factor structure of scores from the conners’ …

251
The Pennsylvania State University The Graduate School College of Education FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ PARENT AND TEACHER RATING SCALES – REVISED AMONG CHILDREN IN NEPAL A Dissertation in School Psychology by Laura L. Pendergast © 2011 Laura L. Pendergast Submitted in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy August, 2011

Upload: others

Post on 02-Dec-2021

6 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

The Pennsylvania State University

The Graduate School

College of Education

FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ PARENT AND

TEACHER RATING SCALES – REVISED AMONG CHILDREN IN NEPAL

A Dissertation in

School Psychology

by

Laura L. Pendergast

© 2011 Laura L. Pendergast

Submitted in Partial Fulfillment

of the Requirements for the Degree of

Doctor of Philosophy

August, 2011

Page 2: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

ii

The dissertation of Laura L. Pendergast was reviewed and approved* by the following:

Beverly J. Vandiver Associate Professor of Education Dissertation Adviser Chair of Committee Barbara A. Schaefer Associate Professor of Education Pamela M. Cole Professor of Psychology Cynthia L. Huang-Pollock Assistant Professor of Psychology James C. DiPerna Associate Professor of Education Professor-in-Charge, School Psychology Program

*Signatures are on file in the Graduate School.

Page 3: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

iii

Abstract

Attention Deficit Hyperactivity Disorder (ADHD) is a psychiatric disorder, which is marked by

symptoms of inattention, hyperactivity, and impulsivity, and has been identified in children

throughout the world (Polanczyk & Rohdes, 2007). The Conners’ Teacher Rating Scale –

Revised (CTRS-R; Conners, 1997) and Conners’ Parent Rating Scale – Revised (CPRS-R;

Conners, 1997) are behavior rating scales that are commonly used to assess symptoms of ADHD

worldwide (Conners, 1997). However, the factor structures of scores from the long forms of the

Revised Conners’ scales have not been independently examined. This study examined the factor

structures of scores from the long forms of the Conners’ Teacher Rating Scale – Revised and the

Conners Parent Rating Scale – Revised. Ratings were provided by 1,387 teachers and 1,835

parents of children living in the Sarlahi district of Nepal. To test the stability of scores and

identify competing models, the overall teacher and parent samples were randomly divided into

subsamples, which were used for exploratory and confirmatory factor analyses, respectively.

Exploratory factor analyses were used to identify competing factor solutions: several reduced,

two-factor solutions for both scales. Ten models (five CTRS-R models and five CPRS-R

models) were tested using confirmatory factor analysis (CFA), and the findings indicated that the

reduced two-factor models of CTRS-R and CPRS-R scores were a better fit to the Nepalese data

than the original models based on a US sample that was identified by Conners. Inattention and

Hyperactivity factors were identified for CTRS-R scores, and Inattention and Oppositional

factors were identified for CPRS-R scores. These findings raise questions regarding the most

appropriate uses of the scales in Nepal and other non-Western nations.

Page 4: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

iv

Table of Contents

List of Tables……………….................................................................................................. vi List of Figures……………………………………………………………………………..... viii Acknowledgements………………………………………………………………………… ix INTRODUCTION………………………………………………………………………….. 1 Cross-Cultural Assessment …………………………………………………….................... 1 The Cultural Context of Nepal…………………………………………………………........ 4 LITERATURE REVIEW………………………………………………………………....... 14 General Information about ADHD ……………………………………………………….... 15 Cultural Differences and ADHD............................................................................................. 22 The Conners’ Rating Scales.................................................................................................... 30 Present Study........................................................................................................................... 46 METHOD…………………………………………………………………………………… 48 Sample………………………………………………………………………………............. 48 Measures…………………………………………………………………………………..... 51 Procedure………………………………………………………………………………….... 55 Data Management and Analysis…………………………………………………………..... 57 RESULTS............................................................................................................................... 62 CONNERS’ TEACHER SCALE............................................................................................ 62 Preliminary Analyses.............................................................................................................. 62 EFA......................................................................................................................................... 62 CFA......................................................................................................................................... 79 CONNERS’ PARENT SCALE.............................................................................................. 93 Preliminary Analyses.............................................................................................................. 93 EFA......................................................................................................................................... 93 CFA......................................................................................................................................... 114 DISCUSSION………………………………………………………………………………. 127 General Discussion................................................................................................................. 127 Conners’ Teacher Scale.......................................................................................................... 130 Conners’ Parent Scale............................................................................................................. 132 Scale Comparisons.................................................................................................................. 134 Implications............................................................................................................................. 135 Limitations.............................................................................................................................. 141 Future Research....................................................................................................................... 142 Conclusions............................................................................................................................. 144 REFERENCES…………………………………………………………………………….... 146 APPENDIX A……………………………………………………………………………..... 169

Page 5: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

v

Back Translation Evaluation Form......................................................................................... 169 APPENDIX B………………………………………………………………………………. 198 Item Translation Evaluation.................................................................................................... 198 APPENDIX C......................................................................................................................... 200 Means, Standard Deviations, and Item-Level Correlation Matrices or Parent and Teacher Items........................................................................................................................................ 200 APPENDIX D......................................................................................................................... 232 Model Identification Criteria.................................................................................................. 232 APPENDIX E......................................................................................................................... 233 Conners’ Items Subscales and Derivation Methods............................................................... 233 APPENDIX F.......................................................................................................................... 238 Glossary of Acronyms............................................................................................................ 238

Page 6: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

vi

List of Tables

Table 1. Demographic Frequencies for Teacher- and Parent-Rated Samples by Caste,

Region of Ancestry, and Sex...................................................................................... 50 Table 2. Frequencies and Percentages for School Status of Children in the Parent-Rated

Sample by Caste, Region of Ancestry, and Sex......................................................... 51 Table 3. Means and Standard Deviations of Items on the Conners’ Teacher Scale................ 63 Table 4. Summary of the Initial Two-Factor Solution of CTRS-R Scores: Pattern

Coefficients and Alternative Solutions....................................................................... 71 Table 5. Pattern Coefficients from EFA of the CTRS-R Rational-2T..................................... 76 Table 6. Pattern Coefficients from EFA of the CTRS-R Statistical-2T Solution.................... 77 Table 7. Pattern Coefficients from EFA of the CTRS-R Hybrid-2T Solution......................... 78 Table 8. Summary of CTRS-R Models Tested through CFA.................................................. 81 Table 9. Standardized Pattern Coefficients from CFA of the Rational-2T Model of the

CTRS-R...................................................................................................................... 86

Table 10. Standardized Pattern Coefficients from CFA for the Statistical-2T Model of the CTRS-R................................................................................................................... 87

Table 11. Standardized Pattern Coefficients from CFA of the Hybrid-2T Model of the

CTRS-R................................................................................................................... 88 Table 12. Summary of Fit Statistics for CFAs of CTRS-R Models......................................... 91 Table 13. Findings from Satorra-Bentler χ2 Difference Tests with Bonferroni Correction..... 92 Table 14. Means and Standard Deviations of Items on the Conners’ Parent Scale................. 94 Table 15. Summary of the Initial Two-Factor Solution of CPRS-R Scores: Pattern

Coefficients and Alternative Solutions.................................................................... 104 Table 16. Pattern Coefficients from EFA of the CPRS-R, Extended-2P Solution.................. 110 Table 17. Pattern Coefficients from EFA of the CPRS-R Abbreviated-2P Solution............... 112 Table 18. Pattern Coefficients from EFA of the CPRS-R Hybrid-2P Solution....................... 113 Table 19. Summary of CPRS-R Models and Rationale........................................................... 115

Page 7: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

vii

Table 20. Standardized Pattern Coefficients from CFA of the Extended-2P Model............... 120 Table 21. Standardized Pattern Coefficients from CFA of the Abbreviated-2P Model.......... 121 Table 22. Fit Indices for CFA of CPRS-R............................................................................... 125 Table 23. Findings from Satorra-Bentler χ2 Difference Tests with Bonferroni Correction..... 126 Table 24. Poorly Translated Items as Identified by Expert Review......................................... 198 Table 25. Correlation Matrix of Items from the EFA Sample of the CTRS-R Scores........................ 200 Table 26. Correlation Matrix of Items from the CFA Sample of the CTRS-R Scores........................ 207 Table 27. Correlation Matrix of Items from the EFA Sample of the CPRS-R Scores........................ 214 Table 28. Correlation Matrix of Items from the CFA Sample of the CPRS-R Scores........................ 223 Table 29. Theoretical Identification Criteria for Conners’ Teacher Models............................ 232 Table 30. Theoretical Identification Criteria for Conners’ Parent Models.............................. 232 Table 31. Conners’ Teacher Items........................................................................................... 233 Table 32. Conners’ Parent Items.............................................................................................. 235 Table 33. Glossary of Acronyms.............................................................................................. 238

Page 8: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

viii

List of Figures

Figure 1. Conners’ teacher models derived through EFAs..................................................... 81 Figure 2. Conners-E6-T depicted at the factor level............................................................... 81 Figure 3. Conners-F11-T depicted at the factor level............................................................. 82 Figure 4. Conners’ parent models derived through EFAs...................................................... 113 Figure 5. Conners-E7-P model depicted at the factor level.................................................... 114 Figure 6. Conners-F12-P model depicted at the factor level.................................................. 114

Page 9: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

ix

Acknowledgements

Although this dissertation bears my name solely, this research could never have come to

fruition without the support and generosity of many including the Bill and Melinda Gates

Foundation, the Office of Health and Nutrition, US Agency for International Development, the

NCOG team, the expert reviewers, my teachers and mentors, my adviser, and my family and

friends. Also, my deepest gratitude goes out to all of the Nepali families who participated in this

research and made this study possible.

In particular, I would like to thank my adviser, Dr. Beverly Vandiver. I am fortunate to

have had the opportunity to work with an advisor who was always willing to question, challenge,

and inspire me, thereby pushing me to achieve beyond the limits of what I believed possible. As

a direct result of my work with Dr. Vandiver, I have learned to balance passion and enthusiasm

with a healthy dose of scientific skepticism – a lesson that I will carry with me as I move forward

in my career and in my life.

It is my pleasure to thank Dr. Barb Schaefer, Dr. Pamela Cole, Dr. Laura Murray-Kolb,

Dr. Parul Christian, and all of the NCOG members for sharing their data with me and including

me as part of their team. Their guidance and assistance made this manuscript possible. I am

honored to have had the opportunity to work with the dedicated professionals that make up the

NCOG team.

I am thankful to Dr. Cynthia Huang-Pollock, Dr. Keith Wilson, Dr. Marley Watkins, and

Dr. Gary Canivez for providing me with opportunities to develop the foundational skills and

knowledge necessary to complete this study and for all of the mentorship they have provided me

over the years.

Page 10: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

x

I could not have survived graduate school much less completed this dissertation without

the friendship, camaraderie, and comic relief provided by my student colleagues. I am

particularly thankful to Amanda Fleming for her companionship during our late nights working

in CEDAR, her encouragement during the inevitable setbacks that occurred throughout this

process, and for lending her organizational skills as necessary. Thanks to Edith Gnanadas for all

of her advice and friendship and to Madhav Kafle for sharing his expertise. Also, special thanks

to Shermayne Moore and Brianne Mintern for all of their help with proofreading and detail

management.

Most of all, I am thankful to my friends and family for their love and encouragement. I

am thankful to my Dad for teaching me to believe in myself and to pursue my goals with

steadfast determination. I truly appreciate all of the support, encouragement, and entertainment

from my Mom and Kerry. I am thankful for my dog, Sammy, who was curled up on my lap as I

wrote this dissertation, thus making it a much more cheerful process. I am grateful to John for

tolerating the mountains of dissertation paraphernalia that reside on our coffee table (and all

other household surfaces), for always knowing the right time to surprise me with ice cream, for

helping me with the riveting and breathtaking tasks of composing correlation matrices and

solving statistical equations by hand, and for his constant love and support throughout this

difficult journey.

Page 11: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

1

Introduction

Cross-Cultural Assessment

Psychological assessments developed in the United States are increasingly being used

worldwide (Hambleton, Merenda, & Speilberger, 2005). To guard against misuse of

psychological assessments, the American Psychological Association (APA, 2002) created ethical

standards for test use. According to APA, whenever possible, psychologists should not use

psychological assessments without first demonstrating that such measures produce reliable and

valid scores for the designated population.

Ethical standards for culturally sensitive test use are in place for good reasons.

Psychological assessments have historically been used in ways that perpetuated discrimination

and led to harmful practices. For example in South Africa, results from a variety of

inappropriately administered psychological tests were used as “evidence” of White superiority

and to justify apartheid and inequitable distribution of resources (Louw, 1997). In the United

States, eugenicists have used the results of inappropriately administered IQ tests to fuel

arguments for involuntary sterilization of so-called feeble-minded individuals and for

immigration policies that prevented immigration of individuals from certain racial and ethnic

backgrounds (e.g., Asians, Africans, and Jews; Reddy, 2008).

One way that psychologists can avoid acting as agents of oppression is by examining

tests used and ensuring that instruments are used appropriately. Adler (2002) states:

To do their job, standards must operate as a set of shared assumptions, the

unexamined background against which we strike agreements and make distinctions.

So it is not surprising that we take measurement for granted and consider it banal.

Yet the use a society makes of its measures expresses its sense of fair dealing.

Page 12: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

2

That is why the balance scale is a widespread symbol of justice… Our methods

of measurement define who we are and what we value. (p. 2)

The measures used by members of a given culture reflect the beliefs and values of that culture.

When values and beliefs differ across cultures, the reliability and validity of scores from

measures may vary accordingly (Artiles & Ortiz, 2002). If the reliability and validity of test

scores vary as a function of cultural differences, then psychologists must adjust their use and

interpretations of test scores and ensure that the scores are not used to perpetuate ethnocentrism,

discrimination, and inequity in education, employment, healthcare, or other domains.

Data obtained from psychological tests often inform high stakes decisions about

numerous issues, such as special education eligibility, employment, and diagnosis of

psychopathology. While many types of tests are used for high stakes decision making, the use of

IQ tests for diagnostic and educational decisions about racial/ethnic minority students has been at

the forefront of debate for several decades (Messick, 1980). The premise of the classic court

case Larry P. v. Riles (1980) was that African American students were disproportionately placed

in classes for students with mental retardation based solely on their IQ test scores. The judge

ruled that IQ tests were biased against African American students and banned the use of IQ tests

for educational decisions about African American students (O’Connell, 2006). Diana v. Board

of Education (1970) is another well known court case in which nine Mexican American families

contended in a suit against their school district that Mexican American students were

disproportionately placed in special education based solely on the results of IQ tests that were

administered in English. The judge ruled that students have the right to be tested in their native

language and ordered that all Spanish-speaking students in the district’s special education

program be retested (Minow, 2001).

Page 13: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

3

The plaintiffs in both cases alleged that IQ test scores were misused. It was indicated that

diagnoses of mental retardation were rendered solely on the basis of a single test scores and that

test users did not have adequate evidence that the test scores obtained were valid for members

the respective Mexican American and African American populations from which they were

obtained. Overall, the literature suggests that IQ test scores can be valid and informative with

minority students (see Lambert, 1978 for review). However, IQ testing should be couched

within a multi-faceted assessment process, and scores should be valid for the members the

population in which they are used (Sattler, 2001).

Systematic examination of potential test bias (defined as cultural differences in the

construct and predictive validity of test scores) is necessary because such analyses can guide test

users’ evaluations about the extent to which a test score provides useful information for making

decisions about an individual (Brown, Reynolds, & Whittaker, 1999). Evaluation of test bias

represents a first step in ensuring that a test is appropriate for use with a particular population.

The majority of research on test bias has focused on racial/ethnic differences in scores on

intelligence tests for children living within the United States. In contrast, a small number of

studies have examined potential bias in scores on psychological tests (e.g., IQ tests and

personality tests) for individuals living outside the United States. Additionally, few researchers

have examined potential bias in scores on behavior rating scales in the United States or

elsewhere. Yet, behavior rating scales are increasingly being adapted and used outside the nation

in which they were developed (often the United States). For example, symptoms of Attention

Deficit Hyperactivity Disorder (ADHD) have been evaluated in more than 30 countries

worldwide, and the majority of non-US studies of ADHD have utilized rating scales that were

imported rather than indigenously developed (Faraone, Sergeant, Gillberg, & Biederman, 2003).

Page 14: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

4

A variety of factors can interfere with the extent to which test scores allow for non-biased

inferences across cultures. For example, poor translation of test items and cultural differences in

interpretations of behavior and expression of psychopathology can influence test scores (Artiles

& Ortiz, 2002). When used appropriately, behavior rating scales can improve the accuracy of

diagnosis, inform treatment, and ultimately benefit the test taker (Achenbach, 2005).

Behavior rating scales, such as the Conners’ Teacher and Parent Rating Scales – Revised

(Conners, 1997), have been used to study children with attention and behavior problems on every

inhabited continent (Polanczyk & Rohde, 2007). However, the psychometric properties of

scores from such scales have disproportionately been based on children living in the United

States and Western Europe. Children growing up in developing nations may be more likely to

face unique challenges (i.e., hunger, disease, and poor health care and education; Shah, 2010).

Thus, the constructs of inattention and hyperactivity may differ in the developing world, and

research in this area is warranted (Barkley, 2006). Validation of test scores from measures of

ADHD symptoms within developing nations can help to prevent test misuse and facilitate quality

research on ADHD worldwide. As the purpose of this study is to examine the validity of scores

from the Conners’ Teacher and Parent Rating Scales – Revised (Conners, 1997) within the

developing nation of Nepal, it is important to understand the culture to contextualize the findings

obtained.

The Cultural Context of Nepal

The Federal Democratic Republic of Nepal (commonly referred to as Nepal) is a

developing nation located in South Asia. Nepal lies between the People’s Republic of China (to

the north) and the Republic of India (to the east, west, and south). Ecologically, Nepal is usually

divided into three primary regions: (a) the mountain region, (b) the hill region, and (c) the Terai.

Page 15: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

5

The mountain region, called Parbat in Nepali, lies to the north along the Chinese border and is

home to the world’s tallest mountain, Sagarmatha (Mount Everest). This region is sparsely

populated, although small farming communities are present in the valleys. The hill region, called

Pahar in Nepali, is located in the center of Nepal and constitutes the majority of the country

(68%). The Pahar region is home to the Kathmandu Valley and is the most fertile and densely

populated region of Nepal. The Terai region is a tropical and subtropical plains region that is

situated along the Indian border in the southern region of the country (US Library of Congress,

2010). Historically, the Terai was primarily inhabited by immigrants from India and indigenous

ethnic groups, with many exhibiting some degree of natural immunity to malaria. In recent

years, travel and permanent relocation to the Terai region has steadily increased due to improved

accessibility to malaria vaccinations, availability of low cost farmland, and state-sponsored

programs to promote relocation into the Terai (United Nations, 2010).

Government. Nepal was formally referred to as the Kingdom of Nepal. Until 2007, the

Kingdom of Nepal was widely known as the world’s only Hindu monarchy. However, after

years of political strife and civil unrest, the Nepali monarchy was overthrown. A democratic

form of government was installed in 2008, and the formal name of the nation was changed to the

Federal Democratic Republic of Nepal (US State Department, 2010).

People. Approximately 28,800,000 people currently reside in Nepal (World Bank,

2008). Nepal is considered to be one of the most impoverished nations of the world. The

majority of the Nepali population is employed in sustenance farming (Central Intelligence

Agency [CIA], 2010). The Nepalese face obstacles similar to those encountered by citizens of

most developing nations: (a) political instability, (b) high rates of illiteracy, infant mortality, and

infectious disease, and (c) limited educational resources (CIA, 2010).

Page 16: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

6

Nepal has suffered from economic hardship, but the nation is developing and has had

many accomplishments over the years. For example, Nepal is considered the birthplace of

Buddhism: Guatam Buddha was born in Nepal. Also, Nepal was the first Asian nation to

legalize same-sex marriage and grant full rights to homo- and bi-sexual individuals (Barnett,

2009).

Nepal is a diverse, multi-ethnic, multi-lingual, multi-faith nation, and it is home to more

than sixty ethnic groups–at least nineteen of which speak distinct languages (Gurung, 1996).

Primary ethnic groups in Nepal include but are not limited to (a) Chhetri (16%), (b) Brahmin

(13%), (c) Magar (7%), (d) Tharu (7%), (e) Tamang (6%), (f) Newar (5%), (g) Muslim (4%), (h)

Kami (4%), and (i) Yadav (4%; Central Intelligence Agency, 2010). Nepali is the official

language of Nepal and is the primary language of roughly 50% of Nepalese residents (Central

Intelligence Agency, 2010). Other commonly spoken languages in Nepal are Maithili, Bhojpuri,

Awadhi, Newar, Magar, and Tamang (Lewis, 2009). Approximately 80% of Nepalese identify

as Hindu, and 11% identify as Buddhist. A small but growing minority of Nepalese identify as

Muslim or Christian (Niraula, 2007).

Region of ancestry. Region of ancestry is a salient cultural characteristic of the

Nepalese. Individuals whose families are from the hill or mountain regions of Nepal are referred

to as Pahadis. Individuals whose families are from the Terai are referred to as Madhesis. The

term “Madhesi” has also been defined as “non-Pahadis with plains languages as their mother

tongue, regardless of place of birth or residence” (International Crisis Group [ICG], 2007, p. 2).

Pahadi and Madhesi are broad terms that include a wide variety of ethnic and caste groups

(Asian Centre for Human Rights, 2009). While linguistic differences are evident between

Page 17: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

7

Pahadis, who primarily speak Nepali, and Madhesis, who usually speak Hindi-derived languages

(e.g., Maithili and Bhojpuri), sociopolitical issues best delineate those who self-identify as either.

Pahadis make up approximately two-thirds and Madhesis make up one-third of the

population of Nepal. Some Pahadis do not view Madhesis as “true Nepalis” because a number

of Madhesis are believed to have immigrated into the Terai from India prior to and during the

twentieth century (ICG, 2007). Madhesis have historically been underrepresented in the Nepali

government and have viewed many laws (e.g., Nepali as the official language of education)

passed by the government as discriminatory affronts to their respective cultures (Haccethu,

2007). Many Madhesis view ethnic/regional identity (e.g., Madhesi, Kami, or Yadav) as their

primary identity. However, most Madhesis report identifying as Nepali as well and are proud of

this heritage (Haccethu, 2007). Tensions between Pahadis and Madhesis have increased in

recent years and have often resulted in violence. Although not always recognized as such by the

Nepali government, the Madhesis of Nepal are recognized as an oppressed group by international

organizations, such as the International Center for Transitional Justice (2008).

Caste. When the Nepali caste system was first implemented, it was loosely based on the

four primary hierarchical levels of the traditional Hindu caste system (Maslak, 2003): (a)

Brahmin, (b) Chhetri, (c) Vaishya, and (d) Shudra. In the Hindu caste system, the Brahmins

represent the top of the caste system and are affiliated with professions related to intellectual and

spiritual matters, such as education, research, and priesthood. The Chhetris, the second tier of

the caste system, are traditionally linked to handling military duties and national defense. The

Vaishyas represent the third tier of the caste system. Members of the Vaishya caste are typically

employed as merchants and traders. Finally, the Shudras occupy the bottom of the caste

hierarchy and usually work as semi-skilled or unskilled laborers.

Page 18: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

8

During the Rana oligarchy in Nepal (1846-1951), an adapted version of the Hindu caste

system was implemented and legally enforced. Numerous subdivisions were created within each

caste categorization. Examples of subdivisions included enslaveable/non-enslaveable, alcohol

consumer/non-alcohol consumer, and food acceptable/non-food acceptable (whether or not

individuals of higher castes can accept food from an individual). Every individual within Nepali

society was formally ranked based on factors such as language and ethnicity (among others).

Thereby, the construct of caste became inextricably entwined with occupation, language, race,

ethnicity, and region of ancestry.

Caste discrimination was legally abolished in 1963, and social mobility is relatively more

prevalent in Nepal today. However, caste continues to be a salient component of Nepali culture,

albeit informally so (Bennett, 2005). Of all of the Nepalese caste groups, the Dalit are

considered the most oppressed, and the term “Dalit” literally translates to “oppressed” (Gurung,

2005, p. 7). In general, members of the Dalit caste have fewer economic resources and are more

likely to be employed in undesirable occupations compared to their higher caste counterparts.

Despite recent government attempts to increase political representation for members of the Dalit

caste, they remain socially excluded, economically exploited, and politically suppressed

(Gurung, 2005).

Sex roles. Sex roles in Nepal vary widely as a function of ethnicity, caste, and region of

ancestry such that any statement about sex roles among the Nepalese will not necessarily apply

to every ethnic, caste, or regional group (Niruala & Morgan, 1996). Overall, Nepalese women

have limited power and legal rights. Under Nepali law, women cannot inherit property from

their natal kin unless they are over 35 years of age and have never married (Ahearn, 2001).

Also, women are underrepresented in schools, the workforce, the government, and community

Page 19: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

9

organizations (Maslak, 2003). While the majority of Nepali society is considered patriarchal, the

rights and autonomy of Madhesi women are particularly restricted. Compared to Pahadi women,

Madhesi women are less likely to be permitted to go to places such as the market, the local

health centre, or the movies without an accompanying male family member, less likely to

influence family decisions about the use of financial resources, less likely to be educated, and

more likely to marry young (sometimes before menarche; Niruala & Morgan, 2007).

Additionally, many girls and women of the Terai (particularly the Dalit) are vulnerable to sex

trafficking, which disrupts education and literacy development, resulting in social ostracism, and

serious physical and mental health needs (Poudel & Carryer, 2000).

Education. Historically, education in Nepal was only available to members of the elite,

ruling class. Guided by a long-held belief that educating the masses would threaten the

monarchy, the government suppressed public education. Therefore, education in Nepal has

traditionally been viewed as an informal process of teaching necessary work skills, and older

family members were responsible for educating younger family members (Niraula, 2007).

Largely in response to international pressure, government initiatives to increase public

accessibility to formal education began in 1971 (Library of Congress, 2010). Today, both

government and private schools are available in Nepal, and most Nepalese consider private

schools superior to government schools (Rothchild, 2006).

Approximately 91% of Nepalese children between the ages of 6 and 14 (94% of boys and

88% of girls) attend school. However, in the Terai (the context of the present study) overall

school attendance rates for 6- to 14-year-old children are lower (77%) and sex differences in

school attendance are more pronounced (83% boys vs. 70% girls; Central Bureau of Statistics

[CBS], 2007). In a survey about children not attending school, parents living in the Terai

Page 20: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

10

frequently cited that it was “too expensive” (33%) to send boys and “parents did not want”

(41%) to send girls (CBS, 2007). Maslak (2003) notes that girls in Nepal complete more

household work than boys. Thus, mothers are often reluctant to send girls to school because of

the associated increases in work load for other females in the family. Because Nepalese women

traditionally leave their natal kin and reside with their husband’s family after marriage, some

Nepalese families do not consider educating girls to be a lucrative investment (Maslak, 2003).

Mental healthcare. Many Nepalese believe that all illnesses (physical and mental) can

cause disruptions in three aspects of being: (a) the body, (b) the mind, and (c) the soul (Kohrt &

Harper, 2008). Because mental illness is highly stigmatized in Nepal, mental and emotional

difficulties are de-emphasized and often reframed as troubles of the body (Jha, 2007). For

example, anxiety might be described in terms of associated physical symptoms, such as

headaches or stomachaches. Pach (1998) used a narrative approach to examine how families

living in a predominantly Brahmin/Chhetri village of Nepal interpreted mental illnesses. The

findings indicated that the Nepalese participants viewed mental illnesses as conditions that cause

individuals to be demanding, to shout and argue with others, to refuse to work, to feel restless

and agitated, and to wander aimlessly. According to Pach (1998), conditions that prevent an

individual from contributing to family goals are believed to be serious, and displays of

oppositional behavior toward family members are considered particularly troublesome. Because

the sample consisted primarily of Brahmins and Chhetris living in a small village east of

Kathmandu, the findings may not be applicable to Nepalese in general.

Traditionally, the Nepalese consult multiple professionals when coping with an illness

(Kohrt & Harper, 2008). Medical doctors provide advice regarding problems with the body,

while traditional healers are consulted for ailments of the mind and soul. However, the influence

Page 21: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

11

of modern medicine in Nepal is increasing, and traditional healers have been marginalized and

are often spuriously viewed as mentally ill themselves (Van Ommeren et al., 2004). Kohrt and

Harper (2008) contend that the stigmatization of traditional healers as mentally ill has

exacerbated the stigma associated with mental illness itself. Understanding the role and

relationship between modern medical techniques, traditional healers, and the stigmatization of

mental disorders is important when considering the consequences of cross-cultural application of

ideas and technologies.

Psychological research. Few psychological studies have been conducted in Nepal. Most

studies have focused on mental health issues in adults (i.e., depression and schizophrenia). The

few researchers who focused on Nepalese children have identified important cultural distinctions

between Nepalese children and those from other nations. Munroe, Hulefeld, Rodgers, Tomeo,

and Yamazaki (2000) examined sex differences in aggressive behavior among children from four

different cultures: (a) the Newari of Nepal, (b) the Logoli of Kenya, (c) the Garifuna of Belize,

and (d) the American Samoa of the South Pacific. The Newari and the Logoli cultures were

identified as patrilinear or male-dominated, while the other two (the Garifuna and the American

Samoa) were identified as relatively gender-neutral. The results indicated that boys from the

patrilinear cultures (the Newari and Logoli) exhibited higher levels of aggression than boys from

the gender-neutral cultures. Further, the findings suggest that aggressive behaviors may be more

or less normative as a function of culture and sex.

Researchers have also identified cultural differences between ethnic subgroups of

Nepalese children. Cole and Tamang (1998) examined differences in the way Tamang and

Brahmin-Chhetri children in Nepal reported reacting to hypothetical, emotionally challenging

situations. Brahmin-Chhetri children were more likely to report having and intentionally

Page 22: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

12

masking negative feelings than their Tamang counterparts. The findings suggest that meaningful

differences may exist in the manner that Tamang and Brahmin-Chhetri children regulate and

express emotions.

Although the above studies did not directly focus on measurement issues, the findings

underscore the importance of evaluating the degree to which behaviors, especially undesirable

ones, are normative among members of a given culture. The findings of these studies have

important implications for studies evaluating the measurement of psychological constructs in

Nepal. For example, it is possible that the ethnic differences in children’s emotional regulation

identified by Cole and Tamang (1998) may influence the way parents and teachers perceive

children’s behavior, thereby differentially altering the accuracy of behavior ratings. Evaluating

the reliability and validity of scores from psychological measures is an important step in

quantifying what is and is not normative behavior within a given society or societal subgroup.

Childhood attention and hyperactivity problems appear to be a cross-cultural

phenomenon. Research indicates that these symptoms exist and are associated with impairment

in children throughout the world (Polanczyk & Rohde, 2007). Thus, it is likely that such

difficulties also exist with Nepalese children. Because problems with attention and hyperactivity

have been shown to disrupt family function and impede a child’s ability to complete tasks

(Johnston & Mash, 2001), it is plausible that such symptoms would cause distress and

impairment for Nepalese children. Before childhood attention and hyperactivity problems

among Nepalese children can be studied or treated, such problems must be accurately measured.

Research indicates that imprecise measurements can lead to inaccurate diagnoses, and many

claim that inaccurate diagnoses contribute to stigmatization of mental illnesses and distrust of

mental health professionals (e.g., Okazaki & Sue, 1995). Given that mental illnesses are already

Page 23: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

13

stigmatized in Nepal, it is crucial that inaccurate diagnoses be avoided to the greatest extent

possible. At present, not a single measure of child psychopathology has been examined or

validated in Nepal. The examination of the psychometric properties of the Conners’ Teacher and

Parent Rating Scales (Conners, 1997) represents an important first step in the process of

facilitating accurate measurement of psychological symptoms in Nepalese children.

Page 24: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

14

Literature Review

Characterized by symptoms of inattention, hyperactivity, and impulsivity, ADHD is a

psychiatric disorder that often results in long-term educational, occupational, and social

disadvantage (Barkley, 2006). Presently, ADHD is among the most commonly diagnosed

psychiatric disorders. A meta-analytic review of international research indicates that the

worldwide prevalence of ADHD in children is approximately 5% (Polanczyk, de Lima, Horta,

Biederman, & Rohde, 2007). According to the Diagnostic and Statistical Manual of Mental

Disorders, Fourth edition text revision (DSM-IV TR; American Psychiatric Association, 2000),

prevalence rates of ADHD in the United States are approximately 3-7%. ADHD is also one of

the most widely researched psychiatric conditions. Prior to 1979, more than 2,000 studies about

ADHD had been conducted (Weiss & Hechtman, 1979), and approximately 400 studies per year

have been published about ADHD since the 1980s (Barkley, 2008).

Although an abundance of literature about ADHD exists, a disproportionate number of

studies have focused on school-age, Caucasian males living in the United States (Waite & Ivey,

2009). ADHD is believed to have a predominantly biological etiology for most affected

individuals (Bush, Valera, & Seidman, 2005). Yet, individual differences (e.g., race, sex,

nationality, and language) may influence the means by which a child’s ADHD symptoms are

perceived and responded to (Polancyzk & Rohde, 2007). Many researchers have called for

increased research on attention and hyperactivity problems with samples that are representative

of United States and world populations (e.g., Achenbach, 2005; Johnston & Mash, 2001; Lahey

et al., 1994; Pelham, Fabiano, & Massetti, 2005; Quinn, 2005). However, instruments designed

to measure behaviors indicative of ADHD are not always useful cross-culturally (see Artilles &

Ortiz, 2002, for a full review). The purpose of this study is to examine whether the factor

Page 25: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

15

structure of scores from translated versions of the long forms of the Conners’ Teacher and Parent

Rating Scales–Revised (Conners, 1997) among children in Nepal are similar to those identified

with the US and Canadian normative samples. A review of general information about ADHD

(i.e., etiology, theory, diagnostic criteria, assessment, and cultural differences) is provided.

Additionally, examples of influential studies on cross-cultural issues in behavioral assessment

are reviewed, and the efficacy and use of the Conners’ Teacher and Parent Rating Scales–

Revised in the assessment of ADHD are discussed.

General Information about ADHD

Diagnostic criteria. The DSM-IV TR criteria (American Psychiatric Association, 2000)

are commonly used standards for ADHD diagnosis. Although the research studies that informed

the development of the DSM-IV TR criteria had samples that were primarily comprised of

children from the United States, the DSM-IV TR criteria are commonly used in nations

throughout the world (Foreman & Ford, 2008). The DSM-IV TR criteria for ADHD are largely

atheoretical, with the majority of ADHD symptoms retained from previous editions. In previous

editions of the Diagnostic and Statistical Manual of Mental Disorders (DSM, American

Psychiatric Association, 1980, 1987, & 1994), symptoms were selected based on (a) findings

from research that indicated specific symptoms were clinically significant, and (b) reports from

parents and teachers that indicated certain symptoms were viewed as particularly salient (Pelham

et al., 2005).

According to the DSM-IV TR (American Psychiatric Association, 2000), “the essential

feature of Attention Deficit Hyperactivity Disorder is a persistent pattern of inattention and/or

hyperactivity/impulsivity that is more frequently displayed and more severe than is typically

observed in individuals at a comparable level of development” (p. 85). The DSM-IV TR criteria

Page 26: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

16

stipulate that for a clinician to render a diagnosis of ADHD, the client must meet the following

criteria: (a) ADHD symptoms must cause impairment in at least two settings, (b) impairment

from symptoms must be evident before age 7, and (c) symptoms must not be better accounted for

by specified alternative disorders or environmental conditions (e.g., mental retardation). An

individual can be diagnosed with one of three ADHD subtypes: (a) predominantly inattentive

(ADHD-PI), (b) predominantly hyperactive-impulsive (ADHD-PH), or (c) combined (ADHD-

C). To be diagnosed with ADHD-PI or ADHD-PH, an individual must exhibit six or more

inattentive or hyperactive symptoms, respectively. Individuals who exhibit both inattentive and

hyperactive symptoms are diagnosed with ADHD-C (American Psychiatric Association, 2000).

Research findings indicate that diagnostic subtype is moderated by age and sex (e.g.,

Lahey et al., 1994; Rucklidge 2008). For example, females, adolescents, and adults with ADHD

are more likely to be diagnosed with ADHD-PI (Lahey et al., 1994). The DSM-IV TR criteria for

ADHD and associated rating scales have been criticized because the criteria and scales are

disconnected from etiological factors associated with ADHD (e.g., Nigg et al., 2005).

Etiology. Findings from functional neuroimaging (see Bush, Valera, & Seidman, 2005,

for full review) and molecular genetic (see Faraone et al., 2005, for full review) studies indicate

that biological factors contribute to the development of ADHD symptoms in most affected

individuals. Although no neuroimaging or genetic analysis technique has demonstrated the

ability to reliably discriminate individuals with ADHD from those without it, group differences

have been identified for several genes and neural regions. Neuroimaging studies indicate that

abnormalities in the fronto-striatal network are associated with ADHD symptoms (Bush et al.,

2005). Specifically, findings have consistently revealed that individuals with ADHD have

smaller brain mass, decreased cerebral metabolism, and decreased blood flow in several regions

Page 27: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

17

within the fronto-striatal network. Additionally, findings from family, twin, adoption, and

molecular genetic studies indicate that ADHD is a polygenetic syndrome in which multiple

genes are implicated (Faraone et al., 2005).

Although genetics and neurological differences are believed to produce ADHD

symptoms for many affected individuals, environmental factors are believed to have a role in the

manifestation of symptoms and developmental course of the disorder (Johnston & Mash, 2001;

Sagvolden, Johansen, Aase, & Russell, 2005). Johnston and Mash (2001) proposed several

pathways by which environmental factors may interact with biological factors to influence

ADHD symptoms. For example, untreated parental ADHD may result in parenting difficulties,

which in turn may lead to exacerbation of a child’s ADHD symptoms. Although biological and

environmental factors are believed to contribute to ADHD, the mechanisms by which these

factors produce symptoms are not well understood. Several theories have been developed in

attempt to explain the causal mechanisms of ADHD and the means by which physiological

differences and environmental factors lead to symptoms and impairment.

Theory. Most researchers contend that ADHD is a heterogeneous condition and that

subgroups of individuals whose symptoms have distinct etiologies exist within each subtype

(inattentive, hyperactive-impulsive, and combined) of ADHD (e.g., Castellanos & Tannock,

2002; Nigg, Wilcutt, Doyle, & Sonuga-Barke, 2004). In other words, it is believed that the

biological, neurological, or environmental mechanisms that produce ADHD symptoms vary even

among individuals diagnosed with the same ADHD subtype. Therefore, no single theory is

believed to explain ADHD symptoms in all affected individuals (Nigg et al., 2004). Three

prominent cognitive-behavioral theories of ADHD are summarized and critiqued here.

Page 28: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

18

Behavioral disinhibition theory. Perhaps the most well-known theory of ADHD is

Barkley’s (2006) theory of behavioral disinhibition. The basis of the theory is that deficits in

behavioral inhibition lead to deficits in executive functioning, which produce ADHD symptoms.

Barkley stated, “Behavioral inhibition delays the decision to respond to an event. This gives

self-control time to act. The self-directed actions occurring during the delay in the response

constitute, I believe, the executive functions” (p. 304). Furthermore, Barkley contends that

inhibitory deficits lead to deficits in four domains of executive functioning: (a) nonverbal

working memory, (b) verbal working memory, (c) self-regulation, and (d) reconstitution.

Because individuals with ADHD have deficient executive functioning, they are less able to

inwardly direct their thoughts, feelings, and behaviors, resulting in a wide array of symptoms and

impairments (Barkley, 2006).

In the literature, response inhibition has been measured in several different ways, but it is

usually measured using computerized tasks on which the participant is expected to refrain from

responding to certain stimuli (Barkley, 1999). For example, on the go-no-go task, the participant

is asked to produce a response (e.g., a mouse click) as quickly as possible when presented with a

response cue (such as a visual signal on a computer screen). After participants become

accustomed to the response cues, an inhibitory cue (such as a tone) is introduced. Then,

participants are instructed to inhibit responses to the response cue if the inhibitory cue is present.

Research suggests that individuals with ADHD are less able to effectively inhibit responses

when presented with inhibitory cues (see Barkley [1999] for a review). As noted previously,

inhibitory deficits (as identified by poor performance on response inhibition measures) are

believed to lead to deficits in executive functioning. However, Barkley (2006) has not specified

the means by which executive functioning should be measured.

Page 29: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

19

Barkley’s (2006) theory of ADHD is well-known, but substantial portions of it have not

been supported by empirical research. To establish causality, the proposed causal variable

should be a necessary condition for the occurrence of the proposed outcome. If executive

functioning deficits were causal agents of ADHD symptoms, then such deficits would be present

in the majority of individuals with ADHD. Findings from meta-analytic studies of children

(Willcutt, Doyle, Nigg, Faraone, & Pennington, 2005) and adults (Boonstra, Oosterlaan,

Sergeant, & Buitelaar, 2005) indicated that some individuals with ADHD exhibited deficits in

executive functioning. The relationship between ADHD diagnosis and the presence of executive

functioning deficits has been shown to have a medium effect size in children and adults (.46 -

.69; Boonstra et al., 2005; Willcutt et al, 2005.) However, inhibition and executive functioning

deficits are absent in a large number of individuals who are diagnosed with ADHD via structured

interviews, behavior rating scales, and observations (Willcutt et al., 2005). Therefore, Willcutt et

al. (2005) and Boonstra et al. (2005) concluded that although executive functioning deficits may

be an important component of ADHD, these deficits cannot be a causal mechanism of the

disorder.

Delay aversion theory. Sonuga-Barke (1994) claimed that neurological deficits in the

reward centers of the brain cause ADHD symptoms. He posited that neurological deficits

produce an intense aversion to waiting, which in turn contributes to a child’s unwillingness to

delay gratification to gain a reward. While behavioral disinhibition theorists believe that ADHD

symptoms are caused by an inability to inhibit responses or behaviors, delay aversion theorists

claim that ADHD symptoms reflect a child’s rational choice to avoid the unpleasant experience

of delay (Solanto et al., 2000). In laboratory settings, delay aversion is typically measured by a

child’s performance on choice-delay tasks (Songua-Barke, Taylor, Sembi, & Smith, 1992). In a

Page 30: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

20

choice-delay task, a child is asked to repeatedly choose between a small reward that is

immediately delivered and a large reward that is delivered after a delay of 30 seconds of more.

Findings from several studies indicated that children with ADHD are more likely to choose the

small, immediate reward rather than the large, delayed reward (e.g., Solanto et al., 2000; Sonuga-

Barke et al., 1992)

Although delay aversion theory has been supported by findings from several studies (e.g.,

Marco et al., 2009), it is unlikely that delay aversion causes ADHD symptoms for most affected

individuals. Delay aversion has been shown to be strongly associated with specific subgroups of

individuals with ADHD (e.g., children with co-morbid conduct disorder, children with lower

IQs, and males). However, many children who are diagnosed with ADHD based on structured

observations, interviews and rating scales do not display delay aversion when presented with

choice-delay tasks (Marco et al., 2009; Paloyelis, Asherton, & Kuntsi, 2009). To establish

causality, an effect should occur exclusively in the presence of the presumed cause. Thus,

because many children with ADHD do not demonstrate delay aversion, delay aversion alone

cannot be the sole causal factor of ADHD.

Dual pathway framework. Sonuga-Barke (2005) developed a dual pathway framework

in response to mixed support for the delay aversion theory. Research suggests that both

inhibition/executive functioning deficits and delay aversion are related to ADHD symptoms, yet

executive functioning deficits and delay aversion are uncorrelated with one another (Sonuga-

Barke, 2004). The basis of the dual pathway framework is that two, independent and

etiologically distinct pathways lead to ADHD symptoms: (a) the inhibitory/executive

dysfunction pathway and (b) the delay aversion pathway.

Page 31: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

21

The dual pathway framework was the first to contain a synthesis of two theories of

ADHD and to explain the heterogeneity commonly found in studies examining

neuropsychological functioning in individuals with ADHD (Johnson, Wiersma, & Kuntsi, 2009).

Although the dual pathway framework provides testable hypotheses, no information specifying

how the proposed constructs should be measured is given (Johnson et al., 2009). Further,

measures that have been used to test the dual pathway model (e.g., serial reaction time tasks)

have not consistently been shown to produce reliable and valid results (Alderson, Rapport, &

Kofler, 2007). Finally, because studies of the dual pathway model have been largely restricted to

the laboratory setting, the ways in which delay aversion and behavioral inhibition deficits might

manifest outside the laboratory setting are unclear (Johnson et al., 2009). While the dual

pathway framework is a plausible explanation of the etiology of ADHD symptoms, further

research is necessary to (a) elucidate the mechanisms by which delay aversion and inhibition

deficits produce impairment in children with ADHD outside of the laboratory setting and (b) to

refine measurement techniques.

Assessment. ADHD symptoms can be assessed in many different ways. Research

findings indicate that the use of multi-informant, multi-method assessment procedures increases

the accuracy of ADHD diagnoses (Johnson & Murray, 2003). Evidence-based assessments of

ADHD symptoms and impairments are (a) broadband rating scales of general psychopathology,

(b) narrow-band ADHD rating scales, (c) structured diagnostic parent interviews, (d) measures of

global impairment, and (e) systematic behavior observations (Pelham et al., 2005). Because full

discussion of all evidenced-based assessment tools for ADHD is beyond the scope of this study,

only broadband and narrowband rating scales are described here. Broadband behavior rating

scales are measures of general psychopathology that can be completed by children and their

Page 32: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

22

parents and teachers. Broadband rating scales are used to rule out alternative diagnoses that may

be associated with symptoms of inattention or hyperactivity (e.g., mood or anxiety disorders) and

to screen for co-morbid conditions (Barkley, 2006). Narrow-band rating scales are tools used to

document the number, frequency, and severity of ADHD symptoms. Although scores from most

narrow-band rating scales have been shown to effectively discriminate between clinical and non-

clinical groups, narrow-band rating scales often yield high rates of false positives because they

do not account for inattentive and hyperactive symptoms that may be better accounted for by

other disorders (Pelham et al., 2005).

The majority of behavior rating scales designed to assess problems associated with

ADHD are based on the DSM-IV TR criteria (American Psychiatric Association, 2000).

However, the samples employed in field trials of the DSM-IV TR criteria contained a

disproportionately high number of Caucasian, school-age, American males (Lahey et al., 1994).

Therefore, the findings may not generalize to other groups, and the DSM-IV TR criteria and

rating scales based on them may not capture ADHD symptoms as effectively for members of

underrepresented groups, including but not limited to, females, racial/ethnic minorities,

individuals living outside the United States, non-native English speakers, pre-school children,

and adults. Although additional research is needed, a growing number of studies have examined

ADHD symptoms in diverse groups of individuals and the findings have varied as a function of

the group studied and the facet of ADHD under examination.

Cultural Differences and ADHD

Race and ADHD. Historically, research on ADHD has been based upon studies that

employed predominantly Caucasian American samples. In recent years, researchers have

increasingly examined issues related to ADHD in African Americans and Hispanic Americans,

Page 33: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

23

but ADHD research continues to focus primarily on Caucasian American children (Miller, Nigg,

& Miller, 2009). ADHD in Asian Americans, Native Americans, and Multiracial Americans

remains largely unexamined. A brief review of key studies on racial/ethnic differences in

behavioral assessment, in general, and on assessment of ADHD in African American and

Hispanic American children is provided.

Racial issues in behavioral assessment. Chang and Sue (2003) hypothesized that racial

bias would influence teacher ratings of student behavior. Participants (197 Caucasian American

teachers in California) were presented with vignettes that described the classroom behaviors of

hypothetical African American, Asian American, and Caucasian American students. A

photograph of a child of the specified race accompanied each vignette. Each vignette described a

child engaging in either over-controlled, under-controlled, or normal behaviors. Race and type

of behavior (over-controlled, under-controlled, or normal) were matched. After viewing the

photograph and reading the behavioral descriptions, teachers provided ratings pertaining to

seriousness of behavioral symptoms, extent to which symptoms warranted a referral, and causal

attributions. When Asian American children were described as engaging in over-controlled

behaviors in the vignettes, teachers were more likely to describe the behaviors as typical and

culturally-related than was true for African or Caucasian American students. No statistically

significant differences between ratings of Caucasian and African American students were

identified. Based on their findings, Chang and Sue (2003) concluded that stereotypes and

student race/ethnicity may influence teacher perceptions about and ratings of student behavior.

In addition to student racial/ethnic background, teacher race/ethnicity and student-

teacher ethnic match may influence the accuracy of behavior ratings. Puig et al. (1999)

compared the accuracy of Black Jamaican teachers’ ratings of Black Jamaican students to those

Page 34: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

24

of White American teachers’ ratings of Black American students. The sample contained 102

children between the ages of six and 11 (54 Black Jamaican children and 48 Black American

children), and the children were matched by sex. Fifty-one teachers provided ratings using the

Child Behavior Checklist (CBCL; Achenbach, 1991) for students who were randomly selected

from their class roster,. Additionally, each child was observed four times within a two week

period by a trained research assistant, and the childrens’ behaviors were recorded using the

CBCL direct observation form (DOF; Achenbach, 1991).

Nine ANCOVAs were used to analyze the data. Age and socio-economic status (SES)

were covariates in all analyses. Nationality and sex were predictor variables in the first six

analyses. Outcome variables for analyses of observational measures were total time-on-task, the

externalizing behavior DOF score, and the internalizing behavior DOF score. Outcome variables

for analyses of teacher report measures were the externalizing behavior CBCL score, the

internalizing behavior CBCL score, and the total problems CBCL score. In the final three

analyses, predictor variables were reporter (teacher or observer), sex, and nationality. Outcome

variables were internalizing, externalizing, and total problems scores on combined versions of

the CBCL and DOF scales (Puig et al., 1999).

For Black Jamaican students, teachers of the same-race reported more problem behaviors

than observers, accounting for a small effect size (3% of the variance). For Black American

students, White American teachers reported more problem behaviors than observers, resulting in

a large effect size (16% of the variance). Overall, observers reported that Black Jamaican

students exhibited more behavioral problems than Black American students (tapping 5-10% of

the variance). Yet, White American teachers of Black American students reported more

behavioral problems than Black Jamaican teachers of students of the same-race (accounting for

Page 35: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

25

8-17% of the variance). According to Puig et al. (1999), the findings suggest that ethnic

similarity between Black Jamaican students and their teachers and ethnic difference between

Black American students and their teachers may influence the threshold at which teachers

identify a particular behavior as a problem. The findings support the hypothesis that teacher-

student ethnic match influenced behavior ratings, and suggest that the use of behavior rating

scales for racial/ethnic minority students within the United States may be problematic in some

circumstances. However, the findings must be interpreted in light of several limitations (e.g.,

numerous univariate analyses and small sample size).

Racial issues in ADHD assessment. Some studies have examined the extent to which

behavior rating scales and structured interviews used to assess symptoms of ADHD are biased

against or unfair to African and Hispanic Americans. Hillemeier, Foster, Heinrichs, and Heier

(2007) examined racial differences on the parent interview form of the Diagnostic Interview

Schedule for Children, Fourth Edition (DISC-IV, Shaffer, Fisher, Lucas, Dulcan, & Schwab-

Stone, 2000) using item response theory (IRT) methodology. Hillemeier et al. reported that

African American parents of at-risk children in the sample responded differently than their

Caucasian American counterparts to four of the 12 DISC-IV items designed to assess

hyperactivity. Specifically, African American parents were less likely than Caucasian American

parents to indicate that their child was exhibiting hyperactivity, though data from other reports

and behavioral observations indicated that African American and Caucasian American children

were exhibiting comparable levels of hyperactive behavior. Hillemeier et al. (2007) claimed that

the discrepancy may be due to a higher tolerance for active and expressive behavior in the

African American community. In other words, African American parents may be more likely to

Page 36: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

26

view highly active behavior as normal and less likely to describe it as hyperactive and

problematic in a structured interview.

Findings from other studies show that teachers rate African American children as more

hyperactive than Caucasian American children (e.g., Conners, 1997; Epstein, March, Conners, &

Jackson, 1998), and group differences remain when the child’s behavior is examined using both

behavioral rating scales and systematic behavioral observations (Epstein et al., 2005).

Hosterman, DuPaul, and Jitendra (2008) examined the extent to which the race of a student

influenced the accuracy of teacher ratings. Caucasian American teachers rated the behavior of

African, Caucasian, and Hispanic American students using the long form of the Conners’

Teacher Rating Scale–Revised (CTRS-R). Scores on the CTRS-R were compared with data

from direct behavior observations (DBOs) conducted by graduate students whose races were not

specified. Findings revealed statistically and practically significant differences between

Caucasian and minority students on CTRS-R ratings but not on DBOs. In addition, the group

differences in the correlations between teacher ratings and DBOs were examined. Teacher

ratings of racial/ethnic minority students had higher correlations with DBOs than were their

ratings of Caucasian students. In essence, Hosterman et al. (2008) claimed that Caucasian

teachers may experience a racial in-group bias, whereby they tend to recognize the hyperactivity

and inattention of minority children and overlook the same behaviors in children of their own

race.

Conversely, Ramirez and Schapiro (2005) found that Latino teachers may be more likely

to identify negative behaviors in children of their own race/ethnicity. Ramirez and Schapiro

used the ADHD-IV Rating Scale (DuPaul, Power, Anastopoulos, & Reid, 1998) to examine

Caucasian (n = 61) and Latino (n = 189) teachers’ ratings of the behavior of one Latino and one

Page 37: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

27

Caucasian child viewed on a videotaped vignette. They conducted systematic behavioral

observations of the students in the vignette and no statistically significant behavioral differences

were identified. Also, the findings revealed no statistically or practically significant differences

in Caucasian teacher’s ratings of Caucasian and Latino students. However Latino teachers rated

the Latino child as more hyperactive than the Caucasian child, but after controlling for the level

of acculturation of the Latino teachers, mean group differences were not statistically significant.

Ramirez and Shapiro concluded that Latino teachers with lower levels of acculturation may hold

children of their own race to a higher behavioral standard. Similarly, other research has

indicated that level of acculturation may also influence Latino mothers’ ratings of their child’s

behavior. Schmitz and Velez (2003) found that Latino mothers with lower levels of

acculturation rated their children as more hyperactive than Latino mothers with higher levels of

acculturation.

In conclusion, research findings indicate that ratings of problematic childhood behaviors

may vary as a function of the student’s race, rater’s race, and level of acculturation. The reasons

for racial differences in behavior ratings have not yet been conclusively determined. Issues such

as true behavioral differences, differences in rater perceptions, or test bias may play a role. The

literature indicates that accurate diagnoses are best made using actuarial formulas (Dawes,

Meehl, & Faust, 1989), and that clinicians who do not use such formulas are more likely to

produce different diagnoses solely as a function of the client’s race (Miller, Nigg, & Miller,

2009; Whaley & Geller, 2007). However, if the components of an actuarial formula are less

accurate for members of a specific group, then it logically follows that the diagnostic formula

itself would be less accurate (Johnson & Mash, 2003). Because parent and teacher ratings are a

central component of ADHD diagnosis, further research must be conducted to determine whether

Page 38: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

28

or not racial bias may be inherent in behaviorial rating scales and the extent to which these

instruments are useful in identifying racial/ethnic minority children who are truly impaired.

Nationality and ADHD. ADHD has been identified in individuals throughout the world

(see Polanczyk et al., 2007, for a full review). However, findings about the prevalence of ADHD

across nations have varied. For example, the prevalence of ADHD in Sweden and Iceland is

estimated to be approximately 3-5%, while the prevalence rates in Columbia and Brazil have

been found to be as high as 16-18% (Faraone, Sergeant, Gillburg, & Biederman, 2003;

Polanczyk et al., 2007). Such variation in prevalence rates of ADHD led Timimi and Taylor

(2004) to contend that ADHD may be a cultural by-product rather than a neurobiological

disorder. In response to Timimi and Taylor’s claim, Polanczyk et al. (2007) conducted a

systematic meta-analysis of international studies of ADHD. The findings were that the

worldwide prevalence of ADHD is approximately 5%, with Polanczyk et al. concluding that the

variation in prevalence rates across countries is primarily an artifact of the various

methodologies used and is not purely the result of cultural influence as Timimi and Taylor

contend. However, examination of potential influence of culture on a disorder should not be

solely based on prevalence rates. Other aspects of a disorder may be reflective of cultural

variations.

The research group called Attention Deficit Hyperactivity Disorder Observational

Research in Europe (ADORE; ADORE Study Group, 2006) examined ADHD diagnosis,

treatment, and outcome for 1,572 newly-diagnosed children living in 10 European countries:

Austria, Denmark, France, Germany, Iceland, Italy, the Netherlands, Norway, Switzerland, and

the UK. Findings from the ADORE (2006) study revealed relative similarity of ADHD

symptoms and levels of impairment across nations. However, differences between nations were

Page 39: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

29

noted in a number of domains: (a) the diagnostic system in use, (b) the incidence of co-morbid

disorders (e.g., learning disabilities, conduct disorder, anxiety disorders, and depression), (c) the

incidence of co-occurring sleep problems, (d) the type of prescribed treatment and, (e) the link

between ADHD diagnosis and family stress. Overall, the results of the ADORE study

underscore the importance of conducting a cross-cultural analysis on all aspects of ADHD in

order to understand the potential influence culture has in the expression of ADHD across various

nationalities.

Examination of cultural differences in ADHD across nations requires systems of

measurement that accurately assess the intended construct cross-culturally. Though difficult,

accurate cross-cultural assessment of psychopathology is possible. For example, Ivanova et al.

(2007) used confirmatory factor analysis (CFA) to test the eight-syndrome factor structure of the

parent version of the Child Behavior Checklist (Achenbach, 1991) in 30 nations throughout the

world. Samples from nations in Asia, Africa, Australia, the Caribbean, Europe, the Middle East,

and North America were included, and CFAs were conducted separately on data from each

nation. The authors reported that root mean square error of approximation (RMSEA) values

were the primary determinant of model fit. However, other fit indices such as the comparative fit

index (CFI) and Tucker-Lewis index (TLI) were calculated. Overall, RMSEA values indicated

acceptable fit (< .07) of the eight-factor model to the data from participants in all nations (.026-

.055). According to Ivanova et al., CFI values were indicative of an acceptable fit (> .90) of

model to data in all nations except Ethiopia, Hong Kong, Germany, and Lithuania (.73-.95). TLI

values indicated acceptable fit (>.90) in all nations except Ethiopia (.79-.96). The authors

concluded that their findings demonstrate that the 8-syndrome model has a good fit to the data in

all societies with minimal fit problems in the Ethiopian sample.

Page 40: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

30

Ivanova et al.’s (2007) study represents an important step in the field of cross-cultural

psychological assessment. The findings indicate that it is possible for the factor structure of

scores from a scale developed in the United States to be replicated in as many as 30 countries

worldwide. However, the limitations of the study must also be considered. In particular, the fact

that a model demonstrates good fit to data in a particular nation does not mean that the proposed

model reflects the best way to measure the construct within that cultural context. In fact, in CFA

it is possible for a model to demonstrate good fit despite the existence of countless equivalent or

superior models (Kline, 2005). Some suggest that the use of CFA is preferable to exploratory

factor analysis (EFA) when researchers have a well supported theory pertaining to the

construct’s factor structure (e.g., Goldberg & Velicer, 2006). However, when the majority of

research guiding a theory about a factor structure was conducted in a particular cultural context

(i.e., the United States), the extent to which the theory is generalizable to other contexts (i.e.,

other countries throughout the world) is unknown. Therefore, it seems that the initial use of a

theory generating procedure (e.g., EFA) would be most appropriate when introducing a scale in a

novel cultural context (Haig, 2005). Additionally, using EFA circumvents the issue of

equivalent models while providing information pertaining to the applicability of a factorial

theory in a novel culture. However, to determine that an instrument produces reliable and valid

scores within a culture, evidence of other types of validity, such as criterion-related validity, is

necessary (Sattler, 2001). Examination of criterion-related validity requires adaptation of a

secondary measure for use within the culture.

The Conners’ Rating Scales

The original Conners’ Rating Scales. The Conners’ Teacher and Parent Rating Scales

(Goyette, Conners, & Ulrich, 1978) were originally unnamed behavior checklists used in the

Page 41: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

31

Harriet Lane Clinic at John’s Hopkins Hospital in the 1960s. Clinicians first used the behavior

checklists to gather information from parents and teachers. Responses to items on the checklists

were used to identify behavioral domains in which further exploration (via semi-structured

parent and student interviews) was warranted. In 1970, Conners published a principal

components analysis (PCA) of the unnamed behavior checklists, which subsequently became

commonly used as pre- and posttest measures in stimulant drug trials. In 1989, the Conners’

Parent Rating Scale (CPRS) and Conners’ Teacher Rating Scale (CTRS) were formally

published, and the instruments quickly gained popularity with both researchers and clinical

professionals (Ginnaris, Golden, & Green, 2001). After their publication, Conners (1997) stated,

The Conners’ Rating Scales have become among the most widely used child

behavior rating scales in the world, with their international use extending to places

such as Australia, Brazil, Hong Kong, Italy, New Zealand, China, Spain, Germany,

Canada, and the United States. (p. 84)

The Revised Conners’ Rating Scales. Revised and restandardized versions of the

Conners’ Rating Scales (CRS-R) were published in 1997 (Conners). According to Conners

(1997), the revised scales differ from the original scales in several ways. The wording of the

response options was changed to increase clarity and precision, the revised scales can be used

with individuals of a wider age range (3-17), and auxiliary scales were added to enhance the

amount of information obtained from the scores. The primary goal in the development of the

CRS-R was to create instruments for the purposes of screening, monitoring treatment,

conducting research, and aiding diagnostic and clinical decision making (Conners, 1997).

The CRS-R (Conners, 1997) are commonly used to assess the behavioral characteristics

associated with ADHD and other disorders. There are three versions of the CRS-R: teacher,

Page 42: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

32

parent, and self-report. Only the teacher and parent versions are examined in this study. When

completing the scales, respondents are instructed to read each item, consider the extent to which

the statement reflects the target child’s behavior, and circle the most appropriate response.

Response options are (a) Not at all true (Never, Seldom), (b) Just a little true (Occasionally), (c)

Pretty much true (Often, quite a bit), and (d) Very much true (Very often, very frequent).

There are long and short forms of each scale version (teacher and parent). The long

forms of the scales are examined in this study and are designed to assess broad ranges of

behavior (e.g., anxiety, social problems, and hyperactivity), while the short forms are intended to

tap only three behavioral dimensions: attentional difficulties, hyperactivity, and oppositional

behavior. The long form of the teacher scale (Conners, 1997) has 59 items distributed across 11

subscales. (Many subscales had overlapping items.) The first six subscales were derived from

factor analytic research (Conners, 1997): (a) Oppositional (6 items), (b) Cognitive

Problems/Inattention (8 items), (c) Hyperactivity (7 items), (d) Anxious-Shy (6 items), (e)

Perfectionism (6 items), and (f) Social Problems (5 items). The remaining scales consist of five

rationally derived subscales–three auxiliary subscales and two DSM scales. The rationally

derived subscales exist as separate scales and the items that make up these scales were analyzed

separately from other items on the teacher scale. The Conners’ Global Index (CGI-P; 10 items)

is derived from a combination of scores on two auxiliary subscales: the Restless-Impulsive

subscale (6 items) and Emotional Liability subscale (4 items). The 12-item ADHD index is an

auxiliary scale that is comprised of items that are predictive of ADHD (Conners, Sitarenios,

Parker, & Epstein, 1998). The final set of teacher subscales (the DSM scales) are made up of

items that correspond directly to ADHD symptoms listed in the DSM-IV TR (American

Psychiatric Association, 2000): (a) the DSM Inattentive subscale (9 items) and (b) the DSM

Page 43: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

33

Hyperactive-Impulsive subscale (9 items). The DSM total symptoms index is derived from a

combination of scores from the DSM Hyperactive-Impulsive and Inattentive subscales. The

majority of the items on the rationally derived DSM and CGI subscales were never tested with

factor analysis, as were the items from the empirically derived subscales.

The long form of the parent scale (Conners, 1997) has 80 items covering 12 subscales.

The structure of the parent scale is nearly identical to that of the teacher scale (Conners, 1997).

Differences in the Conners Parent Rating Scale–Revised (CPRS-R) relative to the teacher scale

are as follows: (a) the parent scale has an additional empirically-derived subscale (six items)

intended to tap psychosomatic symptoms, (b) parent items have slightly different wording, and

(c) four of the parent subscales (Oppositional, Cognitive Problems/ Inattention, Hyperactivity,

and Perfectionism) contain additional items.

Use of the CRS-R. Conners (1997) described the intended use of the CRS-R as follows:

The main use of the Conners’ Rating Scales–Revised will be for the assessment of

ADHD. However, the CRS-R can have a much broader scope, as they also

contain subscales for the assessment of conduct problems, cognitive problems,

family problems, emotional problems, anger control problems, and anxiety

problems. The ability to assess these other facets is crucial given that ADHD is

frequently co-morbid with these other problems. (p. 5)

There is no research on the frequency with which the CRS-R have been used to assess

psychological difficulties other than attention and hyperactivity, but their primary use for

assessing ADHD is evident. Koonce (2007) surveyed school psychologists (N = 246) about

instruments used to evaluate children for ADHD. Approximately 17-19% of the school

psychologists surveyed indicated using the CRS-R as the sole individually administered measure

Page 44: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

34

of ADHD in the test battery. Whether school psychologists generally use the long versus the

short form of the scales was not reported. Given the reported frequency of use of the CRS-R for

diagnosis and educational decision making, as reported by the developer (Conners, 1997) and

independent researchers (e.g., Koonce, 2007), the psychometric adequacy and diagnostic utility

of CRS-R scores should be closely scrutinized.

Development. The development of the revised scales consisted of nine phases: (a)

rationale and goal setting, (b) item selection, (c) pilot study and psychometric analysis, (d) scale

revision, (e) development of new forms, (f) CFA of new forms, (g) development of norms, (h)

collection of psychometric data for each proposed use of the scale, and (i) preparation of testing

manuals. Items on the CRS-R (Conners, 1997) were derived from various sources: (a) previous

versions of the scales, (b) other published instruments, (c) DSM-IV TR criteria, and (d) changes

in clinical practices and conceptions of ADHD. After a large item pool (131 items for CTRS-R

and 193 for CPRS-R) was developed, they were revised or discarded based on comments by an

expert panel. The remaining items were pilot tested using various reliability analyses of the

scores with samples of parents and teachers of 170 school-age children and adolescents from

eight states. Revised versions of the scale were then examined using EFA and subsequently

using CFA with new samples of parents and teachers. Norms were established using data from

the CFA sample, and further research was conducted on the psychometric properties of scores.

During the development of the CTRS-R and CPRS-R, Conners (1997) examined the

psychometric properties (internal consistency and various forms of construct validity) of the

respective scores. Researchers have conducted similar but independent studies of the CRS-R,

and a review and critical analysis of all studies that examined the psychometric properties of the

CTRS-R and CPRS-R (long and short form) scores is provided. Studies that examined earlier

Page 45: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

35

versions of the Conners’ scales are not discussed because previous editions differ substantially

from the current scales in regard to item content and constructs assessed.

Reliability. For the teacher’s long form, the initial internal consistency estimates for the

six empirically derived subscale scores ranged from .80 to .94. Reliability estimates for the

scores of the five rationally derived subscales ranged from .87 to .96 (Conners, 1997). Internal

consistency estimates for scores from the teacher short form are comparable and have ranged

from .88 to .95.

In regard to the long form of the parent scale, Conners (1997) noted that internal

consistency estimates (Cronbach’s α) for scores of the seven empirically derived subscales

(Oppositional, Cognitive Problems/Inattention, Hyperactivity, Anxious-Shy, Perfectionism,

Social Problems, and Psychosomatic) ranged from .77 to .93. Internal consistency values for

scores of the five rationally derived subscales ranged from .87 to .93. Internal consistency

estimates for scores from the parent short form are comparable and have ranged from .88 to .94.

Independent estimates of internal consistency of CRS-R scores have been conducted in

the United States and worldwide, and some studies have yielded similar findings. For example,

Kumar and Steer (2003) examined the internal consistency of scores from the short form of the

parent scale in a sample of 200 children and adolescents, who were psychiatric outpatients in

central New Jersey. The sample was clinic-referred rather than community based (as Conners’,

1997, sample was), but was otherwise similar to the standardization sample (predominantly

Caucasian American participants). Reliability estimates for the scores identified by Kumar and

Steer were .87 for Oppositional Behavior, .89 for Cognitive Problems/Inattention, .87 for

Hyperactivity, and .91 for the ADHD index. Thus, reliability estimates for the CPRS-R scores

were similar to those originally reported by Conners (1997).

Page 46: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

36

Other researchers have examined the internal consistency of CRS-R (Conners, 1997)

scores within the United States using samples quite different from the standardization sample.

Miller, Fee, and Netterville (2004) examined the reliability of the short forms of both the CTRS-

R and CPRS-R scores using a sample of teachers and parents of 48 children with mental

retardation. The range of the estimates was wide. Miller et al. reported that reliability estimates

for scores on the factors of the teacher scale ranged from .51 (Attention Problems) to .96

(Conduct Problems), and alphas for scores on the parent scale factors ranged from .59 (Anxiety)

to .89 (Learning Problems). Overall, the values reported by Miller et al. were lower than those

reported by Conners (1997). Conversely, Deb, Dhaliwal, and Roy (2008) examined the internal

consistency of both CTRS-R and CPRS-R (short forms) scores using a sample of parents and

teachers of 151 children diagnosed with mental retardation or borderline intellectual functioning.

The reliability estimates for scores on a CTRS-R Total score, which included items from every

factor, was .84, and the CPRS-R scores had a reliability estimate equal to .80. The estimates

obtained by Deb et al. were comparable to those reported by Conners. However, Deb et al. did

not report reliability estimate for scores on each of the three factors as Conners did. By

including scores from all three factors and increasing the total number of items used to calculate

the reliability estimates, Deb et al. may have artificially inflated their reliability estimates

(Sattler, 2001). Small sample size was another limitation of this study.

The internal consistency of scores on the Conners’ scales has also been examined outside

the United States and Canada. Gau, Soong, Chiu, and Tsai (2006) examined the internal

consistency of scores from the short forms of the CTRS-R and CPRS-R with a sample comprised

of teachers and parents of 2,584 children living in Taiwan. The reliability estimates for the

scores ranged from .91 (Inattention/Cognitive Problems and Oppositional factors) to .93

Page 47: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

37

(Hyperactivity factor) for the teacher scale and from .84 (Oppositional and Hyperactivity factors)

to .87 (Inattention/Cognitive Problems factor) for the parent scale.

Overall, findings from independent evaluations of the internal consistency of scores from

the Conners’ scales are somewhat similar to the original findings (Conners, 1997). The internal

consistency values reported by Conners (1997) are above the minimum cut-offs specified by

Salvia and Ysseldyke (2007) for research and screening purposes (.60 and .80, respectively) and

many are above the minimum cut-off for diagnostic and educational decision making (.90).

However, independent replications of internal consistency analyses of CRS-R scores have been

variable with some falling below the minimum cut-off for screening purposes and below the cut-

off for diagnostic decision making. Further, all independent studies of the internal consistency of

CTRS-R and CPRS-R scores were conducted on the short form of the scales. The internal

consistency reliabilities of scores from the long forms have not yet been examined by

independent researchers.

Examination of inter-rater reliability of CRS-R scores is not reported in the technical

manual (Conners, 1997). However, Loughran (2003) examined the relationship between regular

and assistant teachers’ ratings of US children using the short form of the teacher scale. The

sample was composed of teachers and assistant teachers, who rated 60 children (33 girls and 27

boys) from a suburban, upper-middle class community. Correlations between teacher and

assistant teacher ratings were .60 for pre-school children and .80 for elementary school children.

Thus, Loughran concluded that CTRS-R scores may have relatively high inter-rater reliability for

elementary school students but somewhat lower inter-rater reliability for pre-school students.

Conners (1997) reported the stability of scores for the CTRS-R and CPRS-R long forms

at six to eight week intervals. Test-retest coefficients of scores were variable, ranging from .47

Page 48: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

38

(Cognitive Problems/ Inattention) to .88 (Anxious-Shy) on the teacher scale and from .47

(Anxious-Shy) to .85 (Hyperactivity) on the parent scale. In an independent study, Loughran

(2003) examined the stability of scores from the short form of the CTRS-R at a four-year

interval, with children in pre-school at the first data collection point and in early elementary

school at the second one. At each data collection points, children were rated by their current

regular and one assistant teacher – first by their pre-school teachers and later by their elementary

school teachers. No statistically significant differences were found between the mean CTRS-R

scores at Time I and Time II, thus lending support to the hypothesis that CTRS-R scores are

stable.

Construct validity. Conners (1997) used EFA and CFA in establishing the factor

structures of empirically-derived subscales of CTRS-R and CPRS-R scores. (Items that are only

on the rationally-derived subscales were not included in the factor analysis.) Initially, Conners

examined the factor structures of the CRS-R using EFA with principal axis extraction (PAF) and

varimax rotation. The number of factors to retain for rotation was based on visual inspection of

scree plots and the eigenvalue rule of one. Items were included on the final versions of the CRS-

R if pattern coefficients were above .30 on only one factor. After conducting EFAs, Conners

used a split-half factor comparabilities technique (Everett, 1983) to verify that the correct

numbers of factors were retained for rotation in the exploratory analyses. The split-half factor

comparabilities coefficient is similar to a congruence coefficient, but Everett (1983) argued that

the comparability coefficient is a more accurate measure of factor stability. Subsequently,

Conners examined the identified models using CFA and evaluated fit based on whether on the

goodness of fit index (GFI) was greater than .85, adjusted goodness of fit index (AGFI) was

greater than .80, and root mean square residual (RMS) was less than .01.

Page 49: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

39

For the long form of the teacher scale (Conners, 1997), 38 of the 59 items were factor

analyzed, whereas the 21 rationally derived items were not included. The sample was divided

into two subsamples: EFA (n = 1,200) and CFA (n = 500). EFA was conducted first, and a six-

factor solution, which accounted for 63.1% of the variance, was identified. Then, the EFA

subsample was split into two smaller subsamples containing 600 participants each. Four-, five-,

and six-factor solutions were derived from each subsample, and comparability coefficients were

examined. Although values were not reported, Conners indicated that the six-factor solution had

the highest comparability coefficient. Therefore, the six-factor model was tested using CFA with

a replication sample of 500 teachers. Conners indicated that the model was a good fit to the data

(fit criteria are described above), but the goodness-of-fit values were not reported. The 38 items

from the CTRS-R long form that had the highest pattern coefficients on the Oppositional

Behavior, Cognitive Problems/Inattention, and Hyperactivity subscales were used to create the

38-item teacher short form. The complete 59-item CTRS-R (long form) has never been factor

analyzed.

Like the teacher scale, not all of the 80 items on the parent long form were included in

the factor analysis. Excluded were the 23 rationally-derived items. Based on a subsample of

1,100 parents, a seven-factor solution, derived via EFA, was retained and accounted for 50.8% of

the total variance. The parent sample was also divided into two subsamples, in which each

contained 550 participants. Thus, the same factor solutions were compared across samples. Of

the solutions (six-, seven-, and eight-factors) compared, the seven-factor solution had the highest

comparability coefficient. Thus, the seven-factor model was examined using CFA with a

separate replication sample of 1,100 parents. Conners (1997) indicated that the model was a

good fit to the data, but goodness-of-fit values were not reported. Finally, the parent items from

Page 50: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

40

the long form that had the highest pattern coefficients on the Oppositional Behavior, Cognitive

Problems/Inattention, and Hyperactivity subscales were used to create the 58-item short form of

the parent scale. Like the teacher scale, the complete 80-item long form of the parent scale has

never been factor analyzed.

Some of the analytic procedures used in the development of the CRS-R are incongruent

with recommendations for best practice in factor analysis and scale development. For example,

Conners (1997) employed a varimax rotation, despite the medium to high factor inter-

correlations reported for many CRS-R subscales. Best practices in EFA indicate that an oblique

rotation be used when factors are correlated, as varimax rotation is based on an assumption of

orthogonality and can result in the loss of important information (Costello & Osbourne, 2005).

Further, findings from Monte Carlo studies (e.g., Zwick & Velicer, 1986) indicate that using

multiple criteria for factor retention (e.g., parallel analysis and minimum average partials) and

setting .40 or .50 as a cut-off for saliency of variables can increase the replicability of findings

and reduce the extent to which findings are the result of chance. Finally, Conners’ criteria for

evaluating goodness-of-fit were somewhat liberal. The current criteria for model fit are higher

(e.g., CFI greater than .95; non-normed fit index (NNFI) greater than .95, and standardized root

mean residual (SRMR) less than .08; Hu & Bentler, 1995; Markland, 2006). In addition to using

outdated procedures and criteria in factor analyses, Conners employed questionable procedures

in the development of both scales. For example, in establishing the factor structures for both

long forms, Conners initially ran an EFA in a large sample to identify a solution only to split the

sample into subsets and run EFAs again to identify the best solution to test with CFA. Why not

split the sample first, run the EFA to identify factor solutions, and use CFAs to compare? These

issues raise questions regarding the utility of CRS-R scores solely on the basis of the research

Page 51: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

41

conducted during the development of the scales. Findings from independent studies raise

additional questions. Further, factor structures of the short forms of the Conners’ scales have not

always replicated in independent studies.

Gergardstein, Lonigan, Cukrowitz, and McGuffy (2003) conducted an EFA on an

adapted version of the teacher short form with a sample of 235 predominantly African American,

low income, pre-school students (no other demographic information was provided). Maximum

likelihood extraction and an unspecified oblique rotational method were used. Factors were

retained using the eigenvalue rule of one, which resulted in the selection of a three-factor

solution. This solution was considered the most interpretable, had the fewest number of cross-

loadings, and had an adequate number of salient pattern coefficients on each factor. Because the

item content of the factors differed from Conners’ (1997) findings, Gergarstein et al. named the

factors: Hyperactivity/Impulsivity, Inattention, and Oppositional Defiant Behavior.

Using the short form of the parent scale with a sample of 200 predominantly Caucasian

children and adolescents referred for psychiatric outpatient services, Kumar and Steer (2003)

found a two-factor structure (PAF extraction and promax rotation) instead of the three-factor

structure identified by Conners (1997). A visual scree test was used to determine the number of

factors to retain. Six items were identified as salient on an Oppositional factor. The remaining

21 items were considered salient on an overall ADHD factor.

Gau et al. (2006) examined the factor structure of a Chinese translation of both the

CTRS-R and CPRS-R short forms with a sample of 2,584 children living in Taiwan. PAF

extraction and a varimax rotation were used, and visual inspection of a scree plot was used to

determine the number of factors to retain for retention. A three-factor structure was identified

Page 52: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

42

for each scale, and the factors were named Inattention/Cognitive Problems, Hyperactivity/

Impulsivity, and Oppositional. Percent of variance accounted for were not reported.

In summary, three independent studies have been conducted on CRS-R scores. Two

studies, one conducted in the US (teacher form; Gergardstein et al., 2003) and one conducted in

Taiwan (both scales; Gau et al., 2006), obtained a three-factor structure comparable to what

Conners (1997) found for both scales with the standardization sample. One study conducted in

the US (parent form; Kumar & Steer, 2003) identified a two-factor structure that bears little

resemblance to the one identified by Conners. Notably, all three studies were conducted on the

short forms of the scales and employed the same criteria Conners used. The extent to which the

factor structures of scores from the long forms of the scales are replicable is unknown. Further,

using the eigenvalue rule of one and the visual scree plot for factor retention can increase the

likelihood of overfactoring (Henson & Roberts, 2006). No research has been conducted on the

CRS-R using current factor retention criteria to determine whether the factor structure is

sufficient or reflective of a reduced factor solution.

Concurrent validity. Conners’ (1997) research on the concurrent validity of the CTRS-R

and CPRS-R is limited to examination of correlations between the two scales. Correlation

coefficients between corresponding factors on the CTRS-R and CPRS-R (long forms) ranged

from .12 to .55. Additionally, other studies (e.g., Mitsis, McKay, Schulz, Newcorn, & Halperin,

2000; Tripp, Schaughency, & Bronwyn, 2006) have examined the concurrent validity of CRS-R

short form scores by examining the relationship between parent and teacher ratings. In general,

research on parent-teacher agreement on ADHD rating scales indicates that scores provided by

teachers often differ from those provided by parents (Pelham, 2001). Giannaris, Golden, and

Greene (2001) conducted a meta-analysis of US studies that had examined the concurrent

Page 53: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

43

validity of the parent scale (Conners, 1997). The findings indicated that CPRS-R (long and short

forms) scores were related to scores on a wide variety of other measures, including other

behavior rating scales (e.g., the Child Behavior Checklist; Achenbach, 1991) and neurological

measures (i.e., continuous performance tasks and vigilance tasks).

Discriminant and predictive validity. Conners, Sitarenios, Parker, and Epstein (1998)

examined the predictive validity of all subscales of the teacher’s long form using a sample of 439

adolescents who were divided into three groups: (a) ADHD, (b) clinical control (students

diagnosed with emotional problems), and (c) non-clinical control. A series of one-way

ANCOVAs that controlled for age were conducted. The ADHD group scored higher than the

non-clinical control group on all subscales except Social Problems. Differences between the

ADHD and clinical control group were in the expected directions and also statistically significant

for all subscales. Conners et al. used discriminant function analysis to examine the extent to

which CTRS-R subscale scores differentiated children with ADHD from non-clinical controls.

CTRS long form subscale scores had 78.1% sensitivity, 91.3% specificity, and an overall correct

classification rate of 84.7%. However, the stability of the findings has been questioned. Snyder,

Drozd, and Xenakis (2004) criticized Conners et al.’s use of the same clinical samples in both the

development and validation of the revised Conners’ scales, as this practice is likely to lead to

increased measurement error. The authors note,

A discriminant analysis must always be validated with a fresh sample [sic],

otherwise the analysis overemphasizes random factors in the original ADHD

sample and results in artificially inflated classification accuracies that reveal

nothing about the actual validity of the rating scale. (p. 1189)

Page 54: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

44

As predicted by Snyder et al., Conners’ findings pertaining to the predictive validity of

CRS-R scores have not always been replicated in independent studies.

Charach, Chen, Hogg-Johnson, and Schacacar (2009) examined the predictive validity of

subscale scores from the long form of the teacher scale (Conners, 1997) with a sample of 1,038

Canadian children ages 6 to 12, who had been referred to an outpatient treatment center for

attentional, learning, or behavioral problems. This study is unique in that a large sample was

used and individual differences (e.g., sex and presence of learning disabilities) that may

influence the likelihood of misclassification (i.e., false positive or negative classification as

ADHD) were examined. Findings from discriminant function analysis indicated that low scores

on the CTRS-R DSM subscales ruled out ADHD diagnosis, but high scores on the same

subscales did not confirm the presence of ADHD. T scores on the teacher’s DSM subscale of 60

or higher yielded high sensitivity estimates (91-94%), but poor specificity estimates (32-53%). T

scores of 80 or higher were necessary to yield adequate specificity estimates (88-93%). Further,

findings revealed that false classifications (positive and negative) were most prevalent among

girls with language or reading impairments and boys with oppositional behaviors (Charach et al.,

2009). Notably, the specificity estimate obtained by Charach et al. was substantially lower than

that reported by Conners et al. (1998).

With a sample of 237 adolescents (55 were female), Conners et al. (1998) examined the

discriminant validity of the long form of the parent scale. The sample had three groups (ADHD,

clinical control, and non-clinical control), which were matched on age and sex. One-way

ANCOVAs, controlling for age, revealed a statistically significant effect for diagnostic group.

The ADHD group scored higher than the non-clinical group on all CPRS-R subscales, except for

Perfectionism and higher than the emotional problems group on all subscales, except Anxious-

Page 55: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

45

Shy, Psychosomatic, and Emotional Lability. Findings from discriminant function analysis for

CPRS-R subscale scores were (a) 92.3% sensitivity, (b) 94.5% specificity, and (c) an overall

classification rate of 93.4%. As was true for the teacher scale and noted by Snyder et al. (2004),

Conners’ et al. use of the same sample in the development and validation of the CPRS-R is a

critical shortcoming, which may influence the extent to which the findings are replicable.

Deb, Dhaliwal, and Roy (2008) used a sample of parents and teachers of 151 children

with mental retardation or borderline intellectual functioning and examined the extent to which

scores on the teacher and parent short forms (Conners, 1997) discriminated between children

with and without co-morbid ADHD. Receiver operating characteristics analysis was used to

identify cut-points with optimal sensitivity and specificity. While parent scores discriminated

between children with and without ADHD, the teacher scores did not. Further, Deb et al. noted

that many of the items on the CTRS-R and CPRS-R were not applicable to nonverbal children in

the sample; thus, a different measure may need to be used to assess ADHD symptomology for

nonverbal children.

Giannaris et al. (2001) examined the predictive validity of the Conners’ Parent Rating

Scale in its various iterations (the revised version, the original scale, and adapted versions) scores

in a meta-analysis of available literature. They concluded, “The CPRS appears to be a reliable

and valid tool in assessing general psychopathology, but seems to fall short in its ability to

discriminate along diagnostic lines” (p. 1085). In other words, the Conners’ Parent Rating

Scale (various versions and forms) may be effective in differentiating children with ADHD from

typical children but not in discriminating children with ADHD from children with other forms of

psychopathology. The long form of the CPRS-R is designed to be used as a broadband measure,

assessing several domains of psychopathology and allowing for the assessment of alternate or

Page 56: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

46

comorbid diagnoses; however, it may be more appropriately used as a narrow band screening

measure.

Summary of research. In summary, some independent studies have replicated Conners’

(1997) findings from research with the standardization sample, while others have reported

differing results. Findings from studies on the construct validity of CTRS-R and CPRS-R scores

have been particularly variable. Pertaining to the factor structure of scores, some studies (Gau et

al., 2006; Gergardstein et al., 2003) identified factor structures similar to those reported by

Conners, while others (Kumar & Steer, 2003) reported finding a reduced factor structure. In

regard to discriminant and predictive validity, it seems that CTRS-R and CPRS-R scores

effectively differentiate children with ADHD from non-clinical controls, but do not discriminate

children with ADHD from those with other disorders as Conners had intended. These findings

are relatively stable despite the shortcomings of Conners’ samples used in discriminant analyses

on the scales.

Substantial gaps in the literature on the CTRS-R and CPRS-R are evident. One, the

majority of independent psychometric research has focused on the short forms of the scales,

leaving the long forms largely unexamined. Two, nearly all existing studies have examined

properties of scores from the empirically-derived CTRS-R and CPRS-R scales, and the

rationally-derived items remain unstudied. Three, all factor analytic studies of CTRS-R and

CPRS-R scores have used outdated factor retention criteria. Thus, whether the factor structures

of the scores can be replicated when current criteria are used has yet to be investigated. Finally,

these gaps are amplified as no cross-cultural research has been conducted to determine whether

the factor structure of the long forms of the CRS-R would emerge in a non-US population.

Page 57: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

47

Present Study

The purpose of the present study was to examine whether the factor structure of scores

from the long forms of the CTRS-R and CPRS-R (Conners, 1997) with seven- and eight-year-old

children living in the Sarlahi district of Nepal was similar to the factor structures identified in the

normative sample (Conners, 1997). However, the present study was the first study to examine

the factor structure of all items included on the CTRS-R and CPRS-R and the first study to

employ the current criteria for factor analysis in examining the structure of the scores. The

primary research questions were the following:

1. Were the factor structures of scores identified in the empirical items of the

CTRS-R (6) and full CTRS-R (11) replicated with Nepalese children?

2. Were the seven-factor structure of scores from the empirical items of the CPRS-

R and the 12-factor structure of scores from the full CPRS-R replicated with a

sample of Nepalese children?

3. Did different models identified through subsample EFAs fit the data better than

those identified with the normative sample?

It was hypothesized that alternative models identified through subsample EFAs would be a better

fit to the data than the models identified using the normative sample.

Page 58: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

48

Method

Sample

The two datasets used in this study were based on seven- to eight-year-old children living

in the Sarlahi district of Nepal. One was obtained from their teachers and the other was obtained

from their parents. All of the data were provided by researchers conducting a larger study titled

Nepal Nutrition Intervention Project–Sarlahi (NNIPS; Tielsch et al., 2006). NNIPS is an

ongoing longitudinal study investigating the effects of prenatal and early childhood

micronutrient supplementation on long-term outcomes such as survival, cognition, and

physiological functioning. The Sarlahi district was selected as the site for the NNIPS study

because it is culturally, socio-economically, and demographically similar to many areas on the

Indian subcontinent (Tielsch et al., 2006)

Teacher-rated sample. At the time of data collection, 1,380 children attended school

and were rated by their teachers on the CTRS-R (Conners, 1997). Data on caste and gender of

teacher raters were not available. Approximately 27% of this sample was randomly drawn to

form an EFA sample (n = 374). The remaining cases served as the CFA sample (n = 1,000).

Teacher-rated EFA subsample. In this teacher-rated sample, approximately 49% of the

children were female. Four caste levels were represented: Brahmin (10%), Chhetri (11%),

Vaishya (66%), and Shudra (6%). Approximately 6% of children were non-Hindu. In regard to

region of ancestry, approximately 37% of the children were considered to be of hills ancestry

(Pahadi), and 63% were of Terai ancestry (Madhesi).

Teacher-rated CFA subsample. Approximately 47% of children in the teachers’ CFA

sample were female. Four caste levels were represented in the sample: 8% Brahmin, 10%

Chhetri, 65% Vaishya, and 9% Shudra level (7% were non-Hindu). Approximately, 33% of the

Page 59: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

49

children were of Pahadi ancestry, and 67% identified as Madhesi. A comparison of the two

teacher-rated samples indicated that their frequency distributions were, as expected, nearly

identical on demographic features (sex, caste, and region of ancestry).

Parent-rated sample. CPRS-R data were available for 1,835 children. Like the teacher-

rated sample, approximately a third of the sample (n = 555) was randomly drawn to form an EFA

parent-rated sample. The remaining cases were designated as the CFA sample (n = 1,253).

Parent-rated EFA subsample. Females represented approximately 51% of the parent-

rated sample. Individuals from four caste levels were represented in the sample: 59% Vaishya,

13% Shudra, 9% Chhetri, and 8% Brahmin. Approximately 10% of children in the sample were

non-Hindu. Children were primarily identified as Madhesi (71%), and 27% were identified as

Pahadi. Most of the children in this sample (81%) started school, but 19% had never attended

school. The majority of parents rating their children were mothers (90%), with the rest

representing a number of other caregivers: 4% fathers, 3% grandparents, 1% aunts or uncles, and

1% other caregivers. In approximately 33% of the cases, one or more additional caregiver

(usually the child’s father or grandparent) was present during data collection. Discussion of

items by additional caregivers was permitted, but only one caregiver (the person who spent the

most time caring for the child) was allowed to provide ratings.

Parent-rated CFA subsample. Approximately 50% of the children in the CFA parent-

rated sample were girls. Regarding caste designation, 63% of children were Vaishya, 7% were

Brahmin, 9% were Chhetri, 11% were Shudra, and 10% were non-Hindu. About 71% were

identified as Madhesi and 29% of children were identified as Pahadi. The majority of the

children in the CFA parent-rated sample (80%) had started school, but 20% had never attended

school. Typically, the mothers (89%) rated their child, with the rest of the children rated by

Page 60: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

50

fathers (5%), grandparents (3%), aunts or uncles (2%), or other caregivers (1%). In

approximately 35% of the cases, more than one caregiver was present during data collection.

Visual inspection of frequency tables indicated that both parent-rated samples were nearly

identical on demographic features (e.g., sex, caste, school entry, and region of ancestry).

Teacher- and parent-rated sample comparisons. The teacher- and parent-rated

samples were not equivalent on demographic features. The parent-rated sample included more

girls, Madhesis, non-Hindus, Shudras, and Vaishyas than did the teacher-rated sample.

Demographic frequencies for the teacher-rated and parent-rated samples are provided in Table 1.

The primary reason for these demographic

Table 1 Demographic Frequencies for Teacher- (N = 1,380) and Parent-Rated (N = 1,822) Samples by Caste, Region of Ancestry, and Sex Sample Characteristics Teacher-Rated Sample n (%) Parent-Rated Sample n (%)

Caste Brahmin 123 (8.91%) 129 (7.08%) Chhetri 143 (10.36%) 160 (8.78%) Vaishya 904 (65.51%) 1130 (62.01%) Shudra 111 (8.04%) 208 (11.42%) Non-Hindu 90 (6.52%) 182 (9.99%) Sex Male 727 (52.68%) 907 (49.78%) Female 653 (47.32%) 915 (50.21%) Region of Ancestry Pahadi 474 (34.34%) 509 (27.94%) Madhesi 895 (64.86%) 1296 (71.13%)

differences was school status. Eighty percent of children in the parent-rated sample versus 100%

in the teacher-rated sample had entered school at the time of data collection.

Two cross-tabulation analyses were conducted on the entire parent-rated sample to

examine differences in schooling status in relation to sex and caste. A 2 (sex) x 2 (school status)

Page 61: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

51

cross-tabulation analysis was statistically significant (χ2 = 18.73, df = 1, p < .001), with a small

effect size (Cramer’s V = .10). Boys were more likely to have been enrolled in school than girls.

A 5 (caste) x 2 (school status) cross-tabulation analysis was also statistically significant (χ2 =

199.77, df = 5, p < .001), with a medium effect size (Cramer’s V = .33). Children of designated

higher castes (Brahmin and Chhetri) were more likely to have been enrolled in school than

children of designated lower castes (Vaishya, Shudra, and non-Hindus). Further, within all caste

groups (excluding Brahmin) fewer girls than boys had entered school. Frequency data for sex,

Hindu caste status, and school entry status are provided in Table 2.

Table 2 Frequencies and Percentages for School Status of Children in the Parent-Rated Sample by Caste, Region of Ancestry, and Sex (N = 1,822) Caste

Sex

Never Started School n (%)

Started School n (%)

n

Brahmin Boys 1 (1.72%) 57 (98.28%) 58

Girls 0 (0.00%) 72 (100.00%) 72 Chhetri Boys 0 (0.00%) 82 (100.00%) 82

Girls 8 (10.25%) 70 (89.74%) 78 Vaishya Boys 73 (12.70%) 502 (87.30%) 575

Girls 111 (19.61%) 455 (80.39%) 566 Shudra Boys 32 (32.99%) 65 (67.01%) 97

Girls 55 (49.12%) 57 (50.89%) 112 Non-Hindu

Boys 37 (39.36%) 57 (60.64%) 94

Girls 48 (54.55%) 40 (45.45%) 88

Measures

The NNIPS study contained several measures used to assess the children’s physical and

psychological functioning (e.g., IQ tests, blood tests, and dietary evaluations). Only the parent

and teacher ratings from Conners’ Rating Scales–Revised were examined in this study. Several

Page 62: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

52

items were also provided by NNIPPS to describe the demographic features of the samples: caste,

region of ancestry, age, sex, and schooling status.

The long forms of the teacher and parent Conners’ Rating Scales–Revised (CRS-R;

Conners, 1997) were used. The CTRS-R (59 items) and CPRS-R (80 items) were designed to be

comprehensive measures of child psychopathology with a focus on ADHD symptomatology

(Conners, 1997). Each item describes a particular behavior (e.g., argues with adults), with

parents and teachers instructed to read each item and indicate the frequency with which the child

has engaged in the behavior within the past month. Response options are “Not at all,”

“Occasionally,” “Often,” and “Very often.” Detailed descriptions, including psychometric

properties, of the CTRS-R and CPRS-R are provided on pages 31-46.

Back translation procedures. The back translation procedure used on the CRS-R was

based on the method recommended by Brislin (1970). The scales were translated from English

into Nepali by a team of three male researchers, who were also the primary investigators of the

study (two Nepalese and one American). Then, the scales were back-translated into English by a

Nepali female who was never exposed to the original English versions of the scales. Additional

information about the translators and their level of fluency was not available.

The quality of the back-translations of each item was evaluated using a multi-prong

approach in which (a) statistical salience, (b) expert review, and (c) cultural congruence were

considered (Brislin, 1970). First, two EFAs were conducted (one on the CTRS-R and one on the

CPRS-R) to evaluate whether the items related to one another in the expected fashion. (EFA

procedures are reported in the data analysis section.) Items that did not have adequate pattern

coefficients on the expected factors were flagged as potentially having translation problems.

Then, CRS-R items were submitted to an expert panel for review of the equivalence of the back-

Page 63: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

53

translated and original items. The expert panel consisted of five psychologists (one educational

psychologist, one clinical psychologist, and three school psychologists), all employed in

university settings and unaffiliated with the study. The expert reviewers reported having

expertise in several areas such as ADHD, scale development, cross-cultural psychological

measurement, assessment of psychopathology, cultural diversity, and psychopathology. Expert

reviewers identified their ethnicity and gender as follows: (a) three Caucasian American males,

(b) one Caucasian American female, and (c) one Asian American male.

Members of the expert panel were asked to review the equivalence of the original and

back-translated items of the CTRS-R and CPRS-R (Conners, 1997) in respect to three domains

(Brislin, 1970): (a) the extent to which the wording of the back-translated item was similar to

that of the original item, (b) the extent to which the intended meaning of the original item was

reflected in the back-translated item, and (c) the extent to which the back-translated item

appeared to reflect the intended construct. Using each domain, expert reviewers rated each back-

translated item relative to the corresponding original item on a five-point Likert-type scale (1 =

Very much to 5 = Not at all). A copy of the form used by expert reviewers is contained in

Appendix A. After the expert review panel was completed, a native Nepali-speaking, male,

doctoral student, who specialized in applied linguistics, was consulted about the CTRS-R and

CPRS-R items (Conners, 1997). The linguistics consultant reviewed the English versions of the

items and commented on the degree to which each item could be easily translated into Nepali

and the extent to which the behaviors specified in the items were likely to occur and be viewed

as troublesome for Nepali families.

Teacher scale. Reviewer responses indicated that most CTRS-R (Conners, 1997) items

were adequately translated. Ten items were identified as problematic in regard to wording,

Page 64: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

54

meaning, or construct reflection. For example, the original wording of one item was “not

reading well.” The corresponding back translated item read “cannot study according to his/her

potential.” Expert reviewers identified the item as problematic because the intended meaning of

the original item was not reflected in the back-translated item. A list of these poorly translated

items on the CTRS-R is provided in Appendix B.

Parent scale. The majority of CPRS-R items were identified as adequately translated by

the expert review panel. Reviewers flagged 13 items as problematic due to issues with wording,

meaning, or perceived construct reflection. For example, the original wording of one item was

“does not listen when spoken to,” but the back translated item read “pretends not to pay attention

to others.” Reviewers indicated that this item was problematic because the back translated item

was incongruent with the original item in regard to intended meaning and construct reflection. A

list of the 13 poorly translated CPRS-R items is also provided in Appendix B.

Although the scales were typically administered in Nepali, an unknown number of parent

respondents were not fluent in the Nepali language. For these parents, the CPRS-R scales were

administered in Maithili, an Indo-Aryan language derived from Sanskrit (Pandey, 2006). Data

were not collected on the frequency in which the Maithili version of the CPRS-R was used.

Furthermore, the Maithili version of the CPRS-R was not back-translated into English as the

Nepali version was. Code-switching (incorporating both Nepali and Maithili in speech) is

common in the Sarlahi district, and the extent to which such code-switching was used could not

be determined. All participating teachers were fluent in Nepali, and all CTRS-R scales were

administered in Nepali.

Page 65: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

55

Procedure

General procedure. The NNIPS study employed a community-based sampling

procedure in which all individuals who lived in the geographic area and met eligibility criteria

were recruited for participation. Children were considered eligible for the study if they were 7-8

years old and their mother had participated in the prior NNIPS micronutrient supplementation

trial during pregnancy. If a child was deemed eligible, an NNIPS team member made a home

visit and invited the family to participate in this follow-up study.

Informed consent was obtained from all adult respondents. Parents provided verbal

consent, and teachers provided written consent. After parental consent was provided, the study

was explained to the child, and verbal assent was obtained. Next, NNIPS research assistants

collected data during two home visits and one school visit, and the family was asked to make one

visit to a central NNIPS site. On the first home visit, research assistants (a) conducted a

structured interview to gather information about socio-economic status (SES), maternal and

child’s diet, and child’s schooling status; (b) tested household salt for iodine; (c) conducted a

child behavioral observation; (d) administered the Ten Questions Plus interview (a screener for

severe disabilities; Durkin et al., 1994) and the CPRS-R; and (e) gave the child an acti-cal watch

designed to measure levels of movement and activity. The first home visit typically lasted

approximately 1.5 hours. During the second home visit, research assistants (a) administered the

Home Observations for Measures of Environment inventory (HOME; Bradley, Caldwell, Rock,

Hamrick, & Harris, 1988) to assess factors such as access to learning materials and parental

support; (b) conducted a child behavior observation; and (c) removed the acti-cal watch. The

second home visit generally took 1-1.5 hours. During the families’ visits to the central NNIPS

site, a variety of assessments were administered: (a) child and maternal IQ tests, (b) child test

Page 66: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

56

session behavior ratings, (c) tests of fine and gross motor skills, (d) measures of child and

maternal anthropometry, (e) child executive functioning tests, and (f) child and maternal

hemoglobin measures. The central site visits typically lasted 5 hours. Finally, a school visit was

completed for children who attended school. The CTRS-R (Conners, 1997) was administered

during the school visit, which typically lasted 1-2 hours. In exchange for their participation in

the study, parents received a small, stainless steel cooking utensil, teachers received a work

diary, and children received a small toy. Data on sex, date of birth (age), caste, and region of

ancestry were collected in prior versions of the NNIPS study and retrieved from extant

databases. Data were not collected on ethnicity and language of participating children and

families.

Conners’ procedures. The CTRS-R and CPRS-R (Conners, 1997) are typically

administered in written form. However, because of high rates of illiteracy among women of

childbearing age in Nepal (76%; Central Bureau of Statistics, 2005), both scales were

administered in a structured interview format. Parent interviews were conducted in families’

homes, and teacher interviews were conducted in the classroom. The average length of the

administration of either scale was not recorded. All interviewers were research assistants who

were male and Nepali; other demographic information was not available (i.e., caste, ancestry,

and age). Prior to data collection, all research assistants completed a two-day training session

and participated in daily supervised practice for two weeks. During data collection, the research

assistants read response choices aloud, the participants provided responses, which were circled

by the research assistants. During the data collection process, the integrity of interview

administration was monitored by project supervisors.

Page 67: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

57

Data Management and Analysis

Parallel steps were used to analyze the data from the teacher and parent scales separately.

In the first step, the data were screened, the samples were randomly divided into EFA and CFA

samples, and preliminary analyses were conducted on each sample. Next, EFAs were conducted

on the CTRS-R or CPRS-R (Conners, 1997) items to derive viable factor solutions. Once viable

solutions were obtained, CFA was used to test the fit of each factor model to the data. An

overview of the theoretical foundations of EFA and CFA, and the criteria used to interpret the

findings are summarized below.

EFA. EFA is a statistical technique commonly used in scale development and

measurement research. In EFA, responses to items or variables are considered observed

variables. Latent constructs that are presumed tapped by observed variables are considered the

factors. The objective of EFA is to evaluate the construct validity of scores from a measure, in

which linear combinations of observed variables are identified. Researchers determine the extent

to which scores on a scale are reflective of latent constructs (factors) based on the extent to

which patterns of linear combinations of observed variables are consistent with expectations

based on theory and prior research on the construct (Byrne, 2005). Although the terms EFA and

principle components analysis (PCA) are sometimes used interchangeably, important differences

exist between the two procedures. In EFA, linear combinations of observed variables are derived

based on common variance among the observed variables, and unique variance is partialled out.

In PCA, all variance is treated as common variance, which can result in artificially inflated

pattern and structure coefficients (Preacher & MacCallum, 2003). For this reason, EFA is

considered to be preferable to PCA when the objective of the research is identification of latent

variables, and PCA is recommended only when the sole objective of the research is data

Page 68: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

58

reduction (e.g., Byrne; Costello & Osborne, 2005; Preacher & MacCallum, 2003). The most

current procedures and criteria were used in running and interpreting the EFAs.

Criteria. Common factor analysis was selected over PCA because the intent of the study

was to identify latent factor structures (Fabrigar, Wegener, MacCallum, & Strahan, 1999). The

extraction method used was principal axis factoring (PAF) because of its relative tolerance of

multivariate non-normality in the data and demonstrated ability to recover weak factors (Briggs

& MacCallum, 2003). Squared multiple correlations were used to estimate the initial

communalities (Gorsuch, 2003). Several procedures were used to determine the appropriate

number of factors for retention and rotation, as this approach is considered best practice (Henson

& Roberts, 2006): (a) parallel analysis (PA; Horn, 1965), (b) minimum average partials (MAP;

Velicer, 1976), and (c) a visual scree test (Cattell, 1966). Due to the nature of the construct

(ADHD), it was assumed that the factors would be correlated (e.g., Barkley, 2006; Conners,

1997). Thus, two correlated rotations (Promax [k = 4; Gorsuch, 1997; Tataryn, Wood, &

Gorsuch, 1999] and Oblimin) were run and compared. (Findings from analyses with Promax

rotation were reported if no meaningful structural differences were evident through the other

rotation method.) Secondary criteria for determining factor adequacy were established a priori.

Pattern coefficients equal to or greater than .40 were considered salient (Stevens, 2002). In the

interest of parsimony and to honor simple structure (Thurstone, 1947), items that had salient

coefficients on more than one factor were not used in interpreting the factors. Factor structures

were considered adequate for interpretation if each factor met the following criteria: (a) a

minimum of 4 items with salient pattern coefficients, (b) internal consistency score reliability

greater than or equal to .70, and (c) the presence of a theoretically meaningful pattern. Viable

EFA solutions were the basis for some of the models tested in CFA.

Page 69: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

59

CFA. CFA is another statistical tool that is used to evaluate the construct validity of

scores from a scale. In CFA, the nature of and relation among latent variables are explicitly

tested through the specification of a priori models that reflect hypothetical patterns of

associations between indicators (test scores or item responses) and latent factors (Jackson, Purc-

Stevenson, & Gillaspy, 2009). Indicators are observed variables measured by responses to

variables (e.g., test items). Latent factors are theoretical constructs, which are measured

indirectly through indicators. In CFA, the term parameter refers to a measurable facet of a

model (factor coefficients, factor correlation coefficients, and error terms). Parameters are

estimated in CFAs, and different models are based on fixing specific parameters and leaving free

others (Gillaspy, 1996). The objective of CFA is to determine the extent to which the estimated

parameters that produce an estimated population covariance matrix reflect a sample covariance

matrix (Tabachnick & Fidell, 1996). An alternative model approach was used to test the

covariance matrices for all specified models because the viability of a scale’s structure is best

attained through such an approach (MacCallum, Wegener, Uchino, & Fabrigar, 1993).

Before a model can be tested, it is necessary to ascertain that the model is theoretically

identified. In other words, “A model is said to be identified if it is theoretically possible to

derive a unique estimate of each parameter” (Kline, 2005, p. 105). According to Kline (2005), a

model is identified if (a) the number of free parameters is less than or equal to the number of

observations, and (b) a metric scale is assigned to every latent variable. The number of

parameters is equal to the number of variances and covariances of exogenous (observed)

variables and loadings. The number of observations equals x (x + 1) / 2, and x is equal to the

number of observed variables. If a model has more parameters than observations, then the model

is considered under-identified, and estimation of the parameters is not theoretically possible. If

Page 70: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

60

the model has an equal number of parameters and observations, it is considered to be just-

identified, and if the number of observations exceeds the number of parameters, the model is

over-identified. If a model is just- or over-identified, then it is theoretically possible to estimate

a unique solution for the model (Kline, 2005). All of the models tested in the present study were

over-identified based on the criteria specified.

Criteria. EQS 5.8 for Windows was used for all CFAs, which were conducted on

covariance matrices based on raw data. Because deviations from normality were evident,

maximum likelihood (ML) estimation using a robust method was employed and the Satorra-

Bentler χ2 statistic was used to test all models (Kline, 2005). Goodness of fit was evaluated

based on multiple criteria (Tanaka, 1993). In particular, three types of fit indices were

considered: incremental, absolute, and predictive (Worthington & Whittaker, 2006). Each type

of fit index provides different information, resulting in different evaluations of model to data fit.

Incremental indices reflect improvements in model fit by comparing the proposed model to a

baseline (usually a null, which means no fit between the data and the model) model. Absolute fit

indices can be used to determine the degree to which the proposed model explains the

relationships identified in the sample data. Finally, the predictive fit indices are used to predict

the extent to which a model would fit data from another sample with the same population

(Worthington & Whittaker, 2006). All three types of fit indices were used to evaluate model fit,

and consistency in findings was considered to be indicative of model fit. Incremental fit indices

that were evaluated were the comparative fit index (CFI), non-normed fit index (NNFI; Bentler

& Bonnet, 1980), and incremental fit index (IFI; Bollen, 1989). CFI, NNFI, and IFI values

above .95 were considered indicative of good fit (Hu & Bentler, 1995). Overall, the extent to

which fit indices were consistently high or low was considered most important in evaluating

Page 71: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

61

model fit (Markland, 2006). The RMSEA is an absolute fit index that represents the lack of

model fit when compared to a perfect model. RMSEA values equal to or less than .06 were

considered indicative of good fit (Hu & Bentler, 1995). The Akaike Information Criterion (AIC;

Akaike, 1987) and Consistent Akaike Information Criterion (CAIC; Bozdogan, 1987) are

predictive fit indices for which lower values were considered indicative of good model fit.

However, the AIC and CAIC were interpreted with caution because research indicates that these

values may artificially favor simpler models and are sensitive to sample characteristics (e.g.,

sample size; Byrne, 2006). The results of the Satorra-Bentler χ2 tests are reported but were not

interpreted because of a high likelihood of error associated with χ2 statistics in analyses with

large samples (Lei & Wu, 2007).

Page 72: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

62

Results

Conners’ Teacher Scale

Preliminary Analyses

Data on the CTRS-R (N = 1,387) were screened for missing values. A total of 7 cases

with one or more missing data points (less than 5% of the total cases) were identified. Using the

Mahalanobis distance test (Tabachnick & Fidell, 1996), six extreme multivariate outliers (p <

.001) were identified. No systematic pattern of missing data and outliers was evident; thus, these

cases were deleted listwise (Roth & Switzer, 1999). Removal of these cases resulted in 1,374

viable cases for statistical analysis. As noted in the Method section, approximately a third (27%;

n = 374) of the cases were randomly selected from the sample and used for EFA, with the

remaining 1,000 cases used for CFA.

EFA

Preliminary analyses. Scores for the items on CTRS-R met the assumptions for EFA

(linearity, presence of moderate correlations [│.01-.67│; Mdn = .20], and absence of

multicollinearity). Scores for 10 items exhibited severe skew (> 3) or kurtosis (> 7; Curran,

West, & Finch, 1996), an indication of some non-normality in the data. Means, standard

deviations, skew and kurtosis for both samples of the CTRS-R items are presented in Table 3.

Linearity was supported through visual inspection of scatterplots, and the presence of moderate

correlations and absence of multicollinearity were confirmed through the standard and

reproduced correlation matrices. A correlation matrix for teacher items in the EFA sample is

provided in Appendix C.

Page 73: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

63 Ta

ble

3

Mea

ns a

nd S

tand

ard

Dev

iatio

ns o

f Ite

ms

on th

e C

onne

rs’ T

each

er S

cale

EFA

Sam

ple

CFA

Sam

ple

Item

#/ I

tem

Su

bsca

le

M

SD

Sk

ew

Kur

tosi

s M

S

D

Skew

K

urto

sis

1. A

cts d

efia

ntly

Opp

ositi

onal

.22

.

59

3.08

9.

73

.18

.

48

2.99

9.

51

2. R

estle

ss a

nd sq

uirm

y H

yper

activ

ity

.

38

.74

2.

08

3.77

.

36

.71

2.

10

3.92

3. F

orge

ts m

ater

ial a

lread

y

le

arne

d

C

ogni

tive

Prob

lem

s/

Inat

tent

ion

DSM

-Ina

ttent

ive

.86

.

88

.89

.

16

.91

.

94

.79

-.3

0

4. U

nacc

epte

d by

gro

up

Soci

al P

robl

ems

.

31

.72

2.

38

4.87

.

28

.70

2.

54

5.55

5. E

asily

hur

t/off

ende

d A

nxie

ty

.

30

.62

2.

29

5.47

.

29

.61

2.

33

5

.59

6. P

erfe

ctio

nist

ic

Perf

ectio

nism

1.60

1.

14

-.11

-1.3

9 1.

65

1.10

-.

22

-1.

28

7. T

empe

r out

burs

ts

Opp

ositi

onal

CG

I- Em

otio

nal L

abili

ty

.30

.

62

2.24

4.8

5 .

22

.55

2.

81

8.58

8. Im

puls

ive,

eas

ily e

xcite

d H

yper

activ

ity

AD

HD

Inde

x

.47

.

77

1.66

2.

12

.45

.

75

1.72

2.

40

9. C

arel

ess m

ista

kes i

n w

ork

DSM

-Ina

ttent

ive

.

64

.86

1.

20

.57

.

63

.88

1.

24

.54

10. O

verly

bol

d O

ppos

ition

al

.

13

.44

4.

01

18.

02

.14

.

48

3.78

15

.01

11. O

n th

e go

/ driv

en b

y m

otor

Hyp

erac

tivity

D

SM-H

yper

activ

e

1.25

1.

10

.21

-1.3

3 1.

29

1.09

.

16

-1.3

1

Page 74: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

64

Tabl

e 3

(con

tinue

d)

EFA

Sam

ple

CFA

Sam

ple

Item

#/ I

tem

Su

bsca

le

M

SD

Sk

ew

Kur

tosi

s M

SD

Sk

ew

Kur

tosi

s 12

. Avo

ids s

usta

ined

men

tal

effo

rt

Cog

nitiv

e Pr

oble

ms/

In

atte

ntio

n

DSM

-Ina

ttent

ive

.66

.

89

1.18

.

42

.67

.

92

1.21

.4

2

13. I

s ch

osen

last

for t

eam

s So

cial

Pro

blem

s

.49

.

89

1.70

1.

64

.48

.

89

1.75

1.

77

14. I

s em

otio

nal

Anx

iety

1.07

.

96

.

33

-1

.02

1.02

.

95

.47

-.8

8

15. E

very

thin

g m

ust b

e pe

rfec

t Pe

rfec

tioni

sm

.

70

.94

1.08

-.0

4 .

71

.93

1.

02

-.17

16. R

estle

ss/ h

ighl

y ac

tive

CG

I- R

estle

ss/

Impu

lsiv

e

.60

.

91

1.35

.

69

.67

.

99

1.23

.1

8

17. F

ails

to fi

nish

thin

gs s

tarte

d C

ogni

tive

Prob

lem

s/

Inat

tent

ion

AD

HD

Inde

x

.72

.

92

1.04

-.0

2 .

64

.88

1.25

.

62

18. D

oes n

ot se

em to

list

en

DSM

-Ina

ttent

ive

.

49

.75

1.

51

1.66

.

51

.81

1.50

1.

33

19. A

ctiv

ely

defie

s/ re

fuse

s to

com

ply

Opp

ositi

onal

.

17

.53

3.

54

12.

90

.16

.5

0 3.

67

14.

15

20. L

eave

s sea

t H

yper

activ

ity

.

14

.43

3.

37

12.

01

.15

.4

7 3.

83

16.

23

21. P

oor s

pelli

ng

Cog

nitiv

e Pr

oble

ms/

In

atte

ntio

n

.95

1.

01

.75

-.6

0 1.

05

1.06

.

61

-.90

Page 75: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

65

Tabl

e 3

(con

tinue

d)

EFA

Sam

ple

CFA

Sam

ple

Item

#/ I

tem

Su

bsca

le

M

SD

Sk

ew

Kur

tosi

s M

S

D

Skew

K

urto

sis

22. D

oesn

’t ha

ve fr

iend

s So

cial

Pro

blem

s

.01

.

10

9.55

8

9.72

.

04

.29

7.76

6

5.79

23. T

imid

, eas

ily sc

ared

A

nxie

ty

.

48

.74

1.

52

1.77

.

52

.83

1.56

1.

58

24. C

heck

s thi

ngs r

epea

tedl

y Pe

rfec

tioni

sm

.

76

.98

.

97

-.32

.79

.9

5 .

87

-.44

25. C

ries f

requ

ently

and

eas

ily

Anx

iety

C

GI-

Emot

iona

l La

bilit

y

.21

.

51

2.71

7.

64

.20

.5

3 3.

01

9.63

26. I

natte

ntiv

e/ d

istra

ctib

le

AD

HD

Inde

x

.64

.

85

1.13

.

34

.62

.8

6 1.

20

.44

27. D

iffic

ulty

org

aniz

ing

task

s D

SM-I

natte

ntiv

e

.59

.

86

1.32

.

78

.54

.8

4 1.

52

1.42

28. D

iffic

ulty

sust

aini

ng a

ttent

ion

D

SM-I

natte

ntiv

e

.40

.

73

1.80

2.

57

.36

.6

8 1.

95

3.31

29. D

iffic

ulty

aw

aitin

g tu

rn

Hyp

erac

tivity

DSM

-hy

pera

ctiv

e

.41

.

72

1.77

2.

57

.44

.7

7 1.

72

2.19

30. N

ot re

adin

g w

ell

Cog

nitiv

e Pr

oble

ms/

In

atte

ntio

n

.72

.

92

1.06

.

07

.79

.9

7 .

96

-.25

31. D

oesn

’t m

ake

frie

nds

Soci

al P

robl

ems

.

15

.49

3.

58

13.2

3 .

16

.53

3.81

1

4.94

32. S

ensi

tive

if cr

itici

zed

Anx

iety

1.16

1.

04

.40

-1

.06

1.09

.9

9 .

53

-.77

33. O

ver-

focu

sed

on d

etai

ls

Perf

ectio

nism

.67

.

91

1.22

.5

1 .

66

.92

1.24

.

50

Page 76: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

66

Ta

ble

3 (c

ontin

ued)

EFA

Sam

ple

CFA

Sam

ple

Item

#/ I

tem

Su

bsca

le

M

SD

Skew

K

urto

sis

M

SD

Skew

K

urto

sis

34. T

alks

too

muc

h

DSM

-hyp

erac

tive

.33

.

76

2.27

4.

08

.34

.7

7 2.

26

4.06

35. F

idge

ts

CG

I-re

stle

ss/

impu

lsiv

e

.76

.

98

1.06

-.0

8 .

75

.97

1.06

-.

05

36. D

istu

rbs p

eers

A

DH

D In

dex

.

30

.65

2.

31

4.86

.

31

.66

2.

36

5.44

37. A

rgue

s with

adu

lts

Opp

ositi

onal

.09

.

38

4.52

2

1.65

.

08

.34

4.

90

27.

30

38. C

anno

t sta

y st

ill

AD

HD

Inde

x

.48

.

75

1.49

1.

45

.44

.

74

1.79

2.

79

39. R

uns/

clim

bs e

xces

sive

ly

Hyp

erac

tivity

.29

.

61

2.20

4.

60

.26

.

59

2.47

6.

18

40. L

acks

inte

rest

in sc

hool

C

ogni

tive

Prob

lem

s/

Inat

tent

ion

.42

.

79

1.85

2.

51

.45

.

83

1.83

2.

38

41. P

oor s

ocia

l ski

lls

Soci

al P

robl

ems

.

41

.72

1.

80

2.80

.

49

.82

1.

64

1.77

42. D

iffic

ulty

pla

ying

qui

etly

H

yper

activ

ity D

SM-

hype

ract

ive

.29

.

64

2.39

5.

34

.28

.

63

2.38

5.

35

43. L

ikes

thin

gs c

lean

/ord

erly

Pe

rfec

tioni

sm

1.

75

1.00

-.

35

-.91

1.74

1.

03

-.37

-1.0

1

44. F

idge

ts w

ith h

ands

or f

eet

AD

HD

Inde

x D

SM-

hype

ract

ive

.51

.

82

1.52

1.

40

.54

.

85

1.49

1.

31

Page 77: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

67 Ta

ble

3 (c

ontin

ued)

EFA

Sam

ple

CFA

Sam

ple

Item

s Su

bsca

le

M

SD

Skew

K

urto

sis

M

SD

Skew

K

urto

sis

45. D

eman

ding

/ eas

ily fr

ustra

ted

CG

I-em

otio

nal

labi

lity

.45

.

79

1.76

2.

38

.47

.

76

1.61

1.

87

46. B

lurts

out

ans

wer

s

DSM

-hyp

erac

tive

.39

.

75

1.99

3.

27

.39

.

76

2.03

3.

36

47. S

pite

ful/

vind

ictiv

e

Opp

ositi

onal

.

17

.50

3.

29

11.

64

.22

.

60

3.12

9.

81

48. P

oor a

ttent

ion

span

AD

HD

Inde

x .

60

.81

1.

18

.56

.

64

.87

1.

15

.31

49. L

oses

thin

gs

Cog

nitiv

e Pr

oble

ms/

In

atte

ntio

n D

SM-

Inat

tent

ive

.51

.

63

1.03

.

92

.52

.

74

1.47

1.

92

50. O

nly

atte

nds i

f int

eres

ted

A

DH

D In

dex

1.

23

1.07

.

26

-1

.23

1.14

1.

04

.41

-1.0

6

51. T

imid

, with

draw

n A

nxie

ty

.

77

.92

.

93

-.17

.

72

.90

1.

09

.23

52. D

istra

ctib

ility

/ attn

. pro

blem

A

DH

D In

dex

.

65

.85

1.

11

.29

.

68

.88

1.

10

.24

53. A

lway

s doe

s thi

ngs s

ame

way

Pe

rfec

tioni

sm

.

49

.84

1.

63

1.60

.

52

.89

1.

56

1.22

54. M

ood

chan

ges q

uick

ly

CG

I-R

estle

ss

Impu

lsiv

e

.48

.73

1.

42

1.26

.

48

.74

1.

48

1.51

55. I

nter

rupt

s/ in

trusi

ve

AD

HD

Inde

x D

SM-

hype

ract

ive

.21

.54

2.

97

9.23

.

25

.62

2.

78

7.96

Page 78: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

68 Ta

ble

3 (c

ontin

ued)

EFA

Sam

ple

CFA

Sam

ple

Item

s Su

bsca

le

M

SD

Item

s Su

bsca

le

M

SD

Item

s Su

bsca

le

56. P

oor i

n m

ath

Cog

nitiv

e Pr

oble

ms/

In

atte

ntio

n

.89

1.03

.

82

-.60

.93

1.

07

.79

-.

72

57. D

oes n

ot fo

llow

thro

ugh

A

DH

D In

dex/

D

SM-I

natte

ntiv

e

.46

.78

1.

72

2.23

.

44

.78

1.

77

2.34

58. E

asily

dis

tract

ed

DSM

-Ina

ttent

ive

.9

1 .

87

.66

-.3

3 1.

02

.96

.

58

-.67

59. R

estle

ss, a

lway

s on

the

go

AD

HD

Inde

x .4

3 .

75

1.70

2.

01

.47

.

82

1.77

2.

21

Not

e. E

FA sa

mpl

e n

= 37

4, a

nd C

FA sa

mpl

e n

= 1,

000.

Page 79: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

69

Initial EFA. Common factor analysis (PAF extraction) was conducted on scores from

the 59-item on the CTRS-R (n = 374). The correlation matrix was factorable (determinant of the

matrix = 4.08 E-010; Kaiser-Meyer-Olkin statistic [KMO; Kaiser, 1974] = .88; Bartlett’s test of

sphericity [Bartlett, 1950]; χ2 = 7620.63, df = 1711, p < .001). The statistical significance of

Bartlett’s test is typically reflective of a large sample size rather than a lack of factorability of the

correlation matrix (Leech, Barrett, & Morgan, 2005). Promax (k = 4; Gorsuch, 1997) and

Oblimin rotations were run and compared. No meaningful differences were evident when

different rotational methods were used; thus, findings from Promax analyses are reported.

For factor retention, MAP (Velicer, 1976) suggested that four factors should be retained,

PA (Horn, 1965) indicated three factors, and visual inspection of the scree plot (Cattell, 1966)

indicated two factors. Conners (1997) reported a six-factor structure for the CTRS-R with the

normative sample. Because the recommended number of factors varied, six solutions were

examined, starting with the six-factor solution and ending with a one-factor solution. All factor

solutions, except two, were rejected because of failure to meet criteria for factor adequacy (i.e., a

minimum of four items with pattern coefficients greater than .40, no item cross-loadings greater

than .20 on multiple factors, internal consistency reliability of scores greater than or equal to .70,

and theoretical convergence). The one- and two-factor solutions met criteria for factor adequacy.

Instead of the one-factor solution, the two-factor solution was retained for interpretation because

it demonstrated greater theoretical convergence, higher communalities and factor coefficients,

and a larger percentage of variance accounted for (27.60%). Additionally, over-factoring is

preferable to under-factoring (Fabrigar, Wegener, MacCallum, & Strahan, 1999).

Page 80: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

70

The two-factor solution contained 33 items, of which 17 items were salient on Factor I.

Pattern coefficients for items on Factor I ranged from .48 to .74 (Mdn = .62), and communalities

ranged from .26 to .55 (Mdn = .41). Examples of item on Factor I included, “Forgets material

already learned,” “Poor in math,” and “Has difficulty organizing tasks.” Content of Factor I

items appeared to reflect an inattention/school problems dimension, and the factor was labeled

accordingly. The Inattention/School Problems factor accounted for 16.63% of the total variance,

and the reliability estimate (Cronbach’s α) of the scores was .91. Sixteen items were salient on

Factor II. Pattern coefficients for items on Factor II ranged from .40 to .78 (Mdn = .54), and

communality estimates ranged from .18 to .58 (Mdn = .30). Examples of items on Factor II

included, “Is restless, always on the go,” “Interrupts, intrusive,” and “Spiteful/ vindictive.”

Content of these items seemed to reflect a hyperactivity/oppositional dimension, which was the

label affixed to Factor II. The Hyperactivity/Oppositional factor accounted for 8.73% of the total

variance, and the Cronbach’s alpha for the scores was .88. A summary of the findings from the

two-factor solution is presented in Table 4.

Secondary EFA solutions. Viability of a scale’s structure is best achieved using an

alternative models approach (MacCallum, Wegener, Uchino, & Fabrigar, 1993). As such, three

different EFA solution criteria, based on rational and statistical considerations, were used to

retain items and identify the best possible models of the CTRS-R that could be tested through

CFA. Three 2-factor solutions for the CTRS-R were identified: (a) moderate rational solution

(hereinafter called the rational-2T solution), (b) the moderate statistical solution (statistical-2T

solution), and (c) the strict hybrid solution (hybrid-2T solution). The two-factor solution of the

CTRS-R derived in the initial EFA was composed of items that

Page 81: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

71

Tabl

e 4

Sum

mar

y of

the

Initi

al T

wo-

Fac

tor

Solu

tion

of C

TRS-

R S

core

s: P

atte

rn C

oeffi

cien

ts (P

AF

ext

ract

ion

with

Pro

max

Rot

atio

n) a

nd

Alte

rnat

ive

Solu

tions

Item

s Fa

ctor

h

2 A

ltern

ativ

e So

lutio

ns

Inat

tent

ion

Hyp

erac

tivity

/ O

ppos

ition

al

R

atio

nal-2

T St

atis

tical

-2T

Hyb

rid-2

T

12. A

void

s sus

tain

ed m

enta

l eff

ort

.74

.01

.55

X

X

X

17. F

ails

to fi

nish

thin

gs s

tarte

d .70

.05

.51

X

X

X

3. F

orge

ts m

ater

ial a

lread

y le

arne

d .69

-.09

.45

X

X

X

9. C

arel

ess m

ista

kes i

n w

ork

.68

.12

.53

X

X

X

30. N

ot re

adin

g w

ell

.68

-.01

.46

26

. Ina

ttent

ive,

dis

tract

ible

.66

.09

.49

X

X

X

56. P

oor i

n m

ath

.66

-.06

.41

X

27. H

as d

iffic

ulty

org

aniz

ing

task

s .64

.04

.42

X

X

X

21. P

oor s

pelli

ng

.62

-.07

.37

X

40. L

acks

inte

rest

in sc

hool

.62

-.03

.37

X

X

X

52. D

istra

ctib

ility

or a

ttent

ion

prob

lem

.59

.11

.41

X

X

X

18. D

oes n

ot se

em to

list

en

.57

.15

.40

13

. Is

chos

en la

st fo

r tea

ms

.54

-.17

.26

57

. Doe

s not

follo

w th

roug

h on

inst

ruct

ions

.52

.11

.32

X

28. H

as d

iffic

ulty

sust

aini

ng a

ttent

ion

.51

.04

.27

X

48. P

oor a

ttent

ion

span

.48

.21

.34

X

51. T

imid

, with

draw

n .48

-.21

.21

6.

Per

fect

ioni

stic

-.41

.08

.18

24

. Che

cks t

hing

repe

ated

ly

-.40

.20

.15

41

. Poo

r soc

ial s

kills

.

39

.05

.17

43

. Lik

es th

ings

cle

an/o

rder

ly

-.36

.02

.12

5.

Eas

ily h

urt/

offe

nded

.3

5 .0

1 .1

3

Page 82: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

72

Tabl

e 4

(con

tinue

d)

Item

s

Fact

or

h2

Alte

rnat

ive

Solu

tions

Inat

tent

ion

Hyp

erac

tivity

/ O

ppos

ition

al

R

atio

nal-2

T S

tatis

tical

-2T

Hyb

rid-2

T

58. E

asily

dis

tract

ed

.35

.21

.21

23

. Tim

id, e

asily

scar

ed

.34

-.04

.11

11

. On

the

go/ d

riven

by

mot

or”

-.3

4 .2

4 .1

2

31. D

oes n

ot m

ake

frie

nds

.27

-.10

.07

4.

Una

ccep

ted

by g

roup

.2

5 -.0

3 .0

6

32. S

ensi

tive

if cr

itici

zed

-.24

.09

.05

49

. Los

es th

ings

.2

4 .0

6 .0

7

25. C

ries f

requ

ently

and

eas

ily

.21

.06

.05

53

. Alw

ays d

oes t

hing

s sam

e w

ay

.17

-.03

.03

22

. Doe

sn’t

have

frie

nds

-.05

-.03

.01

59

. Res

tless

, alw

ays o

n th

e go

-.0

6 .78

.58

X

X

X

16. R

estle

ss/ h

ighl

y ac

tive

-.16

.73

.49

X

X

X

35. F

idge

ts

-.12

.69

.44

X

X

X

36. D

istu

rbs p

eers

-.0

1 .66

.44

55

. Int

erru

pts,

intru

sive

-.0

1 .66

.43

X

X

X

34. T

alks

too

muc

h -.0

9 .69

.34

X

X

39

. Run

s/ c

limbs

exc

essi

vely

.0

3 .57

.34

X

X

38

. Can

not s

tay

still

.0

4 .54

.31

X

2. R

estle

ss a

nd sq

uirm

y -.0

1 .54

.29

X

8. Im

puls

ive,

eas

ily e

xcite

d -.1

1 .52

.24

47

. Spi

tefu

l/ vi

ndic

tive

.06

.50

.27

54

. Moo

d ch

ange

s qui

ckly

-.0

3 .49

.24

1

. Act

s def

iant

ly

.05

.48

.24

44

. Fid

gets

with

han

ds o

r fee

t -.0

3 .47

.21

X

7. T

empe

r out

burs

ts

.13

.41

.22

Page 83: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

73

Not

e. n

= 3

74; h

2 =

com

mun

ality

; Sal

ient

pat

tern

coe

ffic

ient

s are

indi

cate

d in

bol

d.

Tabl

e 4

(con

tinue

d)

Item

s

Fact

or

h2

Alte

rnat

ive

Solu

tions

Inat

tent

ion

Hyp

erac

tivity

/ O

ppos

ition

al

R

atio

nal-2

T St

atis

tical

-2T

Hyb

rid-2

T

10. O

verly

bol

d .0

8 .40

.18

15

. Eve

ryth

ing

mus

t be

perf

ect

-.06

.38

.13

46

. Blu

rts o

ut a

nsw

ers

-.35

.36

.18

37

. Arg

ues

-.0

9 .3

5 .1

1

45. D

eman

ding

/ eas

ily fr

ustra

ted

.07

.35

.14

29

. Diff

icul

ty a

wai

ting

turn

.0

2 .3

3 .1

2

20. L

eave

s sea

t .2

3 .3

0 .1

9

19. A

ctiv

ely

defie

s ad

ults

.2

8 .3

0 .2

2

50. O

nly

atte

nds i

f int

eres

ted

.2

6 .2

7 .1

8

29. H

as d

iffic

ulty

pla

ying

qui

etly

.1

8 .2

6 .1

3

33. O

ver-

focu

sed

on d

etai

ls

-.12

.24

.06

14

. Is

emot

iona

l -.0

4 .0

5 .0

1

Eige

nval

ues

10.4

6 5.

82

% o

f var

ianc

e 17

.73

9.87

C

ronb

ach’

s α

.91

.88

Page 84: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

74

met the minimum criteria for saliency on either factor (pattern coefficient ≥ .40 and with no

cross-loadings ≥ .20) and were adequately translated. This solution was not selected as a model

to test via CFA because the other three solutions were stronger on the factor retention criteria

(i.e., pattern coefficients, communalities, and percent of variance accounted for).

EFAs of each identified solution were conducted in two phases to select the best items to

create models: (a) factor analysis of items on each factor and (b) factor analysis of all items

retained from the individual factor analysis. Separate factor analysis of only items on each factor

was conducted to determine whether all items created a viable one-factor solution or reflected

multiple factor structures. If removal of items was required to strengthen the one-factor

structure, then factor analysis was re-run without them to double-check the stability of the

solution. If the solution was better, these items were excluded from all subsequent analyses.

After each factor was run through EFA, a final EFA was conducted on the remaining items from

both factors. Items that failed to meet saliency criteria in the full solution analysis were excluded

from subsequent analyses. A breakdown of which items aligned with each solution is presented

in Table 4 depicting the original two-factor solution.

Rational-2T solution. CTRS-R items on the rational-2T solution were required to meet

the minimum saliency criteria in the initial EFA, to be adequately translated, and to meet

additional rational criteria. Items retained on Factor I were strictly germane to the construct of

inattention. Items that appeared to reflect other constructs, such as school or social problems,

were not included. Specifically, the items “Poor in math,” “Poor spelling,” “Is chosen for

teams,” and “Timid, withdrawn” were not included in this

solution. Items retained on Factor II were directly related to the construct of hyperactivity.

Items that appeared to reflect other constructs, such as oppositional behavior, were not included.

Page 85: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

75

Specifically, the items “spiteful/ vindictive,” “argues with adults,” and “acts defiantly” were not

included in the solution. Of the 34 items that met saliency criteria in the initial solution, a total

of 20 items made up this solution (11 on Factor I and 9 on Factor II).

EFA of each factor. Separate factor analysis of the two factors supported a one-factor

solution for each factor. All 11 items met retention and criteria on Factor I, and all nine items on

Factor II met criteria. No items were dropped from analyses.

Full EFA. EFA retention criteria supported a two-factor structure for the rational-2T

solution. All items (20) met saliency criteria (11 on Factor I and nine on Factor II). Items on

Factor I, named Inattention, reflected one construct: “Fails to finish tasks started,” “Poor

attention span,” and “Has difficulty organizing tasks and activities.” Items on Factor II appeared

to reflect a hyperactivity dimension (e.g., “Restless/ highly active,” “Talks too much,” and

“Cannot stay still”). Communalities, eigenvalues, percent of variance, and pattern coefficients

are presented in Table 5.

Statistical-2T solution. CTRS-R items for the statistical-2T solution were required to

meet more stringent statistical cutoffs (communalities ≥ .30 and pattern coefficients ≥ .55) than

the minimum criteria (no communality cutoff and pattern coefficients > .40) used in the initial

EFA and to be adequately translated. Of the 34 items that met saliency criteria in the initial two-

factor solution, 16 items made up this solution.

EFA of each factor. Factor analysis of each factor supported a one-factor solution. All

items met retention and saliency criteria on the respective factors (10 on Factor I and 6 on Factor

II). No items were dropped from analyses.

Full EFA. EFA of the 16 CTRS-R (Conners, 1997) items of the statistical-2T solution

supported a two-factor solution (MAP, PA, and scree). The first factor was named

Page 86: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

76

Table 5 Pattern Coefficients from EFA of the CTRS-R Rational-2T Solution (PAF Extraction, Promax Rotation) Items Factor I

Inattention Factor II

Hyperactivity

h2

12. Avoids sustained mental effort .78 -.07 .56 9. Careless mistakes in work .73 .05 .57

17. Fails to finish tasks started .73 -.01 .53 26. Inattentive/ distractible .71 .04 .52 3. Forgets material already learned .70 -.13 .45

27. Difficulty organizing tasks .68 -.02 .45 40. Lacks interest in school .63 -.09 .38 52. Distractibility/ attention problems .62 -.07 .42 57. Does not follow through on instructions .54 .08 .33 48. Poor attention span .54 .18 .29 28. Difficulty sustaining attention .53 .03 .38 59. Restless, always on the go -.01 .81 .65 16. Restless/ highly active -.10 .78 .58 35. Fidgets -.07 .75 .53 55. Interrupts/ intrusive .04 .58 .35 34. Talks too much -.04 .58 .32 39. Runs/ climbs excessively .09 .55 .34 2. Restless and squirmy .05 .53 .30

38. Cannot stay still .08 .52 .31 44. Fidgets with hands and feet .03 .46 .22 Eigenvalue 6.21 3.36 Percent of variance 28.30 14.07 Cronbach’s α .89 .84 Note. n = 374; h2= communality; Salient pattern coefficients are indicated in bold.

Inattention/School Problems, and included 10 items, such as “Fails to finish tasks started,” “Is

inattentive/ distractible,” and “Poor in math.” The six items on Factor II appeared to reflect a

hyperactivity dimension. Examples of items on Factor II are, “Restless/ highly active,”

“Interrupts and intrudes on others,” and “Talks too much” Communalities, eigenvalues, percent

of variance, and pattern coefficients for the statistical-2T solution are presented in Table 6.

Page 87: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

77

Hybrid-2T solution. CTRS-R items on the hybrid-2T solution were required to meet

more stringent statistical cutoffs (communalities ≥ .35 and pattern coefficients > .55) than the

ones used in the initial EFA (no communality cutoff and pattern coefficient > .40 solution), be

adequately translated, and meet additional rational criteria. Items retained on Factor I were

required to be strictly germane to the construct of inattention (not school or social problems), and

items retained on Factor II were strictly related to hyperactivity (not oppositional behavior). Of

the 34 items that met saliency criteria in the initial solution, 12 items made up this solution (eight

items on Factor I and four items on Factor II).

Table 6 Pattern Coefficients from EFA of the CTRS-R Statistical-2T Solution (PAF Extraction, Promax Rotation)

Items Factor I Inattention

Factor II Hyperactivity

h2

12. Avoids sustained mental effort .77 -.03 .58 17. Fails to finish tasks started .73 .04 .54 9. Careless mistakes in work .72 .10 .57 3. Forgets material already learned .70 -.10 .46

26. Inattentive/ distractible .69 .08 .51 56. Poor in math .67 -.05 .44 27. Difficulty organizing tasks .65 .04 .43 40. Lacks interest in school .62 -.08 .37 21. Poor spelling .61 -.06 .36 52. Distractibility/ attention problems .60 .08 .39 16. Restless/ highly active -.07 .82 .65 59. Restless, always on the go .02 .77 .60 35. Fidgets -.05 .76 .57 55. Interrupts/ intrusive .03 .60 .37 34. Talks excessively .02 .57 .32 39. Runs/ climbs excessively .08 .55 .33 Eigenvalue 5.46 3.03 Percent of variance 30.89 50.84 Cronbach’s α .89 .83 Note. n = 374. Salient pattern coefficients are indicated in bold; h2= communality.

Page 88: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

78

EFA of each factor. EFA of each factor supported one-factor structures: eight items on

Factor I and four items on Factor II. All items on each factor met retention and saliency criteria.

Thus, no items were dropped from analyses.

Full EFA. EFA also supported a two-factor structure for the 12 CTRS-R items making

up the hybrid-2T solution. Examples of items on Factor I, called Inattention, were “Fails to

finish tasks he/she starts,” “Is inattentive/ distractible,” and “Has difficulty organizing tasks and

activities.” Named Hyperactivity, Factor II was composed of four items, such as, “Restless/

highly active,” and “Interrupts intrusive.” Communalities, eigenvalues, percent of variance, and

pattern coefficients for this solution are presented in Table 7.

Table 7 Pattern Coefficients from EFA of the CTRS-R Hybrid-2T Solution (PAF Extraction, Promax Rotation Items

Factor I Inattention

Factor II Hyperactivity

h2

12. Avoids sustained mental effort .78 -.05 .59 9. Careless mistakes in work .75 .07 .59

17. Fails to finish tasks started .74 .02 .56 26. Inattentive/ distractible .69 .06 .50 3. Forgets material already learned .69 -.09 .46

27. Has difficulty organizing tasks and activities .66 .01 .43 40. Lacks interest in school .62 -.07 .37 52. Distractibility/ attention span problems .59 .09 .38 59. Restless, always on the go -.06 .84 .69 16. Restless/ highly active -.04 .79 .61 35. Fidgets .04 .75 .58 55. Interrupts/ intrudes on others .06 .57 .35 Eigenvalue 4.58 2.44 Percent of variance 34.03 16.75 Cronbach’s α .88 .82 Note. n = 374. Salient pattern coefficients are indicated in bold; h2 = communality.

Page 89: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

79

EFA summary. Three 2-factor solutions were identified through EFA: rational-2T,

statistical-2T, and hybrid-2T. The solutions were quite similar to one another. All reflected the

same two factors, Inattention and Hyperactivity, even though the solutions varied in the number

of items retained due to the criteria used. The rational solution (minimum cutoffs) contained the

largest number of items and the range of communalities and factor coefficients was wide,

whereas the hybrid-2T solution (the strictest cutoffs) contained the fewest number of items, had

the highest pattern/structure coefficients and communalities, and accounted for the greatest

amount of variance. The statistical solution fell in the middle on the same criteria and solution

output. All three solutions (rational-2T, statistical-2T, and hybrid-2T) were deemed appropriate

for further analysis in CFA.

CFA

Preliminary analyses. The CFA sample contained 1,000 participants. Assumptions of

CFA–linearity, multivariate normality, and specification of a correct model–were tested (Kline,

2006; Tomarken & Waller, 2005). Normality of the data was acceptable for most CTRS-R

(Conners, 1997) items, except for nine that exhibited severe skew (> 3) or kurtosis (> 8).

Descriptive statistics (means, standard deviations, skew, and kurtosis) for all CTRS-R items in

the dataset are provided in Table 3. Multivariate normality was examined based on Mardia’s

coefficient and normalized estimates, and the findings are reported with each model. Linearity,

examined through visual inspection of scatterplots, was met. Although specification of a correct

model cannot be explicitly tested, an alternative models approach was employed to gauge the

extent to which identified models fit the data relative to other models (MacCallum et al., 1993).

A correlation matrix of CTRS-R scores at the item level for the CFA sample is provided in

Appendix C.

Page 90: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

80

Models. CFAs (maximum likelihood extraction, robust) were conducted on the

covariance matrices of raw CTRS-R scores for five non-nested models: (a) 3 two-factor models

(rational-2T1, statistical-2T, and hybrid-2T) and (b) two Conners’ (1997) models (a six-factor

model and an 11-factor one). The six-factor model contained 37 items, was based on the

empirically-derived items of the CTRS-R, and was identical to the model tested by Conners with

the normative sample. This empirical six-factor model is hereinafter referred to as Conners-E6-

T. Conners’ eleven-factor model contained 59 items, was based on both empirically- and

rationally-derived items included on the CTRS-R, and has never been tested by Conners or

others. This eleven-factor model is hereinafter referred to as Conners-F11-T. A list of the

CTRS-R models tested through CFA, including criteria for item selection, is presented in Table

8. The models derived through EFA are depicted in Figure 1, and Conners-E6-T and -F11-T

models are depicted at factor level in Figures 2 and 3, respectively. The items and associated

factors of all the teacher items are provided in Table 3 (see pp. 63-66).

All models met the necessary conditions for identification proposed by Kline (2005).

The numbers of free parameters and observations in each CTRS-R model are listed in Appendix

D. To fix the scale of the models, a single item on each factor was fixed to one. On the

Conners-F11-T model, many items loaded on more than one factor. Start values for items that

loaded on more than one factor were adjusted based on theory and prior research (Byrne, 2006).

1For ease of presentation, the model names were shortened: (a) two-factor moderate rational to rational-2T, (b) two-factor moderate statistical to statistical-2T, (c) two-factor strict hybrid to hybrid-2T, (d) six-factor Conners’ empirical to Conners-E6-T, and (e) eleven-factor Conners’ full model to Conners-F11-T.

Page 91: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

81

Table 8 Summary of CTRS-R Models Tested through CFA Models # of items Item Selection Criteria Two-Factor Models Rational-2T 20 Items selected based on minimum statistical and additional

rational criteria; Factor I = Inattention, Factor II = Hyperactivity

Statistical-2T 16 Items selected based on moderate statistical criteria; Factor I = Inattention, Factor II = Hyperactivity

Hybrid-2T 12 Items selected based on strict statistical and rational criteria; Factor I = Inattention, Factor II = Hyperactivity

Conners’ (1997) Models Conners-E6-T 37 All empirically-derived CTRS-R items

Conners-F11-T 59 All CTRS-R items from all scales

Page 92: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

Rest

less

, on

the

go

SR

SSS SS SS SS

RRR RRR

R R

R

H HH H H

H

HH

R

SSSS

S S RR

R R T R R

R R

HH H H

Inat

tent

ion

Hyp

erac

tivi

ty/

Opp

osit

iona

l

Rest

less

, ov

erac

tive

Fidg

etin

g

Inte

rrup

ts/

Intr

udes

Run/

cl

imbs

Rest

less

, sq

uirm

y

Talk

s ex

cess

ivel

y

Cann

ot

rem

ain

still

Fidg

ets/

sq

uirm

s

Avo

ids

M

enta

l ef

fort

Forg

ets

Mak

es

care

less

err

ors

Fails

to

finis

h

Poor

in

spel

ling

Diff

icul

ty o

rgan

izin

g

No

sust

aini

ng

atte

ntio

n

Lack

s in

tere

st

i Easi

ly

dist

ract

ed

Shor

t att

enti

on

span

Dis

trac

tibi

lity

prob

lem

Poor

inm

ath

No

follo

wth

roug

h

Fig

ure

1. C

onne

rs’ t

each

er m

odel

s de

rive

d th

roug

h E

FA

s. S

= st

atis

tical

-2T,

R =

ratio

nal-2

T, H

= h

ybrid

-2T.

82

Page 93: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

Opp

osit

iona

l

Cogn

itiv

e Pr

oble

ms/

In

atte

ntio

n

Hyp

erac

tivi

ty

Soci

al P

robl

ems

Anx

ious

-Shy

Perf

ecti

onis

m

6 it

ems

8 it

ems

7 it

ems

5 it

ems

6 it

ems

6 it

ems

Fig

ure

2. C

onne

rs-E

6-T

depi

cted

at t

he fa

ctor

leve

l.

83

Page 94: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

Opp

ositi

onal

CP/

I

Hyp

erac

tivi

ty

SP Anx

ious

-Shy

Perf

ecti

onis

m

6 ite

ms*

8 ite

ms*

7 ite

ms*

5 ite

ms

6 ite

ms

6 ite

ms

AD

HD

Inde

x

CGI:

RI

CGI:

EL

DSM

: IA

DSM

: HI

6ite

ms

4 ite

ms*

9ite

ms*

9ite

ms*

12 it

ems

Fig

ure

3. C

onne

rs-F

11-T

dep

icte

d at

the

fact

or le

vel.

CP/

I = C

ogni

tive

Prob

lem

s/ In

atte

ntio

n; S

P =

Soci

al P

robl

ems;

CG

I: R

I =

Con

ners

’ Glo

bal I

ndex

: Res

tless

-Im

puls

ive;

CG

I: EL

= C

onne

rs’ G

loba

l Ind

ex: E

mot

iona

l Lab

ility

; DSM

: IA

= D

SM-I

V In

atte

ntiv

e;

DSM

: HI =

DSM

-IV

Hyp

erac

tive/

Impu

lsiv

e. *

indi

cate

s tha

t one

or m

ore

item

s are

incl

uded

on

mul

tiple

fact

ors.

84

Page 95: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

85

Each model was compared to its respective null and one-factor models. Then, the five

non-nested models were compared to one another to identify the model with the best fit to the

data. Multiple fit indices were examined to evaluate goodness of fit (Markland, 2006). Unless

indicated otherwise, all the chi-square tests were statistically significant, which was possibly

due to sample size (Kline, 1999). As a result, this statistic was not given weight in evaluating

goodness of fit of model to data.

Rational model. The Mardia coefficient indicated multivariate nonnormality in the

CTRS-R (Conners, 1997) data; thus, maximum likelihood extraction, the robust method, was

used. Three cases were identified as outliers based on Mardia’s normalized estimates. No

substantial changes in the model fit to the data were evident when the three cases were removed

and the analyses were re-run without them; therefore, the original analysis with the three cases

was used.

The null rational model was not supported, as the chi-square values were statistically

significantly lower for the one- and two-factor rational models than for the null. Also, the

findings from fit statistics did not support the one-factor rational model (e.g., CFI = .64; RMSEA

= .10). However, the goodness of fit indices provided strong support for the rational-2T (the

two-factor rational model; e.g., CFI = .96; RMSEA = .035). All unstandardized parameter

estimates were statistically significant (α = .01) for the rational-2T model. Standardized

coefficients for this model of the CTRS-R are reported in Table 9. The correlation between the

factors was moderate and statistically significant (r = .36; p < .001).

Statistical model. The Mardia coefficient indicated multivariate nonnormality in the

CTRS-R data; thus, maximum likelihood extraction, robust, was used. Three outlying cases

were identified based on Mardia’s normalized estimates. Removing these cases and re-running

Page 96: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

86

the analyses without them did not substantially change the model fit to the data. Therefore, the

original analysis was used with the three cases.

The null statistical model was not supported, as the chi-square values were statistically

significantly lower for the one- and two-factor statistical models than the null. Also, the fit

Table 9

Standardized Pattern Coefficients from CFA (Maximum Likelihood, Robust) of the Rational-2T Model of the CTRS-R (n = 1,000) Items Factor PC SE R2

12. Avoids sustained mental effort 1 .72 .70 .51 9. Careless mistakes in work 1 .70 .71 .50

17. Fails to finish tasks started 1 .69 .72 .48 26. Inattentive/ distractible 1 .71 .71 .50 3. Forgets material already learned 1 .74 .74 .45

27. Has difficulty organizing tasks 1 .63 .78 .40 40. Lacks interest in school 1 .87 .73 .48 52. Distractibility/ attention problems 1 .68 .74 .46 57. Does not follow through on instructions 1 .51 .79 .37 48. Poor attention span 1 .64 .77 .40 28. Difficulty sustaining attention 1 .48 .88 .27 59. Restless, always on the go 2 .78 .63 .61 16. Restless/ highly active 2 .77 .64 .59 35. Fidgets 2 .77 .64 .60 55. Interrupts/ intrusive 2 .57 .82 .33 34. Talks too much 2 .51 .86 .26 39. Runs/ climbs excessively 2 .51 .84 .29 2. Restless and squirmy 2 .62 .79 .38

38. Cannot stay still 2 .59 .81 .35 44. Fidgets with hands and feet 2 .54 .83 .30

Note. PC = pattern coefficient; SE = standard error; R2 = standardized factor loading squared.

indices for the one-factor model indicated a poor fit (e.g., RMSEA = .10; CFI = .70), whereas the

fit indices supported the statistical-2T (e.g., RMSEA = 03; CFI = .97). All unstandardized

Page 97: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

87

parameter estimates were statistically significant at the .01 level, and standardized coefficients

for the statistical-2T model of the CTRS-R are reported in Table 10. The correlation between the

two factors was small and statistically significant (R = .18; p < .001).

Teacher hybrid model. The Mardia coefficient indicated multivariate nonnormality in

the sample; thus, maximum likelihood extraction, robust, was used. Two outlying cases were

identified based on Mardia’s normalized estimates, the outliers were removed, and the analyses

were run without them for comparative purposes. No meaningful changes in model fit were

Table 10 Standardized Pattern Coefficients from CFA (Maximum Likelihood, Robust) for the Statistical-2T Model of the CTRS-R (n = 1,000) Items

Factor PC SE R2

12. Avoids sustained mental effort 1 .73 .68 .53 17. Fails to finish tasks started 1 .70 .72 .49 9. Careless mistakes in work 1 .70 .71 .49 3. Forgets material already learned 1 .68 .73 .46

36. Inattentive/ distractible 1 .69 .73 .47 56. Poor in math 1 .68 .74 .50 27. Has difficulty organizing tasks 1 .63 .77 .40 40. Lacks interest in school 1 .69 .72 .48 21. Poor spelling 1 .67 .74 .45 52. Distractibility/ attention problems 1 .64 .77 .40 16. Restless/ highly active 2 .80 .60 .64 59. Restless, always on the go 2 .76 .66 .57 35. Fidgets 2 .80 .60 .65 55. Interrupts/ intrusive 2 .55 .84 .30 34. Talks too much 2 .51 .86 .26 39. Runs/ climbs excessively 2 .52 .86 .27 Note. PC = pattern coefficient; SE = standard error; R2 = standardized factor loading squared

evident upon removal of the two outliers; thus, the original analysis was used. The null model

was not supported, as the chi-square values were statistically significantly lower for the one- and

Page 98: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

88

two-factor hybrid models than for the null. Also, the one-factor hybrid model was not supported

by the findings (e.g., CFI = .71; RMSEA = .12), but the fit indices provided strong support for

the hybrid-2T (the two-factor hybrid model; CFI = .99; RMSEA = .03). The Satorra-Bentler chi-

square for the hybrid-2T model was not statistically significant. All unstandardized parameter

estimates were statistically significant (α = .01), and the standardized coefficients for the hybrid-

2T model are reported in Table 11. The correlation between the factors was statistically

significant (R = .33; p < .001).

Teacher empirical model (Conners-E6-T). The Mardia coefficient indicated multivariate

nonnormality in the data; thus, maximum likelihood extraction, robust, was used.

Table 11

Standardized Pattern Coefficients from CFA (Maximum Likelihood, Robust) of the Hybrid-2T Model of the CTRS-R (n = 1,000) Items

Factor PC SE R2

12. Avoids sustained mental effort 1 .72 .69 .52 9. Careless mistakes in work 1 .72 .70 .51

17. Fails to finish tasks started 1 .70 .70 .49 26. Inattentive/ distractible 1 .70 .71 .50 3. Forgets material already learned 1 .67 .67 .46

27. Has difficulty organizing tasks 1 .63 .78 .40 40. Lacks interest in school 1 .69 .72 .48 52. Distractibility/ attention problems 1 .63 .76 .42 58. Restless, always on the go 2 .76 .66 .57 16. Restless/ highly active 2 .81 .59 .65 35. Fidgets 2 .81 .59 .65 55. Interrupts/ intrusive 2 .53 .85 .28 Note. PC = pattern coefficient; SE = standard error; R2 = standardized factor loading squared.

Thirteen outlying cases were identified based on Mardia’s normalized estimates, the outliers

were removed, and the analyses were run without them for comparative purposes. No

Page 99: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

89

meaningful changes in model fit were evident upon removal of the thirteen outliers; thus, the

original analysis was used.

The null empirical model was not supported, as the chi-square values were statistically

significantly lower for the one- and six-factor empirical models than for the null. Also, the

findings did not support the one-factor empirical (e.g., CFI =.63; RMSEA = .04) or the Conners-

E6-T (e.g., CFI = .05; RMSEA = .71) models. Parameter estimates (standardized pattern

coefficients and standard errors) are not reported as the values may be misleading. For

comparative purposes, the models were examined a second time after the removal of six poorly

translated items, but no meaningful changes in model fit were evident after removing these items

(e.g., six-factor model CFI = .83; RMSEA = .05).

Teacher full model (Conners-F11-T). Multivariate non-normality was present in the

data, evidenced by the Mardia coefficient; thus, maximum likelihood extraction, robust, was

used. Thirteen outliers were identified and removed based on the normalized estimates. No

meaningful differences in fit were revealed when the analyses were conducted without the

outlying cases; thus, the findings with the thirteen outlying cases are reported.

The null empirical model was not supported, as the chi-square values were substantially

lower for the one- and 11-factor full models than for the null. Also, the findings did not support

the one-factor (e.g., CFI = .61; RMSEA = .06) or the Conners-F11-T (e.g., CFI = .78; RMSEA =

.04) models. Parameter estimates (standardized pattern coefficients and standard errors) are not

reported as the values may be misleading. For comparative purposes, the models were re-

examined after removal of six poorly translated items from the sample, but no notable changes in

model fit were evident (e.g., eleven-factor model CFI = .80; RMSEA = .04).

Page 100: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

90

Model comparisons. A summary of fit statistics for all CTRS-R models tested is

provided in Table 12. Satorra-Bentler χ2 difference tests with Bonferroni correction were used to

test model distinguishability (Levy & Hancock, 2007; Satorra & Bentler, 2001). Because

analyses of the hybrid-2T model yielded the highest fit statistics (i.e., CFI, RMSEA, IFI, and

NNFI), this model was tested against the other two-factor models, rational-2T and statistical-2T.

Statistically significant differences were found between the hybrid-2T model and each of the

other models (p < .001). Therefore, the fit of hybrid-2T model to the data was distinguishable

from the fit of the other EFA-derived models. The hybrid-2T model had lower Satorra-Bentler

χ2, AIC, and RMSEA values and higher CFI, NNFI, and IFI values than the other models; thus, it

was identified as the best model to fit the data of the CTRS-R. A summary of model comparison

statistics is provided in Table 13.

Page 101: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

91

Tabl

e 12

Sum

mar

y of

Fit

Stat

istic

s fo

r C

FA

s (M

axim

um L

ikel

ihoo

d, R

obus

t) o

f CTR

S-R

Mod

els

(n =

1,0

00)

Mod

el

d

f

S-B

χ2

NN

FI

IFI

CFI

(R

obus

t) R

MSE

A

RM

SEA

(9

0% C

I)

A

IC

C

AIC

Rat

iona

l-T

Nul

l 17

1 4

,502

.02*

-

-

- 4,

160.

02

3,14

9.79

One

-Fac

tor

152

1,7

21.3

9*

.59

.6

4 .6

4 .

102

.097

- .1

06

1,41

7.39

51

9.41

Two-

Fact

or

151

3

37.0

7*

.95

.9

6 .9

6 .

035

.030

- .0

40

35.0

7 -8

57.0

0

St

atis

tical

-T

Nul

l 12

0 3

,939

.08*

-

-

-

3,6

99.0

8 2,

990.

05

O

ne-F

acto

r 10

4 1

,237

.24*

.

66

.71

.70

.10

4 .0

99 -

.110

1,

029.

24

414.

83

Tw

o-Fa

ctor

10

3

211

.55*

.

97

.97

.97

.03

2 .0

26 -

.039

5.

55

-602

.95

Hyb

rid-T

N

ull

66

2,7

94.9

7*

- -

-

2,66

2.47

2,

272.

56

O

ne-F

acto

r 54

865

.54*

.

64

.71

.71

.

122

.115

- .1

29

748.

54

429.

52

Two-Factor

53 88.54

.98

.99

.99

.026

.016 - .035

-17.46

-330.57

Con

ners

’ Em

piric

al

Nul

l 70

3 5

,808

.79*

-

-

- -

-

4,40

2.79

24

9.64

One

-Fac

tor

665

2,5

78.8

0*

.60

.6

3 .6

3 .

054

.051

- .0

56

1,24

7.81

-2

,679

.85

Si

x-Fa

ctor

65

0 1

,748

.05*

.

77

.79

.79

.04

1 .0

39 -

.043

44

8.05

-3

,391

.99

Con

ners

’ Ful

l

N

ull

1,71

1 12

,606

.75*

-

-

- -

-

9,18

4.75

-9

23.4

2

One

-Fac

tor

1,65

2 3

,126

.79*

.

55

- .6

1 .

051

.002

- .0

55

3,12

6.79

-6

,632

.83

El

even

-Fac

tor

1,58

4 4

,034

.27*

.

68

- .7

8 .

041

.039

- .0

43

866.

27

-8,4

91.6

2 N

ote.

CFA

= c

onfir

mat

ory

fact

or a

naly

sis;

S-B

χ2

= Sa

torr

a-B

entle

r chi

-squ

are;

NN

FI =

Ben

tler-

Bon

net N

on-n

orm

ed fi

t ind

ex; I

FI =

B

olle

n’s f

it in

dex;

CFI

= c

ompa

rativ

e fit

inde

x; R

MSE

A =

root

mea

n sq

uare

err

or o

f app

roxi

mat

ion;

CI =

con

fiden

ce in

terv

al; A

IC =

A

kaik

e’s

Info

rmat

ion

Crit

eria

; CA

IC =

Con

sist

ent A

kaik

e In

form

atio

n C

riter

ia. *

p <

.05.

Sel

ecte

d m

odel

is in

bol

d.

Page 102: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

92

Table 13

Findings from Satorra-Bentler χ2 Difference Tests with Bonferroni Correction

Models χ2 df T p

Hybrid-2T 88.54 - - -

Statistical-2T 211.55 50 94.08 <.001

Rational-2T 337.07 98 92.28 <.001

Note. All models were compared with the Hybrid-2T model

Page 103: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

93

Conners’ Parent Scale

Preliminary Analyses

The CPRS-R scores (N = 1,835) were screened for missing data. A total of 13 cases with

one or more missing data points (less than 5% of the cases) were identified. Using the

Mahalanobis distance test (Tabachnick & Fidell, 1996), a total of 14 cases were identified as

extreme outliers. Under both circumstances, no systematic pattern of missing data was evident.

Thus, 27 cases were deleted listwise (Roth & Switzer, 1999). Removal of these cases resulted in

a final CPRS-R sample of 1,808 cases, with 555 cases designated as the EFA subsample and

1,253 cases labeled as the CFA subsample.

EFA

Preliminary analyses. Scores for items in the CPRS-R EFA subsample met the

assumptions of linearity, presence of moderate correlations (i.e., │.01│- │.56│; Mdn = .15), and

absence of extreme multicollinearity. Scores for six CPRS-R items exhibited severe skew (> 3)

or kurtosis (> 7; Curran et al., 1996), an indication of some non-normality in the data.

Descriptive statistics (means, standard deviations, skew, and kurtosis) of scores for the parent

items are reported in Table 14. Visual inspection of scatterplots supported linearity, and visual

inspection of standard and reproduced correlation matrices confirmed the presence of moderate

correlations and the absence of multicollinearity. A correlation matrix for the parent items is

provided in Appendix C.

Initial EFA. Data from the 80 items of the CPRS-R (Conners, 1997) were submitted for

common factor analysis (PAF extraction). The correlation matrix was factorable (determinant of

the matrix = 1.22 to the -11 exponent; KMO = .89, Kaiser, 1974). Bartlett’s test of sphericity

(Bartlett, 1950)

Page 104: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

94 Ta

ble

14

Mea

ns a

nd S

tand

ard

Dev

iatio

ns o

f Ite

ms

on th

e C

onne

rs’ P

aren

t Sca

le

EFA

Sam

ple

CFA

Sam

ple

Item

#/ I

tem

Su

bsca

le

M

SD

S

kew

K

urto

sis

M

SD

Sk

ew

Kur

tosi

s 1.

Ang

ry/ r

esen

tful

Opp

ositi

onal

1.44

1.

05

.26

-1.1

4 1.

43

1.03

.2

9 -1

.07

2. D

iffic

ulty

fini

shin

g ho

mew

ork

Cog

nitiv

e Pr

oble

ms/

In

atte

ntio

n

.83

1.10

.9

4 -.6

2 .7

9 1.

10

1.04

-.4

5

3. "

On

the

go"/

driv

en b

y a

mot

or

Hyp

erac

tivity

D

SM-H

yper

activ

e

1.53

1.

19

-.15

-1.5

1 1.

44

1.18

-.0

2 -1

.50

4. T

imid

, eas

ily sc

ared

A

nxio

us-S

hy

.6

1 .9

0 1.

43

1.10

.5

6 .8

6 1.

54

1.51

5. E

very

thin

g m

ust b

e pe

rfec

t Pe

rfec

tioni

sm

1.

05

.99

.58

-.74

1.11

1.

00

.53

-.78

6. D

oesn

’t ha

ve fr

iend

s So

cial

Pro

blem

s

.01

.10

10.4

2 10

6.98

.0

1 .1

1 9.

25

83.7

1

7. S

tom

ach

pain

Ps

ycho

som

atic

.45

.71

1.74

2.

95

.43

.68

1.71

3.

03

8. F

ight

s O

ppos

ition

al

1.

43

1.03

.1

9 -1

.10

1.37

1.

05

.26

-1.1

2

9. A

void

s sus

tain

ed m

enta

l eff

ort

Cog

nitiv

e Pr

oble

ms/

In

atte

ntio

n A

DH

D In

dex

DSM

-Ina

ttent

ive

.79

1.04

1.

00

-.39

.80

1.05

.9

9 -.4

1

10. D

iffic

ulty

sust

aini

ng a

ttent

ion

D

SM-I

natte

ntiv

e

.51

.87

1.58

1.

37

.41

.78

1.93

2.

87

Page 105: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

95

Tabl

e 14

(con

tinue

d)

EFA

Sam

ple

CFA

Sam

ple

Item

#/ I

tem

Su

bsca

le

M

SD

Skew

K

urto

sis

M

SD

Sk

ew

Kur

tosi

s 11

. Arg

ues

Opp

ositi

onal

.86

.99

.91

-.29

.78

.95

1.01

-.0

2

12. F

ails

to fi

nish

ass

ignm

ents

C

ogni

tive

Prob

lem

s/

Inat

tent

ion

.91

1.07

.8

1 -.7

2 .7

9 1.

00

.99

-.28

13. H

ard

to c

ontro

l in

stor

es

Hyp

erac

tivity

.37

.77

2.19

4.

04

.35

.72

2.21

4.

38

14. S

care

d of

peo

ple

Anx

ious

-Shy

.29

.67

2.66

6.

97

.30

.67

2.50

6.

01

15. C

heck

s thi

ngs r

epea

tedl

y Pe

rfec

tioni

sm

.4

8 .8

0 1.

65

1.93

.5

0 .8

3 1.

60

1.62

16. L

oses

frie

nds q

uick

ly

Soci

al P

robl

ems

.8

6 .9

5 .8

5 -.3

0 .8

6 .9

5 .8

8 -.2

1

17. A

ches

/ pai

ns

Psyc

hoso

mat

ic

.3

4 .6

5 2.

20

5.0

3 .3

5 .6

6 2.

08

4.25

18. R

estle

ss/ h

ighl

y ac

tive

CG

I: R

estle

ss/

Impu

lsiv

e

1.66

1.

09

-.30

-1.2

0 1.

58

1.11

-.1

7 -1

.32

19. T

roub

le c

once

ntra

ting

in c

lass

C

ogni

tive

Prob

lem

s/

Inat

tent

ion

AD

HD

Inde

x

.39

.85

2.11

3.

18

.37

.81

2.23

3.

83

20. D

oes n

ot se

em to

list

en

DSM

-Ina

ttent

ive

.9

9 .9

9 .6

7 -.6

4 1.

05

1.02

.5

8 -.8

4

21. L

oses

tem

per

Opp

ositi

onal

1.30

.9

7 .3

0 -.8

8 1.

27

1.01

.3

5 -.9

4

22. N

eeds

clo

se su

perv

isio

n

Cog

nitiv

e Pr

oble

ms/

In

atte

ntio

n

1.15

1.

13

.40

-1.2

9 1.

13

1.11

.4

6 -1

.19

Page 106: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

96 Ta

ble

14 (c

ontin

ued)

EFA

Sam

ple

CFA

Sam

ple

Item

#/ I

tem

Su

bsca

le

M

SD

Skew

K

urto

sis

M

SD

Sk

ew

Kur

tosi

s 23

. Run

s/cl

imbs

exc

essi

vely

H

yper

activ

ity

DSM

-Hyp

erac

tive

.96

1.02

.6

7 -.8

0 1.

00

1.01

.6

2 -.7

8

24. S

care

d of

new

situ

atio

ns

Anx

ious

-Shy

.30

.70

2.52

5.

80

.28

.64

2.55

6.

41

25. P

icky

abo

ut c

lean

lines

s Pe

rfec

tioni

sm

1.

17

1.14

.3

3 -1

.39

1.07

1.

08

.49

-1.1

3

26. D

oes n

ot m

ake

frie

nds

Soci

al P

robl

ems

.1

1 .4

8 4.

99

25.1

7 .1

1 .4

7 4.

97

25.1

4

27. G

ets a

ches

/ pai

ns b

efor

e

scho

ol

Psyc

hoso

mat

ic

.20

.54

3.08

10

.01

.19

.53

3.23

11

.36

28. I

mpu

lsiv

e, e

asily

exc

ited

Hyp

erac

tivity

C

GI:

Res

tless

/ Im

puls

ive

.83

.91

.81

-.34

.80

.92

.95

-.07

29. N

o fo

llow

thro

ugh

on

inst

ruct

ions

C

ogni

tive

Prob

lem

s/

Inat

tent

ion

AD

HD

Inde

x D

SM-I

natte

ntiv

e

.60

.93

1.45

.9

7 .6

2 .9

3 1.

35

.67

30. H

as d

iffic

ulty

org

aniz

ing

D

SM-I

natte

ntiv

e

.55

.89

1.51

1.

17

.52

.88

1.63

1.

60

31. I

rrita

ble

Opp

ositi

onal

1.00

1.

04

.70

-.72

.92

1.01

.7

6 -.6

2

Page 107: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

97 Ta

ble

14 (c

ontin

ued)

EFA

Sam

ple

CFA

Sam

ple

Item

#/ I

tem

Su

bsca

le

M

S

D

Ske

w

Kur

tosi

s

M

SD

Sk

ew

Kur

tosi

s 32

. Res

tless

and

squi

rmy

Hyp

erac

tivity

.79

.96

.96

-.19

.76

.94

.99

-.10

33. S

care

d of

bei

ng a

lone

A

nxio

us-S

hy

.7

2 1.

04

1.16

-.0

8 .7

2 1.

05

1.18

-.0

4

34. A

lway

s doe

s thi

ngs s

ame

way

Pe

rfec

tioni

sm

.5

2 .8

8 1.

54

1.20

.4

8 .8

3 1.

65

1.68

35. D

oes n

ot g

et in

vite

d

Soci

al P

robl

ems

.2

1 .7

0 3.

25

9.21

.1

7 .6

4 3.

78

13.1

8

36. H

eada

ches

Ps

ycho

som

atic

.37

.61

1.77

3.

66

.38

.62

1.78

3.

58

37. F

ails

to fi

nish

thin

gs s

tarte

d C

GI:

Res

tless

/ Im

puls

ive

.94

1.06

.7

8 -.7

2 .8

6 .9

8 .8

6 -.4

0

38. I

natte

ntiv

e, d

istra

ctib

le

AD

HD

Inde

x C

GI:

Res

tless

/ Im

puls

ive

1.03

.9

8 .5

5 -.8

0 .9

6 .9

4 .6

3 -.5

8

39. T

alks

too

muc

h D

SM-H

yper

activ

e

.94

1.13

.7

6 -.9

2 .9

7 1.

13

.71

-.99

40. A

ctiv

ely

defie

s ad

ults

O

ppos

ition

al

.9

0 1.

04

.85

-.55

.89

1.01

.8

6 -.4

4

41. C

arel

ess m

ista

kes i

n w

ork

Cog

nitiv

e Pr

oble

ms/

In

atte

ntio

n D

SM-I

natte

ntiv

e

.59

.89

1.39

.8

6 .5

9 .8

8 1.

37

.90

42. D

iffic

ulty

aw

aitin

g tu

rn

DSM

-Hyp

erac

tive

.6

0 .8

1 1.

23

.73

.58

.84

1.35

.9

2

Page 108: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

98 Ta

ble

14 (c

ontin

ued)

EFA

Sam

ple

CFA

Sam

ple

Item

#/ I

tem

s Su

bsca

le

M

S

D

Ske

w

Kur

tosi

s

M

SD

Sk

ew

Kur

tosi

s 43

. Has

man

y fe

ars

Anx

ious

-Shy

.42

.78

1.94

3.

06

.36

.72

2.13

4.

04

44. H

as ri

tual

s

Perf

ectio

nism

1.06

.9

7 .5

4 -.7

3 1.

03

.94

.57

-.59

45. D

istra

ctib

ility

/ atte

ntio

n

prob

lem

AD

HD

Inde

x .9

2 .9

9 .7

5 -.5

6 .9

0 .9

6 .7

1 -.6

0

46. C

ompl

ains

/sic

k –

whe

n w

ell

Psyc

hoso

mat

ic

.2

3 .5

7 2.

77

7.94

.2

1 .5

2 2.

78

8.54

47. T

empe

r out

burs

ts

CG

I: Em

otio

nal

Labi

lity

1.10

.9

4 .6

1 -.4

7 1.

07

.92

.58

-.47

48. G

ets d

istra

cted

w/ i

nstru

ctio

ns

AD

HD

Inde

x

1.03

.9

1 .6

1 -.4

1 1.

00

.88

.58

-.38

49. I

nter

rupt

s/ in

trusi

ve

DSM

-Hyp

erac

tive

.9

1 1.

01

.76

-.65

.82

.93

.89

-.22

50. F

orge

tful i

n ev

eryd

ay

activ

ities

C

ogni

tive

Prob

lem

s/

Inat

tent

ion

DSM

-Ina

ttent

ive

.89

1.05

.8

6 -.5

7 .8

4 1.

01

.92

-.36

51. C

anno

t und

erst

and

mat

h C

ogni

tive

Prob

lem

s/

Inat

tent

ion

.97

1.24

.7

4 -1

.18

.99

1.25

.7

2 -1

.21

52. W

ill ru

n be

twee

n m

outh

fuls

H

yper

activ

ity

1.

01

1.10

.6

9 -.9

1 1.

07

1.13

.6

0 -1

.08

53. A

frai

d of

dar

k/ a

nim

als/

bug

s A

nxio

us-S

hy

1.

09

1.13

.5

7 -1

.10

1.07

1.

11

.57

-1.0

8

Page 109: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

99 Ta

ble

14 (c

ontin

ued)

EFA

Sam

ple

CFA

Sam

ple

Item

#/ I

tem

s Su

bsca

le

M

S

D

Ske

w

Kur

tosi

s

M

SD

Sk

ew

Kur

tosi

s 54

. Set

s hig

h go

als f

or s

elf

Perf

ectio

nism

.44

.76

1.87

3.

03

.51

.80

1.58

1.

79

55. F

idge

ts w

ith h

ands

/ fee

t A

DH

D In

dex

DSM

-Hyp

erac

tivity

1.05

1.

02

.50

-.99

1.11

1.

04

.42

-1.0

6

56. P

oor a

ttent

ion

span

A

DH

D In

dex

.8

6 .8

1 .6

5 -.1

6 .8

4 .8

2 .7

5 -.0

3

57. T

ouch

y/ e

asily

ann

oyed

O

ppos

ition

al

.7

7 .9

0 .1

0 .1

1 .8

2 .8

8 .8

8 .0

3

58. H

as m

essy

han

dwrit

ing

Cog

nitiv

e Pr

oble

ms/

In

atte

ntio

n

.60

1.02

1.

42

.53

.60

1.01

1.

42

.53

59. H

as d

iffic

ulty

pla

ying

qui

etly

H

yper

activ

ity

DSM

-Hyp

erac

tive

.40

.79

1.98

3.

02

.32

.69

2.21

4.

23

60. T

imid

, with

draw

n A

nxio

us-S

hy

.9

0 1.

00

.86

-.37

.94

1.02

.7

9 -.5

6

61. B

lam

es o

ther

s fo

r mis

take

s O

ppos

ition

al

.8

4 .9

7 .9

6 -.1

3 .8

7 .9

5 .8

6 -.2

1

62. F

idge

ts

CG

I: R

estle

ss/

Impu

lsiv

e

1.52

1.

08

-.08

-1.2

6 1.

51

1.04

-.0

7 -1

.17

63. M

essy

or d

isor

gani

zed

A

DH

D In

dex

.7

9 .9

6 .9

2 -.3

5 .6

9 .9

3 1.

13

.18

64. G

ets u

pset

if re

arra

nged

th

ings

Perf

ectio

nism

.9

9 .9

5 .7

3 -.3

7 .9

8 .9

5 .7

3 -.3

8

Page 110: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

100 Ta

ble

14 (c

ontin

ued)

EFA

Sam

ple

CFA

Sam

ple

Item

#/ I

tem

s Su

bsca

le

M

S

D

Sk

ew

Kur

tosi

s M

SD

Sk

ew

Kur

tosi

s 65

. Clin

gs to

par

ents

/ adu

lts

Anx

ious

-Shy

1.62

1.

10

-.14

-1.3

0 1.

65

1.11

-.1

3 -1

.35

66. D

istu

rbs p

eers

C

GI:

Res

tless

/ Im

puls

ive

.85

.92

.78

-.40

.86

.94

.76

-.51

67. D

elib

erat

ely

anno

ys o

ther

s O

ppos

ition

al

.6

4 .8

2 1.

13

.54

.65

.84

1.14

.4

9

68. D

eman

ding

eas

ily fr

ustra

ted

CG

I: R

estle

ss/

Impu

lsiv

e

1.11

.9

5 .5

0 -.6

7 1.

12

.98

.50

-.76

69. O

nly

atte

nds i

f int

eres

ted

A

DH

D In

dex

1.

37

.98

.23

-.94

1.38

.9

6 .2

3 -.8

9

70. S

pite

ful/

vind

ictiv

e O

ppos

ition

al

.6

8 .9

1 1.

21

.51

.68

.94

1.22

.3

7

71. L

oses

thin

gs

DSM

-Hyp

erac

tive

.7

8 .9

6 1.

08

.12

.75

.91

1.08

.2

5

72. F

eels

infe

rior

Soci

al P

robl

ems

.1

9 .5

0 3.

01

10.3

4 .2

1 .5

1 2.

80

8.44

73. S

eem

s tire

d/ sl

owed

dow

n

Psyc

hoso

mat

ic

.4

3 .7

5 1.

80

2.74

.4

6 .7

5 1.

70

2.41

74. P

oor s

pelli

ng

Cog

nitiv

e Pr

oble

ms/

In

atte

ntio

n

.66

1.11

1.

34

.12

.66

1.11

1.

34

.09

75. C

ries f

requ

ently

and

eas

ily

CG

I: Em

otio

nal

Labi

lity

.60

.84

1.28

.7

7 .6

3 .8

5 1.

31

.99

76. L

eave

s sea

t in

clas

sroo

m

AD

HD

Inde

x D

SM-H

yper

activ

e

.43

.71

1.67

2.

28

.45

.72

1.61

2.

04

Page 111: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

101 Ta

ble

14 (c

ontin

ued)

EFA

Sam

ple

CFA

Sam

ple

Item

#/ I

tem

s Su

bsca

le

M

S

D

Sk

ew

Kur

tosi

s M

SD

Sk

ew

Kur

tosi

s 77

. Moo

d ch

ange

s qui

ckly

CG

I: Em

otio

nal

Labi

lity

.80

.84

.81

-.06

.84

.82

.71

-.09

78. E

asily

frus

trate

d in

task

s

AD

HD

Inde

x .8

3 .9

2 .8

7 -.2

1 .7

9 .9

1 .9

4 -.0

7

79. E

asily

dis

tract

ed

D

SM-I

natte

ntiv

e 1.

32

.94

.21

-.84

1.34

.9

8 .2

1 -.9

7

80. B

lurts

out

ans

wer

s H

yper

activ

ity

DSM

-Hyp

erac

tivity

.56

.84

1.47

1.

36

.57

.82

1.43

1.

34

N

ote.

EFA

n =

555

; C

FA n

= 1

,253

; A

DH

D =

Atte

ntio

n D

efic

it H

yper

activ

ity D

isor

der;

DSM

= D

iagn

ostic

and

Sta

tistic

al M

anua

l; C

GI=

Con

ners

’ Glo

bal I

ndex

.

Page 112: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

102

was statistically significant (χ2 = 13,232.15, df = 3,160, p < .001), which is typically reflective of

large sample size rather than a lack of factorability of the correlation matrix (Leech et al., 2005).

Promax (k = 4; Gorsuch, 1997) and Oblimin rotations were run and compared. Findings from

analyses with Promax rotation are reported, as no meaningful structural differences were evident

when different rotational methods were used.

Factor retention findings varied. MAP (Velicer, 1967) indicated that seven factors should

be retained, visual inspection of the scree plot suggested one to four factors, and PA (Horn,

1965) indicated eight factors. Research using the normative sample supported a seven-factor

structure (Conners, 1997). Because the suggested number of factors ranged from one to eight,

eight factor structures were examined starting with the eight-factor solution and ending with the

one-factor solution. All factor solutions, except two, failed to meet the a priori criteria for factor

adequacy (e.g., four or more items with salient pattern coefficients). The one- and two-factor

solutions met criteria for factor adequacy. The two-factor solution was retained because research

suggests that over-factoring is preferable to under-factoring (Fabrigar et al., 1999) and because

this solution demonstrated greater theoretical convergence, higher communalities, pattern

coefficients, and structure coefficients, and a larger amount of variance accounted for (18.71%)

than the one-factor solution.

The two-factor solution contained 36 items, with half (18) salient on each factor. For

Factor I, pattern coefficients ranged from .41 to .67 (Mdn = .55), and communalities ranged from

.11 to 38 (Mdn = .30). Examples of Factor I items included, “Difficulty working on/ finishing

homework” and “Inattentive/ distractible.” Factor I was named Inattention, as the items

appeared to reflect the descriptor. The Inattention factor accounted for 15.45% of the total

variance, and the reliability estimate (Cronbach’s α) of the scores was .83. On Factor II, pattern

Page 113: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

103

coefficients ranged from .40 to .62 (Mdn = .45), and communality estimates ranged from .13 to

.40 (Mdn = .29). Examples of items on Factor II included, “Irritable,” “Loses temper,” and

“Runs about and climbs excessively and inappropriately.” Item content on Factor II seemed to

reflect an oppositional/ hyperactivity dimension, which was labeled as such. The Oppositional/

Hyperactivity factor accounted for 3.26% of the total variance, and the Cronbach’s alpha for

scores was .88. A summary of the findings from the two-factor solution is presented in Table 15.

Secondary EFA solutions. As indicated earlier, an alternative models approach was

used to obtain maximum structure viability (MacCallum et al., 1993). Three different sets of

EFA solution criteria were used to retain items and to identify the best possible models of the

CPRS-R that could be tested through CFA: (a) the moderate criteria al extended solution (to be

referred to as extended-2P), (b) the moderate criteria rational abbreviated solution (abbreviated-

2P), and (c) the moderate criteria statistical/rational hybrid solution (hybrid-2P). Unlike the

teacher scale, a strict criteria hybrid solution for the CPRS-R could not be derived because an

insufficient number of items had pattern coefficients and communalities that met the criteria

(pattern coefficients ≥ .55 and communalities ≥ .35). Thus, a moderate criteria hybrid solution

(hybrid-2P) was used instead. However, like the teacher scale, the initial EFA two-factor

minimum criteria solution (pattern coefficient ≥ .40 and absence of cross-loadings ≥ .20) was

determined to be inadequate (e.g., low pattern coefficients, low communalities, and minimal

variance accounted for) relative to the other solutions. Therefore, this solution was not selected

as a model to test via CFA.

EFAs of each identified solution (extended-2P, abbreviated-2P, and hybrid-2P) were

conducted in two steps to select the best items for CFA models: (a) single factor analyses of

Page 114: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

104 Ta

ble

15

Su

mm

ary

of th

e In

itial

Tw

o-F

acto

r So

lutio

n of

CP

RS-

R S

core

s: P

atte

rn C

oeffi

cien

ts (P

AF

ext

ract

ion

with

Pro

max

Rot

atio

n)

and

Alte

rnat

ive

Solu

tions

Ite

m #

/ Ite

m

Fact

or

h2

Alte

rnat

ive

Solu

tions

Inat

tent

ion

Opp

ositi

onal

/H

yper

activ

ity

Ext

ende

d-2P

A

bbre

viat

ed-2

P H

ybrid

-2P

41. C

arel

ess m

ista

kes i

n w

ork

.67

-.13

.35

X

X

X

2. D

iffic

ulty

fini

shin

g ho

mew

ork

.66

-.10

.36

X

X

X

9. A

void

s sus

tain

ed m

enta

l

e

ffor

t .66

-.12

.34

X

X

X

37. F

ails

to fi

nish

thin

gs s

tarte

d .65

-.05

.38

X

X

X

29. D

oes n

ot fo

llow

inst

ruct

ions

.65

-.15

.32

X

X

X

38. I

natte

ntiv

e/ d

istra

ctib

le

.64

-.06

.36

X

X

X

58. H

as m

essy

han

dwrit

ing

.61

-.24

.25

12

. Fai

ls to

com

plet

e

ass

ignm

ents

.57

.01

.33

X

X

51. C

anno

t und

erst

and

mat

h .55

-.11

.24

45

. Dis

tract

ibili

ty p

robl

em

.54

.01

.30

X

X

50

. For

getfu

l in

ever

yday

ac

tiviti

es

.52

-.04

.24

X

X

20. D

oes n

ot se

em to

list

en

.51

.01

.26

22

. Nee

ds c

lose

supe

rvis

ion

to

wor

k .49

-.10

.19

X

X

48. G

ets d

istra

cted

with

in

stru

ctio

ns

.49

.06

.28

74. P

oor s

pelli

ng

.48

-.07

.19

Page 115: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

105 Ta

ble

15 (c

ontin

ued)

Fa

ctor

h

2 A

ltern

ativ

e So

lutio

ns

Ite

m #

/ Ite

m

Inat

tent

ion

Opp

ositi

onal

/ H

yper

activ

ity

Ext

ende

d-2P

Abb

revi

ated

-2P

H

ybrid

-2P

78. E

asily

frus

trate

d in

task

s .48

.13

.32

19

. Tro

uble

con

cent

ratin

g in

c

lass

.47

-.13

.16

15. C

heck

s thi

ngs r

epea

tedl

y -.41

.17

.11

54

. Set

s hig

h go

als f

or s

elf

-.37

.20

.09

30

. Has

diff

icul

ty o

rgan

izin

g

.37

.24

.30

76

. Lea

ves s

eat

.3

3 .2

0 .2

4

63. M

essy

or d

isor

gani

zed

.31

.29

.30

79

. Eas

ily d

istra

cted

.2

8 .0

6 .1

0

25. P

icky

abo

ut c

lean

lines

s -.2

7 .0

1 .0

7

71. L

oses

thin

gs

.25

.07

.09

44

. Has

ritu

als

.2

5 .0

9 .1

0

10. D

iffic

ulty

sust

aini

ng

atte

ntio

n

.23

.20

.15

69. O

nly

atte

nds i

f int

eres

ted

.1

1 .0

9 .0

3

26. D

oes n

ot m

ake

frie

nds

.11

.04

.02

31

. Irr

itabl

e .0

1 .62

.40

X

X

X

21. L

oses

tem

per

-.02

.58

.32

X

X

X

47. T

empe

r out

burs

ts

.11

.56

.40

X

X

X

66. D

istu

rbs p

eers

.0

9 .54

.36

49

. Int

erru

pts/

intru

sive

.1

7 .51

.39

X

X

X

57. T

ouch

y/ e

asily

ann

oyed

-.0

3 .51

.24

39

. Tal

ks e

xces

sive

ly

-.14

.51

.19

X

Page 116: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

106 Ta

ble

15 (c

ontin

ued)

Fa

ctor

h

2 A

ltern

ativ

e So

lutio

ns

Item

#/ I

tem

In

atte

ntio

n O

ppos

ition

al/

Hyp

erac

tivity

Ex

tend

ed-2

P

Abb

revi

ated

-2P

H

ybrid

-2P

43. H

as m

any

fear

s -.2

9 .4

9 .1

5

64. G

ets u

pset

if th

ings

re

arra

nged

-.0

7 .48

.20

8. F

ight

s .1

3 .46

.30

X

X

70

. Spi

tefu

l/ vi

ndic

tive

.14

.44

.29

X

X

67

. Del

iber

atel

y an

noys

oth

ers

.22

.43

.35

X

X

23

. Run

s/cl

imbs

exc

essi

vely

.1

0 .43

.25

X

1. A

ngry

/ res

entfu

l .0

7 .43

.22

68

. Dem

andi

ng e

asily

f

rust

rate

d .0

8 .42

.22

X

X

75. C

ries f

requ

ently

and

eas

ily

-

.09

.41

.13

62

. Fid

gets

.1

8 .40

.29

X

24. S

care

d of

new

situ

atio

ns

-

.21

.40

.10

32

. Res

tless

and

squi

rmy

.17

.40

.28

X

11. A

rgue

s

.13

.40

.24

X

X

61

. Bla

mes

oth

ers

for m

ista

kes

.15

.39

.25

40

. Act

ivel

y de

fies

adul

ts

.25

.38

.33

33

. Sca

red

of b

eing

alo

ne

-

.18

.37

.08

55

. Fid

gets

with

han

ds o

r fee

t .0

6 .3

6 .1

6

28. I

mpu

lsiv

e, e

asily

exc

ited

-

.03

.36

.12

17

. Ach

es/ p

ains

-.0

9 .3

5 .0

9

65. C

lings

to p

aren

ts/ a

dults

-.1

6 .3

3 .0

7

Page 117: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

107 Ta

ble

15 (c

ontin

ued)

Fa

ctor

h

2 A

ltern

ativ

e So

lutio

ns

Ite

m #

/Item

In

atte

ntio

n O

ppos

ition

al/

Hyp

erac

tivity

Ex

tend

ed-2

P

Abb

revi

ated

-2P

H

ybrid

-2P

36. H

eada

ches

-.1

2 .3

3 .0

8

46. C

ompl

ains

abo

ut b

eing

sick

.0

3 .3

3 .1

2

42. D

iffic

ulty

aw

aitin

g tu

rn

.25

.32

.26

16

. Los

es fr

iend

s qui

ckly

.2

3 .3

1 .2

4

4. T

imid

, eas

ily sc

ared

-.2

9 .2

9 .0

6

53. A

frai

d of

dar

k/ a

nim

als

/bug

s

-.2

0 .2

8 .0

5

14. S

care

d of

peo

ple

-

.25

.28

.05

73

. See

ms t

ired/

slow

ed d

own

-.0

9 .2

8 .0

5

52. W

ill ru

n be

twee

n

mou

thfu

ls

.24

.27

.21

13. H

ard

to c

ontro

l in

stor

es

.17

.27

.15

7.

Sto

mac

h pa

in

.

09

.26

.04

77

. Moo

d ch

ange

s qui

ckly

.16

.2

5 .1

4

3. O

n th

e go

/driv

en b

y

m

otor

.01

-.2

5 .0

6

18. R

estle

ss/ h

ighl

y ac

tive

.

05

.23

.07

72

. Fee

ls in

ferio

r

-.01

.2

3 .0

5

59. H

as d

iffic

ulty

pla

ying

q

uiet

ly

.

22

.23

.16

56. P

oor a

ttent

ion

span

.03

.2

1 .0

5

27. G

ets a

ches

and

pai

ns

.

14

.21

.10

80

. Blu

rts o

ut a

nsw

ers

.

10

.20

.07

Page 118: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

108 Ta

ble

15 (c

ontin

ued)

Fa

ctor

h

2 A

ltern

ativ

e So

lutio

ns

Ite

m #

/Item

In

atte

ntio

n O

ppos

ition

al/

Hyp

erac

tivity

Ex

tend

ed-2

P

Abb

revi

ated

-2P

H

ybrid

-2P

60. T

imid

, with

draw

n

-.18

.1

9 .0

3

5. E

very

thin

g m

ust p

erfe

ct

.

06

.17

.04

34

. Alw

ays d

oes t

hing

s sam

e

way

.01

.0

9 .0

1

6. D

oesn

’t ha

ve fr

iend

s

-.02

-.0

8 .0

1

35. D

oes n

ot g

et in

vite

d

-.

02

-.08

.01

Ei

genv

alue

s 1

0.46

5.

82

% o

f var

ianc

e 1

7.73

%

9.87

%

Cro

nbac

h’s ∝

.9

1 .8

8

Not

e. n

= 5

55; h

2 =

com

mun

ality

; sal

ient

pat

tern

coe

ffic

ient

s are

indi

cate

d in

bol

d.

Page 119: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

109

items on each factor and (b) a subsequent factor analysis of all remaining items on both factors.

A breakdown of which parent items aligned with each solution is also presented in Table 14 with

the original two-factor solution.

Extended-2P solution. CPRS-R items on the extended-2P solution were required to have

met minimum saliency criteria in the initial EFA, be adequately translated, and meet additional

rational criteria. Items retained on Factor I were associated with oppositional behavior or

hyperactivity. Items such as “Loses temper” and “Restless and squirmy” were included on

Factor I, whereas items such as “Cries frequently and easily” were not. Items retained on Factor

II were strictly germane to the construct of inattention. Items such as “Poor spelling” and

“Cannot grasp math” were not included on Factor II, as these items appeared to reflect an

academic problems construct that was not directly related to inattention. Of the 36 items that

met saliency criteria in the initial EFA, 24 items met the criteria for this initial solution (13 items

on Factor I and 11 items on Factor II).

EFA of each factor. Separate factor analyses of each factor supported a one-factor

solution. All 13 items met saliency criteria on the Factor I, and all 11 items on Factor II met

saliency criteria. No items were removed from the analyses.

Full EFA. EFA of the items on the extended-2P solution indicated a two-factor solution

was viable (MAP, PA, and scree). Thirteen items remained salient on Factor I. Factor I was

named Oppositional/Hyperactivity, based on the content of the items (e.g., “Irritable,”

“Spiteful/vindictive,” and “Restless and squirmy”). Ten of the 11 items remained salient on

Factor II. The item “Gets distracted when given instructions” did not meet saliency criteria on

either factor and was removed from the solution. EFA was re-run on the remaining 23 items.

Factor II was named Inattention and contained items (described previously) that were reflective

Page 120: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

110

of the name. Communalities, eigenvalues, percent of variance, and pattern coefficients are

presented in Table 16.

Table 16 Pattern Coefficients from EFA of the CPRS-R, Extended-2P Solution (PAF Extraction, Promax Rotation) Items

Factor I Oppositional/ Hyperactivity

Factor II Inattention h2

31. Irritable .67 -.05 .42 21.Loses temper .64 -.07 .36 47. Temper outbursts .63 .03 .42 49. Interrupts/ Intrusive .62 .03 .41 70. Spiteful/ vindictive .56 -.01 .31 23. Runs/ climbs excessively .55 -.03 .28 67. Deliberately annoys others .55 .08 .36 8. Fights .54 .05 .33

62. Fidgets .52 .05 .30 34. Talks too much .51 -.15 .19 32. Restless and squirmy .49 .07 .28 11. Argues .46 .08 .25 68. Demanding, easily frustrated .41 .05 .20 9. Avoids sustained mental effort -.09 .68 .40 2. Difficulty finishing homework -.06 .65 .39

12. Fails to finish things started .03 .65 .44 41. Careless mistakes in work -.03 .63 .37 29. Does not follow through on tasks -.07 .61 .33 38. Inattentive/ distractible -.06 .59 .39 12. Fails to finish assignments .07 .56 .36 45. Distractibility or attention

problem .08 .53 .34

22. Needs close supervision -.01 .44 .19 50. Forgetful in everyday activities -.09 .43 .23 Eigenvalue 6.62 2.26 Percent of variance 25.93% 7.00% Cronbach’s α .85 .83 Note. n = 555; h2 = communality; salient pattern coefficients are indicated in bold.

Page 121: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

111

Abbreviated-2P solution. CPRS-R items on the abbreviated-2P solution were required to

have met minimum saliency criteria in the initial EFA, be adequately translated, and meet

additional rational criteria. Items retained on Factor I were strictly germane to inattention and

were identical to the items included on the Inattention factor of the extended-2P solution. Items

retained on Factor II were strictly germane to the construct of oppositional behavior (not

hyperactivity). Items that appeared to be reflective of oppositional behavior, such as “Loses

temper” were included, and items that were reflective of hyperactivity or other constructs, such

as “Restless and squirmy” and “Cries frequently and easily” were excluded from the factor. Of

the 36 items that met saliency criteria in the initial EFA, a total of 20 items met the additional

criteria to be retained in this solution (11 items on Factor I, and nine items on Factor II).

EFA of each factor. EFA of items on Factor I supported a one-factor solution containing

all items. The same pattern was observed for all items on Factor II (nine). Thus, no items were

dropped from analyses.

Full EFA. The 20 items that made up the abbreviated-2P solution were submitted for

EFA. All indicators (MAP, PA, and scree) suggested a two-factor solution. All items except one

(“Gets distracted when given instructions,”) met saliency criteria on the expected factors,

resulting in its removal. The remaining 19 items were re-submitted for EFA, and all items met

saliency criteria on the expected factors. Nine items, reflective of oppositional behaviors, were

salient on Factor I. Ten items were salient on Factor II and reflected inattention.

Communalities, eigenvalues, percent of variance, and pattern coefficients are presented in Table

17.

Hybrid-2P solution. CPRS-R (Conners, 1997) items on the hybrid-2P solution were

required to meet saliency criteria above the minimum ones in the initial EFA, be adequately

Page 122: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

112

translated, and meet additional statistical and rational criteria. Thus, items had to have both

pattern coefficients .50 or higher and communalities .30 or higher. Items selected for

consideration on Factor I items were strictly related to inattention. Items selected for Factor II

Table 17 Pattern Coefficients from EFA of the CPRS-R Abbreviated-2P Solution (PAF Extraction, Promax Rotation)

Items Factor I Oppositional

Factor II Inattention

h2

47. Temper outbursts .68 -.01 .45 21. Loses temper .66 -.05 .40 70. Spiteful/ vindictive .62 -.07 .34 31. Irritable .57 .05 .36 8. Fights .57 .01 .33

67. Deliberately annoys others .55 .04 .33 49. Interrupts/ intrusive .55 .05 .33 68. Demanding, easily frustrated .49 -.01 .24 11. Argues .48 .03 .25 9. Avoids sustained mental effort -.13 .71 .42 2. Difficulty finishing homework -.10 .63 .34

41. Careless mistakes in work -.40 .60 .34 38. Inattentive/ distractible .09 .57 .39 29. Does not follow through on instructions -.01 .55 .30 37. Fails to finish things started .13 .53 .38 45. Distractibility or attention problem .08 .52 .33 12. Fails to finish assignments .12 .51 .34 22. Needs close supervision -.03 .41 .16 50. Forgetful in everyday activities .11 .41 .23 Eigenvalue 4.95 1.29 Percent of variance 26.03 6.79 Cronbach’s α .82 .82 Note. n = 555; h2 = communality; salient pattern coefficients are indicated in bold.

were required to be strictly germane to the construct of oppositional behavior (not hyperactivity).

Of the 36 items that met saliency criteria in the initial EFA, 12 items were selected for the

hybrid-2P solution (eight items on Factor I and four items on Factor II).

Page 123: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

113

EFA of each factor. EFA of items for each hybrid-2P factor supported a one-factor

structure: six items on Factor I and four items on Factor II. All items on each factor met saliency

criteria. Thus, no items were dropped from analyses.

Full EFA. The retention criteria for an EFA of all items on the hybrid-2P solution

supported a two-factor structure for the CPRS-R items. All items, except one (“interrupts/

intrudes”), met saliency criteria on the expected factors. The item was deleted, the factor

structure of the remaining 11 items was examined. Eight items reflective of inattention were

salient on Factor I, called Inattention. Three items were salient on Factor II, which was named

Temperamental, based on the content of the items. Communalities, eigenvalues, percent of

variance, and pattern coefficients are presented in Table 18.

Table 18 Pattern Coefficients from EFA of the CPRS-R Hybrid-2P Solution (PAF Extraction, Promax Rotation) Items

Factor I Inattention

Factor II Temperamental

h2

9. Avoids sustained mental effort .72 -.11 .44 2. Difficulty finishing homework .66 -.13 .37

41.Careless mistakes in work .57 -.01 .32 29. No follow through on instructions .56 .01 .31

38. Inattentive/ distractible .54 .12 .37 12. Fails to finish assignments .54 .08 .34 37. Fails to finish things he started .53 .12 .36 45. Distractibility or attention problem .48 .14 .32 21. Loses temper -.07 .76 .52 47. Temper outbursts -.01 .74 .54 31. Irritable .08 .56 .37

Eigenvalue 3.98 1.44 Percent of variance 30.61 10.24 Cronbach’s α .81 .73 Note. n = 555; h2 = communality; salient pattern coefficients are indicated in bold.

Page 124: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

114

EFA summary. Three 2-factor solutions were identified through EFA: an extended-2P,

an abbreviated-2P, and a hybrid-2P solution. The solutions varied in the item retention criteria

used and in the number of items retained. However, the extended-2P solution had an

Oppositional/ Hyperactivity factor, the abbreviated 2-P solution reflected a strictly Oppositional

factor, and the hybrid-2P solution had a Temperamental factor. All solutions reflected an

Inattention factor. In comparison to the other two solutions, the hybrid-2P solution contained the

fewest number of items, had the highest pattern/structure coefficients and communalities, and

accounted for the greatest amount of variance. All three solutions were determined to be

appropriate for further analysis in CFA.

CFA

Preliminary analyses. The CFA sample contained 1,253 cases. Assumptions of CFA

(linearity, multivariate normality, and specification of a correct model; Kline, 2006; Tomarken &

Waller, 2005) for the CPRS-R items were tested. Normality of the data was acceptable for most

CPRS-R items, except for six items, which were severely skewed (> 3) or kurtotic (> 8).

Descriptive statistics (means, standard deviations, skew, and kurtosis) of CPRS-R items are also

presented in Table 11. Findings of multivariate normality, based on the Mardia’s coefficient and

normalized estimates, are reported by model. Linearity, examined through visual inspection of

scatterplots, was met. Although specification of a correct model cannot be conclusively

determined, EFAs and an alternative models CFA approach were used to enhance the likelihood

of selection of the correct model (MacCallum et al., 1993). A correlation matrix for the CFA

sample of CPRS-R items is provided in Appendix C.

Models. CFAs (maximum likelihood extraction, robust) of the CPRS-R scores were

conducted on the conversion of raw scores to covariance matrices for five non-nested models: (a)

Page 125: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

115

3 two-factor models–extended-2P2, abbreviated-2P, and hybrid-2P models, (b) Conners’ seven-

factor empirical model, and (c) Conners’ twelve-factor full model. All CPRS-R models tested in

this study and the criteria through which they were derived are summarized in Table 19.

Table 19 Summary of CPRS-R Models and Tested through CFA Models # of Items Item Selection Criteria

Two-Factor Models Extended-2P 23 Items selected based on minimum statistical and

additional rational criteria; Factor I = Inattention, Factor II = Oppositional/Hyperactivity

Abbreviated-2P 19 Items selected based on minimum statistical and additional rational criteria; Factor I = Inattention, Factor II = Oppositional

Hybrid-2P 11 Items selected based on above moderate statistical and additional rational criteria

Conners’ (1997) Models Conners-E7-P 59 All empirically-derived CPRS-R items

Conners-F12-P 80 All CPRS-R items from all scales

The EFA-derived models are depicted in Figure 4, and Conners’ models are depicted at the

factor level in Figures 5 (Conners-E7-P) and 6 (Conners-F12-P).

All models met the identification criteria proposed by Kline (2005); a delineation of the

numbers of free parameters and observations in each CPRS-R model is provided in Appendix D.

To fix the scale of the models, a single item on each factor was fixed to one. On the Conners- 2 For ease of presentation the model names were shortened: (a) two-factor moderate rational extended to extended-

2P, (b) two-factor moderate rational abbreviated to abbreviated-2P, (c) two-factor moderate hybrid to hybrid-2P,

(d) seven-factor Conners’ empirical to Conners-E7-P, and (e) twelve-factor Conners’ full model to Conners-F12-P.

Page 126: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

Spit

eful

or

vind

icti

veE

A

E EE EE EE E

AA AAA

E

H HHA

EE

E EA

A E

AA

AH H

Inat

tent

ion

Hyp

erac

tivi

ty/

Opp

osit

iona

l

Lose

s te

mpe

r

Tem

per

outb

urst

s

Inte

rrup

ts/

intr

udes

Talk

s ex

cess

ivel

y

Figh

ts

Nee

ds cl

ose

supe

rvis

ion

Care

less

mis

take

s

Diff

icul

ty

hom

ewor

k

Avo

ids

men

tal e

ffor

t

Fails

to fi

nish

Doe

s not

follo

w

inst

ruct

ions

Inat

tent

ive,

ea

sily

dis

trac

ted

Can’

t com

plet

e w

ork

Dis

trac

tibi

lity

prob

lem

Forg

etfu

l

AH

AH

A

Ann

oys

othe

rs

Runs

/ cl

imbs

Dem

ands

fr

ustr

ated

Fidg

etin

gRe

stle

ss /

squi

rmy

Arg

ues

wit

h ad

ults

EH

EA E

EA

EE

EA

Irri

tabl

e

H

HA

E

Fig

ure

4. C

onne

rs’ p

aren

t mod

els

deri

ved

thro

ugh

EF

As.

E =

exte

nded

-2P,

A =

abb

revi

ated

-2P,

H =

hyb

rid-2

P m

odel

.

116

Page 127: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

Opp

osit

iona

l

Cog

niti

ve

Prob

lem

s/

Inat

tent

ion

Hyp

erac

tivi

ty

Soci

al P

robl

ems

Anx

ious

-Shy

Perf

ecti

onis

m

6 it

ems

8 it

ems

7 it

ems

5 it

ems

6 it

ems

6 it

ems

5 it

ems

Psyc

hom

atic

Fig

ure

5. C

onne

rs-E

7-P

mod

el d

epic

ted

at th

e fa

ctor

leve

l.

117

Page 128: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

Opp

osit

iona

l

CP/

I

Hyp

erac

tivi

ty

SP Anx

ious

-Shy

Perf

ecti

onis

m

6 it

ems*

8 it

ems*

7 it

ems*

5 it

ems

6 it

ems

6 it

ems

AD

HD

Inde

x

CGI:

RI

CGI:

EL

DSM

: IA

DSM

: HI

6it

ems

4 it

ems*

9it

ems*

9it

ems*

12 it

ems

5 it

ems

Psyc

hoso

mat

ic

Fig

ure

6. C

onne

rs-F

12-P

mod

el d

epic

ted

at th

e fa

ctor

leve

l. C

P/I =

Cog

nitiv

e Pr

oble

ms/

Inat

tent

ion;

SP

= So

cial

Pro

blem

s; C

GI:

RI =

Con

ners

’ Glo

bal I

ndex

: Res

tless

-Im

puls

ive;

CG

I: EL

= C

onne

rs’ G

loba

l Ind

ex: E

mot

iona

l Lab

ility

; DSM

: IA

= D

SM-I

V In

atte

ntiv

e;

DSM

: HI =

DSM

-IV

Hyp

erac

tive/

Impu

lsiv

e. *

indi

cate

s tha

t one

or m

ore

item

s are

incl

uded

on

mul

tiple

fact

ors.

118

Page 129: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

119

F12-P model, many items loaded on more than one factor. Start values for items that loaded on

more than one factor were adjusted based on theory and prior research. Each of the models was

compared to its respective null and one-factor models. Then, the five non-nested models were

compared to one another to identify the model with the best fit to the data. Because an

alternative models approach was employed in this study, no post hoc model re-specifications

were conducted. However, model re-specification was used when a model would not run as

initially established (redundancy of factors, Heywood case removal, etc.). Unless indicated

otherwise, all the chi-square tests were statistically significant, which possibly was due to sample

size (Kline, 199). As a result, this statistics was not given weight in evaluating goodness of fit of

model to data.

Extended model. Mardia coefficients indicated multivariate nonnormality in the CPRS-

R data; thus, maximum likelihood extraction, robust method, was used. Four outlying cases

were identified based on Mardia’s normalized estimates. Removing these cases and re-running

the analyses without them did not substantially change the model fit to the data. Therefore, the

original analysis was used with the four cases included.

The null model was not supported, as the chi-square was statistically significantly lower

for the one- and two-factor models than the null. Also, the fit indices for the one-factor model

indicated a poor fit (e.g., RMSEA = .07; CFI = .77), whereas the fit indices supported the two-

factor model (the extended-2P; e.g., RMSEA = 04; CFI = .91). All unstandardized parameter

estimates were statistically significant at the .01 level and standardized coefficients for the

extended-2P model of the Conners’ Parent Scale are reported in Table 20. The correlation

between the two factors was moderate and statistically significant (R = .63; p < .001).

Page 130: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

120

Abbreviated model. CFA with maximum likelihood extraction, robust, was used because

Mardia coefficients indicated multivariate nonnormality in the CPRS-R data. Two cases were

identified as outliers based on Mardia’s normalized estimates. No substantial changes in the

model fit to the data were evident when the two cases were removed from the analysis; therefore,

the original analysis with the two cases was used.

The null model was not supported, as the chi-square was statistically significantly lower

for the one- and two-factor models than the null. Also, the findings from fit statistics did not

Table 20 Standardized Pattern Coefficients from the CFA (Maximum Likelihood, Robust) of the Extended-2P Model for the Conners’ Parent Scale (n = 1,253)

Items Factor PC SE R2

31. Irritable 1 .59 .81 .35 21. Loses temper 1 .63 .78 .40 47. Temper outbursts 1 .67 .74 .45 49. Interrupts/ intrusive 1 .56 .83 .31 70. Spiteful/ vindictive 1 .55 .84 .30 23. Runs/ climbs excessively 1 .49 .87 .24 67. Deliberately annoys others 1 .57 .82 .33 8. Fights 1 .56 .83 .31

62. Fidgets 1 .52 .85 .27 39. Talks too much 1 .39 .92 .16 32. Restless and squirmy 1 .59 .81 .35 11. Argues with adults 1 .49 .87 .24 68. Demanding, easily frustrated 1 .50 .87 .25 9. Avoids sustained mental effort 2 .54 .84 .29 2. Difficulty finishing homework 2 .51 .86 .26

27. Fails to finish things he started 2 .61 .79 .38 37. Careless mistakes in work 2 .54 .84 .29 29. Does not follow through on

instructions 2 .52 .85 .28

38. Inattentive/ disractible 2 .65 .76 .42 12. Fails to complete assignments 2 .59 .81 .34 45. Distractibility/attention problem 2 .59 .81 .35 22. Needs close supervision to work 2 .38 .93 .14 50. Forgetful in everyday activities 2 .53 .85 .28

Note. PC = parameter estimate; SE = standard error. R2 = standardized factor loading squared

Page 131: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

121

support the one-factor model (e.g., CFI = .77; RMSEA = .08). However, the fit indices provided

support for the two-factor (abbreviated-2P) model (e.g., CFI = .90; RMSEA = .05). All

unstandardized parameter estimates were statistically significant (α = .01) for the abbreviated-2P

model. Standardized coefficients for this model of the parent scale are reported in Table 21. The

correlation between the factors was statistically significant (R = .61, p < .001).

Table 21

Standardized Pattern Coefficients from Maximum Likelihood (Robust) CFA of the Abbreviated-2P Model for the Conners’ Parent Scale (n = 1,253)

Items Factor PC SE R2

47. Temper outbursts 1 .69 .73 .47 71. Loses temper 1 .64 .76 .42 70. Spiteful/ vindictive 1 .55 .83 .31 31. Irritable 1 .60 .80 .36 8. Fights 1 .57 .82 .32

67. Deliberately annoys others 1 .56 .83 .31 49. Interrupts/ intrusive 1 .55 .84 .30 68. Demanding, easily frustrated 1 .51 .86 .36 11. Argues 1 .49 .87 .24 9. Avoids sustained mental effort 2 .60 .80 .36 2. Difficulty finishing homework 2 .52 .86 .27

41. Careless mistakes in work 2 .54 .84 .29 38. Inattentive/ distractible 2 .65 .76 .42 29. Does not follow through on instructions

2 .53 .85 .28

37. Fails to finish things started 2 .61 .79 .37 45. Distractibility or attention problem 2 .58 .81 .34 12. Fails to finish assignments 2 .59 .81 .34 22. Needs close supervision to work 2 .37 .93 .14 50. Forgetful in everyday activities 2 .48 .88 .23 Note. PC = parameter estimate; SE = standard error. R2 = standardized factor loading squared

Page 132: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

122

Parent hybrid model. CFA with maximum likelihood extraction, robust, was used

because Mardia coefficients indicated multivariate nonnormality in the CPRS-R data. One

outlying case was identified based on the Mardia’s normalized estimate, the outlier was

removed, and the analyses were run without it. No meaningful changes in model fit were evident

upon removal of the outlying case; thus the original analysis was used.

The null model was not supported, as the chi-square was statistically significantly lower

for the one- factor and two-factor models than the null. Also, findings did not support the one-

factor model (CFI = .80; RMSEA = .10) or the two-factor (hybrid-2P) model (CFI = .85;

RMSEA = .09). Parameter estimates (i.e., standardized coefficients and standard error) for this

model are not reported as they could be misleading.

Parent empirical model. Seven factors (59-items) made up the parent empirical model

(Conners-E7-P): Oppositional, Hyperactivity, Inattention/Cognitive problems, Social Problems,

Anxious-Shy, Psychosomatic, and Perfectionism. CFA with maximum likelihood extraction,

robust, was used because Mardia coefficients indicated multivariate nonnormality in the CPRS-R

data. Thirteen cases were identified as outliers based on Mardia’s normalized estimates. No

substantial changes in the model fit to the data were evident when the 13 cases were removed

and the analyses were re-run without them; therefore, the original analysis with the thirteen cases

was used.

The null model was not supported, as the chi-square was statistically significantly lower

for the one-factor and Conners-E7-P models than the null. Also, findings for the Conners-E7-P

model did not support the one-factor model (CFI = .59; RMSEA = .05) or the seven-factor model

(CFI = .75; RMSEA = .04). Thus, the parameter estimates were not interpreted for any of the

models. For comparative purposes, the models were examined a second time after the removal

Page 133: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

123

of six poorly translated items from the sample, and no meaningful changes in model fit were

evident after removal of the items (e.g., seven-factor model CFI = .61; RMSEA = .05).

Parent full model. Conners’ full model (Conners-F12-P) was based on twelve factors

(80 items): Oppositional, Hyperactivity, Inattention/Cognitive problems, Social Problems,

Anxious-Shy, Psychosomatic, Perfectionism, ADHD Index, CGI- Restless/ Impusive, CGI-

Emotional Lability, DSM-Inattentive, and DSM-Hyperactive. CFA with maximum likelihood

extraction, robust, was used because Mardia coefficients indicated multivariate nonnormality in

the CPRS-R data. An initial CFA of the full model would not run, as the covariance matrix of

the data was identified as not positive definite (zero or negative values; Wothke, 1993). After

excluding errors in the model language or data, possible reasons for a non-positive definite

matrix of the full model were (a) poor start values, (b) linear dependency of between variables

and factors, and (c) limited variability of the data. Estimating start values based on prior

research did not result in the model running. Scale and item correlations computed outside of

CFA indicated multicollinearity of some factors (R ≥ .90). Inter-item correlations were not

indicative of inter-item multicollinearity (across or within factors), as all correlations were below

.60. Also, one item, “Doesn’t have friends,” was identified as a zero value and had a mean of .01

and a standard deviation of .11, which indicates that the item was not effective in this study and

could not load on any factor.

To determine to what extent scale revision might be needed to get a model to run

containing as many of the 80 items on the full Conners’ Parent Scale as possible, the following

was conducted: (a) removing 2 items identified as Heywood cases (constraining the values to

zero did not work), (b) removing fifteen cases identified as multivariate outliers based on

Mardia’s normalized estimate, (c) removing 11 poorly translated items, (d) removing 10 items

Page 134: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

124

with severe kurtosis or extremely low R2 values (< .10), and (e) collapsing factors due to

multicollinearity. Two items (“Always on the go, as if driven by a motor” and “Picky about

cleanliness”) were identified as Heywood cases and were removed. Both items had low

communalities in EFAs (< .08) and low correlations with other items, which may explain why

the items emerged as Heywood cases. Additionally, highly correlated factors were combined to

eliminate multicollinearity (R > .90). The Hyperactivity, Oppositional, ADHD Index, Conners’

Global Index – Restless Impulsive, and DSM Hyperactivity factors were combined into one

factor, and the Cognitive Problems/Inattention and DSM Inattention factors were combined into

one factor. Removing poorly translated and highly kurtotic items resulted in two factors (Social

Problems and Perfectionism) that had fewer than three items each; thus, the Social Problems and

Perfectionism factors and corresponding items were removed from the analysis. The changes

reduced the 12-factor (80 items) model to four factors containing 57 items, based on a reduced

sample size of 1,238 cases. The four-factor model did result in an appropriate running of CFA,

but model fit to the data was not evident (RMSEA = .05; CFI = .69). Factor inter-correlations

ranged from .15 to .68 and all were statistically significant (p < .001).

Model comparisons. A summary of fit statistics (RMSEA, CFI, NNFI, IFI, AIC, and

CAIC) and Satorra-Bentler chi-square values for all CPRS-R models tested is provided in Table

22. Satorra-Bentler χ2 difference tests with Bonferroni correction were used to test model

distinguishability (Levy & Hancock, 2007; Satorra & Bentler, 2001). Only the extended-2P and

abbreviated-2P models approached the established fit criteria. Based on findings from the

Satorra-Bentler χ2 difference test and examination of fit statistics, the extended-2P and

abbreviated-2P models were determined to be indistinguishable and were deemed equivalent

models.

Page 135: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

125 Ta

ble

22

Fit

Indi

ces

for

CF

A (M

axim

um L

ikel

ihoo

d, R

obus

t) o

f CP

RS-

R M

odel

s (n

= 1

,253

) M

odel

df

S-

B χ

2

NN

FI

Rob

ust

IFI

Rob

ust

C

FI

Rob

ust

R

MSE

A

Rob

ust

R

MSE

A

90%

CI

A

IC

C

AIC

Exte

nded

N

ull

153

4,86

6.53

* -

- -

- -

4,56

0.53

3,

622.

25

O

ne-F

acto

r 13

5 1,

211.

39*

.74

.77

.77

.080

.0

76 -

.084

4

,548

.33

3,

609.

93

Tw

o-Fa

ctor

13

4 59

7.54

* .8

9 .9

0 .9

0 .0

53

.048

- .0

57

3

29.5

4 -4

92.2

1

A

bbre

viat

ed

Nul

l 17

1 5,

144.

86*

- -

- -

- 4

,802

.86

3,

754.

07

O

ne-F

acto

r 15

3 1,

284.

93*

.75

.77

.77

.077

.0

73 -

.081

978

.93

40.5

4

Two-

Fact

or

151

628.

29*

.89

.90

.90

.050

.0

46 -

.054

326

.29

-599

.84

Hyb

rid

Nul

l 55

2

,846

.37*

-

- -

- -

2,7

36.3

7 2,

399.

04

O

ne-F

acto

r 46

61

0.71

* .7

5 .8

0 .8

0 .1

01

.094

- .1

09

5

22.1

3

25

2.85

Two-

Fact

or

44

467.

41*

.85

.85

.85

.088

.0

80 -

.095

379

.41

109.

54

Empi

rical

N

ull

1

,596

10

,086

.25*

-

- -

- -

6,8

94.2

5 -2

,532

.93

O

ne-F

acto

r

1,5

39

5,04

8.62

* .5

7 .5

9 .5

9 .0

48

.046

- .0

49

1,9

70.6

2 -7

,119

.87

Se

ven-

Fact

or

1

,518

3,

671.

91*

.73

.75

.75

.038

.0

36 -

.039

635

.91

-8

,330

.54

Full

Nul

l 1,

431

12,9

32.0

6*

- -

- -

- 1

0,07

0.06

1,32

3.32

One

-Fac

tor

1,37

7 5,

737.

81*

.61

.62

.62

.051

.0

49 -

.052

2,98

3.67

-

5.43

2.86

Four

-Fac

tor

1,37

7 4,

937.

95*

.68

.69

.69

.046

.0

45 -

.047

2,18

3.95

-

6,23

2.72

N

ote.

CFA

= c

onfir

mat

ory

fact

or a

naly

sis;

S-B

χ2

= Sa

torr

a-B

entle

r chi

-squ

are;

NN

FI =

Ben

tler-

Bon

net N

on-n

orm

ed fi

t ind

ex; I

FI =

Bol

len’

s fit

in

dex;

CFI

= c

ompa

rativ

e fit

inde

x; R

MSE

A =

root

mea

n sq

uare

err

or o

f app

roxi

mat

ion;

CI =

con

fiden

ce in

terv

al; A

IC =

Aka

ike’

s In

form

atio

n C

riter

ia; C

AIC

= C

onsi

sten

t Aka

ike

Info

rmat

ion

Crit

eria

. *p

< .0

5.

Page 136: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

126

Table 23

Findings from Satorra-Bentler χ2 Difference Tests with Bonferroni Correction

Models χ2 df T Critical Value (p < .001)

Extended-2P 597.54

Abbreviated-2P 628.29 31 57.71 61.10

Note. The models were compared to one another.

Page 137: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

127

Discussion

The purpose of this study was to examine whether the factor structures of CTRS-R and

CPRS-R scores from a sample of Nepalese children were similar to the structures identified with

the normative samples composed of US and Canadian children (Conners, 1997). Factor analytic

findings do not support the factor structures Conners initially reported for scores from the CTRS-

R and CPRS-R. For both scales, a reduced dimensionality was found to be a better fit to the

data. In a Nepali context, the CTRS-R and CPRS-R scores seem to reflect a more narrow range

of symptoms than Conners intended. Two-factor models derived through EFAs and tested

through CFA were a better fit to the Nepali data than Conners’ models for CTRS-R and CPRS-R

scores. For the teacher scale, Inattention and Hyperactivity dimensions defined the factor

structure, whereas Oppositional and Inattention factors made up the parent scale. The

differences in factor structures underscore potential similarities and differences of parents and

teachers’ perceptions of Nepalese children on the CTRS-R and CPRS-R.

General Discussion

Two-factor structures with substantially reduced item content were identified for both the

teacher and parent scales. The original teacher scale contains 59 items, six empirically-derived

subscales, and five rationally-derived subscales. In this study, the best CTRS-R model, hybrid-

2T, had two factors and a total of 12 items. The original parent scale has 80 items: (a) seven

empirically-derived subscales, and (b) five rationally-derived subscales. Two equivalent, two-

factor models (the 23-item extended-2P and the 19-item abbreviated-2P) were identified as the

best models in the present study. Two possible reasons for the current findings of a reduced

factor structure for both scales are (a) cultural issues and (b) factor retention criteria.

Page 138: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

128

Cultural issues. Research suggests that some psychological constructs, such as anxiety

and somatization, are expressed differently among the Nepalese (Tol, Jordans, Regmi, &

Sharma, 2005), which seemed to evidence itself in the reduced factor structures on both scales.

Constructs that some of the factors (e.g., social problems, anxiety, and perfectionism) were

designed to tap may not be suitable for cross-cultural assessment in Nepal. For example, the

majority of items that loaded on the Social Problems factor in the normative sample were

problematic in this study. Scores on the “social problems” items had limited variability and were

severely skewed and kurtotic. On the CTRS-R item “Doesn’t have friends,” almost all the

teachers rated children the same. Few teachers felt that a particular child did not have friends;

thus, the item did not have a salient pattern coefficient on any factor. In contrast, this item

“Doesn’t have friends” in the US and Canadian normative sample had at least enough variability

to meet saliency criteria on the Social Problems factor (Conners, 1997). Nepal is a relatively

collectivist society, and the Nepalese value social harmony and group cohesion (Skinner et al.,

1998). It is possible that excluding a peer from social interaction would be culturally

inappropriate, thereby hindering the extent to which items such as “Doesn’t have friends” could

reflect social problems among Nepalese children. Further, “social problems” as viewed in the

US and Canada may not exist in Nepal or may be manifested differently such that assessment

instruments need to be indigenously developed to be of clinical utility (Leung & Wong, 2003).

Factor retention criteria. As noted earlier, Conners (1997) used minimum criteria (e.g.,

eigenvalues > 1, visual inspection of a scree plot, and pattern coefficients > .30) to determine the

numbers of factors to retain for the empirically-derived scales, and the rationally-derived scales

were not factor analyzed. Other researchers have examined the factor structures of scores from

the short forms of the revised scales (i.e., Gau et al., 2006; Gergardstein et al., 2003; Kumar &

Page 139: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

129

Steer, 2003) and also retained factors using similar criteria as Conners. As reported in the

literature review, findings have varied across studies with some researchers (Gau et al., 2006;

Gergardstein et al., 2003) reporting findings comparable to Conners and others (Kumar & Steer,

2003) reporting reduced structures. Inconsistencies in findings may be the result of the factor

retention methods employed, as certain techniques have been shown to result in over-extraction

and inconsistent findings.

Research indicates that the eigenvalue rule of one and visual inspection of the scree plot

commonly result in over-extraction (e.g., Henson & Roberts, 2006). Other factor retention

methods (parallel analysis and MAP) have been shown to yield more accurate estimates for

factor retention. Best practices for EFA dictate that a variety of factor retention methods be

used, including parallel analysis and MAP (Henson & Roberts, 2006). Thus, in the process of

scale development, the factors of the CTRS-R and CPRS-R may have been over-extracted,

thereby limiting the extent to which the normative findings are generalizable or stable (regardless

of the cultural background of the sample or version of the scale). As a result, the findings of this

study are noteworthy. One, this is the first independent study to examine the factor structure of

the long forms of the revised teacher and parent scales. Two, best practices for factor analysis

were used, resulting in a reduced factor structure for both scales. Three, these findings were

cross-validated with separate samples. However, research using data from the US normative

sample would be required to explicitly test whether the reduced factor structures identified in this

study are due to sampling differences or over-factoring of the original scales. Data from the

CTRS-R and CPRS-R normative samples were requested for the present study, but Multi-Health

Systems (MHS) has yet to respond to requests for the normative sample data.

Page 140: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

130

Conners’ Teacher Scale

Reliability. The reliability estimates of scores for the two factors (Hyperactivity and

Inattention) of the teacher scale were good to excellent (in the low to high .80 range). Reliability

estimates obtained in this study were comparable to those for factors obtained from the

normative sample (high .70 to mid .90 range; Conners, 1997). Because accuracy of

measurement is crucial for the development of instruments and their use in high stakes

assessment (Salvia & Ysseldyke, 2009), these initial reliability estimates obtained for the scores

of the teacher model are promising.

Validity. The validity of the CTRS-R scores as a broadband measure of childhood

psychopathology for Nepalese children is not supported by these findings, evidenced by poor fit

of the six- and eleven-factor models in CFA. The CTRS-R is comprised of empirically-derived

subscales (subscales included in Conners’ [1997] factor analyses) and rationally-derived

subscales (subscales not included in Conners’ factor analyses). Neither Conners (1997) nor other

researchers have tested the factor structure of items from both empirically- and rationally-derived

subscales on the long form of the CTRS-R. The current evidence indicates that the six- and

eleven-factor structures of Conners’ Teacher Scales may not be viable from a statistical or

theoretical perspective. These factor structures definitely are not viable for the Nepalese children

sampled and may not be viable for any Nepalese children. If the CTRS-R is not viable as a

broadband measure of childhood psychopathology for Nepalese children (and perhaps for

children elsewhere in the world), then what is the potential usefulness of this scale?

Findings from analyses using stringent factor selection criteria of items indicate that the

hybrid-2T model may be most appropriate in a Nepali context. The hybrid-2T model represents

Page 141: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

131

a reduced number of items from the CTRS-R and reflects two specific aspects of childhood

psychopathology: inattention (12 items) and hyperactivity (4 items). The majority of items that

met the criteria for inclusion on the hybrid-2T model are from the rationally-derived subscales

(see Appendix E for a listing of the items by scale type). Specifically, several items from the

CGI Restless/ Impulsive subscale, the ADHD Index, and the DSM-IV Inattentive subscale are

included on the hybrid-2T model. The majority of researchers use the short form of the CTRS-

R, which consists of three empirically-derived subscales (inattention, hyperactivity, and

oppositional) and the ADHD index (a rationally-derived subscale). Although items on

rationally-derived scales were never factor analyzed (Conners, 1997), these items have still been

used in studies of predictive validity and may be particularly effective at discriminating children

with and without ADHD (Charach et al., 2009; Conners, 1999). The current findings provide

support that certain items, not all, on the rationally- and empirically-derived scales may be

reflective of the ADHD construct. Using all 59 items may be unnecessary, as some items may

reflect error more than the actual construct. Thus, the reduced hybrid-2T model of the CTRS-R

may be a preferable alternative as a short, narrowband measure of inattention and hyperactivity

symptoms.

Diagnosticians, however, must exercise caution when using narrowband rating scales to

evaluate ADHD symptoms (Pelham et al., 2005). Symptoms of inattention and hyperactivity are

not specific to ADHD and are present in a wide array of childhood disorders (e.g., depression

and autism; American Psychiatric Association, 2000). Thus, exclusive use of narrowband

measures in ADHD diagnosis would inevitably result in an over-identification of children with

ADHD, because inattention and hyperactivity symptoms are not specific to ADHD and exist also

within the context of other psychological disorders (Barkley, 2006; Polanczyk & Rohdes, 2007).

Page 142: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

132

While the hybrid-2T model of the CTRS-R has the potential to be a viable narrowband ADHD

scale for children in Nepal, clinical and diagnostic use should be deferred until tools are available

to assess competing and comorbid diagnoses in a Nepali context.

Conners’ Parent Scale

Reliability. Reliability estimates of scores for CPRS-R factors (Inattention and

Oppositional) were adequate to good (in the low .70 to .80 range). The reliability estimates

obtained in this study fell on the low end relative to estimates from factors obtained from the

normative sample (low .70 to mid .90 range). Further, the reliability estimates of scores fell

below the recommended threshold for individual decision making (.90; Salvia & Ysseldyke,

2009).

Validity. The findings do not support the validity of CPRS-R scores as broadband

measures of child psychopathology, based on the poor fit of the seven- and 11-factor models.

This study was the first to examine the factor structure of all 80 items on the long form of the

CPRS-R, and the findings underscore that the factor structure of the CPRS-R as proposed by

Conners (1997) may not be viable. Specifically, the factor structures are certainly not viable for

the sample of Nepalese children and may not be viable for any population of Nepalese children.

While, CPRS-R scores may not be appropriate broadband measures of childhood

psychopathology, the scores might be useful for other purposes.

Analyses of alternative models of the CPRS-R indicate that the extended- and

abbreviated-2P models may be more appropriate versions of the CPRS-R in a Nepali context.

The CPRS-R was designed primarily as a measure of ADHD, but in the Nepalese sample,

ADHD symptoms as typically construed in the United States were not found. In the US,

hyperactivity is a primary factor of ADHD in children. Instead, the abbreviated-2P model in this

Page 143: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

133

study reflects Oppositional and Inattentive factors, and the extended-2P model taps an

Inattention factor and an Oppositional/Hyperactivity factor that predominantly reflects

oppositional behaviors with some hyperactivity items included.

Although the extended- and abbreviated-2P models were deemed the most appropriate

models tested in this study, a series of psychometric and fit problems were evident in the models.

First, neither solution accounted for more than 50% of the variance in the EFA as is

recommended for factor solutions (Streiner, 1994). Also, the reliability estimates of the factors

were in the .80 range and below the recommended level for individual decision making (Salvia

& Ysseldyke, 2009). In CFA, the extended- and abbreviated-2P models approached but did not

meet criteria for good fit, suggesting additional modifications to the models (e.g., allowing error

variance to correlate) may improve model fit. Finally, other studies of the CPRS-R identified a

factor that solely or predominantly reflected hyperactivity (e.g., Conners, 1997; Epstein et al.,

1998), but no such factor was identified in this study.

It is unclear why a Hyperactivity factor did not emerge in this study. An examination of

the effect sizes (communalities and pattern coefficients) of items on the CPRS-R shows that the

values were lower than expected. Each item contributed only a small proportion of variance to

the overall factor structure, and the majority of the variance in most items was accounted for by

error or unique variance. Thus, a low proportion of common variance and high degree of error

and unique variance may have prevented derivation of a Hyperactivity factor and other factors as

well. The present findings, withstanding further research, indicate that the CPRS-R (as is) is not

an appropriate assessment of the most common form of ADHD (ADHD-Combined subtype) for

Nepalese youth. It is unclear if the extended- and abbreviated-2P models are viable measures of

Nepalese parents’ perceptions of attention problems (e.g., ADHD-Predominantly Inattentive

Page 144: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

134

subtype) and oppositional behaviors of their children. Cultural differences between Nepalese

children and the adults who raise and teach them in comparison to the US and Canadian children

included in the standardization sample may have contributed to the absence of a hyperactivity

factor on the Nepalese version of this scale.

Scale Comparisons

Similarities and differences are evident between the teacher and parent scales. Reduced

(two-factor) structures best defined both scales. The Inattentive factors on both scales have

markedly similar item content and reliability estimates of the scores. It is not surprising that

similar findings were obtained for the teacher and parent scales because both scales were

designed to assess symptoms of ADHD, which are generally expected to be present in both the

home and school environment (American Psychiatric Association, 2001).

Differences in findings were also identified between the parent and the teacher scales.

Several teacher models met criteria for good fit, and the hybrid-2T model exceeded all specified

goodness of fit criteria. On the parent scale, two models approached a good fit to the data, but

specified goodness of fit criteria were not met. Also, a Hyperactivity factor was identified on the

teacher scale, while a predominantly Oppositional factor was identified on the parent scale. Why

would a Hyperactivity factor emerge on the teacher scale, but an Oppositional factor emerge on

the parent scale? Why would the parent models have problems with fitting to the data that were

not evident on the teacher scale? Prior studies have reported that parent and teacher responses

differ in regard to ratings of the severity of children’s behaviors, but structural differences in the

scales have not been reported (see Giannaris et al., 2001, for review). Methodological or

environmental factors may have contributed to the structural differences between the parent and

teacher scales.

Page 145: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

135

Methodological factors. A major methodological concern in this study is the language

used for administration and its subsequent influence on the translation and inclusion of items in

the analyses. Nepali is the language of education in Nepal and all teachers are fluent in Nepali.

Thus, administration of all teacher scales was in Nepali and subsequent translations were based

on this language. However, the language used in the administration of the parent scale was not

clear-cut and possibly had an impact on the findings obtained for the parent sample. Many

people in Nepal (40-50%; Eagle, 2010) do not speak Nepali as a first language. Several

languages, such as Bhojpuri, Tharu, Tamang, and Newari, are commonly spoken in the Terai

(Eagle). Some parents in the sample did not speak Nepali, but data on how many non-Nepali

speakers were in the sample were not collected (P. Christian, personal communication,

December 2, 2009). In Nepal, members of some caste groups (Vaishyas, Shudras, and non-

Hindus) are less likely to be fluent in Nepali (Eagle, 2010), and members of these castes made up

the majority of the sample (see pp. 49-51). Thus, a language barrier may have existed between

the administrators and the parents, potentially diminishing the accuracy of parent responses.

A number of parent scales were translated in Maithili during the interview session to

parents who did not speak Nepali. While this practice may have facilitated communication with

these parents, accuracy may have been compromised in evaluating the adequacy of items for use.

Unlike the Nepali version of the CPRS-R, the Maithili version was translated using a forward-

only translation approach, which means that no back translations were conducted

(Maneesriwongul & Dixon, 2004). A drawback to the forward-only translation approach is that

it is impossible to determine whether responses reflect the intended construct or are a function of

errors in translation (Maneesriwongul & Dixon, 2004). Moreover, adequacy of translation was

an important consideration in decisions made about whether to retain items for the teacher and

Page 146: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

136

parent models. Because data on adequacy of translation was only available for the Nepali

versions of the scales, it is possible that any number of items on the Maithili version of parent

scale may have been poorly translated. Data on frequency of use of the Nepali and Maithili

CPRS-R forms were also not available. In essence, differences in language of administration and

the possible inclusion of poorly translated Maithili items on the parent models may have

impeded derivation of a logical, accurate factor structure for CPRS-R scores.

Sex of the parent respondents and interviewers is another methodological issue that may

have influenced the factor structure of the parent scores. Most of the primary parent respondents

were mothers, and male research assistants conducted all of the interviews. Prior research (e.g.,

Axinn, 1991), including a study in the Nepali Terai (Robinson-Pant, 2001), indicates that

Nepalese women may underreport negative information about their families if the interviewer is

male. Therefore, it is possible that mothers in the sample may have been uncomfortable

reporting certain behaviors to male interviewers, which thereby contributed to altered response

patterns and a different factor structure for the parent scores. Data on the sex of teacher

respondents were not collected. Thus, the extent to which sex roles may have influenced teacher

responses could not be examined.

Environmental factors. Environmental differences between Nepalese homes and

schools may also account for differences in the factor structures of parent and teacher scores. In

general, strict disciplinary procedures are used in Nepalese classrooms (UNICEF, 2004).

Consequences for negative behaviors are often severe by US standards (e.g., corporal

punishment) and are applied immediately (Bartlett et al., 2007). While corporal punishment may

also be used in Nepalese homes, the home environment is generally less structured than the

school environment. Research conducted in the United States indicates that environmental

Page 147: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

137

context may influence behaviors of children with ADHD. Specifically, behaviors of children

with ADHD improve in highly structured environments and are exacerbated in unstructured

environments (e.g., Barkley, 2006; Granger, Whalen, Henker, & Cantwell, 1996). It is possible

that children in the present sample were more likely to display minor externalized behaviors

(fidgeting, restlessness, etc.) in the school environment and more severe behaviors (losing

temper, defiance, etc.) at home, which may have influenced finding the Hyperactivity factor with

the teacher-rated sample and the Oppositional factor with the parent-rated sample.

Alternatively, it is possible that children are engaging in similar behaviors in the home

and school environment but that parents and teachers perceive the behaviors differently. ADHD

is characterized by externalizing behaviors (American Psychiatric Association, 2000). Findings

from several studies conducted in the US indicate that teacher ratings more accurately reflect

externalized behavior problems than do parent ratings (e.g., DuPaul, Power, McGoey, Ikeda, &

Anastopoulos, 1998). Teacher ratings are believed to be more accurate than parent ratings

because of differences in available bases of comparison: teachers compare student behavior to

that of many other students, while parents typically have fewer opportunities to observe the

behavior of children other than their own. However, Conners (1997) reported finding the same

six factors on both the teacher and parent scales, whereas structural generalizability across scales

was not obtained in the present study. There is no research on teacher versus parent bases for

behavioral comparison and accuracy of ratings within a Nepali context. Additionally, it is

possible that other differences between parents and teachers, such as educational differences,

may have led teachers to be more adept at recognizing and reporting hyperactivity symptoms.

However, further research is needed before such conclusions can be drawn.

Page 148: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

138

Implications

ADHD theory. Over the past several decades, controversy regarding whether ADHD is

truly a disorder of attention per se has persisted(see Nigg, 2005, for review). Some researchers

(e.g., Barkley, 2006) contend that ADHD symptoms are largely a function of deficits in

behavioral inhibition, and attentional difficulties are viewed as secondary impairments. Other

researchers (e.g., Robertson, Ward, Ridgeway, & Nimmo-Smith, 1994) claim that ADHD

symptoms are largely attributable to deficits in attentional processes, such as sustained and

selective attention. In this study, an inattention factor (similar to that identified in research based

in the US) emerged on both the teacher and parent scales. Thus, this study suggests that

attention problems exist across settings (home and school) and across cultures (US and Nepal).

Thereby, these findings provide additional support for theories suggesting that attentional

difficulties are central components of ADHD.

Use of CRS-R. Several issues pertaining to the appropriateness of using the Conners’

teacher and parent scales in a Nepali context warrant discussion. When introducing products of a

field to a new cultural context, professionals must consider potential ramifications and attempt to

address them. Childhood mental health problems (such as ADHD) are largely unrecognized in

Nepal, and the implications of introducing a measure of ADHD and the concept of ADHD must

be carefully considered. Care should be taken to consider the role of traditional healing methods

and other cultural practices and values prior to medical or psychological treatments for ADHD to

ensure appropriate understanding of a child’s cognition and behavior. Scholars of multi-cultural

psychology recommend that culturally knowledgeable individuals such as traditional healers and

religious leaders as well as local cultural workers be consulted and potentially used as mediators

in the diagnostic and treatment processes (Nastasi, Varjas, Bernstein, & Jayasena, 2000).

Page 149: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

139

It is also important to consider the widespread stigmatization of mental illness in Nepal

(Skinner et al., 1998). Because research suggests that disorders of the body may be less

stigmatized in Nepal, education and emphasis on the well known biological components of

ADHD may be appropriate (Skinner et al, 1998). Further, some mental health professionals

working in Nepal report that problem-focused, behavioral explanations of problems and

proposed solutions tend to be relatively well-received by Nepalese families (Tol et al., 2005).

Given that multi-modal treatment approaches for ADHD (including medication and behavioral

strategies) have been effective within the United States (Jensen et al., 2001), it is possible that

such treatment approaches could be effective and culturally sensitive in a Nepali context.

However, additional research is necessary to verify the efficacy of such treatments in Nepal.

Finally, it is critical that diagnostic measurements of ADHD symptoms not be misused or

employed prematurely. Inadequate assessment of a psychological disorder is likely to lead to

misdiagnosis. Some (e.g., Barkley, 2008; Hinshaw & Ciccheti, 2000) contend that misdiagnosis

can result in skepticism about the existence of a disorder and further stigmatization of affected

individuals. Misdiagnosis of ADHD is believed to have contributed to skepticism and

stigmatization even within the United States (Barkley, 2008). Given that mental illnesses are

already severely stigmatized in Nepal, care must be taken to ensure that inappropriate

measurement and premature diagnosis of ADHD do not cause or exacerbate stigmatization of

Nepalese children and adults with attention and hyperactivity problems.

CTRS-R. The hybrid-2T model of CTRS-R has the potential to be a useful tool in

assessing symptoms of inattention and hyperactivity for screening or research purposes.

However, further research on the stability of this two-factor structure and its discriminant and

predictive validity for Nepali children is necessary. Even if additional evidence for validity of

Page 150: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

140

CTRS-R scores is produced, the hybrid-2T model is, at best, a narrowband measure of ADHD

symptoms (inattention and hyperactivity), because symptoms of inattention and hyperactivity

may be evident in many childhood psychological disorders (e.g., autism, mental retardation, and

anxiety; American Psychiatric Association, 2001). However, the CTRS-R could potentially be

useful as a diagnostic measure of ADHD symptoms in Nepalese children when used in

conjunction with other measures that have been validated in the cultural context of Nepal.

CPRS-R. The CPRS-R is not an adequate measure of ADHD-Combined subtype

symptoms in Nepalese children. Rather than assessing inattention and hyperactivity, the CPRS-

R appears to be tapping inattention and oppositional behaviors in this Nepali sample. The

extended- and abbreviated-2P models of the CPRS-R identified in this study may be viable

measures of the inattentive dimension of ADHD as well as opposition, but further measurement

refinement and research on the factor structures is necessary because of identified problems with

model fit.

NNIPS research. The findings of this study yield implications for NCOG studies on the

effects of prenatal and early childhood iron/folic acid and zinc supplementation on attention,

hyperactivity, and other constructs among school-age, Nepalese children. For the teacher scale,

it is recommended that scores from the CTRS-R Hybrid-2T model Inattention and Hyperactivity

factors be used to evaluate the outcomes of supplementation trials. Additional analyses

examining the extent to which scores from the Inattention and Hyperactivity factors correlate

with scores from other measures administered in this study (e.g., behavioral observations, neuro-

psychological tests, measures of test session behaviors) may be valuable precursors to analyses

evaluating the outcome of supplementation trials.

Page 151: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

141

In regard to the parent scale, decisions about which model to use may depend on

theoretical consideration of which aspects of behavior are most likely to be influenced by

supplementation. Scores from either the Extended-2P (Inattention and Oppositional/

Hyperactivity factors) or Abbreviated-2P models (Inattention and Oppositional factors) may be

useful as part of multi-component assessments of supplementation outcomes. Findings from the

CPRS-R should be interpreted carefully and in conjunction with findings from other measures.

Further, studies examining the concurrent and predictive validity of CPRS-R scores would be

useful.

Limitations

Findings of this study must be evaluated in light of several limitations: (a) demographic

aspects, (b) methodological, and (c) statistical issues. Several demographic features of the

sample limit the external validity of the findings. Only seven- and eight-year-old children were

in the sample; thus, the findings may not be applicable to children of different ages and stages of

development (e.g., pre-school children, older children, or adolescents). Additionally, all data

were collected in the Sarlahi district of Nepal, and as a result, the findings may not generalize to

individuals living in other regions of Nepal. Although Nepal is a geographically small nation,

the regions of Nepal are highly diverse on factors such as ethnicity, language, and religious

practices (Central Intelligence Agency, 2010). The present findings may not generalize to

regions of Nepal wherein the socio-cultural norms differ substantially from those of the Sarlahi

district. Additionally, the findings may not be applicable to members of demographic subgroups

(e.g., caste, region of ancestry, or ethnic) within the sample.

As mentioned previously, methodological and measurement issues may have influenced

the findings of this study. Sex differences between parent respondents (mothers) and male

Page 152: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

142

interviewers may have influenced the extent to which a logical factor structure could be derived

from CPRS-R scores. Further, language and the translation of items are measurement issues that

may have affected the outcome of the findings.

Several statistical issues may have limited the strength of the findings obtained.

Specifically, some problems have been associated with the statistics used for CFA model

comparison. The Satorra-Bentler χ2 difference test was used to evaluate model

distinguishability. Like the Satorra-Bentler χ2 statistic, the difference test is sensitive to sample

size (Satorra & Bentler, 2001). Therefore, it is possible that models were falsely identified as

distinguishable due to large sample size. Also, the AIC and CAIC (predictive fit statistics) tend

to favor simple models more than other statistics (e.g., RMSEA and CFI); thus, it is possible that

more parsimonious models were artificially identified as having a better fit (Byrne, 2006).

Statistics such as the Satorra-Bentler χ2 difference test, AIC, and CAIC were used because they

have been shown to be the most appropriate statistics for comparison between non-nested models

(Kline, 2005; Levy & Hancock, 2007).

Future Research

One set of future studies should focus on the factor structure of the scales. No study prior

to this one has examined the factor structure of the long forms of the revised Conners’ scales

with empirically and rationally derived items included in the analyses. Thus, further

examination of the long forms is needed, not only for Nepalese children, but for US children as

well. Only the factor structure of the empirical scales has been examined, whereas the factor

structure of the rational items has been assumed based on theory and prior research. Substantial

overlap appears to exist between the two sets of items and additional structural work could assist

in eliminating the redundancy of items and creating a potentially more efficient and effective

Page 153: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

143

scale. Also, any future research needs to use more stringent criteria in selecting factors. The

current factor structure of the CRS-R has been based on EFA and CFA, but minimum criteria

(i.e., eigenvalue rule of one, visual inspection of the scree plot, and pattern coefficients of .30 or

greater; Conners, 1997) were used. In a Nepali context, research examining the stability of the

factor structures identified in this study and the extent to which the structures are generalizable to

demographically diverse groups of Nepali children is necessary. Further, a study of factorial

invariance of the factor structures of CTRS-R and CPRS-R scores across new samples of

Nepalese and US citizens would illuminate the potential influence of culture on the scales.

As a prerequisite for validity, reliability of scores must also be examined. In the present

study, internal consistency of scores was evaluated, but additional research on the inter-rater

reliability and stability of CTRS-R and CPRS-R scores in Nepal is necessary. Because factor

structure is only one aspect of validity, it is important that other aspects of validity also be

examined. For example, future research should focus on the predictive and discriminant validity

of scores from the CTRS-R and CPRS-R in Nepalese children.

Aside from further examination of psychometric properties of scores from the CTRS-R

and CPRS-R (Conners, 1997), future research should also focus on the assessment of ADHD in

Nepal. Before a true diagnostic measure of ADHD can be developed for the Nepali context,

research is needed on assessment of childhood disorders that are commonly comorbid with

ADHD and disorders that may produce similar symptoms (e.g., depression, anxiety, and autism).

Also, research should examine the extent to which hyperactivity and inattention cause

impairment (e.g., school problems, family difficulties, problems in social interaction, and

adaptive behavior deficits; Barkley, 2006) in Nepalese children. It is possible that the extent to

which symptoms of inattention and hyperactivity result in impairment varies as a function of

Page 154: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

144

culture. Research examining the extent to which the DSM-IV TR criteria for ADHD are

applicable in a Nepali context is warranted. The Conners’ Rating Scales (Conners, 1997) contain

a set of items (the DSM-IV subscales) that are nearly identical in wording to the symptoms listed

in the DSM-IV TR criteria for ADHD. Some, but not all, of the DSM-IV TR items were

identified as viable in the present study.

Multi-lingual administration of the CPRS-R (Nepali and Maithili) was a critical issue in

this study. Additional research examining the factor structure of the Maithili versions of the

Conners’ scales is warranted. Also, the translation of the CRS-R into other languages of Nepal,

such as Newari, Tamang, and Tharu, and evaluation of scales translated into these languages

would expand the potential usefulness of the Conners’ scales in Nepal.

Conclusions

The psychometric properties of scores from an instrument should be examined with

members of the population before using the instrument in a novel cultural context (Sattler, 2001).

The CRS-R are used for research and clinical purposes throughout the world (Conners, 1997;

Koonce, 2007). Thus, research examining the factor structures of scores from the CTRS-R and

CPRS-R across cultures is necessary. In two samples of Nepalese children, scores from the

CTRS-R and CPRS-R reflected a narrower range of factors than originally reported in the

normative sample (Conners, 1997). Findings indicate that the CTRS-R scores reflect Inattention

and Hyperactivity factors, and the 12-item hybrid-2T model exceeded all specified goodness of

fit criteria in CFA. Thus, CTRS-R scores may be viable measures of teacher perceptions of

inattention and hyperactivity in Nepalese children. However, Inattention and Oppositional

factors were identified for the CPRS-R, and all models fell short of the specified goodness of fit

criteria in CFA. The findings indicate that CPRS-R scores may not adequately measure parent

Page 155: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

145

perceptions of childhood behavior problems in a Nepali context or that childhood behavior

problems (as perceived by parents) may differ in a Nepali context. This information regarding

the differences in and viability of factor structures of scores from the CTRS-R and CPRS-R

should be considered in future research on childhood behavior problems in Nepalese children.

Page 156: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

146

References

Achenbach, T. M. (1991). Integrative guide for the 1991 CBCL/ 4-18, YSR and TRF profiles.

Burlington, VT: University of Vermont.

Achenbach, T. M. (2005). Advancing assessment of children and adolescents: Commentary on

evidence-based assessment of child and adolescent disorder. Journal of Clinical Child

and Adolescent Psychology, 34, 541-547. doi: 10.1207/s15374424jccp3403_9

Achenbach, T. M., & Rescorla, L. A. (2001). Manual for the ASEBA school-age forms &

profiles. Burlington, VT: University of Vermont.

Adler, K. (2002). The measure of all things: The seven year odyssey and hidden error that

changed the world. New York, NY: Free Press.

ADORE Study Group. (2006). Study design, baseline patient characteristics, and intervention in

a cross-cultural framework: Results from the ADORE study. European Child and

Adolescent Psychiatry, 15, 14-26. doi: 10.1007/s00787-006-1002-0

Ahearn, L. M. (2001). Invitations to love: Literacy, love, and social change in Nepal. Ann Arbor,

MI: The University of Michigan Press.

Akaike, H. (1987). Factor analysis and AIC. Psychometrica, 52, 317-332. doi:

10.1007/BF02294359

Alderson, R. M., Rapport, M. D., & Kofler, M. J. (2007). Attention deficit hyperactivity disorder

and behavioral inhibition: A meta-analytic review of the stop-signal paradigm. Journal of

Abnormal Child Psychology, 35, 745-758. doi: 10.1007/s10802-007-9131-6

American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorder,

third edition. Washington, DC: Author.

Page 157: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

147

American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorder,

third edition, revised. Washington, DC: Author.

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorder,

fourth edition. Washington, DC: Author.

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorder,

fourth edition, text revision. Washington, DC: Author.

American Psychological Association. (2002). American Psychological Association ethical

principles of psychologists and code of conduct. Retrieved from http://www.apa.org

Arminio, J. (2010). Waking up White: What it means to accept your legacy, for better or for

worse. In M. Adams, W. J. Blumenfeld, C. R. Castaneda, H. W. Hackman, M. L. Peters,

& X. Zuniga (Eds.), Readings for diversity and social justice (pp. 125-127). New York,

NY: Routledge, Taylor and Francis.

Artiles, A. J., & Ortiz, A. A. (2002). English language learners with special education needs:

Identification, assessment, and instruction. McHenry, IL: Delta Publishing Company.

Asian Centre for Human Rights (2009). Madhes: The challenges and opportunities for a stable

Nepal. Briefing Papers on Nepal. Retrieved from http://www.achrweb.org

Axinn, W. (1991). The influence of interviewer sex on responses to sensitive questions in Nepal.

Social Science Research, 20, 303-318. doi: 10.1016/0049-089X(91)90009-R

Barkley, R. A. (1999). Response inhibition in Attention Deficit Hyperactivity Disorder. Mental

Retardation and Developmental Disabilities Research Reviews, 5, 177-184. doi:

10.1002/(SICI)1098-2779(1999)5:3<177::AID-MRDD3>3.0.CO;2-G

Barkley, R. A. (2006). Attention deficit hyperactivity disorder: A handbook for diagnosis and

treatment (3rd ed.). New York, NY: The Guilford Press.

Page 158: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

148

Barkley, R. A. (2008, November). Advances in ADHD theory, diagnosis, and management.

Paper presented at J & K seminars in Lancaster, PA.

Bartlett, M. S. (1950). Tests of significance in factor analysis. British Journal of

Psychology (Statistical Section), 3, 77-85.

Bartlett, S., Sunar, B., Malla, H., Pandey, G., Pradhanaunga, U., & Sapkota, P. (2007). Finding

hope in troubled times: Education and protection for children in Nepal. Retrieved from

www.norad.no/en

BELLA Study Group. (2008). Psychometric properties of two ADHD questionnaires: Comparing

the Conners’ scale and the FBB-HKS in the general population of German children and

adolescents – results from the BELLA study. European Child and Adolescent Psychiatry,

17, 106-115. doi: 10.1007/s00787-008-1012-1

Bennett, L. (2005, May). New frontiers of social policy. Paper presented at the World Bank

Conference in Arusha, Tanzania. Retrieved from http://www.worldbank.org

Bentler, P. M., & Bonnet, D. G. (1980). Significance tests and goodness of fit in the analysis of

covariance structures. Psychological Bulletin, 88, 588-606. doi: 0033-2909/80/8803-

0588S00.75

Biederman, J., Mick, E., Faraone, S. V., Braaten, E., Doyle, A., Spencer, T.,… Johnson, M. A.

(2002). Influence of gender on attention deficit hyperactivity disorder in children

referred to a psychiatric clinic. American Journal of Psychiatry, 159, 36-42.

Bollen, K. A. (1989). A new incremental fit index for general structural models. Sociological

Methods & Research, 17, 303-316.

Page 159: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

149

Boonstra, A. M., Oosterlaan, J., Sergeant, J. A., & Buitelaar, J. K. (2005). Executive functioning

in adult ADHD: A meta-analytic review. Psychological Medicine, 1097-1108. doi:

10.1017/S003329170500499X

Bozdogan, H. (1987). Model selection and Akaike’s information criteria (AIC): The general

theory and its analytical extensions. Psychometrika, 52, 345-370. doi: 0033-

3123/87/0900-SS03

Bradley, R. H., Caldwell, B. M., Rock, S. L., Hamrick, H. M., & Harris, P. (1988). Home

Observation for Measurement of the Environment: Development of a home inventory for

use with families having children 6 to 10 years old. Contemporary Educational

Psychology, 13, 58-71. doi: 10.1016/0361-476X(88)90006-9

Briggs, N. E., & MacCallum, R. C. (2003). Recovery of weak common factors by maximum

likelihood and ordinary least squares estimation. Multivariate Behavioral Research, 38,

25-56. doi: 10.1207/S15327906MBR3801_2

Brislin, R. W. (1970). Back translation for cross-cultural research. Journal of International

Psychology, 1, 185-216. doi: 10.1177/135910457000100301

Brown, R. T., Reynolds, C. R., & Whitaker, J. S. (1999). Bias in mental testing since Bias in

mental testing. School Psychology Quarterly, 14, 208-238. doi: 10.1037/h0089007

Bush, G., Valera, E. M., & Seidman, L. J. (2005). Functional neuroimaging of attention deficit

hyperactivity disorder: A review and suggested future directions. Biological Psychiatry,

57, 1273-1284. doi: 10.1016/j.biopsych.2005.01.034

Byrne, B. M. (2005). Factor analytic models: Viewing the structure of an assessment instrument

from three perspectives. Journal of Personality Assessment, 85, 17-32. doi:

10.1207/s15327752jpa8501_02

Page 160: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

150

Byrne, B. M. (2006). Structural equation modeling with EQS: Basic concepts, applications, and

Programming (Vol. 2). Thousand Oaks, CA: Sage Publications, Inc.

Castellanos, F. X., & Tannock, R. (2002). Neuroscience of attention deficit hyperactivity

disorder: The search for endophenotypes. Nature Reviews, 3, 617-628. doi:

10.1038/nrn896

Cattell, R. B. (1966). The scree test for the number of factors. Multivariate Behavioral

Research,1, 245-276. doi: 10.1207/s15327906mbr0102_10

Central Bureau of Statistics (2005). Nepal. Retrieved from: http://www.cbs.gov.np.

Central Bureau of Statistics (2007). Nepal Living Standards Survey. Retrieved from

http://www.cbs.gov

Central Intelligence Agency (2010). The CIA world factbook: Nepal. Retrieved from

https://www.cia.gov

Chang, D. F., & Sue, S. (2003). The effects of race and problem type on teachers’ assessments of

student behavior. Journal of Consulting and Clinical Psychology, 71, 235-242. doi:

10.1037/0022-006X.71.2.235

Charach, A., Chen, S., Hogg-Johnson, S., & Schachar, R. J. (2009). Using the Conners’ Teacher

Rating Scale–Revised in school children referred for assessment. Canadian Journal of

Psychiatry, 54, 232-241.

Cole, P. M., & Tamang, B. L. (1998). Nepali children’s ideas about emotional displays in

hypothetical challenges. Developmental Psychology, 34, 640-646. doi: 10.1037/0012-

1649.34.4.640

Cole, P. M., Walker, A. R., & Lama-Tamang, M. S. (2006). Emotional aspects of peer relations

among children in rural Nepal. In X. Chen, D. C. French, & B. H. Schneider (Eds.), Peer

Page 161: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

151

relationships in social context (pp. 148-169). New York, NY: Cambridge University

Press.

Conners, C. K. (1989). Conners’ Rating Scales manual. New York: Multi Health Systems.

Conners, C. K. (1997). Conners’ Rating Scales-Revised technical manual. North Tonawanda,

NY: Multi-Health Systems.

Conners, C. K., Sitarenios, G., Parker, J. D., & Epstein, J. N. (1998). The revised Conners’

parent rating scale (CPRS-R): Factor structure, reliability, and criterion validity. Journal

of Abnormal Child Psychology, 26, 257-268. doi: 10.1023/A:1022602400621

Costello, A. B., & Osborne, J. W. (2005). Best practices in exploratory factor analysis: Four

recommendations for getting the most from your analysis. Practical Assessment,

Research & Evaluation, 10, 2-9. Retrieved from: http://pareonline.net

Conway, J. M., & Huffcutt, A. I. (2003). A review and evaluation of exploratory factor analysis

practices in organizational research. Organizational Research Methods, 6, 147-168. doi:

10.1177/1094428103251541

Curran, P. J., West, S. G., & Finch, J. F. (1996). The robustness of test statistics to nonnormality

and specification error in confirmatory factor analysis. Psychological Methods, 1, 16-29.

doi: 1082-989X/96/

Dawes, R. M., Faust, D., & Meehl, P. E. (1989). Clinical versus actuarial judgment. Science,

243, 1668-1674. doi: 10.1126/science.2648573

Deb, S., Dhaliwal, A. J., & Roy, M. (2008). The usefulness of Conners’ Rating Scales–Revised

in screening for attention deficit hyperactivity disorder in children with intellectual

disabilities and borderline intelligence. Journal of Intellectual Disability Research, 52,

950-965. doi: 10.1111/j.1365-2788.2007.01035.x

Page 162: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

152

Dereboy, C., Senol, S., Sener, S., & Dereboy, F. (2007). Validation of the Turkish versions of the

short-form Conners’ Teacher and Parent Rating Scales. Turkish Journal of Psychiatry,

18, 48-58.

Diana v. State Board of Education, CA 70 RFT (N.D. Cal. 1970).

DuPaul, G. J., Power, T. J., Anastopoulos, A. D., & Reid, R. (1998). ADHD Rating Scale–IV.

New York: Guilford Press.

DuPaul, G. J., Power, T. J., McGoey, K. E., Ikeda, M. J., & Anastopoulos, A. D. (1998).

Reliability and validity of parent and teacher ratings of Attention Deficit Hyperactivity

Disorder symptoms. Journal of Psychoeducational Assessment, 16, 55-68. doi:

10.1177/073428299801600104

Durkin, M.S., Davidson, L. L., Desai, P., Hasan, M., Kahn, N., Shrout, P. E., … Zaman, S. S.

(1994). Validity of the Ten Questions Screen for childhood disability: Results from

population-based studies Bangladesh, Jamaica, and Pakistan. Epidemiology, 5, 283-289.

Eagle, S. (1999). The language situation in Nepal. Journal of Multilingual and Multicultural

Development, 20, 272-327. doi: 10.1080/01434639908666382

Edelbrock, C. S., & Achenbach, T. M. (1984). The teacher version of the Child Behavior Profile:

Boys ages 6-11. Journal of Consulting and Clinical Psychology, 52, 207-217. doi:

10.1037/0022-006X.52.2.207

El-Hassan Al-Awad, A. M., & Sonuga-Barke, E. J. (2002). The application of the Conners’

Rating Scales to a Sudanese sample: An analysis of parents’ and teachers’ ratings of

childhood behavior problems. Psychology and Psychotherapy: Theory, Practice, and

Research, 75, 177-187.

Page 163: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

153

Epstein, J. N., March, J. S., Conners, C. K., & Jackson, D. L. (1998). Racial differences on the

Conners’ Teacher Rating Scale. Journal of Abnormal Child Psychology, 26, 109-118.

doi: 10.1023/A:1022617821422

Everett, J. E. (1983). Factor comparability as a means of determining the number of factors and

their rotation. Multivariate Behavioral Research, 18, 197-212. doi:

10.1207/s15327906mbr1802_5

Fabrigar, L. R., Wegener, D. T., MacCallum, R. C., & Strahan, E. J. (1999). Evaluating the use

of exploratory factor analysis in psychological research. Psychological Methods, 4, 272-

299. doi: 10.1037/1082-989X.4.3.272

Fan, X. (2003). Two approaches for correcting correlation attenuation caused by measurement

error: Implications for research practice. Educational and Psychological Measurement, 6,

915-930. doi: 10.1177/0013164403251319

Faraone, S. V., Perlis, R. H., Doyle, A. E., Smoller, J. W., Goralnick, J. J., Holmgren, M. A., &

Sklar, P. (2005). Molecular genetics of attention deficit hyperactivity disorder. Biological

Psychiatry, 57, 1,313-1,323. doi: 10.1016/j.biopsych.2004.11.024

Faraone, S. V., Sergeant, J., Gillberg, C., & Biederman, J. (2003). The worldwide prevalence of

ADHD: Is it an American condition? World Psychiatry, 2, 104-113.

Foreman, D. M., & Ford, T. (2008). Assessing the diagnostic accuracy of the identification of

hyperkinetic disorders following the introduction of government guidelines in England.

Child and Adolescent Psychiatry and Mental Health, 2, 2-7. doi: 10.1186/1753-2000-2-

32

Page 164: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

154

Gau, S. S., Soong, W., Chiu, Y., & Tsai, W. (2006). Psychometric properties of the Chinese

version of the Conners’ Parent and Teacher Rating Scales–Revised: Short Forms.

Journal of Attention Disorders, 9, 648-659. doi: 10.1177/1087054705284241

Gerhardstein, R. R., Lonigan, C. J., Cukrowicz, K. C., & McGuffey, J. A. (2003). Factor

structure of the Conners’ Teacher Rating Scale–Short Form in a low-income preschool

sample. Journal of Psychoeducational Assessment, 21, 223-233. doi:

10.1177/073428290302100301

Gianarris, W. J., Golden, C. J., & Greene, L. (2001). The Conners’ Parent Rating Scales: A

critical review of the literature. Clinical Psychology Review, 21, 1,061-1,092.

doi:10.1016/S0272-7358(00)00085-4

Gillaspy, J. A. (1996, January). A primer on confirmatory factor analysis. Paper presented at the

annual meeting of the southwest educational research association, New Orleans, LA.

Retrieved from http://www.eric.ed.gov

Goldberg, L. R., & Velicer, W. F. (2006). Principles of exploratory factor analysis. In S. Strack

(Ed.), Differentiating normal and abnormal personality (2nd ed.; pp. 209-237). New

York, NY: Springer.

Gorsuch, R. L. (1997). Exploratory factor analysis: Its role in item analysis. Journal of

Personality Assessment, 68, 532-560. doi: 10.1207/s15327752jpa6803_5

Goyette, C. H., Conners, C. K., & Ulrich, R. F. (1978). Normative data on revised Conners’

Parent and Teacher Rating Scales. Journal of Abnormal Child Psychology, 6, 221-236.

doi: 10.1007/BF00919127.

Granger, D. A., Whalen, C. K., Henker, B., & Cantwell, C. (1996). ADHD boys’ behavior

during structured classroom social activities: Effects of social demands, teacher

Page 165: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

155

proximity, and methylphenidate. Journal of Attention Disorders, 1, 16-30. doi:

10.1177/108705479600100102

Guadagnoli, E., & Velicer, W. F. (1988). Relation of sample size to the stability of component

patterns. Psychological Bulletin, 103, 265-275. doi: 10.1037/0033-2909.103.2.265

Gurung, H. (1996). Ethnic demography of Nepal. Retrieved from

http://www.nepaldemocracy.org

Gurung, H. (2005). The Dalit context. Occasional Papers in Sociology and Anthropology.

Retrieved from http://www.nepjol.info.

Haccethu, K. (2007). Madhesi nationalism and restructuring the Nepali state. Centre for Nepal

and Asian Studies. Retrieved from http://www.cnastu.org.np

Haig, B. D. (2005). Exploratory factor analysis, theory generation, and scientific method.

Multivariate Behavioral Research, 40, 303-329. doi: 10.1207/s15327906mbr4003_2

Halsall, P. (1997). Modern history sourcebook: Summary of Wallerstein on world system theory.

Internet Modern History Sourcebook. Retrieved from

http://www.fordham.edu/halsall/mod/wallerstein

Hambelton, R. K., Merenda, C. F., & Spielberger, P. D. (2005). Adapting educational and

psychological tests for cross cultural assessment. Mahwah, NJ: Lawrence Erlbaum.

Hardy, K. K., Kollins, S. H., Murray, D. W., Riddle, M. A., Greenhill, L. G., Cunningham, C.,

… Chuang, S. Z. (2007). Factor structure of parent and teacher rated Attention Deficit

Hyperactivity Disorder Symptoms in pre-schoolers with Attention Deficit Hyperactivity

Disorder Treatment Study (PATS). Journal of Child and Adolescent

Psychopharmacology, 17, 621-633. doi: 10.1089/cap.2007.0073

Page 166: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

156

Helton, S. C., Corwyn, R. F., Bonner, M. J., Brown, R. T., & Mulhern, R. K. (2006). Factor

analysis and validity of the Conners’ Parent and Teacher Rating Scales in childhood

cancer survivors. Journal of Pediatric Psychology, 31, 200-208. doi:

10.1093/jpepsy/jsj010

Henson, R. K., & Roberts, J. K. (2006). Use of exploratory analysis in published research:

Common errors and some comments on improved practice. Educational and

Psychological Measurement, 66, 393-416. doi: 10.1177/0013164405282485

Hillemeier, M. M., Foster, E. M., Heinrichs, B., & Heier, B. (2007). Racial differences in

parental reports of attention deficit hyperactivity disorder behaviors. Journal of

Developmental & Behavioral Pediatrics, 28, 353-361. doi:

10.1097/DBP.0b013e31811ff8b8

Hinshaw, S. P., & Cicchetti, D. (2000). Stigma and mental disorder: Conceptions of illness,

public attitudes, personal disclosure, and social policy. Development and

Psychopathology, 12, 555-598. doi: 10.1017/S0954579400004028

Horn, J. L. (1965). A rationale and test for the number of factors in factor analysis.

Psychometrika, 30, 179-185. doi: 10.1007/BF02289447

Hosterman, S. J., DuPaul, G. J., & Jitendra, A. K. (2008). Teacher ratings of ADHD symptoms

in ethnic minority students: Bias or behavioral difference? School Psychology Quarterly,

23, 418-435. doi: 10.1037/a0012668

Hu, L. T., & Bentler, P. M. (1995). Evaluating model fit. In Hoyle, R. H. (Ed.), Structural

equation modeling: Concepts, issues, and applications (pp. 76-99).Thousand Oaks, CA:

Sage Publications.

Page 167: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

157

International Center for Transitional Justice (2008). Nepali voices: Perceptions of truth, justice,

reconciliation, reparation, and the transition in Nepal. Retrieved from

http://www.ictj.org

International Crisis Group (2007). Nepal’s troubled Terai region. Retrieved from:

http://himalaya.socanth.cam.ac.uk

Ivanova, M. Y., Achenbach, T. M., Dumenci, L., Rescorla, L. A., Almgvist, F., Weintraub, S.,

…Verhulst, F. C. (2007). Testing the 8-syndrome structure of the Child Behavior

Checklist in 30 societies. Journal of Clinical Child & Adolescent Psychology, 36, 405-

417. doi: 10.1080/15374410701444363

Jackson, D. L., Purc-Stevenson, R., & Gillaspy, J. A. (2009). Reporting practices in confirmatory

factor analysis: An overview and some recommendations. Psychological Methods, 14, 6-

23. doi: 10.1037/a0014694

Jensen, P. S., Hinshaw, S. P., Swanson, J. M., Greenhill, L. L., Conners, C. K., Arnold, L., …

Widal, T. (2001). Findings from the NIMH multimodal treatment study of ADHD

(MTA): Implications and applications for primary care providers. Journal of

Developmental and Behavioral Pediatrics, 22, 60-73.

Jha, A. (2007). Nepalese psychiatrists struggle for evolution. Psychiatric Bulletin, 31, 348-350.

doi: 10.1192/pb.bp.107.014571

Johnson, C., & Murray, C. (2003). Incremental validity in psychological assessment of children

and adolescents. Psychological Assessment, 15, 496-507. doi: 10.1037/1040-

3590.15.4.496

Page 168: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

158

Johnson, K. A., Wiersma, J. R., & Kunsi, J. (2009). What would Karl Popper say? Are current

theories of ADHD falsifiable? Behavior and Brain Functions, 5, 15-26. doi:

10.1186/1744-9081-5-15

Johnston, C., & Mash, E. J. (2001). Families of children with attention-deficit/hyperactivity

disorder: Review and recommendations for future research. Clinical Child and Family

Psychology Review, 4, 183-207. doi: 10.1023/A:1017592030434

Jones, P. S., Lee, J. W., Phillips, L. R., Zhang, X. E., & Jaceldo, K. B. (2001). An adaptation of

Brislin’s translation model for cross-cultural research. Nursing Research, 50, 300-304.

doi: 10.1097/00006199-200109000-00008

Kaiser, H. F. (1974). An index of factorial simplicity. Psychometrika, 39, 31-36. doi:

10.1007/BF02291575

Kline, R. B. (2005). Principles and practice of structural equation modeling. New York, NY:

The Guilford Press.

Kohrt, B. A., & Harper, I. (2008). Navigating diagnoses: Understanding mind-body relations,

mental health, and stigma in Nepal. Culture, Medicine, and Psychiatry, 32, 1573-1576.

doi: 10.1007/s11013-008-9110-6

Koonce, D. A. (2007). Attention deficit hyperactivity disorder assessment practices by practicing

school psychologists. Journal of Psychoeducational Assessment, 25, 319-333. doi:

10.1177/0734282906298264

Kumar, G., & Steer, R. A. (2003). Factorial validity of the Conners’ Parent Rating Scale–

Revised: Short Form with psychiatric outpatients. Journal of Personality Assessment, 80,

252-259. doi: 10.1207/S15327752JPA8003_04

Page 169: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

159

Lahey, B. B., Applegate, B., McBurnett, K., Biederman, J., Greenhill, L., Hynd, G. W., ...

Richters, J. (1994). DSM-IV field trials for attention deficit/hyperactivity disorder in

children and adolescents. American Journal of Psychiatry, 151, 1,673-1,685.

Lambert, N. M. (1978). Challenges to testing – Larry P.: A case in point. Paper presented at the

annual meeting of the American Psychological Association, Toronto, Ontario. Retrived

from www. eric.ed.gov.

Larry P. v. Riles, 793 F. 2d 969. (9th Cir. 1984).

Leech, N. L., Barrett, K. C., & Morgan, G. A. (2005). SPSS for intermediate statistics: Uses and

interpretation (2nd ed.). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.

Lei, P. W., & Wu, Q. (2007). Introduction to structural equation modeling: Issues and practical

considerations. Educational Measurement: Issues and Practice, 26, 33-43. doi:

10.1111/j.1745-3992.2007.00099.x

Leung, T. W., & Wong, M. M. (2003). Measures of child and adolescent psychopathology in

Asia. Psychological Assessment, 15, 268-279. doi: 10.1037/1040-3590.15.3.268

Levy, R., & Hancock, G. R. (2007). A framework of statistical tests for comparing mean and

covariance structure models. Multivariate Behavioral Research, 42, 33-66. doi:

10.1080/00273170701329112

Lewis, P. M. (2009). Ethnologue: Languages of the world (16th ed.; Online version). Retrieved

from http://www.ethnologue.com.

Library of Congress (2010). A country study: Nepal. Retrieved from

http://lcweb2.loc.gov/frd/cs/nptoc.html.

Page 170: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

160

Loughran, S. B. (2003). Agreement and stability of teacher rating scales for assessing ADHD in

preschoolers. Early Childhood Education Journal, 30, 247-253. doi:

10.1023/A:1023391708850

Louw, J. (1997). Social context and psychological testing in South Africa. Theory and

Psychology, 7, 235-256. doi: 10.1177/0959354397072006

MacCallum, R. C., Wegener, D. T., Uchino, B. N., & Fabrigar, L. R. (1993). The problem of

equivalent models in applications of covariance structure analysis. Psychological

Bulletin, 114, 185-199. doi: 10.1037/0033-2909.114.1.185

MacCallum, R. C., Widaman, K. F., Preacher, K. J., & Hong, S. (2001). Sample size in factor

analysis: The role of model error. Multivariate Behavioral Research, 36, 611-637. doi:

10.1207/S15327906MBR3604_06

Maneesriwongul, W., & Dixon, J. K. (2004). Instrument translation process: Methods review.

Journal of Advanced Nursing, 48, 175-186. doi: 10.1111/j.1365-2648.2004.03185.x

Marco, R., Miranda, A., Schlotz, W., Melia, A., Mulligan, A., Muller, U., ... Sonuga-Barke, E. J.

(2009). Delay and reward choice in ADHD: An experimental test of the role of delay

aversion. Neuropsychology, 23, 367-380. doi: 10.1037/a0014914

Markland, D. (2006). The golden rule is that there are no golden rules: A commentary on Paul

Barrett’s recommendations for reporting model fit in structural equation modeling.

Personality and Individual Differences, 42, 851-858. doi: 10.1016/j.paid.2006.09.023

Maslak, M. A. (2003). Daughters of Tharu: Ethnicity, religion, and the education of Nepali girls.

New York, NY: Routledge Falmer.

Messick, S. (1980). Test validity and the ethics of assessment. Diagnostica, 28, 1-25. doi:

10.1037/0003-066X.35.11.1012

Page 171: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

161

Miller, M. L., Fee, V. E., & Netterville, A. K. (2004). Psychometric properties of ADHD rating

scales for children with mental retardation I: Reliability. Research in Developmental

Disabilities, 25, 459-476. doi:10.1016/j.ridd.2003.11.003

Miller, T. W., Nigg, J. T., & Miller, R. L. (2009). Attention Deficit Hyperactivity Disorder in

African American children: What can be concluded from the past ten years? Clinical

Psychology Review, 29, 77-86. doi: 10.1016/j.cpr.2008.10.001

Minow, M. L. (2001). Limited English proficient students and special education. Retrieved from

http://www.cast.org/publications/ncac/ncac_limited.html

Mitsis, E. M., McKay, K. E., Schulz, K. P., Newcorn, J. H., & Halperin, J. M. (2000). Parent-

teacher concordance for DSM-IV Attention Deficit Hyperactivity Disorder in a clinic-

referred sample. Journal of the American Academy of Child and Adolescent Psychiatry,

39, 308-313. doi: 10.1097/00004583-200003000-00012

Munroe, R. L., Hulefeld, R., Rodgers, J. M., Tomeo, D. L., & Yamazaki, S. K. (2000).

Aggression among children in four cultures. Cross-Cultural Research, 34, 3-25. doi:

10.1177/106939710003400101

Nastasi, B. K., Varjas, J., Bernstein, R., & Jayasena, A. (2000). Conducting participatory culture-

specific consultation: A global perspective on multicultural consultation. School

Psychology Review, 29, 401-413. doi: 61076457

Nigg, J. T. (2005). Attention, task difficulty, and ADHD. British Journal of Developmental

Psychology, 23, 513-516. doi: 10.1348/026151005X55848

Nigg, J. T., Willcutt, E. G., Doyle, A. E., & Songua-Barke, J. S. (2005). Causal heterogeneity in

Attention Deficit Hyperactivity Disorder: Do we need neuropsychological impaired

subtypes? Biological Psychiatry, 57, 1,224-1,230. doi: 10.1016/j.biopsych.2004.08.025

Page 172: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

162

Niraula, B. B., & Morgan, S. P. (1996). Marriage formation, post-marital contact with natal kin,

and autonomy of women: Evidence from two Nepali settings. Population Studies, 50, 36-

50. doi: 10.1080/0032472031000149036

Niraula, T. (2007). Schooling in Nepal. In A. Gupta (Eds.), Going to school in South Asia. (pp.

126–142). Westport, CT: Greenwood Press.

O’Connell, J. (2006). Disproportionality and IDEA 2004: Disproportionate representation of

racial and ethnic groups in special education. Retrieved from http://www.casponline.org.

Okazaki, S., & Sue, S. (1995). Methodological issues with assessment research in ethnic

minorities. Psychological Assessment, 7, 367-375. doi: 10.1037/1040-3590.7.3.367

Pach, A. (1998). Narrative constructions of madness in a Hindu village of Nepal. In D. Skinner,

A. Pach, & D. C. Holland (Eds.), Selves in time and place: Identities, experience, and

history in Nepal. (pp. 111-125). Lanham, MD: Rowan & Littlefield, Inc.

Paloyelis, Y., Asherton, P., & Kuntsi, J. (2009). Are ADHD symptoms associated with delay

aversion or choice impulsivity? A general population study. Journal of the Academy of

Child and Adolescent Psychiatry, 48, 837-846. doi: 10.1097/CHI.0b013e3181ab8c97

Pandey, A. (2006). Request to allocate the Maithili script in the Unicode roadmap. Retrieved

from http://en.wikipedia.org/wiki/Maithili_language.

Pedhazur, E. J., & Schmelkin, L. P. (1991). Measurement, design, and analysis: An integrated

approach. Hillsdale, NJ: Lawrence Erlbaum.

Pelham, W. E., Fabiano, G. A., & Massetti, G. M. (2005). Evidence-based assessment of

attention deficit hyperactivity disorder in children and adolescents. Journal of Clinical

Child and Adolescent Psychology, 34, 499-476. doi: 10.1207/s15374424jccp3403_5

Page 173: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

163

Polanczyk, G., de Lima, M. S., Horta, B. L., Biederman, J., & Rohde, L. A. (2007). The

worldwide prevalence of ADHD: A systematic review and metaregression analysis. The

American Journal of Psychiatry, 164, 942-949. doi: 10.1176/appi.ajp.164.6.942

Polanczyk, G., & Rohde, L. A. (2007). “Psychiatric disorders as social constructs: ADHD as

case in point”: Drs. Polancyzk and Rohde reply. The American Journal of Psychiatry,

164, 1,612-1,613. doi: 10.1176/appi.ajp.2007.07060942r

Poudel, P., & Carryer, J. (2000). Girl-trafficking, HIV/AIDS, and the position of women in

Nepal. Gender and Development, 8, 74-79. doi: 10.1080/741923626

Power, T. J. (2008). Editorial note: Promoting social justice. School Psychology Review, 37, 451-

452. doi: 1638374721

Preacher, K. J., & MacCallum, R. C. (2003). Repairing Tom Swift’s electric factor analysis

machine. Understanding Statistics, 2, 13-43. doi: 10.1207/S15328031US0201_02

Puig, M., Lambert, M. C., Rowan, G. T., Winfrey, T., Lyubansky, M., Hannah, S. D., & Hill, M.

F. (1999). Behavioral and emotional problems among Jamaican and African American

children, ages 6 to 11: Teacher reports versus direct observations. Journal of Emotional

and Behavioral Disorders, 7, 240-250. doi: 10.1177/106342669900700406

Quinn, P. O. (2005). Treating adolescent girls and women with ADHD: Gender specific issues.

Journal of Clinical Psychology, 61, 579-587. doi: 10.1002/jclp.20121

Ramirez, R. D., & Shapiro, E. S. (2005). Effects of student ethnicity on judgments of ADHD

symptoms among Hispanic and White teachers. School Psychology Quarterly, 20, 268-

287. doi: 10.1023/A:1021934705392

Reddy, A. (2008). The eugenic origins of IQ testing: Implications for post-Atkins litigation.

DePaul Law Review. Retrieved from: http://www.law.depaul.edu

Page 174: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

164

Robertson, I. H., Ward, A., Ridgeway, V., & Nimmo-Smith, I. (1994). Test of everyday

attention. Bury St Edmunds, UK: Thames Valley Test Company

Robinson-Pant, A. (2001). Women’s literacy and health: Can an ethnographic researcher find the

links? In B. V. Street (Eds.), Literacy and development: Ethnographic perspectives (152-

171). New York, NY: Routledge, Taylor & Francis Group.

Roth, P. L., & Switzer, F. S. (1999). Missing data: Instrument-level heffalumps and item-level

woozles. Retrieved from: http://www.aom.pace.edu

Rothchild, J. (2006). Gender trouble makers: Education and empowerment in Nepal. New York,

NY: Taylor and Francis Group.

Rucklidge, J. (2008). Gender differences in ADHD: Implications for psychosocial treatments.

Expert Review of Neurotherapeutics, 8, 643-655. doi: 10.1586/14737175.8.4.643

Sagvolden, T., Johansen, E. B., Aase, H., & Russell, V. A. (2005). A dynamic developmental

theory of Attention Deficit Hyperactivity Disorder predominately hyperactive/impulsive

and combined subtypes. Behavioral and Brain Sciences, 28, 397-468. doi:

10.1017/S0140525X05000075

Salvia, J., & Ysseldyke, J. E. (2009). Assessment: In special and inclusive education (10th ed.)

New York, NY: Houghton Mifflin.

Satorra, A., & Bentler, P. M. (2001). A scaled difference chi-square test statistic for moment

structure analysis. Psychometrika, 66, 507-514. doi 10.1007/BF02296192

Sattler, J. M. (2001). Assessment of children: Cognitive applications (4th ed.). San Diego, CA:

Jerome M. Sattler, Publisher, Inc.

Page 175: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

165

Schmitz, M. F., & Velez, M. (2003). Latino cultural differences in maternal assessments of

attention deficit hyperactivity symptoms in children. Hispanic Journal of Behavioral

Sciences, 25, 110-122. doi: 10.1177/0739986303251700

Shaffer, D., Fisher, P., Lucas, C. P., Dulcan, M. K., & Schwab-Stone, M. E. (2000). National

institute of mental health Diagnostic Interview Schedule for Children: Version IV. New

York, NY: DISC Development Group.

Shah, A. (2010). Poverty facts and stats. Retrieved from http://www.globalissues

Snyder, S. M., Drozd, J. F., & Xenakis, S. N. (2004). Validation of ADHD rating scales. Journal

of the American Academy of Child and Adolescent Psychiatry, 43, 1,189-1,190. doi:

10.1097/01.chi.0000135631.10911.fd

Solanto, M. V, Abikoff, H., Sonuga-Barke, E., Schachar, R., Logan, G. D., Wigal, T., …Turkel,

E. (2004). The ecological validity of delay aversion and response inhibition as measures

of impulsivity in ADHD: A supplement to the NIMH multimodal treatment study of

ADHD. Journal of Abnormal Child Psychology, 29, 1573-2835. doi:

10.1023/A:1010329714819

Sonuga-Barke, E. J. (1994). On dysfunction and function in psychological accounts of childhood

disorder. Journal of Child Psychology and Psychaitry, 42, 199-210. doi: 10.1111/j.1469-

7610.1994.tb02296.x

Sonuga-Barke, E. J. (2005). Causal models of attention deficit hyperactivity disorder: From

common simple deficits to multiple developmental pathways. Biological Psychiatry, 57,

1,231-1,238. doi: 10.1016/j.biopsych.2004.09.008

Page 176: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

166

Sonuga-Barke, E. J., Taylor, S., Sembi, E., & Smith, J. (1992). Hyperactivity and delay

aversion–The effect of delay on choice. Journal of Child Psychology and Psychiatry, 33,

387-392.

Stash, S., & Hannum, E. (2001). Who goes to school? Education stratification by gender caste, &

ethnicity in Nepal. Comparative Education Review, 45, 354-378. doi: 10.1086/447676

Stevens, J. (2002). Applied multivariate statistics for the social sciences. Mahwah, NJ: Lawrence

Erlbaum Associates, Inc.

Streiner, D. L. (1994). Figuring out factors: The use and misuse of factor analysis. Canadian

Journal of Psychiatry, 39, 135-140.

Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics (3rd ed.). New York, NY:

Harper Collins.

Tanaka, J. S. (1993). Multifaceted conceptions of fit in structural equation models. In K. A.

Bollen, & J. S. Long (Eds.), Testing structural equation models (pp. 10-39). Newbury

Park, CA: Sage.

Tataryn, D. J., Wood, J. M., & Gorsuch, R. L. (1999). Setting the value of k in promax: A Monte

Carlo study. Educational and Psychological Measurement, 59, 384-391. doi:

10.1177/00131649921969938

Thurstone, L. L. (1947). Multiple factor analysis. Chicago, IL: University of Chicago.

Tielsch, J., Stoltzfus, R., Katz, J., West, K. P., Black, R., LeClerq, S., et al. (2006). NNIPS-4:

Impact of zinc and/or iron-folic acid supplementation on child mortality, southern Nepal.

Retrieved from:http://www.jhsph.edu/GRA/Research/Micronutrient_Dietary_

Interventions

Page 177: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

167

Timimi, S., & Taylor, E. (2004). ADHD is best understood as a cultural construct. Psychiatry,

184(1), 8-9.

Tol, W. A., Jordans, M. J., Regmi, S., & Sharma, B. (2005). Cultural challenges to psychosocial

counseling in Nepal. Transcultural Psychiatry, 42, 317-335. doi:

10.1177/1363461505052670

Tomarken, A. J., & Waller, N. G. (2005). Structural equation modeling: Strengths, limitations,

and misconceptions. Annual Review of Clinical Psychology, 1, 31-65. doi:

10.1146/annurev.clinpsy.1.102803.144239

Tripp, G., Schaughency, E. A., & Bronwyn, C. (2006). Parent and teacher ratings in the

evaluation of attention deficit hyperactivity disorder: Contribution to diagnosis and

differential diagnosis in clinically referred children. Journal of Developmental and

Behavioral Pediatrics, 27, 209-218. doi: 10.1097/00004703-200606000-00006

United Nations. (2010). United Nations mission in Nepal. Retrieved from http://www.unmin.org

United Nations Children’s Fund (UNICEF; 2004). Violence against children in Nepal: A study of

the system of school discipline in Nepal. Retrieved from www.unicef.org

United States Department of State (2010). Background note: Nepal. Retrieved from

http://www.state.gov

Van Ommeren, M., Komproe, I., Cardeña, E., Thapa, S. B., Praisain, D., de Jong, J. T., &

Sharma, B. (2004). Mental illness among Bhutanese Shamans in Nepal. The Journal of

Nervous and Mental Disease, 192, 313–317.

Velicer, W. F. (1976). Determining the number of components from the matrix of partial

correlations. Psychometrika, 41, 321-327. doi: 10.1007/BF02293557

Page 178: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

168

Velicer, W. F., & Fava, J. L. (1998). Effects of variable and subject sampling on factor pattern

recovery. Psychological Methods, 3, 231-251. doi: 10.1037/1082-989X.3.2.231

Waite, R., & Ivey, N. (2009). Promoting culturally sensitive ADHD services for women: An

individual example and call for action. Journal of Psychosocial Nursing and Mental

Health Services, 47, 26-33.

Weiss, G., & Hechtman, L. (1979). The hyperactive child syndrome. Science, 205, 1,348-1,354.

doi: 10.1126/science.472752

Whaley, A. L., & Gellar, P. A. (2007). Toward a cognitive process model of ethnic/ racial biases

in clinical judgment. Review of General Psychology, 11, 75-96. doi: 10.1037/1089-

2680.11.1.75

Willcutt, E. G., Doyle, A. E., Nigg, J. T., Faraone, S. V., & Pennington, B. F. (2005). Validity of

the executive function theory of attention-deficit/hyperactivity disorder: A meta-analytic

review. Biological psychiatry, 57, 1136-1147. doi: 10.1016/j.biopsych.2005.02.006

World Bank (2008). Midyear estimates of Nepal resident population. Retrieved from

http://www.google.com/publicdata

Worthington, R. L., & Whittaker, T. A. (2006). Scale development research: A content analysis

and recommendations for best practices. The Counseling Psychologist, 34, 806-838. doi:

10.1177/0011000006288127

Zwick, W. R., & Velicer, W. F. (1986). Comparison of five rules for determining the number of

components to retain. Psychological Bulletin, 99, 432-442. doi: 10.1037/0033-

2909.99.3.432

Page 179: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

169

Appendix A

Back Translation Evaluation Form

Directions: The Conners’ Parent and Teacher Rating Scales are designed to assess behaviors associated with ADHD and other forms of psychopathology. On these scales, parents and teachers are presented with items and instructed to determine the extent to which the stated behavior is true for the respective child and circle the corresponding response. Response options include: “Not true at all (Never, Seldom)”, “Just a little true (Occasionally)”, “Pretty much true (Often, quite a bit)”, and “Very much true (Very often, very frequently)”. Read the description of each scale. Then, read each original item and back translated item and complete the corresponding questions by clicking the box next to the response you feel is most appropriate. Please check only one box per item, answer every item, and do not leave any items blank. If you have any questions or comments regarding the items, feel free to email the primary investigator at [email protected]. ______________________________________________________________________________ Original Subscale: Cognitive Problems/ Inattention Conners (1997) reported the following regarding the Cognitive Problems/ Inattention subscale: “High scores may be inattentive. They may have more academic difficulties than most individuals their age, have problems organizing their work, have difficulty completing tasks or schoolwork, and appear to have trouble concentrating on tasks that require sustained mental effort.” Original Item: Needs close supervision when completing assignments Back Translation: Need strong supervision to do work

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, cognitive problems/ inattention?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 180: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

170

Original Item: Cannot grasp math Back Translation: Cannot understand math

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, cognitive problems/ inattention?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Fails to complete assignments Back Translation: Not complete assigned work

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, cognitive problems/ inattention?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Difficulty working on/ finishing homework Back Translation: Have difficult doing or completing homework

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, cognitive problems/ inattention?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 181: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

171

Original Item: Has messy handwriting Back Translation: Write alphabets carelessly and untidily

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, cognitive problems/ inattention?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Not reading well Back Translation: Cannot study according to his/her potential

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, cognitive problems/ inattention?

Very much Somewhat Neutral/Unsure Only a little Not at all

Original Item: Poor in math Back Translation: Is weak in math

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, cognitive problems/ inattention?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 182: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

172

Original Item: Poor spelling Back Translation: Weak in spelling

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, cognitive problems/ inattention?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Lack of interest in school Back Translation: Is not interested in studies

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, cognitive problems/ inattention?

Very much Somewhat Neutral/Unsure Only a little Not at all ______________________________________________________________________________ Original Subscale: Anxious-Shy Conners (1997) reported the following regarding the Anxious-Shy subscale: “High scorers generally have more worries and fears than most individuals their age – they are prone to be emotional, are very sensitive to criticism, are particularly anxious in new situations, and appear to be very shy and withdrawn.”

Page 183: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

173

Original Item: Has many fears Back Translation: Is afraid of many things

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, anxious-shy?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Afraid of being alone Back Translation: Afraid to be alone

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, anxious-shy?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Timid, withdrawn Back Translation: Shy, keep quiet

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, anxious-shy?

Very much Somewhat Neutral/Unsure Only a little Not at all ______________________________________________________________________________

Page 184: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

174

Original Scale: Perfectionism Conners (1997) reported the following regarding the Perfectionism subscale: “High scorers are likely to set high goals for themselves, are very fastidious about the way they do things at home or at school, and may be more obsessive about their work or tasks than most individuals their age.” Original Item: Gets upset when someone rearranges belongings Back Translation: Gets angry when other’s rearrange things he/she arranged

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, perfectionism?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Perfectionistic Back Translation: Is detail oriented

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, perfectionism?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 185: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

175

Original Item: Checks things repeatedly Back Translation: Regularly checks work he/she completed

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, perfectionism?

Very much Somewhat Neutral/Unsure Only a little Not at all ______________________________________________________________________________ Original Scale: Social Problems Conners (1997) reported the following regarding the Social Problems subscale: “High scorers are likely to perceive that they have few friends, are likely to have low self-esteem and little confidence, and will likely feel more socially detached from their peers than most individuals their age.” Original Item: Loses friends quickly Back Translation: Break relationship with friends quickly

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, social problems?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 186: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

176

Original Item: Is chosen last for teams/ games Back Translation: Get picked last for teams and games

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, social problems?

Very much Somewhat Neutral/Unsure Only a little Not at all ______________________________________________________________________________ Original Scale: Hyperactivity Conners (1997) reported the following regarding the Hyperactivity subscale: “High scorers have difficulty sitting still, feel more restless and impulsive than most individuals their age, and have the need to always be “on the go”. Original Item: Will run between mouthfuls while eating Back Translation: Move around while eating

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, hyperactivity?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 187: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

177

Original Item: Has difficulty waiting (e.g., in lines or awaiting turn in games) Back Translation: While waiting in line, playing or in group situations difficulty waiting for one’s turn

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, hyperactivity?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Has difficulty playing/ participating in leisure quietly Back Translation: Difficulty engaging in peaceful work or play alone

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, hyperactivity?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Restless and squirmy Back Translation: Get restless (In negative sense)

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, hyperactivity?

Very much Somewhat Neutral/Unsure Only a little Not at all ______________________________________________________________________________

Page 188: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

178

Original Scale: Oppositional Conners (1997) reported the following regarding the Oppositional subscale: “Individuals scoring highly on this subscale are likely to break rules, have problems with persons in authority, and are more easily annoyed and angered than most individuals their age.” Original Item: Blames others for own mistakes/ misbehavior Back Translation: Blame others for one’s behaviors and mistakes

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, oppositional?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Fights Back Translation: Has fights

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, oppositional?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 189: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

179

Original Item: Argues with adults Back Translation: Argues with older people

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, oppositional?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Spiteful/ vindictive Back Translation: Knowingly hurt others (take revenge)

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, oppositional?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Acts defiantly Back Translation: Is arrogant

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, oppositional?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 190: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

180

Original Item: Irritable Back Translation: Get irritated

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, oppositional?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Actively defies/ refuses to comply with adults Back Translation: Not listen to elder stubbornly

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, oppositional?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Loses temper Back Translation: Quick to get angry

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, oppositional?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 191: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

181

Original Item: Touchy or easily annoyed Back Translation: Get upset by small things others do

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, oppositional?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Deliberately does things to annoy others Back Translation: Knowingly do things that get others angry

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, oppositional?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Angry/ resentful Back Translation: Be angry and sad

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, oppositional?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 192: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

182

Original Item: Overly bold Back Translation: Impolite behavior

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, oppositional?

Very much Somewhat Neutral/Unsure Only a little Not at all ______________________________________________________________________________ Original Scale: DSM-Inattentive ADHD The DSM-Inattentive ADHD scale is designed to reflect the inattentive symptoms of attention deficit hyperactivity disorder (ADHD) as they are listed in the Diagnostic and Statistical Manual of Mental Disorders – Fourth Edition, Text Revision (DSM-IV TR). Original Item: Has difficulty organizing tasks or activities Back Translation: Has difficulty organizing work

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured on the DSM-Inattentive ADHD scale?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 193: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

183

Original Item: Often distracted by extraneous stimuli Back Translation: Get distracted easily by surrounding situation

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured on the DSM-Inattentive ADHD scale?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Does not seem to listen when spoken to Back Translation: Pretends not to pay attention to others

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured on the DSM-Inattentive ADHD scale?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Has difficulty sustaining attention in tasks or play Back Translation: Has difficulty concentrating while working or playing

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured on the DSM-Inattentive ADHD scale?

Very much Somewhat Neutral/Unsure Only a little Not at all ______________________________________________________________________________

Page 194: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

184

Original Scale: DSM-Hyperactive ADHD The DSM-Hyperactive ADHD scale is designed to reflect the hyperactive symptoms of attention deficit hyperactive disorder (ADHD) as they are listed in the Diagnostic and Statistical Manual of Mental Disorders – Fourth Edition, Text Revision (DSM-IV TR). Original Item: Talks too much Back Translation: Very talkative

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured on the DSM-Hyperactive ADHD scale?

Very much Somewhat Neutral/Unsure Only a little Not at all ______________________________________________________________________________ Original Scale: Conners’ Global Index The Conners’ Global Index, referred to on former versions of the scale as the Hyperactivity index, was designed to be a short, efficient measure of general psychopathology. Original Item: Cries frequently and easily Back Translation: Cry frequently and easily

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured on the Conners’ Global Index?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 195: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

185

Original Item: Wants demands met immediately – easily frustrated Back Translation: Have to deliver things he/she asked for immediately – easily disappointed

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured on the Conners’ Global Index?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Restless/ highly active Back Translation: Very restless and is constantly active

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured on the Conners’ Global Index?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Fidgets Back Translation: Is constantly active

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured on the Conners’ Global Index?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 196: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

186

Original Item: Mood changes quickly/ dramatically Back Translation: Talk quickly and out of habit

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured on the Conners’ Global Index?

Very much Somewhat Neutral/Unsure Only a little Not at all ______________________________________________________________________________ Original Scale: ADHD Index According to Conners (1997), the items on the ADHD Index have repeatedly been shown to be the best set of items for distinguishing children with ADHD from children in nonclinical control groups. Original Item: Inattentive, distractible Back Translation: Cannot concentrate, easily distracted

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured in the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 197: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

187

Original Item: Only pays attention when interested Back Translation: Pay attention only to things of high interest

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured in the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Distractibility/ attention span problem Back Translation: Cannot concentrate or cannot sustain concentration for necessary time

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured in the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Gets distracted when given instructions Back Translation: When others give instruction, easily distracted

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured in the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 198: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

188

Original Item: Easily frustrated in efforts Back Translation: Gets disappointed if he/she needs to put in effort

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured in the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Poor attention span Back Translation: Cannot sustain attention for required amount of time

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured in the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Restless, always on the go Back Translation: Is restless, always moving

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured in the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 199: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

189

Original Item: Messy or disorganized across settings Back Translation: Be unruly and wild at home or school

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured in the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Cannot stay still Back Translation: Cannot sit still

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured in the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all ______________________________________________________________________________ Multiple subscale items Many of the items on the CRS-R are included on more than one subscale. These items are listed below. Please review each original and back-translated item and answer the corresponding question. Refer to subscale descriptions listed in previous sections if necessary.

Page 200: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

190

Original Item: Disturbs peers Back Translation: Upsets other children

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured in the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all

4. To what extend does the back-translated item appear to reflect the intended construct measured in the Conners’ Global Index – Restless/Impulsive?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Avoids/has difficulty with tasks requiring sustained mental effort Back Translation: Does not do things that requires concentration, gets discouraged or has difficulty (for example, class work or home work)

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all 2. To what extent is the intended meaning of the original item reflected in the back-

translated item? Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct,

cognitive problems/ inattention? Very much Somewhat Neutral/Unsure Only a little Not at all

4. To what extent does the back-translated item appear to reflect the intended construct

measured in the DSM-Inattentive ADHD scale? Very much Somewhat Neutral/Unsure Only a little Not at all

Page 201: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

191

Original Item: Fails to finish things he/she starts Back Translation: Cannot complete task

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, cognitive problems/ inattention?

Very much Somewhat Neutral/Unsure Only a little Not at all

4. To what extent does the back-translated item appear to reflect the intended construct measured in the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all

5. To what extend does the back-translated item appear to reflect the intended construct measured in the Conners’ Global Index – Restless/Impulsive?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Forgets material already learned Back Translation: Forgets things already learnt

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, cognitive problems/ inattention?

Very much Somewhat Neutral/Unsure Only a little Not at all

4. To what extent does the back-translated item appear to reflect the intended construct measured in the DSM-Inattentive ADHD scale?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 202: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

192

Original Item: Does not follow through on instruction/ does not finish assignments Back Translation: Cannot follow direction in schoolwork (cannot follow order or cannot understand direction)

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured in the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all

4. To what extent does the back-translated item appear to reflect the intended construct measured in the DSM-Inattentive ADHD scale?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Interrupts/ intrudes on others Back Translation: Get intrusive or get stubborn (intrusive in other’s conversation or play)

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured in the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all

4. To what extent does the back-translated item appear to reflect the intended construct measured in the DSM-Hyperactive ADHD scale?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 203: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

193

Original Item: Impulsive, easily excited Back Translation: Gets excited, anxious

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, hyperactivity?

Very much Somewhat Neutral/Unsure Only a little Not at all

4. To what extent does the back-translated item appear to reflect the intended construct measured in the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Fidgets with hands and feet or squirms in seat Back Translation: Constantly moving arms and legs or twisting body while seated

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured in the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all

4. To what extent does the back-translated item appear to reflect the intended construct measured in the DSM-Hyperactive ADHD scale?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 204: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

194

Original Item: Temper outbursts Back Translation: Get angry quickly, change behavior uncharacteristically

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, Oppositional?

Very much Somewhat Neutral/Unsure Only a little Not at all

4. To what extent does the back-translated item appear to reflect the intended construct measured in the Conners’ Global Index – Emotional Lability scale?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Fails to attend to details/ makes careless mistakes Back Translation: Cannot concentrate or make careless mistake while completing tasks or in schoolwork

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, cognitive problems/ inattention?

Very much Somewhat Neutral/Unsure Only a little Not at all

4. To what extent does the back-translated item appear to reflect the intended construct measured on the DSM-Inattentive ADHD scale?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 205: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

195

Original Item: Avoids, shows reluctance, or has difficulties with tasks that require sustained mental effort Back Translation: Move away from things that require concentration, gets discouraged or has difficulty (for example, class work or home work)

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, cognitive problems/ inattention?

Very much Somewhat Neutral/Unsure Only a little Not at all

4. To what extent does the back-translated item appear to reflect the intended construct measured on the DSM-Inattentive ADHD scale?

Very much Somewhat Neutral/Unsure Only a little Not at all

5. To what extend does the back-translated item appear to reflect the intended construct measured on the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Forgetful in everyday activities Back Translation: Forget everyday activities (chore)

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, cognitive problems/ inattention?

Very much Somewhat Neutral/Unsure Only a little Not at all

4. To what extent does the back-translated item appear to reflect the intended construct measured on the DSM-Inattentive ADHD scale?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 206: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

196

Original Item: Difficulty concentrating in class Back Translation: Difficulty concentrating at school

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, cognitive problems/ inattention?

Very much Somewhat Neutral/Unsure Only a little Not at all

4. To what extent does the back-translated item appear to reflect the intended construct measured on the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all Original Item: Runs about or climbs excessively and inappropriately Back Translation: Run and climb when not appropriate

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct, hyperactivity?

Very much Somewhat Neutral/Unsure Only a little Not at all

4. To what extent does the back-translated item appear to reflect the intended construct measured on the DSM-Hyperactive ADHD scale?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 207: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

197

Original Item: Leaves seat in class or at other inappropriate times Back Translation: Leave in the middle from class or other situations that require to sit for a while

1. To what extent is the wording of the back-translated item similar to the wording of the original item?

Very much Somewhat Neutral/Unsure Only a little Not at all

2. To what extent is the intended meaning of the original item reflected in the back- translated item?

Very much Somewhat Neutral/Unsure Only a little Not at all

3. To what extent does the back-translated item appear to reflect the intended construct measured on the ADHD Index?

Very much Somewhat Neutral/Unsure Only a little Not at all

4. To what extent does the back-translated item appear to reflect the intended construct measured on the DSM-Hyperactive ADHD scale?

Very much Somewhat Neutral/Unsure Only a little Not at all

Page 208: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

App

endix B

Item

Translation Evaluation

Tabl

e 24

Poo

rly

Tran

slat

ed It

ems

as Id

entif

ied

by E

xper

t Rev

iew

Orig

inal

Item

B

ack-

Tran

slat

ion

Expe

rt Id

entif

ied

Prob

lem

Are

a Sc

ale

Poor

W

ordi

ng

Con

grue

nce

Po

or

Mea

ning

C

ongr

uenc

e

Po

or

Con

stru

ct

Ref

lect

ion

Not

read

ing

wel

l C

anno

t stu

dy a

ccor

ding

to h

is/h

er p

oten

tial

X

X

X

Teac

her

Doe

s not

list

en

Pret

end

not t

o pa

y at

tent

ion

to o

ther

s X

X

X

B

oth

Perf

ectio

nist

istic

Is

det

ail o

rient

ed

X

X

X

Teac

her

Che

cks t

hing

s re

peat

edly

R

egul

arly

che

ck w

ork

he/s

he c

ompl

eted

X

X

X

B

oth

Dis

turb

s pee

rs

Ups

et o

ther

chi

ldre

n X

X

X

B

oth

Impu

lsiv

e/ e

asily

ex

cite

d G

et e

xcite

d, a

nxio

us

X

X

X

Bot

h

Moo

d ch

ange

s qu

ickl

y Ta

lk q

uick

ly a

nd fr

eque

ntly

out

of h

abit

X

X

X

Bot

h

Ove

rly b

old

Impo

lite

beha

vior

X

X

Teac

her

Thin

g m

ust b

e pe

rfec

t N

eeds

to h

ave

ever

ythi

ng h

is/h

er w

ay

X

X

X

Bot

h

198

Page 209: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

Ta

ble

24 (c

ontin

ued)

Orig

inal

Item

B

ack-

Tran

slat

ion

Expe

rt Id

entif

ied

Prob

lem

Are

a Sc

ale

Poor

W

ordi

ng

Con

grue

nce

Po

or

Mea

ning

C

ongr

uenc

e

Po

or

Con

stru

ct

Ref

lect

ion

Easi

ly fr

ustra

ted

G

ets d

isap

poin

ted

if he

/she

nee

ds to

put

in

effo

rt X

X

X

Pa

rent

Mes

sy/ d

isor

gani

zed

acro

ss se

tting

s B

e un

ruly

or w

ild a

t hom

e or

scho

ol

X

X

X

Pare

nt

Touc

hy/ e

asily

an

noye

d

Get

hur

t and

wor

ry b

y sm

all t

hing

s oth

ers

do

X

X

Pa

rent

Ang

ry/ r

esen

tful

Be

angr

y an

d sa

d

X

X

Pa

rent

Diff

icul

ty p

layi

ng

quie

tly

Diff

icul

ty e

ngag

ing

in p

eace

ful w

ork

or

play

alo

ne

X

X

X

Pare

nt

Not

e. O

rigin

al it

ems w

ere

para

phra

sed

slig

htly

in th

is ta

ble.

199

Page 210: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

200

Appendix C

Item-Level Correlation Matrices

Table 25 Item Correlation Matrix of the CTRS-R Scores for EFA Sample (N = 374)

T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11 T12 T13

M .22 .38 .86 .31 .30 1.60 .30 .47 .64 .13 1.25 .66 .49 SD .59 .74 .88 .72 .62 1.14 .62 .77 .86 .44 1.10 .89 .89

T1 - .37 .11 -.02 .20 -.09 .39 .21 .26 .21 -.02 .15 -.01 T2 - .07 -.06 -.01 .02 .23 .26 .21 .23 -.01 .11 -.02 T3 - .14 .27 -.25 .15 .05 .50 .05 -.24 .55 .33 T4 - .13 -.12 .04 -.03 .13 .08 .01 .10 .16 T5 - -.11 .22 .07 .22 .09 -.08 .24 .23 T6 - -.11 -.01 -.30 -.06 .26 -.32 -.13 T7 - .29 .28 .20 -.05 .19 .05 T8 - .11 .24 .04 .05 -.12 T9 - .11 -.20 .65 .29 T10 - .08 .11 .01 T11 - -.24 -.15 T12 - .42

T14 T15 T16 T17 T18 T19 T20 T21 T22 T23 T24 T25 T26 M 1.07 .70 .60 .72 .49 .17 .14 .95 .01 .48 .76 .21 .64 SD .96 .94 .91 .92 .75 .53 .43 1.01 .10 .74 .98 .51 .85

T1 .03 .11 .27 .18 .23 .34 .30 .12 .01 -.01 .01 .12 .17 T2 -.04 .14 .33 .08 .18 .15 .17 .05 -.02 .01 .02 .01 .14 T3 -.09 -.01 -.02 .52 .38 .15 .17 .43 -.04 .21 -.30 .16 .45 T4 .04 .12 .01 .10 .13 .12 .15 .15 -.01 .12 .02 .02 .13 T5 -.01 .07 -.05 .22 .19 .32 .18 .17 -.01 .31 -.07 .31 .20 T6 .10 .04 -.01 -.34 -.22 -.17 -.04 -.23 .06 .03 .19 -.02 -.27 T7 .03 .27 .23 .20 .17 .23 .22 .14 -.05 .10 .01 .07 .22

Page 211: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

201

Table 25 (continued)

T14 T15 T16 T17 T18 T19 T20 T21 T22 T23 T24 T25 T26 T8 .02 .19 .34 .09 .10 .19 .15 -.01 .01 -.04 .01 .06 .10 T9 -.07 .03 .14 .60 .47 .29 .28 .38 -.05 .15 -.22 .16 .54 T10 -.01 .13 .18 .12 .18 .24 .31 .08 -.03 .06 -.02 .05 .24 T11 .05 .05 .17 -.15 -.15 -.03 -.03 -.25 -.05 -.02 .21 .03 -.17 T12 -.05 .04 .08 .59 .48 .29 .25 .42 .01 .22 -.23 .17 .52 T13 .10 -.02 -.05 .35 .30 .10 .11 .25 .06 .31 -.10 .17 .31 T14 - .17 -.04 .02 .04 -.01 -.02 -.04 .10 .04 .18 .03 -.01 T15 - .26 .03 .03 .18 .04 .01 -.08 .05 .15 .05 .07 T16 - .14 .15 .15 .17 .02 -.04 -.03 .10 -.01 .15 T17 - .54 .23 .18 .44 -.05 .23 -.31 .14 .53 T18 - .30 .30 .33 -.03 .14 -.20 .13 .47 T19 - .33 .24 -.03 .11 -.10 .23 .23 T20 - .16 -.03 .08 -.08 .17 .28 T21 - -.05 .19 -.20 .08 .44 T22 - -.03 .05 .01 -.05 T23 - -.07 .25 .20 T24 - .05 -.29 T25 - .15

T27 T28 T29 T30 T31 T32 T33 T34 T35 T36 T37 T38 T39 M .59 .40 .41 .72 .15 1.16 .67 .33 .76 .30 .09 .48 .29 SD .86 .73 .72 .92 .49 1.04 .91 .76 .98 .65 .38 .75 .61

T1 .14 -.03 .06 .15 -.04 .09 .19 .28 .30 .23 .16 .25 .36 T2 .12 .08 .16 .12 -.07 -.01 .07 .34 .43 .27 .03 .29 .41 T3 .39 .31 .02 .43 .15 -.15 -.07 .04 .12 -.06 -.04 .10 .11 T4 .25 .15 .12 .14 .28 -.09 -.01 .03 .01 -.01 .05 -.07 -.02 T5 .20 .20 .05 .22 .18 .07 -.03 .01 .04 -.06 .02 .09 .04 T6 -.24 -.15 .01 -.24 -.02 .15 .02 -.01 -.06 .01 .01 -.06 -.04 T7 .15 .07 .09 .24 .02 -.04 .21 .21 .28 .21 .06 .22 .21 T8 .01 -.05 .16 .08 -.04 .12 .17 .30 .35 .24 .23 .28 .22 T9 .54 .33 .08 .48 .11 -.10 .01 .13 .23 .13 .06 .17 .21

Page 212: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

202

Table 25 (continued)

T27 T28 T29 T30 T31 T32 T33 T34 T35 T36 T37 T38 T39 T10 .19 .13 .12 .05 -.01 -.09 .04 .16 .25 .28 .15 .25 .22 T11 -.15 -.09 .08 -.22 -.09 .16 .05 .13 .11 .09 .09 .04 .02 T12 .47 .33 .09 .50 .13 -.14 -.04 .08 .19 .09 .02 .18 .14 T13 .33 .27 -.01 .25 .21 -.11 -.05 -.02 -.01 -.03 .02 .04 .01 T14 -.04 .03 -.01 .05 .11 .05 .14 -.04 .04 .06 .09 -.04 .11 T15 .03 .07 .16 .01 .01 .01 .17 .23 .16 .20 .16 .21 .21 T16 .12 .03 .17 .03 -.09 .05 .15 .74 .52 .44 .21 .37 .39 T17 .51 .31 .07 .48 .12 -.11 -.03 .12 .20 .09 .03 .18 .19 T18 .42 .29 .13 .41 .13 -.11 -.01 .09 .21 .19 .13 .20 .23 T19 .13 .18 .10 .26 .14 -.01 .04 .19 .21 .21 .30 .23 .20 T20 .19 .12 .21 .22 .04 -.02 -.01 .19 .22 .16 .12 .24 .20 T21 .37 .29 .01 .49 .05 -.06 .01 .02 .09 .11 -.02 .09 .05 T22 -.07 -.06 -.02 -.03 .18 .03 .04 -.05 -.05 -.01 .04 -.07 -.01 T23 .12 .20 .07 .20 .20 .01 .01 -.01 .01 -.01 -.02 .12 -.01 T24 -.19 -.12 -.01 -.27 -.09 .16 .16 .10 -.02 .09 .09 .05 -.05 T25 .12 .10 .01 .11 .12 .01 -.08 .03 .07 .01 .08 .01 -.01 T26 .49 .45 .08 .39 .07 -.16 .01 .15 .18 .11 .04 .15 .20 T27 - .43 .15 .42 .14 -.12 -.04 .08 .14 .13 .04 .10 .18 T28 - .10 .28 .17 -.18 -.05 .08 .09 .13 .06 .20 .15 T29 - .10 .07 .03 .01 .20 .24 .17 .05 .17 .22 T30 - .17 -.11 .01 .07 .15 .05 -.02 .11 .15 T31 - -.02 -.06 -.06 .03 -.05 -.02 -.05 -.03 T32 - .16 .02 .03 -.04 -.01 -.02 .01 T33 - .15 .10 .09 .15 .09 .09 T34 - .44 .40 .15 .39 .36 T35 - .46 .17 .35 .34 T36 - .31 .33 .37 T37 - .23 .23 T38 - .34

Page 213: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

203

Table 25 (continued)

T40 T41 T42 T43 T44 T45 T46 T47 T48 T49 T50 T50 T51 M .42 .41 .29 1.75 .51 .45 .39 .17 .60 .51 1.23 .77 .65 SD .79 .72 .64 1.00 .82 .79 .75 .50 .81 .63 1.07 .92 .85 T1 .12 -.02 .08 -.11 .13 .25 .08 .30 .13 .01 .14 -.03 .09 T2 .08 .03 .19 -.16 .29 .17 .09 .30 .21 .07 .14 -.04 .20 T3 .43 .23 .15 -.24 .06 .09 -.21 .07 .31 .26 .18 .26 .40 T4 .08 .21 .11 -.02 -.01 .02 -.07 -.02 .12 .05 .23 .15 .24 T5 .21 .06 .09 .01 .08 .27 -.01 .03 .14 .06 .11 .30 .12 T6 -.17 -.12 -.03 .25 .03 -.07 .21 -.09 -.22 -.05 -.10 -.18 -.21 T7 .15 .15 .15 -.04 .15 .32 .08 .17 .20 .13 .20 .03 .13 T8 .01 .02 .16 -.04 .19 .21 .16 .27 .12 .04 .06 -.07 .06 T9 .38 .24 .10 -.29 .15 .16 -.13 .18 .40 .24 .24 .25 .41 T10 .10 .17 .24 -.02 .21 .13 .03 .29 .16 .07 .21 .08 .13 T11 -.12 -.10 .03 .17 .14 .03 .21 .02 -.08 -.03 .11 -.05 -.11 T12 .46 .25 .14 -.25 .08 .11 -.13 .19 .40 .22 .17 .30 .42 T13 .29 .28 .06 -.10 -.07 .11 -.11 .02 .18 .07 .17 .34 .25 T14 .04 .13 .07 .08 -.12 .09 .09 .04 -.04 -.08 .04 .08 .03 T15 .05 .15 .18 .05 .13 .33 .13 .14 .08 .01 .26 -.01 .10 T16 -.01 .03 .15 -.01 .32 .21 .25 .30 .20 .04 .21 -.11 .13 T17 .38 .23 .15 -.28 .16 .10 -.17 .12 .36 .19 .26 .18 .43 T18 .38 .24 .10 -.22 .13 .08 -.14 .15 .38 .10 .23 .25 .39 T19 .27 .09 .17 -.18 .09 .20 .02 .19 .24 -.01 .20 .10 .24 T20 .19 .07 .13 -.14 .22 .06 -.08 .26 .22 .12 .20 .04 .23 T21 .34 .29 .18 -.24 -.03 .03 -.20 .04 .35 .14 .19 .26 .38 T22 -.02 -.06 -.05 .03 .03 .11 .15 -.04 -.08 -.04 -.12 -.03 -.05 T23 .24 .20 .09 -.06 .09 .13 -.07 .10 .14 .13 .21 .27 .19 T24 -.16 -.03 -.10 .28 .01 .10 .36 .06 -.14 -.05 .03 -.09 -.14 T25 .14 .10 .07 -.05 .18 .21 .06 .19 .09 .12 .07 .18 .04 T26 .42 .26 .22 -.22 .14 .13 -.15 .18 .48 .12 .22 .24 .46 T27 .37 .27 .20 -.18 .05 .12 -.11 .15 .38 .19 .25 .23 .41 T28 .33 .39 .27 -.15 .15 .14 -.05 .05 .36 .09 .23 .23 .35

Page 214: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

204

Table 25 (continued)

T40 T41 T42 T43 T44 T45 T46 T47 T48 T49 T50 T50 T51 T29 .03 .09 .20 -.03 .22 .12 .12 .18 .15 .10 .17 -.01 .16 T30 .49 .27 .20 -.21 .01 .05 -.20 .16 .35 .22 .20 .25 .46 T31 .17 .19 .16 -.13 .01 .03 -.06 .05 .08 .08 .05 .20 .16 T32 -.15 -.15 -.01 .12 .03 .06 .12 -.01 -.17 -.08 -.05 -.16 -.16 T33 -.01 .06 -.05 .14 -.03 .20 .25 .09 .07 .03 .07 .01 -.01 T34 .04 .01 .18 -.02 .30 .19 .23 .25 .24 .08 .18 -.16 .17 T35 .04 .06 .19 -.16 .30 .19 .11 .38 .22 .09 .19 -.06 .22 T36 .02 .13 .19 .01 .21 .12 .19 .24 .16 .14 .16 -.13 .11 T37 .01 .07 .04 -.01 .09 .04 .15 .19 .03 -.06 .05 -.10 .05 T38 .16 .12 .30 -.01 .30 .06 .08 .29 .20 .06 .19 -.07 .22 T39 .03 .11 .15 -.10 .26 .24 .04 .39 .29 .06 .19 -.07 .14 T40 - .27 .26 -.15 .06 .08 -.12 .10 .34 .19 .19 .23 .51 T41 - .19 -.14 .08 .13 -.01 .16 .19 .09 .21 .25 .28 T42 - -.08 .16 .14 -.02 .12 .16 .01 .25 .04 .19 T43 - -.01 -.01 .21 -.14 -.13 -.08 -.08 -.18 -.21 T44 - .22 .13 .22 .11 .15 .12 -.09 .12 T45 - .22 .21 .17 .04 .15 .10 .17 T46 - .07 -.09 .02 -.01 -.08 -.11 T47 - .22 .12 .18 .03 .13 T48 - .11 .26 .14 .47 T49 - .10 .09 .20 T50 - .16 .23 T51 - .22

T53 T54 T55 T56 T57 T58 T59

M .49 .48 .21 .89 .46 .91 .43 SD .84 .73 .54 1.03 .78 .87 .75

T1 -.05 .30 .38 .13 .09 .15 .29 T2 .01 .23 .36 -.01 .19 .12 .47 T3 .11 .01 .13 .41 .34 .28 .08 T4 .05 .01 .10 .16 .08 .09 -.04

Page 215: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

205

Table 25 (continued)

T53 T54 T55 T56 T57 T58 T59 T5 .05 .05 .05 .23 .25 .20 .01 T6 -.12 .03 -.04 -.25 -.21 -.19 .04 T7 -.06 .22 .38 .12 .05 .18 .32 T8 -.01 .18 .33 .01 .03 .11 .40 T9 .15 .15 .25 .46 .39 .34 .23 T10 -.03 .27 .36 .08 .14 .14 .25 T11 .10 .08 .08 -.19 -.12 -.02 .11 T12 .13 .04 .11 .49 .39 .29 .12 T13 .13 .01 -.08 .31 .28 .16 -.08 T14 -.01 .12 .02 -.05 -.07 -.03 -.05 T15 .07 .22 .18 .02 .02 .02 .24 T16 .04 .27 .39 .06 .12 .14 .57 T17 .16 .12 .17 .43 .38 .32 .13 T18 .09 .14 .22 .34 .41 .24 .25 T19 .04 .17 .26 .20 .31 .20 .25 T20 -.02 .16 .25 .18 .19 .14 .26 T21 .04 .02 .10 .52 .28 .23 .06 T22 -.03 .01 -.04 .01 -.03 -.02 -.02 T23 .04 .03 -.01 .21 .19 .15 .03 T24 -.03 .15 -.03 -.19 -.22 -.12 .03 T25 .06 .09 .01 .10 .14 .18 .04 T26 .11 .11 .14 .44 .34 .21 .20 T27 .11 .14 .15 .41 .38 .27 .09 T28 .08 .13 .07 .31 .39 .16 .13 T29 .02 .26 .23 .01 .09 .17 .26 T30 .11 .06 .17 .54 .31 .38 .15 T31 .09 .07 .02 .09 .13 .05 -.04 T32 -.01 -.03 -.03 -.10 -.09 -.05 -.03 T33 -.11 .14 .15 .08 -.03 -.06 .05 T34 .06 .27 .36 .06 .13 .17 .54

Page 216: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

206

Table 25 (continued)

T53 T54 T55 T56 T57 T58 T59

T35 .04 .28 .50 .03 .19 .19 .51 T36 .03 .34 .35 .06 .10 .13 .42 T37 -.04 .21 .23 .01 .05 .04 .23 T38 -.01 .27 .30 .15 .15 .17 .44 T39 -.05 .26 .37 .10 .23 .18 .42 T40 .06 .08 .05 .41 .33 .22 .10 T41 .05 .07 .17 .32 .17 .12 .11 T42 .13 .18 .20 .09 .16 .12 .25 T43 -.01 .04 -.06 -.19 -.22 -.17 -.11 T44 -.03 .25 .25 -.01 .15 .28 .45 T45 .06 .21 .22 .09 .17 .15 .21 T46 -.02 .18 .16 -.11 -.09 -.06 .20 T47 .04 .22 .39 .08 .22 .21 .38 T48 -.01 .14 .22 .34 .43 .27 .24 T49 -.03 .07 .08 .20 .13 .23 .06 T50 .14 .19 .28 .22 .21 .28 .23 T51 .16 .04 -.06 .27 .17 .16 -.11 T52 .19 .14 .15 .44 .39 .24 .22 T53 - -.06 -.06 .05 .09 .10 .04 T54 - .27 .01 .17 .20 .34 T55 - .04 .13 .19 .56 T56 - .34 .20 .10 T57 - .29 .15 T58 - .26

Page 217: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

207

Table 26

Item Correlation Matrix of the CTRS-R Scores for CFA Sample (N = 1,000)

T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11 T12 T13

M .18 .36 .91 .28 .29 1.65 .22 .45 .63 .14 1.29 .67 .48 SD .48 .71 .94 .70 .61 1.10 .55 .75 .88 .48 1.09 .92 .89

T1 - .32 .09 .08 .19 -.05 .37 .24 .13 .24 -.01 .12 .11 T2 - .19 .02 .16 -.13 .43 .31 .24 .30 -.01 .14 .04 T3 - .09 .19 -.35 .16 .05 .51 .18 -.22 .51 .29 T4 - .08 -.09 .06 .02 .08 .09 -.06 .17 .29 T5 - -.11 .26 .12 .16 .10 -.07 .17 .17 T6 - -.07 -.01 -.34 -.13 .27 -.33 -.17 T7 - .26 .17 .29 -.03 .13 .07 T8 - .17 .20 .05 .02 .01 T9 - .21 -.16 .50 .27 T10 - -.01 .20 .06 T11 - -.21 -.18 T12 - .37 T14 T15 T16 T17 T18 T19 T20 T21 T22 T23 T24 T25 T26 M 1.02 .71 .67 .64 .51 .16 .15 1.05 .04 .52 .79 .20 .62 SD .95 .93 .99 .88 .81 .50 .47 1.06 .29 .83 .95 .53 .86

T1 -.05 .14 .25 .15 .20 .18 .17 .09 .04 .02 -.05 .10 .17 T2 -.09 .16 .45 .24 .23 .29 .26 .12 .04 -.03 -.08 .10 .15 T3 -.14 -.02 .14 .50 .38 .20 .21 .46 .08 .26 -.33 .17 .45 T4 .13 .09 .05 .08 .11 .06 .05 .13 .12 .09 -.01 .11 .16 T5 .01 .08 .10 .14 .20 .16 .12 .13 .05 .29 .04 .36 .16 T6 .13 .06 -.13 -.31 -.25 -.16 -.15 -.34 -.08 -.12 .28 -.12 -.26 T7 -.04 .18 .32 .22 .19 .21 .23 .10 .04 .03 -.06 .17 .18 T8 .02 .17 .37 .12 .17 .11 .16 .01 .01 .09 .05 .11 .08 T9 -.15 .03 .22 .50 .46 .27 .26 .46 .07 .19 -.29 .11 .51 T10 -.04 .09 .23 .18 .18 .31 .30 .15 .07 .01 -.06 .09 .20 T11 .14 .10 .05 -.21 -.12 -.01 -.03 -.22 -.09 -.08 .18 -.08 -.15 T12 -.10 -.02 .12 .51 .43 .24 .22 .49 .10 .23 -.31 .10 .50

Page 218: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

208

Table 26 (continued) T14 T15 T16 T17 T18 T19 T20 T21 T22 T23 T24 T25 T26 T13 .02 .01 .08 .34 .28 .19 .11 .29 .15 .26 -.09 .16 .29 T14 - .15 -.03 -.11 -.07 -.06 -.07 -.09 -.02 .11 .17 .02 -.04 T15 - .18 .01 .06 .09 .05 -.05 .03 .07 .13 .09 .03 T16 - .16 .26 .22 .24 .13 .04 .05 -.05 .07 .18 T17 - .47 .27 .18 .46 .09 .20 -.30 .16 .47 T18 - .35 .22 .36 .12 .23 -.23 .15 .44 T19 - .20 .21 .06 .09 -.13 .15 .25 T20 - .15 .04 .08 -.16 .08 .22 T21 - .02 .28 -.30 .12 .44 T22 - .03 -.01 .08 .03 T23 - -.07 .31 .21 T24 - -.03 -.26 T25 - .13 T26 - T27 T28 T29 T30 T31 T32 T33 T34 T35 T36 T37 T38 T39 M .54 .36 .44 .79 .16 1.09 .66 .34 .75 .16 1.09 .66 .34 SD .84 .68 .77 .97 .53 .99 .92 .77 .97 .53 .99 .92 .77

T1 .17 .03 .18 .09 .02 -.01 .09 .20 .26 .18 .22 .26 .19 T2 .21 .11 .25 .15 .09 -.10 .11 .44 .45 .32 .17 .43 .36 T3 .39 .27 .14 .49 .22 -.21 -.02 .13 .23 .03 .04 .16 .16 T4 .19 .19 .09 .12 .22 -.06 .12 .02 .04 .10 .04 .04 .01 T5 .15 .11 .05 .15 .17 .03 .05 .06 .14 .04 .05 .12 .13 T6 -.29 -.18 -.06 -.34 -.14 .18 -.02 -.13 -.14 -.02 -.01 -.10 -.15 T7 .19 .14 .19 .13 .04 -.08 .09 .32 .33 .21 .13 .24 .24 T8 .05 .09 .30 .07 -.01 .04 .13 .34 .31 .27 .12 .26 .23 T9 .43 .29 .24 .46 .22 -.13 .03 .21 .25 .09 .07 .21 .26 T10 .21 .17 .18 .18 .06 -.09 .05 .24 .28 .17 .14 .24 .22 T11 -.14 -.10 .01 -.20 -.09 .19 .04 .02 -.06 .01 .09 .03 -.03 T12 .48 .37 .19 .47 .22 -.18 .05 .11 .15 .07 .01 .18 .13 T13 .33 .28 .16 .29 .23 -.08 .13 .04 .12 .07 -.03 .03 .07

Page 219: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

209

Table 26 (continued)

T27 T28 T29 T30 T31 T32 T33 T34 T35 T36 T37 T38 T39

T14 -.02 .03 -.04 -.11 .03 .11 .12 -.04 -.08 .03 .01 -.08 -.10 T15 .07 .10 .12 .01 .01 .06 .20 .18 .16 .19 .17 .13 .17 T16 .21 .12 .25 .15 .01 -.06 .15 .67 .52 .42 .20 .42 .37 T17 .46 .29 .20 .51 .23 -.19 .04 .19 .28 .10 .03 .17 .19 T18 .47 .29 .22 .38 .23 -.14 .05 .25 .29 .16 .07 .25 .26 T19 .26 .22 .15 .24 .25 -.12 .02 .23 .30 .16 .13 .20 .27 T20 .18 .07 .22 .22 .17 -.13 .07 .27 .27 .11 .18 .19 .29 T21 .43 .30 .11 .50 .18 -.16 .01 .14 .17 .05 -.02 .10 .15 T22 .06 .12 .03 .07 .28 -.10 .01 .01 .05 .05 -.01 .08 .03 T23 .25 .22 .13 .24 .20 .01 .09 .02 .02 -.05 -.04 .01 .06 T24 -.25 -.17 -.08 -.31 -.08 .27 .18 -.06 -.13 .07 .03 -.09 -.06 T25 .18 .16 .12 .17 .13 .01 .10 .08 .13 .02 .04 .04 .12 T26 .43 .34 .18 .46 .22 -.18 .06 .18 .22 .08 .06 .16 .17 T27 - .47 .27 .41 .22 -.15 .10 .20 .22 .13 .06 .17 .20 T28 - .26 .32 .23 -.19 .03 .13 .16 .14 .04 .13 .14 T29 - .12 .08 -.02 .16 .28 .31 .20 .11 .26 .25 T30 - .23 -.21 .04 .15 .20 .05 .01 .14 .21 T31 - -.09 .01 .05 .07 -.01 -.01 .04 .06 T32 - .16 -.06 -.06 -.07 .02 -.09 -.07 T33 - .21 .10 .17 .07 .12 .08 T34 - .51 .41 .19 .44 .40 T35 - .40 .25 .44 .43 T36 - .25 .30 .30 T37 - .16 .29 T38 - .30 T40 T41 T42 T43 T44 T45 T46 T47 T48 T49 T50 T51 T52

M .75 .31 .08 .44 .26 .45 .49 .28 1.74 .54 .47 .39 .22 SD .97 .66 .34 .74 .59 .83 .82 .63 1.03 .85 .76 .76 .60

T1 .15 .16 .03 -.08 .12 .11 .11 .28 .13 .13 .11 .03 .13 T2 .17 .15 .15 -.12 .35 .11 .09 .40 .18 .18 .15 -.04 .17

Page 220: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

210

Table 26 (continued)

T40 T41 T42 T43 T44 T45 T46 T47 T48 T49 T50 T51 T52

T3 .44 .32 .22 -.25 .23 .09 -.19 .16 .43 .28 .15 .24 .43 T4 .10 .20 .16 -.05 .04 .06 -.02 .05 .16 .02 .16 .11 .21 T5 .13 .13 .05 -.02 .13 .17 -.02 .15 .12 .07 .09 .18 .14 T6 -.27 -.23 -.16 .26 -.13 -.06 .15 -.10 -.23 -.21 -.09 -.16 -.26 T7 .17 .14 .08 -.14 .25 .15 .11 .36 .16 .15 .14 -.02 .17 T8 .04 .05 .08 .01 .23 .22 .22 .24 .09 .11 .17 -.02 .07 T9 .50 .34 .30 -.26 .27 .12 -.09 .20 .42 .29 .23 .21 .46 T10 .19 .12 .11 -.13 .21 .18 .06 .27 .17 .15 .11 .08 .19 T11 -.16 -.10 -.08 .20 .01 .01 .16 -.02 -.11 -.11 .05 -.08 -.18 T12 .51 .40 .25 -.22 .18 .08 -.17 .15 .45 .30 .22 .21 .44 T13 .29 .30 .15 -.12 .15 .11 -.12 .08 .26 .15 .18 .22 .27 T14 -.06 .02 .01 .15 -.10 .11 .14 -.10 -.08 -.08 .06 .05 -.06 T15 .04 .08 .11 .09 .10 .31 .25 .16 .04 .02 .22 .02 .08 T16 .16 .17 .13 -.07 .37 .17 .13 .39 .16 .21 .24 -.01 .18 T17 .49 .37 .24 -.23 .24 .09 -.14 .20 .39 .26 .17 .18 .42 T18 .44 .33 .25 -.17 .30 .12 -.06 .25 .37 .26 .19 .18 .43 T19 .29 .21 .18 -.16 .23 .16 .01 .23 .26 .21 .13 .19 .27 T20 .28 .19 .12 -.13 .25 .13 -.05 .30 .23 .25 .13 .10 .22 T21 .44 .35 .24 -.21 .17 .03 -.20 .13 .37 .35 .17 .26 .39 T22 .11 .09 .09 -.05 .07 -.01 -.06 .01 .11 -.04 .01 .03 .05 T23 .17 .18 .08 -.06 .07 .13 -.05 .04 .19 .15 .13 .41 .23 T24 -.27 -.14 -.15 .28 -.17 .09 .26 -.12 -.23 -.23 -.10 -.08 -.24 T25 .12 .13 .10 -.07 .10 .15 .05 .16 .17 .09 .08 .16 .17 T26 .48 .36 .24 -.24 .18 .12 -.12 .21 .49 .24 .27 .22 .53 T27 .43 .38 .31 -.20 .21 .13 -.11 .21 .32 .30 .23 .26 .43 T28 .28 .27 .33 -.15 .20 .09 -.07 .16 .32 .13 .24 .19 .37 T29 .13 .17 .21 -.09 .27 .23 .08 .23 .18 .23 .24 .09 .20 T30 .49 .37 .25 -.26 .15 .10 -.20 .19 .40 .34 .20 .26 .40 T31 .26 .28 .12 -.20 .14 .10 -.07 .06 .14 .07 .08 .21 .23 T32 -.20 -.11 -.11 .18 -.09 .10 .16 -.05 -.16 -.09 -.07 .03 -.15

Page 221: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

211

Table 26 (continued)

T40 T41 T42 T43 T44 T45 T46 T47 T48 T49 T50 T51 T52

T33 .02 .10 .06 .07 .07 .18 .17 .11 .02 .09 .08 .05 .04 T34 .17 .15 .13 -.09 .38 .20 .15 .41 .16 .24 .18 -.04 .18 T35 .23 .17 .19 -.15 .45 .16 .07 .49 .18 .27 .15 .01 .21 T36 .04 .08 .12 -.01 .26 .15 .19 .28 .09 .10 .16 -.09 .09 T37 .05 .07 .05 -.01 .17 .21 .22 .14 .09 .06 .06 -.07 .03 T38 .19 .19 .18 -.07 .34 .16 .05 .28 .22 .20 .14 -.03 .18 T39 .21 .19 .19 -.11 .37 .17 .05 .38 .14 .20 .10 .01 .15 T40 - .39 .31 -.24 .24 .11 -.15 .21 .43 .31 .20 .23 .44 T41 - .27 -.20 .16 .14 -.07 .14 .34 .19 .18 .19 .36 T42 - -.16 .21 .09 -.03 .18 .24 .12 .20 .11 .29 T43 - -.13 .08 .13 -.15 -.17 -.19 -.02 -.09 -.22 T44 - .16 .04 .34 .18 .23 .19 .06 .24 T45 - .23 .14 .11 .12 .15 .13 .20 T46 - .07 -.10 -.05 .02 -.09 -.09 T47 - .19 .21 .17 .03 .20 T48 - .25 .26 .15 .53 T49 - .17 .15 .26 T50 - .13 .29 T51 - .24

T53 T54 T55 T56 T57 T58 T59

M .25 .48 .25 .93 .44 1.02 .47 SD .62 .74 .62 1.07 .78 .96 .82

T1 .06 .22 .31 .11 .14 .19 .24 T2 .07 .25 .36 .15 .22 .18 .49 T3 .10 .12 .08 .47 .40 .32 .18 T4 .11 .07 .02 .11 .09 .03 .04 T5 .14 .09 .09 .12 .09 .17 .11 T6 -.02 -.05 -.10 -.33 -.27 -.23 -.17 T7 .07 .27 .31 .14 .19 .16 .30 T8 .04 .26 .24 .03 .11 .14 .29

Page 222: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

212

Table 26 (continued) T53 T54 T55 T56 T57 T58 T59

T9 .17 .17 .18 .43 .43 .34 .23 T10 .08 .18 .22 .13 .20 .13 .23 T11 .10 .05 .02 -.22 -.14 -.08 .03 T12 .13 .09 .10 .51 .41 .29 .15 T13 .10 .05 .03 .29 .18 .16 .09 T14 .07 .06 -.04 -.09 -.09 -.11 -.04 T15 .22 .20 .13 .03 .01 .07 .16 T16 .04 .35 .41 .17 .17 .28 .60 T17 .09 .16 .17 .46 .45 .32 .19 T18 .17 .18 .24 .35 .41 .32 .24 T19 .11 .16 .21 .17 .29 .20 .22 T20 .06 .17 .23 .22 .21 .23 .22 T21 .07 .11 .10 .57 .36 .33 .14 T22 .05 -.05 -.03 .07 .06 .04 .02 T23 .20 .10 -.06 .25 .12 .18 .02 T24 -.02 -.04 -.10 -.28 -.25 -.28 -.10 T25 .06 .07 .05 .14 .11 .13 .10 T26 .15 .16 .15 .42 .42 .34 .23 T27 .18 .13 .17 .42 .37 .30 .20 T28 .21 .17 .10 .28 .24 .18 .16 T29 .15 .22 .24 .13 .21 .18 .24 T30 .08 .12 .12 .54 .37 .38 .18 T31 .12 .04 .03 .18 .18 .07 .05 T32 .08 -.03 -.06 -.21 -.16 -.06 -.11 T33 .04 .14 .05 .04 .04 .04 .06 T34 .01 .32 .40 .15 .22 .27 .60 T35 .10 .29 .50 .19 .30 .28 .51 T36 .02 .20 .27 .07 .13 .08 .36 T37 .07 .20 .24 .02 .10 .08 .20 T38 .05 .24 .33 .14 .23 .21 .48

Page 223: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

213

Table 26 (continued) T53 T54 T55 T56 T57 T58 T59

T39 .12 .28 .39 .20 .22 .25 .39 T40 .10 .12 .20 .50 .47 .38 .22 T41 .10 .15 .14 .42 .35 .24 .19 T42 .17 .15 .15 .20 .25 .21 .18 T43 .02 -.02 -.10 -.24 -.23 -.15 -.10 T44 .09 .31 .37 .12 .24 .25 .44 T45 .19 .22 .16 .06 .14 .19 .16 T46 .04 .16 .10 -.19 -.07 -.05 .07 T47 .10 .26 .53 .14 .25 .24 .40 T48 .10 .14 .16 .38 .39 .32 .21 T49 .01 .18 .24 .35 .31 .29 .21 T50 .21 .19 .14 .19 .15 .25 .20 T51 .18 .06 -.06 .24 .16 .16 -.03 T52 .19 .16 .18 .39 .41 .36 .21 T53 - .01 .14 .09 .08 .14 .09 T54 - .28 .16 .21 .20 .37 T55 - .12 .29 .21 .46 T56 - .42 .36 .18 T57 - .34 .21 T58 - .24 T59 -

Page 224: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

214

Table 27

Item Correlation Matrix of the CPRS-R Scores for EFA Sample (N = 555)

P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 P13 P14 P15 P16 P17

M 1.44 .83 1.53 .61 1.05 .01 .45 1.43 .79 .51 .86 .91 .37 .29 .48 .86 .34

SD 1.05 1.10 1.19 .90 .99 .10 .71 1.03 1.04 .87 .99 1.07 .77 .67 .80 .95 .65

P1 - .16 -.03 -.02 .19 -.06 .04 .37 .15 .11 .31 .24 .15 .03 -.04 .22 .11

P2 - -.17 .01 .12 -.06 .01 .22 .57 .22 .20 .34 .17 .04 -.17 .19 .09

P3 - -.06 .07 .01 -.14 -.10 -.14 -.18 -.17 -.06 -.07 .04 -.08 .01 -.15

P4 - .01 -.02 .13 .01 .05 .05 -.07 -.05 .01 .31 .07 -.08 .18

P5 - -.06 -.01 .15 .11 .05 .08 .03 .05 -.03 .07 .12 -.04

P6 - -.01 -.08 -.02 -.06 -.03 -.03 -.05 -.01 -.03 -.09 -.02

P7 - .09 .04 .02 .11 .12 .05 .13 -.08 .07 .37

P8 - .20 .17 .35 .24 .23 -.02 -.08 .35 .18

P9 - .25 .19 .41 .13 .03 -.20 .17 .07

P10 - .15 .17 .16 .02 -.05 .18 .11

P11 - .27 .19 .01 -.03 .33 .08

P12 - .18 -.03 -.26 .22 .14

P13 - -.04 -.13 .25 .06

P14 - .06 .03 .14

P15 - -.11 -.03

P16 - .06

P18 P19 P20 P21 P22 P23 P24 P25 P26 P27 P28 P29 P30 P31 P32 P33 P34 1.58 .37 1.05 1.27 1.13 1.00 .28 1.07 .11 .19 .80 .62 .52 .92 .76 .72 .48 1.11 .81 1.02 1.01 1.11 1.01 .64 1.08 .47 .53 .92 .93 .88 1.01 .94 1.05 .83 P1 .14 .06 .14 .43 .10 .24 .11 -.10 .03 .12 .13 .18 .20 .35 .26 .11 .05 P2 .10 .42 .21 .14 .24 .14 .08 -.15 .10 .17 .08 .39 .34 .18 .22 .13 .01 P3 .11 -.05 -.10 -.11 .11 -.08 -.06 .08 -.05 -.14 -.07 -.08 -.15 -.16 -.05 -.03 .13 P4 -.06 .01 -.04 .02 .03 -.05 .26 .09 .06 .01 .10 -.04 .02 .04 .05 .28 .08 P5 .19 .11 .09 .18 .19 .06 .01 .14 .01 .03 .08 .07 .07 .15 .10 .09 .16

Page 225: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

215

Table 27 (continued) P18 P19 P20 P21 P22 P23 P24 P25 P26 P27 P28 P29 P30 P31 P32 P33 P34

P6 -.09 -.04 -.06 -.05 -.03 -.07 -.04 -.03 .02 -.04 -.07 -.02 .01 -.07 -.06 -.01 -.06

P7 -.01 -.01 .06 .08 -.03 .01 .14 -.02 -.01 .29 .08 .05 .09 .09 -.02 .07 .02

P8 .17 .11 .18 .35 .15 .32 .10 -.19 .05 .15 .21 .21 .23 .40 .30 .10 .02

P9 .06 .35 .27 .17 .22 .13 .01 -.11 .04 .19 .11 .34 .29 .16 .22 .10 .04

P10 .11 .26 .22 .19 .10 .13 .01 -.03 .04 .11 .19 .15 .35 .11 .27 .09 .09

P11 .15 .07 .19 .31 .05 .24 .02 -.13 .04 .16 .19 .22 .26 .28 .20 .04 .06

P12 .03 .15 .33 .22 .21 .21 .05 -.21 .09 .21 .12 .38 .30 .24 .22 .05 -.02

P13 .06 .09 .21 .16 .13 .24 .01 -.11 .02 .16 .08 .10 .23 .20 .18 .10 -.05

P14 -.04 .03 -.03 .06 -.01 .01 .34 -.04 .01 .03 -.01 -.02 .08 .06 .01 .22 .03

P15 -.05 -.06 -.12 .03 -.10 -.03 .04 .13 .02 -.08 -.07 -.17 -.15 -.03 -.13 .05 .01

P16 .17 .08 .27 .22 .16 .21 .02 -.11 -.03 .11 .14 .15 .19 .26 .29 -.06 .04

P17 .06 .08 .02 .15 .03 .10 .13 .01 .06 .18 .16 .07 .13 .16 .14 .12 -.07

P18 - .09 .14 .15 .18 .29 .01 .06 -.07 -.02 .17 .04 .06 .15 .21 .01 .08

P19 - .17 .02 .21 .11 .09 -.06 .04 .22 -.05 .24 .32 .08 .16 .06 .07

P20 - .18 .27 .27 .06 -.18 .01 .10 .04 .21 .28 .21 .19 .09 .05

P21 - .12 .28 .17 -.11 .06 .13 .20 .14 .19 .48 .26 .07 -.01

P22 - .12 .05 -.01 .06 .02 .10 .26 .15 .09 .19 .01 .14

P23 - .09 -.15 .06 .08 .20 .15 .26 .37 .32 .09 .05

P24 - -.01 .11 .08 .01 .01 .20 .22 .09 .33 .04

P25 - -.11 -.09 .01 -.21 -.18 -.12 -.12 -.05 .15

P26 - .01 .01 .09 .20 .09 .01 .13 -.06

P27 - .05 .12 .25 .17 .08 .13 -.06

P28 - .09 .05 .14 .17 .07 .01

P29 - .29 .20 .16 .03 .04

P30 - .34 .26 .17 .04

P31 - .36 .12 .10

P32 - .03 .06

P33 - .06

Page 226: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

216

Table 27 (continued)

P35 P36 P37 P38 P39 P40 P41 P42

P43 P44 P45 P46 P47 P48

P49

P50 P51

M .17 .38 .86 .96 .97 .89 .59 .58 .36 1.03 .90 .21 1.07 1.00 .90 .21 1.07 SD .64 .62 .98 .94 1.13 1.01 .88 .84 .72 .94 .96 .52 .92 .88 .96 .52 .92 P1 -.05 .16 .16 .16 .20 .35 .22 .20 .04 .15 .11 .19 .42 .21 .25 .23 .14 P2 .01 .05 .35 .28 .08 .22 .39 .24 .08 .12 .35 .15 .24 .27 .23 .27 .33 P3 .06 -.15 -.09 -.04 -.14 -.13 -.04 -.12 -.15 -.07 -.15 -.15 -.14 -.06 -.13 -.13 -.12 P4 .03 .08 -.04 -.02 .01 -.03 -.04 .01 .31 -.04 -.03 -.02 -.02 -.06 -.03 -.02 -.05 P5 -.03 -.04 .05 .03 .10 .09 .16 .11 .01 .07 .10 .02 .15 .10 .12 .05 .04 P6 -.03 -.06 -.05 -.06 -.08 -.06 -.02 -.05 .02 -.03 -.03 -.01 -.01 -.02 -.05 -.03 -.06 P7 -.07 .30 .14 .08 .08 .13 .08 .01 .12 .01 .06 .06 .06 .02 .08 .03 .10 P8 -.09 .17 .21 .23 .21 .32 .22 .24 .04 .13 .21 .14 .35 .24 .30 .23 .14 P9 -.04 .05 .37 .32 .07 .22 .45 .17 .04 .16 .27 .14 .22 .24 .16 .27 .25 P10 -.02 .06 .14 .25 .12 .17 .17 .27 .15 .06 .19 .08 .19 .19 .19 .17 .14 P11 .04 .14 .26 .19 .29 .46 .24 .20 .01 .14 .18 .22 .33 .19 .34 .16 .10 P12 -.05 .16 .48 .36 .04 .36 .30 .23 -.03 .17 .31 .20 .24 .32 .27 .30 .24 P13 -.06 .04 .20 .17 .06 .27 .09 .24 .04 .04 .16 .12 .17 .18 .25 .11 .17 P14 .01 .08 -.02 -.06 -.02 .01 -.02 .05 .27 -.10 -.04 .07 .07 -.02 -.02 -.05 -.01 P15 -.01 -.01 -.22 -.18 .07 -.09 -.12 -.17 .13 -.01 -.10 -.06 -.04 -.21 -.13 -.21 -.15 P16 -.05 .08 .23 .27 .07 .30 .17 .28 .03 .17 .20 .05 .31 .24 .42 .12 .17 P17 -.04 .29 .12 .12 .14 .11 .09 .07 .25 .02 .12 .17 .16 .06 .13 .05 .10 P18 .08 -.02 .09 .15 .22 .14 .10 .10 .06 .14 .08 .01 .23 .12 .17 .06 .03 P19 -.03 .03 .15 .17 .06 .11 .34 .15 .12 .06 .17 .09 .12 .10 .09 .13 .20 P20 -.10 .08 .43 .44 .07 .32 .23 .18 .04 .16 .26 .08 .23 .41 .23 .26 .20 P21 .05 .15 .20 .16 .29 .44 .19 .20 .11 .10 .19 .19 .52 .22 .34 .20 .13 P22 -.06 -.03 .29 .37 .02 .16 .26 .21 .04 .11 .31 .06 .18 .26 .13 .19 .22 P23 -.03 .05 .22 .20 .27 .29 .14 .24 .09 .21 .16 .18 .34 .25 .34 .20 .10 P24 -.03 .08 .05 .04 .14 .08 -.05 .11 .30 -.04 .08 .06 .12 .06 .08 .06 .09 P25 .07 -.04 -.19 -.14 .01 -.19 -.11 -.16 .08 .01 -.11 -.15 -.11 -.16 -.13 -.18 -.07 P26 .02 .04 .04 .13 .13 .07 .10 .09 -.04 .05 .09 .11 .11 .14 .02 .14 .16 P27 -.07 .22 .15 .06 .04 .20 .22 .08 .02 .01 .10 .32 .16 .13 .15 .07 .15 P28 .05 .06 .09 .18 .17 .15 .07 .23 .11 .11 .14 .08 .20 .06 .25 .08 .01 P29 -.02 .05 .35 .31 .06 .29 .40 .18 -.02 .22 .32 .13 .21 .26 .17 .26 .28

Page 227: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

217

Table 27 (continued)

P35 P36 P37 P38 P39 P40 P41 P42 P43 P44 P45 P46 P47 P48 P49 P50 P51 P30 -.07 .11 .30 .23 .14 .30 .27 .35 .14 .13 .23 .22 .29 .31 .30 .26 .24 P31 -.07 .13 .22 .18 .28 .39 .18 .26 .20 .23 .26 .16 .45 .23 .33 .24 .26 P32 -.07 .09 .24 .25 .20 .24 .17 .37 .19 .20 .19 .17 .33 .20 .38 .14 .14 P33 -.07 .12 .05 .01 .10 .03 .06 .09 .30 -.06 .03 .13 .11 .05 .04 .04 .07 P34 .06 -.02 .06 .06 .07 .05 .03 .01 .09 .19 .03 -.03 .01 .11 .04 .05 .06 P35 - -.06 -.02 .02 .02 -.04 .01 -.12 -.03 .01 -.07 -.05 -.02 -.07 .02 .02 -.03 P36 - .15 .04 .13 .14 .06 .04 .08 .01 .03 .19 .14 -.04 .13 .07 .05 P37 - .47 .05 .31 .34 .23 .01 .17 .44 .17 .26 .31 .27 .37 .29 P38 - .13 .30 .41 .28 .05 .20 .43 .08 .25 .40 .32 .33 .24 P39 - .26 .14 .09 .08 .09 .09 .19 .28 .13 .27 .12 .02 P40 - .32 .28 .07 .15 .22 .22 .37 .34 .37 .24 .14 P41 - .18 .02 .16 .35 .15 .24 .25 .22 .28 .30 P42 - .10 .18 .31 .10 .26 .26 .37 .16 .11 P43 - .01 .05 .15 .19 -.02 .17 .01 .06 P44 - .22 .04 .18 .20 .22 .20 .14 P45 - .04 .25 .25 .26 .29 .27 P46 - .21 .11 .23 .21 .10 P47 - .32 .38 .20 .20 P48 - .28 .26 .21 P49 - .20 .16 P50 - .30

P52

P53 P54 P55 P56 P57

P58 P59

P60 P61 P62

P63 P64

P65 P66

P67 P68 M 1.00 .88 .84 .99 1.07 1.07 .51 1.11 .84 .82 .60 .32 .94 .87 1.51 .69 .98 SD .82 .93 1.01 1.25 1.13 1.11 .80 1.04 .82 .88 1.01 .69 1.02 .95 1.04 .93 .95 P1 .19 .02 -.06 .18 .13 .20 .10 .14 -.03 .22 .28 .19 .19 .08 .25 .24 .20 P2 .25 .01 -.14 .16 .16 .15 .36 .15 -.03 .17 .16 .29 .13 .04 .17 .21 .17 P3 -.17 .06 .08 -.01 .02 -.23 -.01 -.03 -.12 -.06 -.15 -.24 -.23 -.04 -.10 -.11 -.11 P4 -.11 .32 .02 .07 .11 .07 -.07 -.04 .17 -.02 -.03 -.03 .08 .01 .02 -.06 .06 P5 .05 .02 .09 .13 .15 .03 .03 .14 -.08 .07 .20 .05 .09 .08 .10 .06 .20 P6 -.09 -.02 .05 -.02 -.03 -.02 -.02 -.02 -.07 .01 -.06 -.06 -.02 -.05 -.07 -.05 -.05 P7 .07 .09 .07 .10 .03 .07 -.06 -.02 .05 .12 .05 .07 .07 -.05 .01 .09 .08

Page 228: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

218

Table 27 (continued)

P52 P53 P54 P55 P56 P57 P58 P59 P60 P61 P62 P63 P64 P65 P66 P67 P68 P8 .26 .03 .01 .17 .10 .25 .10 .17 -.06 .30 .31 .29 .20 .12 .34 .33 .28 P9 .20 .04 -.15 .13 .14 .17 .32 .15 .01 .21 .21 .26 .10 .01 .17 .21 .20 P10 .10 .03 -.10 .09 .22 .24 .17 .37 .04 .13 .15 .24 .17 .04 .23 .20 .18 P11 .23 .01 -.08 .22 .09 .23 .12 .16 -.09 .25 .21 .23 .22 .13 .24 .29 .25 P12 .29 .02 -.15 .14 .05 .17 .15 .15 -.04 .23 .17 .26 .14 .01 .18 .24 .18 P13 .15 .01 -.10 .20 .11 .21 .12 .17 -.02 .24 .18 .18 .21 .08 .25 .30 .22 P14 -.03 .30 -.04 .01 .01 .03 -.02 -.05 .21 -.03 -.07 .04 .04 .06 .01 -.01 -.01 P15 -.08 -.01 .20 -.11 -.04 -.06 -.08 -.09 .02 -.10 -.12 -.10 .05 .08 -.13 -.15 -.06 P16 .29 .02 -.12 .23 .11 .20 .14 .32 -.09 .30 .24 .22 .19 .11 .29 .39 .23 P17 .10 .09 .03 .05 .03 .13 .05 .08 .06 .06 .11 .07 .13 .09 .11 .09 .12 P18 .10 .03 .04 .25 .12 .05 .08 .06 -.03 .11 .33 .12 .04 .17 .14 .11 .18 P19 .08 -.05 -.08 .07 .09 .04 .32 .10 .02 .12 .15 .15 .07 .04 .07 .07 .13 P20 .17 .02 -.16 .11 .04 .10 .19 .24 -.04 .19 .26 .24 .12 .01 .17 .19 .19 P21 .20 -.01 -.07 .18 .14 .25 .10 .09 .06 .24 .29 .23 .28 .10 .25 .28 .25 P22 .02 .09 -.08 .19 .18 .05 .26 .14 -.04 .10 .23 .21 .03 .02 .12 .20 .18 P23 .25 -.03 .01 .17 .03 .16 .12 .17 .09 .24 .34 .28 .24 .15 .33 .25 .21 P24 .06 .23 .07 -.01 .01 .12 .02 .04 .20 .08 .06 .07 .14 .07 .01 .06 .17 P25 -.17 .04 .16 -.05 .01 -.06 -.06 -.08 .01 -.16 -.08 -.23 -.06 .03 -.16 -.21 -.05 P26 .07 .13 -.03 -.11 .01 .08 .08 .02 .03 .04 .03 .16 .09 .03 .02 .07 .08 P27 .17 .01 -.01 .15 .08 .19 .15 .11 .03 .20 .11 .11 .14 .04 .13 .19 .16 P28 .06 .06 .01 .18 .23 .14 .02 .20 -.04 .07 .25 .21 .16 .01 .22 .20 .15 P29 .24 -.02 -.10 .12 .07 .13 .31 .10 .01 .15 .15 .30 .10 -.01 .17 .16 .09 P30 .22 .04 -.12 .13 .06 .25 .24 .25 .07 .30 .17 .31 .21 .08 .28 .32 .22 P31 .31 .06 -.01 .26 .06 .30 .04 .14 .08 .31 .34 .33 .26 .18 .33 .31 .28 P32 .21 .03 -.04 .23 .10 .25 .09 .35 -.06 .25 .36 .30 .13 .07 .38 .32 .23 P33 .07 .37 .07 .02 .10 .14 .06 .01 .13 .12 .03 .05 .13 .13 .04 .01 .09 P34 -.01 .18 .04 .09 .06 .01 .02 .07 .06 .04 .03 .04 .01 .08 .05 .07 .06 P35 -.07 .06 -.03 -.10 -.03 -.07 .01 -.09 .06 .02 -.02 -.09 -.03 -.06 -.08 -.03 -.02 P36 .10 .04 .02 .11 -.01 .15 .01 .01 -.02 .11 .05 .09 .11 .04 .08 .08 .15 P37 .25 .03 -.15 .20 .02 .14 .24 .17 -.08 .23 .19 .27 .17 -.05 .22 .27 .20 P38 .21 .04 -.14 .12 .09 .10 .24 .25 -.03 .23 .25 .23 .16 -.02 .21 .26 .17

Page 229: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

219

Table 27 (continued)

P52 P53 P54 P55 P56 P57 P58 P59 P60 P61 P62 P63 P64 P65 P66 P67 P68 P39 .09 -.04 .08 .12 .07 .20 .04 .08 .01 .15 .24 .21 .14 .22 .25 .24 .19 P40 .26 .02 -.14 .20 .12 .22 .10 .16 -.05 .29 .31 .32 .22 .07 .22 .34 .27 P41 .15 .01 -.13 .13 .14 .14 .31 .17 -.06 .21 .21 .24 .07 .02 .15 .22 .14 P42 .21 .05 -.11 .18 .19 .23 .17 .27 -.10 .29 .31 .29 .19 .11 .38 .32 .19 P43 .05 .21 .05 .05 .05 .19 -.06 .11 .21 .09 .09 .20 .22 .09 .16 .09 .10 P44 .14 .01 -.02 .12 -.04 .09 .10 .14 .02 .13 .22 .19 .11 .09 .16 .17 .17 P45 .23 .02 -.17 .16 .13 .25 .26 .20 -.02 .17 .26 .30 .18 .06 .25 .26 .20 P46 .19 .04 .01 .09 .06 .15 .04 .09 .07 .23 .13 .22 .19 .04 .18 .26 .12 P47 .32 .05 -.04 .25 .16 .29 .19 .14 .08 .28 .35 .34 .30 .14 .38 .33 .29 P48 .23 -.02 -.14 .13 .07 .17 .19 .15 .06 .14 .28 .26 .11 .03 .26 .32 .16 P49 .30 .01 -.12 .28 .04 .30 .14 .32 -.05 .34 .35 .37 .23 .15 .46 .47 .28 P50 .24 -.03 -.17 .05 -.04 .10 .23 .10 -.02 .17 .20 .28 .15 -.03 .17 .28 .14 P51 .21 -.01 -.18 .07 .02 .08 .40 .15 .04 .09 .15 .27 .12 .02 .13 .18 .12 P52 - .01 -.10 .17 -.01 .17 .10 .14 .06 .30 .24 .28 .16 .12 .29 .31 .17 P53 - -.03 .09 .06 .08 -.11 .03 .17 .06 .01 .03 .03 .02 .02 .03 .08 P54 - -.01 .06 -.04 -.11 -.05 -.03 -.06 -.05 -.15 .06 .02 -.03 -.12 .01 P55 - .27 .23 .07 .15 -.07 .21 .26 .24 .16 .09 .28 .26 .18 P56 - .13 .11 .11 .01 .08 .13 .13 .12 .03 .16 .02 .12 P57 - .13 .21 .07 .20 .26 .24 .31 .11 .34 .31 .25 P58 - .08 -.03 .12 .12 .14 .07 .04 .12 .13 .10 P59 - -.03 .19 .27 .24 .15 .09 .29 .26 .10 P60 - .01 -.06 .04 .08 .12 .02 .01 .01 P61 - .29 .22 .18 .04 .34 .44 .25 P62 - .32 .17 .12 .38 .35 .28 P63 - .16 .02 .35 .30 .18 P64 - .10 .26 .19 .23 P65 - .14 .08 .13 P66 - .51 .21 P67 - .24

Page 230: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

220

Table 27 (continued)

P69 P70 P71 P72 P73 P74 P75 P76 P77 P78 P79 P80 M 1.37 .68 .78 .19 .43 .66 .60 .43 .80 .83 1.32 .56 SD .98 .91 .96 .50 .75 1.11 .84 .71 .84 .92 .94 .84 P1 .04 .21 .12 .03 .09 .08 .17 .23 .14 .19 .09 .10 P2 .14 .20 .25 .07 .07 .33 .15 .21 .16 .35 .15 .18 P3 .06 -.07 -.01 -.17 -.06 -.17 -.13 -.12 .06 -.03 .13 -.12 P4 .05 -.08 .01 .07 .06 .01 .08 -.07 -.01 .01 -.01 -.04 P5 .19 .05 .16 -.06 .06 .07 .14 .08 .13 .10 .10 .13 P6 .02 -.07 -.04 .01 .01 -.06 .05 -.03 -.05 -.07 -.01 -.06 P7 -.08 .05 .06 .02 .19 .10 .11 .10 .02 .01 -.01 .06 P8 .06 .34 .13 -.01 .05 .12 .17 .26 .23 .26 .12 .08 P9 .17 .12 .19 -.01 .09 .25 .06 .22 .16 .38 .15 .09 P10 .18 .17 .11 .12 .15 .18 .13 .21 .11 .23 .03 .15 P11 .05 .31 .06 .08 .06 .07 .11 .26 .14 .16 .07 .21 P12 -.02 .27 .10 .06 .08 .15 .09 .30 .15 .33 .20 .09 P13 .07 .29 .08 .13 .04 .04 .18 .24 .17 .27 .10 .19 P14 -.03 .02 -.01 .03 .09 .05 .09 .01 .03 -.02 -.07 -.10 P15 .02 -.15 -.07 -.02 .01 -.07 -.04 -.12 -.11 -.17 -.16 -.05 P16 .01 .45 .01 -.02 .04 .09 .10 .25 .24 .24 .19 .25 P17 -.02 .07 .16 -.01 .17 .07 .10 .07 .05 .11 .05 .01 P18 .14 .14 .12 .01 -.09 .04 .05 .13 .10 .09 .09 .10 P19 .14 .07 .29 .05 -.02 .31 .07 .14 .07 .22 .03 .07 P20 .07 .22 .06 .07 .03 .21 .09 .25 .15 .31 .15 .11 P21 .12 .27 .14 .10 .10 .10 .27 .21 .22 .18 .11 .07 P22 .26 .11 .11 .01 .05 .15 .06 .19 .23 .30 .22 .03 P23 .08 .27 .16 .08 .05 .17 .16 .23 .15 .23 .12 .09 P24 -.05 .08 .02 .18 .08 .15 .15 .13 .03 .07 .08 -.02 P25 .14 -.14 -.03 -.04 .01 -.13 .01 -.15 -.04 -.10 -.02 .02 P26 -.03 .01 .07 .03 .03 .16 .09 .08 .02 .08 -.10 .02 P27 .02 .18 .25 .09 .08 .18 .14 .16 .16 .20 .13 .10 P28 .16 .14 .07 .02 .08 .03 .09 .17 .14 .14 .14 .19 P29 .05 .13 .19 .06 .07 .30 .06 .26 .11 .29 .17 .11

Page 231: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

221

Table 27 (continued)

P69 P70 P71 P72 P73 P74 P75 P76 P77 P78 P79 P80 P30 .11 .26 .20 .18 .10 .32 .15 .37 .16 .29 .06 .13 P31 .03 .31 .14 .16 .09 .18 .23 .33 .23 .28 .15 .06 P32 .17 .29 .07 .08 .11 .22 .11 .17 .19 .29 .23 .08 P33 -.02 .03 .07 .14 .14 .09 .18 .07 .10 .06 .07 .04 P34 .02 .05 -.03 .01 .01 .01 .07 .03 .10 .07 .10 -.02 P35 -.05 -.09 .02 .05 -.07 .01 -.11 -.06 -.11 -.05 -.04 .01 P36 -.04 .08 -.01 .13 .21 .06 .10 .08 .07 .04 .02 .10 P37 -.03 .23 .11 .08 .10 .20 .10 .29 .20 .33 .22 .15 P38 .10 .22 .14 .10 .04 .20 .03 .22 .20 .33 .23 .07 P39 .06 .12 .17 .11 .10 .02 .21 .14 .16 .05 .03 .11 P40 .09 .26 .13 .09 .09 .04 .16 .27 .18 .23 .13 .14 P41 .07 .15 .29 .08 .03 .3 .05 .22 .19 .33 .23 .07 P42 .15 .35 .13 .17 .04 .12 .15 .23 .19 .24 .18 .20 P43 .06 .05 .01 .11 .06 .11 .13 .05 .01 .09 .05 -.02 P44 .12 .10 .06 .03 .01 .06 .05 .14 .10 .21 .14 .11 P45 .10 .24 .12 .09 .01 .23 .13 .27 .17 .36 .12 .10 P46 -.02 .19 .13 .15 .17 .10 .14 .13 .13 .13 .05 .04 P47 .12 .31 .18 .08 .14 .20 .21 .23 .21 .29 .13 .08 P48 .09 .27 .20 .09 .02 .26 .10 .22 .19 .25 .15 .07 P49 .11 .47 .11 .12 .09 .17 .13 .27 .22 .29 .18 .21 P50 -.01 .08 .12 .11 .02 .24 .06 .30 .14 .23 .03 .13 P51 .02 .18 .16 .11 .09 .52 .11 .25 .12 .21 .18 .01 P52 -.08 .37 .18 .07 .13 .18 .14 .25 .19 .22 .13 .15 P53 .05 .03 -.09 -.02 .14 -.05 .07 .06 .10 .08 .06 -.03 P54 .01 -.08 .03 .03 .08 -.07 .09 -.13 .01 -.11 -.01 .02 P55 .11 .26 .12 .06 .14 -.02 .08 .14 .22 .18 .14 .17 P56 .20 .07 .11 .10 .18 .04 .16 -.02 .08 .13 .08 .13 P57 .06 .27 .07 .17 .16 .14 .16 .14 .13 .27 .05 .20 P58 .12 .08 .19 .05 .02 .38 .03 .13 .08 .25 .13 .05 P59 .23 .25 .07 .05 .15 .21 .15 .27 .13 .21 .18 .19 P60 .01 .01 .08 .11 .09 .14 .04 -.02 -.05 .02 -.06 -.16

Page 232: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

222

Table 27 (continued)

P69 P70 P71 P72 P73 P74 P75 P76 P77 P78 P79 P80 P61 .03 .39 .08 .07 .05 .06 .14 .23 .20 .26 .21 .21 P62 .23 .28 .22 .08 .03 .16 .13 .16 .20 .26 .20 .21 P63 .15 .23 .11 .12 .08 .25 .21 .31 .15 .29 .11 .12 P64 .05 .16 .09 .17 .08 .13 .18 .13 .12 .21 -.06 .16 P65 .08 .11 .06 .04 .04 .03 .10 .08 .04 .07 .01 -.01 P66 .17 .46 .13 .12 .05 .14 .19 .30 .22 .31 .13 .18 P67 .01 .54 .09 .13 .03 .15 .10 .26 .29 .29 .21 .16 P68 .15 .23 .14 .13 .07 .11 .21 .21 .19 .37 .18 .14 P69 - -.04 .12 -.03 .08 .04 .09 .06 .13 .19 .17 .06 P70 - .14 .08 .05 .09 .10 .18 .26 .21 .15 .13 P71 - .06 .08 .22 .13 .07 .08 .08 .10 .09 P72 - .21 .21 .08 .09 .13 .10 .06 .10 P73 - .07 .07 .07 .12 .09 .07 .09 P74 - .06 .17 .02 .18 .16 .01 P75 - .16 .13 .12 .15 .10 P76 - .18 .28 .14 .16 P77 - .27 .30 .12 P78 - .23 .17 P79 - .08

Page 233: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

223

Table 28

Item Correlation Matrix of the CPRS-R Scores for CFA Sample (N = 1,253)

P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 P13 P14 P15 P16 P17 M 1.43 .79 1.44 .56 1.11 .01 .43 1.37 .80 .41 .78 .79 .35 .30 .50 .86 .35 SD 1.03 1.10 1.18 .86 1.00 .11 .68 1.05 1.05 .78 .95 1.00 .72 .67 .83 .95 .66 P1 - .18 -.01 .07 .20 -.03 .17 .35 .17 .09 .26 .17 .15 .06 -.06 .19 .11 P2 - -.18 .07 .05 -.02 .05 .19 .54 .17 .12 .31 .11 .09 -.16 .15 .07 P3 - -.08 .02 .01 -.13 -.14 -.12 -.15 -.02 -.07 -.05 -.02 -.05 -.02 -.10 P4 - .01 -.02 .11 -.01 .04 .10 -.01 .02 .04 .30 -.01 -.04 .12 P5 - .01 -.02 .11 .05 .12 .13 .09 .13 -.03 .12 .08 .01 P6 - .01 .04 .01 .01 -.02 -.02 -.04 .01 .01 .01 .03 P7 - .13 .03 .07 .13 .06 .03 .10 -.01 .05 .35 P8 - .15 .17 .35 .24 .19 .05 -.06 .35 .12 P9 - .18 .17 .37 .10 .03 -.19 .17 .07 P10 - .13 .20 .17 .11 -.06 .19 .14 P11 - .26 .16 -.03 -.06 .29 .07 P12 - .17 .05 -.18 .25 .09 P13 - .04 .01 .18 .06 P14 - -.05 .05 .06 P15 - -.13 -.01 P16 - .03

P18 P19 P20 P21 P22 P23 P24 P25 P26 P27 P28 P29 P30 P31 P32 P33 P34 M 1.58 .37 1.05 1.27 1.13 1.00 .28 1.07 .11 .19 .80 .62 .52 .92 .76 .72 .43 SD 1.11 .81 1.02 1.01 1.11 1.01 .64 1.08 .47 .53 .92 .93 .88 1.01 .94 1.05 .83 P1 .17 .10 .19 .40 .10 .22 .13 -.02 .12 .10 .13 .17 .17 .34 .27 .10 .03 P2 .09 .30 .26 .13 .17 .15 .07 -.12 .03 .21 .07 .31 .29 .18 .17 .08 -.03 P3 .06 -.09 -.09 -.16 .09 -.08 -.07 .10 -.09 -.08 -.05 -.11 -.21 -.13 -.01 -.08 .12 P4 .01 .03 .01 .09 .04 .05 .20 .01 .05 .09 .04 .05 .07 .05 .02 .28 .03 P5 .14 .06 .07 .17 .06 .08 .02 .15 .08 .01 .09 .10 .03 .13 .17 .04 .14 P6 .01 -.04 -.02 -.01 .05 -.02 .02 .03 .09 -.01 -.05 -.03 -.03 -.05 -.01 .01 .03 P7 .01 .02 .05 .15 -.05 .09 .12 -.08 .06 .23 .09 .07 .06 .18 .04 .14 -.07

Page 234: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

224

Table 28 (continued)

P18 P19 P20 P21 P22 P23 P24 P25 P26 P27 P28 P29 P30 P31 P32 P33 P34 P8 .20 .10 .28 .37 .17 .23 .05 -.11 .01 .08 .19 .14 .16 .35 .32 .05 -.04 P9 .09 .29 .31 .18 .23 .22 .08 -.13 .10 .18 .08 .36 .26 .20 .18 .07 -.02 P10 .12 .21 .21 .21 .10 .11 .08 -.01 .07 .03 .21 .17 .28 .16 .30 .10 .03 P11 .20 .10 .24 .29 .08 .24 .01 -.01 .09 .14 .21 .19 .14 .26 .24 .11 -.02 P12 .10 .17 .30 .23 .20 .25 .07 -.14 .05 .08 .10 .29 .30 .25 .27 .08 .01 P13 .13 .09 .14 .21 .12 .16 .04 -.04 .01 .09 .08 .14 .12 .18 .16 .08 .01 P14 -.03 .04 .06 .13 .07 .01 .29 -.05 -.01 .05 -.01 .03 .04 .08 .04 .18 .01 P15 .01 -.07 -.12 -.03 -.06 -.07 .04 .19 -.01 -.07 .02 -.12 -.08 -.01 -.13 .01 .06 P16 .15 .06 .25 .26 .17 .21 .01 -.07 .01 .06 .19 .14 .11 .26 .32 .04 .03 P17 .02 .06 .08 .17 .03 .07 .11 -.06 .05 .16 .11 .07 .10 .11 .11 .12 -.05 P18 - .08 .11 .22 .15 .30 -.01 .04 -.05 .02 .17 .13 .07 .20 .25 .04 .06 P19 - .18 .15 .07 .19 .08 -.07 .04 .13 .09 .28 .30 .12 .17 .04 -.01 P20 - .30 .17 .24 .10 -.13 .09 .08 .10 .30 .19 .22 .27 .03 .01 P21 - .12 .26 .11 -.05 .10 .18 .22 .19 .22 .42 .35 .13 .02 P22 - .10 .01 -.01 -.03 .03 .11 .16 .10 .14 .17 -.04 .04 P23 - .03 -.07 .04 .12 .11 .21 .19 .25 .28 .06 .04 P24 - -.08 .09 .08 .01 .01 .14 .11 .07 .23 .04 P25 - .01 -.03 .01 -.09 -.14 -.05 -.13 -.06 .12 P26 - .02 -.01 .08 .14 .10 .05 .10 .02 P27 - .06 .17 .14 .13 .09 .13 -.06 P28 - .05 .12 .19 .23 .03 -.02 P29 - .35 .24 .24 .04 .01 P30 - .30 .27 .10 .05 P31 - .36 .15 .07 P32 - .08 .08 P33 - -.01

P35 P36 P37 P38 P39 P40 P41 P42 P43 P44 P45 P46 P47 P48 P49 P50 P51 M .17 .38 .86 .96 .97 .89 .59 .58 .36 1.03 .90 .21 1.07 1.00 .82 .84 .99 SD .64 .62 .98 .94 1.13 1.01 .88 .84 .72 .94 .96 .52 .92 .88 .93 1.01 1.25

P1 -.02 .12 .17 .15 .18 .27 .14

.19 .09 .16 .13 .10 .41 .18 .20 .13 .08

Page 235: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

225

Table 28 (continued)

P35 P36 P37 P38 P39 P40 P41 P42 P43 P44 P45 P46 P47 P48 P49 P50 P51 P2 -.01 .08 .25 .28 .01 .13 .32 .13 .09 .08 .24 .14 .13 .23 .18 .21 .26 P3 .06 -.13 -.05 -.08 -.06 -.05 .01 -.06 -.16 -.07 -.09 -.06 -.14 -.02 -.07 -.10 -.22 P4 -.07 .11 .05 .03 .05 .01 .02 .02 .27 .05 .04 .02 .08 .07 -.01 .05 .01 P5 -.01 .01 .07 .09 .15 .15 .08 .14 .02 .14 .10 .02 .17 .08 .11 -.01 -.04 P6 .02 .01 .01 -.02 -.02 -.01 .02 -.04 .01 .01 -.01 .01 -.02 -.01 -.02 -.01 -.01 P7 .01 .34 .03 .01 .02 .10 .01 .02 .13 .01 .03 .06 .15 .07 .07 .05 .04 P8 -.08 .13 .27 .23 .19 .28 .11 .21 .07 .17 .18 .12 .38 .25 .31 .19 .16 P9 .02 .07 .29 .32 .06 .19 .38 .21 .05 .07 .29 .16 .21 .29 .16 .26 .24 P10 .03 .06 .17 .24 .15 .20 .19 .24 .15 .12 .23 .06 .26 .17 .23 .11 .11 P11 .03 .07 .25 .20 .23 .42 .14 .12 .02 .14 .14 .11 .32 .28 .28 .16 .09 P12 -.08 .06 .43 .38 .14 .32 .26 .27 -.01 .14 .29 .14 .26 .36 .23 .26 .18 P13 -.04 .01 .11 .12 .12 .15 .09 .21 .05 .04 .10 .10 .20 .17 .18 .10 .04 P14 -.02 .12 .06 .03 -.01 .01 .01 .07 .21 .03 .06 .09 .08 .03 .01 .04 .01 P15 .05 .04 -.16 -.15 .02 -.13 -.18 -.04 .12 .06 -.14 -.10 -.07 -.15 -.07 -.18 -.09 P16 .01 .05 .31 .23 .12 .23 .13 .25 .02 .19 .22 .06 .27 .23 .35 .18 .15 P17 .02 .28 .09 .07 -.01 .08 .07 .10 .12 .01 .07 .08 .15 .13 .08 .06 .01 P18 .02 .02 .15 .17 .26 .19 .13 .12 .10 .16 .14 .08 .23 .18 .27 .07 .06 P19 .03 -.01 .14 .23 .11 .17 .26 .09 .08 .06 .19 .10 .14 .16 .13 .12 .15 P20 -.01 .03 .35 .37 .14 .32 .18 .24 .09 .19 .28 .15 .28 .37 .24 .23 .21 P21 -.03 .16 .23 .26 .24 .33 .18 .25 .15 .22 .23 .18 .56 .24 .28 .22 .10 P22 -.01 -.04 .24 .26 .01 .12 .24 .15 .11 .10 .28 .07 .09 .24 .12 .16 .12 P23 -.04 .06 .21 .22 .25 .27 .15 .18 .11 .20 .19 .14 .26 .25 .32 .18 .16 P24 .01 .18 .04 .03 .02 .04 .05 .08 .20 .04 .02 .08 .10 .05 .03 .05 .03 P25 .06 -.08 -.13 -.11 .11 -.05 -.07 -.05 .01 .07 -.11 -.05 -.02 -.11 -.07 -.20 -.08 P26 .18 .06 .10 .05 .03 .09 .05 .03 .04 .08 .05 .03 .13 .04 .02 .05 .09 P27 .01 .16 .10 .06 .07 .12 .20 .09 .05 -.01 .05 .30 .13 .12 .09 .11 .08 P28 -.02 .08 .11 .13 .09 .21 .16 .20 .08 .13 .18 .12 .23 .15 .19 .11 -.01 P29 -.01 .03 .29 .32 .13 .31 .38 .18 .07 .14 .28 .17 .20 .25 .20 .22 .23 P30 .02 .15 .23 .28 .12 .25 .24 .18 .17 .14 .22 .16 .28 .24 .21 .23 .21 P31 -.02 .19 .27 .26 .21 .31 .22 .25 .17 .28 .28 .15 .43 .29 .29 .26 .22 P32 -.04 .09 .28 .28 .22 .32 .24 .29 .11 .25 .28 .15 .37 .32 .32 .17 .12

Page 236: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

226

Table 28 (continued)

P35 P36 P37 P38 P39 P40 P41 P42 P43 P44 P45 P46 P47 P48 P49 P50 P51 P33 -.03 .15 .05 .05 .06 .14 .09 .15 .22 .05 .06 .07 .17 .12 .05 .07 .03 P34 .12 -.01 .02 .02 .07 .07 .05 .08 .06 .25 .07 .01 .05 .01 .05 .01 .01 P35 - .01 .01 -.01 -.03 .01 .01 -.01 -.03 .07 -.01 -.01 -.03 -.05 -.03 .01 -.02 P36 - .12 -.01 -.01 .05 .04 .05 .17 .08 .07 .09 .14 .10 .05 .11 .09 P37 - .45 .13 .31 .25 .21 .04 .22 .37 .14 .26 .34 .28 .32 .24 P38 - .12 .31 .31 .25 .07 .23 .47 .16 .30 .36 .26 .30 .24 P39 - .29 .10 .13 .09 .13 .10 .11 .28 .17 .26 .06 .01 P40 - .28 .29 .01 .20 .29 .20 .40 .41 .34 .26 .09 P41 - .17 -.01 .12 .31 .19 .18 .25 .19 .31 .24 P42 - .09 .18 .25 .15 .27 .30 .28 .23 .04 P43 - .11 .05 .02 .12 .05 .06 -.02 .08 P44 - .20 .06 .25 .21 .21 .14 .15 P45 - .11 .25 .31 .24 .31 .28 P46 - .17 .14 .18 .13 .09 P47 - .31 .33 .25 .14 P48 - .31 .28 .18 P49 - .22 .11 P50 - .26

P52 P53 P54 P55 P56 P57 P58 P59 P60 P61 P62 P63 P64 P65 P66 P67 P68 M 1.07 1.07 .51 1.11 .84 .82 .60 .32 .94 .87 1.51 .69 .98 1.65 .86 .65 1.12 SD 1.13 1.11 .80 1.04 .82 .88 1.01 .69 1.02 .95 1.04 .93 .95 1.11 .94 .84 .98 P1 .18 .05 .03 .12 .03 .22 .01 .07 .02 .19 .21 .17 .15 .09 .15 .21 .28 P2 .19 -.01 -.09 .12 .05 .13 .3 .15 .07 .19 .16 .17 .10 .03 .14 .16 .13 P3 -.12 .05 .06 -.04 .01 -.15 -.06 -.08 -.12 -.09 -.08 -.22 -.21 -.14 -.10 -.05 -.13 P4 .04 .26 .02 .01 -.01 .07 .05 .02 .17 -.02 -.01 .07 .02 .06 .01 .03 .07 P5 -.01 .05 .12 .08 .14 .13 .04 .15 -.05 .07 .13 .04 .12 .03 .08 .09 .25 P6 -.05 .02 -.01 .01 .02 0 -.04 -.03 .01 .02 -.02 -.04 -.03 -.01 .01 -.03 -.01 P7 .08 .07 .02 .04 -.05 .1 -.03 .03 .13 .1 .02 .09 .14 .05 .03 .04 .10 P8 .27 -.01 -.06 .21 .04 .28 .04 .11 -.05 .31 .27 .22 .22 .08 .29 .28 .25 P9 .18 0 -.13 .14 .06 .16 .27 .15 .07 .19 .16 .18 .07 -.01 .17 .20 .10 P10 .12 .01 -.07 .18 .23 .21 .11 .34 .04 .11 .19 .25 .15 .10 .21 .19 .15

Page 237: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

227

Table 28 (continued)

P52 P53 P54 P55 P56 P57 P58 P59 P60 P61 P62 P63 P64 P65 P66 P67 P68 P11 .23 .02 -.02 .15 .07 .25 .04 .07 -.06 .28 .24 .11 .19 .05 .25 .25 .25 P12 .27 .01 -.14 .15 .03 .26 .11 .20 .04 .23 .24 .21 .13 .07 .21 .26 .18 P13 .13 .04 -.03 .11 .08 .12 .02 .16 .02 .15 .19 .12 .10 .07 .19 .20 .21 P14 .06 .16 -.05 .01 -.01 .05 .01 .04 .18 .02 -.02 .02 .07 .02 .01 .07 .04 P15 -.16 -.03 .23 -.08 -.01 -.03 -.06 -.12 -.05 -.10 -.02 -.06 .10 .07 -.09 -.13 .01 P16 .29 .03 -.11 .20 .05 .24 .10 .18 -.07 .28 .19 .12 .13 .05 .30 .33 .21 P17 .10 .04 .01 .06 .01 .13 .01 .06 .09 .13 .03 .09 .16 .06 .05 .14 .12 P18 .12 .01 .01 .19 .14 .12 .05 .06 -.06 .16 .41 .15 .09 .16 .23 .16 .21 P19 .08 .01 -.04 .08 .15 .15 .22 .20 .06 .12 .15 .17 .09 .05 .17 .17 .10 P20 .24 -.03 -.09 .13 .09 .21 .13 .11 .04 .19 .21 .23 .16 .02 .20 .21 .20 P21 .24 .08 -.02 .20 .12 .34 .06 .14 .06 .28 .31 .27 .28 .12 .29 .30 .37 P22 .16 .03 -.13 .11 .11 .05 .14 .10 .02 .13 .20 .13 .01 -.02 .16 .12 .08 P23 .28 -.04 -.01 .21 .02 .25 .08 .10 .02 .22 .35 .29 .18 .15 .29 .34 .21 P24 .05 .16 -.02 .01 .01 .11 .01 -.02 .23 .02 -.02 .07 .07 .07 .01 .07 .07 P25 -.19 .06 .15 .01 .06 -.06 .01 -.04 .01 -.10 -.06 -.20 -.02 .02 -.07 -.13 -.02 P26 .01 .02 .06 .03 .04 .10 .01 .01 .07 .07 .02 -.01 .06 .06 .07 .03 .12 P27 .12 .08 .04 .07 -.05 .12 .11 .06 .05 .14 .07 .05 .07 .04 .10 .11 .13 P28 .12 .03 -.02 .19 .19 .19 .05 .20 -.04 .20 .21 .14 .19 .05 .22 .20 .20 P29 .19 -.02 -.08 .07 .04 .18 .25 .10 .02 .18 .19 .26 .11 .07 .17 .19 .20 P30 .21 .01 -.03 .09 .08 .21 .12 .16 .03 .20 .21 .29 .18 .13 .19 .20 .21 P31 .28 .06 -.02 .19 .09 .32 .08 .15 .06 .29 .30 .30 .25 .17 .28 .26 .31 P32 .31 .05 -.05 .24 .14 .33 .06 .27 -.03 .25 .34 .29 .16 .09 .32 .38 .27 P33 .07 .27 .01 .09 .01 .09 .10 .12 .15 .06 .10 .04 .14 .10 .07 .08 .12 P34 -.01 .06 .03 .01 .06 .01 -.02 .01 .01 -.03 .02 .06 -.02 .05 .04 .04 .07 P35 -.01 .01 -.01 -.02 .06 -.04 .01 -.05 .04 -.04 -.02 -.08 -.01 -.04 -.01 -.03 .01 P36 .10 .10 .01 .07 .01 .17 .01 .01 .13 .09 .05 .12 .15 .03 .02 .02 .11 P37 .27 .03 -.12 .22 .05 .25 .14 .15 .03 .24 .20 .22 .15 .04 .21 .24 .20 P38 .23 .03 -.14 .22 .14 .20 .18 .19 .02 .23 .24 .31 .14 .08 .26 .24 .23 P39 .13 -.02 .10 .19 .09 .21 .02 .10 -.03 .15 .30 .16 .18 .15 .20 .20 .17 P40 .25 .03 -.05 .21 .08 .3 .09 .13 -.02 .29 .28 .25 .21 .09 .28 .30 .28 P41 .19 .01 -.11 .15 .13 .16 .33 .14 -.02 .19 .18 .18 .06 -.01 .15 .19 .15

Page 238: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

228

Table 28 (continued)

P52 P53 P54 P55 P56 P57 P58 P59 P60 P61 P62 P63 P64 P65 P66 P67 P68 P42 .22 .03 -.02 .24 .14 .28 .04 .31 -.02 .24 .22 .21 .23 .06 .28 .25 .27 P43 .06 .17 .04 .02 .09 .12 .03 .08 .16 .02 .11 .16 .10 .13 .11 .08 .14 P44 .20 .09 .03 .17 .08 .15 .07 .09 .02 .06 .17 .18 .15 .13 .15 .17 .21 P45 .23 .04 -.10 .18 .15 .23 .20 .23 .02 .19 .27 .28 .17 .04 .22 .19 .22 P46 .13 .05 -.04 .13 .05 .16 .12 .09 .04 .24 .13 .12 .12 .06 .13 .16 .16 P47 .21 .07 -.01 .26 .12 .38 .06 .19 .08 .30 .32 .26 .35 .15 .35 .33 .35 P48 .28 .08 -.11 .24 .11 .32 .15 .24 .03 .26 .27 .26 .17 .10 .27 .31 .28 P49 .26 .01 -.01 .26 .16 .32 .13 .24 -.07 .35 .30 .24 .27 .07 .43 .40 .30 P50 .28 .01 -.14 .16 -.01 .17 .19 .09 .06 .23 .18 .25 .12 .01 .20 .19 .15 P51 .22 -.08 -.15 .08 .01 .11 .33 .01 .12 .17 .12 .25 .05 .06 .09 .10 .09 P52 - .04 -.12 .25 .01 .28 .07 .14 .04 .33 .30 .28 .17 .15 .29 .31 .21 P53 - -.03 .12 .04 .03 .01 .06 .12 .06 .03 -.02 .04 .01 .01 .01 .09 P54 - .01 .05 -.05 -.07 -.06 -.08 -.04 -.02 -.12 .10 .06 -.06 -.04 .07 P55 - .21 .24 .14 .13 .01 .25 .29 .17 .19 .11 .29 .26 .18 P56 - .12 .08 .18 -.07 .10 .16 .14 .13 .02 .16 .05 .14 P57 - .06 .28 .07 .30 .20 .28 .32 .17 .30 .31 .32 P58 - .08 .07 .12 .08 .08 .06 -.05 .09 .09 .08 P59 - .02 .15 .17 .22 .11 -.02 .27 .20 .13 P60 - -.04 -.07 .13 .08 .08 -.06 -.06 .02 P61 - .29 .25 .27 .09 .35 .37 .27 P62 - .31 .21 .13 .37 .28 .25 P63 - .15 .15 .28 .23 .24 P64 - .12 .23 .25 .32 P65 - .14 .05 .16 P66 - .50 .26 P67 - .29

P69 P70 P71 P72 P73 P74 P75 P76 P77 P78 P79 P80 M 1.38 .68 .75 .21 .46 .66 .63 .45 .84 .79 1.34 .57 SD .96 .94 .91 .51 .75 1.11 .85 .72 .82 .91 .98 .82 P1 .06 .21 .15 .02 .11 .11 .24 .12 .21 .18 .14 .14 P2 .02 .12 .23 .07 .07 .25 .14 .18 .13 .28 .13 .05

Page 239: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

229

Table 28 (continued)

P69 P70 P71 P72 P73 P74 P75 P76 P77 P78 P79 P80 P3 .04 -.06 -.11 -.13 -.05 -.20 -.09 -.15 .01 -.06 .15 .02 P4 .02 -.01 .06 .07 .07 .04 .10 .03 .08 .07 -.03 -.09 P5 .21 .07 .06 -.01 -.01 -.01 .11 .06 .14 .09 .06 .13 P6 .01 -.01 -.04 .02 .02 -.02 -.07 -.06 -.05 -.03 -.05 -.02 P7 -.03 .08 .03 .03 .19 .07 .18 .07 .04 .04 -.04 -.01 P8 .06 .33 .15 .08 .11 .15 .20 .19 .22 .22 .14 .12 P9 .07 .14 .24 .06 .09 .22 .07 .18 .15 .35 .16 .04 P10 .14 .12 .16 .13 .16 .13 .16 .15 .15 .22 .01 .08 P11 .06 .27 .16 .05 .11 .08 .17 .15 .16 .14 .15 .18 P12 .11 .20 .16 .06 .13 .14 .13 .17 .14 .29 .2 .13 P13 .10 .13 .08 .04 .04 .03 .10 .11 .15 .19 .09 .11 P14 -.03 .04 -.01 .05 .11 .01 .10 .05 .09 .06 .02 .01 P15 -.01 -.09 -.07 .05 -.05 -.07 -.03 -.10 -.11 -.11 -.17 -.03 P16 .04 .40 .05 .07 .14 .12 .15 .17 .22 .24 .21 .21 P17 .04 .10 .05 .09 .16 .08 .19 .04 .03 .10 -.04 .05 P18 .13 .17 .09 -.01 .02 .04 .09 .09 .18 .05 .15 .10 P19 .06 .10 .32 .04 .01 .23 .04 .10 .13 .14 .02 .06 P20 .11 .16 .16 .10 .08 .17 .14 .21 .18 .22 .15 .11 P21 .14 .28 .14 .07 .14 .10 .27 .24 .30 .22 .11 .16 P22 .20 .11 .10 -.03 .05 .06 .05 .09 .17 .25 .26 -.01 P23 .08 .28 .20 .05 .06 .18 .14 .16 .17 .15 .12 .12 P24 -.04 .07 .05 .07 .04 .06 .11 .02 .06 .05 -.01 -.01 P25 .19 -.06 -.01 -.02 -.02 -.14 -.02 -.11 -.01 -.08 -.03 .05 P26 -.02 .03 .06 .11 .09 .09 .09 .04 .01 .02 -.04 .01 P27 .01 .10 .17 .01 .14 .09 .17 .09 .07 .10 -.01 .02 P28 .11 .20 .06 .01 .12 .01 .12 .08 .13 .14 .05 .08 P29 .11 .15 .25 .04 .09 .26 .11 .24 .13 .26 .17 .11 P30 .05 .15 .19 .06 .07 .21 .15 .21 .11 .18 .04 .08 P31 .10 .31 .16 .06 .11 .21 .24 .22 .26 .28 .18 .12 P32 .11 .35 .10 .07 .17 .19 .18 .21 .25 .23 .19 .15 P33 .01 .05 .10 .08 .09 .04 .15 .06 .10 .10 .01 .06

Page 240: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

230

Table 28 (continued)

P69 P70 P71 P72 P73 P74 P75 P76 P77 P78 P79 P80 P34 .08 .02 .01 .05 .01 .01 .04 -.02 .02 .04 .09 .04 P35 -.04 -.03 -.03 .04 -.02 -.03 -.04 .03 -.04 -.04 .01 -.01 P36 -.04 .09 .04 .05 .17 .11 .13 .11 .01 .12 .02 -.03 P37 .09 .22 .14 .09 .12 .23 .18 .21 .20 .29 .21 .10 P38 .19 .16 .17 .08 .08 .18 .11 .25 .24 .30 .27 .07 P39 .13 .18 .11 .08 .03 .06 .14 .15 .16 .11 .04 .13 P40 .12 .31 .14 .02 .07 .11 .18 .24 .22 .22 .15 .18 P41 .09 .15 .29 -.01 .06 .22 .11 .19 .16 .28 .14 .06 P42 .15 .25 .08 .11 .16 .07 .13 .21 .20 .22 .19 .17 P43 .12 .07 .03 .13 .08 .13 .14 .10 .10 .07 .02 -.03 P44 .17 .13 .03 .08 .04 .08 .08 .20 .15 .11 .12 .12 P45 .13 .16 .19 .08 .14 .22 .17 .27 .23 .28 .20 .08 P46 .03 .14 .12 .01 .11 .03 .14 .10 .12 .11 .01 .08 P47 .12 .32 .17 .11 .12 .13 .28 .22 .31 .23 .06 .15 P48 .16 .32 .14 .01 .08 .19 .20 .27 .23 .32 .22 .09 P49 .14 .40 .09 .06 .08 .07 .18 .20 .24 .22 .17 .27 P50 .02 .16 .13 .08 .08 .19 .16 .23 .18 .23 .17 .07 P51 .04 .10 .19 .06 .05 .51 .11 .23 .08 .17 .13 -.04 P52 .01 .30 .10 .05 .13 .19 .15 .28 .21 .24 .18 .10 P53 .07 .01 .02 .03 .15 -.10 .09 -.02 .14 .06 .04 -.02 P54 .04 -.04 .03 .03 -.02 -.08 .04 -.06 -.02 -.12 -.11 .10 P55 .12 .21 .14 .05 .13 .08 .12 .16 .21 .17 .10 .18 P56 .21 .02 .07 .04 .05 -.01 .05 .13 .14 .11 .11 .08 P57 .09 .34 .12 .12 .20 .17 .27 .22 .19 .25 .06 .17 P58 .07 .11 .26 .02 .03 .29 .07 .12 .14 .15 .12 .04 P59 .18 .15 .07 .04 .16 .05 .11 .13 .17 .20 .05 .17 P60 -.01 -.05 .01 .09 .10 .10 .06 .04 .04 .07 -.06 -.12 P61 .07 .38 .17 .04 .08 .14 .24 .21 .26 .16 .16 .17 P62 .21 .23 .20 .05 .04 .13 .15 .19 .22 .20 .17 .17 P63 .12 .19 .14 .07 .10 .28 .13 .30 .16 .20 .12 -.03 P64 .10 .20 .08 .13 .10 .07 .21 .15 .15 .13 .01 .18

Page 241: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

231

Table 28 (continued) P69 P70 P71 P72 P73 P74 P75 P76 P77 P78 P79 P80

P65 .05 .11 .07 .08 .06 .10 .10 .06 .09 .04 .01 .03 P66 .14 .38 .16 .07 .10 .10 .21 .20 .21 .23 .16 .19 P67 .09 .46 .13 .04 .11 .13 .19 .21 .23 .21 .16 .19 P68 .19 .27 .10 .10 .12 .10 .28 .18 .26 .24 .15 .17 P69 - .01 .07 .01 .09 .04 .09 .08 .18 .18 .18 .08 P70 - .11 .01 .12 .15 .18 .16 .21 .22 .19 .20 P71 - .09 .11 .25 .12 .11 .07 .19 .07 .03 P72 - .21 .11 .06 .12 .01 .12 -.01 .02 P73 - .10 .20 .04 .12 .21 .10 .05 P74 - .11 .17 .07 .19 .12 -.05 P75 - .18 .22 .19 .11 .13 P76 - .19 .23 .11 .11 P77 - .26 .24 .09 P78 - .26 .13 P79 - .13

Page 242: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

232

Appendix D

Model Identification Criteria

Table 29 Theoretical Identification Criteria for Conners’ Teacher Models Conners’ Teacher Model Total

Factors Total Items

Free Parameters Observations

Conners-F11-T 11 59 204 1,170 Conners-E6-T 6 38 85 741 Statistical-2T 2 16 31 136 Rational-2T 2 20 39 210 Hybrid-2T 2 12 23 78

Table 30 Theoretical Identification Criteria for Conners’ Parent Models Conners’ Parent Model Total

Factors Total Items

Free Parameters Observations

Conners-F12-P 12 80 214 3,645 Conners-E7-P 7 57 128 1,653 Extended-2P 2 23 45 276 Abbreviated-2P 2 16 31 136 Hybrid-2P 2 11 21 66

Page 243: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

233

Appendix E

Overview of Conners’ Items

Table 31

Overview of CTRS-R Items

Items Subscale Derivation 1. Acts defiantly Oppositional Empirical 2. Restless and squirmy Hyperactivity Empirical 3. Forgets material already learned

Cognitive Problems/ Inattention; DSM-Inattentive

Both

4. Unaccepted by group Social Problems Empirical 5. Easily hurt/offended Anxiety Empirical 6. Perfectionistic Perfectionism Empirical 7. Temper outbursts Oppositional; CGI- Emotional Lability Both 8. Impulsive, easily excited Hyperactivity; ADHD Index Rational 9. Careless mistakes in work DSM-Inattentive Rational 10. Overly bold Oppositional Empirical 11. On the go/ driven by motor Hyperactivity; DSM-Hyperactive Both 12. Avoids sustained mental effort

Cognitive Problems/ Inattention; DSM-Inattentive

Both

13. Is chosen last for teams Social Problems Empirical 14. Is emotional Anxiety Empirical 15. Everything must be perfect Perfectionism Empirical 16. Restless/ highly active CGI- Restless/ Impulsive Rational 17. Fails to finish things started Cognitive Problems/ Inattention; ADHD Index Both 18. Does not seem to listen DSM-Inattentive Rational 19. Actively defies/ refuses to comply

Oppositional Empirical

20. Leaves seat Hyperactivity Empirical 21. Poor spelling Cognitive Problems/ Inattention Empirical 22. Doesn’t have friends Social Problems Empirical 23. Timid, easily scared Anxiety Empirical 24. Checks things repeatedly Perfectionism Empirical 25. Cries frequently and easily Anxiety; CGI-Emotional Lability Both 26. Inattentive, distractible ADHD Index Rational 27. Difficulty organizing tasks DSM-Inattentive Rational 28. Difficulty sustaining attention

DSM-Inattentive Rational

29. Difficulty awaiting turn Hyperactivity; DSM-hyperactive Both 30. Not reading well Cognitive Problems/ Inattention Empirical 31. Does not make friends Social Problems Empirical 32. Sensitive if criticized Anxiety Empirical

Page 244: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

234

Table 31 (continued)

Items Subscale Derivation 33. Over-focused on details Perfectionism Empirical 34. Talks too much DSM-hyperactive Rational 35. Fidgets CGI-restless/ impulsive Rational 36. Disturbs peers ADHD Index Rational 37. Argues with adults Oppositional Empirical 38. Cannot stay still ADHD Index Rational 39. Runs/climbs excessively Hyperactivity Empirical 40. Lacks interest in school Cognitive Problems/ Inattention Empirical 41. Poor social skills Social Problems Empirical 42. Difficulty playing quietly Hyperactivity; DSM-hyperactive Both 43. Likes things clean/orderly Perfectionism Empirical 44. Fidgets with hands or feet ADHD Index; DSM-hyperactive Both 45. Demanding/ easily frustrated

CGI-emotional lability Rational

46. Blurts out answers DSM-hyperactive Rational 47. Spiteful/ vindictive Oppositional Empirical 48. Poor attention span ADHD Index Rational 49. Loses things Cognitive Problems/ Inattention; DSM-

Inattentive Both

50. Only attends if interested ADHD Index Rational 51. Timid, withdrawn Anxiety Empirical 52. Distractibility/ attn. problem

ADHD Index Rational

53. Always does things same way

Perfectionism Empirical

54. Mood changes quickly CGI-Restless Impulsive Rational 55. Interrupts/ intrusive ADHD Index; DSM-hyperactive Rational 56. Poor in math Cognitive Problems/ Inattention Empirical 57. Does not follow through ADHD Index/ DSM-Inattentive Rational 58. Easily distracted DSM-Inattentive Rational 59. Restless, always on the go ADHD Index Rational

Page 245: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

235

Table 32

Overview of CPRS-R Items Items Subscale Derivation 1. Angry/ resentful Oppositional Empirical 2. Difficulty finishing homework

Cognitive Problems/ Inattention Empirical

3. On the go/driven by a motor Hyperactivity; DSM-Hyperactive Both 4. Timid, easily scared Anxious-Shy Empirical 5. Everything must be perfect Perfectionism Empirical 6. Doesn’t have friends Social Problems Empirical 7. Stomach pain Psychosomatic Empirical 8. Fights Oppositional Empirical 9. Avoids sustained mental effort

Cognitive Problems/ Inattention; ADHD Index; DSM-Inattentive

Both

10.Difficulty sustaining attention

DSM-Inattentive Rational

11. Argues Oppositional Empirical 12. Fails to finish assignments Cognitive Problems/ Inattention Empirical 13. Hard to control in stores Hyperactivity Empirical 14. Scared of people Anxious-Shy Empirical 15. Checks things repeatedly Perfectionism Empirical 16. Loses friends quickly Social Problems Empirical 17. Aches/ pains Psychosomatic Empirical 18. Restless/ highly active CGI: Restless/ Impulsive Rational 19. Trouble concentrating in class

Cognitive Problems/ Inattention; ADHD Index

Both

20. Does not seem to listen DSM-Inattentive Rational 21. Loses temper Oppositional Empirical 22. Needs close supervision Cognitive Problems/ Inattention Empirical 23. Runs/climbs excessively Hyperactivity; DSM-Hyperactive Both 24. Scared of new situations Anxious-Shy Empirical 25. Picky about cleanliness Perfectionism Empirical 26. Does not make friends Social Problems Empirical 27. Gets aches/ pains before school

Psychosomatic Empirical

28. Impulsive, easily excited Hyperactivity; CGI: Restless/ Impulsive Both

Page 246: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

236

Table 32 (continued)

Items Subscale Derivation 29. No follow through on instructions, fails to finish work

Cognitive Problems/ Inattention; ADHD Index; DSM-Inattentive

Both

30. Has difficulty organizing DSM-Inattentive Rational 31. Irritable Oppositional Empirical 32. Restless and squirmy Hyperactivity Empirical 33. Scared of being alone Anxious-Shy Empirical 34. Always does things same way

Perfectionism Empirical

35. Does not get invited Social Problems Empirical 36. Headaches Psychosomatic Empirical 37. Fails to finish things started

CGI: Restless/ Impulsive Rational

38. Inattentive, distractible ADHD Index; CGI: Restless/ Impulsive Rational 39. Talks too much DSM-Hyperactive Rational 40. Actively defies adults Oppositional Empirical 41. Careless mistakes in work Cognitive Problems/ Inattention; DSM-

Inattentive Both

42. Difficulty awaiting turn DSM-Hyperactive Rational 43. Has many fears Anxious-Shy Empirical 44. Has rituals Perfectionism Empirical 45. Distractibility/ attention problem

ADHD Index Rational

46. Complains/sick – when well

Psychosomatic Empirical

47. Temper outbursts CGI: Emotional Lability Rational 48. Gets distracted w/ instructions

ADHD Index Rational

49. Interrupts/ intrusive DSM-Hyperactive Rational 50. Forgetful in everyday activities

Cognitive Problems/ Inattention; DSM-Inattentive

Both

51. Cannot understand math Cognitive Problems/ Inattention Empirical 52. Will run between mouthfuls

Hyperactivity Empirical

53. Afraid of dark/ animals/ bugs

Anxious-Shy Empirical

54. Sets high goals for self Perfectionism Empirical

Page 247: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

237

Table 32 (continued)

Items Subscale Derivation 55. Fidgets with hands/ feet ADHD Index; DSM-Hyperactivity Rational 56. Poor attention span ADHD Index Rational 57. Touchy/ easily annoyed Oppositional Empirical 58. Has messy handwriting Cognitive Problems/ Inattention Empirical 59. Has difficulty playing quietly

Hyperactivity; DSM-Hyperactive Both

60. Timid, withdrawn Anxious-Shy Empirical 61. Blames others for mistakes Oppositional Empirical 62. Fidgets CGI: Restless/ Impulsive Rational 63. Messy or disorganized ADHD Index Rational 64. Gets upset if rearranged things

Perfectionism Empirical

65. Clings to parents/ adults Anxious-Shy Empirical 66. Disturbs peers CGI: Restless/ Impulsive Rational 67. Deliberately annoys others Oppositional Empirical 68. Demanding easily frustrated

CGI: Restless/ Impulsive Rational

69. Only attends if interested ADHD Index Rational 70. Spiteful/ vindictive Oppositional Empirical 71. Loses things DSM-Hyperactive Rational 72. Feels inferior Social Problems Empirical 73. Seems tired/ slowed down Psychosomatic Empirical 73. Poor spelling Cognitive Problems/ Inattention Empirical 75. Cries frequently and easily CGI: Emotional Lability Rational 76. Leaves seat in classroom ADHD Index; DSM-Hyperactive Rational 77. Mood changes quickly CGI: Emotional Lability Rational 78. Easily frustrated in tasks ADHD Index Rational 79. Easily distracted DSM-Inattentive Rational 80. Blurts out answers Hyperactivity; DSM-Hyperactivity Both

Page 248: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

238

Appendix F

Glossary of Acronyms

Table 33

Glossary of Acronyms

Abbreviated-2P Moderate Criteria Rational Abbreviated Solution, Parent Scale

ADHD Attention Deficit Hyperactivity Disorder

ADHD-C Attention Deficit Hyperactivity Disorder – Combined Subtype

ADHD-PH Attention Deficit Hyperactivity Disorder – Predominantly Hyperactive/Impulsive Subtype

ADHD-PI Attention Deficit Hyperactivity Disorder – Predominantly Inattentive Subtype

ADORE Attention Deficit Hyperactivity Disorder Observational Research in Europe

AGFI Adjusted Goodness of Fit Index

AIC Akaike Information Criterion

APA American Psychological Association

CAIC Consistent Akaike Information Criteria

CBCL Child Behavior Checklist

CBS Central Bureau of Statistics

CGI:EL Conners’ Global Index: Emotional Lability

CGI:RI Conners’ Global Index: Restless/Impulsive

CFA Confirmatory Factor Analysis

CFI Comparative Fit Index

CIA Central Intelligence Agency

Page 249: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

239

Table 33 (cont.)

Conners-E6-T

Conners’ Teacher Empirical Six-Factor Model

Conners-E7-P Conners’ Parent Empirical Seven-Factor Model

Conners-F11-T Conners’ Teacher Full Eleven-Factor Model

Conners-F12-P Conners’ Parent Full Twelve-Factor Model

CP/I Cognitive Problems/Inattention

CPRS-R Conners’ Parent Rating Scale – Revised

CRS-R Conners’ Rating Scales – Revised

CTRS-R Conners’ Teacher Rating Scale – Revised

DISC-IV Diagnostic Interview Schedule for Children, Fourth Edition

DOF Direct Observation Form of Child Behavior Checklist

DSM:HI DSM-IV Hyperactive/Impulsive Subscale

DSM:IA DSM-IV Inattentive Subscale

DSM-IV TR Diagnostic and Statistical Manual of Mental Disorders – Fourth Edition, Text Revision

EFA Exploratory Factor Analysis

Extended-2P Moderate Criteria Rational Extended Solution, Parent Scale

GFI Goodness of Fit Index

Hybrid-2P Moderate Criteria Statistical/Rational Hybrid Solution, Parent Scale

Hybrid-2T Moderate Criteria Statistical/Rational Hybrid Solution, Teacher Scale

ICG International Crisis Group

Page 250: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

240

Table 33 (continued)

IFI

Incremental Fit Index

IRT Item Response Theory

KMO Kaiser-Meyer-Olkin Statistic

MAP Minimum Average Partials

NNFI Non-Normed Fit Index

NNIPS Nepal Nutrition Intervention Project – Sarlahi

PA Parallel Analysis

PAF Principal Axis Factoring

PCA Principal Components Analysis

Rational-2T Moderate Criteria Rational Solution, Teacher Scale

RMS Root Mean Square Residual

RMSEA Root Mean Error of Approximation

S-B χ2 Satorra-Bentler Chi-Square

SP Social Problems

Statistical-2T Moderate Criteria Statistical Solution, Teacher Scale

TLI Tucker-Lewis Index

Page 251: FACTOR STRUCTURE OF SCORES FROM THE CONNERS’ …

VITA

4150 Harbour Dr. Palmyra, NJ 08065 [email protected]

217-549-0234

Education: 2006 – Present The Pennsylvania State University M.Ed. (May 2008), PhD. (exp, August, 2011) University Park, PA School Psychology 2002 – 2006

Eastern Illinois University BA (May, 2006; GPA – 3.7) Charleston, IL Psychology

Publications: Pendergast, L., & Watkins, M. W. (2009). Development of an Electronic Version of the Homework

Performance Questionnaire. Journal of Educational Computing Research, 40, 323 – 335. Research Experience:

• Research Assistant, Penn State Child Attention and Learning Lab, 2008 – 2009 • Research Assistant, Nepal Nutrition Intervention Project – Sarlahi, cognition studies division,

Penn State University, 2007 - 2008 • Research Assistant, School Psychology Program, Penn State University, 2006 - 2007 • Presenter, Annual National Association of School Psychologists (NASP) conferences: 2007,

2009, 2010. Clinical Experience:

• Doctoral School Psychology Intern, CORA Services, Inc., 2010 - present • CEDAR Clinic Student Supervisor, Penn State CEDAR Clinic, 2009 - 2010 • School Psychology Practicum Intern, State College Area School District, 2008 – 2009 • School Psychology Student Clinician, Penn State CEDAR Clinic, 2007 - 2009

Teaching Experience: • Co-instructor, Multi-cultural counseling course, The Pennsylvania State University, 2010 • Guest instructor, Multivariate statistics and professional writing course, Penn State University,

2008 – 2009 (as needed) • Guest instructor, Academic writing course for adult English Language Learners, Penn State

University, 2008 Work Experience:

• Developmental Therapy Intern, CCAR Industries – Birth through Three Program, Charleston, IL, 2004 - 2005

• Day Care Teacher, Living Waters Children’s Center, Manhattan, IL, 2003 - 2006 Awards/Grants:

• Specialization in Culture and Language Education (SCALE) fellowship, Penn State, 2008 • Fred L. Yaffe scholarship for achievement in research, Eastern Illinois University, 2006 • Eastern Illinois Honors College research grant, 2004

Research Interests: Psychoeducational assessment, scale development, diagnosis and assessment of ADHD, diversity issues in individuals with ADHD, education of English Language Learners, multicultural education, the role of homework in education, learning behaviors and academic achievement

Laura L. Pendergast