) found that in one of the four schools in her study, audi-

tory mcas,-es provided significantly better predictions of speci-

fic reading abilities than were provided by mental age.

In general the literature suggests.that with increasing

age there is a decreasing number of .subjects with poor auditory

13 9

10

discrimination (Wepman, 1960 - a decrease from first to secondgrade; Templin, 1954 - a ceiling effect at the fourth grade level;and Thompson, 19(i3 - a decrease from first to second grade).

Also, Thompson (1963) found that for good readers, a highauditory discrimination score was not characteristic of subjectswhose reading level surpassed their mental age level, and thathalf the subjects with inaccurate auditory discrimination werepoor readers. Dykstra (1964), and Thompson (1963) found signi-ficant relationships between intelligence and auditory discrimin-ation ability on the primary grade level. Dykstra (1964) andWheeler & Wheeler (195(1-) found low significant intercorrelationsamong auditory discrimination tests.

The few studies investigating different SES levels in re-lationship to auditory discrimination were all with nursery schoolage subjects (Clark & Richards, 1966; McArdle, 1965; and Giebink &

Marden, 1968). They all found that the auditory discriminationskills of advantaged subjects were superior to those of disadvan-taged subjects.

Several investigators found that the positive relationshipbetween auditory discrimination and reading for middle-class sub-jects obtained for lower-class subjects as well (Katz & Deutsch,1963b - with second and fourth graders; Katz & Deutsch, 1964 -with first, and fifth graders; Robinson & Hanson, 1968 -with second anel third graders; Butenica, 1968 - with first graders;Bruininks, 1969 - with third graders; and Weiner, Wepman & Morency,1965 - with fourth graders).

The methods and findings relating auditory and visual per-ceptual functioning to each other and to reading were varied.Generally the relationship between auditory discrimination andvisual discrimination was either very slight or absent. Poorerreaders among both lower and middle-class groups had more diffi-culty with cross-modal shifts (Katz & Deutsch, 1963a; Raab,Deutsch & Freedman, 1960).

A positive relationship between auditory discriminationand reading, as well as between visual discrimination and readingbut not between auditory discrimination skills and visual discrim-ination skills was found by Weiner, Wepman & Morency (1965), and

Butenica (1968).

In the research on auditory perceptual functioning andreading performance of subjects from advantaged environments, thefindings have been inconclusive, perhaps as a result of variabil-ity among the studies in methods of sample selection and stimulus

presentation. In general, the results indicate that factors such

as age, sex, mental age, and auditory perceptual functioning seemto be interacting with reading skill; however, additional re-search is needed to determine precise relationships.

14

Studies i11vestialiu:4 tiff, auditory pereeptual shills ordisadvantaged subjects are Lewer in numbeP, but the findings are

More consistent. Results indicate that subjucts rvom impover-

ished environment,; achieve lower auditory discrimination scoresthan-middle-class subjects, cYhibit poorer auditory modality func-

tioning in comparison to visual functioning, and reveal inadequate

auditory perceptual skills coexistent with retarded reading

achievement.

15 11

Chapter 2

Method

SubjuetThe subjects were 180 white and black males from lower and

middle socioeffmomie status (SES) backgrounds as measured by theInstitute's SES scale. Thuy were drown from the lirst, third,and fifth gradcs in New York City public schools. There weretwelve groups of 15 subjects each, categoriv:ed as follows:

1.

2.Lower SES, white first gradeLower SES, black first grade

3. Middle SES, white first grade4. Middle SES, black first grade5. Lower SES, white third grade6. Lower SES, black third grade7. Middle SES, white third grade8. Middle SES, black third grade9. Lower SES, white fifth grade

10. Lower SES, black fifth grade11. Middie SES, white fifth grade12. Middle SES, black fifth grade

Extensive surveying of a large number of schools in differ-ent areas of New York City was necessary to obtain a sufficient

e,:r.flarison groups. Subjectswere located in lifteen different schools in the Bronx, Manhattan,and Queens.

Assessment ProceduresSince project emphasis was on the auditory modality, consid-

erable attention was given to auditory testing. The testingproceeded from the level.of simply responding to the presence orabsence of auditory stimulation (audiometric testing), to thelevel of perceptual recognition of complex auditory stimuli (IESAuditory Masking Test), to auditory discrimination among similarsounding words (Wepman Auditory Discrimination Test).

A. Allatallx_aatiEL

1, Auditory Screening: A standard audiometric test,to ascertain whether hearing was normal, was administered to thosechildren for whom school authorities had no audiometric record.

2. IDS Pitch Discrimination Test (Pitch Test):After pilot study results indicated that the Seashore Test, whichoriginally was to have been used, would not be appropriate (prin-cipally because of the comparative language required in responseto test items), the Pitch Discrimination Test was constructedwhich required the subject to discriminate between two tones 8,12, 17, or 2S cycles apart. ln 15 of the 110 comparisons, the2nd Lone was hiOler, and in 15 it was lower (8 cycle and 17 cycle

separation blocks of 10 items-- higher and 4 lower 2nd tones,2 equal tone items; 25 cycle block of 10 items-3 higher and 4lower 2nd ones, 3 equal tones; 12 cycle block of 10 items--4higher and 3 lower 2nd tones, 3 equal tones) . The subject had togive a response of "same" or "different" which made the test con-sistent with the Wepman Auditory Discrimination Test. The scoresfor the IDS Pitch Disorimination Test included: (1) - (4) thenumber of correct responses made for each of the four cycle separ-ations (8, 12, 17, and 25 cycles); (5) the total number of correctresponses over the four cycle separations (Total cycles); (6) thetotal number of correct responses made when bpth tones were thesame (Same score); (7) the total number of correct responses madewhen the second tone was higher than the first (Hi score); (8) thetotal number of correct responses made when the second tone waslower than the first (Lo score); and (9) the total number of cor-rect responses over scores (6), (7), and (8) (Total Hi/Lo/Samescore) . Scores 6-9 were corrected for guessing.

3. IDS Auditory Masking Test (Masking Test): Afterprevious masking tests failed to meet certain minimum requirements,three masking tests were constructed for this project. Theyincluded:

a. Masking Test I - Word x Noise (WxN): A word wasplayed simultaneously with a white noise mask. The intensity ofthe white noise was initially at its highest level to make theword unintelligible. The starting point of the mask (the loudestmask) was based on the point at which the word was first recognizedin the pilot sample, so that the number of masking steps variedfor each item, ranging over the three Masking Tests from 10-16steps. The noise intensity was reduced by 2 decibel steps untilit disappeared altogether and only the stimulus was audible (thiswas taped for ease of administration). Each stimulus-noise pre-sentation was followed by a 4 second silent period during whichthe subject was encouraged to identify the stimulus. The testconsisted of 12 test items and two practice items. Words weremonosyllables from Rineland (1945), and were among the 1,000 mostfrequently used words by first graders. A subject's scorerepresented the total number of steps over all items required foreach item to correctly identify the masked stimulus.

B. Masking Test II - Word x Word (WxW): A word wasmasked by a word. The procedure was the same as for the previousMasking Test. The test consisted of 13 test items and one prac-tice item. A total score representing the total number of stepsover all items required for each item to correctly identify themasked stimulus was again obtained.

c. Masking Test III - Nonsense x Noise (NxN):Nonsense syllables were masked by white noise. CVC consonant-vowel-consonant) trigrams of high meaningfulness (Glaze, 1928;Kreuger, 1934) were employed so as to eliminate age differencesdue to differential comprehension of words. There were 13 testitems and one practice item, and a total score was obtained.

1814

4. The Wepman Auditory Dierimination Test (WADT):The Wopman Auditory Discrimination Test (MDT-Special), rorm Jl,was administered to all subjects. This test evaluates ability toperceive Fine dillerenecf; between similap-soundin English words.Forty word pairs arc used. Ten of these are ;;ame word pairs. Ofthe 30 diflerent word pairs, 13 differ in initial phoneme only(e.g., map, nap), 13 difler in Ulna] phoneme only (e.g., shot,shop.), and 4 pairs differ in medial phoneme only (e.g., pet, pit).For the present study, the word paies were prerecorded on tape toinsure standardization of procedure and to avoid contamination bythe varied speech characteristics or different examiners. Ear-phones were used for each subject, who was instructed to listencarefully, and to indicate whether the two words in each pairwere the same or different. Four scores were obtained: (1) thenumber oi correct responses when the word pairs were difterent inthe initial phoneme; (2) the number of correct responses when theword pairs were different in the final phoneme; (3) the number ofcorrect responses when the word pairs were different in the middlephoneme; and (4) the total number of correct responses over (1),(2) , and (3).

B. Visual Discrimination and Memory MeasureThe Multiple-Choice Bender-Gestalt Test (Chicago

adaptation): In the first part of the test (the memory portion),after exposure to two practice items, 15 designs were shown oneat a timP to thp snhinot_ After each design was exposed from 3

LL Wetht reIllUVUd and replaced by a card containingfour designs, one of which was identical to and three of whichwere similar to the presented standard. The S was instructed toselect the one that was the same as the one he had just seen. Thesecond part of the test was a Matching task. After one practicetrial, the same 15 stimuli used in the Memory subtest were presented.Here, however, the standard and the choices were presented simulta-neously. The total numbeli of correct responses was computed foreach part of the test and could be compared with the tentative normsavailable. This test requires no drawing or other reproductivemotor skills--only visual perception and memory are involved.

C. Attention MeasureThe 1DS Revision of the Continuous Performance Test

(CPT): The Institute version of the Continuous Performance Testis used to measure vigilance or attention. Previous Institutestudies (Deutsch, 1964 reviews this research), have shown a dis-tinct positive relationship between performance on this test andgood reading performance. This test presents, via tape recordings,a 72 item series of color names. The subject is required to re-spond only to the one color name, "red," by pressing a button whichis linked to a timer which records his response latency. Fourscores were obtained for each subject: (1) median of reactiontimes that fell between 100-1000 milliseconds; (2) median of re-action times that fell between 100-2000 milliseconds; (3) percentof reaction times that fell between 100-1000 milliseconds; and(q) percent of reaction times that lell between 100-2000 milli-seconds.

19 15

.16

D. ReadilT Tests,I. Gotes Diagnostic Battery: These are widely avail-

able standard tests. The appropriate forms were selected for thethird and fifth grade children. The third grade was given theGates Advanced Primary Reading Test, Type APR, rorm 2, "Readingfor Grade 2 (second half) and Grade 3." This test consists of24 paragraphs accompanied by illustrations which are to be markedin such a way as to indicate the meaning of the paragraph. Amaximum of 32 correct markings is possible. The fifth grade wasgiven the Gates Basic Reading Survey, Type GS, "Reading toAppreciate General Significance for Grade 3 (second half) through8," which measures speed and accuracy of reading. Three scoreswere obtained for each subject in the third and fifth grades:Raw Score, Reading Age, and Reading Grade.

2. The IDS Reading Prognosis Test: This test wasused with the first grade children who had not yet learned toread. It was designed to tap skills that are involved in thereading process for use as a predictor of reading success. Itconsists of six subtests grouped in three areas as follows:(1) Beginning Reading (alphabet letters, sight vocabulary);(2) Perceptual Discrimination (auditory discrimination, visualdiscrimination); and (3) Languae (vocabulary, story telling).Total scores for each subject for each of the three areas wereobtained.

E. Iat.f11.412.2.2111-PilThe Lorge-Thorndike Intelligence Tests: The Lorge-

Thorndike Intelligence Tests were administered to all subjects.The first grade was given Primary Battery, Level 1, Form A; thethird grade received Primary Battery, Level 2, Form A; the fifthgrade was given Verbal Battery, Level 3, Form A.

Using the preceding instruments, six test batteries wereadministered to all subjects. The batteries are as follows:

Battery IIDS Pitch Discrimination TestMultiple-Choice Bender-Gestalt TestWcpman Auditory Discrimination Test

Battery IIThe Gates Diagnostic Battery (3rd & 5th grades)IDS Reading Prognosis Test (1st grade)

Battery IIIIDS Auditory Masking Test IInstitute Revision of the Continuous Performance Test

Battery IVIDS Auditory Masking Test II

Battery VLorge-Thorndike Intelligence Test

Battery VIIDS Auditory Masking Test III

20

Bat teri cs were _LI R1J Vi dually adminitercd excel) L :Cor -thc(atcs 1).1 o:;uo:.; zinc! the 1.1o1 ::;e -Thurndi n 3-,cnec.!Te:it v:716 wure adndni s Lercd to groups ol: five. Much buttery re -qui red approximately thirty minutes.

a*

21

Statjstical Analy,;os

Results

The data 1,..4.!P(' analyzed

ysis of. variance techniquos.

by means oi correlational and anal-

C000clations. A number or inieoeorrolatiou matrices wasconstruetc,d An °PdPU to analyzei the data: (1) an oveoall matrixof intereoroeiatiol:s between the individual variables (See Table1 for a listing 01 Lhcse) -- a total_ of (J11 correlations; (2) two

matri('es (ono kir each SES group) intereorvelating all variableswithin SES--a toLai of .1922 correlations; (3) thrf_ie intercorrela-

tion matrices (one ror eac) i :wado) comparing all variables withingradea total of 23 corri.lations; and (q) two intercorrclationmatrices (one for each race) comparing all variables within race--

a total or 3922 correlations. Thus, there were literally thou-szmil, of intercorrelations.

Since by chance alone there wou3d be some si,,vificant our-rchitions out of so large a number, in Or(ler to evaluate the sig-nificance of those obtained, patterns ol correlations were loohed

for which made psychological sense. Certain criteria were used asguides Lo the seicctin of such patterns. Correlations which re-lated to the hypotheses advanced were focused on fi-:st. Then,those correlations si:Jniiicant at p4:.05 between demographic para-

La .1,i l) k %,-.1-05, and between reading scores and othervariables were examined.

Analyses nr variance. Several analyses of variance werecarried out on the Pitch Test, the Wepman Auditory DiscriminationTest, and the Masking Test. These included: 1) two four-factoranalyses of variance with repeated measurements on the PitchTest; 2) a three-factor analysis of variance with repeated measure-ments oa the Wepman Auditory Discrimination Test; and 3) nine two-factor analyses of varianee on the Mashing Test.

Findincrs

Socioeconomic status and auditory discrimination abilit .It will be recalled that a major objective of this study was todetermine if there were social class difrerences lit auditory dis-

crimination ability. Several significant correlations from theoverall intercorrulation matrix (See Appendix 1, Table A) tendedto support the existence of differences In performanee on auditorydiscrimination tasks according to social class. Socioceonomic Sta-

tus (SES) was significanay correlated with initial phoneme (10and middle phoneme (MP) scores on the t-JADT (r=.18, pe_.05 In both

cases). There was a correlation of .18 () < .05) between SU, and

Total scores on the WADT. SES was also significantly and positively

22

20

Table 1

Variables Included in CorrelationalMatrices

A. Multiple Choice Bender Gestalt Test:1. Memory Score2. Matching Score

B. Wepman Auditory Discrimination Test (WADT):1. Initial Phoneme Score (IP)

Middle Phoneme Score (MP)3. Final Phoneme Score (FP)4. Total Score

C. IDS Pitch Discrimination Test:1. 3 Cycle Score2. 12 Cycle Score3. 17 Cycle Score4. 25 Cycle Score5. Total Cycle Score6. Hi Score7. Lo Score8. Same Score9. Total Hi/Lo/Same ScoreIDS Auditory Masking Test (MT):1. Word x Noise Score (WxN)2. Word x-Word Score OWO3. Nonsense x Noise Score (NxN)

E. IDS Revision of the Continuous Performance Test (CPT):1. Median Reaction Time: 100-1000 Milliseconds (RT 100-1000)2. Median Reaction Time: 100-2000 Milliseconds (RT 100-2000)3. Percent Reaction Time: 100-1000 Milliseconds (A 100-1000)4. Percent Reaction Time: 100-2000 Milliseconds (% 100-2009)

F. IDS Reading Prognosis Test (UT):1. Beginning Reading Score2. Language Score3. Perep':ual Discrimination Score4, Tot,.' 3core

G. Gates Diagnostic Battery:1. Raw Score2. Reading Grade Score3. Reading Age Score

H. Lorge-Thorndike IQ ScoreI. Chronological AgeJ. RaceK. Socioeconomic Status (SES)L. Grade in School

23

correlated with a number or Pitch Test measuves (25 cycles1:r = .16, p

r = .30, p 4..01; Hi:2 r = .21, p4 .05; Lo: r = .28, p4,C.05).Thus, middle SES black children did better on these tests thandid lower SES black children. There were no significant correla-tions on the intereorrelation matrix within the black group be-tween SES and any ol the Masking Test or CPT measures.

The, only significant correlation between SES and any ofthe auditory discrimination measures for whites was between SESand the CPT %100-2000 scores (r = .22, p

At the Fir!:t Ic I.cv c1 there were simfl'ipant correlations be-tween SES and WADT Initial phoneme and Total SCOPOS (P = .25,p.05; dn(1 P '4: .27, pe.0'), respectively), as well as betweenSES and one Pitch Test score (ho: r = .31, p .(15), and two CPTscores (RT 100-L000: P .25, p .05: RT 100-200(1: r = .27,p

The analyses of variance for the Pitch Test also sup-porVed thc hypothesis. Two four-factor analyses ol variance--grafle by socioeconomic status by race by pjtch cycle (Table 5),und grade by socioeconomic status by race by Hi/Lo trials(Table 6) --were periormed on the data. SES was one of the mainsignificant effects: F = 8.44, p

Table 6

Analysis of Variance For IDS Pitch Discrimination Test Scores inDiffc.rent Grades, SES Groups and Raoial Gronps OVQP Hi/Lo Trials

Souree d1 NSBuLween SsA (Grade)B (SES)C (Race)AB

1792

112

7.2460.84

.404.94

417.70***e-

AC 2 24.W 3.04'BC 1 14.00 1.82ABC 2 27.30 3.46*Error (between) 168 7.00

Within Ss 180D (Hi/Lo) 1 24.54 13.89***AD 2 6.01 3.11-0*BD 1 2.50 1.42CD 1 .28 4 1ABD 2 .53ACD 2 1.48 z. 1BCD 1 1.34 4 1ABCD 2 1.74 <Error (within) 168 1.77

*p< .05***p < 001

Table 7

Mean Number of Correct Rpsponses for the IDS Pitch DiscriminationTest for Lower and Middle Socioeconomic Groups

Socioeconomic GrouIDS Pitch Discrimination Test: Lower Middle

25 cycles 7.09 7.8717 cycles 5.93 7.0412 cycles 5.64 6.578 cycles 4.2P 4.27

Total over cycles 22.87 25.76Hi trials 1.72 2.33Le) trials 2.14 3.06Same trials 7.67 8.76Total over Hi/Lo/Same Vrials 11.53 14.20

25

Reading skill and auditory discrimination ability. A re-lated hypothesis stated that in order for a child to begin tolearn to read, a basic level of auditory discrimination must beattained. It was not possible to include reading performance asone of the variables in the overall intureorrelation matrix be-cause different reading tests were used in each of the threegrades (the IDS Reading Prognosis Test at the first grade lovel;the Gates Advanced Primary Reading Test, Type ArR, Form 2 at thethird grade level; and the Gates Basic Reading Survey, Type GS atthe fifth grade level). Reading performance, however, was inclu-ded as a variable in each of the three grade intercorrelation ma-trices, so that the relationship between reading and other vari-ables could be evaluated within each grade level.

Examining the grade intercorrelation matrices confirmed arelationship between reading and auditory discrimination abilityin all three grades, however, this relationship was strongest inthe first grade (See Table 8). Here, the language scores on theIDS Reading Prognosis Test (RPT) were significantly correlatedwith scores on the MT Word x Word (WxW) (r = -.32, p .05) andNonsense x Noise (NxN) (r = -.34, pe:: .01) Tests; WADT IP scores(r = .42, 13 .01), FP scores (r = .26, E) .05), and Total scores(r = .36, p

Tab

l8

CO

rrel

atio

ns B

etw

een

the

I 3 Reading Prognosis

Test

(First Grade) and Audito

/ Discrimination Scores

Begin

Perceptual

Total

Read"

12......_ka-LE122aL....aa9s12.1221121:1§.22...._

Wepman Auditory

Discrimination Test:

Initial Phoneme

.32*t

.48**

.49**

Middle Phoneme

.10

.21

.23

.21

Final Phoneme

.18

.26*

.26*

.28*

Total

.39**

.40**

Auditory Masking Test:

Word x Noise

.15

-.08

-.01

.05

Word x Word

.05

-.38**

-.23

Nonsense x Noise

-.06

-.28*

IDS Pitch Discrimination

Test:

25 cycles

.25*

.15

.19

.25*

17 cycles

.07

.04

.03

.05

12 cycles

.12

.18

.22

.20

8 cycles

-.04

.07

.07

..03

Total cycles

.04

.14

.13

.12

Hi

-.06

-.04

-.04

-.06

Lo

-.09

-.03

.02

-.05

Same

.29*

.11

.3V.

.30*

Total Hi/Lo/Same

.18

.06

.23

.20

Continuous Performance

Test:

R T 100-1000

msecs.

-.11

-.04

-.15

-.13

R T 100-2000

msecs.

-.15

-.09

-.21

-.19

513100-1000 msecs.

.38**

.26*

.46**

.46**

100-2000 msecs.

36**

20

48**

.43**

*p< .05

**p< .01

p .05) ; with Pitch Test_ 25 eyele (r = .25, p,.05) and Samescores (r = .30, p

Table 9

Correlations Between the C-tes Diagnostic Batterya

(Third and Fifth Grades) and Au.:itory Discrimination Scores

T,ird Grade

Fifth Grade

Reading

Grade

Raw

Score

Reading

Age

Reading

Grade

Raw

Score

Reading

Ace

Wepman Auditory Discrimination Test:

Initial Phoneme

.25*

.27*

.02

.10

.14

.22

Middle Phoneme

.26*

.25*

.06

.07

.07

.03

Final Phoneme

.18

.17

-.10

.02

.04

.08

Total

.29*

.29*

.14

.18

.27

Auditory Masking Test:

Word x Noise

-.25*

-.22

.01

.01

-.01

.14

Word x Word

-.25*

-.22

-.14

-.05

-.07

-.01

Nonsense x Noise

.01

.01

-.18

.07

.05

.08

IDS Pitch Discrimination Test:

23 cycles

.11

.10

-.14

.24

.32**

.10

17 cycles

.09

.11

.-.05

.23*

.34**

.03

12 cycles

.01

.01

-.21

.22

.32*

-.06

8 cycles

-.03

.13

.13

.02

-.04

Total cycles

.07

.07

-.11

.33**

41**

-.01

Hi

-.05

-.06

.04

.22

.32*

-.10

Lo

.09

.10

.00

.34**

.38**

.09

Same

-.03

.29*

.36'.:;*

.02

Total Hino/Same

-.01

.00

-.13

.33**

41**

.00

Continuous Performance Test:

R T 100-1000 msecs.

-.32*

-.36**

-.06

-.08

.06

.06

R T 100-2000 msecs.

-.34**

-.39**

-.06

-.08

.06

.05

100-1000 msecs.

.15

.19

-.04

.16

.13

.15

100-2000 msecs.

12

.15

-,02

.16

.10

.22

aThe Gates Advanced Primary Reading Test, Type APR, Form 2 was administered at

the first grade level; the Gates Basic Reading Survey, Type GS was administered

at the fifth grade level.

'10

Relationship betw('en Vit,mal and Auditory ')iseriminationSkills. Another ohjeetive or this study was concerned with ex-ploring the possible relationship between levels uf auditory dis-crimination sktlls and the visual pereeptuat skills involved inreading. Examining the overall intercorreiation matrix again(Appendix 1, Table A), it can be seen that all of the correla-tions between the Multiple Choice Bender-Gestalt and the WADTwere signiFicant: the range of correlations was from r = .19,p .05 (Matching with MP) to r = .45, p .01 (Matching with IP).All the c:orrelations between the Multiple Choice Bender-Gestaltand the CPT were also significant, ranging from r = .22 to .37,p < .01; as were all the correlations with the Masking Test(r = .23 to .38, pe.;.01). There were only four significant corre-lations between the Multiple Choice Bender-Gestalt and the PitchTest and these all involved the Matching slbtest (25 cycles:r = .22, p4.01; Total cycles: r = .19, pcf.05; Same scores:r = .18, p4:.05; and Total Hi/Lo/Same scores: r = .16, p4C.05).

At the first grade level (See Table 10), there were threesignificant correlations between the Memory subtest of the Mul-tiple Choice Bender-Gestalt and the WADT (IP: r = .30, FP: r = .26,and Total: r = .31, p,:.05); and one signiricant correlation be-tween the Matching subtest and the WADT (IP: r = .30, 1)4.05).There were three significant correlations between the Memory sub-test of the Multiple Choice Bender-Gestalt and the CPT (RT 100-2000:r = .29, L).05; % 100-1000: r = .32, p.(` .05; % 100-2000: r = .26,p( .05). There were no significant correlations at the firstgrade level between the Multiple Choice Bender-Gestalt and any ofthe Masking Test or Pitch Test scores. The other nonsignificantcorrelations were the Memory subtest scores with the WADT MPscores, Matching scores with WADT MP, FP, and Total scores, Mem-ory scores with CPT RT 100-1000 scores and Matching scoreswith all CPT scores.

At the third grade-level (Table 10) there were two signifi-cant correlations between the Multiple Choice Bender-GestaltMatching scores and the WADT IP and MP scores (r = .27, p4.05,r = .36, p.'.01 respectively), and one between the Memory subtestscores and the CPT RT 100-2000 scores (r = .27, pd.!. .05). Therewere no significant correlations between the Multiple Choice Ben-der-Gestalt scores and either the Masking Test or Pitch Testscores, nor were the Memory subtest scores significantly correla-ted with any of the WADT scores, the CPT RT 100-1000 scores oreither of the CPT % scores. Matching scores were not significant-ly related to WADT IT and Total scores n.31, to any of the CPTscores.

In the fifth grade, three WADT scores were significantlyrelated to the Multiple Choice Bender-Gestalt: IP scores withMemory scores and Matching scores (r = .28, p.:".05; and r = .50,

respectively) and FP scores with Memory scores (r = .29,p .05) The Memory subtest scores were also significantly corre-

Tab): 10

CLrrelations BetweenAuditory Discriminatior Measures and Visual

Discrimination Measures

(Multiple Choice Bender-:estait) for Each

Grade

Paci

eGrac.e

5

Memory Match lg

Mwc,ory

Matching

Y:atchin-

Wepman Auditory Discrimination Test:

Initial Phoneme

.30*

.30*

.21

.27*

.28*

Middle Phoneme

.17

A2

.10

.36**

.09

,07

Final Phoneme

.26*

.10

.14

.02

.29*

.09

Total

.31*

.19

.06

.23

.1!:-

.19

Auditory Masking Test:

Word x Noise

.01

.04

.21

.14

.03

.06

Word x Word

.02

.01

.16

.16

.20

.22

Nonsense x Noise

.08

.06

.03

.13

.13

.06

IDS Pitch Discrimination Test:

25 cycles

.02

.23

.04

.03

.22

.22

17 cycles

.24

.02

.12

.04

.30*

.22

12 cycles

.15

.07

.08

.03

.23

.07

8 cycles

.13

.07

.08

.09

.21

.20

Total cycles

.11

.09

.07

.01

.24

.19

Hi

.17

.02

.10

.03

.19

.13

Lo

.23

.00

.06

.03

.27*

.10

Same

.02

.17

.00

.03

.22

.21

Total Hi/La/Same

.11

.13

.05

.01

.26*

.20

Continuous Performance Test:

R T100-1000 msecs.

.21

.14

.22

.04

.16

.13

R T 100-2000 msecs.

.29*

.18

.27*

.07

.15

.12

% 100-1000 msecs.

.32*

.24

.07

.15

.16

.13

% 100-2000 msecs.

.26*

.21

.01

.14

.24

.21

*p G. .U5

**p< .01

lated with three Pitch Test scores: 17 cycle scores, TA) scores,and Total Hi/to/Same scores (r = n30, .27, and .26, respectively,p < .05) . There were no significant correlations between the Mul-tiple Choice Bender-Gestalt and either the Masking Test or CPTscores. Neither Memory nor Matching scores were significantlycorrelated with either WADT MP or Total scores. Matching scoreswere, in addition_ nut significantly correlated with WADT FPscores. There were no significant correlations between the Match-ing subtest scores and any of the Pitch Test scores. Memory sub-test scores were not significantly correlated with either 25, 12,or 8 cycle scores, Iii scores, or Same scores.

The relationship between auditory and visual discriminationwas about as strong for middle as for lower SES children (SeeTable 11). There were 21 significant correlations out of 22 forlower SES children and 17 out of 22 for middle SES children be-tween the Memory and Matching scores of the Multiple Choice Bender-Gestalt and the WADT, Masking Test, and CPT scores. However, there

were only three significant correlations between Pitch Test scoresand the Matching scores for middle SES children and none betweenPitch Test scores and either the Memory or Matching scores for

lower SES children.

The relationship between auditory and visual discriminationwas also about equally as strong for black as for white subjects(Table 12). There were 21 significant correlations out of 22 forblacks and 19 out of 22 for whites between the Memory and Matchingscores of the Multiple Choice Bender-Gestalt and the WADT, MaskingTest, and CPT scores. Again, however, while there were no signi-ficant correlations between the Multiple Choice Bender-Gestalt andthe Pitch Test for blacks, there were eight significant correla-tions between the Multiple Choice Bender-Gestalt and the Pitch

Test for whites.

It would seem from'the results cited above that the rela-tionship between auditory and visual discrimination skills is alinear and positive one, at least for the subjects in this study.

Related findings. Although not central to the hypothesesof this study, a number of other findings are of peripheral in-

terest. There were several significant correlations betweenreading test scores and visual discrimination scores, 1Q, Race,

and SES (Table 13). At the first grade level there were signifi-cant correlations between all of the reading test scores and bothMultiple Choice Bender-Gestalt subtests, i.e., eight significanttorrelations ranging from r = .36 to .58, 1)4:1..01). There werealso significant correlations between all of the reading testscores and SES (range: r = .26, p.L.05 to r = .38, p.01);Race (range: r = .27, p 4.05 to r = .47, p.::..01); and IQ(range: r = .33 to r = .45, p At the third grade levelthere were four significant correlations out of six correlationsbetween the Multiple Choice Bender-Gestalt and the reading test

2.5

1?

Tab:

11

Correlations Between Auditory Discriminatioi Measures

and Visual Discrimination Measures

(Multiple Choice Bender-Gesi lt) According to SES

Groups

Momol

iddle SES

Matehincr

Lower SES

Lower SES

Iiiddle SES

Wepman Auditory Discrimination Test:

Initial Phoneme

.38**

.38**

.52**

Middle Phoneme

.24*

.12

.22*

.01

Final Phoneme

.23*

.21

.27*

Total

.27*

.38**

.23*

.48**

Auditory Masking Test:

Word x Noise

.25*

.22*

.23*

Word x Word

.36**

Nonsense x Noise

.26*

.25*

.27*

24*

IDS Pitch Discrimination Test:

25 cycles

.13

.11

.16

.25*

17 cycles

.01

.08

04

.17

12 cycles

.01

.11

.09

.08

8 cycles

.03

.03

.07

.13

Total cycles

.05

.18

.14

.20

Hi

.05

.06

..04

.02

La

.00

.01

.08

.06

::ame

.12

.12

.02

Total Hi/La/Same

.08

.10

.05

.24*

Continuous Performance Test:

T 100-1000 msecs.

.17

.30**

.09

R T 100-2000 msecs.

.23*

.38**

.13

% 100-1000 msecs.

.40**

.41**

.40**

% 100-2000 msecs.

.30**

.34**

.40**

wp 4. 0

**p el .01

Table 12

Correlations Between Auditory Discrimination Measures and Visual Discrimination

Measures

(Multiple Choice Bender-Gestalt) According to Racial Groups

Memory

Matching

Blacks

Whites

Blacks

Whites

Wepman Auditory Discrimination Test:

Initial Phcneme

.33**

.50**

.43**

.46**

Middle Phoneme

.23*

.22*

.14

.29*

Final Phoneme

.30**

.23*

.25*

.19

Total

.35**

.25*

.36**

.24*

Auditory Masking Test:

Word x Noise

.34**

.23*

.31**

.16

Word x Word

.30**

CA,

Akj

Nonsensta x Noise

IDS PiLch Discrimination Test:

.34**

.18

.26*

.27*

25 cycles

.09

.16

.15

.28*

17 cycles

.03

.18

.03

.26*

12 cycles

04

.22*

.02

.22*

8 cycles

.00

.00

.09

.08

Total cycles

.06

.21

..12

.28*

Hi

.06

.14

.00

.12

Lo

.08

.18

.08

.25*

Same

.04

.18

.12

.20

Total Hi/Lo/Same

.02

.22*

.06

.25*

Continuous Performance Test:

R T 100-1000 msecs.

.28*

.34**

.23*

.24*

R T 100-2000 msecs.

.35**

42**

.30**

% 100-1000 msecs.

.38**

.47**

.44**

36**

% 100-2000 msecs.

29**

43**

.38**

34**

*p4 .05

**p 4: .01

Table 13

Correlations Between Reading Measures

and Visual Discrimination

Measures

(MUltiple Choice Bender-Gestalt),

IQ, Race, and SES for Each Grade

Memor

Matchin

10

Race

SES

First Grade:

Beginning Reading

.44**

.39**

.27*

.26*

Language

.38**

.37**

.47**

.33**

Perceptual Discrimination

.58**

.33**

.37**

.34**

Total

.50**

.58**

.45**

.44**

.38**

Third Grade:

Reading Grade

.26*

.27*

.54**

.04

Raw Score

.24

.25*

.51**

.04

Reading Age

15

.31*

.00

.16

.05

Fifth Grade:

Reading Grade

.11

.04

.56**

34**

.28*

Raw Score

.06

.10

61**

.42**

.21

Readincr Ate

07

02

.17

35**

.16

*p.05

**p

scores (range: P = .25 to r = .31, p....05). Two of the three eor-roLtLiens hetween SES and reading test :;COPOS were sinilicant(o = ,49 and .47, pz:.1.01), as wore two or the three euerelations

twt,en I (2 (old toNt(U )):_ Les L scores (11 , am. .51, p . 0.1.) .There were no sinilicant correlations between race and readingtest scores at the third grade level. At the fifth grade levelthere wore no signiiicant correlations between the MultipleChciec Bender-Gestalt and any of the reading test measures. Therewa!-; one sinificant correlation, however, between SES and readir:.;test scores (r .28, pz=. .05) , three significant correlations be-tween race and readinf.; Lest scores (range: r = .34 to r = .42,pz.(11)0 and two significant correlations between 1(2 and readingtest scores (r = .56 and .61, p,,..01).

On the overall intercorrelation matrix (Appendix I, Table A)there was a number of significant correlations between auditorydiscrimination measures and age, grade, and 10, while there wereonly a few between auditory dicrimination measures and race.Older children, as opposed to younger ones, uid better on allfour WADT scores (range: r = .21 to r = .35, p/ .01) , on thethree Hlsking Test scores (range: r = .35 to r = .51, p:1.01),on the Total cycles Pitch Test scores (r = .22, p .01) , and onthe four scores of the CPT (z'ange: r = .39, to r = .52, p,::..01).This same cluster occurred for the grade variable: children inthe higher grades had higher scovez; than children in the lowergrades on all WADT scores (range: r = .22 to r = .34, pZ..01),all Masking Test scores (range: r = .36 to r = .52, p4.:.01), the

- . 1 p,.01), and all CPTscores(range: r = .38 to r = .52, p4.01). IQ was related toauditory discrimination as indicated by the correlations betweenIQ and all four WADT scores (range: r = .16 to r = .19, 1)4405),one of the Masking Test scores (td/: r = .25, p/:'..01), the Totalcycles Pitch Test scores (r = .25, pz::.01), and the Total Hi/1,o/Same Pitch Test scores (r = .29, p,.;1.01), and the two % CPTscores (,6 100-1000: r = 418, p.--.05; % 100-2000: r = .16, p-:.05).The only significant correlations between race and auditory dis-crimination scores were with three subtest scores of the PitchTest (25 cycle and 8 cycle scores: r = .17, p.(Z.05 in both cases)and Same scores: r = .23, pz:.01).

There were also several significant correlations betweenvisual discrimination measures and age, grade, IQ, and SES (Ap-pendix I, Table A). The correlations between age and the Memoryand Matching subtests of the Multiple Choice Bender-Gestalt wererespectively r = .54 and .52, p-:.01; between the Memory andMatching subtests and grade the correlations were r = .56 and.53, p.C.01. Similarly, IQ and the Memory and Matehing subtestswere significantly correlated (r in both cases = .36, p/:.01).While there were no signilicant correlations between the MultipleChoice Bender-Gestalt and race, the Matching test of the MultipleChoice Bender-Gestalt was significantly correlated with SES(r = .18, p

The r(q;ults Were also examined in a very general way toSQ0 if, over all differenees between SES rroups' scoreswere_ greater within tho black groups or within white samples, andolso to see whether the ditlerences between black and white ;1!oupswero greater for the lower SES Samples or for the middle SESsamples.

Confli2ting results are to be found in the literature forsuch comparisons. A study which attempted to tease out ethnicand class variables in level and patterning of mental ability wasreported by Lesser, fifer, and Clark (1965) . They studied middleand lower-class samples of first grade children in each of fourdifferent ethnic-racial groups: Blacks, Chinese, Jews, and PuertoRicans. They found that both SES and cultural background wererelated to level of score on the Hunter tests, a specially con-strueted scale that yielded ratings of verbal ability, space con-ceptualization, reasoning and numerical ability. Within each ethnicgroup the middle-class children did better than the lower-elasschildren. While the ethnic groups differed in their performancelevels, for the entire sample, middle-class children were signi-ficantly superior to lower-class children on all scales and sub-tests. Within the black group there was a greater difference inscores between class levels than was true for any of the otherthree groups. Within class groups and across ethnic groups therewas a greater similarity among the four ethnic groups than amongthe lower-class groups. In a replication of this study by Stodol-sky and Lesser (1967) with lower and middle-class Chinese, black,and IL,isn Catholic children in Boston, results were again that SESdifferences within the black group were greater than those withineither of the other two groups.

These findings are in contrast to the results of studiesreported by the Institute for Developmental Studies (Deutsch &Brown, 1964; Whiteman, Brown & Deutsch, 1967; Whiteman & Deutsch,1967) which compared clases and races on IQ scores and on a num-ber of skills thought to underlie and/or relate to intellectivefunctioning. Using the Lorge-Thorndike Test (non-verbal form)and the Vocabularly subtest from the Wechsler Intelligence Scalefor Children (WISC), Deutsch and his associates gathered data on543 children in a random stratified sample including equal sub-samples of black and white, first grade and fifth grade childrenat SES levels I, II, and Il on the Institute's SES scale. Theyfound, in contrast to the Lesser, Fifer, and Clark (1965) andthe Stodolsky and Lesser (1967) findings, that black and whitegroups were farther apart at each higher SES level, and that therewas less differentiation between SES groups in the black, asopposed to the white sample.

Results for the present project are more in accord withthe institute's previous findings (Table 14). At the firstgrade level the differences between lower and middle-class groupswere greater for whites than for blacks on 18 out of 27 scores and

40

at the third :%rade level on 18 oF the 28 scores. Thus on approN-111.ately Lwo-thirds ot Lhe scores thetie were greater differences

between lower and middle-elass groups' scores For whites than. loebiNcks, i.e., there was less difFereutiatiou between SLS groupsin the bl.aek. as opposed to the white sample at both the first and

third grade levels. At the I wade level, however, on 19 olthe 28 scores, differences between lower and middle SLS groups'scores were greater for blacks than for whites.

Table 1/1

Frequencies of Differences Between SES Groups Within Race,and Between Racial Groups Within SESa

FirstGrade

Fat).ThirdNumber of Differences Between SESGroups Within Racial Groups

Black (bvsM) White (LvsM) 9/27 10/28 19/28

White (LvsM) Black (LvsM) 18/27 18/28 9/28

X2 3.00 2.28 3.58

Number of Differences BetweenRacial Groups Within SES Groups

Lower-Class (BvsW) Middle-Class (BvsW) 11/27 8/28 15/28

Middle-Class (Bvsh) Lower-Class (BvsW) 16/27 20/28 13/28

X2 92 5 14* .14

*p

PCVPPS(ud. OH fifteen OHL or the 28 scores, diftcrenecs be-tween racial groups Ire greater at the lower SES level than atthe mi d die SES level.

Most of the auditory discrimination measures were inter-eorrefaLed. The Wepman Auditory Discrimination Test was includedin this study us a basic measure ol auditory discrimination.Standardization data are available for the WADT, and in earlierstudies at: the 3nstitute for Developmental Studies (Deutsch(1964) reviews some of this research) this test significantly dis-criminated good from poor readers. The reliability (Kuder-Richardson formula) for first graders5 on the WADT was .85. TheWepman was signiFleantly correlated (Appendix I, Table A) withscores on the Masking Test (11 of the 12 correlations were signi-ficant -- range: r = .16, .05 to r = .32, p< .01) ; with theCPT scores (12 of the 16 correlations were significant -- range:r = 17, p.05 to r = .48, 1).01); and with the Pitch Test scores(15 of the 36 correlations were significant -- range: r = .16,p. .05 to r = .28, L) .01).

The IDS Pitch Discrimination Test was developed specificallyfor this study. It was included in the test battery so that audi-tory discrimination could be broken down into "purer" sound dis-crimination ability us distinguished from the verbal discrimina-tion tapped by the WADT. However, as was just noted, there was anumber oi significant and positive correlations between subtestson the WADT and subtests on the Pitch Test (15 out of the 36 cor-relations -- Appendix T. Table A). Nine of the 27 correlations

j.L;.:)L Wid LAIL! Nasking Test were significant(range: r = .16, r:.'.05 to r = .26, p

-Led by 25 cycles, mean number of correct responses diminishinsuccesiveJy from 17, to 12, to 8 cycles (the mean scores were7.48, 6.4) , 6.12 and 4.23 respectively -- Appendix 11, Table A.Many subtest scores on the Pitch Test were signilicantly correla-ted with each other (Appendix f, Table A). So, for example, amongcycle responses, the number of correct responses made when twotones were separated by 12 cycles was significantly correlatedwith the number of correct responses made when two tones wereseparated by 17 cycles (r = .75, p4.01), and by 25 cycles(r = .62, pc.. .01) . Seventeen and 25 cycle responses were alsosignificantly correlated with each other(r = .68, pe:.01). ThereWas a noticeable lack of significant correlations, however,between 8 cycle responses and responses made at any other cyclelevel. From this it may be inferred that it wus very difficultfor subjects to discriminate two tones only 8 cycles apart. Theinteraction between SES and Cycle was 41so statistically signi-ficant (F = 2.91, p

As has already been reported, there werecorrolatiors between the M;iskin;!, Test and the MDT :-;eues(11 out ol 1;0; and the Maskilw, Test and the Piteh Test 0 out oF27) . There were also 11 sinilicant cocrefations (nit of the 12correlations between the Masking Test and the CPT (range:r = .16, p.05 to r = .31, p.:.0]). These eurrelalions offeredsome evidence For the validity oi the Masking Test as a measnreof auditory discrimination.

As has been mentioned, scores on the CPT were significantlycorrelated with WADT scores (11 out of 1(i), Masking Test scores(11 out of 12) and Pitch Test scores (16 out of 36). Its correla-tions with other measm!es oL auditory discrimination would seemto indicate that the CPT, while a measure of attention and vigi-lance, is also a good measuru of auditory discrimination.

Table 15Analysis of Variance of IDS Auditory Masking Test Scores,

Word x Noise, First Gradea

Source dFBetween

Within

.05

51

rr, rr

ms615.620 2.638*

233.360

aFour subjects were omitted from the first grade Masking Testanalyses of variance: one because scores were not obtainable onone of the three Masking Tests, and three because they were theonly subjects from one of the schools represented.

Table 16Analysis of Variance of IDS Auditory Masking Test Scores,

Word x Word, First Grade

SourceBetween

Within

Total

di

51

55

MS275.247

418.120

0.658

44

Table 17Analysis of Variance of IDS Auditory Masking Test Scores,

Nonsense x Noise, First Grade

SourceBetween

Within

df ms r4 317.453 .419

756.90951

Total 55

Table 18Analysis of Variance of IDS Auditory Masking Test Scores,

Word x Noise, Third Grade

Source MSBetween 5 69.04 .296

Within 54- 232.84

Total 59

Table 19Analysis of Variance of IDS Auditory Masking Test Scores,

Word x Noise, Third Grade

Source di MSBetween 4 325.17 1..134Within 55 286.67

Total 59

Table 20Analysis of Variance of IDS Auditory Masking Test Scores,

Nonsense x Noise, Third Grade

SourceBetween

Within

cif

55

Total 59

MS127.78

264.69

F.482

45112

Table 21Analysis or Variance or IDS Auditopy Masking Test SOOPCS,

Word x Noise, Fifth Gradea

Sotwee MSBetween 3 346.113 .5611.

Within 50 204.372

Total 53USix subjects WeVe omitted from the fift:h grade Husking Testanalyses of variance because they rep'resented two schools withonly four subjects in one and two subjects in the other.

Table 22Analysis of Variance of IDS Auditory Masking Test Scores,

Word x Word, Fifth Grade

11M1.11..1,41/.....M.M. =111.

Source df MSBetween 3 681.891 3.017*

Wittuin SO 225.963

Total 53**1) .05

Table 23Analysis of Variance of IDS Auditory Masking Test Scores,

Nonsense x Noise, Fifth Grade

Source df MS rBetween 3 19.235 .081

Within 50 237.195

Total 53 IMM.IMMI11.=111.70.1.11111441.W.M..

46 113

Chapter 4

Discilssjon and Concluslons

In view of the large number of correlations reported inthe results section. it may aid the reader to glance at the fol-lowing summary of results principally connected with the hypo-theses of this study.

Res tilts Supporting a Relations14j etweeii S ES and Audi torDi scrimi nation

Number of significant correlations between SES and audi-tory discrimination:

WADT Masking Pitch CPT TotalAll Subjects: _0 5 0 8First Griide: 2 0 2 2 6Third Grade: 1 0 1 0 2Fifth Grade: 0 0 1 0 1i; / .s, ,1.. .. , ,..! .

1 0 4 0 5White Subjects: 0 0 0 1 1

Total 7/24 0/18 13/54 3/24 23/120

Analyses of Variance:

WADT: SES main effect for first grade analysisPitch Test: SES main effect -- Grade x SES x Race x Cyoles

SES main effect -- Grade x SES x Race x HYLo Trials

Results Sup:porting a Relationship Between Auditory Discriminationand Reading

Number of significant correlations between reading scoresand auditory measures:

WADT Masking Pitch CPT TotalFirst Grade: 11 5 3 7 26Third Grade: 6 1 0 4 11Fifth Grade: 1 0 13 0 14

Total 18/40 6/30 16/90 11/40 51/200

4745

Results Stipporting a TZ(lationship 13(!tcen Visual and AuditoryDiscrimination

Number of significant correlations between theChoice Beuder-GestalL and auditory measures:

All Subjects:First Grade:Third Grade:Fifth Grade:Black Subjects:White Subjects:Lower-Class Subjects:Middle-Class Subjects:

ltipie

WADT Maskin Pitch CPT Total8 6 4 8 40

o 0 3 70 0 1 30 0 o 36 0 8 214 8 8 266 0 8 21.6 3 5 20

28/48 15/144 41/64 127/320

42

3

7

6

7

6

43/64

Results Supportina a Relationship Between Auditory Discriminationand IQ, Race, and Grade

Number of significant correlations between auditory mea-sures and IQ, Race, and Grade:

All Subjects: 9/20Race3/20

Grade12/20

Results Supportinu a Relationshi, Between Visual Discriminationand I Race, SES, and Grade

Number of significant correlations between visual dis-crimination measures and IQ, Race, SES, and Grade:

All Subjects:Race SES0/2 1/2

Grade2/2

Results Supportinga Relationship Between Reading, and I9L Race,S'3S, and Visual Discrimination

Number of significant correla-Wns between reading scoresand IQ, Race, SES, and the Multiple CRoice Bender-Gestalt:

Race SESMultiple ChoiceBender-Gestalt

First Grade: 4 4 4 8Third Grade: 2 0 2 4Fifth Grade: 2 3 1 0

Total 8/10 7/10 7/10 12/20

4646

Valle the experimenfal design ol this study does nut purmitthe dclinition UI eanse-eireet relationships, it dues make POS--sible some pinpointin of attrihutes mor charactertle or oneexperimentally delinv(i !,rocip than of another, wnd of those vari-ables ,Iore closely associated with one area of ability than withanother.

The results tend to conFirm the expectation that poorauditory uiscriminotion is one factor distinuishing lower SESchildren from middle SES children. That a relationship betweenSES and auditory discrimination exists in the data js seen in thesignilicant correlations on the overall matrix between SES and theWepman Test ot Auditory Discrimination (WADT), and the 11)8 PitchDiscrimination Test, as well as in the significant main effect ofSES in the analyses of variance of the.Pitch Test and the MDT.These Findings are consistent with the expectation of greaterdifficulty with these tests for children Irom noisy slum environ-ments with a low rate of verbal interchange, and a high noise/lowsignal ratio. Such conditions are not conducive to the earlydevelopment of good auditory discrimination skills. The findingsare also consistent with the literature. Those studies reviewedin the Introduction which investigated the'relationship betweenSES and auditory discrimination ability (Clark & Richards, 196(;McArdle, 1965; Giebink & Marden, 1968), though few in number, allfound that middle-class nursery school children did better onauditory discrimination tasks than did their disadvantaged counter-parts. The current results lend support to those earlier findings])v tPM tO older age

The stronger relationship between SES and auditory discrim-ination for blacks than for whites may reflect the possibilitythat a disadvantaged environment has more deleterious effects forthe black child than for the white child. Also possible, andmore likely, is that the disadvantaged black child's environmentmay be more disadvantaged.than the disadvantaged white child's,creating a greater difference between SES groups for blacks thanfor whites, and thus the possibility of a stronger correlation be-tween SES and auditory discrimination for black children.

The decrease with age of the relationship between SES andauditory discrimination may reflect gains in general discrimina-tion ability acquired perhaps through school learning experiences.The ceiling effect on the WADT also suggests this interpretation.

Neither the Masking Test nor the Continuous PerformanceTest yielded the SES discrimination that had been anticipated.The WADT is more dependent than the Masking Test or the CPT onverbal discrimination as opposed to sound discrimination, andthis dependence on verbal stimuli might account for its differen-tiating between social class groups. In addition, the WADT dif-

49It"

fers From the CPT and Maskiiv; Tests in its demand For reFineddiscriminations between vowel and consonant sounds as opposed todiscriminations wIlich locus on different_ 1.evels of meaningiutness otmaterials. The Masking Test, for example, was constructed in partin fact_ along a dimension oF meaningLniness (word by noise, wordby nonsense syllable, word by word) Lower SES children, as com-pared to middle SES children, then, do not have greater difficultydiscriminating between words of highly differentiated meaningswhich also sound quite different. On the CPT,