Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Construct validity of the K-ABC for Cantonese, English and Punjabi speaking Canadians Gardner, JoAnne Marie 1986

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1986_A2 G37.pdf [ 12.04MB ]
Metadata
JSON: 831-1.0054558.json
JSON-LD: 831-1.0054558-ld.json
RDF/XML (Pretty): 831-1.0054558-rdf.xml
RDF/JSON: 831-1.0054558-rdf.json
Turtle: 831-1.0054558-turtle.txt
N-Triples: 831-1.0054558-rdf-ntriples.txt
Original Record: 831-1.0054558-source.json
Full Text
831-1.0054558-fulltext.txt
Citation
831-1.0054558.ris

Full Text

CONSTRUCT V A L I D I T Y OF THE K - A B C FOR C A N T O N E S E , E N G L I S H , AND P U N J A B I S P E A K I N G CANADIAN C H I L D R E N By JOANNE MARIE GARDNER B . S c , The U n i v e r s i t y o f V i c t o r i a , 1977 D i p . E d u c . , The U n i v e r s i t y o f V i c t o r i a , 1979 M . A . , The U n i v e r s i t y o f B r i t i s h C o l u m b i a , 1981 A T H E S I S SUBMITTED I N P A R T I A L F U L F I L L M E N T OF THE REQUIREMENTS FOR THE DEGREE OF DOCTORATE OF EDUCATION i n THE F A C U L T Y OF GRADUATE S T U D I E S D e p a r t m e n t o f E d u c a t i o n a l P s y c h o l o g y a n d S p e c i a l E d u c a t i o n We a c c e p t t h i s t h e s i s a s c o n f o r m i n g t o t he - r e a u i r e d s t a n d a r d THE U N I V E R S I T Y OF B R I T I S H COLUMBIA May, 1986 © J o A n n e M a r i e G a r d n e r , 1986 In p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f the requirements f o r an advanced degree a t the U n i v e r s i t y o f B r i t i s h Columbia, I agree t h a t the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and study. I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e copying o f t h i s t h e s i s f o r s c h o l a r l y purposes may be granted by the head of my department or by h i s o r her r e p r e s e n t a t i v e s . I t i s understood t h a t copying o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l gain s h a l l not be allowed without my w r i t t e n p e r m i s s i o n . Department of £clui_ ^ H-cWc The U n i v e r s i t y of B r i t i s h Columbia 2075 Wesbrook P l a c e Vancouver, Canada, V6T 1W5 D a t e r i i ABSTRACT The purpose of t h i s study was to examine the c o n s t r u c t v a l i d i t y of the Kaufman Assessment B a t t e r y f o r C h i l d r e n (K-ABC) f o r use with Cantonese, E n g l i s h , and Punjabi speaking Canadians. The K-ABC i s a r e l a t i v e l y new, i n d i v i d u a l l y a d m i n i s t e r e d t e s t of i n t e l l i g e n c e and achievement. The i n t e l l i g e n c e s c a l e i s being promoted as measuring a mental p r o c e s s i n g dichotomy (Se q u e n t i a l / S i m u l t a n e o u s ) . A sample of 210 students (70 i n each of the three groups) between the ages of 8 years, 1 month and 10 years, 5 months v o l u n t e e r e d t o p a r t i c i p a t e i n the study. A l l were e n r o l l e d i n grade 3 c l a s s e s (n = 34) i n a l a r g e urban c i t y i n Western Canada. The s u b j e c t s (sexes e q u a l l y represented w i t h i n each of the three groups) were a l l Canadian born, attended E n g l i s h s c h o o l s , were not Native Indians, and had not been p r e v i o u s l y diagnosed as having emotional, mental, p h y s i c a l or sensory handicaps. Each student was admi n i s t e r e d the K-ABC and WISC-R. Information on t h e i r biodemographic c h a r a c t e r i s t i c s was c o l l e c t e d from t h e i r parents and te a c h e r s . S p e c i f i c a l l y , the parents completed a q u e s t i o n n a i r e addressing such i s s u e s as the language(s) spoken i n the home, t h e i r b i r t h p l a c e , family size, and socioeconomic status. Teachers were required to rate the students' English fluency and learning style. Confirmatory and exploratory factor analyses were performed on the K-ABC, for each group, to investigate its internal structure. Pearson's correlation coefficients and dependent t test comparisons were performed to identify the relationship between the K-ABC and the WISC-R. The differences found among the groups on the two cognitive tests and the significant discrepancies found between each test for specific individuals were explained in relation to group and individual biodemographic characteristics. The scales on both tests were found to be reliable measures for each group. The subtest and scale means on the K-ABC and WISC-R differed significantly among the groups. Factor Analyses on the K-ABC indicated the English and Punjabi data as supporting the theoretical underpinnings (Sequential Processing and Simultanous Processing) of the K-ABC while the Cantonese data did not. High correlations between the K-ABC Mental Processing Composite and WISC-R Full Scale IQ suggests the two tests are measuring similar constructs for English and Punjabi children. The moderate correlation between these two tests iv f o r the Cantonese suggests the K-ABC and WISC-R may not be measuring i n t e l l i g e n c e the same way. An i n v e s t i g a t i o n o f the biodemographic c h a r a c t e r i s t i c s of each group i n d i c a t e d t h a t c u l t u r a l and l i n g u i s t i c f a c t o r s might be c o n t r i b u t i n g t o the d i f f e r e n t i a l performance of the three groups on the K-ABC and WISC-R. Moreover, the i m p l i c a t i o n s of s i g n i f i c a n t d i s c r e p a n c i e s between the K-ABC and WISC-R i n t e l l i g e n c e s c a l e s are d i s c u s s e d . F i n a l l y , the c o n s t r u c t v a l i d i t y (concurrent v a l i d i t y and i n t e r n a l s t r u c t u r e ) o f the K-ABC f o r use with E n g l i s h and Punjabi Canadians was c o n s i d e r e d a c c e p t a b l e ; however, i t s use with Cantonese Canadians remains q u e s t i o n a b l e . V TABLE OF CONTENTS Page ABSTRACT i i LIST OF TABLES v i i LIST OF APPENDICES X ACKNOWLEDGEMENT x i i CHAPTER I INTRODUCTION 1 K-ABC: Basis for Development 3 Definitions 9 Purpose of Study 9 Delimitations of Study 11 Organization of Thesis 11 CHAPTER II REVIEW OF LITERATURE 12 Theoretical Basis for Measuring Intelligence 13 Problem-Solving versus Acquired Factual Knowledge 24 Educational Intervention . 32 Ease of Administration and Scoring 36 Accommodation of Diverse Populations 39 CHAPTER III INSTRUMENTATION 66 Kaufman Assessment Battery for Children 66 Wechsler Intelligence Scale for Children -Revised 71 Parent Questionnaire 72 Teacher Questionnaire 77 Teacher Rating Scale 79 • CHAPTER IV METHODOLOGY 80 Population 80 Samples 81 Data Collection 84 Scoring and Data Preparation 88 Preliminary Analysis 90 Statistical Methods 93 CHAPTER V DESCRIPTION OF SAMPLES 95 Rate of Response 95 Biodemographic Characteristics of Samples 97 CHAPTER VI PSYCHOMETRIC PROPERTIES OF THE K-ABC 117 Central Tendency and Variability 117 Reliability 122 Internal Structure 129 v i TABLE OF CONTENTS, cont. Page CHAPTER VII RELATIONSHIP BETWEEN THE K-ABC and WISC-R 151 WISC-R: Psychometric P r o p e r t i e s 151 K-ABC versus WISC-R 156 CHAPTER VII I INTERPRETATION OF GROUP AND TEST DIFFERENCES 175 I n t e r p r e t a t i o n o f Group D i f f e r e n c e s 176 I n t e r p r e t a t i o n o f Tes t D i f f e r e n c e s 183 CHAPTER IX SUMMARY, CONCLUSIONS AND RECOMMENDATIONS ..189 Summary of the Study 189 I m p l i c a t i o n s f o r Using the K-ABC 193 I m p l i c a t i o n s f o r A s s e s s i n g M i n o r i t y C h i l d r e n 197 L i m i t a t i o n s o f the Study 204 Recommendations f o r F u r t h e r Research 205 REFERENCES 209 APPENDICES 224 v i i LIST OF TABLES Table Page 1 D e s c r i p t i o n o f K-ABC s u b t e s t s 67 2 K-ABC su b t e s t s by age admi n i s t e r e d 68 3 D e s c r i p t i o n o f WISC-R s u b t e s t s 73 4 Number of s u b j e c t s each examiner t e s t e d by language group 91 5 Response r a t e 96 6 Biodemographic c h a r a c t e r i s t i c s of the students by language group 99 7 Number of parents and grandparents born i n Canada by language group 101 8 Parents' l e n g t h o f r e s i d e n c e i n Canada by language group . 102 9 Community s i z e of pa r e n t s ' b i r t h l o c a t i o n by language group 104 10 Socioeconomic s t u d i e s (SES) by language group 105 11 Languages spoken by f a m i l y member 107 12 Frequency o f spoken E n g l i s h by language group 108 13 Means, standard d e v i a t i o n s and Tukey comparisons f o r E n g l i s h Fluency items by language group 110 14 Grade s u b j e c t s entered p r e s e n t s c h o o l by language group 112 15 Means, standard d e v i a t i o n s and Tukey comparisons f o r the Teacher Rating S c a l e by language group 113 16 Number and percentage o f s u b j e c t s r e c e i v i n g remediation by language group 115 v i i i LIST OF TABLES, cont Table Page 17 K-ABC means, standard d e v i a t i o n s , and Tukey comparisons f o r each group 118 18 K-ABC i n t e r n a l c o n s i s t e n c y r e l i a b i l i t i e s ( r x x ) and standard e r r o r s of measurement (SEM) f o r each group 124 19 C o r r e c t e d and s e l e c t e d uncorrected c o r r e l a t i o n s between noncomposite K-ABC G l o b a l S c a l e Standard Scores f o r each group ...128 20 C o r r e l a t i o n s between s e l e c t e d K-ABC Composite Sca l e s f o r each group 130 21 T h e o r e t i c a l t a r g e t matrix f o r the two f a c t o r and t h r e e f a c t o r model 133 22 Confirmatory F a c t o r A n a l y s i s o f Mental P r o c e s s i n g s u b t e s t s f o r each group 134 23 Confirmatory F a c t o r A n a l y s i s of Mental P r o c e s s i n g and Achievement s u b t e s t s f o r each group 136 24 Number of f a c t o r s i d e n t i f i e d f o r Mental P r o c e s s i n g and Mental P r o c e s s i n g and Achievement s u b t e s t s f o r each group ........ 140 25 F a c t o r l o a d i n g s f o r the two f a c t o r unweighted l e a s t squares a n a l y s i s with a varimax r o t a t i o n f o r the Mental P r o c e s s i n g s u b t e s t s f o r each group 143 26 F a c t o r l o a d i n g s f o r the three f a c t o r unweighted l e a s t squares a n a l y s i s with a varimax r o t a t i o n f o r the Mental P r o c e s s i n g and Achievement s u b t e s t s f o r each group 147 27 WISC-R means, standard d e v i a t i o n s and i n t e r n a l c o n s i s t e n c y r e l i a b i l i t i e s f o r each group 152 i x LIST OF TABLES, cont Table Page 28 C o r r e l a t i o n s between the K-ABC and WISC-R sub t e s t s s c a l e d scores f o r Cantonese s u b j e c t s 158 29 C o r r e l a t i o n s between the K-ABC and WISC-R su b t e s t s s c a l e d scores f o r E n g l i s h s u b j e c t s 159 30 C o r r e l a t i o n s between the K-ABC and WISC-R su b t e s t s s c a l e d scores f o r Punjabi s u b j e c t s 160 31 Means and standard d e v i a t i o n s f o r the K-ABC G l o b a l S c a l e s and WISC-R IQs by language group 164 32 Dependent t - t e s t comparisons between K-ABC Standard Scores and WISC-R IQs 166 33 D i s t r i b u t i o n a l i n d i v i d u a l d i s c r e p a n c i e s between the Mental P r o c e s s i n g Composite and F u l l S c a l e IQ f o r each group 167 34 C o r r e l a t i o n s between the K-ABC G l o b a l S c a l e s and WISC-R IQs f o r each group 169 X LIST OF APPENDICES Appendix Page A. INSTRUMENTS 224 A - l Parent Q u e s t i o n n a i r e 225 A-2 Teacher Q u e s t i o n n a i r e 229 A-3 Teacher Rat i n g S c a l e 230 B. LETTERS TO PRINCIPALS AND TEACHERS 231 B - l P r i n c i p a l Information L e t t e r 232 B-2 P r i n c i p a l Thank-You L e t t e r 234 B-3 Teacher Thank-You L e t t e r 235 C PARENT INFORMATION PACKAGE 236 C - l E n g l i s h Package 237 C-2 Cantonese Package 244 C-3 Punjabi Package 252 D. OUTLINE OF K-ABC TRAINING WORKSHOP 260 D-1 K-ABC T r a i n i n g Workshop 261 E. ITEM CHANGES 262 E - l K-ABC 263 E-2 WISC-R 264 F. MATERIAL CONTAINED WITHIN TESTER'S PACKAGE 265 F - l T e s t i n g Procedure Sheet 266 F-2 Cover Sheet 270 F-3 Request f o r P a r t i c i p a t i o n 271 F-4 Parent Consent Form 272 F-5 C h e c k l i s t 273 F-6 Teacher Q u e s t i o n n a i r e 274 F-7 Teacher Rat i n g S c a l e 275 F-8 Parent L e t t e r 276 G. TEST ORDER EFFECT 277 G-l K-ABC 278 G-2 WISC-R 279 H. GROUP DIFFERENCES ON K-ABC 280 H-l A n a l y s i s o f Va r i a n c e 281 I. K-ABC SUBTEST INTERCORRELATIONS 282 1-1 Cantonese 283 1-2 E n g l i s h 284 1-3 Punjabi 285 x i LIST OF APPENDICES, cont Appendix Page J . SCREE TESTS 286 Menta l P r o c e s s i n g Composite J - l Cantonese 287 J - 2 E n g l i s h 288 J - 3 P u n j a b i 289 Menta l P r o c e s s i n g and Achievement J - 4 Cantonese 290 J - 5 E n g l i s h 291 J - 6 P u n j a b i 292 K. FACTOR ANALYSIS - CANTONESE 293 K - l P r i n c i p a l Components - 2 f a c t o r 294 K-2 Unweighted L e a s t Squares - 3 f a c t o r 295 K-3 P r i n c i p a l Components and Unweighted L e a s t Squares - 4 f a c t o r 296 L . GROUP DIFFERENCES ON WISC-R 297 L - l A n a l y s i s o f V a r i a n c e 298 x i i ACKNOWLEDGEMENT To Drs. Stan Blank, J u l i e Conry, and Vin c e n t D'Oyley, my committee members, I would l i k e to express my a p p r e c i a t i o n f o r t h e i r support and guidance. I would l i k e to thank the c o n s u l t a n t s who c o n t r i b u t e d to t h i s study: Drs. Bob Conry, Ron Jarman, Bernie Mohan, Todd Rogers, Jim S t e i g e r , Anne Watson-Russell, Mr. Bob Prossor and Ms. Marsha Shroeder. The f i n a n c i a l support from the E d u c a t i o n a l Research I n s t i t u t e of B r i t i s h Columbia was g r e a t l y a p p r e c i a t e d . The p a r t i c i p a t i o n of the 210 students, t h e i r p a r e n t s , teachers, and p r i n c i p a l s made t h i s study happen. A s p e c i a l thanks to Nona Kent f o r her pat i e n c e i n typi n g t h i s t h e s i s . And, to my f r i e n d s f o r t h e i r encouragement, I thank you. 1 CHAPTER 1 I n t r o d u c t i o n The Kaufman Assessment B a t t e r y f o r C h i l d r e n (K-ABC) (Kaufman & Kaufman, 1983a) i s a r e c e n t l y developed, i n d i v i d u a l l y a d m i n i s t e r e d t e s t of i n t e l l i g e n c e and achievement f o r c h i l d r e n from 2 1/2 t o 12 1/2 years of age. U n l i k e p r e v i o u s l y developed t e s t s , the i n t e l l i g e n c e s c a l e of the K-ABC was e x p l i c i t l y based upon a theory o f i n f o r m a t i o n p r o c e s s i n g , namely: s e q u e n t i a l and simultaneous p r o c e s s i n g . F u r t h e r , the i n c l u s i o n o f an achievement s c a l e , which p r o v i d e s a measure of ac q u i r e d knowledge, language a c q u i s i t i o n , and sch o o l l e a r n i n g , p r o v i d e s an added advantage of being able t o i n v e s t i g a t e the r e l a t i o n s h i p between i n t e l l i g e n c e and achievement. As such, t h i s novel approach t o c o g n i t i v e assessment has i m p l i c a t i o n s f o r d i a g n o s i n g and dev e l o p i n g e d u c a t i o n a l i n t e r v e n t i o n programs f o r c h i l d r e n . In Canada, the K-ABC has r e c e i v e d c o n s i d e r a b l e i n t e r e s t and a t t e n t i o n f o r use i n both academic and a p p l i e d s e t t i n g s . However, the K-ABC was not developed or s t a n d a r d i z e d i n Canada. Moreover, none of the 43 v a l i d i t y s t u d i e s r e p o r t e d i n the K-ABC's I n t e r p r e t i v e Manual (IM) (Kaufman & Kaufman, 1983b) were conducted on Canadian c h i l d r e n . Consequently, the lack o f data on the performance of Canadian c h i l d r e n on the K-ABC 2 r a i s e s q u e s t i o n s about the t e s t ' s u t i l i t y f o r use with Canadian c h i l d r e n . There i s evidence to suggest t h a t Canadian c h i l d r e n perform d i f f e r e n t l y from American c h i l d r e n on other measures of i n t e l l i g e n c e , such as the Wechsler I n t e l l i g e n c e S cale f o r C h i l d r e n - Revised (WISC-R) (Holmes, 1981; Pet e r s , 1976; Vernon, 1977) and The Lorge-Thorndike I n t e l l i g e n c e T e s t (Wright & H a r r i s , 1972). S p e c i f i c a l l y , Canadian c h i l d r e n have been found on these two t e s t s to have s i g n i f i c a n t l y h i g h e r means and s m a l l e r v a r i a n c e s than c h i l d r e n i n the American s t a n d a r d i z a t i o n group. P o s s i b l e reasons f o r the d i f f e r e n c e i n performance between the c h i l d r e n i n the two c o u n t r i e s may i n c l u d e c u l t u r a l , demographic, economic, e d u c a t i o n a l , and s o c i a l f a c t o r s t h a t a f f e c t the c h i l d r e n ' s knowledge and understanding of the t e s t c ontent. The concern f o r the a p p r o p r i a t e n e s s of using American t e s t s f o r a s s e s s i n g Canadian c h i l d r e n has r e s u l t e d i n c l a i m s t h a t American t e s t s may be b i a s e d a g a i n s t Canadian c h i l d r e n (Wormeli, 1984). Subsequently, attempts have been made to Canadianize q u e s t i o n a b l e items (Vernon, 1977; V i o l a t o , 1984), develop Canadian norms (Holmes, 1981) and develop t e s t s with Canadian content (Wormeli, 1984). However, when a new t e s t l i k e the K-ABC i s developed, i t s v a l i d i t y ( i . e . , the accuracy with which i t measures what i t i s purported to assess) needs to be determined ( A n a s t a s i , 1976), w e l l b e f o r e changes are made to i t s content, and, thus, b e f o r e new norms are developed. 3 Cronbach (1970) states that validity is the most important characteristic of a test and i t needs to be examined for a l l populations with which the test wi l l be employed. Since Canada is a multicultural country (Bhatnagar, 1981; Burnet, 1984; McLeod, 1984) consisting of many diverse populations, ultimately the validity of the K-ABC should be determined for a l l cultural groups with whom i t wi l l be used. In Vancouver, British Columbia, a large city (population approximately 450,000) in Western Canada, LaTorre (1983) identified over 50 cultural/linguistic groups. He found that of the 29,700 children enrolled in elementary schools within the public school system, 51% were identified as speaking English at home, 14% speaking Cantonese, and 5% speaking Punjabi. These three groups comprise the majority of children in the Vancouver school district. The validity of the K-ABC for use with these children has yet to be investigated, i t is towards this aim that the present study was directed. K-ABC: Basis for Development The K-ABC has been heralded as controversial (Kamphaus & Reynolds, 1984), novel (Das, 1984a), and revolutionary (Bolen & Chidlers, 1985). Subsequently, i t has received a considerable amount of cr it ical "discussion" in the literature. For example, The Journal of Special Education devoted a Special Issue to the 4 K-ABC ( M i l l e r , 1984). The reason for i t s attractiveness may be related to the c r i t e r i a set for i t s development. Kaufman and Kaufman (1983b) designed the K-ABC to meet the following c r i t e r i a , namely, 1. to measure i n t e l l i g e n c e from a strong t h e o r e t i c a l and research basis 2. to separate acquired factual knowledge from the a b i l i t y to solve unfamiliar problems 3. to y i e l d scores that translate to educational intervention 4. to include novel tasks 5. to be easy to administer and objective to score 6. to be sensitive to the diverse needs of preschool, minority groups, and exceptional children, (p. 5) Theory Based. The K-ABC's th e o r e t i c a l model i s based on a dichotomous model of information processing (sequential and simultaneous processing). This dichotomy represents one part of the Das/Luria Information Processing Theory (Das, Kirby, Jarman, 1975, 1979; Luria, 1966). Sequential processing refers to the synthesis of information i n a s e r i a l or temporal order. Simultaneous processing refers to the synthesis of information in a gestalt or h o l i s t i c fashion. In the K-ABC, those two processing modes constitute separate scales, and when combined, they comprise the Mental Processing Composite or in t e l l i g e n c e scale. This process-oriented approach to assessing i n t e l l i g e n c e 5 o f f e r s a departure from the c o n t e n t - o r i e n t e d approach of i n t e l l i g e n c e t e s t s (Kaufman & Kaufman, 1983b) such as the S t a n f o r d - B i n e t I n t e l l i g e n c e S c a l e (Terman & M e r r i l l , 1973), Wechsler I n t e l l i g e n c e S c a l e f o r C h i l d r e n - Revised (WISC-R) (Wechsler, 1974), and Woodcock-Johnson Psychoeducational B a t t e r y (Woodcock & Johnson, 1977). T h i s t h e o r e t i c a l model upon which the K-ABC i s based i s , a c c o r d i n g to Kamphaus and Reynolds (1984), the most d i s t i n g u i s h i n g c h a r a c t e r i s t i c of t h i s t e s t . The importance of d e v e l o p i n g a t e s t with a s t r o n g t h e o r e t i c a l and r e s e a r c h base i s w e l l supported i n the l i t e r a t u r e (Das, 1984ab; Kaufman & Kaufman, 1983b; Majovski, 1984; Mehrens, 1984; Sternberg, 1984). "The reason i s not o n l y to g i v e adequate t h e o r e t i c a l i n t e r p r e t a t i o n , but i n a d d i t i o n t o a i d i n making p r a c t i c a l d e c i s i o n s r e g a r d i n g the c h i l d ' s p a r t i c u l a r need" (Majovski, 1984, p. 263). However, some reviewers have expressed concern f o r the apparent l a c k of complexity of the S e q u e n t i a l P r o c e s s i n g S c a l e (Bracken, 1985; H e s s l e r , 1985; Sternberg, 1984) and the i n e q u a l i t y of the s e q u e n t i a l and simultaneous dichotomy i n c o n t r i b u t i n g t o the t o t a l i n t e l l i g e n c e score (Bracken, 1985; Jensen, 1984; K e i t h , 1985). Problem-solving versus A c q u i r e d Knowledge. On the K-ABC pro b l e m - s o l v i n g or i n t e l l i g e n c e i s assessed s e p a r a t e l y from a c q u i r e d knowledge. I n t e l l i g e n c e (problem-solving) i s assessed 6 on the Mental Processing Composite and acquired knowledge i s assessed on the Achievement Scale. There i s no evidence that the K-ABC has been successful in separating problem-solving from acquired knowledge (Goetz & H a l l , 1984; Sternberg, 1984). Tra d i t i o n a l i n t e l l i g e n c e tests have not distinguished between acquired knowledge and problem-solving. The a b i l i t y to separate acquired knowledge from i n t e l l i g e n c e i s not strongly supported in the l i t e r a t u r e . Subsequently, i t i s believed by some (Anastasi, 1984; Goetz & H a l l , 1984; Sternberg, 1984) that they are neither d i s t i n c t nor separable. Further, the K-ABC d i f f e r s i n i t s measurement of achieve-ment compared with more t r a d i t i o n a l measures. The K-ABC Achievement Scale assesses school achievement (Arithmetic, Reading/Decoding, Reading/understanding), and also verbal a b i l i t y (Riddles) and general factual knowledge (Faces & Places). This provides for a more general measure of acquired learning. Anastasi (1984) expressed concern that confusion may r e s u l t i n how to interpret the K-ABC Achievement Scale i n r e l a t i o n to t r a d i t i o n a l tests that measure just school achievement. Educational Intervention. The aim of most contemporary test developers i s to create an instrument which w i l l lead to a v a l i d diagnosis of a c h i l d ' s learning strengths and l i m i t a t i o n s to serve i n the educational intervention processes. The authors 7 of the K-ABC have attempted to make a d i r e c t l i n k between t e s t scores and i n t e r v e n t i o n by p r o v i d i n g educators with an i n t e r v e n t i o n procedure based on i d e n t i f i e d K-ABC p r o f i l e s . Kaufman and Kaufman (1983b) advocate the d i r e c t t e a c h i n g of academic areas, of which three are assessed by the Achievement Sc a l e ( v i z . , a r i t h m e t i c and re a d i n g decoding and comprehension), v i a the c h i l d ' s most e f f i c i e n t mode of p r o c e s s i n g ( s e q u e n t i a l versus simultaneous). As an example, i f a c h i l d ' s r e a d i n g decoding s k i l l s are low and he or she has a simultaneous p r o c e s s i n g s t r e n g t h , a whole word approach t o te a c h i n g decoding s k i l l s i s recommended. The authors of the K-ABC devoted 58 pages i n t h e i r IM t o p r o v i d i n g such e d u c a t i o n a l s u g g e s t i o n s . However, there are no r e p o r t s i n the IM on the e f f i c a c y of u s i n g the K-ABC based i n t e r v e n t i o n model compared with other models. Novel Tasks. The i n c l u s i o n of novel tasks i n the K-ABC has stemmed from i t s authors' concern f o r the l i t t l e o r i g i n a l i t y shown i n the tasks i n t r a d i t i o n a l i n t e l l i g e n c e measures. However, A n a s t a s i (1984) commented t h a t the main c o n s i d e r a t i o n should be i n p r e p a r i n g t e s t items t o f i t the t h e o r e t i c a l d e f i n i t i o n o f the t r a i t : T h e r e f o r e , f u r t h e r comments r e l a t e d to the n o v e l t y o f the K-ABC tasks w i l l be d i r e c t e d to t h e i r t h e o r e t i c a l r e l e v a n c e . 8 Administration and Scoring. Errors in administering and scoring t r a d i t i o n a l measures of in t e l l i g e n c e , such as the WISC-R (Freides, 1978; S a t t l e r , 1982) also concerned the authors of the K-ABC. As such, they i d e n t i f i e d "easy administration and simple, objective scoring" (Kaufman & Kaufman, 1983b, p. 7) as a p r i o r i t y i n the development of the K-ABC. This goal was considered important because errors made in the process of administration and scoring can a f f e c t the r e l i a b i l i t y and v a l i d i t y of the test results (Kaufman & Kaufman, 1983b). Accommodates Diverse Populations. The f i n a l goal set for the development of the K-ABC was to construct the test to be "an e f f e c t i v e and powerful tool for important r e f e r r a l populations (Kaufman & Kaufman, 1983b, p. 8). As an example, "teaching" items were included to provide the examiner with f l e x i b i l i t y i n establishing that children (especially those from minority groups) understand the demands of the task. These six c r i t e r i a are c r i t i q u e d i n Chapter II and empirical evaluations of the v a l i d i t y of the K-ABC for various populations are included i n the c r i t i q u e . Because of the paucity of published research on thi s instrument due to i t s recent release, results from the research review are 9 tentative. However, there i s agreement among researchers that further investigations of the K-ABC's v a l i d i t y are required (Saklofske & J e d l i c k i , 1985; Zucker, 1985). Defin i t i o n s This study i s one such v a l i d i t y investigation. In t h i s study three groups were defined as follows: Cantonese Oriental children who, while at home, spoke and/or were spoken to i n Cantonese. English Caucasian children (non-native Indian) who, while at home, spoke and were spoken to i n English. Punjabi Asian Indian children who, while at home, spoke and/or were spoken to i n Punjabi. Purpose of Study The purpose of thi s study was to investigate the construct v a l i d i t y of the K-ABC for nonimmigrant t h i r d graders from three 10 Canadian c u l t u r a l / l i n g u i s t i c groups, namely, Cantonese, English and Punjabi speaking. The two methods employed to determine the construct v a l i d i t y of the K-ABC for these three groups of Canadian children were the examination of i t s underlying factor structure and i t s c o r r e l a t i o n with the WISC-R. The r e p l i c a t i o n of the va l i d a t i o n methods reported i n the K-ABC IM permitted comparisons between the Canadians i n t h i s study and the Americans i n the standardization sample. More s p e c i f i c a l l y , the following questions were addressed. Group Differences 1. What differences e x i s t among the groups i n terms of the children's mean test scores, variances, and r e l i a b i l i t i e s on the K-ABC? Factor Structure 2. For each group, how well does the theoreti c a l model of the K-ABC (sequential/simultaneous/achievement) support the data? 3. For each group, what factors describe the internal structure of the K-ABC? Relationship between K-ABC and WISC-R 4. For each group, how do the performance of the subjects on the K-ABC and WISC-R compare? 11 Delimitations of Study The Cantonese, English and Punjabi groups were r e s t r i c t e d to grade three children of nonimmigrant status. Children with documented emotional, mental, physical and sensory handicaps were excluded from the study. Furthermore, the children were representative of the middle to lower-middle socioeconomic stratum. These r e s t r i c t i o n s on the sample l i m i t the generaliz-a b i l i t y of the results to beyond such s p e c i f i c groups of children. Organization of Thesis Contained within the following eight chapters are: a review of the l i t e r a t u r e on the v a l i d i t y of the K-ABC (Chapter I I ) ; a description of the instruments (Chapter I I I ) ; a detailed outline of the methods used i n c o l l e c t i n g and processing the data (Chapter IV); a description of the ch a r a c t e r i s t i c s of each sample (Chapter V); the analysis of the psychometric properties of the K-ABC (Chapter VI); the comparison of the K-ABC and WISC-R (Chapter VII); an interpretative discussion of group and te s t differences (Chapter VIII); arid a summary of the study, a discussion of the implications of the findings, and recommendations for future avenues of research (Chapter IX). CHAPTER II Review of the Literature According to Cronbach (19 71), "validating examines the soundness of a l l [ i t a l i c s included] the interpretations of a test" (p. 443). In other words, i t assesses how e f f i c i e n t l y the test measures what i t i s reported to assess. " I t [ v a l i d i t y ] i s the most essential c h a r a c t e r i s t i c of an 'assessment instrument'" (Brown, 1980, p. 3). Sp e c i f i c to the present study i s construct v a l i d i t y which refers to the degree to which a test measures the t r a i t or psychological construct i t was intended to assess. The process of construct val i d a t i o n , according to Anastasi (1984), i s a gradual one and represents the accumulation of data from many research investigations. Anastasi (1984) believes that i n the course of developing a test i t s authors should follow a multi-stage procedure for establishing construct v a l i d i t y , which includes: formulating a construct based on a theory and research findings, developing items to represent the construct, empirically evaluating the items to determine t h e i r appropriateness for inclusion i n the item pool, factor analyzing the items and/or subtests to determine i f the psychological constructs emerge, and f i n a l l y , cross-validating the test with other tests hypothesized to assess the construct i n question. Given t h i s multi-stage procedure, Anastasi concluded that the val i d a t i o n of the K-ABC i s procedurally i n t a c t . This does not imply support for the construct v a l i d i t y of the K-ABC, but i t does speak favourably for the d i l i g e n t methodological care that went into developing t h i s t e s t . Kaufman and Kaufman (1983b) focused on the following major areas when developing the K-ABC: 1) measuring i n t e l l i g e n c e from a t h e o r e t i c a l model, 2) separating acquired knowledge from problem-solving a b i l i t y , 3) y i e l d i n g test scores that provide for appropriate educational t r a n s l a t i o n , 4) making the test easy to administer and score, and 5) accommodating the testing needs of exceptional children, preschoolers, and minority groups. In reviewing the l i t e r a t u r e pertaining to the construct v a l i d i t y of the K-ABC, the Kaufmans' major test development goals w i l l be addressed. Theoretical Basis for Measuring Intelligence Intelligence as measured by the K-ABC i s defined i n terms of an individual's s t y l e of solving problems and processing information. (Kaufman & Kaufman, 1983b, p. 2) Style of processing i s assessed on the K-ABC by dichotomous scales, namely, Sequential Processing and Simultaneous Processing. When combined these two scales form the Mental Processing Composite, a "measure of t o t a l i n t e l l i g e n c e " (Kaufman & Kaufman, 1983b, p. 31). According to Kaufman (1984) t h i s dichotomous model i s not a r e f l e c t i o n of 1 4 one theory, but rather, i t represents the convergence of a number of theoreti c a l models. Kaufman and Kaufman (1983b) provided the following dichotomous models as examples: Sequential versus p a r a l l e l or s e r i a l versus multiple (Neisser, 1967), successive versus simultaneous (Das, Kirby & Jarman, 1975; Luria, 1966), analytic versus g e s t a l t / h o l i s t i c (Levy, 1982), propositional versus appositional (Bogen, 1969), verbal versus imagery or sequential versus synchronous (Paivio, 1975, 1976), controlled versus automatic (Schneider & S h i f f r i n , 1977; S h i f f r i n & Schneider, 1977), time-ordered versus time-independent (Gordon & Bogen, 1974), and other dichotomous labels associated with individuals such as Freud, Pavlov, Maslow, and James. (Bogen, 1969). (p. 25) Given the K-ABC's i n t e l l i g e n c e scale i s not based on any one theory, the construct v a l i d i t y of i t s i n t e l l i g e n c e scale w i l l be discussed i n terms of what the authors of the K-ABC have defined t h i s t e s t to represent. Sequential and Simultaneous Dichotomy Factor analysis i s a commonly employed s t a t i s t i c a l technique for id e n t i f y i n g psychological t r a i t s . I t was used by the authors of the K-ABC to determine i f the subtests hypothesized to measure the two processing modes did so for the K-ABC standardization sample. They employed both exploratory and confirmatory factor analyses procedures. For the exploratory analyses the data from the K-ABC standardization sample were subjected to a p r i n c i p a l components and p r i n c i p a l factor analysis with a varimax (orthogonal) rotation. Two factors ( l a b e l l e d "Sequential and Simultaneous") were retained for each of the three age groups (3, 6 & 10 year olds) reported i n the IM. When employing a .35 salience c r i t e r i o n the subtests loaded on the hypothesized factor, except for the Hand Movements subtest for the 10 year olds which i s a Sequential Scale subtest and requires the examinee to copy a series of hand patterns. I t had a double loading (above .35) on both factors for the 10 year olds, which may indicate that a developmental s h i f t i n processing occurs on t h i s task. Moreover, Kaufman and Kaufman (1983b) added that for 5, 9, 10 and 12 year olds, Hand Movements loaded higher on the simultaneous factor than on the sequential. Given these findings, i t appears that Hand Movements i s not a pure measure of Sequential Processing for these age groups, yet i t remained on the Sequential Processing Scale. This i s a major flaw i n the construct v a l i d i t y of the K-ABC. One can not help questioning why t h i s subtest was not deleted or moved to the Simultaneous Processing Scale for the appropriate age l e v e l s . According to Mehrens (1984), Photo Series (placing pictures i n chronological order) was hypothesized to be a measure of sequential processing and after the standardization of the K-ABC and the r e s u l t i n g factor analysis, i t was moved to the Simultaneous Processing Scale. Now only three subtests remained on the Sequential Scale. As such, the removel of Hand 16 Movements from th i s scale would have resulted i n the Sequential Processing Scale having only two subtests, both measures of short-term auditory memory. This c a l l s into question the test's a b i l i t y to d i f f e r e n t i a t e sequential processing tasks from pure memory tasks. Because Hand Movements loaded on the Simultaneous factor for the four age groups, other theorists and researchers have proposed other t h e o r e t i c a l models for interpreting the two K-ABC factors. Das (1984b) concluded that a verbal/nonverbal dichotomy may be an acceptable explanation. Similarly, Keith and Dunbar (1984) proposed a verbal-memory and nonverbal reasoning dichotomy. Empirical investigations w i l l need to be conducted to validate the models proposed by Das and Keith and Dunbar. Since the development of the K-ABC was based on a the o r e t i c a l model (sequential/simultaneous) i t i s more appropriate to determine i f t h i s model i s supported or confirmed by the K-ABC standardization data. Therefore, a confirmatory factor analysis was performed and i t s results were reported i n the IM. The factor model was spec i f i e d i n advance and was representative of the subtest-scale match. Kaufman and Kaufman (1983b) concluded that "large highly s i g n i f i c a n t values of chi-square were obtained for a l l analyses, and substantial factor loadings (usually i n excess of .55) were found for the subtests on each factor" (p. 107). This i s a common misinter-pretation of thi s goodness of f i t s t a t i s t i c . According to jOreskog and Sttrbom (1979) a high chi square suggests a s i g n i f -icant difference between the model and data, which i s the opposite of the interpretation applied i n the case of the K-ABC. However, an investigation of the factor loadings from the confirmatory factor analysis, for the 5, 7 and 9 year olds reported i n the IM, suggests the data-model f i t was at least acceptable. Since the goodness of f i t s t a t i s t i c s were not provided i n the IM, a more informed decision could not be made as to the cause of the inconsistency i n the reporting of the confirmatory factor analysis solution. Keith (1985) performed a confirmatory factor analysis on the K-ABC standardization data at three age l e v e l s : 5, 7 and 10 year olds. He concluded that the Sequential/Simultaneous model was a very good f i t for the 7 year olds and a reasonable f i t for the 5 and 10 year olds. Hand Movements appeared central to the determiner of goodness f i t . Kaufman (1983) stated that "the two types of mental processing bear a nonhierarchial relationship to each other and are equally important" (p. 212). If the two scales representing these processes have equal importance and are nonhierarchial, as Kaufman maintains, then they should have equal weight. However, because there are three Sequential and fi v e Simultaneous subtests (ages 6 through 12 1/2), and because they contribute a d d i t i v e l y rather than i n a pro-rated r a t i o to the 18 Mental Processing Composite, the two K-ABC processing scales do not contribute equally to the Mental Processing Composite, or " t o t a l i n t e l l i g e n c e " score (Bracken, 1985; Jensen, 1984; Keith, 1985). According to Bracken (1985) "the MPC [Mental Processing Composite] i s weighted 60% Simultaneous ages 2 1/2 through 3 years, 57% ages 4-0 through 5-11 and 63% between ages 6-0 and 12-6" (p. 23). This a r t i f a c t of the data negates the Kaufmans' claim that the two processing scales contribute equally to the in t e l l i g e n c e score. As an example, Bracken observed that i f a 10 year old c h i l d received a Simultaneous Processing Score of 85 and a Sequential Processing Score of 100 (discrepancy of 1 standard deviation), t h i s c h i l d would receive a Composite Score of 89. On the other hand, i f the c h i l d scored 100 on the Simultaneous Processing Scale and 85 on the Sequential Processing Scale, his or her Composite would equal 95. Jensen (1984) concluded that there was no th e o r e t i c a l basis for the unequal weights of the two scales. Consequently, t h i s psychometric flaw may have serious implications when interpreting test results or when developing educational interventions. Kaufman (1983), aware that the disproportionate number of subtests may concern many psychologists, reported This lack of equality probably i s due to the constructs themselves: Simultaneous Processing 19 seems to be more multifaceted, as unique aspects of the dimension are assessed by tests that are highly perceptual (Gestalt Closure), s p a t i a l (Triangles), and analogic (Matrix Analogies) i n nature; sequential processing seems more unidimensional, as a variety of tasks developed during e a r l i e r stages of K-ABC test construction did not add enough unique information to warrant inclusion i n the K-ABC. (pp. 212-213) Das et a l . (1975, 1979) stated that successive (labelled sequential i n K-ABC) and simultaneous synthesis can both be of the perceptual, memory and conceptual variety - hence multifaceted. I t appears that the three Sequential Processing subtests chosen by the Kaufmans assess only sequential memory and do not tap the multifaceted t h e o r e t i c a l underpinnings of sequential processing. Although the unequal weights and unequal l e v e l of i n t e l l e c t u a l complexity assessed by the two processing scales may indicate that the K-ABC i s not representative of the theory behind the test, Kaufman (1983) provided c o r r e l a t i o n a l evidence that he believes indicates the two processing scales are equally important i n th e i r r e l a t i o n to 30 other measures of in t e l l i g e n c e and achievement. He found both processing scales had comparable correlations with these 30 external variables. Nevertheless, caution i n interpreting the Mental Processing Composite needs to be extended because children performing better on the Simultaneous Scale w i l l have a higher composite i n t e l l i g e n c e score than children with a superior Sequential Scale Score. 20 Furthermore, attention needs to be directed at how to interpret a K-ABC processing score. Das (1984b) expressed concern that "the battery [K-ABC] does not provide a procedure for scoring the performance on the task according to the strategies used by the c h i l d - the tasks are scored as defined, a p r i o r i , by the i r placement i n one of the two coding [processing] categories" (p. 233). It would appear that the child's score i s a r e f l e c t i o n of hi s or her a b i l i t y to perform on subtests hypothesized to measure sequential or simultaneous processing, and i s not a measure of the child's cognitive style (Sternberg, 1984). Gunnison's (1984ab) c l i n i c a l interpretations of strategies used and error patterns produced by an ind i v i d u a l c h i l d on each subtest may prove to be the most informative data obtained from an administration of the K-ABC. I t may also provide the K-ABC with the construct v a l i d i t y i t seeks vis a vis the individual's st y l e of processing. Mental processing Composite Although the K-ABC measures the two types of mental processing with separate nonhierarchial scales, we believe that i n t e l l i g e n c e i s complex [ i t a l i c s added] and that probably the most i n t e l l i g e n t behavior results from an integration [ i t a l i c s added] of sequential and simultaneous processing. (Kaufman & Kaufman, 1983b, p. 31) 21 Although t h i s d e f i n i t i o n does not state that the authors of the K-ABC necessarily believe that the Mental Processing Composite i s a complex measure of integrated mental behavior, they do state that t h i s i s th e i r b e l i e f l a t e r i n the IM. Therefore, the l e v e l of complexity of the Mental Processing Composite needs to be addressed. Some c r i t i c s believe that the K-ABC may not be a complex measure of i n t e l l i g e n c e (Bracken, 1985; Das, 1984ab, Goetz & H a l l , 1984; Hessler, 1985). The unidimensional nature of the Sequential Scale and a number of the Simultaneous subtests i s thought to be an ind i c a t i o n of the tests' limited complexity (Bracken, 1985; Das & Jarman, i n press). According to Bracken, the only complex measure of i n t e l l e c t u a l behavior of the K-ABC are three subtests on the Simultaneous Processing Scale (Triangles, Matrix Analogies, Photo Serie s ) . These three subtests require s k i l l i n problem-solving and appear to require planning and judgement. The a b i l i t y to plan and make decisions i s considered to be the most complex form of i n t e l l i g e n t behavior (Das & Jarman, 1981; Das & N a g l i e r i , 1985). Although three of the K-ABC subtests appear to involve complex mental a b i l i t y , Sternberg (1983) concluded that " i t would seem highly desirable to have one or more subtests e x p l i c i t l y measuring planning a b i l i t y " (p. 201). He added that planning i s an important aspect of Luria's theory. More s p e c i f i c a l l y , i t i s represented i n the t h i r d block of the Das-Luria Information Processing Model (Das et a l . , 1975, 1979; Luria, 1966). Although the K-ABC was not designed as a measure of the Das-Luria Model (Kaufman, 1984), as a comparative model i t appears to be the most frequently referenced (Das, 1984ab; Majovski, 1984; Sternberg, 1984). The Das-Luria model consists of three blocks. The f i r s t block regulates c o r t i c a l tone (Luria, 1973). For mental a c t i v i t y to occur an optimal l e v e l of c o r t i c a l tone i s necessary. Affected by t h i s unit are such behaviors as attention, drive, and motivation. The K-ABC does not measure thi s block d i r e c t l y . Although inferences about these behaviors can be drawn from the child' s test-taking behavior, the K-ABC offe r s no formal evaluative c r i t e r i a for them. The second functional unit i s primarily for reception, storage and analysis of information. The information which enters t h i s block i s coded or processed simultaneously or successively. Respectively, these two processing modes are defined the same as the Simultaneous and Sequential Processing Scales on the K-ABC (Das, 1984a; Kaufman, 1984). In ef f e c t , the K-ABC represents just one unit of the Das-Luria Model. Decision-making, planning, and programming are functions of the t h i r d and l a s t block. This unit i s not d i r e c t l y assessed by the K-ABC, however three subtests (Triangles, Matrix Analogies, Photo Series) are more complex measures because they do require functions s p e c i f i e d i n thi s block. 23 Given these three blocks, Luria (1973) stressed that i t i s not correct to think of each unit carrying out an a c t i v i t y independent of the other units. Das and Jarman (in press) provided the following example. Input from the external environment reaches both Block 1 and Block 2. From that point on, there i s a continuous i n t e r a c t i o n between the three blocks. The arousal functions of Block 1 c e r t a i n l y influence both coding and planning. On the one hand, i n Block 2 the coding processes themselves provide a basis for planned action; on the other hand, plans and decisions and strategies influence the way we code information. These planning functions also modulate our arousal response, (in press) Although these units are considered to be interdependent they s t a t i s t i c a l l y have been shown to be d i s t i n c t . Das and Heemsbergen (1981) found planning emerged as an independent factor "possibly because of the presence of additional variance over and above that which can be explained by coding" (p. 2). Therefore, the concern for the K-ABC's degree of complexity may be more related to the absence of a scale that measures planning. The l e v e l of complexity of the K-ABC subtests i s an external v a l i d a t i o n issue, hence requiring an empirical investigation of t h e i r r e l a t i o n s h i p to existing measures of complex i n t e l l i g e n c e . According to Jarman (personal communication, June 19, 1985) there has not been enough cross-validation of the K-ABC to determine i t s l e v e l of 24 complexity. T h e r e f o r e , he concluded that t h i s issue i s i n c o n c l u s i v e . In summary, two years a f t e r the r e l e a s e of the K-ABC there remains mixed support f o r the c o n s t r u c t v a l i d i t y of the i n t e l l i g e n c e s c a l e . When c o n s i d e r i n g the f a c t o r a n a l y s e s ' evidence, there i s support f o r a Sequential/Simultaneous dichotomous model of i n t e l l i g e n c e . A d d i t i o n a l models and i n t e r p r e t a t i o n s of the K-ABC 1s data need f u r t h e r v a l i d a t i o n . A fundamental concern r e s t s with the unequal weighting of the two pr o c e s s i n g s c a l e s i n c o n t r i b u t i n g to the t o t a l i n t e l l i g e n c e s c o r e . The d i a g n o s t i c i m p l i c a t i o n s of t h i s i n e q u i t y r e q u i r e f u r t h e r i n v e s t i g a t i o n . F i n a l l y , the l e v e l of complexity of the i n t e l l i g e n c e s c a l e remains a moot p o i n t . As such, e m p i r i c a l i n v e s t i g a t i o n s need to be conducted to determine i f i t s apparent l a c k of complexity a f f e c t s the p r e d i c t i v e v a l i d i t y of the s c a l e . Problem-Solving versus Acq u i r e d F a c t u a l Knowledge The second c r i t e r i o n s e t f o r the development of the K-ABC invo l v e d s e p a r a t i n g the a b i l i t y to sol v e novel problems from acq u i r e d f a c t u a l knowledge. Kaufman and Kaufman (1983b) s t a t e d t h a t , i n the K-ABC, problem-solving a b i l i t y was " i n t e r p r e t e d as 25 i n t e l l i g e n c e " (p. 2) and acquired knowledge was "defined as achievement" (p. 2 ) . They recognized this as a departure from the more t r a d i t i o n a l tests of i n t e l l i g e n c e where acquired knowledge i s considered part of i n t e l l i g e n c e . Elsewhere, Kaufman (1984) also refers to this d i s t i n c t i o n as a b i l i t y ( i ntelligence) versus achievement. The rationale behind separating acquired knowledge and problem-solving stems from the b e l i e f of the K-ABC's authors that acquired knowledge i s related to educational, environ-mental and motivational factors. As such, they reported that i t is an achieved s k i l l and not to be equated with i n t e l l e c t u a l functioning (Kaufman & Kaufman, 1983b). This appears to infer that i n t e l l e c t u a l functioning i s not related to environmental factors. However, this stance i s negated l a t e r : " i n t e l l i g e n c e involves a dynamic interaction of heredity and environment" (Kaufman & Kaufman, 1983b, p. 20 ) . One cannot help but question the extent of a clear rationale for the specified separations. Anastasi (1984), Goetz and H a l l (1984) and Sternberg (1984) hold that acquired knowledge cannot be separated from problem-solving a b i l i t y . Kaufman (1984) agrees that the d i s t i n c t i o n i s not a clean one. He added that the K-ABC has to function within the "real world" where testing is done, which requires meeting testing guidelines which specify testing a 26 c h i l d on both i n t e l l i g e n c e and achievement tests (Kaufman, 1984). Sternberg (1984) provided an i n s i g h t f u l explanation of why problem-solving and acquired knowledge are both knowledge based. The former he relates to procedural knowledge, "the knowledge of the strategies and procedures one can use to solve problems of various kinds" (p. 273), and the l a t t e r he spe c i f i e s as declarative knowledge, "the knowledge of facts, ideas, and certain p r i n c i p a l s " (p. 273). Kaufman (1984) agreed with Sternberg (1984) that procedural knowledge i s the more sophisticated of the two knowledge forms. However, these two authors appear to disagree as to the relationship each of the two types of knowledge has with i n t e l l i g e n c e . While Kaufman (1984) believes the procedural/declarative d i s t i n c t i o n supports his ability/achievement dichotomy, Sternberg (1984) believes both forms are representative of i n t e l l e c t u a l functioning. However, what Kaufman and Sternberg appear to be avoiding i s that the most sophisticated form of i n t e l l i g e n t behavior may be the a b i l i t y of an in d i v i d u a l to fin d (Blank, 1982) or generate (Das & jarman, 1981) problems. This i s not assessed on the K-ABC or any of the t r a d i t i o n a l i n t e l l i g e n c e tests. Sternberg (1984) praised the authors of the K-ABC i n attempting to ensure t h e i r battery was a " f a i r " assessment instrument for minority children. However, he added that separating acquired knowledge from the in t e l l i g e n c e scale i s not necessarily going to promote a f a i r e r assessment. Sternberg added that because the abstract stimuli on the in t e l l i g e n c e scale of the K-ABC requires a more sophisticated form of knowledge (procedural knowledge) many minority children may not have the understanding of the strategies needed to successfully complete these novel and abstract tasks. This i s why minority children have been found to be disadvantaged on c u l t u r e - f a i r tests which t r a d i t i o n a l l y consist of abstract stimuli (Jensen, 1980) . In addition to the knowledge d i s t i n c t i o n provided by Sternberg, the K-ABC's intelligence/achievement dichotomy i s related to the f l u i d / c r y s t a l l i z e d dichotomy of the Cattell-Horn theory ( C a t t e l l , 1963). F l u i d a b i l i t y refers to solving problems with unfamiliar s t i m u l i and c r y s t a l l i z e d a b i l i t y refers to solving problems which emphasize previous t r a i n i n g and education. Although Kaufman (1984) acknowledged that the K-ABC's intelligence/achievement dichotomy p a r a l l e l s the f l u i d / c r y s t a l l i z e d dichotomy, he was quick to add that the K-ABC was not developed to represent the Cattell-Horn theory. Moreover, these two dichotomous models d i f f e r i n how they relate to i n t e l l i g e n c e . As was true for Sternberg's (1984) knowledge dichotomy, the f l u i d / c r y s t a l l i z e d a b i l i t i e s are both considered to represent i n t e l l e c t u a l functioning. This i s evident i n the WISC-R, for example, where the Verbal IQ ( c r y s t a l l i z e d ) and the 28 Performance IQ ( f l u i d ) equally contribute to the F u l l Scale IQ. However, i n the K-ABC the achievement and i n t e l l i g e n c e portions of the K-ABC together o f f e r a composite of the child's o v e r a l l present l e v e l of functioning i n both novel and t r a d i t i o n a l learning tasks, the t r a d i t i o n a l (Achievement) subtests are never used to in f e r a child's i n t e l l e c t u a l p o t e n t i a l or capacity, mental a b i l i t y , general learning aptitude, or "IQ" by any other name. (Kaufman & Kaufman, 1983b, p. 33) The rationale behind separating acquired knowledge from i n t e l l i g e n c e i s not a clear one. Nevertheless, given that t h i s was one of the c r i t e r i a set for the development of the K-ABC, i t i s important to determine i f the K-ABC i s a v a l i d representation of the proposed dichotomy. In the K-ABC, the problem-solving, or i n t e l l i g e n c e , construct i s represented by the Mental Processing Composite, and the acquired knowledge or achievement construct i s represented by the Achievement Scale. The Achievement Scale measures acquired knowledge i n the form of general information, language acquisition, and school achievement (arithmetic, reading). Anastasi (1984) stated that the label "Achievement" may be misleading for t h i s scale does not measure what has t r a d i t i o n a l l y been referred to as school achievement. As such, she believes that "Achievement" i s an "unfortunate choice as a label for the Achievement Scale" (p. 364). Hessler (1985) pointed out the Achievement Scale i s not a comprehensive measure of achievement and w i l l need to be 29 supplemented with measures of mathematical computations, sp e l l i n g , and written expression, whereby, bringing this scale's u t i l i t y into question. Salvia and Hritcko (1984) added that the Achievement score i s not meaningful because i t i s a composite of diverse content areas. Moreover, Keith and Dunbar (1984) i d e n t i f i e d t h i s scale to be a measure of verbal reasoning and reading achievement. Kaufman (1984), however, believes the Achievement Scale compares favorably with other measures of achievement. He added that i t was given the label because of testing guidelines that specify testing a c h i l d on both i n t e l l i g e n c e and achievement measures. Given that the in t e l l i g e n c e scale was c a l l e d "Mental Processing Composite", Kaufman believed that to c a l l achievement by another name might prove confusing. A t o t a l investigation of the properties of the Mental Processing and Achievement scales indicates that there are problem-solving and knowledge elements i n each scale. For example, Photo Series (Mental Processing subtest), which requires a c h i l d to place photographs i n chronological order, expects the c h i l d to have p r i o r knowledge of the event pictured for successful completion of the task (Goetz & H a l l , 1984). Another example, Riddles (Achievement subtest), which requires the c h i l d to name an object a f t e r being given a l i s t of i t s c h a r a c t e r i s t i c s , has verbal content presented sequentially and 30 requires simultaneous processing for solving the problem (Kaufman, 1984). Empirical investigations of the 13 K-ABC subtests provides additional insight into the effectiveness of the K-ABC in separating acquired knowledge and problem-solving. Kaufman and Kamphaus (1984) factor analyzed the standardization data (principal factor with a varimax rotation) and found three independent factors hypothesized to measure sequential processing, simultaneous processing and achievement, for children 8 years of age and older. While the factors may be independent, the sequential and simultaneous factor scores were found to have a moderate correlation with the Achievement subtest raw scores. As an example, the Sequential factor scores correlated between .40 (Faces & Places) and .47 (Reading/Decod-ing) with the Achievement subtest raw scores. Similarly, the Simultaneous factor score correlated between .41 (Faces & Places) and .52 (Riddles) with the Achievement subtest raw scores. The authors did not indicate i f these correlations were significant. The .74 mean correlation between the Mental Processing Composite and Achievement Scale is further evidence that a significant relationship exists between these two scales for school-aged children. This correlation is lower than the Achievement Scales correlation with the Total Score from the Peabody Individual Achievement Test (PIAT) (.89) for 31 subjects studied (Study #13, IM) and from the California 31 Achievement Test (.86) for 44 subjects (Study #9, IM). However, the Mental Processing Composite/Achievement Scale co r r e l a t i o n was somewhat higher than the Achievement Scales c o r r e l a t i o n with the Total Score from the Stanford Achievement Test (.69) for 109 subjects (Study #36, IM). The reason the correlations between the Mental Processing Composite and Achievement Scale are s i m i l a r i n magnitude to correlations between the Achievement Scale with other measures of achievement may be related to the prediction of school achievement as the main c r i t e r i o n for the development of i n t e l l i g e n c e tests (Kaufman, 1984). I t appears that the Mental Processing Composite and Achievement Scale are both measuring reproductive thinking (learned material) (Parnes, Moller & Biondi, 1977) rather than productive thinking ( c r i t i c a l , creative, innovative thinking) (S. Blank, personal communication, August 26, 1985). According to Blank (personal communication, August 26, 1985) the type of problem-solving measured by the K-ABC (Mental Processing Composite and Achievement Scale) i s reproductive. He added that because productive problem-solving i s not assessed by the K-ABC or WISC-R both tests are not r e a l l y measuring i n t e l l i g e n c e , and the v a l i d i t y these tests may possess i s i n r e l a t i o n to t h e i r c o r r e l a t i o n with each other and th e i r prediction of school achievement. Hessler (1985) reported that one of the advantages of standardizing a cognitive and academic achievement measures on the same p o p u l a t i o n i s to allow f o r comparisons to be made while e l i m i n a t i n g the e r r o r v a r i a n c e i n v o l v e d when comparing two measures s t a n d a r d i z e d on d i f f e r e n t p o p u l a t i o n s . However, the Kaufmans d i d not c o r r e c t f o r " r e g r e s s i o n e r r o r " ( H e s s l e r , 1985, p. 146) r e s u l t i n g from the imperfect c o r r e l a t i o n between the achievement and i n t e l l i g e n c e measures. As a r e s u l t , s u b j e c t s , such as the g i f t e d , who score high on the i n t e l -l i g e n c e s c a l e , w i l l tend to score lower on the achievement s c a l e . T h i s may r e s u l t i n them being l a b e l l e d as underachievers or even l e a r n i n g d i s a b l e d . The e f f e c t s of r e g r e s s i o n e r r o r on the assessment of v a r i o u s p o p u l a t i o n s r e q u i r e s e m p i r i c a l i n v e s t i g a t i o n . There i s l i t t l e support f o r s e p a r a t i n g acquired knowledge from r e p r o d u c t i v e p r o b l e m - s o l v i n g . Moreover, there i s evidence to suggest that these two a b i l i t i e s , as measured by the K-ABC, are not separate e n t i t i e s and may r e s u l t i n m i s l a b e l l i n g c h i l d r e n . E d u c a t i o n a l I n t e r v e n t i o n One of the main purposes f o r conducting an assessment i s to have scores that r e a d i l y t r a n s l a t e i n t o e d u c a t i o n a l programs of remediation ( B e r n a l , 1977; Das, 1984a; Goetz & H a l l , 1984; Grover, 1981; Kaufman Se Kaufman, 1983b). T h i s was one of the s t a t e d g oals f o r the K-ABC's development. Das et a l . (1979) i d e n t i f i e d three approaches to remedia-t i o n . The f i r s t involved attempting to improve the processes through d i r e c t t r a i n i n g . There i s no evidence that t h i s can be done. It i s possible that, for brain damaged children for example, improving processes may be impossible. A second approach, and the one promoted as being the most e f f e c t i v e by Das et a l . , involves teaching individuals strategies so they can employ the most e f f i c i e n t process. The assumption underlying t h i s approach i s that the strategies are weak not the processes. Employing strategies i s a function of the t h i r d block of the Luria-Das Model, while t r a i n i n g processes would involve the second block. The t h i r d approach, and that advocated by the Kaufmans, involves designing an educational program that u t i l i z e s an i n d i v i d u a l ' s process strength. More s p e c i f i c a l l y , the K-ABC's strength model of remediation involves i d e n t i f y i n g an individual's processing strength (sequential or simultaneous) and using i t to remediate the individual's i d e n t i f i e d academic d e f i c i t areas (Gunnison, 1984ab; Kamphaus & Reynolds, 1984; Kaufman & Kaufman, 1983b). While the strength model of intervention adopted by the authors of the K-ABC has i t s followers (Gunnison, 1984b; Reynolds, 1981), there i s no post-publication research investigating t h i s model of educational intervention. Salvia and Hritcko (1984) c r i t i c i z e d the authors of the K-ABC for not empirically val i d a t i n g t h e i r recommendations for t r a n s l a t i n g 34 test scores into an educational intervention program before they published them. The various intervention models and their advantages and limit a t i o n s are not relevant to the present study. However, what i s relevant to t h i s study i s the v a l i d i t y of the K-ABC scores from which any educational program can be developed. As previously mentioned, when interpreting a ch i l d ' s test score the unequal weighting of the two processing scales, the reduction i n complexity i n the Sequential Scale, and the inconsistency of Hand Movements i n loading on i t s designated factor should be considered. Further, since the discrepancy between the various Global scales ( i . e . , Sequential/ Simultaneous, Sequential/Achievement, Simultaneous/Achievement, Mental Processing/Achievement) i s suggested as the basis for developing an educational program for a given c h i l d , i t i s important to r e a l i z e that at least 50% of the children i n the standardization sample did not have s i g n i f i c a n t l y discrepant scores. In addition, of the reported mean scores for the exceptional children and minority groups (17 studies) presented i n the IM (Table 4.19) only the Navajo children (Study #7) evidenced a Sequential score that was s i g n i f i c a n t l y lower (12 points; jo < .05) than t h e i r Simultaneous score. Although comparisons w i l l ultimately be made for individuals, the mean group scores do suggest that the majority of exceptional children studied were not found to have discrepancies between the various Global Scale Scores, r a i s i n g questions as to the diagnostic u t i l i t y of the K-ABC. Since the model of intervention reported i n the K-ABC IM appears based on the finding of discrepant scores, i t i s not clear how referred children with no discrepant Global Scores should be treated. F i n a l l y , to determine i f an educational intervention program i s e f f e c t i v e , children are often retested af t e r a designated period of time. Improvement i s often s p e c i f i e d i n terms of gain scores. It i s important to determine i f there i s a practice e f f e c t upon retesting with the K-ABC. An investigation of Table 4.3 (IM) reveals that for the 92 children between the ages of 5-0 years and 8-11 years who were restested on the K-ABC at a 2 to 4 week i n t e r v a l , the Simultaneous Processing score showed a gain of 6.4 points. This was more than the 1.0 point gain for the Sequential Scale, 4.8 for Mental Processing Composite, or 1.8 for the Achievement Scale. This suggests that the Simultaneous Scale i s more prone, to a practice e f f e c t than the other scales. The practice e f f e c t found for the Simultaneous Scale i s a factor i f the object of the intervention i s to improve processing. However, since the object of the K-ABC's intervention approach i s to improve academic areas by providing i n s t r u c t i o n to the child's process strength, the practice 36 e f f e c t o n t h e a c h i e v e m e n t s c a l e ( 1 . 8 p o i n t s ) i s m o r e r e l e v a n t t o t h i s r e m e d i a t i o n a p p r o a c h . When i n t e r p r e t i n g a g a i n s c o r e a s e v i d e n c e o f i m p r o v e m e n t r e l e v a n t t o t h e s p e c i f i e d e d u c a t i o n a l p r o g r a m i t s h o u l d f i r s t b e d e t e r m i n e d i f t h e g a i n s c o r e i s s i g n i f i c a n t l y g r e a t e r t h a n w o u l d b e e x p e c t e d d u e t o t h e t e s t ' s p r a c t i c e e f f e c t . E a s e o f A d m i n i s t r a t i o n a n d S c o r i n g T h e c o n c e r n e x p r e s s e d f o r t h e e r r o r s made w h e n a d m i n i s t e r -i n g a n d s c o r i n g t e s t s s u c h a s t h e W I S C - R ( F r e i d e s , 1 9 7 8 ; S a t t l e r , 1 9 8 2 ) p r o m p t e d t h e a u t h o r s o f t h e K - A B C t o m a k e e a s e o f a d m i n i s t r a t i o n a n d s c o r i n g a p r i o r i t y f o r t h e d e v e l o p m e n t o f t h e i r t e s t . E r r o r s made i n t h e p r o c e s s o f a d m i n i s t e r i n g a n d s c o r i n g a t e s t c a n r e d u c e t h e r e l i a b i l i t y a n d v a l i d i t y o f t h e t e s t r e s u l t s ( K a u f m a n & K a u f m a n , 1 9 8 3 b ) . T o e n s u r e t h a t t h e K - A B C m e t t h i s c r i t e r i o n , s u b t e s t s w e r e s e l e c t e d b a s e d o n t h e i r e a s e o f a d m i n i s t r a t i o n a s r e p o r t e d b y t e s t e r s i n f i e l d s t u d i e s d u r i n g t h e p r e p u b l i c a t i o n s t a g e o f t h e K - A B C . " E a s e o f a d m i n i s t r a t i o n C w a s ] f u r t h e r a i d e d b y u s i n g e a s e l s , b y a d o p t i n g a h i g h l y s i m i l a r f o r m a t f o r e a c h p r o c e s s i n g t a s k , b y k e e p i n g t h e e x a m i n e r ' s v e r b i a g e t o a m i n i m u m , a n d b y a s i n g l e d i s c o n t i n u e r u l e f o r a l l s u b t e s t s " ( K a u f m a n & K a u f m a n , 1 9 8 3 b , p . 7 ) . F u r t h e r , t h e t i m e b o n u s p o i n t s f o u n d i n o t h e r t e s t s w e r e e l i m i n a t e d t o r e d u c e s c o r i n g i n v o l v e m e n t . 37 In published reviews of the K-ABC, administration has been reported as clear (Das, 1984a) and the scoring as straight-forward (Kamphaus & Reynolds, 1984). However, Thomas (1984) identified four Mental Processing subtests that are prone to scoring errors, namely, Spatial Memory, Hand Movements, Matrix Analogies, and Triangles. For these subtests the examiner must have well developed "visual spatial representation in order to make the transition from the child's reproduction on the presented item to the scoring key provided on the back of the test easel" (Thomas, 1984, p. 3). Kaufman and Kaufman (1983c) admitted that the examiners involved in the national standardization program reported problems in scoring the more difficult items on the Spatial Memory subtest. Further, research needs to be conducted to determine whether the decision to disallow time-bonus points has the effect of not discriminating between bright and superior functioning children. Sternberg (1979) believes speeded items are a valuable source of information, i t is the subjects with the faster problem solving skil ls who usually have superior performance on IQ tests.Vernon (1983) elaborated that There is now evidence that the speed of execution of basic cognitive processes, as measured by a variety of RT [Reaction Time] tests, is an important aspect of intelligence, and that individual differences in processing efficiency account for a significant amount of the variance in g. To some extent, this may be attributable to the limiting properties of the 38 working-memory system: Faster processing enables an i n d i v i d u a l to overcome the lim i t a t i o n s , or at least to make a more e f f i c i e n t (and successful) use of his working memory, (p. 398) Vernon also acknowledged that personality variables may contribute to the individual's arousal l e v e l . The authors of the K-ABC wanted the K-ABC to promote a f a i r assessment for a l l children. The f i r s t item of every subtest i n the Mental Processing Composite i s an unscored "sample" item. On thi s item the examiner i s allowed to c l a r i f y the procedure for the c h i l d . I f the c h i l d needs further c l a r i f i c a t i o n the f i r s t two scored items may act as "teaching items"; however, only the c h i l d ' s i n i t i a l response i s scored. When teaching or t r a i n i n g a c h i l d on these items, the examiner may reword the question, use gestures, or the chi l d ' s f i r s t language to c l a r i f y . The examiner i s not permitted to use additional materials or teach the c h i l d strategies. The authors of the K-ABC reported that some preschoolers, mentally handicapped and minority children, perform poorly on a task because they are not i n i t i a l l y clear as to what i s expected. Kaufman and Kaufman (1983b) believe that the teaching items eliminate t h i s problem. Sternberg (1979) reported that measurements taken early during testing may be subject to a variety of extraneous influences -pr i o r experience with the problem type, a b i l i t y to s e t t l e on the task, f a m i l i a r i t y with test situations - that may cease to function l a t e r on. It may not be u n t i l a l l people have had a f a i r chance to f a m i l i a r i z e themselves with the task that 39 measurements of performance can be considered v a l i d predictors for other kinds of tasks, (p. 50) The reviews to date have supported the inclusion of teaching items i n the K-ABC (Hessler, 1985; Kamphaus & Reynolds, 1984; Mehrens, 1984; Sternberg, 1984). Nevertheless, these items may not be used e f f e c t i v e l y . For example, no matter where the examiner st a r t s testing, only the f i r s t two items can be administered as teaching items. The child's age determines the s t a r t i n g point for a given c h i l d . There i s , however, an exception. According to Kaufman and Kaufman (1983c), i f the examiner i s aware that the c h i l d i s mentally handicapped or emotionally disturbed, he or she can choose to s t a r t at an e a r l i e r item, giving these children the advantage of learning the demands of a task at a l e v e l they are capable of. Since the examiners have a choice, t h i s procedure implicates the lack of f u l l standardization. Research needs to be conducted to investigate the effects of the teaching procedure unevenly applied. Accommodation of Diverse Populations Kaufman and Kaufman (1983b) stated that the f i n a l goal for the development of the K-ABC was to make i t sensitive to the diverse needs of the various populations with which i t w i l l be employed. The c r i t e r i a discussed i n the preceding sections 40 (e.g., theory based, teaching items) are relevant to determining the u t i l i t y of the K-ABC for use with preschoolers, exceptional children and minorities. Moreover, additional e f f o r t s were made by the test's developers to accommodate d i f f e r e n t groups of children. For example: 1) a Nonverbal Scale was included for children with communication disorders who tend to be penalized on t r a d i t i o n a l measures of i n t e l l i g e n c e ; 2) subtests were constructed to be c o l o u r f u l , enjoyable and game-like to aid i n maintaining rapport with young children; 3) subtests were t a i l o r e d to accommodate the attentional and developmental needs of young children by reducing the length of the test battery and the number of items administered i n each subtest; and 4) exceptional children were included i n the standardization sample to allow for confidence i n interpreting the performance of these children on the K-ABC. Research has yet to focus on the effects the above have had on the performance of exceptional children on the K-ABC. Nevertheless, research has been conducted to investigate the v a l i d i t y of the K-ABC for use with special groups of children. Because the research on preschoolers and exceptional children does not d i r e c t l y apply to the non-referred, school-aged population investigated i n the present study, i t w i l l not be discussed i n d e t a i l . However, s p e c i f i c findings as they pertain 41 to the v a l i d i t y of the K-ABC w i l l be summarized. More relevant to the present study i s the research on c u l t u r a l l y and l i n g u i s t i c a l l y diverse children. This research w i l l be discussed i n d e t a i l . Preschoolers and Exceptional Children The research available on the construct v a l i d i t y of the K-ABC for preschoolers and exceptional children are summarized, where available, in terms of evidence directed at the K-ABC's 1) factor structure, 2) co r r e l a t i o n with other measures of in t e l l i g e n c e and achievement, and 3) diagnostic properties. Interpretations of findings are also provided from a c l i n i c a l perspective. Preschoolers Author Findings/Conclusions Kaufman & Kamphaus (1984) a) One factor emerged for the Mental Processing Composite for 2 1/2, 3 and 5 year olds. b) Two factors emerged for the entire battery (Sequential vs. Simultaneous & Achievement). McLoughlin & E l l i s o n (1984) c) K-ABC Achievement Scale was s i g n i f i c a n t l y correlated (r.66) with the PPVT-R Form L . Bing & Bing (1984) d) K-ABC Achievement Scale had a mean score 13 points higher than the PPVT-R. 42 C l i n i c a l Interpretation: Given that the Sequential and Simultaneous subtests did not emerge as independent factors for preschoolers, c l i n i c i a n s should cautiously apply Telzrow's (1984) recommendations for using the Sequential and Simultaneous Processing Scales " i n the d i f f e r e n t i a l diagnoses of s p e c i f i c learning d i s a b i l i t i e s i n preschool children" (p. 316). Furthermore, the comparative evidence on the K-ABC Achievement Scale and PPVT-R suggests that they are not measuring verbal a b i l i t y i n the same manner. As such, they should not be used interchangeably, but rather i n a complementary fashion. Learning Disabled Klanderman & a) Factor structure supports the Kroeschell (1984) construct v a l i d i t y of the K-ABC. Keith, Hood, Eberhart & b) Hand Movements loaded on the Pottebaum (1985) Simultaneous factor for some age groups. c) Other models, such as verbal-memory/ nonverbal reasoning may explain the K-ABC factor dichotomy. Haddad (1983) d) K-ABC Mental Processing Haddad, Carey, Culver, Composite and WISC-R F u l l Eckelcamp, Parker, Scale IQ correlated between Schwartz, Smith & Webb (1984) .30 and .70 depending on the Snyder, Leask & study. The mean score A l l i s o n (1983) discrepancy did not exceed 5 Lyon & Smith (1985) points, but i t was always i n Stoiber, Bracken & favor of the WISC-R. Gissal (1983) 43 Haddad (1983) Na g l i e r i & Haddad (1984) Hooper & Hynd (1985) Lyon & Smith (1985) Stoiber et a l . (1983) e) K-ABC Achievement Scale and PIAT Total Test Score had a .84 co r r e l a t i o n with only a 2 point difference between the two measures (K-ABC favored). f) K-ABC Sequential Scale correlated s i g n i f i c a n t l y and in a po s i t i v e d i r e c t i o n with a l l the subtests and the t o t a l score on the Visual-Aural D i g i t Span Test (VADS). The Simultaneous and Achievement Scales did not correlate s i g n i f i c a n t l y with the VADS. These results support the concurrent v a l i d i t y of the Sequential Processing Scale. g) No consistent evidence that children c l a s s i f i e d as learning disabled have lower Sequential than Simultaneous Processing scores. h) Rank order of K-ABC subtests suggested a l i n g u i s t i c -sequential d e f i c i t with the children performing the lowest on Number Recall and Word Order. C l i n i c a l Interpretation: Given the loading of Hand Movements on the Simultaneous Processing Scale, for which i t was not hypothesized to represent, caution needs to be extended i n interpreting a learning disabled c h i l d ' s Sequential vs. Simultaneous p r o f i l e . The varied c o r r e l a t i o n a l findings between the K-ABC Mental Processing Composite and WISC-R F u l l Scale IQ may be a re s u l t of the small samples resu l t i n g i n unstable c o e f f i c i e n t s . Moreover, the studies reported were not always well defined i n terms of how children were diagnosed as 44 learning disabled, therefore, t h e i r comparability can not be determined. The K-ABC Achievement Scale and PIAT Total Test score appear to be measuring achievement i n a similar fashion which speaks to the concurrent v a l i d i t y of the K-ABC Achievement Scale. F i n a l l y , possible reasons why Sequential/ Simultaneous discrepancies were not consistently evidenced by the K-ABC may be due to the small samples, heterogeneous nature of the samples (learning d i s a b i l i t i e s not distinguished from learning problems), r e s t r i c t e d a b i l i t y range, and the impurity of Hand Movements as a measure of Sequential Processing. C l i n i c a l Interpretation: Mentally handicapped children score, on average, higher on the K-ABC than the WISC-R. Bracken (1985) and Thomas (1983) concluded that the Mental Processing Composite does not have an adequate basal or downward extension for younger children, r e s u l t i n g i n fewer children being Mentally Handicapped N a g l i e r i ( i n press a) Obrzut, Obrzut & Shaw (1984) a) K-ABC Mental Processing Composite has a cor r e l a t i o n i n excess of .80 with the WISC-R F u l l Scale IQ with the two scales d i f f e r i n g by as much as 7 points (in favor of the K-ABC). Kaplan & Klanderman (1984) b) No scores provided, however, the K-ABC i d e n t i f i e d previously diagnosed TMR. diagnosed as mentally handicapped. Therefore, the predictive 45 v a l i d i t y of the K-ABC as a diagnostic measure for id e n t i f y i n g mentally handicapped children requires investigation. C l i n i c a l Interpretation; The magnitude of the discrepancy between the two i n t e l l i g e n c e scales suggests s i g n i f i c a n t l y fewer children would be i d e n t i f i e d as g i f t e d on the K-ABC than on the WISC-R. Bracken (1984), Hessler (1985) and Thomas (1984) attr i b u t e t h i s 'discrepancy to the low c e i l i n g e f f e c t on the K-ABC. The predictive v a l i d i t y of the K-ABC w i l l need to be determined as i t applies to i d e n t i f y i n g children for g i f t e d programs, esp e c i a l l y for programs emphasizing verbal s k i l l s . According to Hessler (1985), when assessing the g i f t e d on the K-ABC i t i s important to consider what eff e c t the low c e i l i n g on some of the subtests and the regression error on the in t e l l i g e n c e and achievement comparison may have on the child' s score on t h i s t e s t . Gifted (1984) McCallum, Karnes & Edwards (1984) McCallum & Karnes a) A moderate cor r e l a t i o n between the K-ABC Mental Processing Composite and WISC-R F u l l Scale IQ was observed along with a discrepancy of 13 points (WISC-R was higher) between these two scales. Hearing Impaired Courtney, Hayes, Walkins & Frick (study #11 i n IM) a) Mean scores for the K-ABC Simultaneous and Nonverbal Scales and the WISC-R Performance IQ were a l l within 46 2 points, however, correlations were only within the moderate range (.60's). C l i n i c a l Interpretation: The correlations indicate that the K-ABC and WISC-R are not measuring i n t e l l i g e n c e the same way, however, with only 40 subjects studied, the results are tentative. The moderate correlations may be unstable due to the small sample. V i s u a l l y Impaired Kaufman (1983) Because the majority (13 of 16) of the subtests are presented v i s u a l l y , the K-ABC would penalize v i s u a l l y impaired children. C l i n i c a l i n terpretation: The K-ABC i s not a recommended test for use with v i s u a l l y impaired children. Minority Groups Providing a f a i r and unbiased assessment for minority children i s a major concern of most psychologists given the number of minority children i n our schools, i s ste a d i l y increasing (Esquirel, 1985). While there i s more than one d e f i n i t i o n of bias i t generally refers to a test's v a l i d i t y (measuring what i t i s supposed to) across groups (Shepard, C a m i l l i , A v e r i l l , 1981). There are three types of v a l i d i t y , namely, content 47 v a l i d i t y , c r i t e r i o n - r e l a t e d v a l i d i t y , and construct v a l i d i t y (American Psychological Association, 1974). Bias as i t relates to these three forms of v a l i d i t y has been defined comprehensively by Reynolds (1982). These d e f i n i t i o n s are as follows: Bias i n Content V a l i d i t y . An item or subscale of a test i s considered to be biased i n content when i t i s demonstrated to be r e l a t i v e l y more d i f f i c u l t for members of one group than another when the general a b i l i t y l e v e l of the groups being compared i s held constant and no reasonable t h e o r e t i c a l rationale exists to explain group differences on the item (or subscale) i n question. (Reynolds, 1982, p. 188) Bias i n Criterion-Related or Predictive V a l i d i t y . A test i s considered biased with respect to predictive v a l i d i t y when the inference drawn from the test score i s not made with the smallest feasible random error or i f there i s constant error i n an inference or prediction as a function of membership i n a p a r t i c u l a r group. (Reynolds, 1982, p. 201) Bias i n Construct V a l i d i t y . Bias exists i n regard to construct v a l i d i t y when a test i s shown to measure d i f f e r e n t hypothetical t r a i t s (psychological constructs) for one group than another or to measure the same t r a i t but with d i f f e r i n g degrees of accuracy. (Reynolds, 1982, p. 194) Numerous s t a t i s t i c a l techniques have been employed to determine the extent to which t r a d i t i o n a l i n t e l l i g e n c e tests are biased against minorities. Mishra (1983) and Murray and Mishra (1983) found the evidence of bias related to the 4 8 s t a t i s t i c a l method used. However, Reschly (1979) concluded that "analyses of data u s u a l l y r e s u l t s i n c o n c l u s i o n s of l i t t l e or no b i a s i n c u r r e n t t e s t s " (p. 230). Jensen (1980) has reported s i m i l a r f i n d i n g s . Regardless of the c o n c l u s i o n s as to the presence of s t a t i s t i c a l b i a s i n t e s t s , i t has been w e l l documented that the use of s t a n d a r d i z e d t e s t s f o r a s s e s s i n g m i n o r i t y c h i l d r e n has r e s u l t e d i n g r e a t e r numbers of m i n o r i t y c h i l d r e n c l a s s i f i e d as mentally r e t a r d e d (Samuda, 1983) and subsequently placed i n s p e c i a l education programs (Mercer, 1973; More & O l d r i d g e , 1980); an o v e r r e p r e s e n t a t i o n of m i n o r i t y c h i l d r e n on non-academic t r a c k s (Samuda, 1983); and a l i m i t e d number of minor-i t y c h i l d r e n being placed i n programs f o r the g i f t e d (Samuda, 1983). Given the assumption that one c u l t u r a l group i s n ' t b r i g h t e r than another, and each has i t s own p a t t e r n of a b i l i t i e s , i t i s o n l y l o g i c a l t h a t the o v e r r e p r e s e n t a t i o n of m i n o r i t y c h i l d r e n i n c l a s s e s f o r the m e n t a l l y handicapped i n d i c a t e s they have not been e f f e c t i v e l y served by s t a n d a r d i z e d t e s t s . Standardized t e s t s r e f l e c t the c u l t u r a l demands of the m i d d l e - c l a s s , m a j o r i t y group (Samuda, 1983). The use of i n t e l l i g e n c e t e s t s i n a c r o s s - c u l t u r a l environment r a i s e s the q u e s t i o n of how a p p r o p r i a t e Western-type t e s t s are f o r a s s e s s i n g c h i l d r e n from non-Western c u l t u r e s (Bhatnagar, 1970, p. 121). According to Reynolds (1982) the reasons most f r e q u e n t l y c i t e d f o r why m i n o r i t y c h i l d r e n , on average, perform 49 less well on standardized i n t e l l i g e n c e tests than majority children includes the test's unfamiliar content (minority children not exposed to the same material); inadequate standardization procedures (minority children underrepresented); language bias (minority children unable to communicate i n p r o f i c i e n t English); lack of construct v a l i d i t y (test i s not measuring same at t r i b u t e for minority children); d i f f e r e n t i a l p r edictive v a l i d i t y (the test i s not predicting a relevant c r i t e r i o n for minority children for academic attainment may be a biased c r i t e r i o n for minority children). As a r e s u l t of the growing f r u s t r a t i o n of minority people with being inappropriately l a b e l l e d as mentally handicapped and in c o r r e c t l y placed i n special classes, the j u d i c i a l system i n the United States began to examine the assessment processes. The outcome of cases, such as Hobson versus Hansen 1967, Diana Versus C a l i f o r n i a State Board of Education 1968, Larry P. versus Riles 1972, 1974, was Public-Laws 94-142 (Education of a l l Handicapped Act). This law mandated that a l l minority •children be tested by non-biased assessment procedures and then provided with appropriate programs of ins t r u c t i o n . While a moratorium on testing minority children was being discussed by the National Education Association (Cbffman, 1974) i t did not receive widespread adoption. Rather, psychologists started to look towards c u l t u r a l - s p e c i f i c and c u l t u r a l - f a i r 50 tests. One problem with c u l t u r a l - s p e c i f i c tests, such as The  BITCH-100 (Williams, 1972), i s that they do not provide information on how a c h i l d i s functioning within the majority culture. C u l t u r a l - f a i r tests were usually nonverbal and abstract i n content. Anastasi (1976) and Jensen (1980) both concluded that nonverbal tests are c u l t u r a l l y bound. Mercer's (1979) approach to providing a f a i r assessment o minority children involved using The System of M u l t i c u l t u r a l  P l u r a l i s t i c Assessment (SOMPA) to compute normal d i s t r i b u t i o n s (by way of regression equations) for various so c i o c u l t u r a l groups based on th e i r performance on the WISC-R and s o c i a l l y descriptive variables. For example, i f a Black c h i l d achieved WISC-R IQ of 85 and came from a large, low-income family, he o she by way of s t a t i s t i c a l l y manipulating t h i s IQ score could achieve an Estimated Learning Potentional (ELP) score of 100 (mean 100, SD 15). According to J i r s a (1983) "The ELP process is descriptive, not pr e s c r i p t i v e - i t does not provide any strategies, by i t s e l f , for increasing a child's school-related competency" (p. 19). J i r s a (1983) concluded that by s t a t i s t i c a l l y manipulating a child' s achieved WISC-R score i t may succeed i n having him or her removed from a special class but i t does not change the child' s current functioning. However, one benefit of the SOMPA was that i t provided psychologists with a model for investigating environmental 51 factors that contributed to a ch i l d ' s performance on a cognitive measure. Much of the p u b l i c i t y the K-ABC has generated has focused on the claim, made by i t s authors, that i t too i s a " f a i r e r " t e s t for use with minority children, implying the K-ABC i s less biased than conventional i n t e l l i g e n c e tests. Mehrens (1984) c r i t i c i z e d the authors of the K-ABC for not i d e n t i f y i n g t h e i r d e f i n i t i o n of bias. However, based on the discussion i n the IM, Mehrens concluded that i t appears the Kaufmans were r e f e r r i n g to diminished white-minority differences on the Mental Processing Composite. He added that "most psychometric experts have long rejected t h i s d e f i n i t i o n of bias" (Mehrens, 1984, p. 308). The main reason i s that although mean white-minority differences have been found on t r a d i t i o n a l i n t e l l i g e n c e tests, these tests have been found to "predict future achievement equally well for the two groups" (Bracken, 1985, p. 31). That i s , the tests are not "biased"; they f a i r l y r e f l e c t the a b i l i t i e s of the tested subjects. However, i t can be argued that i n t e l l i g e n c e tests predict the dysfunctional educational system minority children are forced to function within (Reschly, 1979). In t h e i r attempts to design the K-ABC as a f a i r test for use with minority children, the Kaufmans excluded acquired knowledge items from the " i n t e l l i g e n c e " scale and placed them 52 in the Achievement Scale; reduced the language requirement i n the i n t e l l i g e n c e scale; and subjected the t r i a l items to the scrutiny of Black and Hispanic psychologists to judge for offensive content. They concluded that the K-ABC i s a more f a i r t e s t than the WISC-R, basing t h e i r conclusions on mean group differences. Blacks i n the K-ABC standardization sample were found to score on average 7 points lower than Whites on the Mental Processing Composite. This i s i n contrast to the WISC-R standardization sample where Blacks performed, on average, 15.9 points below the Whites on the F u l l Scale IQ (Kaufman & Doppelt, 1976). As previously mentioned, Mehrens (1984) concluded that reduced mean group difference i s i n i t s e l f not evidence that a tes t i s less biased. The three d e f i n i t i o n s of bias (content, c r i t e r i o n - r e l a t e d , construct) previously presented also do not i d e n t i f y reduced mean differences as evidence that a test i s less biased for certain groups. In addition, a careful examination of the results presented i n the IM indicates that other facts might explain the smaller mean discrepancy found for the K-ABC. SES Sampling A r t i f a c t s . The highest l e v e l of education for the parents of each c h i l d i n the standardization program was the index of socioeconomic status (see Table 3.7, IM). Compared 53 with the 1979 U.S. census data, the Blacks and Hispanics in the standardization program were disproportionately represented. There were fewer Blacks (25.7%) and Hispanics (25.5%) i n the standardization sample who had less than high school education than was true of the Blacks (35.8%) and Hispanics (49.7%) i n the U.S. population. Conversely, there were more Blacks (14.5%) and Hispanics (17.2%) i n the standardization sample with four or more years of university than was true of the Blacks (8.8%) and Hispanics (6.4%) i n the U.S. population. The Whites were not disproportionately represented. It has been widely reported that children with better educated parents achieve higher scores on IQ tests (Bracken, 1985; S a t t l e r , 1982). Data reported i n the IM (Table 4.34) appears to support t h i s claim. As an example, school-aged children with parents having less than a high school education achieved a mean Mental Processing Composite of 93.9, while children with parents having four or more years of college were found to have a mean Mental Processing Composite of 109.2. Since higher SES minority children are over-sampled and lower SES under-sampled i n the K-ABC standardization sample (Bracken, 1985; Jensen, 1984) t h i s sampling a r t i f a c t may have contributed to increasing the mean K-ABC score for minority children. Kaufman (1984) agreed with Kamphaus and Reynolds (1984) that the re a l Black-White discrepancy may be closer to 9 54 p o i n t s given the d i s p r o p o r t i o n a t e number of Blacks sampled at the v a r i o u s e d u c a t i o n a l l e v e l s . A b i l i t y Sampling A r t i f a c t . The authors of the K-ABC attempted to make the s t a n d a r d i z a t i o n sample r e p r e s e n t a t i v e of the a b i l i t y range of the American p o p u l a t i o n . In doing so they i n c l u d e d e x c e p t i o n a l c h i l d r e n i n the s t a n d a r d i z a t i o n program who were i d e n t i f i e d by t h e i r enrollment i n v a r i o u s s p e c i a l education programs. The i n c l u s i o n of these c h i l d r e n was based on the numbers i d e n t i f i e d by the 1980 data from the U.S. Department of Education. Although t h i s i s not a c r i t i c i s m of the s t a n d a r d i z a t i o n procedures, i t has served to make the sample more heterogeneous. T h i s i n turn has c o n t r i b u t e d to reducing the i n t e r g r o u p d i f f e r e n c e s . Jensen (1984) e l a b o r a t e d that i n c r e a s i n g the he t e r o g e n i t y of the K-ABC s t a n d a r d i z a t i o n sample would r e s u l t i n a l a r g e r raw score v a r i a n c e . Although the WISC-R has a set standard score of 100 and s t a n d a r d - d e v i a t i o n of 15 - as does the K-ABC -when both t e s t s are administered to the same groups of c h i l d r e n the K-ABC should have a l a r g e r standard d e v i a t i o n . Among these i n s t a n c e s i n which the a p p r o p r i a t e comparison could be made (given the evidence of the IM, only s i x (or 21%) of the s t u d i e s show a l a r g e r SD [ i t a l i c s included] on the K-ABC than on the comparison t e s t , while 22 (or 79%) of the s t u d i e s show a s m a l l e r SD [ i t a l i c s included] on the K-ABC than the comparison t e s t - a h i g h l y s i g n i f i c a n t ( X 2 = 9.14, 1 d f , p_ < .01) 55 difference, favoring the heterogeneity-hypothesis, (p. 399). The e f f e c t t h i s has on Black-White differences would need to be investigated by comparing the raw score variance on the K-ABC and WISC-R across these two groups. Age Selection A r t i f a c t . Bracken (1985), Jensen (1984) and Na g l i e r i (1985) reported that the difference i n the age range between the K-ABC (2-6 to 12-6 years of age) and the WISC-R (6 to 16-11 years of age) may contribute to the smaller observed White-minority differences, since the magnitude of the discrepancy increases as a function of age. Because the K-ABC has a younger age range than the WISC-R, smaller discrepancies between c u l t u r a l groups i s an expected outcome. Bracken (1985) reported that for the upper age l e v e l on the K-ABC the discrepancy i s closer to 12 points. Subtest Selection A r t i f a c t s . Kaufman and Kaufman (1983b) reported: In selecting items and tasks for the K-ABC, much weight was given to (a) the empirical results of items bias s t a t i s t i c s , using methods developed by Angoff and Rasch; (b) the subjective perceptions and attitudes of two black and two Hispanic educators who were hired to review tasks that have repeatedly been shown to be f a i r c r o s s - c u l t u r a l l y (Kagan & Klein, 1973) or to produce minimal black-white or Hispanic-white differences (Bogen, DeZure, Tenhouten & Marsh, 1972; Gerken, 1978; Jensen & Figueroa, 1975). (p. 15) 56 Jensen (1980, 1984) c r i t i c i z e d the Kaufmans f o r attempting to minimize group d i f f e r e n c e s by s e l e c t i n g s u b t e s t s that have shown s m a l l e r White-minority d i f f e r e n c e s . Kaufman (1984) agreed with Jensen that g i v i n g p r e f e r e n c e to s u b t e s t s demonstrating dimi n i s h e d White-minority d i f f e r e n c e s was not t h e o r e t i c a l l y or p s y c h o m e t r i c a l l y j u s t i f i e d , however, he r a t i o n a l i z e d that from a humanistic p o i n t of view i t was a c c e p t a b l e . In c o n c l u d i n g , Kaufman s t a t e d t h a t s e l e c t i n g s u b t e s t s with diminished r a c i a l d i f f e r e n c e s i s no more biased i n i t s approach than c o n t i n u i n g to use s u b t e s t s that have shown l a r g e r a c i a l d i f f e r e n c e s . The i s s u e , however, i s one of d i f f e r e n t i a l p r e d i c t i v e v a l i d i t y . As p r e v i o u s l y mentioned, Bracken (1985) concluded that although t r a d i t i o n a l IQ t e s t s show s i g n i f i c a n t White-m i n o r i t y d i f f e r e n c e s they s t i l l " p r e d i c t f u t u r e achievement e q u a l l y w e l l f o r the two groups" (p. 31). As p r e v i o u s l y mentioned, perhaps these t r a d i t i o n a l IQ t e s t s p r e d i c t a c h i l d ' s achievement i n a d y s f u n c t i o n a l e d i c a t i o n a l system (Reschly, 1979) and not the c h i l d ' s p o t e n t i a l f o r l e a r n i n g . To date, there are no K-ABC white-nonwhite long range p r e d i c t i v e v a l i d i t y s t u d i e s that have been p u b l i s h e d . T h e r e f o r e , the i s s u e of d i f f e r e n t i a l p r e d i c t i v e v a l i d i t y with the K-ABC i s u n s e t t l e d . (Bracken, 1985, p. 31) 57 Content A r t i f a c t s . Separating the acquired knowledge, comprehension and verbal reasoning subtests from the Mental Processing Composite and placing them i n an Achievement Scale i s another explanation for the reduced White-minority discrepancy on the K-ABC (Bracken, 1985; Jensen, 1984), for some minority children perform less well on these tasks. As an example, the Navajo children studied by N a g l i e r i (in pressb) scored 13 points lower on the Achievement Scale than the Mental Processing Composite. A 14 point difference i n the same di r e c t i o n was found by Valencia (1984a) for Mexican American children. However, N a g l i e r i (1985) found less than a one point difference between these two scales for Blacks. Scale A r t i f a c t s . The i n s u f f i c i e n t basal and c e i l i n g levels for some age groups has the p o t e n t i a l of making the K-ABC less discriminatory at the lower and upper levels (Bracken, 1985; Jensen, 1984; Thomas, 1984). For example, Bracken reported that the Triangles, Matrix Analogies, and Photo Series subtests do not have a s u f f i c i e n t basal u n t i l a f t e r eight years of age. Similarly, Bracken (1985) observed that " i n more than hal f (56%) of the subtests entries from age 9-0 through 12-6 the maximum attainable score i s between 1 1/3 to 2 SD above the mean" (p. 27). Therefore, the i n s u f f i c i e n t basal may work to the advantage of the minority children who have t r a d i t i o n a l l y 5 8 performed lower than Whites on i n t e l l i g e n c e tests. In addition, the i n s u f f i c i e n t c e i l i n g may serve to put the higher functioning white children at a disadvantage. It appears that the smaller discrepancy between the performance of r a c i a l groups on the Mental Processing Composite compared with the WISC-R F u l l Scale IQ may be a function of the K-ABC's construction and standardization procedures. Jensen (1984) stated that to test for Black-White discrepancies on these two tests an investigation would have to employ a matched design with both tests being administered to the same sample of children. N a g l i e r i (1985) followed-up on Jensen's suggestion and administered the K-ABC and WISC-R within one week of each other to 86 pairs of Black and White children. Each pair was matched on age (+ 3 months), gender, and socioeconomic status (highest parent occupation level) and school attended. A l l the children were enrolled i n regular grade 4 and 5 classes and there was no mention of the children having any handicaps. Instead of the 15.9 point Black-White discrepancy found on the WISC-R standardization data (Kaufman & Doppelt, 1976), N a g l i e r i found a 9.08 difference on that test (Blacks: mean 92.30; Whites: mean 101.38). Although the discrepancy was smaller i t was s t i l l s i g n i f i c a n t (£ < .0001). The discrepancy between the two groups on the Mental Processing Composite was 59 6.03 points (Blacks: mean 91.53; Whites: mean 97.56). This discrepancy was s i g n i f i c a n t (p < .0005) and s i m i l i a r to the 7.0 point Black-White difference reported i n the IM for K-ABC standardization sample. N a g l i e r i did not report i f there was a s i g n i f i c a n t difference between the 9 point WISC-R Black-White discrepancy and the 6 point K-ABC Black-White discrepancy. Nevertheless, the discrepancy i s smaller compared to that reported i n the IM suggesting much of the reported reduction i n Black-White differences for the K-ABC may be a re s u l t of comparing unmatched samples of children. Of further inte r e s t i s the finding by N a g l i e r i that the mean difference between the Mental Processing Composite and F u l l Scale IQ for Black children was less than 1 point. As such, p r a c t i t i o n e r s should not assume that the K-ABC w i l l y i e l d higher estimates of o v e r a l l i n t e l l e c t u a l a b i l i t y than the WISC-R children but rather, blacks w i l l l i k e l y earn s i m i l a r WISC-R F u l l Scale IQ and K-ABC Mental Processing Composite means. (Naglieri, 1985, p. 4) Na g l i e r i (1985) recommended that further investigations be conducted to determine differences i n estimates of mental a b i l i t i e s on the K-ABC and WISC-R for other populations of minority children. 60 Mexican American. Valencia (1984ab) investigated the performance on the K-ABC and WPPSI of 42 Mexican American Pre-schoolers (age 53 to 67 months) enrolled i n Head Start Programs. A l l children came primarily from low SES homes and a l l were English-speaking. The K-ABC and WPSSI were administered i n English i n a counterbalanced order from 1 to 51 days apart. There was not a s i g n i f i c a n t difference between the subjects' mean Mental Processing Composite (104.07) and F u l l Scale IQ (102.43). These children did, however, perform better on nonverbal tasks than on verbal tasks. Although a l e v e l of significance was not provided they achieved higher means on the Mental Processing Composite (104.07) than on the Achievement Scale (90.60). The difference between t h e i r mean performance on the Sequential (100.10) and Simultaneous (106.50) Processing Scales was s i g n i f i c a n t (£ < .05). Given, the children did not perform as well on the more verbally oriented Sequential Scale, the Sequential/Simultaneous discrepancy needs to be interpreted cautiously. Navajo. N a g l i e r i (in press* 3) investigated the performance of 35 Navajo children on the K-ABC, WISC-R and PIAT. These children were l i v i n g on a reservation, came from low SES homes, were between the ages of 6 and 12 1/2 years, and were b i l i n g u a l 61 (Navajo was the dominant language). The K-ABC and WISC-R were administered i n a counterbalanced order and i n English. The predictive v a l i d i t y of the K-ABC was determined by administering the PIAT 10 1/2 months afte r the K-ABC. The Navajo children performed s i g n i f i c a n t l y higher (£ < .001) on the K-ABCs Mental Processing Composite (mean 95.0) than on the WISC-R's F u l l Scale IQ (mean 86.9). This difference may, i n part, be attributed to the heavy verbal loading on the WISC-R, for there was a s i g n i f i c a n t discrepancy (p < .001) between the mean subjects' Verbal IQ (75.0) and the i r Performance IQ (102.8). There was also a s i g n i f i c a n t difference (p_ < .001) between t h e i r mean performance on the K-ABC' s Sequential Scale (85.5) and i t s Simultaneous Scale (101.1). N a g l i e r i commented that the Sequential/Simultaneous discrepancy should be interpreted with caution for i t may be an ind i c a t i o n of English language d i f f i c u l t i e s as opposed to a processing d e f i c i t , on the three subtests on the Sequential Scale the children had the lowest mean scores on the two subtests requiring v e r b a l i z a t i o n (Number Recall, 7.2; Word Order, 6.9) compared with the nonverbal Hand Movement subtest (9.9), thus supporting the contention that the Sequential Scale has a verbal requirement. The Achievement Scale was found to be the strongest predictor (R2 = . 6 2 ) of the PIAT Total Test Score, compared to 62 the other K-ABC scales. However, the children performed s i g n i f i c a n t l y lower (p_ < .001) on the Achievement Scale (mean 82.1) than on the PIAT Total Test Score (mean 89.2). N a g l i e r i concluded that the difference may be attributed to the language ac q u i s i t i o n tested on the Achievement Scale. That i s , of the Achievement subtests, the lowest mean score achieved by these Navajo children was 75.7 on Riddles, which i s a measure of verbal analogies. Another possible reason for the difference between the K-ABC Achievement and PIAT scales may be related to the PIAT being standardized approximately 12 years before the K-ABC. Because children are becoming more sophisticated over the years they achieve lower scores on revised tests or tests more recently standardized than on older measures (Doppelt & Kaufman, 1977; Sa t t l e r , 1982; Thorndike, 1977). This also applies to the WISC-R which was normed 10 years before the K-ABC. In summary, i t i s unclear whether the K-ABC has any more relevance or "less bias" for testing Navajo children than does the WISC-R and PIAT combination. At least with the WISC-R examiners are aware of the verbal content and how to interpret i t . With the K-ABC one can not be sure i f a low performance on the Sequential Scale i s in d i c a t i v e of a processing d e f i c i t or a language d e f i c i t . In addition, research on the factor structure of the K-ABC for Native Americans and other c u l t u r a l groups 63 needs to be conducted to determine i f sequential and simultaneous factors emerge. Canadian Children Related to the present study i s the issue of how three subpopulations of Canadian children perform on the K-ABC. Only one summary study has emerged in the l i t e r a t u r e s p e c i f i c to Canadian subjects. Saklofske and J e d l i c k i (1985) investigated the concurrent v a l i d i t y of the K-ABC for 105 English speaking Canadian children l i v i n g in an urban c i t y in Western Canada. Approximately half of the children were 8 years 6 months and the other half were 10 years 6 months. No information was provided on the socioeconomic status of the subjects. Although no scaled scores were provided in this summary study, i t s authors concluded that (consistent with the majority of studies investigating the WISC-R performance of English Canadians - Hardman, 1984; Holmes, 1981; Peters, 1976) the Canadian children studied tended to score higher than the U.S. standardization sample on almost a l l the subtests. The exception was the Faces & Places subtest, which requires the examinees to name pictures of people and places familiar to American children. The finding that English speaking Canadians 64 performed higher on the K-ABC than American children may indicate that the English children i n the present study may also have higher mean scores r e l a t i v e to the American children i n the standardization sample. Vernon (1984) reported Canadian children of Chinese descent (he did not specify language spoken) tend to have higher v i s u a l - s p a t i a l s k i l l s as assessed by the WISC-R Performance IQ than verbal comprehension s k i l l s as assessed by the Verbal IQ. He added that although the Chinese generally do not perform as high on verbal analogies and other English usage tasks as Anglophones, t h e i r performance on the WISC-R Verbal IQ i s generally within the average range and not s i g n i f i c a n t l y lower than Anglophones. From Vernon's description of Chinese Canadians, i t i s hypothesized that the Cantonese i n the present study may show a Simultaneous Processing (Visual-Spatial) strength on the K-ABC. Because of the verbal content on the Sequential Processing Scale and Achievement Scale the performance of the Cantonese on these scales may not be as high as t h e i r Simultaneous Processing score. The performance of Punjabi speaking Canadians on the K-ABC or WISC-R has yet to emerge i n the l i t e r a t u r e . Therefore, no predictions w i l l be made as to the i r expected performance on the K-ABC or WISC-R. There i s a strong l i k e l i h o o d that K-ABC scores across 65 these three groups w i l l show differences among groups, as well as differences between these Canadian children and American children. This i s based on the psychological d i f f e r e n t i a t i o n theory (Vyas, 1983) which assumes that ethnic groups are d i f f e r e n t i n terms of cognitive s t y l e which often manifests as differences i n performance on cognitive measures. 66 CHAPTER III Instrumentation The instruments used i n the present study were the K-ABC, the WISC-R, a Parent Questionnaire, a Teacher Questionnaire, and a Teacher Rating Scale. The following i s a description of each measure. Kaufman Assessment Battery for Children The K-ABC i s comprised of 16 subtests organized i n terms of the three major scales of the K-ABC (viz., Sequential/ Simultaneous/Achievement). These subtests are l i s t e d i n Table 1 together with a b r i e f description of the required examinee response and again i n Table 2 with the i d e n t i f i e d age for which each subtest applies. As shown in Table 2, the maximum number of subtests administered to a given c h i l d i s 13. Further, only those subtests marked (in both Tables) with an asterisk are administered to hearing impaired, language disordered, and/or non-English speaking children forming the Nonverbal Scale. For the 8 to 10 year olds i n t h i s study the test requires approximately 75 minutes to administer and i s recommended to be done i n one setting. Scoring. A l l items administered are scored either pass (1) 67 Table 1 Description of K-ABC subtests Subtests Required Examinee Response Sequential Scale Hand Movements* Number Recall Word Order Simultaneous Scale Magic Window Face Recognition* Gestalt Closure Triangles* Matrix Analogies* Spatial Memory* Photo Series* Achievement Scale Expressive Vocabulary Faces & Places Arithmetic Riddles Reading/Decoding Reading/Understanding Copies a series of handmovements Repeats a series of digits Points to a series of named silhouettes Identifies slowly a partially exposed picture through a narrow s l i t Identifies from a group photograph one or two people previously pictured Identifies incomplete inkblot drawing Assembles triangle pictures to match picture model Selects the best picture or design to complete a visual analogy Identifies placement of previously exposed pictures on an unmarked grid Places photographs in chronological order Names pictures of objects Names pictures of well known people, places, and fictional characters Demonstrates knowledge and understanding of school-related arithmetic problems Names an object after being given a l i s t of its characteristics Identifies letters and reads words Reads words or sentences and performs the command * Nonverbal Scale subtests 68 Table 2 K-ABC subtests by age administered Ages i n Years Scales/Subtests 2.5 7-12.5 Mental Processing Composite Sequential Hand Movements* Number Recall Word order x x X X X X X X X X X X X X X X Simultaneous Magic Window Face Recognition* Gestalt Closure Triangles* Matrix Analogies* Spatial Memory* Photo Series* x x X X X X X X X X X X X X X X X X X X X X X X Achievement Expressive Vocabulary Faces and Places Arithmetic Riddles Reading/Decoding Reading/Understanding x x X X X X X X X X X X X X X X X X X X X X X Total Number Tests Administered: 11 11 12 13 * Subtests i n Nonverbal Scale or f a i l (0). The t o t a l raw score for each subtest consists of the number of passed items and assumes that the items before the s t a r t i n g point would have been answered co r r e c t l y . The raw scores for the Sequential and Simultaneous subtests are converted to scaled scores (mean 10, standard deviation 3), while the raw scores for the Achievement subtests are converted to standard scores (mean 100, standard deviation 15). To obtain the standard scores for the Sequential and Simultaneous Processing Scales, t h e i r respective subtests' scaled scores are summed and then converted to a mean of 100 and a standard deviation of 15. Simi l a r l y , the Achievement standard scores for each subtest are summed and converted to a Global Standard Score (mean 100, standard deviation 15). To obtain the Mental Processing Composite, the Sequential and Simultaneous Processing Standard Scores are added and converted to a standard score with a mean of 100 and a standard deviation of 15. As previously mentioned, the Nonverbal Scale consists of the subtests i d e n t i f i e d i n Tables 1 and 2. Its standard score i s computed by summing the appropriate subtest scaled scores and then converting them to a standard score with a mean of 100 and a standard deviation of 15. Norms. The battery was normed i n 1981 on American children separately for the ages l i s t e d i n Table 2. At each age the 70 standardization sample was s t r a t i f i e d by age, community size, educational placement (regular versus special c l a s s ) , geographic region, sex, socioeconomic status (parental educational attainment), and race (e.g., Black, Hispanic, Native Indian, P a c i f i c Islander, White). Kamphaus and Reynolds (1984) concluded that the o v e r a l l match between the standardization sample and U.S. census data was "quite good, although high SES [socioeconomic status] minorities ( s p e c i f i c a l l y blacks and Hispanics) were s t a t i s t i c a l l y s i g n i f i c a n t l y oversampled" (p. 220). R e l i a b i l i t y . The r e l i a b i l i t y estimates reported i n the IM appear comparable with those of other respected i n t e l l i g e n c e and achievement tests (Kamphaus & Reynolds, 1984). As an example,' the r e l i a b i l i t y c o e f f i c i e n t s for the K-ABC Global Scales for school aged children ranged from a mean (Fisher's Z transformation) of .89 (Sequential Processing) to .97 (Achievement). Siegel and Piottrowski (1985) provided additional evidence of the r e l i a b i l i t i e s for the i d e n t i f i e d a b i l i t y c lusters reported i n the LM. They concluded that the composite r e l i a b i l i t i e s for the K-ABC were generally higher than the WISC-R. These researchers cautioned that although these a b i l i t y clusters may be r e l i a b l e they require empirical v a l i d a t i o n . V a l i d i t y . A t o t a l of 43 v a l i d i t y studies were reported in the IM. According to Kamphaus and Reynolds (1984) this represents "an impressive amount of prepublication research, that i s a l l too uncommon in test manuals" (p. 221). Although these studies were conducted by independent researchers in the United States, the authors of the K-ABC interpreted the re s u l t s . They concluded that, taken together, these studies c i t e d support the v a l i d i t y of the K-ABC for use with a variety of normal and exceptional groups of American children. However, as previously discussed in Chapter II, there i s not general agreement among researchers and theorists for the K-ABC's v a l i d i t y . As such, further studies of the v a l i d i t y of the K-ABC are required. Wechsler Intelligence Scale for Children-Revised The WISC-R i s an i n d i v i d u a l l y administered test of int e l l i g e n c e for children 6-0 through 16-11 years of age. Wechsler (1974) defined i n t e l l i g e n c e in the WISC-R as "the ov e r a l l capacity of an individual to understand and cope with the world around him" (p. 5). The WISC-R, the most frequently used psychoeducational test (Cummins, 1984a), is regularly used as the c r i t e r i o n measure against which other measures of in t e l l i g e n c e , including the K-ABC, are assessed. I t i s both r e l i a b l e and v a l i d (Salvia 72 & Ysseldyke, 1981; Sa t t l e r , 1982). Further, relevant to the present study, Canadian data are available for this instrument (Holmes, 1981; Peters, 1976; Vernon, 1974). Given the widespread use and popularity of the WISC-R, i t is not discussed further here. However, to f a c i l i t a t e the discussion of the relat i o n s h i p between the K-ABC and the WISC-R in the present study a description of the required examinee responses for the 12 WISC-R subtests, organized i n terms of the Verbal and Performance Scales, i s provided i n Table 3. Parent Questionnaire A self-administered parent Questionnaire was constructed to obtain information descriptive of the family background of each tested c h i l d , and to a s s i s t i n explaining differences, i f any, i n performance among the three groups. Seventeen items, organized i n fi v e scales, were developed, and then reviewed by the school board research committee, m u l t i c u l t u r a l workers, and pr i n c i p a l s i n the cooperating school system. The following i s a description of these items. Appendix A contains a copy of the questionnaire written i n the three languages. Languages Spoken i n the Home. Sa t t l e r (1982) and Vernon (1984) concluded that ethnic differences found on cognitive tests can often be attributed to a foreign language spoken i n 7 3 Table 3 Description of WISC-R subtests Subtests Required Examinee Response Verbal Scale Information Similiarities Arithmetic Vocabulary Comprehension Digit Span Performance Scale Picture Completion Picture Arrangement Block Design Object Assembly Coding Mazes Demonstrates knowledge of general facts Identifies commonalities in verbally presented stimuli Demonstrates conceptual and computational understanding of arithmetic problems Defines words Demonstrates understanding of specific customs and mores Recalls digits presented orally Identifies parts missing from pictures Places pictures in a correct sequence Assembles blocks to resemble a pictured model Assembles puzzle pieces Matches symbols then copies them Traces a path through a maze 74 the homes of the children. Therefore, the f i r s t four questions of the questionnaire were asked to obtain a description of the languages spoken i n the homes of the subjects and the frequency with which English was spoken. The m u l t i c u l t u r a l workers commented that i t was not uncommon for adults (e.g., aunts, grandparents, parents) i n the home to speak t h e i r mother tongue while the children respond i n English. Therefore,information related to the language(s) i n the home was e l i c i t e d separately for both the parents and c h i l d . C hild Urbanization. Children raised i n r u r a l communities do not perform as well on i n t e l l i g e n c e tests as urban raised children (Kaufman & Kaufman, 1983b; Sat t l e r , 1982; Vernon, 1984). Questions f i v e and six provided information on the urban status of the subjects by addressing the location of the chi l d ' s birthplace i n Canada along with the length of time the c h i l d has resided in Vancouver. Family s i z e . There i s evidence to suggest that elevated family size may a f f e c t academic success (Mercer, 1979; Sattler, 1982) because parents have less time to spend with each c h i l d (Brody & Brody, 1976). Question seven addressed the number of children l i v i n g i n the home, th e i r gender, and th e i r b i r t h order. This information was to be part of a more involved question dealing with the structure and size of the family. However, questions dealing with the number of adults i n the home and th e i r relationship to the subject were not accepted by the school board research committee for inclus i o n i n the Parent Questionnaire. Consequently, the information col l e c t e d i s a limited indicator of family s i z e . Parent Acculturation. I t i s commonly believed that parents transmit t h e i r b e l i e f s and values to the i r children. Unless l i v i n g i n is o l a t e d communities, the longer a family l i v e s i n the host country the more acculturated they become (Vernon, 1984). Goldman (1973) added, however, that i t takes approximately 20 years for a newcomer to become s o c i a l i z e d . Thus a series of six questions (8 through 13) were asked to obtain information on the parents' birthplace, the population size of th e i r birthplace, the number of years they resided i n Canada and i n Vancouver. As an index of how many generations each family resided i n Canada, the birthplace of the grandparents was also s o l i c i t e d . Socioeconomic Status (SES) . Socioeconomic status has been found to be a predictor of a chi l d ' s performance on int e l l i g e n c e and achievement measures (Mercer, 1979; S a t t l e r , 1982). Kaufman and Kaufman (1983b) used educational attainment as t h e i r measure of SES i n the K-ABC standardization, while 76 Wechsler ( 1 9 7 4 ) used employment and occupational status in the WISC-R. Questions 14, 16 and 17 addressed the educational attainment, employment status, and occupational status of both parents, respectively. Country Educated. At the suggestion of the mul t i c u l t u r a l workers i n the school d i s t r i c t , parents were asked to report the country i n which they received t h e i r highest l e v e l of education (Question 1 5 ) . The m u l t i c u l t u r a l workers added that often educational standards d i f f e r from country to country. Therefore, they f e l t t h i s question would provide additional insight into the qu a l i t y of education the parents received. However, because they had no information on the educational standards of the various schools within each country, t h i s question could not be used comparatively. Religion. Rees ( 1 9 8 3 ) concluded that there are c u l t u r a l and l i n g u i s t i c differences among Punjabi speaking Sikh and Muslim children i n terms of t h e i r b e l i e f s and the language used when p r a c t i c i n g t h e i r r e l i g i o n . However, asking the r e l i g i o n of the children i n the present study was not permitted by the school board committee. Teacher Questionnaire Since the purpose of t h i s study was to investigate the v a l i d i t y of the K-ABC (a test administered i n English, for three language groups), a measure of English fluency for each tested c h i l d was considered. However, because of the two hours required to administer the K-ABC and WISC-R, additional testing time was not permitted. As such, a teacher questionnaire dealing with t h e i r perceptions of the subjects' fluency was included. Moreover, because of the time r e s t r i c t i o n s , the inclusion of an additional achievement test to validate the K-ABC Achievement Scale was not granted. Unfortunately, the children did not have grades or achievement te s t scores on f i l e . Wormeli (1984), having the same problem, assessed the v a l i d i t y of his test - the B r i t i s h Columbia Quick Individual Achievement Test -by determining i f i t discriminated between children receiving remedial i n s t r u c t i o n and those not. This operates under the premise that children receiving assistance i n Arithmetic, for example, should perform lower on the Arithmetic subtest than t h e i r peers not requiring such assistance. This procedure was adopted i n the present study. Thus English Fluency and Academic Remediation data were obtained from teachers using the Teacher Questionnaire (see Appendix A). 78 English Fluency. Ashworth and Wakefield (1978) developed a scale to e l i c i t information on the l e v e l of English language learning of English as a second language (ESL) students. Their scale was adapted for use i n the study. However, some of the wording was changed to make i t s p e c i f i c to the present study. On t h i s questionnaire the teachers were asked to i d e n t i f y the best description of each c h i l d ' s proficiency i n Listening, Speaking, Reading and Writing English. The f i r s t two response levels represented the "beginning and intermediate stages of language learning" while the next two levels were considered "the advanced and t r a n s i t i o n a l stages" (Ashworth & Wakefield, 1978, p. 1). The highest l e v e l was considered to represent fluency comparable to a native English speaker. Academic Remediation. The various forms of remediation or learning assistance offered i n the school d i s t r i c t were i d e n t i f i e d by one of the language consultants. The forms i d e n t i f i e d were: English language i n s t r u c t i o n , arithmetic remediation, written language remediation, and perceptual remediation. The teachers were then asked to indicate which of these forms, i f any, each tested c h i l d was receiving outside of the i r regular classroom. In addition, the teachers were asked to specify, i n hours, the amount of assistance these children were receiving. 79 Teacher Rating Scale A Teacher Rating Scale, adapted from Mercer's (1980) Teacher Scale, was used to provide descriptive information on the subject's academic a b i l i t y . Mercer's Teacher Scale consisted of six, five-point semantic-differential readings. The bipolar adjectives included: I n t e l l i g e n t - Dull-minded, Quick - Slow, Able to Concentrate - Subject to D i s t r a c t i o n , Organized - Disorganized, Good memory - Poor memory, and Persevering - Quitting. The Teacher Rating Scale used i n the present study consisted of f i v e of these items written i n question form. The adjective p a i r i n t e l l i g e n t - dull-minded was deleted since the mu l t i c u l t u r a l workers were concerned that the teachers may consider the wording of th i s p a i r offensive. The remaining items were rewritten as a question and accompanied by a five-point L i k e r t response scale. This format was adopted to ensure a more clear evaluation of each of the behaviours considered. For example, the adjective pair Quick - Slow was reformulated as follows: What i s t h i s student's a b i l i t y to  master new material? The five-point L i k e r t response scale ranged from poor to superior. 80 CHAPTER IV Methodology The main objective of thi s study was to assess the construct v a l i d i t y of the K-ABC for use with three Canadian populations of t h i r d graders. Described i n thi s chapter are the procedures adhered to when selecting and testing the subjects as well as preparing the data for subsequent s t a t i s t i c a l analyses. The preliminary analyses of the K-ABC and WISC-R are also detailed. Population The subjects i n the study were selected from grade three classes i n the Vancouver public school system i n B r i t i s h Columbia, the largest coastal c i t y i n Western Canada, having a population of approximately 425,883 (Vancouver Enumeration Data, 1984). In 1982 there were 29,700 elementary l e v e l students (grades K - 7) attending 74 public schools i n Vancouver, with 14,377 (48.4%) i d e n t i f i e d as speaking English as a second language (ESL) (LaTorre, 1983). Of these ESL elementary pupils, 4,165 (29.0%) spoke Cantonese as their f i r s t language while 1,357 (9.4%) spoke Punjabi. The majority of the Cantonese speaking (57.0%) and Punjabi speaking (69.4%) children were Canadian born. 81 Sample A t o t a l of 210 subjects, 70 i n each of the three language groups, was the sample s i z e . Subject Selection C r i t e r i a 1) The subjects were selected based on the language of the i r home: Cantonese, English or Punjabi. Canadian Native Indian children were not e l i g i b l e for membership i n the English group because there i s some evidence that Canadian Indian children [A. More, personal communication, December, 1984J and American Indian children (Brokerly & Bryde, study #7, c i t e d i n IM) perform d i f f e r e n t l y on the K-ABC than non-Indian children. 2) The subjects were enrolled i n grade three, and ranged from 8 to 10 years of age. This age/grade l e v e l was selected based on the findings of Das et a l . (1979), that " c u l t u r a l preference i n preferred mode of processing was exhibited as early as ages 8 and 9" (p. 31). 3) The subjects were Canadian born. 4) The subjects attended t h e i r present school since the commencement of the academic year. Since the teachers were asked to evaluate t h e i r students' English fluency and learning s t y l e , t h i s allowed the teachers ample time to f a m i l i a r i z e themselves with each subject. 82 5) The children were enrolled in regular, grade three classes, and none of the subjects had documented emotional, mental, physical, or sensory handicaps. There i s evidence that mentally handicapped children, for example, code information d i f f e r e n t l y than average a b i l i t y children (Das, 1972). Children attending remedial assistance classes for part of the school day were not excluded, for these were generally average a b i l i t y children needing extra assistance with English and/or t h e i r academic subjects. At the commencement of the study, i t was not known how many of the minority children were receiving English as a second language i n s t r u c t i o n . As such, i t was decided not to r e s t r i c t t h e i r numbers any further. In addition, there was no consistent procedure for i d e n t i f y i n g children for remedial assistance at the d i s t r i c t l e v e l . The type of remedial assistance received by the subjects ( i f any) was employed as a descriptive variable. 6) The subjects and t h e i r parents agreed to p a r t i c i p a t e i n the study. Selection Procedures The schools having the highest representation of Cantonese and Punjabi speaking children were i d e n t i f i e d by the school board s t a f f i n charge of research. The English children were then selected from these schools. The p r i n c i p a l s of the 29 schools having the highest 83 representation of Cantonese and Punjabi speaking children were asked by the school board s t a f f i f they would consent to the tes t i n g of the children i n t h e i r school. Twenty-one agreed, and were sent a l e t t e r from the P r i n c i p a l Investigator (see Appendix B) i n which the study procedures were described. This was followed by a school v i s i t to answer any questions and ask for the cooperation of the grade three teachers. A l l the p r i n c i p a l s and a l l of the 34 teachers confirmed t h e i r willingness to cooperate. The teachers provided a class l i s t i d e n t i f y i n g t h e i r Cantonese, English and Punjabi speaking students. In addition, the following information was obtained from the children's school record cards: birthdate, birthplace, gender, grade entered present school, i d e n t i f i e d handicaps, names of parents and telephone number. From t h i s information a l i s t of 318 students meeting the selection c r i t e r i a was obtained. Information packages (see Appendix C) were i n d i v i d u a l l y prepared for a l l e l i g i b l e subjects. The contents of t h i s package included: an explanatory l e t t e r , consent form, Parent Questionnaire, and a stamped addressed envelope for the parents to return the information d i r e c t l y to the p r i n c i p a l investigator. The telephone number of the Cantonese or Punjabi speaking m u l t i c u l t u r a l worker was also provided so the parents could address any concerns they had i n their native language. A l l the m u l t i c u l t u r a l workers had previously been consulted i n 84 the development of the Parent Questionnaire and were provided with an information package. The English parents were only-provided with the telephone number of the p r i n c i p a l investigator. A l l parents received an English version of the above l i s t e d material. The Cantonese and Punjabi parents also received a copy of the same information translated into t h e i r native language. The t r a n s l a t i o n of the English information was done by a foreign language service and v e r i f i e d by mu l t i c u l t u r a l workers. Data C o l l e c t i o n Outlined i n t h i s section are the procedures adhered to i n tr a i n i n g the testers and i n the administering of the instruments. Training Testers Eight graduate students i n c l i n i c a l and school psychology were hired and two university professors of school psychology volunteered to a s s i s t the p r i n c i p a l investigator with the tes t i n g . A l l testers were English speaking and had previous t r a i n i n g and experience i n administering the WISC-R. None, however, had administered the K-ABC. Therefore, two, three-hour t r a i n i n g sessions were conducted by the p r i n c i p a l investigator. The f i r s t of these sessions was devoted to 85 the administration of the K-ABC, while the second concentrated on scoring and a review of the administration procedures. An outline of the information covered i n the two tr a i n i n g sessions i s provided i n Appendix D. Test Item Changes Given that the metric system i s taught i n Canadian schools, items stating or requiring responses i n the B r i t i s h engineering system were rewritten i n metric form. This i s a common practice i n Canada (Holmes, 1981; Vernon, 1977). For the K-ABC, two items required changing. On the Arithmetic subtest items 28 and 29 referenced 650 pounds and 550 pounds. These were changed to read 650 kilograms and 550  kilograms (see Appendix E). Four items on the WISC-R (v i z . , Information #20, #24, #27 and S i m i l a r i t i e s #10) were changed to meet the metric c r i t e r i o n (see Appendix E). Answers were accepted i n B r i t i s h or metric form. For Information #20, How many pounds make a ton? was changed to read How many kilograms make a tonne? Information #24, How t a l l i s the average American man? was changed to reference a Canadian man. Information #27, How far i s i t from  New York to Los Angeles? did not require rewording, however the answer was accepted i n metric form. S i m i l a r i t i e s #10, In what  way are a pound and a yard alike? was read f i r s t so the c h i l d had the opportunity to respond "a place to keep a dog" or "both 8 6 measures." If the c h i l d did not answer correctly, he or she was asked, In what way are a kilogram and a metre alike? Testing Material Each tester was provided with a package (see Appendix F) containing: 1) A Testing Procedure Sheet on which were l i s t e d the contents of the test package, as well as the testing procedures to be followed. 2) A Cover Sheet i d e n t i f y i n g , by name, the school, p r i n c i p a l , teacher, and subjects. 3) Request form for Subject P a r t i c i p a t i o n to be read to the subjects before the commencement of tes t i n g . 4) A test Package for each subject which contained: a) a consent form signed by the subjects' parents or guardian; b) a che c k l i s t d e t a i l i n g the data to be collected; c) a Teacher  Questionnaire; d) a Teacher Rating Scale; a K-ABC record form; e) a WISC-R record form; and f) a l e t t e r thanking the c h i l d and his or her guardians for p a r t i c i p a t i n g i n the study. The l e t t e r was to be given to the c h i l d at the end of the tes t i n g . Administration of Tests To avoid an order e f f e c t the WISC-R and K-ABC were administered i n a counterbalanced order. The order i n which each te s t was to be administered to a given c h i l d was 87 previously coded on the test record forms by the p r i n c i p a l investigator. Testing Procedures Testers contacted the schools to which they were assigned to arrange a mutually convenient time with the teachers to s t a r t the tes t i n g . The teachers were responsible for consulting with the p r i n c i p a l to reserve a quiet testing room. The following procedures were to be followed by the te s t e r s : 1) Confirm the subject's birthdate written on the record forms; 2) Allow the teachers to specify the most convenient time for the c h i l d to be removed from the classroom; 3) Encourage the subjects not to discuss the test questions or t h e i r answers with t h e i r peers u n t i l the testing was completed i n t h e i r classroom; 4) Administer the K-ABC and WISC-R i n the order coded ( f i r s t or second) on the record forms; 5) Administer both tests preferably on d i f f e r e n t days and no more than one week apart to avoid fatigue and carry-over from one test to the next; 6) Administer both tests following the directions i n the i r respective manuals, but adhering to the metric s p e c i f i c a t i o n s ; 7) Score only the items and not the t o t a l test to avoid a 88 halo e f f e c t as a re s u l t of both tests being administered within one week of each other by the same tester; and 8) Answer any questions the teachers might have in completing the Teacher Questionnaire and the Teacher Rating Scale. Check to make sure the teachers completed a l l questions. After the testing was completed i n a school the p r i n c i p a l investigator and p a r t i c i p a t i n g teachers were sent a thank-you l e t t e r (see Appendix B). Scoring and Data Preparation Scoring For the K-ABC and WISC-R a multi-step scoring procedure was followed to ensure the accuracy of the scoring. As well, the open-ended response for parent occupation required rating. The remaining items i n the Parent Questionnaire and the two teacher scales could be coded d i r e c t l y ; as such, no scoring was required. K-ABC and WISC-R. Below are the steps followed i n scoring the K-ABC and WISC-R. 1) K-ABC and WISC-R items scored i n the f i e l d were rescored by the p r i n c i p a l investigator. Then a 10% random sample was examined by a q u a l i f i e d school psychologist. A one percent error rate was accepted with items as the unit of 89 analysis. A l l i d e n t i f i e d errors were corrected. 2) The standard and scaled scores for both tests were computed by the p r i n c i p a l investigator with 100% v e r i f i c a t i o n by a q u a l i f i e d school psychologist. Again a l l errors were corrected. 3) The Prorated Achievement Score (PACH) was calculated by summing the standard scores for the Arithmetic, Riddles, Reading/Decoding and Reading/Understanding subtests and divi d i n g by four to obtain a prorated subtest score to replace Faces & Places. The prorated subtest score was summed with the four remaining Achievement subtest standard scores to equal the sum of subtest scores. The PACH could then be obtained by looking at the ACH norm tables. Although this procedure has not been standardized, A. Kaufman (personal communication, A p r i l , 1983) said i t was acceptable for research purposes. Parent Questionnaire. The open-ended response to occupation status (Question 17) required categorizing. The occupations were c l a s s i f i e d into the following f i v e categories s p e c i f i e d by Wechsler (1974): 1. Professional and technical workers. 2. Managers, o f f i c i a l s , proprietors, c l e r i c a l workers, and sales workers. 3. Craftsmen and foremen. 90 4. Operatives, service workers (including private household), farmers and farm managers. 5. Laborers, farm laborers, and farm foremen. (p. 18) Data Preparation The data were coded with 100% v e r i f i c a t i o n . A l l errors were corrected. These data were entered onto computer cards with 100% v e r i f i c a t i o n by a private firm, Elan Data Makers Ltd. Preliminary Analysis Before performing the psychometric analyses to assess the v a l i d i t y of the K-ABC within the three Canadian subpopulations considered i n t h i s study, two preliminary issues were addressed. The f i r s t concerned possible examiner effects, and the second possible differences i n performance on the K-ABC and WISC-R due to order of administration. Examiner E f f e c t . Shown i n Table 4 are the number of subjects tested by each of the 11 examiners. Examination of th i s table reveals that the numbers of subjects tested by the examiners were not equal and i n many instances i n s u f f i c i e n t to s t a t i s t i c a l l y t e s t for an examiner e f f e c t . In addition, any 91 Table 4 Number of subjects each examiner tested by language group Language Group Examiner Cantonese English Punjabi Total n % n % n % n % 1 19 27.1 13 18.6 28 40.0 60 28.6 2 22 31.4 33 47.1 13 18.6 68 32.4 3 2 2.9 1 1.4 2 2.9 5 2.4 4 9 12.9 1 1.4 4 5.7 14 6.7 5 2 2.9 5 7.1 7 10.0 14 6.7 6 0 0.0 1 1.4 1 1.4 2 .9 7 1 1.4 1 1.4 1 1.4 3 1.4 8 2 2.9 7 10.0 1 1.4 10 4.8 9 2 2.9 1 1.4 1 1.4 9 4.3 10 2 2.9 2 2.9 5 7.1 10 4.8 11 9 12.9 5 7.1 7 10 .0 11 5.2 Total 70 70 70 210 differences observed between examiner 1 and examiner 2 (they tested the majority of the subjects) i s confounded by the fact that the children were not a l l taken from the same school and the three groups were not equally represented within each school. Given these situations, possible effects due to examiners were not determined. However, debriefing of each examiner revealed that a l l experienced no d i f f i c u l t y i n administering the test within the usual time l i m i t s , and no one reported any adverse or abnormal test behavior on the part of the c h i l d on the second of the two testing occasions, regardless of the t e s t . Nevertheless, the d i f f i c u l t y with s t a t i s t i c a l l y t esting for an examiner e f f e c t i s a l i m i t a t i o n of t h i s study. Test Order: K-ABC. Test order e f f e c t was examined separately within the three groups using a multivariate analysis and employing the Wilks c r i t e r i o n l e v e l . Tabachnick & F i d d e l l (1983) reported that Wilks' Lambda i s the most frequently employed c r i t e r i o n for i n f e r r i n g population differences. Given the scales were formed by aggregating the subtest scores, t h i s analysis was r e s t r i c t e d to subtests only. The results of the multivariate analysis performed using the computer program S t a t i s t i c a l Package for the Social Sciences (SPSS X) (Nie, 1983), are shown i n Table G-l, Appendix G together with the corresponding univariate F values. To guard 93 against the p r o b a b i l i t y of a Type II error ( f a i l i n g to reject the n u l l hypothesis when i t i s r e a l l y f a l s e ) , the .25 l e v e l of significance was accepted for t h i s analysis. As shown, there was no s i g n i f i c a n t difference between the test order of the two mean vectors for the Cantonese, however, for the English and Punjabi s i g n i f i c a n t differences i n the mean vectors were found at the .15 and .06 levels of significance, respectively. Examination of the corresponding univariate F values for the English and Punjabi revealed s i g n i f i c a n t differences (£ < .25) between the means on three subtests. Given t h i s number was expected by chance at the .25 significance l e v e l , the decision was taken to disregard order and combine the two samples for further analysis. Test Order: WISC-R. Following the same procedures outlined above for the K-ABC, the e f f e c t of test order upon performance on the WISC-R was examined. An examination of the results of t h i s analysis, presented i n G-2, Appendix G, revealed the number of s i g n i f i c a n t subtests was expected by chance. Consequently, the two samples were collapsed for further analysis. S t a t i s t i c a l Methods Given the sequential nature of the s t a t i s t i c a l tests, with 94 each step somewhat determined by the results of the preceding step, the s t a t i s t i c a l methods used i n analysing the data are described together with the results i n the following three chapters. However, the strategy for analyzing the data involved f i r s t describing the biodemographic c h a r a c t e r i s t i c s of the three groups by employing multivariate and univariate analyses, along with Chi Square analysis where appropriate. Investigation of the psychometric properties of the K-ABC followed ( i . e . , central tendency, v a r i a b i l i t y , r e l i a b i l i t y , and in t e r n a l structure), and multivariate and univariate analysis of variance, Pearson c o r r e l a t i o n a l , confirmatory and exploratory factor analyses were employed. F i n a l l y , the relationship between the K-ABC and WISC-R was explored through dependent t-t e s t , Pearson correlation, and q u a l i t a t i v e analyses. 9 5 CHAPTER V Description of Samples Described i n t h i s chapter are the response rates and the biodemographic c h a r a c t e r i s t i c s of the three groups. The description of the samples i s based on biodemographic information c o l l e c t e d from the subjects' school f i l e s , parents and teachers. Rate of Response As reported i n the previous chapter, the desired number of subjects for each of the three groups were secured from 34 classes i n 21 schools. Reported i n Table 5 i s a summary of the response rate at the student l e v e l . Altogether 318 students -115 Cantonese, 108 English, and 95 Punjabi - were i n i t i a l l y i d e n t i f i e d i n the 34 classes as meeting the selection c r i t e r i a . Of these numbers, consent was given for 7 5 Cantonese, 71 English, and 70 Punjabi students. The remainder were accounted for by no response, r e f u s a l , or i n e l i g i b l e (as determined through further screening). The f i r s t 70 e l i g i b l e students tested i n the case of the Cantonese and English and a l l 70 of the Punjabi students formed the f i n a l samples. As shown i n Table 5, completed Parent Questionnaires were received for a l l but one student; completed 96 Table 5 Response rate by language group Response Cantonese Language Group English Punjabi Subjects Approached 115 no response 17 refused 18 i n e l i g i b l e 3 5 E l i g i b l e subjects secured*3 75 Subjects tested 70 Parent Questionnaires' 2 70 Teacher Rating Scales 70 Teacher Questionnaires 70 108 95 18 11 8 71 70 70 70 70 13 7 5 70 70 69 70 70 a Subjects were i n e l i g i b l e because they were found to either speak a language other than the three i d e n t i f i e d for this study or were immigrants. b The f i r s t 210 secured subjects were tested. c One parent refused to complete the questionnaire, however, information on his daughter's e l i g i b i l i t y for t h i s study was secured over the phone. 97 Teacher Rating Scales and Teacher Questionnaires were returned for a l l 210 students. Biodemographic Charact e r i s t i c s of Samples The description of the biodemographic c h a r a c t e r i s t i c s of the three groups i s divided into the following four sections: biodemographic c h a r a c t e r i s t i c s of the students; biodemographic c h a r a c t e r i s t i c s of the students' parents; English language experience of the students; and educational backgrounds of the students. These data were taken from the Parent Questionnaire and teacher scales (See Appendix A) as well as school record cards. In examining the s i m i l a r i t i e s and differences among the three samples, s t a t i s t i c a l procedures appropriate to the type of variable scale were employed. Interval, quasi-interval and r a t i o scales, were subjected to a multivariate and/or a one-way analysis of variance while nominal and ordinal scales were analyzed by a Chi Square ( X 2 ) . A l l analyses were completed using the SPSS X computer program (Nie, 1983) and employing the .05 l e v e l of si g n i f i c a n c e . The s t a t i s t i c a l equation used for each analysis i s i d e n t i f i e d i n each table. Biodemographic Characteristics of Students The biodemographic c h a r a c t e r i s t i c s of the students i n the 98 three samples are summarized separately in Table 6. Gender. The number of females and males was equal i n a l l three samples, as c a l l e d for in the selection of the subjects. Age. The age d i s t r i b u t i o n s of the three samples was comparable. With the exception of two 10 year olds, the students were 8 or 9 years of age. Neither the mean age nor the variance of ages d i f f e r e d s i g n i f i c a n t l y among the groups. Years Resided i n Vancouver. A l l the children i n the present study were Canadian born. The mean length of residency did not d i f f e r s i g n i f i c a n t l y among the groups. Since the students were 8 and 9 years of age, the number of years they l i v e d i n Vancouver indicates that the majority of the children had l i v e d i n Vancouver since b i r t h or infancy. S i b l i n g s . The mean number of s i b l i n g s (children l i v i n g with subjects) did not d i f f e r s i g n i f i c a n t l y among the three samples. While the Cantonese and Punjabi subjects had a maximum of 6 and 5 s i b l i n g s , respectively, one English subject had 11. However, the maximum number of s i b l i n g s for the remaining 69 English subjects was 4. 99 Table 6 Biodemographic c h a r a c t e r i s t i c s of the students by language group Language Group  Demographic Cantonese English Punjabi Variables Gender female male (n) (n) 35 35 35 35 35 35 Age£ 8- 0-0 to 8-11-30 9- 0-0 to 9-11-30 10-0-0 to 10-11-30 mean (years) SD (months) Years l i v e d i n Vancouver 0 mean (years) SD (years) S i b l i n g s 0 maximum minimum median mean 52 18 0 8. 4, 8.07 1.05 6 0 2 1.86 52 17 1 8. 7 4.7 7, 1, 11 0 1 1. 50 76 43 26 1 8.8 5.6 54 5 1 2 1, 87 .65 88 a F (2,207) = 1.84, £ > .05. b F (2,206) = 2.54, p > .05. c F (2,206) = 1.73, £ > .05. 100 Biodemographic Cha r a c t e r i s t i c s of Students' Parents Presented i n Tables 7, 8, 9 and 10 are the biodemographic c h a r a c t e r i s t i c s of the students' parents for each group. B i r t h Location of Parents and Grandparents. As shown i n Table 7, the majority of English fathers (68.8%) and mothers (82.6%) were born i n Canada while less than 3% of the parents in the other two groups were Canadian born. Likewise, while the number of Canadian-born English grandparents was hot as great as1 the English parents, i t was greater than the number of Cantonese and Punjabi grandparents. Examination of the data reveals that the majority of the Cantonese and Punjabi subjects were f i r s t generation Canadian. Parents' Length of Residence i n Canada. Of the three duration periods considered, the majority of the English fathers (78.0%) and mothers (88.4%) l i v e d i n Canada for 21 or more years (see Table 8). The majority of the Cantonese fathers and mothers and the Punjabi fathers had resided i n Canada from 11 to 20 years. In the case of the Punjabi mothers, however, a more equal percentage had l i v e d i n Canada 1 to 10 years and 11 to 20 years. The difference i n years of residency i n Canada between the English parents and the parents i n the other two groups, suggests that the English parents have had longer to acculturate than those i n the two minority groups. 101 Table 7 Number of parents and grandparents born in Canada by language group Language Group Relative Cantonese n % n English % n Punjabi % Father-a 1 1.5 42 68.8 1 1.5 Mother13 1 1.5 57 82.6 2 2.9 Maternal Grandfather0 1 1.5 47 69.1 1 1.5 Paternal Grandfather^ 0 0 32 52.5 0 0 Maternal Grandmother6 2 2.9 49 71.0 0 0 Paternal Grandmother^ 1 1.5 35 57.4 0 0 a X 2 = 110.2, df = 2; £ < .001. ^  X 2 = 1 4 1 . 4 , df = 2; £ < .001. c = 89.3, df = 2; £ < .001. d X 2 = 119.3, df = 2; £ < .001. 6 x Z = 84-1* M = 2; £ < .001. f X 2 = 112.3, df = 2; £ < .001. 102 Table 8 Parents' length of residence in Canada by language group Language Group  Cantonese English Punjabi Years a Father Mother Father Mother Father Mother n % n % n % n % n % n % I to 10 10 14.7 17 24.3 4* 6.8 2 2.9 9 13.8 30 45.5 II to 20 44 64.7 48 68.6 9 15.2 6 8.7 52 80.0 33 50.0 21 or more 14 20.6 5 7.1 46 78.0 61 88.4 4 6.1 3 4.5 Note: Two English mothers reported that a Father: X 2 = 84.5, df = 4; p_ < .001. Mother: X 2 = 148.9, df = 4; £ < .001. the father never resided in Canada. 103 Community Size of Parents' B i r t h Location. Within each language group, the b i r t h location categorized i n terms of population si z e , was sim i l a r for the Cantonese and English fathers and mothers. However, as shown i n Table 9, greater proportions of Punjabi parents were born i n smaller towns or r u r a l areas i n contrast to the greater proportions of Cantonese and English parents born i n larger c i t i e s . The difference among the groups was s i g n i f i c a n t . Socioeconomic Status. Shown i n Table 10 are the representative socioeconomic variables for the three groups. The fathers i n the three groups did not d i f f e r s i g n i f i c a n t l y i n t h e i r highest l e v e l of education. A v a l i d X 2 could not be performed on employment status and occupational status. However, an examination of Table 10 reveals that more of the Cantonese fathers were employed f u l l - t i m e (97.1%) than the English (80.3%) or Punjabi (83.8%). However, a greater number of English fathers were employed i n higher status jobs than the other two groups. The X 2 for mothers' levels of education and occupational status could not be legitimately computed. Nevertheless, an investigation of Table 10 reveals that while a similar percentage of mothers i n each group achieved a university degree, over 50% of the Cantonese and Punjabi mothers did not 104 Table 9 Community size of parents' b i r t h location by language group Language Group  Cantonese E n g l i s h 5 Punjabi Father Mother Father Mother Father Mother Community S i z e a n % n % n % n % n % n % large c i t y > 500,000 28 41 35 51 30 51 37 53 8 12 11 16 small c i t y < 500,000 9 13 14 20 10 17 12 17 7 10 10 15 town < 20,000 15 22 10 15 14 25 12 17 25 37 20 29 farm or r u r a l area 16 24 10 15 5 9 9 13 27 40 27 40 a Father: X 2 = Mother: X 2 = 32 33 • 8, • 3, df = df = 6; 6; £ < E < .001. .001. 105 Table 10 Socioeconomic status (SES) by language group Language Group Cantonese English Punjabi SES Variables Father Mother Father Mother Father Mother % % % % % % Education3 no high school 40.3 51.4 29.3 24.3 31.8 57.3 degree high school degree 19.4 28.6 29.3 32.9 18.2 26.5 some university 19.4 12.9 28.6 34.3 31.8 10.3 university degree 20.9 7.1 13.8 8.6 18.2 5.9 n 67 70 58 70 66 68 Employment0 none 2.9 21.4 18.0 : 47.1 11.8 40.6 part time 0.0 18.6 1.6 25.7 4.4 27.5 ful l time 97.1 60.0 80.3 27.1 83.8 31.9 n 68 70 61 70 68 69 Occupation0 laborer 47.7 51.8 28.0 10.8 51.7 70.3 operator 20.0 14.8 30.0 18.0 27.6 18.9 craftsperson 12.3 22.2 12.0 21.6 10.3 0.0 manager 12.3 11.1 16.0 29.7 8.6 5.4 professional 7.7 0.0 14.0 18.9 1.7 5.4 n 65 54 50 37 58 37 a Father: X 2 Mother: X 2 D Father: X 2 Mother: X 2 c Father: X 2 Mother: X 2 6.21, df = 6; £ > .05. 8.3% of expected cell frequencies < 5. 44.4% of cells with expected cell frequency < 5. 19.30, df = 4; £ < .001. 20.0% of expected cell frequencies < 5. 20% of expected cell frequencies < 5. 106 have a high school diploma compared with 24.3% of the English mothers. More of the English mothers were professionals than the mothers i n the other two groups, however, s i g n i f i c a n t l y more of the Cantonese mothers were employed. English Language Experience of Subjects Presented i n Tables 11, 12 and 13 are the languages spoken in the homes of the subjects, the frequency with which English was spoken by the subjects and th e i r parents, and the l e v e l of the students' English fluency as evaluated by th e i r teachers, respectively. Languages Spoken. As shown i n Table 11, English was the only language spoken i n the English homes. However, over half of the Cantonese parents and students spoke Cantonese and English at home and over h a l f of the Punjabi parents and students spoke Punjabi and English at home. Frequency of English Spoken. As evidenced i n Table 12, the English students and t h e i r parents always spoke English at home. In contrast, the Cantonese and Punjabi parents spoke English ranging from never to most of the time. None of the parents i n these two groups always spoke English at home. The students on the other hand spoke English ranging from never to 107 Table 11 Languages spoken by family member Family Member Group/Language(s) Adults Subject n % n % Cantonese (n=70) Cantonese 15 21.4 1 1.4 Cantonese & English 54 77.1 65 92.9 Cantonese; English & Mandarin 1 1.4 1 1.4 English 0 0.0 3 4.3 English (n=70) EnglTsh 70 100.0 70 100.0 Punjabi (n=69) Punjabi 6 8.7 1 1.4 Punjabi & English 62 89.8 64 92.7 Punjabi, English & Hindi 1 1.4 1 1.4 English 0 0.0 3 4.3 Table 12 Frequency of spoken English by language group 108 Language Group  Frequency o f a Cantonese English Punjabi English Spoken (n=70) (n=70) (n=69) n % n % n % Adults always 0 0.0 70 100. 0 0 0.0 most of the time 8 11.4 0 5 7.2 ha l f of the time 14 20.0 0 23 33. 3 some of the time 34 48.6 0 34 49. 3 never 14 20.0 0 7 10.1 :udents always 12 17.1 70 100. 0 5 7.2 most of the time 18 25.7 0 25 36.2 h a l f of the time 27 38.6 0 29 42.0 some of the time 12 17.1 0 9 13 .0 never 1 1.4 0 1 1.4 a Adults: F (2,206) = 362.31, £ < .001. Students: F (2,206) = 105.32, £ < .001. 109 always. As suggested by the mu l t i c u l t u r a l workers, the students in the two minority groups tended to speak English more often than t h e i r parents. As expected the frequency with which the parents and the subjects spoke English at home d i f f e r e d s i g n i f i c a n t l y among the groups. A Tukey test i d e n t i f i e d the English parents and students as s i g n i f i c a n t l y d i f f e r e n t from the other two groups. There was not a s i g n i f i c a n t difference between the frequency with which the Cantonese and Punjabi families spoke English at home. English Fluency. The means and standard deviations of the English fluency ratings provided by the teachers on the students are reported i n Table 13. The teachers were required to rate the students' pr o f i c i e n c y i n understanding, speaking, reading, and writing English on a f i v e point scale where (0) referred to no proficiency and (4) referred to age appropriate proficiency (see Teacher Questionnaire, Appendix A). Multivariate analysis of variance employing the Wilks c r i t e r i o n revealed a s i g n i f i c a n t difference (F (8.408) = 3.86, £ < .001) among the mean vectors of the three groups. An examination of the corresponding univariate Fs revealed s i g n i f i c a n t differences (at least at .05 level) among the groups for three of the four items (Understanding, Speaking, Writing). The students' fluency with reading English did not d i f f e r s i g n i f i c a n t l y among the groups. A Tukey range test among 110 Table 13 Means, standard deviations and Tukey comparisons for English Fluency items by  language group Test Items c Cantonese Language Group 3 English Punjabi Tukey13 Ccmparisons English Fluency Understands English^ M 3.24 3.69 3.14 E > CP SD .82 .67 .75 Speaks English e M 3.29 3.77 3.21 E > CP Reads English^ SD .82 .59 .78 M 3.67 3.77 3.60 SD .54 .42 .52 Writes English^ M 3.23 3.39 3.03 E > P SD .84 .77 .85 Note: MANOVA. accompanied by a univariate analysis was performed. a n = 70 i n each group. b C = Cantonese, E = English, P = Punjabi. c Responses ranged from: 0 = not fluent to 4 = fluent. d F (2,207) = 10.38, £ < .001. e F (2,207) = 11.85, £ < .001. f F (2,207) = 2.17, £ > .05. 9 F (2,207) = 3.34, £ < .05. I l l the mean pairs revealed that the English students were better at speaking and understanding English than the other two groups. In addition, while the English outperformed the Punjabi i n written English they did not d i f f e r s i g n i f i c a n t l y from the Cantonese. Educational Background The grade the subjects entered t h e i r present school, the subjects' learning s t y l e as assessed by t h e i r teachers, and the type of remedial assistance they were receiving are presented in Tables 14, 15 and 16, respectively. Grade. As shown in Table 14, the majority of the Cantonese (75.7%), English (64.3%) and Punjabi (82.9%) had attended only the school they were presently enrolled i n . There was not a s i g n i f i c a n t difference among the groups on this variable. Learning Style. Presented i n Table 15 are the means, standard deviations, and Tukey comparisons for the teacher ratings of the subjects' learning s t y l e . Multivariate analysis of variance (Wilks c r i t e r i o n ) revealed a s i g n i f i c a n t difference (F (10,406) = 3.09, £ < .001) among the groups. Univariate analysis indicated that the groups d i f f e r e d on each of the f i v e items. Tukey's test of significance among the pairs showed the 112 Table 14 Grade subjects entered present school by language group Language Group5 Grade5 Cantonese English Punjabi n % n % n % Kindergarten K 53 75.7 45 64.3 58 82.9 1 6 8.6 8 11.4 5 7.1 2 6 8.6 8 11.4 1 1.4 3 5 7.1 9 12.9 6 8.6 Totals: 70 70 70 n = 70 in each group. a X 2 = 8.89, df = b; £ < .05. 113 Table 15 Means, standard deviations and Tukey comparisons for the Teacher Rating  Scale by language group Language Group5 Tukey0 Ability 0 Cantonese English Punjabi Comparisons 1. Master new material^ M 3.44 SD .83 2. Concentratee M 3.54 SD .97 3. Retain material^ M 3.51 SD .79 4. Persevere^ M 3.69 SD .97 5. Plan & Organize" M 3.49 SD .96 3.24 3.01 C > P .96 .89 3.07 3.07 C > EP 1.09 .92 3.43 3.01 CE > P .96 .88 3.11 3.13 C > EP 1.08 .95 2.94 3.01 C > EP 1.14 .93 a n = 70/group. ° C = Cantonese, E = English, P = Punjabi. c scale: 1 (poor) to 5 (superior). d F (2,207) = 4.03, £ < .05. e F (2,207) = 5.19, £ < .01. f F (2,207) = 6.49, £ < .01. 9 F (2,207) = 7.40, £ < .001. h F (2,207) = 5.94, £ < .01. 114 Cantonese as having superior a b i l i t y to concentrate, persevere, and plan/organize than the other two groups. In addition, the Cantonese were rated higher than the Punjabi on t h e i r a b i l i t y to master new material. Both the Cantonese and the English were rated s i g n i f i c a n t l y higher than the Punjabi on t h e i r a b i l i t y to retain material. Remedial Assistance. As shown i n Table 16, 15.7% of the Cantonese, 8.6% of the English, and 27.1% of the Punjabi were receiving some form of remedial assistance. S p e c i f i c a l l y , 18.6% of the Punjabi compared with 8.6% of the Cantonese were receiving remedial i n s t r u c t i o n i n English. In addition, fewer Cantonese (2.9%) were receiving assistance i n reading compared with 8.6% of the English and 15.7% of the Punjabi. It should be noted that the children i n each school usually received remedial assistance based on: 1) the i r need r e l a t i v e to others i n t h e i r school; and 2) the a v a i l a b i l i t y of time with the learning assistance teacher. Furthermore, there was no d i s t r i c t t esting of a l l children to i d e n t i f y the children for remedial assistance. As such, i t i s possible that a l l those i n need were not attending remedial assistance classes. One teacher commented that because a l l of her children spoke English as a second language only those most i n need received English as a second language assistance. Therefore, the v a l i d i t y of th i s variable (what i t i s measuring), i s Table 16 Number and percentage of subjects receiving remediation by  language group Language Group a Cantonese English Punjabi Types of remediation n % n % n % English remediation 6 8.6 1 1.4 13 18.6 Reading remediation 2 2.9 6 8.6 11 15.7 Written language remediation 2 2.9 1 1.4 8 11.4 Arithmetic remediation 1 I - 4 2 2.9 3 4.3 "Other" remediation 0 0.0 1 1.4 1 1.4 Total Receiving Remediation 0 c 11 15.7 6 8.6 19 27 .1 n = 70/group. ° Total Receiving Remediation refers to the o v e r a l l t o t a l of subjects receiving remediation i n each group. Some subjects were, however, receiving remediation i n more than one of the s p e c i f i c areas. 116 questionable. The teachers were also asked to document the amount of assistance ( i n hours) each c h i l d was receiving ( i f any). A number of teachers reported that they were either unsure of the amount of time or that the time varied depending on the a c t i v i t y schedule of the week. Due to the questionable v a l i d i t y of the question dealing with the type of remediation a c h i l d was receiving, and the amount of incomplete data on the number of hours of remediation he or she was receiving, the remediation variables w i l l not be dealt with further. CHAPTER VI Psychometric Properties of the K-ABC Presented i n t h i s chapter are the psychometric character-i s t i c s of the K-ABC as determined separately for the three language groups i n t h i s study. Included are the central tendency and v a r i a b i l i t y estimates, the r e l i a b i l i t y estimates, and the in t e r n a l structure (confirmatory and exploratory factor analyses). Interpretation of group differences noted are discussed i n Chapter VIII i n r e l a t i o n to the biodemographic variables previously presented. Central Tendency and V a r i a b i l i t y The means and standard deviations for the subtests and scales are reported i n Table 17 for each group. These were computed using the SPSS X computer program (Nie, 1983). Multivariate analysis of variance, using Wilks' c r i t e r i o n , of the subtest scaled scores revealed that there was a s i g n i f i c a n t difference (F (26,390) = 7.47, p < .001) among the mean vectors of the three groups. To determine where the three groups d i f f e r e d , each subtest was then analyzed separately by employing a one-way analysis of variance, and where differences were found, Tukey's test of the s i g n i f i c a n t difference among pairs of means was performed. The F ratios and F p r o b a b i l i t i e s for each subtest can be found i n Appendix H. 118 Table 17 K-ABC means, standard deviations, and Tukey comparisons for each group Language Group Scales/Subtests Cantonese x SD _English x SD _Punjabi Tukeyb x SD Comparisons Sequential Hand Movements 9. 47 2. 99 9. 79 2. 47 9. .73 2. 52 NS Number Recall0 9. 87 2. 39 10. 86 2. 46 9. 61 2. 67 E > P Word Order^ 10. 19 2. 75 10. 76 2. 11 9. 47 2. 33 E > P Simultaneous Gestalt Closure 10. 63 2. 64 10. 31 2. 97 8. 79 2. 84 CE > P Triangles 12. 70 2. 08 11. 67 2. .38 9. 54 2. 70 C > E > P Matrix Analogies 10. 97 2-45 10. 41 2. 19 9. 44 2. 26 CE > P Spatial Memory 11. 10 2. 47 10. 23 2. 10 9. 27 1. 85 C > E > P Photo Series 11. 56 2. 38 11. 46 2. 69 9. 54 1. 98 CE > P Achievement Faces & places 90. 10 11. 04 91. 53 11. 62 80. 97 10. 78 CE > P Arithmetic 101. 79 11. 57 103. 56 12. 45 95. 32 10. 83 CE > P Riddles 94. 24 11. 41 105. 14 10. 82 89. 13 9. 30 E > C > p Reading/E>ecoding 105. 06 8. 85 104. 04 10. 58 102. 96 7. 63 NS Reading/Understanding 102. 79 9. 39 103. 91 11. 19 97. 84 6. 81 CE : > P Global Scales Sequential® 98. 81 12. 56 102. 76 10. 90 97. 33 10. 75 E > P Simultaneous 109. 54 10. 26 105. 53 11. 78 95. 41 10. 68 CE : > P Mental Processing 106. 07 9. 95 104. 73 10. 98 95. 43 9. 50 CE : > P Achievement 98. 14 9. 72 101. 73 10. 59 91. 91 7. 08 CE : > P ' Prorated Achievement 100. 84 9. 78 104. 64 10. 46 95. 39 7. 31 E > c > p Nonverbal 107. 87 10. 84 104. 59 11. 37 96. 04 10. 64 CE : > p a n = 70 for each group. D NS = not significant, C = Cantonese, E = English, P = Punjabi. c d e When controlling for the Type 1 error rate at the subtest (.05/13 = .004) and scale (.05/6 = .008) level, Word Order and Number Recall (subtests) and Sequential Processing (scale) were no longer significant. In addition, the Bonferroni method (Harris, 1975) was employed to control for the effects of the Type 1 error rate (rejecting the null hypothesis when it is really true) on a group of dependent variables. Timm (1975) advocated when finding a significant overall MANOVA test, the Bonferroni method should be applied on each of these variables. However, Stevens (1972) concluded the Bonferroni just adds an additional restriction on the significance level. Nevertheless, given the general acceptance of the Bonferroni method (Bray & Maxwell, 1985) it was also employed in this study. As shown in Table 17, significant differences (ANOVA) were found among the mean performance of the three groups on 11 of the 13 subtests. However, when controlling for Type 1 error rate (Bonferroni method), Number Recall and Word Order were no longer significant. As a result none of the Sequential Processing (SEQ) subtests differed significantly among the groups. On a l l five Simultaneous Processing (SIM) subtests the English and Cantonese had means greater than those of the Punjabi. Further, the Cantonese outperformed the English on Triangles and Spatial Memory. On the four significant Achievement (ACH) subtests, the Cantonese and English outperformed the Punjabi. On the Riddles subtest the English also outperformed the Cantonese. The groups did not differ significantly in their performance on Reading/Decoding. The Bartlett-Box Homogeneity of Dispersion Test did not evidence a significant difference (0^ > .05) among the groups in 120 thei r corresponding variance-covariance matrices. The standard deviation for the Mental Processing subtests ranged from 1.98 to 2.99, and the Achievement subtests ranged from 6.81 to 11.62. Multivariate analyses of variance (Wilks 1 c r i t e r i o n ) of the noncomposite Global Scales ( v i z . , SEQ, SIM, ACH) revealed a s i g n i f i c a n t difference (F(6,110) = 5.79, £ < .001) among the mean vectors of the three groups. Subsequently, a l l Global Scales were subjected to a one-way analysis of variance, and where differences were found, Tukey's test of s i g n i f i c a n t differences among pairs of means was performed. The F rati o s and F p r o b a b i l i t i e s for differences among the groups for each scale can be found i n Appendix H. As shown i n Table 17, s i g n i f i c a n t differences (ANOVA) were observed on a l l Global scales. The English means were s i g n i f i c a n t l y greater than the corresponding means for the Punjabi. Moreover, the Cantonese outperformed the Punjabi on a l l scales except for the SEQ scale. However, when c o n t r o l l i n g for the effects of a Type 1 error, the SEQ scale no longer showed s i g n i f i c a n t differences among the groups. Noteworthy i s the 10 point discrepancy between the SEQ and SIM Scales for the Cantonese, the l a t t e r being the higher (t(69) = 6.19, £ < .001). The discrepancy between these two scales for the English and Punjabi was approximately 3 and 2 points, respectively. Their discrepancies were similar to the 2 point discrepancy found for the 182 normal children reported i n 121 the IM (p. 113). The 10 point discrepancy found i n the Canton-ese group may suggest that they have r e l a t i v e l y superior simultaneous and/or v i s u a l s p a t i a l a b i l i t i e s compared with t h e i r sequential a b i l i t i e s , for they performed higher, on average, on a l l the SIM subtests than the SEQ subtests. This may be related to a c u l t u r a l l y s p e c i f i c cognitive strength or an a r t i f a c t of the te s t . I t w i l l be discussed in more depth i n Chapter IX. The mean ACH Scale score was s i g n i f i c a n t l y lower than the mean MPC for the Cantonese (t(69) = 6.33, £ < .001), for the English (t(69) = 2.56, £ < .013), and for the Punjabi (t(69) = 3.48, £ < .001). For the 182 normal children reported i n the IM, there was less than a 1 point discrepancy (not s i g n i f i c a n t ) between t h e i r mean MPC and ACH scores. The lower performance of the three Canadian groups on the ACH Scale compared with t h e i r American peers appears related to th e i r poor performance on Faces & Places. This subtest has content s p e c i f i c to the American culture. When Faces & places was excluded from the ACH Scale and the Prorated Achievement (PACH) Scale score computed, a MPC/PACH discrepancy was not evidenced for the English or the Punjabi. However, for the Cantonese, the 6 point discrepancy between the MPC and PACH scores may indicate that they have better v i s u a l - s p a t i a l s k i l l s than verbal a b i l i t i e s , which was also found by Lesser, F i f e r and Clark (1965) and Vernon (1984). Compared with the set standard deviation of 15 for the 122 K-ABC norm group, the highest standard deviation found for the three groups i n thi s study was 12.56 (Cantonese, SEQ) and the lowest was 7.08 (Punjabi ACH). Given that the K-ABC norm group included exceptional children (children with emotional, mental, motor, sensory handicaps) and these children were excluded from p a r t i c i p a t i n g i n the present study, the more homogenous nature of the three samples i n the present investigation was an expected outcome. Further, Das (1972) found that retarded and nonretarded children have d i s t i n c t processing modes, that i s , not only do the retarded children perform less well on many of these measures, but th e i r psychometric p r o f i l e s also d i f f e r from the nonretarded children. Consequently, generalizations, with regard to processing mode, from nonretarded to retarded children can not be made. In addition to the three groups being more homogenous than the standardization sample because of this study's imposed r e s t r i c t i o n i n a b i l i t y range, the data for each group has been analyzed separately. This also introduced homogeneity, i n the c u l t u r a l sense. Nevertheless, given the purpose of t h i s study was to examine the v a l i d i t y of the K-ABC for three c u l t u r a l groups of nonretarded children, the data were neither pooled nor corrected for r e s t r i c t i o n of range. R e l i a b i l i t y The r e l i a b i l i t y of the K-ABC was examined through i t s 123 internal consistency estimates, standard errors of measurement and i n t e r c o r r e l a t i o n s of i t s subtests and scales. Internal Consistency The int e r n a l consistency estimates reported in Table 18 for the subtests were computed using the odd-even cor r e l a t i o n corrected with the Spearman-Brown Prophecy Formula (Ferguson, 1981, p. 438). To avoid spurious re s u l t s , only the items attempted were used; items below the basal l e v e l and items above the c e i l i n g were discarded for each student. The use of the common s p l i t - h a l f procedure d i f f e r s from the procedure used i n the development of the K-ABC. There, the Rasch-Wright procedure (Robertson & Eisenberg, 1981) was adopted. Because of the small samples, t h i s procedure could not be replicated i n the present study. As a res u l t , these c o e f f i c i e n t s are not d i r e c t l y comparable. Examination of the i n t e r n a l consistency c o e f f i c i e n t s reported i n Table 18 reveals the subtest c o e f f i c i e n t s ranged between .61 and .89 for the Cantonese sample, .52 and .91 for the English sample, and .64 and .82 for the Punjabi sample. Further, of the 13 subtests - 3 subtests for the Cantonese, 6 subtests for the English, and 2 subtests for the Punjabi were below .70. These low c o e f f i c i e n t s may be related to the homogeneous performance of the groups on these subtests. Of the inte r n a l consistencies of the composites (Guilford, 124 Table 18 K-ABC internal consistency reliabilities ( r C T ) and standard errors of measurement.  (SEM) for each group Language Group5 Subtests/Scales Cantonese English Punjabi IM° IM° rxx SEM rxx SEM r x x SEM r x x SEM Sequential Hand Movements .81 1.30 .67 1.42 .72 1.33 .79 1.4 Number Recall .72 1.26 .68 1.39 .70 1.46 .80 1.3 Word Order .80 1.23 .52 1.46 .77 1.12 .88 1.0 Simultaneous Gestalt Glosure .61 1.65 .63 1.81 .75 1.42 .71 1.6 Triangles .73 1.08 .86 .89 .79 1.24 .84 1.2 Matrix Analogies .81 1.07 .83 .90 .68 1.28 .87 1.1 Spatial Memory .76 1.21 .69 1.16 .64 1.11 .85 1.2 Photo. Series .74 1.21 .73 1.09 .80 .85 .82 1.3 Achievement Faces and Places .84 4.42 .87 4.19 .82 4.57 .86 5.6 Arithmetic .60 7.32 .85 4.82 .81 4.72 .86 5.6 Riddles .80 5.10 .63 6.58 .74 4.74 .87 5.4 Reading/Decoding .67 5.08 .84 4.23 .75 3.81 .93 4.0 Reading/Understanding .89 3.11 .91 3.36 .79 3.12 .95 3.4 Sequential Processing .86 4.70 .76 5.33 .80 4.81 .90 4.7 Simultaneous Processing .87 3.70 .89 3.91 .88 3.70 .93 4.0 Mental Processing Composite .89 3.30 .89 3.64 .89 3.15 .95 3.4 Achievement .92 2.75 .94 2.59 .91 2.12 .97 2.6 Prorated Achievement .90 3.13 .92 2.93 .90 2.34 Nonverbal .87 3.90 .90 3.64 .88 3.72 .94 3.7 Note: The IM were computed using the Rasch-Wright method; as a result, they are not directly comparable with the r x x for the three groups in this study, which were computed by an odd-even method. a n = 70 in each group. r x x for 8 year olds in K-ABC standardization sample. c SEM for 8 year olds in K-ABC standardization sample. 1954, p. 393), shown in Table 18, only one was less than .80 (English, .76). The r e l i a b i l i t y of the MPC ( i n t e l l i g e n c e scale) was .89 for a l l three groups. The magnitude of this scale's c o e f f i c i e n t substantiates the high r e l i a b i l i t y of the MPC. Si m i l a r l y , high r e l i a b i l i t y c o e f f i c i e n t s , i n excess of .90, were found for the Achievement Scale for a l l three groups. For the most part the i n t e r n a l consistency c o e f f i c i e n t s for the three c u l t u r a l groups are lower than those reported i n the IM. As shown in Table 18, a l l of the subtest c o e f f i c i e n t s for the 8 year olds i n the standardization sample were above .700 with a l l of the scale c o e f f i c i e n t s above .900. As previously mentioned, the more homogeneous nature of the three c u l t u r a l groups i n t h i s study compared with the 8 year olds in the standarization sample i s probably the main contributing factor . Standard Error of Measurement The standard error of measurement (SEM) (Ferguson, 1981, p. 442) for the subtests and scales are presented i n Table 18. These were computed using the standard deviations of the respective samples (Table 17) and the internal consistency c o e f f i c i e n t s (Table 18). For the MPC subtests the SEM ranged between 1.07 and 1.65 for the Cantonese, .89 and 1.81 for the English, and .85 and 1.46 for the Punjabi. For the ACH subtests the SEMs were a l l higher than on the MPC as a r e s u l t of the 126 d i f f e r e n t standard score metric. Overall, they ranged between 3.12 and 7.32. These are comparable to those reported in the IM for the 8 years olds (see Table 18). I t i s probable that the 200, 8 year olds i n the standardization sample did not have a s u f f i c i e n t number of handicapped children to r e s t r i c t s i g n i f i c a n t l y t h e i r a b i l i t y range. The SEMs for the 8 year olds i n the standardization are somewhat lower than the SEMs reported for the 1500 school-aged children. The SEM for the scales ranged from 2.75 to 4.70 for the Cantonese, 2.59 to 5.33 for the English, and 2.12 to 4.81 for the Punjabi. These are generally comparable i n magnitude to those for the 8 year olds reported i n the IM (see Table 18). Intercorrelations The degree of relationship among the K-ABC subtests and among the K-ABC scales was determined by i n t e r c o r r e l a t i n g the components. As presented i n Tables 1-1, 1-2, and 1-3 i n Appendix I, the in t e r c o r r e l a t i o n s among the subtests were within the low to moderate range. For the Cantonese and Punjabi groups a t o t a l of eight negative c o e f f i c i e n t s were observed. These are lower than those i n the IM for the entire school-aged population, perhaps as a re s u l t of the homogeneity induced i n the present study. Anastasi (1982) reported that low int e r c o r r e l a t i o n s among the subtests are desirable i n a multisubtest battery. 127 The correlations between the noncomposite K-ABC Global Scales (scales which do not have overlapping subtests) for each group are shown i n Table 19. As shown, the SEQ and SIM scales had low c o e f f i c i e n t s (.20 Cantonese, .34 English, .21 Punjabi), with each other, however, both scales correlated more strongly with the ACH scale. These c o e f f i c i e n t s were s i g n i f i c a n t (at least £ < .05). The MPC and ACH Scale correlated s i g n i f i c a n t l y (£ < .001) with each other for the Cantonese (.43), English (.59) and Punjabi (.51). This suggests that there i s s i g n i f i c a n t shared variance between the MPC and ACH to indicate that the K-ABC i s not completely successful in separating problem-solving from acquired knowledge. As previously mentioned, i t was not the intent of th i s study to correct the c o e f f i c i e n t s for r e s t r i c t i o n of range. Nevertheless, as an example of what the r e s u l t might be, selected correlations between the noncomposite K-ABC Global Scales have been corrected. These corrected c o e f f i c i e n t s resulted by f i r s t correcting the r e l i a b i l i t y c o e f f i c i e n t s for each comparative scale and then correcting the c o e f f i c i e n t r e s u l t i n g from the co r r e l a t i o n between these scales (Gullicksen, 1950; Nunnaly, 1970). As shown i n Table 19, while the corrected c o r r e l a t i o n c o e f f i c i e n t s were higher than the uncorrected c o r r e l a t i o n c o e f f i c i e n t s , they were not substa n t i a l l y higher. 128 Table 19 Uncorrected and selected corrected correlations between noncomposite K-ABC  Global Scale Standard Scores for each group Language Group3 Comparisons Cantonese English Punjabi Coefficients 0 r l r r l Sequential Processing with -Simultaneous Processing -Achievement -Prorated Achievement .20 .35** .39*** .21 .37 .34** .41*** .41*** .37 .45 .21** .23 .34** .37 .40*** Simultaneous Processing with -Achievement .34** -Prorated Achievement .34** .36 .50*** .53 .50*** .50*** Mental Processing with -Achievement -Prorated Achievement .43*** .46 .46*** .59*** .58*** ,62 .56*** .56*** Nonverbal with -Achievement -Prorated Achievement .38*** .40*** .50*** .50*** !55*** Note; Cantonese corrected reliabilities: SEQ (.90), SIM (.94), ACH (.97) and MPC (.95). English corrected reliabilities: SEQ (.87), SIM (.93), ACH (.97) and MPC (.94). Punjabi corrected reliabilities: SEM (.90), SIM (.94), ACH (.98) and MPC (.95). a n = 70 in each group. ° Coefficients r = uncorrected, r^ = corrected. * £ < .05. ** p < .01. *** £ < .001. 129 Although not composite scales, the Nonverbal (NVER) Scale was correlated with the MPC and the Prorated Achievement (PACH) Scale with the ACH scale as both comparisons are relevant to the present investigation. As shown i n Table 20, the NVER-MPC relati o n s h i p for the three groups exceeded .830 and was s i g n i f i c a n t (£ < .001). Kaufman and Kaufman (1983b) concluded that high a cor r e l a t i o n between these two scales provides an indi c a t i o n that the NVER i s a good estimate of MPC. Simi l a r l y , the ACH-PACH relationship was s i g n i f i c a n t (£ < .001) and higher than .940 for each group. As such, the PACH may be a good estimate of ACH. However, further investigations w i l l be needed to determine how v a l i d the PACH score i s for Canadian children. Given the strong relationships between the MPC-NVER Scale and the ACH-PACH Scales, only the MPC and ACH Scale w i l l be used i n the following analyses. Internal Structure To test the in t e r n a l structure of the K-ABC for the subjects i n the standardization program, Kaufman and Kaufman (1983b) performed confirmatory and exploratory factor analyses. More recently, Keith (1985) reported on the results of a confirmatory factor analysis on the K-ABC standardization data for three age groups - 5, 7 and 10 year olds - and Kaufman and Kamphaus (1984) published a detailed description•of the 130 Table 20 Correlations between selected K-ABC composite scales for  each group Language Group a Comparisons" Cantonese English Punjabi Nonverbal with Mental Processing .86* .91* .83* Prorated Achievement with Achievement .98* .98* .94* a n = 70 i n each group. * p < .001. 131 procedures and outcomes of the various exploratory factor analyses performed on the K-ABC standardization data. The confirmatory procedures reported by Keith (1985) and exploratory procedures reported by Kaufman and Kamphaus (1984) were approximated i n the present study. The use of confirmatory analyses allowed an examination of the extent to which the data f i t t e d the th e o r e t i c a l model; the exploratory analyses allowed for further investigation of the factor loadings of s p e c i f i c subtests. The COSAN program (Fraser, 1980) was used to perform the confirmatory analyses. In t h i s program the output yi e l d s a maximum l i k e l i h o o d s t a t i s t i c referencing the goodness of f i t of the data with the model. A res u l t i n g matrix i l l u s t r a t i n g the loadings of the subtests on th e i r hypothesized factor i s produced. The exploratory factor analyses were computed using the Alberta General Factor Analytic program (AGFAP) (Hakstian & Bay, 1973). In each case, the K-ABC subtests were analyzed i n two stages. F i r s t , the eight Mental processing subtests were analyzed to determine the magnitude of the i r respective loadings on the i d e n t i f i e d factors. Next, the 13 Mental Processing and Achievement subtests were analyzed together to determine the influence of the Achievement subtests upon the factor structure of the Sequential and Simultaneous subtests (Kaufman & Kamphaus, 1984). Because the purpose of this study was to investigate the v a l i d i t y of the K-ABC for use with three groups, each group was analyzed separately. Child (1973) and Tabachnick and F i d e l l (1983) pointed out that pooling data from d i f f e r e n t groups may obscure factors and factor loadings for a p a r t i c u l a r group. Confirmatory Factor Analysis The purpose of confirmatory factor analysis was to determine whether the data f i t the sp e c i f i e d t h e o r e t i c a l model. In t h i s analysis the model (target matrix) was s p e c i f i e d a p r i o r i and represents the K-ABC subtest-scale match. As shown in Table 21, there are three subtests on the hypothesized Sequential factor, and f i v e subtests on the hypothesized Simultaneous factor. In the COSAN program the target matrix i s entered i n simple structure with the subtests hypothesized to load on the expected factor given a loading of .5 and the subtests not hypothesized to load on the i d e n t i f i e d factor given a loading of .0. The factor variance was set at 1.0. The target matrix for the 13 subtest solution i s also shown i n Table 21. These factors are independent. The .05 l e v e l of significance was i d e n t i f i e d as being acceptable as a determiner of the goodness of f i t of the data-model (Keith, 1985). Factor loadings above .350 were considered s a l i e n t . Mental processing Subtests. Shown i n Table 22 are the re s u l t s for the confirmatory analyses for the eight subtest/two 133 Table 21 Theoretical target matrix for the two factor and three factor model Scales/Subtests SEQ Target Matrix  2 factors 3 factors SIM SEQ SIM ACH Sequential Hand Movements Number Recall Word Order Simultanaeous Gestalt Closure Triangles Matrix Analogies Photo Series Achievement Faces and Places Arithmetic Riddles Reading/Decoding Reading/Understanding .5 .5 .5 .0 .0 .0 .0 .0 .0 .0 .5 .5 .5 .5 .5 .5 .5 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .5 .5 .5 .5 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .5 .5 .5 .5 .5 134 Table 22 Confirmatory Factor Analysis 5 of Mental Processing subtests for each group Language Group0 Cantonese English Factors 0 Subtests SEQ SIM SEQ SIM Punjabi SEQ SIM Sequential Hand Movements Number Recall Word Order 30 96 51 35 66 66 09 39 97 Simultaneous Gestalt Closure Triangles Matrix Analogies Spatial Memory Photo Series 31 89 51 23 37 38 70 61 46 65 51 57 70 50 45 Goodness of F i t Statistics Chi square df probability 30.86 19 < .05 20.89 19 > .05 21.31 19 > .05 Notel; Decimals have been omitted. Note^: Only loadings for subtests on targeted factor are produced. a Maximum likelihood estimation. D n = 70 for each group. c SEQ = Sequential, SIM = Simultaneous. f a c t o r s o l u t i o n . For the Cantonese group, the Hand Movements, G e s t a l t C l o s u r e , and S p a t i a l Memory subt e s t s had f a c t o r l o a d i n g s below the s e l e c t e d .35 s a l i e n c e c r i t e r i o n . T h i s may have c o n t r i b u t e d to the s i g n i f i c a n t X 2 value (£ < .05). The high X 2 value suggests that the Cantonese data does not f i t the s p e c i f i e d t a r g e t matrix b r i n g i n g i n t o q u e s t i o n the v a l i d i t y of the K-ABC Sequential/Simultaneous dichotomy f o r t h i s group. The E n g l i s h data f i t t e d the Sequential/Simultaneous model. A l l f a c t o r l o a d i n g s met the .35 s a l i e n c e c r i t e r i o n . The X 2 was not s i g n i f i c a n t at the .05 l e v e l i n d i c a t i n g that the r e s u l t s support the v a l i d i t y of the Sequential/Simultaneous dichotomy f o r t h i s group. Although the Hand Movements subt e s t d i d not f i t the s p e c i f i e d model f o r the P u n j a b i , t h i s d i d not r e s u l t i n a r e j e c t i o n (or approach one) of the data-model f i t . The X 2 was not s i g n i f i c a n t at the .05 l e v e l , hence, support f o r the v a l i d i t y of the Sequential/Simultaneous dichotomy was p r o v i d e d . N e v e r t h e l e s s , the f a i l u r e of the Hand Movements sub t e s t to load on i t s s p e c i f i e d f a c t o r f o r the Cantonese and Punjabi i n d i c a t e s the need f o r f u r t h e r i n v e s t i g a t i o n s of t h i s s u b t e s t ' s p r o p e r t i e s . Mental P r o c e s s i n g and Achievement Subtests. Shown i n Table 23 are the f a c t o r l o a d i n g s and Goodness of F i t s t a t i s t i c s f o r the c o n f i r m a t o r y a n a l y s i s performed on the 13 K-ABC s u b t e s t s . L36 Table 23 Confirmatory Factor Analysis 3 of Mental Processing and Achievement subtests  for each group Language Group*3 Cantonese English Punjabi Factors c :  Subtests SEQ SIM ACH SEQ SIM ACH SEQ SIM ACH Sequential Hand Movements 30 35 23 Number Recall 69 61 54 WOrd Order 70 70 71 Simultaneous Gestalt Closure 26 40 49 Triangles 74 67 60 Matrix Analogies 60 61 66 Spatial Memory 24 48 54 Photo Series 47 65 45 Achievement Faces and Places 72 84 42 Arithmetic 64 66 65 Riddles 73 58 61 Reading/Decoding 75 71 52 Reading/ Understanding 83 86 81 Goodness of Fit Statistics Chi square 107.79 61.92 84.25 df 62 62 62 probability < .001 > .05 < .05 Notel; Decimals have been emitted. Note"*: Only loadings for subtests on targeted factor are produced. a Maximum likelihood estimation. n = 70 for each group, c SEQ - Sequential, SIM = Simultaneous, ACH = Achievement. 137 As was observed for the two factor (eight subtest) solutions, the data for the Cantonese did not f i t the specified model (£ < .05). Examination of the factor loadings reported i n Table 23 reveals that the Hand Movements, Gestalt Closure, and Spatial Memory subtests again did not meet salience on t h e i r hypothesized factors for t h i s group. In the case of the English group, the hypothesized three factor model was confirmed. Each subtest met salience on i t s speci f i e d , hypothesized factor. Keith (1985) reported that for the 5 year olds, 7 year olds and 10 year olds (n = 200 in each group) i n the standardization sample, the factor loadings i n the confirmatory factor analyses supported the v a l i d i t y of the K-ABC. He added, however, that Hand Movements loaded s i g n i f i c a n t l y on both factors. For the Punjabi, the hypothesized model was rejected (p < •05). Examination of the factor loadings revealed that Hand Movements was nonsalient on the Sequential factor. In summary, the confirmatory factor analyses revealed that the hypothesized two factor model was confirmed by the English and Punjabi data. In contrast, the Cantonese data did not f i t th i s model. Moreover, the three factor model was confirmed by the English data but not by the Cantonese and Punjabi data. This suggests that there i s support for the v a l i d i t y of Sequential/Simultaneous dichotomy for the English and Punjabi data but not for the Cantonese data. The rejec t i o n of the 3 138 factor model for the Punjabi (2 factor model was not rejected) may indicate that the ACH subtests are highly correlated with the Sequential and/or Simultaneous factors. This was investigated using exploratory analysis. Exploratory Factor Analysis To c l a r i f y the nature of the factor structure within the Cantonese and Punjabi samples, exploratory factor analytic procedures were employed. Further, although the confirmatory factor analysis evidenced a good data-model f i t for the English sample, Keith (1985) (for the ages similar to those researched i n t h i s study) found that Hand Movements did not load on the hypothesized Sequential factor when an exploratory analysis was performed. For comparative purposes, exploratory factor analysis was performed on a l l three groups. The exploratory procedures were applied separately to each sample. F i r s t , the number of factors to retain for the analysis was determined using three r u l e s : 1) Kaiser-Guttman unity root c r i t e r i o n (Guttman, 1954; Kaiser, 1970); 2) Scree test ( C a t t e l l , 1966); and 3) s t a t i s t i c a l l i k e l i h o o d or maximum li k e l i h o o d method (Lawley & Maxwell, 1963). In the Kaiser-Guttman c r i t e r i o n , eigenvalues of 1.0 or greater are suggestive of the number of factors to r e t a i n . For the Scree t e s t the eigenvalues are plotted i n descending order. In the 139 r e s u l t i n g graph, the plotted eigenvalues are connected by li n e s which represent a scree. More than two points must be connected by a scree. I t i s possible for one graph to have more than one scree. The plots not connected represent the number of factors to be retained. In the maximum l i k e l i h o o d method the n u l l hypothesis tested i s that no more than k common factors are required to f i t the data. I f the n u l l hypothesis i s rejected (Chi square), the conclusion i s that some number of factors greater than k i s required. Secondly, the data were subjected to a p r i n c i p a l components analysis followed by an orthogonal (Varimax) rotation (Kaufman & Kamphaus, 1984). The most interpretable or c l i n i c a l l y meaningful solution was retained. F i n a l l y , an unweighted least squares analysis with an orthogonal rotation (Varimax) was performed on the number of factors retained i n the previous analysis (Kaufman & Kamphaus, 1984; and Keith, 1985). Factor loadings of .35 or greater were considered s a l i e n t (Kaufman & Kamphaus, 1984; Kaufman & Kaufman, 1983b; Keith, 1985). The factor solutions produced were inspected for i n t e r p r e t a b i l i t y and c l i n i c a l meaningfulness (a factor with loadings above .350 and not producing a singleton). Mental Procesing subtests. As shown i n Table 24, the number of factors i d e n t i f i e d by each of the three rules was two for both the English and Punjabi samples. In the case of the 140 Table 24 Number of factors i d e n t i f i e d for Mental Processing and Mental  Processing and Achievement subtests for each group Solution/Method Language Group' Cantonese English Punjabi Mental Processing Subtests DKaiser-Guttman rule cScree t e s t S t a t i s t i c a l l i k e l i h o o d Mental Processing and Achievement Kaiser-Guttman rule Scree t e s t S t a t i s t i c a l l i k e l i h o o d 4 4 2 5 3 4 2 2 2 2 2 2 3 4 3 a n = 70 i n each group. D Eigenvalues on Scree t e s t i n Appendix J. c Scree t e s t i n Appendix J . 141 Cantonese, both the Kaiser-Guttman rule and Scree test i d e n t i f i e d four factors while maximum l i k e l i h o o d suggested two. The correponding graphs for the Scree tests are displayed i n Appendix J. To c l a r i f y the number of factors for the Cantonese sample, the two, three and four factors were orthogonally rotated following p r i n c i p a l components extraction. Of the solutions, the two and three factors proved the most interpretable. As shown in Table K - l , Appendix K, the two factor solution had a l l the memory subtests on one factor and the reasoning subtests on another. The three factor solution appeared to have the auditory memory/sequencing subtests on one factor, the v i s u a l / s p a t i a l tasks on another, and the v i s u a l sequencing tasks on yet another factor. Spatial Memory may not "look" l i k e a sequencing task, but many of the testers commented that the Cantonese appeared to remember the pictures i n a set sequence usually moving from l e f t to r i g h t . Although the three factor solution for the Cantonese does not appear to be as well defined as the two factor solution, i t i s worthy of discussion. Kaufman and Kamphaus (1984) reported that for the 7, 8, 9, 10 and 12 year olds i n the standardiza-t i o n sample, Hand Movements had a high loading on the t h i r d factor and was inconsistently joined by Spatial Memory, Photo Series, or Matrix Analogies. "Thus, the extra factors i n the three-factor solutions seems inconsistent, t r i v i a l , and of 142 l i t t l e apparent c l i n i c a l meaning" (Kaufman & Kamphaus, 1984, p. 632). Keith (1985) also expressed concern about these " t r i v i a l " subtests when he observed for the 10 year olds i n his study the 2 factor model did not adequately account for the c o r r e l a t i o n between Hand Movements and Matrix Analogies, Spatial Memory and Photo Series. Given the concern expressed by Keith (1985) and the results i n t h i s study for the Cantonese, i t appears that the t h i r d factor loadings should not be dismissed as t r i v i a l . In f act these loadings may signal an underlying factor s p e c i f i c to the cognitive style of a s p e c i f i c c u l t u r a l group. Given a l l of the c u l t u r a l groups i n the standardization sample were pooled for factor analysis, the r e s u l t i n g factor patterns may have been d i f f e r e n t i f each c u l t u r a l group were analyzed separately. The four factor solution was not well defined or meaning-f u l . The two factor patterns r e s u l t i n g from the unweighted least squares extraction followed by an orthogonal rotation are shown i n Table 25 for each of the three groups. For the Cantonese, four subtests ( v i z . , Hand Movements, Number Recall, Word Order, Spatial Memory) loaded on the hypothesized Sequential factor. These four subtests have a short-term memory component. In addition, three subtests (vi z . , Triangles, Matrix Analogies, Photo Series) loaded on the hypothesized Simultaneous factor and a l l have a reasoning component. Gestalt 143 Table 25 Factor loadings for the two factor, unweighted least squares analysis with a varimax  rotation for the Mental Processing subtests for each group Factors Sequential Simultaneous Language Group 5 0 Language Group 3 0 Subtests C E' P C E P Hand Movements 38 21 15 04 40 29 Number Recall 89 99 65 -02 01 -08 Word Order 48 46 65 06 20 11 Gestalt Glosure 03 07 -12 24 38 54 Triangles 08 10 17 87 67 61 Matrix Analogies 08 22 09 51 56 68 Spatial Memory 35 06 34 21 50 45 Photo Series 06 04 09 54 66 47 Variances 1.30 1.32 1.05 1.30 1.80 1.65 Note: Decimals have been omitted. Factor loadings > .350 are underlined. a n = 70 in each group. b C = Cantonese, E = English, P = Punjabi. 144 Closure did not meet the .35 salience c r i t e r i o n . Not a l l the K-ABC subtests loaded on t h e i r hypothesized factors. Hence, as was evidenced i n the confirmatory analysis, the Sequential/ Simultaneous dichotomy i s not c l e a r l y i d e n t i f i e d by the Cantonese K-ABC data. A "memory/reasoning" dichotomy may be the more accurate description of the factor dichotomy for the eight MPC subtests. As previously mentioned the three factor solution for the Cantonese appears to have some c l i n i c a l merit. The unweighted least squares solution i s displayed i n Appendix K-2. This solution d i f f e r e d from the P r i n c i p a l Components solution, previously discussed, i n that Photo Series did not achieve salience on the t h i r d factor. As such, an auditory memory and sequencing/visual-spatial/visual memory t r i a d may be appropriate labels for the r e s u l t i n g factors. The factor pattern for the English data evidenced the sub-test composition of the SEQ and SIM Scales with one exception, Hand Movements, which had a loading of .40 on the hypothesized Simultaneous factor. For the standardization sample Kaufman and Kamphaus (1984); Kaufman and Kaufman (1983b); and Keith (1985) found the Hand Movements subtest loaded substantially, for some age groups, on either both factors or on the Simultaneous factor. An investigation of the exploratory analyses done by Kaufman and Kamphaus (1984) across a l l ages for the standardization sample revealed a developmental trend on t h i s subtest. For ages 2 1/2 - 4, th i s subtest was strongly associated with the Sequential factor (mean loading of .60 vs. .19 on Simultaneous). At age 5 a sudden s h i f t occurred, and this subtest became about equally dependent on both mental processes for ages 5 - 12 1/2 (mean loading of .37 and .43 on Sequential and Simultaneous dimensions, r e s p e c t i v e l y ) . (Kaufman & Kamphaus, 1983b, p. 699) The factor pattern for the Punjabi was similar to the English except that Hand Movements did not meet the salience c r i t e r i o n (.35) on any factor. However, i t did load higher (.29) on the Simultaneous factor than the Sequential factor ( .15) . Mental Processing and Achievement Subtests. The same three tests (Kaiser-Guttman unity root c r i t e r i o n , Scree test, s t a t i s t i c a l likelihood) used to determine the number of factors to r e t a i n for the 8 MPC subtests were employed for the 13 MPC and ACH subtests. The outcomes of these three methods are summarized i n Table 24. The corresponding Scree test graphs are displayed i n Appendix I. As shown i n Table 24, the English data were i d e n t i f i e d by a l l three methods as having three factors. For the Punjabi three factors were i d e n t i f i e d by the Kaiser-Guttman rule and s t a t i s t i c a l l i k e l i h o o d test, but four factors emerged on the Scree t e s t . The factors i d e n t i f i e d by the three methods were even more discrepant for the Cantonese with three, four and 146 fiv e factors emerging. Because of the inconsistencies i n the outcome of these methods for the Cantonese and Punjabi, a p r i n c i p a l components analysis with an orthogonal rotation was conducted to determine the i n t e r p r e t a b i l i t y of the three, four and f i v e factor solutions for these two groups. A .35 salience loading was accepted (Kaufman & Kamphaus, 1984) and interpret-a b i l i t y was determined through the c l i n i c a l meaningfulness of factor patterns. For the Punjabi the three factor solution was the most interpretable. Their four factor solution was not well defined (double loadings) or e a s i l y interpretable. The f i f t h factor for each group was not s u f f i c i e n t l y defined. Therefore, three factors for the Punjabi was retained for further analysis. The four factor solution for the p r i n c i p a l components analysis for the Cantonese i s shown i n Appendix K, Table K-3. This solution was not as e a s i l y interpretable as the three factor solution for the subtests. Hand Movements, for example, loaded on a factor with Triangles, Arithmetic, and Reading/ Understanding. I t appears that the three factor solution for the 8 subtests does not r e t a i n i t s i d e n t i t y when the Achievement subtests are included i n the factor analysis. As shown i n Table 26, three factors were retained. For the Cantonese the four "short-term memory" subtests grouped together on one factor and the three subtests (Triangles, Matrix Analogies, Photo Series), purported to measure reasoning 147 Table 26 Factor loadings for the three factor, unweighted least squares analysis with a varimax  rotation for the Mental Processing and Achievement subtests for each group Sequential Factors  Simultaneous Language Group3*5 Achievement Subtests C E P C E P C E P Sequential Hand Movements 48 17 22 06 38 35 04 16 03 Number Recall 63 99 58 -07 -04 -04 19 08 -04 Word order 44 41 56 -15 14 11 39 35 08 Simultaneous Gestalt Closure -02 00 -22 29 34 48 -02 23 21 Triangles 06 06 20 66 70 54 15 08 09 Matrix Analogies 06 19 -02 39 54 65 27 18 18 Spatial Memory 56 02 30 34 47 50 -12 19 02 Photo Series 08 00 -11 47 65 47 24 14 07 Achievement Faces and Places -12 10 03 28 31 09 74 77 70 Arithmetic 33 -01 34 19 48 53 54 55 14 Riddles 09 09 07 24 03 43 65 64 60 Reading/Decoding 22 15 51 -13 27 05 85 63 46 Reading/ 07 18 47 20 33 50 76 73 35 Understanding Variances 1.33 1.30 1.49 1.27 2.23 2.29 2.93 2.57 l . : Note; Decimals have been emitted. Factor loadings > .350 are underlined. a n = 70 in each group. b C = Cantonese, E = English, P = Punjabi. 148 a b i l i t y (Kaufman & Kaufman, 1983b), formed a second factor. The f i v e ACH subtests formed one factor with loadings above .54. One additional subtest, Word Order, also had a s a l i e n t loading (.39) on the Achievement factor. The four factor solution i s referenced i n Appendix K, Table K-3. Although i t does not appear as easy to interpret as the three factor solution, Hand Movements did load on the same factor as Spatial Memory and Photo Series. This again suggests the presence of a v i s u a l sequencing factor. The English factor pattern resembled the solution reported by Kaufman and Kamphaus (1984). Even Hand Movements loaded more substantially (.38) on the hypothesized Simultaneous factor. For the English group, as for the Cantonese, the Gestalt Closure subtest f a i l e d to reach the salience c r i t e r i o n of .35 on any of the three factors. This was not the case for the children i n the standardization sample. This may be a r e s u l t of the smaller samples i n the present study (making results less stable), lack of homogeneity in performance among the children within the Cantonese and English groups, and/or th i s subtest's lower r e l i a b i l i t y for these two groups. Further investigation of the pattern loadings for the English reveals that the A r i t h -metic subtest loaded sub s t a n t i a l l y on both the Simultaneous factor (.48) and the Achievement factor (.55). Since achieve-ment i s dependent upon processing a b i l i t y , double loadings were expected. This may indicate that the children are employing 149 both processing modes to complete this task. Hand Movements achieved salience on the Simultaneous factor for the Punjabi (and English). Four of the f i v e Achieve-ment subtests loaded above .43 on either or both the Sequential and Simultaneous factor for the Punjabi. Also for t h i s group the Arithmetic had a low loading (.14) on the Achievement factor. Kaufman and Kamphaus (1984) factor analyzed the ACH subtests with the SEQ and SIM subtests to determine whether Sequential and Simultanous factors were retained. They found, as did Keith (1985), that for the standardization data three factors emerged and some of the ACH subtests loaded on one or both of the Mental Processing factors. Kaufman and Kaufman (1983b) believe on the one hand that mental processing i s an important variable for school learning; on the other hand i t can be distinguished from achievement (acquired school learning). The substantial loadings of a l l the ACH subtests (except Faces and Places) for the Punjabi and Arithmetic for the English on the mental processing factors, suggests that problem-solving and acquired knowledge, as measured by the K-ABC, may not be independent constructs. Furthermore, the substantial loadings of four of the f i v e ACH subtests for the Punjabi on the mental processing factors might be why, when conducting the confirmatory factor analysis, the three-factor model was not confirmed by t h e i r data. This may be the r e s u l t 150 of the ACH subtests correlating highly with the MPC subtests. Given only one of the English group's subtests loaded on the Simultaneous factor, this was not enough to reject the data-model f i t . Not only are some of the loadings lower than one might expect, the results of the exploratory analysis provides support for the v a l i d i t y of the K-ABC Sequential/Simultaneous dichotomy for the English and Punjabi with the exception of the Hand Movements subtest. Similar findings for Hand Movements by other researchers has resulted in proposing labels, such as, verbal/nonverbal (Das, 1984b) and verbal-memory/nonverbal reasoning (Keith & Dunbar, 1984) to interpret the two factors. The c l i n i c a l interpretation of the Cantonese dichotomy may suggest that a short-term memory/reasoning dichotomy is a more appropriate l a b e l . However, i f the three factor solution is interpreted, auditory sequential memory, visua l sequencing and v i s u a l s p a t i a l i z a t i o n may apply as appropriate labels to this solution. The rejection of the data-model f i t for the Cantonese when confirmatory factor analysis was performed lends support for an alternate model to explain the factor pattern for this group. This also brings into question the interpretation of the SEQ and SIM Scale Scores for the Cantonese. The inclusion of the f i v e ACH subtests does not a l t e r s i g n i f i c a n t l y the factor pattern of the MPC subtests. This suggests that the factor pattern for these Mental Processing subtests for a l l three groups i s f a i r l y stable. 151 CHAPTER VII Relationship Between the K-ABC and WISC-R Presented i n thi s chapter are the outcomes of the analyses investigating the relationship between the K-ABC and WISC-R. Since the WISC-R was selected as the c r i t e r i o n measure of i n t e l l i g e n c e , with which the K-ABC was compared, the WISC-R's psychometric properties (central tendency, v a r i a b i l i t y , and r e l i a b i l i t y ) are discussed f i r s t . WISC-R: Psychometric Properties The means, standard deviations, and in t e r n a l consist-ency estimates for the WISC-R subtest and scales are report-ed i n Table 27. As was the case for the K-ABC, each group was analyzed separately using the SPSS X computer program (Nie, 1983). Central Tendency and V a r i a b i l i t y Multivariate analysis of variance, using Wilks' c r i t e r -ion, of the 12 subtest scaled scores revealed that there was a s i g n i f i c a n t difference (F (24,392) = 7.28, £ < .001) among the mean vectors of the three groups. One-way analyses of variance i d e n t i f i e d s i g n i f i c a n t differences (£ <_ .001) among the groups on 10 of the 12 subtests (see Appendix L 152 Table 27 WISC-R means, standard deviations and internal consistency reliabilities for each  9 r o uP Language Group3 Subtests/Scales Cantonese ic SD r*x English x SD rxx Punjabi x SD rxx TukeyD Compar-sons Verbal Information 10 .04 2. 57 .58 10. 44 2. 65 .64 7.57 2. 14 .25 CE>P Similarities 10 .96 2. 95 .72 11. 74 2. 58 .56 8.36 3. 28 .82 CE>P Arithmetic 11 .17 2. 00 .68 10. 59 2. 29 .59 10.49 2. 78 .73 NS Vocabulary 10 .14 2. 77 .71 11 91 2. 91 .60 8.67 2. 24 .58 E>C>P Comprehension 9 .14 2. 82 • 61 9. 81 2. 74 .65 7.49 1. 86 .39 CE>P Digit Span 9 .56 2. 07 9. 86 2. 70 9.40 2. 33 NS Performance Picture Completion 11 .14 2. 74 .43 11 46 2. 52 .64 9.17 2. 60 .71 CE>P Picture Arrangement 12 .39 2. 89 .63 12. 07 2. 43 .28 10.63 2. 94 .63 CE>P Block Design 13 .70 2. 83 .79 12. 36 3. 34 .84 9.40 3-06 .73 C>E>P Object Assembly 12 .33 2. 91 .36 11. 63 2. 90 .64 9.40 2. 58 .41 CE>P Coding 12 .24 2. 89 10. 20 2. 95 10.60 2. 56 C>EP Mazes 12 .83 2. 94 .64 12. 04 2. 80 .75 10.31 3. 21 .61 CE>P Verbal IQ 101 .41 11. 49 .87 105. 51 11. 78 .85 90.19 10 .88 .84 CE>P Performance IQ 116 .60 13. 12 .79 110. 69 12. 33 .80 98.74 12 .06 .81 CE>P Full Scale IQ 109 .23 11. 15 .87 108. 73 11. 31 .87 93.80 10 .34 .87 CE>P a n = 70 for each group. D C = Cantonese. E = English. P = Punjabi. 153 for F ratios and F p r o b a b i l i t i e s ) . The Bonferroni method was employed to control for the effects of a Type 1 error rate on the subtests. S i g n i f i c a n t differences found among the groups by the one-way analysis of variance remained s i g n i f i c a n t after c o n t r o l l i n g for the experimentwise alpha. As shown i n Table 27, Tukey's test of s i g n i f i c a n t difference between pairs of means revealed that the groups did not d i f f e r s i g n i f i c a n t l y on Arithmetic or D i g i t Span. On the ten remaining subtests, the Cantonese had means s i g n i f i c a n t l y greater than the Punjabi. Moreover, the English performed s i g n i f i c a n t l y higher than the Punjabi on 9 of the 10 s i g n i f i c a n t subtests (Coding was the exception). F i n a l l y , the English outperformed the Cantonese on Vocabulary while the Cantonese outperformed the English on Block Design and Coding. Multivariate analyses of variance (Wilks' c r i t e r i o n ) was performed on the two noncomposite WISC-R scales ( v i z . , Verbal IQ, Performance IQ) and i t revealed a s i g n i f i c a n t difference (F (4, 412) = 26.95, £ < .0001) among the mean vectors of the three groups. Moreover, one-way analysis of variance evidenced a s i g n i f i c a n t difference (£ _< .001) among the groups on the Verbal, Performance and F u l l Scale IQs (see Appendix L for F ratios and F p r o b a b i l i t i e s ) . These scales remained s i g n i f i c a n t a f t e r c o n t r o l l i n g for the experimentwise alpha (Bonferroni method). A Tukey range test (£ < .05) i d e n t i f i e d the Cantonese and English as performing 154 s i g n i f i c a n t l y higher than the Punjabi on a l l three scales. On the Verbal IQ (VIQ) the mean Punjabi score (90.19) was 10 and 16 points lower than the mean VIQ for the Cantonese (101.41) and English (105.51), respectively. On the Performance IQ (PIQ), the Punjabi (x=98.74) were again performed s i g n i f i c a n t l y higher than the Cantonese (x=116.60) (18 points) and English (x 110.60) (12 points). These differences between the Punjabi and the other two groups on the VIQ and the PIQ resulted i n the Punjabi scoring over one standard deviation lower on the F u l l Scale IQ (FSIQ) than the Cantonese and the English. A test for homogeneity of variance covariance matrices through SPSS X (Nie, 1983) produced F (156, 114,173) = 1.28, p_ < .01 for Box's M, showing a s t a t i s t i c a l l y s i g n i f i c a n t deviation from homogeneity of covariance matrices. One-way analysis of variance was performed to determine the corresponding homogeneity of variances for each subtest. Of the 12 subtests Arithmetic and Comprehension had variances that s t a t i s t i c a l l y deviated (£ <^  .05) from homogeneity. The groups were therefore not pooled for the following analyses. Internal Consistency WISC-R s p l i t - h a l f c o e f f i c i e n t s corrected for length by the Spearman-Brown formula (Ferguson, 1981, p. 438) were computed. This procedure for estimating r e l i a b i l i t y "was not appropriate for Coding, because i t i s a speeded test, or D i g i t Span, because i t i s given as two separate subtests" (Wechsler, 1974, p. 27). As shown in Table 27, the c o e f f i c -ients ranged from a low of .25 (Information-Punjabi) to a high of .84 (Block Design-English). Even among the three groups for the same subtest there was immense v a r i a b i l i t y . As a case i n point, on the Picture Arrangement subtest the English had a c o e f f i c i e n t of .28 whereas the Cantonese and Punjabi had c o e f f i c i e n t s of .63. On t h i s subtest, many of the English performed better on the two odd numbered subtests than the two even numbered subtests. The reason for them performing less well on the second item than the t h i r d item was not r e a d i l y apparent. Their age would suggest that they might have some problems with the l a s t item, but t h i s i s true for the other two groups. Moreover, on other subtests where low c o e f f i c i e n t s were found, either the same was true or the subjects reached the c e i l i n g quickly. Compared with the 200, 8 l/2 year olds i n the WISC-R standardization sample where r e l i a b i l i t y c o e f f i c i e n t s ranged from a low of .66 (Object Assembly) to a high of .86 (Vocabulary), generally the c o e f f i c i e n t s reported for each group i n the present study are lower. The more homogenous a b i l i t y range within each group may have contributed to the lower r e l i a b i l i t y c o e f f i c i e n t s found i n the present study. Nevertheless, the low r e l i a b i l i t y estimates for some of the 156 subtests suggests caution needs to be extended i n making interpretations at the subtest l e v e l . Guilford's formula (Guilford, 1954, p. 393) was employed to compute the composite r e l i a b i l i t i e s of the VIQ, PIQ and FSIQ. Because a s p l i t - h a l f r e l i a b i l i t y could not be computed for the Coding subtest, the PIQ composite r e l i a b i l -i t y was comprised of four subtests. Only nine subtests were included i n the computation of the FSIQ r e l i a b i l i t y . The supplementary subtests ( i . e . , D i g i t Span, Mazes) are not included i n computing the IQs, therefore, they were not included i n the inter n a l consistency estimates. As shown i n Table 27, the r e l i a b i l i t y c o e f f i c i e n t s for the WISC-R scales ranged from a low of .79 to a high of .87. For each group the FSIQ had a r e l i a b i l i t y c o e f f i c i e n t of .87. This was lower than the .95 and .96 reported i n the WISC-R Manual for 8 1/2 and 9 1/2 year olds, respectively, i n the standardization sample. Possible reasons for the lower r e l i a b i l i t y estimates i n the present study may be the more homogeneous a b i l i t y range within the three groups. K-ABC versus WISC-R The SPSS X computer program (Nie, 1983) was used to obtain Pearson co r r e l a t i o n c o e f f i c i e n t s between K-ABC and WISC-R subtests and scales. Dependent t - t e s t comparisons between K-ABC and WISC-R scales were also obtained. The 157 subtest analyses w i l l preceed the discussion of the scale analyses. Subtests Pearson correlations were computed to compare the K-ABC subtests with the WISC-R subtests. Shown i n Tables 28, 29, and 30 are these c o r r e l a t i o n c o e f f i c i e n t s for the Cantonese, English and Punjabi, respectively. A l l three SEQ subtests for each group correlated s i g n i f i c a n t l y with D i g i t Span (£ <_ .05). For the English, Word order correlated more strongly with vocabulary than D i g i t Span. As one would expect, D i g i t Span, which requires the examinee to repeat a series of st i m u l i given aurally, converged with the SEQ subtests, which also have the stimulus presented sequentially. Kaufman (1979) reported that Picture Arrangement, Coding, and Mazes are the three PIQ subtests that q u a l i f y as successive (sequential) tasks. However, for the Cantonese and English these' three WISC-R subtests were not s i g n i f i c -antly correlated with any of the SEQ subtests. For the Pun-ja b i , only one comparison (Hand Movements-Coding) was found to correlate (.29) s i g n i f i c a n t l y (£ < .01). It appears that the three WISC-R "successive" tasks are not measuring se-quential processing the same way as the three SEQ subtests. Picture Arrangement did not correlate s i g n i f i c a n t l y with Photo Series. Given that both subtests involve placing Table 28 Correlations between the K-ABC and WISC-R subtests scaled scares for Cantonese subjects WISO-R Subtests K-ABC inform- similar- Arith- Vocab- ctxnpre- Digit Picture Picture Block Object Qxiing Mazes ation ities metic ulary hension Span completion Arrangement Design Assembly Sequential ffindnovsiients 04 03 21* 02 05 22* 01 01 09 03 00 03 Number Recall 09 24* 22* 21* 21* 53** 08 - 03 01 - 01 10 01 Word order 28** 08 22* 12 - 01 44** - 03 - 13 04 - 08 15 - 19 Simul taneous Gsstalt Closure 04 08 - 10 05 21* 05 35** 25* 17 12 25* 14 Triangles 21* 10 03 10 20 24* 17 25* 62** 25* 13 21* Matrix Analogies 29** 14 - 06 14 25* 13 21* 24* 38* 20 - 01 08 Spatial Memory 02 07 22* 12 - 11 14 05 29** 29** 15 29** 17 Photo Series 23* 33** 15 - 01 00 03 15 01 31** 23* 08 20* Achievement faces & places 57** 45** 25* 49** 32** 10 25* 24* 28** 19 07 20* Arithmetic 57** 48** 44** 36** 37** 25* 18 13 21* 15 21* - 00 Riddles 57** 50** 25* 63** 51** 21* 23* 19 20* 20 07 06 Reading/Decoding 45** 41** 38** 40** 35** 29** 09 02 05 02 - 01 - 08 Reading/ Understanding 48** 45** 31** 54** 44** 20* 20* 31** 26* 18 07 - 01 Note; Because of rounding seme coefficients of .20 (£ < .05) and .27 (p_ < .01) are note identified as significant. * £ < .05, ** £ < .01. Table 29 Oorrelaticns between the K-ABC and WISC-R subtests scaled scores f o r English subjects WISOR Subtests K-ABC Inform- similar- A r i t h - Vocab- Ctnpre- D i g i t Picture picture Block Cfoject Ooding Mazes atian i t i e s metic ulary hension Span Ctmpletion Arrangement Design Assembly Sequential Handmcvements .12 .30** .32** .27* .21* 32** 12 .01 .17 .17 --.09 .14 Nimber Recall .14 .14 .21* 19 03 61** - 02 - 10 - 06 02 11 05 Word Order 37** 27* 25* 45** 18 38** 14 09 12 16 02 17 Simultaneous Gsstalt Closure 20 36** 31** 30** 21* 18 33** 12 28** 34** 13 01 Triangles 14 26* 25* 17 02 18 27** 11 63** 38** 02 28** Matrix Analogies 05 31** 37** 34** 32** 27* 29** - 04 40** 23* 02 34** Spatial Manory 16 22* 40** 20* 14 25* 27* 06 26* 24* 29** 18 photo Series 09 22* 39** 34** 29** 19 38** 08 39** 26* 07 43** Adhievanent Faces & places 50** 53** 40** 63** 40** 33** 11 - 03 25* 30** 13 21* Arithmetic 42** 43** 46** 53** 49** 20* 19 13 47** 38** 22* 32** Riddles 64** 46** 36** 63** 49** 16 06 - 01 26* 22* 13 28** Reading/Decoding 51** 46** 37** 53** 30** 14 22* - 08 19 30** 17 13 Reading/ Understanding 46** 41** 48** 57** 34** 38** 13 - 08 28** 28** 20 21* Note; Because of rounding sane coefficients of .20 (p_ < .05) and .27 (ja < .01) do not appear as s i g n i f i c a n t . * £ < .05, ** £ < .01. Table 30 Correlations between the K-ABC and WISC-R subtests scaled scores f o r Punjabi subjects WISC-R Subtests K-ABC Inform- Similar- A r i t h - Vocab- Ccrapre- D i g i t Picture picture Block Object Coding Masses ation i t i e s metic ulary hension Span Carpletion Arrangement Design Assembly Sequential BBradmovements 35* 10 24* 14 08 Number Recall 20* 02 15 05 - 10 Word Order 21* 06 23* 24* 16 Simultaneous Qastalt Closure 14 05 12 14 27** Triangles 26* 16 07 07 10 Matrix Analogies 26* 20* 25* 27* 30** Spatial Manory 19 10 28** 16 07 photo series 26* 24* 36** 20* 16 Achievement Faces & places 50** 29** 16 33** 21* Arithmetic 42** 31** 54** 25* 33** Riddles 50** 56** 34** 58** 46** Reading/Decoding 35** 27* 15 22* 01 Reading/ Understanding 46** 26* 48** 37** 42** 38** 11 23* 20* 05 29** - 02 54** 14 - 08 05 - 10 16 - 12 43** 13 03 26* - 03 05 02 01 29** 36** 44** 30** 02 28** 41** 44** - 00 50** 31** 09 33** 20* 34** 26* 46** 22* 20* 22* 30** 26* 37** 42** 27* 27* 17 11 31** 19 40** 30** 14 20* 18 20* 36** 22* - 03 - 01 - 05 31** 28** 21* 36** 09 34** - 04 20* 36** 35** 31** 07 13 02 39** 26* - 03 13 - 28** - 00 - 22* 45** 27* 25* 33** 06 40** 09 Note; Because of rounding seme coefficients of .20 (£ < .05) and .27 (£ < .01) do not appear as s i g n i f i c a n t . * £ < .05, ** £ < .01. 161 pictures i n a sequence, i t seems l o g i c a l that Photo Series should be a sequential task. Incidentally, Kaufman and Kaufman (1983b) also thought Photo Series was a sequential task, for "Photo Series was placed on the Simultaneous rather than the Sequential Processing Scale based on compel-l i n g factor analytic data" (p. 63). Kaufman and Kamphaus (1984) elaborated, saying Photo Series "seems to be solved best by good h o l i s t i c processors who can organize a large array of v i s u a l - s p a t i a l s t i m u l i i n t h e i r minds and maintain th i s simultaneous integration of the entire sequence, while responding vi a a sequential format" (p. 628). Unlike Photo Series, Picture Arrangement involves the understanding of the sequential nature of a story l i n e . Photo Series just involves putting pictures i n a chronological order. Of the SIM subtests, Triangles had the highest c o r r e l a -t i o n with Block Design (Cantonese, .62; English, .63; Punjabi, .50) for a l l three groups. Both subtests require the examinee to manipulate either blocks (Block Design) or triangles (Triangles) to reproduce a stimulus design. Block Design was also i d e n t i f i e d by Kaufman (1979) as one of the three WISC-R PIQ subtests that q u a l i f y as a simultaneous task. The other two were Picture Completion and Object Assembly. For the English and Punjabi these two subtests correlated s i g n i f i c a n t l y (£ < .05) with the SIM subtests. However, Picture Completion and Object Assembly for the Cantonese were not as highly correlated with the SIM 162 subtests. Only Gestalt Closure correlated s i g n i f i c a n t l y (£ < .01) with Picture Completion. This may suggest that the Cantonese are not performing the same on the K-ABC SIM Scale as the other two groups. The results of the factor analyses also suggested t h i s . There i s some evidence from t h i s anal-yses that some of the SIM subtests for the Cantonese may be measuring memory, reasoning, and/or v i s u a l sequencing. Research w i l l need to be conducted to investigate these alternative interpretations. The f i v e ACH subtests for the Cantonese and English group correlated s i g n i f i c a n t l y (£ _< .05) with the f i v e mand-atory VIQ subtests (Digit Span i s supplementary). For the Punjabi the majority of the ACH-VIQ subtest comparisons were s i g n i f i c a n t l y correlated (£ < .05), however, Reading Decod-ing had a co r r e l a t i o n of only .01 with Comprehension. Children do not need to have good judgment or a good receptive vocabulary to decode words. In fact, the Punjabi did not d i f f e r from the other two groups i n the i r a b i l i t y to decode words, however, they did perform s i g n i f i c a n t l y lower than the Cantonese and English on Comprehension. The s i g n i f i c a n t correlations among the ACH subtests and the VIQ subtests are commensurate with the evidence present-ed i n the IM showing the ACH subtests as having a moderate to strong c o r r e l a t i o n with the VIQ Scale. Given that the K-ABC ACH Scale and the WISC-R Verbal Scale emphasize verbal conceptualization and acquired learning, t h e i r convergence 163 was expected. However, i t does indicate that what Wechsler (1974) refers to as verbal i n t e l l i g e n c e and Kaufman and Kaufman (1983b) refer to as achievement are not independent constructs. Many of the PIQ subtests also correlated s i g n i f i c a n t l y with the ACH subtests. This i s further evidence that even Vis u a l - S p a t i a l a b i l i t y i s not independent of academic achievement and verbal learning. Scales For ease of comparison the means and standard devia-tions for the four K-ABC scales and the three WISC-R scales have been reproduced i n Table 31. Mental Processing Composite versus F u l l Scale IQ. As shown i n Table 31, the MPC-FSIQ discrepancy did not exceed four points for any of the three groups. This was similar to the three point difference found between the same scales for the 182 normal children reported i n the IM (p. 113). For the Punjabi, the K-ABC MPC was 1.63 points higher than the WISC-R FSIQ. For the normal children i n the IM and the Cantonese and English i n the present study, the discrepancy favored the WISC-R FSIQ. This i s consistent with normative trends i n that children score lower on more recently developed and standardized tests than on existing measures 164 Table 31 Means and standard deviations for the K-ABC Global Scales and WISC-R IQs by  language group Language Group3 Cantonese x SD English x SD _ Punjabi x SD K-ABC Mental processing Sequential Simultaneous Achievement 106.07 (9.95) 98.81 (12.56) 109.54 (10-26) 98.14 (9.72) 104.73 (10.98) 102.76 (10.90) 105.54 (11.78) 101.73 (10.59) 95.43 (9.50) 97.33 (10.75) 95.41 (10.68) 91.91 (7.08) WISC-R Full Scale IQ Verbal IQ performance IQ 109.23 (11.15) 101.41 (11.49) 116.60 (13.12) 108.73 (11.31) 105.51 (11.78) 110.69 (12.33) 93.80 110.34) 90.19 (10.88) 98.74 (12.06) n = 70 in each group. 165 (Doppelt & Kaufman, 1977; Thorndike, 1977). The WISC-R was standardized approximately 10 years before the K-ABC. The significance of the MPC-FSIQ discrepancy was obtained by conducting a dependent t - t e s t comparison between these two i n t e l l i g e n c e scales for each group. As shown i n Table 32, the MPC-FSIQ discrepancy was s i g n i f i c a n t for the Cantonese (t(69) = -2.35, £ < .05) and for the English (t(69) = 3.68, £ < .001), but not for the Punjabi (t(69) = 1.59, £ < .12). When c o n t r o l l i n g for Type 1 error rate as a res u l t of multiple t comparisons, only the English discrep-ancy was s i g n i f i c a n t . Although the mean discrepancy between the MPC and FSIQ did not d i f f e r by more than four points for each group, ind i v i d u a l differences between these two i n t e l l i g e n c e scales were i n many cases much larger. Presented i n Table 33 are the number and magnitude of the discrepancies between the MPC and FSIQ for each group. The difference between the two in t e l l i g e n c e scales ranged from the MPC being 21 points higher to 32 points lower than the FSIQ. A discrepancy of 15 or more points (1 standard deviation) between these two measures was observed for 21.4% of the Cantonese, 12.8% of the English, and 11.4% of the Punjabi. This discrepancy may be related to the d i f f e r e n t d e f i n i t i o n s of i n t e l l i g e n c e these two tests were based on. Nevertheless, c l i n i c a l l y these discrepancies can have serious implications when the purpose of the assessment i s to determine the most 166 Table 32 Dependent t-test comparisons between K-ABC Standard Scores and WISC-R IQs Language Group 3 Cantonese English Punjabi Comparisons 0 t £ t p_ t p_ Mental Processing F u l l Scale IQ -2.35 ns -3.68 .001 1.59 ns Verbal IQ 3.13 .002 -0.64 ns 4.03 .001 Performance IQ -6.81 .001 -4.31 .001 -2.99 ns Achievement F u l l Scale IQ -11.13 .001 -7.27 .001 -2.00 ns Verbal IQ - 3.95 .001 -4.26 .001 1.87 ns Performance IQ -11.19 .001 -5.92 .001 -5.14 .001 Note: When controlling for Type 1 error rate due to multiple t comparisons (.05/18 = .0027), only £ values less than .0027 were significant. a n = 70 i n each group. b df = 69. 167 Table 33 Distribution of individual discxepancies between the Mental Processing  Composite and Full Scale IQ for each group Language Group5 Difference 5 Cantonese English Punjabi Scores n % n % n % + 21 - 25 2 2.9 0 0.0 0 0.0 + 16 - 20 0 0.0 0 0.0 4 5.7 + 11 - 15 2 2.9 2 2.9 6 8.6 + 6 - 10 S 11.4 12 17.1 11 15.7 + 1 - 5 15 21.4 6 8.6 17 24.3 0 4 5.7 2 2.9 4 5.7 - 1 - 5 11 15.7 16 22.8 17 24.3 - 6 - 10 10 14.3 11 15.7 4 5.7 - 11 - 15 7 10.0 12 17.1 5 7.1 - 16 - 20 3 4.3 6 8.6 1 1.4 - 21 - 25 6 8.6 2 2.9 1 1.4 - 26 - 30 1 1.4 1 1.4 0 0.0 - 31 - 35 1 1.4 0 0.0 0 0.0 5 + = Mental Processing Composite greater than Full Scale IQ, and - = Mental Processing Composite less than Full Scale IQ. 168 appropriate educational placement for a given c h i l d . For example, a discrepancy of 15 or more points could res u l t in placing a c h i l d i n a regular program (IQ 115) versus a g i f t e d program (IQ 130), depending on the test used. Further evidence to suggest these two scales (MPC and FSIQ) are not measuring the same construct to the same degree i s shown i n Table 34. Here the c o r r e l a t i o n c o e f f i c i e n t s between these two measures ranged from .44 to .67. The variance shared by the relationship of the MPC and FSIQ was 19% for the Cantonese, 45% for the English, and 40% for the Punjabi. According to Anastasi (1982) 50% shared variance (that found for 182 normals i n IM) i s high enough to support the construct v a l i d i t y of a test but low enough to suggest the new test i s not a duplication of another. Kaufman and Kaufman (1983b) also concluded that "this degree of overlap [50%3 supports the construct v a l i d i t y of the K-ABC Mental Processing Composite, while leaving enough unexplained variance to j u s t i f y the assertion of the K-ABC's unique contribution to the measurement of children's i n t e l l i g e n c e " (p. 111). In the case of the English and Punjabi, i t would appear there i s s u f f i c i e n t shared variance (close to 50%) between the two measures to warrant t h i s conclusion, hence, providing support for the v a l i d i t y of the K-ABC by Anastasi's (1982) c r i t e r i o n . However, with less than 20% shared variance between these two measures for the Cantonese, there i s less support for this conclusion. It 169 Table 34 Correlations between the K-ABC Global Scales and WISC-R IQs for each language group WISC-R Full Scale IQ Verbal IQ Performance IQ Language Group3*3 K-ABC Can. Eng. Pun. Can. Eng. Pun. Can. Eng. Pun. Mental Processing , 44** .67** .63** .33** .59** , 44** .40** .51** .65** Sequential Processing .16 .35** .29** .24* .43** .26* .02 .12 .22** Simultaneous Processing .50** .65** .63** .27* .51** .40** .55** .59** .70** Achievement .64** .73** .65** .80** .78** .71** .30** .40** .42** a n = 70 in each group. D Can = Cantonese, Eng = English, Pun = Punjabi. * £ < .05, ** £ < .01. 170 appears that for the Cantonese group the K-ABC MPC i s measuring a b i l i t i e s not assessed by the WISC-R FSIQ. Additional K-ABC and WISC-R Scale Comparisons.- The SEQ Scale correlated more strongly with the VIQ than the PIQ for the Cantonese and English (see Table 34). The Punjabi did not evidence a difference i n t h e i r SEQ-VIQ and SEQ-PIQ corre l a t i o n s . Of the K-ABC scales, SEQ had the lowest correlations with the WISC-R IQs. This may suggest that the SEQ Scale i s assessing a b i l i t i e s not measured by the WISC-R scales or that i t i s a less complex measure of i n t e l l i g e n c e than the other scales. The SIM Scale correlated more strongly with the WISC-R IQs than did the SEQ Scale (see Table 34). As expected the SIM Scale correlated more strongly with the PIQ than the VIQ with c o e f f i c i e n t s ranging from .55 (Cantonese) to .70 (Punjabi). The Cantonese c o e f f i c i e n t of .55 represents 30% shared variance between the SIM-PIQ relationship. Both scales have a v i s u a l - s p a t i a l component and some of the PIQ subtests have a simultaneous component. However, the Cantonese had a mean PIQ of 116.60 and a mean SIM Scale score of 109.54. The SIM Scale was more than 10 points lower than the PIQ - more than expected due to normative trends. This discrepancy i s also larger than was observed for the English (5.13 points), Punjabi (3.33 points) and the 182 normal subjects (1.5 points) reported on i n the IM. This substantial mean difference between these two scales for the Cantonese may be related to the two scales measuring d i f f e r -ent types of s p a t i a l a b i l i t i e s , or i t may be related to the lower c e i l i n g (Bracken, 1985; Thomas, 1984) on the SIM Scale for superior functioning children (McCallum et a l . , 1984). In the present study, the lower c e i l i n g e f f e c t resulted i n the Cantonese scoring lower on the SIM Scale than the PIQ. As an example, the Cantonese achieved t h e i r highest mean score on Triangles (SIM) and Block Design (PIQ). Of a l l the K-ABC and WISC-R subtests these two subtests correlated the highest (r=62). On Triangles the maximum attainable raw score i s 18. Five subjects achieved t h i s score, however, the maximum scaled score the 8 to 10 year old i n thi s study could achieve was 17. On the other hand, a maximum attainable raw score for Block Design was 62. The highest raw score received was 51 which was equivalent to a maximum scaled score of 19. Given that both subtests have the same mean and standard deviation (X 10, SD 3) and the same range of scaled scores (0 to 19) there was not an adequate upward extension for the superior functioning Cantonese children on Triangles. The ACH Scale had a higher co r r e l a t i o n with the FSIQ than did the MPC for the Cantonese and English. For the 172 Punjabi, the ACH-FSIQ and MPC-FSIQ correlations were nearly equivalent (see Table 34). The c o r r e l a t i o n between the ACH Scale and the FSIQ provides for shared variance ranging from 42% (Punjabi) to 53% (English). The c o e f f i c i e n t s found for the ACH-FSIQ relat i o n s h i p i n t h i s study were similar to the .76 c o e f f i c i e n t found for the 182 normals reported on i n the IM. Kaufman and Kaufman (1983b) concluded that the high ACH--FSIQ relationship "was anticipated because of the heavy weight given to verbal a b i l i t y and factual knowledge i n determining a child's global IQ on the WISC-R" (p. 111). However, the ACH Scale and the MPC correlated .43 for the Cantonese, .59 for the English, and .51 for the Punjabi. Although these correlations are not as high as the ACH-FSIQ correlations they are s t i l l s i g n i f i c a n t (£ < .001) and within the moderate range. This brings into question the a b i l i t y of a test to separate acquired knowledge from problem-solving a b i l i t y — one of the goals set for developing the K-ABC. In summary, the investigation of the performance of the three groups on the K-ABC and WISC-R resulted i n the following findings: 1) The mean MPC and FSIQ scores did not d i f f e r by more than four points for any group. However, the magnitude of many of the i n d i v i d u a l discrepancies indicates that s i g n i f -173 ic a n t l y d i f f e r e n t conclusions can be drawn with regard to educational placement depending on the test used. This suggests that research needs to be conducted to investigate what confidence i n t e r v a l i s appropriate to determine when an individual c h i l d ' s performance on the FSIQ i s s i g n i f i c a n t l y d i f f e r e n t from his or her performance on the MPC. 2) The MPC had s u f f i c i e n t shared variance with the FSIQ to suggest they are measuring similar constructs for the English and Punjabi. For the Cantonese, however, the MPC and FSIQ are not measuring the same construct to the same degree as the other two groups. 3) The SEQ Scale appears to be measuring a construct not assessed by the WISC-R. D i g i t Span i s the subtest that comes closest to measuring a b i l i t i e s measured by the SEQ Scale and i t may not be included in the computation of the WISC-R FSIQ. There are inconsistencies, for the Hand Move-ments subtest i s not c l e a r l y a sequential processing task, and the Vocabulary subtest for the English group correlated s i g n i f i c a n t l y with the Word Order subtest. 4) The SIM Scale and PIQ both have v i s u a l - s p a t i a l and simultaneous components. The superior v i s u a l - s p a t i a l s k i l l s of the Cantonese are better measured by the PIQ due to i t s higher c e i l i n g . S i m i l a r l y , some of the PIQ subtests may be better measures of simultaneous processing for 8 and 9 year 174 olds -than the MPC subtests. 5) The Kaufmans' attempts to separate acquired knowledge from problem-solving a b i l i t y have not, by the c o r r e l a t i o n a l evidence, proven successful on the K-ABC. This may, however, be further evidence that the two constructs are not independent. 175 CHAPTER VIII Interpretation of Group and Test Differences As previously i d e n t i f i e d i n Chapters V through VII, differences were observed among the groups on a number of the biodemographic variables, and on K-ABC and WISC-R subtests and scales. It was not the focus of the present investigation to determine the contributing factors to the group differences found on the K-ABC or the WISC-R. Never-theless, due to the number and the magnitude of differences among groups on these two cognitive measures, some explica-t i o n of possible contributors i s i n order. A canonical analysis was performed using the Biomedical Computer Programs P6M series (BMDP6M) (Dixon, 1981). For many of the biodemographic variables unequal sample sizes, o u t l i e r s , skewed variables and m u l t i c o l l i n e a r variables were evidenced. The r e s u l t i n g solutions from the multivariate analyses of covariance and multiple regression analyses were unstable. Hence, i t was decided to attempt to describe group differences by l o g i c a l l y integrating the subjects' test performance with their biodemographic c h a r a c t e r i s t i c s . S i m i l a r l y , i n d i v i d u a l differences i n excess of 15 points between the K-ABC Mental Processing Composite and WISC-R F u l l Scale IQ are discussed i n terms of the 176 performance of these subjects on additional K-ABC scales and subtests, and the i r biodemographic c h a r a c t e r i s t i c s . Interpretation of Group Differences Cantonese The Cantonese performed s i g n i f i c a n t l y higher than the other two groups on Triangles (K-ABC), Block Design (WISC-R) and Coding (WISC-R). These three subtests require a c h i l d to work e f f e c t i v e l y under time pressure as well as to have strong v i s u a l - s p a t i a l and visual-motor s k i l l s . At the scale l e v e l , the Cantonese also achieved higher scores on scales having a v i s u a l - s p a t i a l component ( i . e . , Simultaneous Processing and Performance IQ) than the more verbal or l i n g u i s t i c a l l y oriented scales ( i . e . , Achievement, Sequential Processing and verbal IQ). Vernon (1984) also found that Chinese children (language spoken not i d e n t i f i e d ) generally have superior v i s u a l - s p a t i a l s k i l l s as measured by the WISC-R Performance IQ and lower (yet average) verbal a b i l i t i e s as assessed by the WISC-R Verbal IQ. Lesser, F i f e r and Clark (1965) also reported a similar p r o f i l e . This pattern was replicated by the Cantonese children i n the present study. Cummins (1984a) investigated the performance of 264 English as a Second Language (ESL) children on the Perform-177 ance IQ and 234 ESL children on the Verbal IQ ( c u l t u r a l / l i n g u i s t i c background was not provided). These ESL children performed the lowest on Information and Vocabulary. However, the Cantonese i n t h i s study had a mean scaled score of 10 (average) on both of these subtests, and they performed the lowest on Comprehension and D i g i t Span. D i g i t Span was the verbal subtest which the ESL children i n Cummins' (1984a) study performed the highest on. Compared with the ESL children studied by Cummins, the Cantonese i n t h i s study appeared to perform d i f f e r e n t l y on the WISC-R. However, the WISC-R p r o f i l e for the ESL children in Cummins' study suggests that h i s subjects were less fluent i n English, on average, than the Cantonese subjects i n t h i s study. This may indicate that the test p r o f i l e for the Cantonese i s related more to t h e i r cognitive s t y l e than English fluency. Some of the biodemographic variables provide insight into the performance of the Cantonese on the K-ABC and WISC-R. S p e c i f i c a l l y , the teachers rated the Cantonese students as having superior a b i l i t y to concentrate, persevere, and plan/organize compared with the English and Punjabi students. A debriefing of the examiners also revealed that compared with the other two groups, the Cantonese were, on average, quicker to learn from th e i r errors (especially on the Triangles and Block Design 178 subtests), better able to concentrate, and more goal directed i n t h e i r test taking behavior. This may be a further i n d i c a t i o n that the cognitive p r o f i l e emerging for the Cantonese i s c u l t u r a l s p e c i f i c . The other biodemographic variables, such as socio-economic status, did not d i f f e r e n t i a t e the Cantonese from the English and the Punjabi. As such, these variables are not as e a s i l y interpreted as contributors to the v i s u a l -s p a t i a l strength found for the Cantonese on the K-ABC and WISC-R. English The English outperformed the Cantonese on the Riddles (K-ABC) and the Vocabulary (WISC-R) subtests. Both subtests are measures of verbal conceptualization and are influenced by c u l t u r a l background. Given that the English children and th e i r parents spoke only English at home, the English had more experience with the English language than the other two groups. Furthermore, the English children had s i g n i f i c a n t l y more Canadian born parents and grandparents than the other two groups. As such, the English group had longer to "Canadianize" than the Cantonese and Punjabi. Not surprisingly, the teachers also rated the English as more fluent i n communicating i n English ( i . e . , speaking, 179 understanding) than the Cantonese and Punjabi. The teachers did not rate the English as having a more p r o f i c i e n t learning s t y l e than the other two groups. There-fore, the superior performance of the English on two verbal tasks may be more related to t h e i r p r o f i c i e n c y with the English language than a c u l t u r a l l y s p e c i f i c cognitive s t y l e . The English achieved a mean Mental Processing Composite of 104.73 and F u l l Scale IQ of 108.73. Since these mean scores are above the mean of 100 set for both these tests, the results appear to support the findings that show English Canadian children as having higher mean WISC-R Standard Scores compared to the American standardization sample (Hardman, 1984; Holmes, 1981; Peters, 1976). However, a report i n the IM on the performance of 182 normal children showed them to have a mean Mental Processing Composite of 113.6 and a mean F u l l Scale IQ of 116.7. These American children scored even higher than the English children i n the present study. This suggests that sampling a r t i f a c t s may be contributing to the higher mean performance of some English speaking Canadians and Americans on cognitive tests when compared with the more hetergeneous standardization samples. On the K-ABC the English performed s i g n i f i c a n t l y lower on the Faces and places subtest than they did on the other Achievement subtests. As previously mentioned, this subtest does not appear to f a i r l y assess the range of general 180 factual knowledge possessed by Canadian children (Saklofske & J e d l i c k i , 1985). Rather, i t appears to be a more s p e c i f i c measure of general knowledge possessed by American children. Although not a l l items are s p e c i f i c to the American culture, enough of them are to place Canadian children at a disadvantage on t h i s subtest. Punjabi Except for the SEQ subtests and Reading/Decoding on the K-ABC, and the Arithmetic and D i g i t Span subtests on the WISC-R, the Punjabi had mean subtest scores s i g n i f i c a n t l y lower than one or both of the other two groups. Sternberg (1984) commented that tasks requiring rote memorization (e.g., Sequential Memory tasks) generally show lower cor-relations with other measures of i n t e l l i g e n c e and smaller r a c i a l and ethnic differences. The early items on the A r i t h -metic and Reading/Decoding subtests are similar to D i g i t Span and Hand Movements, for example, i n that they too are rote memory tasks. On the K-ABC Mental Processing subtests the Punjabi did not demonstrate a s i g n i f i c a n t spread i n t h e i r scores, for t h e i r means ranged from 8.79 (Gestalt Closure) to 9.73 (Hand Movements). It should be noted that these means do camou-flage i n d i v i d u a l differences. On the WISC-R subtests s i g n i f -icant v a r i a b i l i t y (3 point spread) was evident with subtest 181 means ranging from 7.57 (Information) to 10.63 (Picture Arrangement). They performed the poorest on the four sub-tests which measure verbal conceptualization and are influenced by c u l t u r a l background ( i . e . , Information, Comprehension, S i m i l a r i t i e s , and Vocabulary). Coincidental-l y , the ESL children studied by Cummins (1984a) also performed the poorest on these four subtests. On four other subtests, namely: Arithmetic, Coding, Mazes and Picture Arrangement, the Punjabi group performed better, obtaining a mean standard score of 10 on each of these. Not only do these subtests have low verbal content, they were i d e n t i f i e d by Kaufman (1979) as being sequential tasks. However, the way Coding and Picture Arrangement tasks are processed fluctuates with a b i l i t y l e v e l ( N a g l i e r i , Kamphaus & Kaufman, 1983). The ESL children studied by Cummins (1984) also performed better on these subtests than those requiring verbal conceptualization. An investigation of the biodemographic variables showed the Punjabi children had parents who were, on average, born in more rur a l areas than the more urban born Cantonese and English parents. However, the Punjabi children were raised in an urban setting and the number of years they had li v e d in their present community did not d i f f e r s i g n i f i c a n t l y from the other two groups. However, their parents' early environ-mental upbringing can not be discounted as a variable having an e f f e c t on the way their children perform on cognitive 182 measures. The Punjabi were rated by the teachers as being lower than the Cantonese in their a b i l i t y to master new material, concentrate, retain what has been taught, persevere, and plan/organize. This may suggest that there exist c u l t u r a l variations in the learning styles of these two groups. These two groups did not d i f f e r s i g n i f i c a n t l y in terms of the frequency with which they spoke English at home or in their teachers' perceptions of their fluency in communicating in English ( i . e . , speaking, understanding). However, no measure of the proficiency with which the parents used English was included. From the comments made by the examiners, i t would appear that, compared with the Cantonese, the Punjabi were not as strongly motivated to get the correct solution to the test items. In conclusion, group differences were observed on the K-ABC and WISC-R. While some possible explanations for these differences have been provided, this i s a complex issue with no d e f i n i t i v e solution. In addition to those hypothesized above, other reasons for the differences found among the groups on the cognitive measures could feasibly include: d i f f e r e n t c h i l d rearing practices (Vernon, 1984); dys-functional educational processes (e.g., teacher style does not match learner style) for some groups (Burke, 1984); academic pursuit not stressed by certain cultures' (Samuda, 183 1984); unfamiliarity with the testing procedures (Samuda, 1984); r e l i g i o u s b e l i e f s (Akhilanada, 1951; Burke, 1984; Ferron, 1973); temperament (Garth, 1931), and/or concept of speed unimportant to s p e c i f i c cultures (Samuda, 1984). No data were collected in this study to investigate these p o s s i b i l i t i e s . Interpretation of Test Differences As i d e n t i f i e d in Chapter VII, 15 of the Cantonese, 9 of the English, and 8 of the Punjabi achieved a K-ABC Mental Processing Composite (MPC) and a WISC-R F u l l Scale IQ (FSIQ) which d i f f e r e d by 15 or more points. Educationally the number of subjects showing this magnitude of a discrepancy is cause for concern - e s p e c i a l l y i f class placement is the objective of an i n t e l l i g e n c e assessment. For example, a 15 point difference between performance on two int e l l i g e n c e tests which are supposed to measure the same learning p o t e n t i a l , could mean the difference between a c h i l d receiving a special class placement versus a regular class placement - depending on which of the two tests is administered. Therefore, a description of the character-i s t i c s of the children with between-test discrepancies of .15 or more points w i l l be given in an attempt to provide psychologists with an assessment p r o f i l e of these children in each group. 184 Cantonese Of the 15 Cantonese (9 boys, 6 g i r l s ) with a discrepancy of 15 or more points (between these two t e s t s ) , 12 were found with a lower MPC than FSIQ. No consistent pattern emerged which could d i f f e r e n t i a t e the 12 children with superior FSIQs from the 3 children with superior MPCs with respect to: t h e i r teachers' rating of th e i r English fluency and learning s t y l e ; and the i r Sequential/Simul-taneous and Verbal IQ/Performance IQ p r o f i l e s . However, of these 15 children showing a s i g n i f i c a n t MPC and FSIQ discrepancy, 13 scored higher on the Performance IQ than the Simultaneous Processing Scale. This pattern may be related to the lower c e i l i n g e f f e c t on the Simultaneous Scale (Bracken, 1985; Thomas, 1984) perhaps r e s u l t i n g from the lack of time bonus points (Sternberg, 1984) on the Simultaneous Processing Scale compared to time bonus points given on Performance IQ items. There was a tendency for the Cantonese to c e i l i n g out on Triangles (K-ABC) and not on Block Design (WISC-R). There are two d i s t i n c t differences between these two subtests. F i r s t the WISC-R offers time-bonus points which are not offered on the K-ABC. The examiners remarked at how quickly many of the Cantonese 185 children could manipulate the triangles and blocks to rep l i c a t e a design. It appears that the WISC-R (with i t s time-bonus points) may be a more sensitive or rewarding measure of t h i s . Secondly, the WISC-R serves children up to and including 16 years - 11 months, while the K-ABC has an upper age l i m i t of 12 years - 6 months. As a result, the K-ABC does not appear to have an adequate upward extension: Triangles, for example, for Cantonese children 8 years of age or older. Other evidence to suggest the WISC-R PIQ i s a more sensitive measure of v i s u a l - s p a t i a l a b i l i t i e s than the K-ABC SIM Scale comes from an investigation of the teacher ratings of the Cantonese children's learning s t y l e s . They were rated by t h e i r teachers as having superior a b i l i t y to concentrate, persevere, and plan and organize than the other two groups. Strong v i s u a l - s p a t i a l s k i l l s are dependent upon these a b i l i t i e s . Although there appears to be evidence that the Cantonese generally have superior v i s u a l - s p a t i a l a b i l i t i e s than other c u l t u r a l groups, the predictive v a l i d i t y of both the K-ABC and WISC-R for the Cantonese requires investigation. 186 English A consistent p r o f i l e emerged for the 9 English children (4 boys, 5 g i r l s ) with respect to the i r MPC and FSIQ discrepancies. A l l 9 children were found to have a FSIQ greater than a MPC. This was not related just to superior verbal a b i l i t y , for 7 of the 9 children had Performance IQs greater than Verbal IQs. Moreover, for a l l 9 children the Performance IQ was superior to the Simultaneous Scale score. This may be related to the higher c e i l i n g on the Performance Scale. Similarly, the majority (n = 6) of the children scored higher on the Verbal IQ than the Achievement Scale. This may be related to the effects of the Faces & places subtest lowering the o v e r a l l Achievement Scale score for these children, for the mean Prorated Achievement Scale scores were 2.70 (Cantonese), 3.00 (English) and 3.48 (Punjabi) points higher than the Achievement Scale score. For a l l 9 children the Verbal IQ was higher than the Sequential Scale score. It appears that when the K-ABC and WISC-R are admin-ist e r e d to average or higher functioning English children there i s a tendency for them to perform better on the WISC-R. In contrast, lower functioning children have been known to perform higher on the K-ABC than the WISC-R (Naglieri, i n press a; Obrzut et a l . , 1984). Research w i l l need to be conducted to investigate the performance of lower 1 8 7 functioning English Canadians on the K-ABC and WISC-R to determine the predictive v a l i d i t y of each instrument. Punjabi There were 8 Punjabi children ( 7 boys, 1 g i r l ) showing a discrepancy of 15 or more points between the two i n t e l -ligence t e s t s . Unlike the other two groups r e l a t i v e l y more Punjabi had a superior MPC (n = 5) than FSIQ (n = 3). The three children showing superior FSIQs consistently performed higher on a l l the WISC-R scales than the K-ABC scales. These children a l l had FSIQs greater than 90 and average or higher Verbal IQs. In contrast, the f i v e children with superior MPCs a l l had FSIQs less than 90. This suggests that lower functioning Punjabi children (as assessed by the FSIQ) do not perform as well on the WISC-R as they do on the K-ABC. These children did not appear to d i f f e r from the children with higher FSIQs in terms of th e i r parents' l e v e l of education, employment status, or b i r t h place. Also there was no difference among these children in terms of the frequency with which English was spoken i n the home. The predictive v a l i d i t y of both tests for the Punjabi requires investigation. In summary, the WISC-R appears to be more sensitive to making fine discriminations between performances given by higher functioning children, while the K-ABC w i l l give lower 188 (WISC-R) functioning children a higher i n t e l l i g e n c e score. The predictive v a l i d i t y of both measures, therefore, needs to be investigated. In addition, a detailed analyses of the ethnic factors which contribute to the performance of the three groups on these two measures needs to be explored. CHAPTER IX 189 Summary, Conclusions, and Recommendations The purpose, procedure and results are summarized i n th i s f i n a l chapter. The li m i t a t i o n s of the study are detailed and the p r a c t i c a l implications of the findings are discussed. Recommendations for future avenues of research are suggested. Summary of the Study The purpose of this study was to assess the construct v a l i d i t y of the K-ABC for use with three groups of Canadian children. The subpopulations i d e n t i f i e d were Cantonese, English and Punjabi speaking t h i r d graders attending public school i n Vancouver, a large c i t y i n B r i t i s h Columbia. The samples were further r e s t r i c t e d i n that a l l subjects were Canadian born, none were Native Indians, and none had been previously diagnosed as having emotional, mental, physical, or sensory handicaps. A l l subjects were volunteers. The students selected were taken from 34 classes i n 21 schools within the c i t y . The genders were equally represent-ed within each group. The children ranged i n age from 8 years, 1 month to 10 years, 5 months (mean age, 8 years, 8 months). The mean age did not vary s i g n i f i c a n t l y among the groups. 190 Biodemographic data were co l l e c t e d on each subject from th e i r student record cards, t h e i r parents (Parent Questionnaire), and t h e i r teachers (Teacher Questionnaire and Teacher Rating Scale). The WISC-R was administered as a c r i t e r i o n measure of i n t e l l i g e n c e with which the K-ABC was compared. The two cognitive measures were administered i n a counter-balanced fashion within a one-week period. The order i n which the K-ABC and WISC-R were administered did not s i g n i f i c a n t l y a f f e c t the mean scores for each group. The findings of the study are summarized below as they pertain to the s p e c i f i c research issues. Group Differences Differences existed among the groups i n terms of the i r mean test score, for 11 of the 13 K-ABC subtests (when co n t r o l l i n g for the type 1 error rate only 9 of 13 subtests) showed s i g n i f i c a n t differences among the groups. The groups also d i f f e r e d s i g n i f i c a n t l y on a l l of the K-ABC scales. More s p e c i f i c a l l y , the Cantonese demonstrated they were superior to both the English and the Punjabi i n the i r a b i l i t y to r e c a l l the placement of pictures on a page (Spatial Memory) and t h e i r a b i l i t y to assemble triangles (Triangles). The English, however, performed better than the Cantonese and Punjabi on Riddles - a language task requiring the c h i l d to 191 name a concept after given a l i s t of i t s c h a r a c t e r i s t i c s . Except for the SEQ subtests and Reading/Decoding, the Punjabi performed s i g n i f i c a n t l y lower than the Cantonese and/or English on the remaining 9 subtests and on a l l of the K-ABC scales except for Sequential Processing. The standard deviations for the subtests and scales of the K-ABC were generally smaller than the IM standardization sample. Given the r e s t r i c t e d a b i l i t y range in this study and the c u l t u r a l s p e c i f i c i t y of each group, this was an expected outcome. The r e l i a b i l i t y (internal consistency) estimates were somewhat lower than those reported in the IM (probably a res u l t of the r e s t r i c t e d range and homogeneity in the present study), however, they were strong enough to indicate the test had good r e l i a b i l i t y . The internal consistency estimate of the Mental Processing Composite for each group was .89 while the Achievement Scale had a r e l i a b i l i t y estimate in excess of .90 for each group. Factor Structure A confirmatory factor analysis revealed that the Sequential/Simultaneous theoretical model was supported by the English and Punjabi data. I t could also support a Verbal/Spatial hypothesis. However, the Cantonese data did not exhibit a good f i t with this model. S i m i l a r l y , the 192 results of the exploratory factor analysis suggested that Sequential and Simultaneous factors could apply when des-cribing the internal structure of the K-ABC for the English and the Punjabi data. Consistent with other research (Kauf-man & Kamphaus, 1984; Keith, 1985), Hand Movements (a Sequential subtest) f a i l e d to load on the Sequential factor for the English and Punjabi. Das (1984b) suggested a verbal/ nonverbal dichotomy may explain this factor pattern while Keith and Dunbar (1984) proposed a verbal-memory/nonverbal reasoning dichotomy. These alternate models require empirical v a l i d a t i o n . The internal structure of the K-ABC data for the Cantonese was not as c l e a r l y explained by a Sequential/ Simultaneous factor pattern. Rather, a memory/reasoning dichotomy may apply. There is also some support for an auditory memory/visual memory/visual-spatial trichotomy. A l l proposed models require empirical v a l i d a t i o n . For a l l three groups the addition of the five Achieve-ment subtests in the factor analysis did not a l t e r the factor pattern of the Sequential and Simultaneous subtests. This suggests that the factor pattern for the Mental Processing subtests was r e l a t i v e l y stable. Relationship between K-ABC and WISC-R The mean MPC and mean FSIQ did not d i f f e r by more than 193 four points for each group. However, 15 Cantonese, 9 English, and 8 Punjabi were found to have discrepancies of 15 or more points between these two measures. The Cantonese were i d e n t i f i e d on the K-ABC and WISC-R as having strong v i s u a l - s p a t i a l s k i l l s . However, they performed the highest on the WISC-R. This resulted in more of the Cantonese having superior FSIQs than MPCs. Of the English children showing discrepant i n t e l l i g e n c e scores, a l l had superior FSIQs. The Punjabi, however, were more i n c l i n e d to perform higher on the MPC than the FSIQ, es p e c i a l l y i f they had an FSIQ less than 90. The s u f f i c i e n t shared variance between the MPC and FSIQ for the English and the Punjabi provides support for the v a l i d i t y of the K-ABC as measuring a construct similar to the WISC-R for these two groups. However, with less than 20% shared variance between the MPC and FSIQ for the Cantonese, i t would appear that the two tests are not measuring the same construct for t h i s group. Implications for Using the K-ABC The results of t h i s study support the construct v a l i d i t y of the K-ABC, as i t pertains to the test's i n t e r n a l structure, for use with English and Punjabi children. The f a i l u r e of Hand Movements, however, to load on the 194 hypothesized Sequential factor, indicates that psychologists should be cautious in interpreting this subtest as an indicator of Sequential processing unless a number of observations of an individual child's test performance suggests otherwise. Given the evidence of a developmental or age s h i f t in performance of children on Hand Movements (Kaufman & Kamphaus, 1984), various factor models across the age groups should be explored to explain the age s h i f t on Hand Movements. Since the Sequential/Simultaneous factor model was not supported by the Cantonese data, caution should be extended in interpreting a Cantonese child's performance on the K-ABC as indicative of this cognitive style dichotomy. While other models have been proposed (e.g., memory/reasoning), these also require v a l i d a t i o n . The correlations between the K-ABC Mental Processing Composite and the WISC-R F u l l Scale IQ for the English and Punjabi supports the v a l i d i t y of the K-ABC as measuring a construct similar to the WISC-R. One should expect average-to-higher functioning English children to do less well on the K-ABC. As such, fewer of them may be e l i g i b l e for g i f t e d programs i f the K-ABC i s the primary diagnostic c r i t e r i o n . Although there was not a s i g n i f i c a n t difference between the mean performance of the Punjabi on the Mental Processing Composite and F u l l Scale IQ, di a g n o s t i c a l l y there are some 195 s i g n i f i c a n t trends. For example, Punjabi children with FSIQ 15 points or greater than their Mental Processing Composite had a F u l l Scale IQ in excess of 90 and average to better Verbal IQs. On the other hand, Punjabi children with a Mental Processing Composite higher (15 or more points) than their F u l l Scale IQ had a F u l l Scale IQ less than 90 and a below average Verbal IQ. The moderate co r r e l a t i o n between the two int e l l i g e n c e scales for the Cantonese suggests that the Mental Processing Scale is measuring a b i l i t i e s not d i r e c t l y measured by the F u l l Scale IQ. U n t i l psychologists can be more confident as to what the K-ABC i s measuring for Cantonese children, they should not use the K-ABC as the only measure of "i n t e l l i g e n c e " for these children. This is not to say that the K-ABC does not have diagnostic relevance — for this w i l l need to be further investigated. As an example, observing the individual c h i l d ' s performance on each subtest to determine the strategies he or she i s employing may be one method. In addition, the predictive v a l i d i t y of the K-ABC and WISC-R requires investigation for a l l groups of children, for i t has yet to be empirically evidenced that the WISC-R is a more v a l i d measure of inte l l i g e n c e than the K-ABC. However, in this study, the Cantonese children did perform higher on the Performance IQ than the Simultaneous Processing scale, p a r t l y because not a l l of the Simultaneous 196 subtests had s u f f i c i e n t c e i l i n g s . The fact that the K-ABC appears to be a r e l i a b l e test for the English and Punjabi further supports i t s use with these groups. The K-ABC was found to be a r e l i a b l e measure for use with the Cantonese, however, because of i t s questionable v a l i d i t y i t should be used cautiously with these children. Psychologists should be careful when interpreting Canadian children's performance on the Achievement Scale. The Canadians i n thi s study performed s i g n i f i c a n t l y lower on th i s scale than on the i n t e l l i g e n c e scale. This may have serious implications i n terms of making diagnostic decisions based on the intelligence/achievement discrepancies. The performance of a Canadian c h i l d on the Faces and Places subtest should not be interpreted as evidence of general factual knowledge, since the face v a l i d i t y of thi s subtest for Canadian children has been found to be questionable (Saklofske & J e d l i c k i , 1985). The difference i n performance among the groups on the K-ABC and the WISC-R strongly suggests the importance of vali d a t i n g a l l tests for the various c u l t u r a l groups for whom they w i l l be administered. Considering the investigation of the factor structure of the K-ABC standardization sample has been done on pooled ethnic groups ( i . e . , Blacks, Hispanics, P a c i f i c Islanders/Asians, and Whites) by separating out the groups and analyzing them 1 9 7 separately, i t may be found that the various groups have c u l t u r a l l y s p e c i f i c patterns which w i l l provide insight into the cognitive styles of the various groups. Implications of Assessing Minority Children The Kaufmans need to be commended on the i r attempts to provide the i n t e l l i g e n c e testing movement with an alternative approach to investigating cognition in children. Moreover, they have provided researchers and psychologists with a detailed account of t h e i r rationale, res u l t s , and inter p r e t i v e procedures i n the IM. American Guidance Service also needs to be recognized i n t h e i r willingness to f i n a n c i a l l y support ongoing research on the K-ABC. The Kaufmans' attempts at developing a test which investigates psychological processes i s an exciting approach which provides an alte r n a t i v e to t r a d i t i o n a l i n t e l l i g e n c e t e s t s . Bagley, Iwawaki, and Young (1983) stated that "comparing psychological processes across cultures can provide valuable information on how culture influences cognition" (p. 27). However, the K-ABC has not met i t s p o t e n t i a l . There are issues that need to be considered when using the K-ABC: 198 1) The concern for whether the K-ABC measures sequential and simultaneous processing has been directed at the evidence of i t s factor structure (Das, 1984b; Keith & Dunbar, 1984). Even more c l e a r l y evidenced in the present study i s the c u l t u r a l v a r i a b i l i t y i n the int e r n a l structure of the K-ABC. 2) Although the authors of the K-ABC intended i t to be a measure of an individual's s t y l e of processing informa-tion, i t appears to only measure how a c h i l d performs on tasks hypothesized to measure Sequential and Simultaneous processing. As Frederiksen (1977) pointed out, even i f two children have the same score on the same test, these children may be employing d i f f e r e n t processing methods. As such, the tests do not necessarily measure the same constructs between groups (Brody & Brody, 1976) or ind i v i d u a l s . 3) The complexity of the Sequential Processing Scale remains questionable (Bracken, 1984). It appears to be just a measure of Sequential memory. 4) There i s no evidence that summing two content areas (Sequential/simultaneous) can give a t o t a l i n t e l l i g e n c e score. This also applies to the WISC-R (Verbal/performance). 5) Although advocating a new approach to testing, Kaufman and Kaufman have made the K-ABC a less discrimina-tin g measure of i n t e l l i g e n c e which they appear to believe 199 makes the test more " f a i r " for testing minority children. Instead of recognizing that c u l t u r a l groups d i f f e r and trying to design an instrument that w i l l i d e n t i f y t h e various differences i n cognitive style among the various groups, they have limited t h e i r test to just two measures of processing. If they did not f e e l bound to come up with a " t o t a l i n t e l l i g e n c e " score, but rather a sampling of select i v e measures of cognitive style, i t would not be necessary to minimize c u l t u r a l , developmental, and gender differences. 6) Related to the l e v e l of complexity of the K-ABC i s the type of problem solving i t assesses. Generally, i t appears to be a measure of how children solve problems for which they have s p e c i f i c knowledge to do so. While the content may d i f f e r and the materials may be novel, generally the problem-solving i s reproductive i n nature. Therefore, i t does not adequately assess creative problem-solvers who can not only generate solutions but can generate problems. At the expense of making lower functioning children appear brighter, the K-ABC appears to make g i f t e d children appear less so. 7) F i n a l l y , while the Kaufmans deserve cre d i t for at least recognizing that test developers need to be sensitive and thoughtful i n the i r e f f o r t s at recognizing the assessment needs of minority children, they do not develop 200 this idea to the f u l l e s t . I t would appear that what is required are awareness, knowledge, and understanding of the various ethnic factors that relate to a s p e c i f i c group's a b i l i t y to perform and achieve i n the majority culture. Differences among ethnic groups may be related to c u l t u r a l , demographic, economic, educational, and s o c i a l variables that a f f e c t the c h i l d ' s cognitive s t y l e , i n t e l l e c t u a l growth, academic achievement, and test performance. More s p e c i f i c a l l y , culture i s one determiner of how people interact with and perceive the world (Bagley, 1 9 8 4 ) . People generally adapt to the demands of th e i r culture (Berry, 1983; Williams, 1970) and cognitive style i s a r e f l e c t i o n of the demands of a culture (Maccoby & Modiano, 1971). For example, "The modern i n d u s t r i a l i z e d world demands abstraction by i t s very arrangements, i t s stimuli, i t s contrasts, i t laws of j u s t i c e and exchange. What i s demanded of the peasant, on the other hand, i s that he pay attention to his crops, the weather, and the p a r t i c u l a r people around him" (Maccoby & Modiano, 1971, p. 293). They added that even the way a c h i l d i s reared a f f e c t s his or her a b i l i t y to deal with abstraction. Even children raised outside of their parents' country of b i r t h i d e n t i f y to a certain extent with t h e i r parents' c u l t u r a l background (Vyas, 1983). The family i s a s o c i a l unit and cognitive style i s the 201 end product of thi s s o c i a l i z a t i o n process (ponnuswami, 1977). Many aspects of thi s family s o c i a l i z a t i o n process a f f e c t cognitive growth. Some of these include parental aspirations (Cummins, 1982, 1984c), qua l i t y of mother-child in t e r a c t i o n (Ponnaswami, 1977; Vyas, 1983), family size and b i r t h order (Brody & Brody, 1976). The s o c i a l class membership (Brody & Brody, 1976) or socioeconomic status (Mercer, 1979; Sattler, 1982) of the family has been found to be a contributing factor i n the performance of Chinese, Jews, Puerto Rican and Black children on i n t e l l i g e n c e measures (Lesser, F i f e r & Clark, 1965). The differences i n s o c i a l class within each ethnic group can be as large as differences between groups (Laosa, 1977). However, minority children are generally poorer than majority group children (Esquire, 1985; Samuda, 1983), and lower class children score lower on i n t e l l i g e n c e measures regardless of ethnic group (Laosa, 1977). This c u l t u r a l deprivation results from inadequacies i n the child's home learning environment (Marjoribanks, 1980). Another factor related to a child's s o c i a l i z a t i o n i s the degree of ambivalence between the home culture and the majority culture (Cummins, 1984b). This has an impact on the chil d ' s adjustment and academic achievement (Bhatnagar, 1970). The school environment also affects the academic 202 achievement of the c h i l d (Marjoribanks, 1980). A school r e f l e c t s the c u l t u r a l t r a i t s of the majority group. For example, i n an urban/industrial society more abstraction i s dealt with i n school (Maccoby & Modiano, 1971) than i n a r u r a l society. In addition to the material taught, a teacher has a cert a i n s t y l e that may a f f e c t a child's achievement (Bagley & Verma, 1983) such as the type of reinforcement used (Bagley, 1983). Intelligence tests are generally designed to predict school achievement (Reschly, 1979) and are representative of middle class majority culture values (Samuda, 1983). Children schooled i n i n d u s t r i a l cultures are more familiar with t e s t material (Rogoff, 1981), te s t demands (Rogaff, 1981), and t e s t language (Rogoff, 1981; Van der F l i e r , 1977) than children raised i n t h i r d world countries. Subsequently, an i n t e l l i g e n c e test developed in a Western society may not measure the same construct between various c u l t u r a l groups (Brody & Brody, 1976). Rather, the tests may be a r e f l e c t i o n of differences i n cognitive styles between c u l t u r a l groups (Edgerton & Langness, 1974; Frederiksen, 1977). The perceptual and language experiences of a c h i l d are also related to a chil d ' s c u l t u r a l , economic, educational, and s o c i a l environment and can a f f e c t the cognitive style and cognitive growth of a c h i l d . Perceptual a b i l i t i e s have been found to vary across cultures (Engerton & Langress, 203 1979; Shade, 1981). Sinha (1977) concluded that i n India, c u l t u r a l and economic deprivation have had an e f f e c t on the perceptual growth of children. The lower perceptual competence i n disadvantaged Indian children i s related to quality of schooling and lack of motivation of parents towards education th e i r children (Sinha, 1977). Poor language s k i l l s also i n h i b i t i n t e l l e c t u a l growth (Alleyne, 1977), for language proficiency relates to a chi l d ' s a b i l i t y to interact with his or her environment. According to Samuda (1983) and S a v i l l e (1977) ethnic groups have languages which d i f f e r i n t h e i r l e x i c a l , morphological, phonological, s y n t a c t i c a l , and intonational structure. Level of language prof i c i e n c y i s an important determiner of how a c h i l d w i l l perform co g n i t i v e l y (Cummins, 1982; 1984c; Rees, 1982). Children who are inadequate monolinguals w i l l perform poorly on cognitive measures (Rees, 1982) as w i l l children who have limited profiency i n two languages (Cummins, 1982; 1984c). In conclusion, Schludermann and Schludermann (1976) concluded that "Cross-cultural researchers have shown that a variety of needs, such as a f f i l i a t i o n , approval-seeking, power, avoidance of shame, are relevant motivating forces for achievement" (p. 156). They added that what i s needed " i s a thorough, comprehensive and detailed knowledge of work-related motives, attitudes, and values in a p a r t i c u l a r 2 0 4 society" (p. 157). In the future, test developers should look towards developing instruments related to various processing a b i l i t i e s or cognitive s t y l e s . Psychologists, on the other hand, should become more aware of the various c u l t u r a l factors that relate to a c h i l d ' s cognitive s t y l e . Together, they w i l l be able to e f f e c t i v e l y assess children so appropriate educational programs can be developed to increase the children's a b i l i t y to meet the demands of their new culture without having them lose their ethnic i d e n t i t y . Limitations of the Study 1. The study involved only volunteer subjects of Cantonese, English and Punjabi language backgrounds who were enrolled in grade 3 classes i n a large urban Canadian c i t y . 2. The g e n e r a l i z a b i l i t y of the findings i s further limited to Canadian born children with no emotional, mental, physical or sensory handicaps. 3. Given the three groups of children were not equally represented within each school, and i t was not feasible for each examiner to test equal numbers of children from each c u l t u r a l group, an estimate of examiner ef f e c t could not be computed. 205 Recommendations for Future Research Based on the findings of thi s study the following research questions are provided as possible avenues of future research. 1. What i s the construct v a l i d i t y (e.g., inte r n a l structure) of the K-ABC for a l l c u l t u r a l groups t o whom i t w i l l be administered? 2. What i s the predictive v a l i d i t y of the K-ABC f o r Canadian children? 3. What i s the content v a l i d i t y of the K-ABC Faces & Places subtest for Canadian children? 4. How v a l i d are the K-ABC Achievement and Prorated Achievement Scales for Canadian children? 5. What i s the d i f f e r e n t i a l v a l i d i t y of the Sequential Processing and Simultaneous Processing Scales for use with Canadian children? 6. What are the diagnostic implications of the inequity i n the Sequential/Simultaneous Processing Scales as they contribute to the t o t a l i n t e l l i g e n c e scale (Mental Processing Composite)? 7. Given there i s some doubt as to the complexity of the Sequential Processing Scale as a measure of 206 i n t e l l i g e n c e , what other tasks might be incorporated as part of an assessment battery to enhance the complexity of the Sequential Processing Scale? 8. What other t h e o r e t i c a l models might explain the data for Cantonese, English and Punjabi children? 9. Since the K-ABC does not have a scale that just measures planning a b i l i t y , what other measures might be used to complement the K-ABC and provide s p e c i f i c information on planning a b i l i t y ? 10. Based on their K-ABC tes t p r o f i l e s , how v a l i d are the educational intervention recommendations outlined i n the IM for Cantonese, English, and Punjabi children? 11. How v a l i d i s the K-ABC Nonverbal Scale for assessing immigrant children with limited English mastery? 12. Because the Kaufmans have presented the K-ABC "as having the capacity to answer to many c l i n i c a l questions" (Sewell, 1983), how v a l i d are each of the i r claims? For example, how v a l i d a projective instrument i s the Gestalt Closure subtest? According to Salvia and Ysseldyke (1985) "No data are presented to validate the K-ABC as a measure of learning p o t e n t i a l , for use i n educational placement and planning, for c l i n i c a l assessment, or neurological assessment. These are also avenues for future research. 13. How does English fluency (as measured by a standardized 207 test) a f f e c t the performance of various c u l t u r a l / l i n g u i s t i c groups on the K-ABC and WISC-R. 14. Why i s English fluency in idiomatic speech important to o v e r a l l i n t e l l i g e n c e development? What are the implica-tions of thi s for t r a i n i n g English as a second language students? 15. Why are the Punjabi, as a group, less motivated to concentrate on and persevere at academic and cognitive tasks than the Cantonese and English children? What are the implications of t h i s pervasiveness for curriculum development? 16. What are the ef f e c t s of regression error on using the Mental Processing Composite and Achievement Scale i n diagnosing children? 17. What generation of Canadian does a c h i l d have to be before he or she acquires the cognitive style of the majority group? What variables a f f e c t t h i s assimilation process? 18. What i s the indigenous v a l i d i t y of the K-ABC? S p e c i f i c a l l y , what " i s the extent to which c u l t u r a l group members both i d e n t i f y the dimension of behavior under scrutiny and place individuals along that dimension s i m i l a r l y to the invesigator's placement" (Irwin, Klein & Townsend, 1982). 19. When matched on such variables as socioeconomic status, 208 school, gender, English fluency, etc., what differences exist among the Cantonese, English, and Punjabi on the K-ABC and WISC-R. 209 Bibliography Akhilananda, S. (1951). Mental health and Hindu psychology. Boston: Branden Press. Alleyne, M. (1977). L i n g u i s t i c dimensions of culture and the implications for urban education. In V. d'Oyley (Ed.), The Impact of m u l t i - e t h n i c i t y on Canadian education (pp. 47-53). Toronto, Ontario: The Urban A l l i a n c e on Race Relations. American Psychological Association (1974). Standards for educational and psychological t e s t s . Washington, D.C. Anastasi, A. (1976). Psychological t e s t i n g (4th ed.). New York: Macmillan. Anastasi, A. (1982). Psychological testing (5th ed.). New York: Macmillan. Anastasi, A. (1984). The K-ABC i n h i s t o r i c a l and contemporary perspective. Journal of Special Education, 18, 357-366. Ashworth, M. & Wakefield, P. (1978). Guidelines for  administrative notes and suggestions for the  i n s t r u c t i o n of English as-a-second-language. V i c t o r i a : B r i t i s h Columbia Ministry of Education. Bagley, C. (1983). Cognitive styles, ethnicity, s o c i a l class and s o c i a l i z a t i o n i n c r o s s - c u l t u r a l perspective. In C. Bagley & G. Verma (Ed.), M u l t i c u l t u r a l childhood (pp. 3-15). Hampshire, England: Gower Publishing. Bagley, C. (1984). Cultural d i v e r s i t y , migration, and cognitive s t y l e s : A study of B r i t i s h , Japanese, Jamaican, and Indian children. In R.J. Samuda, J.W. Berry & M. L a f e r r i e r e (Eds.), Multiculturalism i n  Canada: Social and educational perspectives (pp. 368-381). Toronto: A l l y n & Bacon, Inc. Bagley, C. & Verma, G. (1983). Introduction: Issues i n m u l t i c u l t u r a l education. In C. Bagley & G. Verma (Eds.), M u l t i c u l t u r a l childhood (pp. x v i i i - x x i ) . Hampshire, England: Gower Publishing. Bagley, C. Iwawaki, S. & Young, L. (1983). Japanese Children: Group-oriented but not f i e l d dependent? In 210 Bagley, C. & G. Verma (Eds.), M u l t i c u l t u r a l childhood (pp. 27-37). Hampshire, England: Gower Publishing. Bernal, E.M. (1977). Introduction: Perspectives on nondiscriminatory assessment. In T. Oakland (Ed.), Psychological and educational assessment of minority  children (pp. x i - x i v ) . New York: Brunner/Mazel. Berry, J.W. (1983). Comparative studies of cognitive s t y l e s . In R.J. Samuda & S.L. Woods (Eds.), Perspectives i n immigrant and minority education (pp. 247-257). New York: University Press of America. Bhatnagar, J. (1970). Immigrants at school. London: Cornmarket Press. Bhatnagar, J . (1981). Multiculturalism and education of immigrants in Canada. In J . Bhatnagar (Ed.), Educating  immigrants (pp. 69-95). London: Croom Helm Ltd. Bing, S. & Bing, J . (1984, A p r i l ) . Relationship between the  K-ABC and PPVT-R for preschoolers. Paper presented at the National Association of School Psychologists Convention, Philadelphia, Pennsylvania. Blank, S.S. (1982). The challenge: Encouraging creative thinking and problem solving i n the g i f t e d . San Diego Unified School D i s t r i c t , San Diego. Bolen, L.M. & Childers, J.S. (1985). Test-Retest r e l i a b i l i t y and practice effects on the Kaufman Assessment. [Summary]. Proceedings of the 17th Annual Convention of  the National Association of School Psychologists, 49-50. Bracken, B. (1985). A c r i t i c a l review of the Kaufman Assessment Battery for Children (K-ABC). School  Psychology Review, 14, 21-36. Bray, J.H. & Maxwell, S.C. (1982). Analyzing and interpreting s i g n i f i c a n t MANOVAs. Review of Educational  Research, 52, 340-367. Brody, E.B. & Brody, N. (1976). Intelligence: Nature, determinants, and consequences. New York: Academic Press. Brown, F. (1980). Perspectives on v a l i d i t y . Washington, D.C. American Council on Measurement i n Education. 211 Burke, M.E. (1984). Educational implications of c u l t u r a l d i v e r s i t y . In R.J. Samuda, J.W. Berry & M. La f e r r i e r e (Eds.), Multiculturalism i n Canada: S o c i a l and  educational perspectives (pp. 3-17). Toronto: A l l y n & Bacon, Inc. Burnet, J . (1984). Myths and multiculturalism. In R.J. Samuda, J.W. Berry Se M. L a f f e r i e r e (Eds.), Multiculturalism i n Canada: So c i a l and educational  perspectives (pp. 18-29). Toronto: A l l y n Se Bacon, Inc. C a t t e l l , R.B. (1963). Theory of f l u i d and c r y s t a l l i z e d i n t e l l i g e n c e : A c r i t i c a l experiment. Journal of  Educational Psychology, 54, 1-23. C a t t e l l , R.B. (1966). The Scree test for the number of factors. Multivariate Behavioral Research, 1, 245-276. Child, D. (1973). The essentials of factor analysis. New York: Holt, Rinehart Se Winston. Coffman, W.E. (1974, Spring). A moratorium? What kind? NCME:  Measurement i n Education, j i , 1-7. Cronbach, L. (1970). Essentials of psychological t e s t i n g . (3rd ed.). New York: Harper and Row. Cronbach, L. (1971). Test v a l i d a t i o n . In R.L. Thorndike (Ed.), Educational measurements (pp. 443-507). Washington, D.C: American Council of Education. Cummins, J. (1982). Bilingualism and cognitive functioning. In S. Shapson, V. d'Oyley, Se A. Lloyd (Eds.), Bilingualism and multiculturalism i n Canadian education (pp. 39-49). Vancouver, B.C.: University of B r i t i s h Columbia, Centre for the Study of Curriculum and Instruction. Cummins, J . (1984a). Psychological assessment of minority language students. In R.J. Samuda, J.W. Berry Se M. L a f e r r i e r (Eds.), Multiculturalism i n Canada: So c i a l  and educational perspectives (pp". 238-249) . Toronto: A l l y n Se Bacon, Inc. Cummins, J . (1984b). The minority language c h i l d . In S. Shapson Se V. d'Oyley (Eds.), B i l i n g u a l and  mu l t i c u l t u r a l education: Canadian perspectives (pp. 71-92). Clevedon, England: M u l t i l i n g u a l Matters. 212 Cummins, J. (1984c). Bilingualism and cognitive functioning. In S. Shapson & V. d'Oyley (Eds.), B i l i n g u a l and m u l t i c u l t u r a l education: Canadian  perspectives (pp. 55-67). Clevedon, England: Mu l t i l i n g u a l Matters. Das, J.P. (1972). Patterns of cognitive a b i l i t y i n nonretarded and retarded children. American Journal of  Mental Deficiency, 77, 6-12. Das, J.P. (1984a). Test review. Journal of Psychoeducational  Assessment, 2, 83-88. Das, J.P. (1984b). Simultaneous and successive process and K-ABC. Journal of Special Education, 18, 229-238. Das, J.P. & Heemsbergen, D.B. (1981). Planning as a factor i n the assessment of cognitive processes. Journal of  Psychoeducational Assessment, _1, 1-15. Das, J.P. & Jarman, R.F. (1981). Coding and planning processes. In M.P. Friendman, J.P. Das Sc N. O'Connor (Eds.), Intelligence and learning (pp. 297-315). New York: Plenum Publishing Corp. Das, J.P. Sc Jarman, R.F. ( i n press). The essence of human i n t e l l i g e n c e . American Psychologists. Das, J.P., Kirby, J. & Jarman, R.F. (1975). Simultaneous and successive synthesis: An alternative model for cognitive a b i l i t i e s . Psychological B u l l e t i n , 82, 87-103. Das, J.P., Kirby, J.R. Sc Jarman, R.F. (1979). Simultaneous  and successive cognitive processes. New York: Academic Press. Das, J.P. Sc N a g l i e r i , J.A. (1985, A p r i l ) . Simultaneous - successive - planning and arousal: A model for  assessment and intervention. Paper presented at the National Association of School Psychologists Convention, Las Vegas, Nevada. Dixon, W.J., (Ed.), (1981). BMDP s t a t i s t i c a l software. Berkeley: University of C a l i f o r n i a Press. Doppelt, J.E. Sc Kaufman, A.S. (1977). Estimation of the difference between WISC-R and WISC-IQs. Education and  Psychological Measurement, 37, 417-427. 213 Edgerton, R. & Landness, L.L. (1974). Methods and style i n  the study of culture. San Francisco: Chandler & sharp Publishers, Inc. Esquirel, G. (1985). Best practices in the assessment of limited English p r o f i c i e n t and b i l i n g u a l children. In A. Thomas & j . Grimes (Eds.), Best Practices i n School  Psychology (pp. 113-123). Kent, Ohio: The National Association of School Psychologists. Ferguson, G.A. (1981). S t a t i s t i c a l analysis i n psychology  and education. New York: McGraw-Hill, Inc. Ferron, 0. (1973). Family, marital and child-rearing patterns i n d i f f e r e n t ethnic groups. In P. Watson (Ed.), Psychology and race (pp. 330-341). Chicago: Aldine Publishing. Fraser, C. (1980). COSAN: User's guide. Toronto: Ontario Institute for Studies i n Education. Frederikesen, W. (1977). How to t e l l i f a test measures the same thing i n d i f f e r e n t cultures. In Y.H. Poortinga (Ed.), Basic problems i n cr o s s - c u l t u r a l psychology (pp. 14-18). Amsterdam: Swetsland Z e i t l i n g e r , B.U. Freides, D. (1978). Review of WISC-R. In 0-K. Buros (Ed.). Mental measurements year book. New Jersey: The Gryphon Press. Garth, T.R. (1931). Race psychology: A study of r a c i a l  mental differences. New York: Whittlesey House. Goetz, E.T. & H a l l , R.J. (1984). Evaluation of the Kaufman Assessment Battery for Children from an information-processing perspective, journal of Special  Education, 18, 281-296. Goldman, R. (1973). Education and immigrants. In P. Watson (Ed.). Psychology and race (pp. 343-359). Chicago: Aldine Publishing Co. Grover, S. (1981). The cognitive basis of the i n t e l l e c t : A  response to Jensen's "Bias i n Mental Testing". Washington, D.C: University Press of America. Guilford, J.P. (1954). Psychometric methods. (2nd ed.). New York: McGraw-Hill, Inc. Gulliksen, H. (1950). Theory of mental t e s t s . New York: John Wiley & Sons Inc. 214 Gunnison, J . (1984a, J u l y ) . K-ABC - Administration -Translation. Paper presented at the Seventh Annual U.B.C. Education C l i n i c , Summer Symposium, Vancouver, B.C. Gunnison, J.A. (1984b). Developing educational intervention from assessments involving the K-ABC. Journal of  Special Education, 18, 325-343. Guttman, L. (1954). Some necessary conditions for common-factor analysis. Psychometrika, 19, 149-61. Haddad, F. (1983, March). The v a l i d i t y study of the K-ABC with learning disabled children. Paper presented at the convention of the National Association of School Psychologists, Detroit, Michigan. Haddad, F.A., Carey, K., Culver, C , Eckelcamp, M., Parker, M., Schwartz, L., Smith, R. & Webb, P. (1984, A p r i l ) . Learning disabled children's performance on the K-ABC:  r e p l i c a t i o n and extention. Paper presented at the National Association of School Psychologists Convention, Philadelphia, Pennsylvania. Hakstian, A.K. & Bay, K.S. (1973). User's manual to accompany the Alberta General Factor Analysis Program. Edmonton, Alberta. D i v i s i o n of Educational Research Services, University of Alberta. Hardman, S.A. (1984). Relationship between the WISC-R and  the Woodcock-Johnson Achievement Battery, the B.C.  Quiet and teacher rankings. Unpublished master's thesis, University of B r i t i s h Columbia, Vancouver, B.C. Harris, R.J. (1975). A primer of multivariate s t a t i s t i c s . New York: Academic Press. Hessler, G. (1985). Review of the Kaufman Assessment Battery for Children: implications for assessment of the gi f t e d . Journal for the Education of the Gifted, VIII, 131-147. Holmes, B. (1981). Ind i v i d u a l l y administered i n t e l l i g e n c e  t e s t s : An application of anchor test norming and  equating procedures i n B r i t i s h Columbia. Unpublished doctoral d i s s e r t a t i o n , University of B r i t i s h Columbia. Hooper, S.R. & Hynd, G.W. (1985). The relationship between the K-ABC and the VADS with reading-disabled children: 215 V a l i d i t y of the K-ABC. Journal of Psychoeducational  Assessment, 3^  77-87. Irwin, M.f Klein, R.E. & Townsend, J.W. (1982). Indigenous versus Construct V a l i d i t y in Cross-Cultural Research. In L.L. Adler (Ed.), Cross-cultural research at issue (pp. 23-29). New York: Academic Press. Jensen, A.R. (1980). Bias in mental testing. New York: Free Press. Jensen, A.R. (1984). The Black-White difference on the K-ABC: Implications for future tests. Journal of Special Education, 18, 377-408. J i r s a , J.E. (1983). The SOMPA: A brief examination of technical considerations, philosophical rationale, and implications for practice. Journal of School  Psychology, 21, 13-21. JiJreskog, K.G., & Sflrbom, D. (1979). Advances in factor  analysis and st r u c t u r a l equation models. Cambridge, M.A.: Abt. Kaiser, H.F. (1970). A second generation L i t t l e J i f f y . Psychometrika, 35, 401-415. Kamphaus, R.W. & Reynolds, C.R. (1984). Development and structure of the Kaufman Assessment Battery for Children. Journal of Special Education, 18, 213-228. Kaplan, R.J. & Klanderman, J.W. (1984). Neuropsychological p r o f i l e of T.M.H. youngsters assessed with the K-ABC. [Summary]. Proceedings of the 16th Annual Convention of  the National Association of School Psychologists, 31-32. Kaufman, A. (1979). I n t e l l i g e n t testing with the WISC-R. New York: John Wiley & Sons. Kaufman, A.S. (1983). Some questions and answers about the Kaufman Assessment Battery for Children (K-ABC). The  Journal of Psychoeducational Assessment, .1 , 205-218. Kaufman, A.S. (1984). K-ABC and controversy. Journal of Special Education, 18, 410-444. Kaufman, A.S. & Doppelt, J.E. (1976). Analysis of WISC-R standardization data in terms of the s t r a t i f i c a t i o n variables, Child Development, 47, 165-171. ' 216 Kaufman, A. & Kamphaus, R. (1984). Factor analysis of the Kaufman Assessment Battery for Children (K-ABC) for ages 2 1/2 through 12 l/2 years. Journal of Educational  Psychology, 76, 623-637. Kaufman, A.S., & Kaufman, N.L. (1983a). Kaufman Assessment  Battery for Children. C i r c l e Pines, MN: American Guidance Service. Kaufman, A. & Kaufman, N. (1983b). Kaufman Assessment Battery for Children: Interpretation Manual. C i r c l e Pines, MN: American Guidance Service. Kaufman, A. & Kaufman, N. (1983c). Kaufman Assessment  Battery for Children: Administration and scoring  manual. C i r c l e Pines, MN: American Guidance Service. Keith, T.Z. (1985). Questioning the K-ABC: What does i t measure? School Psychology Review, 14, 9-20. Keith, T.Z. & Dunbar, S.B. (1984). Hierarchial factor analysis of the K-ABC: Testing alternate models. Journal of Special Education, 18, 367-375. Keith, T.Z., Hood, C., Eberhart, S. & Pottebaum, S.M. (1985, A p r i l ) . Factor structure of the K-ABC for referred school children. Paper presented at the meeting of the National Association of School Psychologists, Las Vegas, Nevada. Klanderman, J . & Kroeschell, Z. (1984, A p r i l ) . Comparisons  of the K-ABC and WISC-R for LD children. Paper presented at the National Association of School Psychologists Convention, Philadelphia, Pennsylvania. LaTorre, R. (1983). The 1982 survey of pupils i n Vancouver  Schools for whom English i s a second language. (Report No. 83-02). Vancouver, B.C.: Evaluation and Research Services; School Board Trustees. Laosa, L.M. (1977). Nonbiased assessment of children's a b i l i t i e s : H i s t o r i c a l antecedents and current issues. In T. Oakland (Ed.), Psychological and educational  assessment of minority children (pp~. 1-20) . New Yor jL Brunner/Mazel. Lawley, D.N. & Maxwell, A.E. (1963). Factor analysis as a s t a t i s t i c a l method. London: Butterworth. 217 Lesser, G.S., F i f e r , G. & Clark, D.H. (1965). Mental a b i l i t i e s of children from d i f f e r e n t s o c i a l class and cu l t u r a l groups. Monographs of the Society for research  i n c h i l d development, 30, (24). Luria, A.R. (1966). Human brain and psychological  processes• New York: Harper and Row. Luria, A.R. (1973). The working brain. London: Penguin. Lyon, M.A. & Smith, D.K. (1985, A p r i l ) . Referred students'  performance on the K-ABC, WISC-R, and  Woodcock-Johnson. Paper presented at the annual convention of the National Association of School Psychologists, Las Vegas, Nevada. Maccoby, M. & Modiano, N. (1971). On culture and equivalence. In P. Sears (Ed.), I n t e l l e c t u a l  development (pp. 283-301). New York: John Wiley & Sons, Inc. Majovski, L.V. (1984). The K-ABC: Theory and applications for c h i l d neuropsychological assessment and research. Journal of Special Education, 18, 257-265. Marjoribanks, K. (1980). Ethnic families and children's achievements. Sydney, A u s t r a l i a : George A l l e n & Unwin. McCallum, R.S. & Karnes, F.A. (1984). K-ABC, S-B, WISC-R; Test of choice for evaluating g i f t e d children?? [Summary]. Proceedings of the 16th Annual Convention of  the National Association of School Psychologists, 371-372. McCallum, R., Karnes, F., & Edwards, R. (1984). The test of choice for assessment of g i f t e d children: A comparison of the K-ABC, WISC-R, and Stanford-Binet. Journal of  Psychoeducational Assessment, 2, 57-63. McLeod, K.A. (1984). Multiculturalism and mu l t i c u l t u r a l education: Policy and pra c t i c e . In R.J. Samuda, J.W. Berry & Mr. La f e r r i e r e (Eds.), Multiculturalism i n  Canada: So c i a l and educational perspectives (pp. 30-49). Toronto: A l l y n St Bacon, Inc. Mcloughlin, C. St E l l i s o n , C. (1984, A p r i l ) . Comparing preschoolers' scores on the revised Peabody and K-ABC  Achievement. Paper presented at the National Association of School Psychologists Convention, Philadelphia, Pennsylvania. 218 Mehrens, W.A. (1984). A c r i t i c a l analysis of the psychometric properties of the K-ABC. Journal of  Special Education, 18, 297-310. Mercer, J. (1973). Labeling the mentally retarded. Berkeley: University of C a l i f o r n i a Press. Mercer, J . (1979). System of M u l t i c u l t u r a l P l u r a l i s t i c Assessment: Technical manual. New York: Psychological Corporation. Mercer, J.R. (1980). Workbook for the SOMPA workshops. C a l i f o r n i a : I n s t i t u t e for p l u r a l i s t i c assessment, research, t r a i n i n g . M i l l e r , T. (1984). Special Issue: The Kaufman Assessment Battery for Children. Journal of Special Education, 18, 212. Mishra, S.P. (1983). Evidence of item bias i n the verbal subtests of the WISC-R for Mexican-American children. Journal of Psychoeducational Assessment, 1_, 321-328. More, A.J. & Oldridge, B. (1980). An approach to non-discriminatory assessment of Native Indian children. B r i t i s h Columbia Journal of Special  Education, 4 (1), 51-59. Murray, A.M. & Mishra, S.P. (1983). The bias i n i n d i v i d u a l l y adminsistered a b i l i t y t e s t s : Implications for nondiscriminatory assessment. Journal of  Psychoeducational Assessment, 1_, 353-366. Na g l i e r i , J . (1985, A p r i l ) . WISC-R and K-ABC with matched samples of Black and White children. Paper presented at the annual convention of National Association of School Psychologists, Las Vegas, Nevada. Na g l i e r i , J . (in p r e s s a ) . Assessment of mentally retarded children with the Kaufman Assessment Battery for Children. American Journal of Mental Deficiency. N a g l i e r i , J . (in p r e s s D ) . Concurrent and predictive v a l i d i t y of the Kaufman Assessment Battery for Children with a Native American Navajo sample. Journal of School  Psychology. N a g l i e r i , J . & Haddad, F. (1984). Learning disabled children's performance on the Kaufman Assessment 219 Battery for Children: A concurrent v a l i d i t y study. journal of Psychoeducational Assessment, 2, 49-56. Na g l i e r i , J., Kamphaus, R. & Kaufman, A. (1983). The Luria-Das simultaneous-successive model applied to the WISC-R. Journal of Psychoeducational Assessment, _1, 25-34. Nie, N.H. (1983). Users guide for the S t a t i s t i c a l Package of  the Social Sciences (SPSS X). New York: McGraw-Hill. Nunnaly, J.G. (1970). Introduction to psychological measurement. New York: McGraw H i l l Book Company. Obrzut, A., Obrzut, J . & Shaw, D. (1984). Construct v a l i d i t y of the Kaufman Assessment Battery for Children with learning disabled and mentally retarded. Psychology i n  the Schools, 21, 417-424. Parnes, S.J., Noller, R.B. Se Biondi, A. (1977). Guide to creative action: Revised edition of creative behavior  guidebook. New York: Charles Scribner's Sons. Peters, H. (1976). The v a l i d i t y of the Wechsler Intelligence Scale for Children - Revised. Canadian Journal of  Behavioural Science, 8, 414-417. Ponnuswami, S. (1977). Mother c h i l d interactions and cognitive s t y l e . Indian Journal of Psychology, 52, 250-255. Rees, 0. (1983). Ethnic group, bilingualism and attainment. In C. Bagley Se G. Verma (Eds.), M u l t i c u l t u r a l childhood (pp. 38-62). Hampshire, England: Grower Publishing. Reschly, D. (1979). Nonbiased assessment. In G. Phye Se D. Reschly (Eds.), School Psychology: Perspectives and  issues. New York: Academic Press. Reynolds, C.R. (1981). Neuropsychological assessment and the h a b i l i t a t i o n of learning: Considerations in the search for the aptitude X treatment interaction. School  Psychology Review, 10, 343-349. Reynolds, C.R. (1982). The problem of bias i n psychological assessment. In C.R. Reynolds Se T.B. Gutkin (Eds.), The  Handbook of School Psychology (pp. 178-208). New York: John Wiley Se Sons. Robertson, G.J. Se Eisenberg, J . (1981). Peabody Picture Vocabulary Test - Revised technical supplement. C i r c l e Pines, MN: American Guidance Service. 220 Rogoff, B. (1981). Schooling and the development of cog-n i t i v e s k i l l s . In H. Triandis & A. Heron (Eds.), Hand- book of c r o s s - c u l t u r a l psychology: Developmental psych- ology, Volume 4 (295-341). Boston: A l l y n and Bacon, Inc. Saklofske, D.H. & J e d l i c k i , G. (1985, A p r i l ) . Concurrent and  content v a l i d i t y of the K-ABC. Paper presented at the meeting of the National Association of School Psych-olo g i s t s , Las Vegas, Nevada. Salvia, J . & Hritcko, T. (1984). The K-ABC and a b i l i t y t r a i n i n g . Journal of Special Education, 18, 345-356. Salvia, J. & Ysseldyke, J.E. (1981). Assessment i n special  and remedial education (2nd ed.). Boston: Houghton M i f f l i n , Co. Salvia, J . & Ysseldyke, J.E. (1985). Assessment i n special  and remedial education (3rd ed.). Boston: Houghton M i f f l i n , Co. Samuda, R.J. (1983). C l a s s i f y i n g and programming ethnic minority immigrants i n the public schools. In R.J. Samuda and S.L. Woods, Perspectives i n immigrant and  minority education (pp. 175-186). New York: University Press of America. Samuda, R.J. (1984). Assessing the a b i l i t i e s of minority students with a multiethnic milieu. In R.J. Samuda, J.W. Berry & M. Laferniere (Eds.), Multiculturalism i n  Canada: So c i a l educational perspective (pp. 353-367). Toronto: A l l y n & Bacon, Inc. Sat t l e r , j . (1982). Assessment of children's i n t e l l i g e n c e and special a b i l i t i e s (2nd ed.). Boston: A l l y n & Bacon, Inc. S a v i l l e , M.R. 11977). Language and the Disadvantaged. In R.D. Abrahams and R.C. Troike (Eds.), Language and  c u l t u r a l d i v e r s i t y i n American education (pp. 310-318). Englewood C l i f f s , New Jersey: Prentice-Hall, Inc. Schluderman, S. and Schluderman, E. (1977). Achievement motivation: Cross c u l t u r a l and developmental issues. In Y.H. Poortinga (Ed.), Basic problems i n cros s - c u l t u r a l  psychology (pp. 149-159). Amsterdam: Swets and Zeitlingen B.V. Sewell, T. (1983). New thinking cast i n an old mold: The Kaufman Assessment Battery for Children. Contemporary  Education Review, 2, 199-207. 221 Shade, B. (1981). R a t i a l v a r i a t i o n i n perceptual d i f f e r e n t i a t i o n , Perceptual and Motor S k i l l s , 52, 243-248. Shepard, L.A., C a m i l l i , G., & A v e r i l l , M. (1981). Comparison of six procedures for detecting test item bias using both i n t e r n a l and external a b i l i t y c r i t e r i a . Journal of  Educational S t a t i s t i c s , 6, 317-375. Siegel, P.J. & Piotrowski, R.J. (1985). R e l i a b i l i t y of K-ABC subtest composites. Journal of Psychoeducational  Assessment, 3^, 73-76. Sinha, D. (1977). Some s o c i a l disadvantages and development of ce r t a i n perceptual s k i l l s . Indian Journal of  Psychology, 52, 115-132. Snyder, T., Leask, R. & A l l i s o n , R. (1983, March). Correlation of the K-ABC, WISC-R and Lur i a - Nebraska  Children's Battery for exceptional Children. Paper presented at the annual meeting of the National Association of School Psychologists, Detroit, Michigan. Sternberg, R.J. (1979, September). Stalking the IQ quark. Psychology Today, pp. 42-54. Sternberg, R.J. (1983). Should K come before A, B, C? A review of the Kaufman Assessment Battery for Children (K-ABC). Contemporary Education Review, 2, (3), 199-207. Sternberg, R.J. (1984). The Kaufman Assessment Battery for Children: An information processing analysis and c r i t i q u e . Journal of Special Education, 18, 269-279. Stevens, J.P. (1972). Four methods of analyzing between va r i a t i o n for the K-Group MANOVA problem. Multivariate  Behavioral Research, 1_, 499-522. Stoiber, K., Bracken, B. & G i s s a l , T. (1983). Cognitive processing styles i n reading-disabled and a matched group of normal children. Journal of Psychoeducational  Assessment, 1, 219-233. Tabachnick, B.G. & Fidel1, L.S. (1983). Using multivariate  s t a t i s t i c s . New York: Harper and Row, Publishers. Telzrow, G. (1984). P r a c t i c a l applications of the K-ABC i n the i d e n t i f i c a t i o n of handicapped preschoolers. Journal  of Special Education, 18, 311-324. 222 Terman, L.M. & M e r r i l l , M.A. (1973). Stanford-Binet  Intelligence Scale: 1972 norms edition. Boston: Houghton M i f f l i n . Thomas, A. (1984, A p r i l ) . Adaptions of the K-ABC. Paper presented at the convention of the National Association of School Psychologists, Philadelphia, Pennsylvania. Thorndike, R.L. (1977). Causation of Binet IQ decrements. Journal of Educational Measurement, 14, 197-202. Timm, N.H. (1975). Multivariate analysis: With applications  i n education and psychology. Monterey, C a l i f o r n a i : Brooks/Cole. Valencia, R.R. (1984a). Concurrent v a l i d i t y of the Kaufman Assessment Battery for Children i n a sample of Mexican-American Children. Educational and  Psychological Measurement, 1984, 44, 365-372. Valencia, R.R. (1984b, A p r i l ) . Psychometric properties of  the Kaufman Assessment Battery for Children: V a l i d i t y  research with Mexican American children, New Orleans, LA. Vancouver Enumeration Data (1984). Vancouver, B.C. Canada. Van der F l i e r , H. (1977). Environmental factors and deviant response patterns. In Y.H. Poortinga (Ed.), Basic  problems i n cr o s s - c u l t u r a l psychology, (pp. 30-35). Amsterdam: Swets and Z e i t l i n g e r BV. Vernon, P.E. (1974). WISC-R. Canadian Psychological  Association B u l l e t i n , 4 (4), 8-9. Vernon, P.E. (1975). Intelligence across cultures. I. G. Verma and C. Bagley (Eds.), Race and education across  culture (pp. 7-29). London: Heinemann Educational Books. Vernon, P.E. (1977). F i n a l report on modifications of WISC-R for Canadian use. Canadian Psychological Association  B u l l e t i n , 5_, 5-7. Vernon, P.E. (1983). Recent findings on the nature of g. Journal of Special Education, 17, 389-400. Vernon, P.E. (1984). A b i l i t i e s and achievements of ethnic groups i n Canada with special reference to Canadian 223 natives and o r i e n t a l s . In R. Samuda, j . Berry & M. Lafe r r i e r e (Eds.). Multiculturalism i n Canada: Social  and educational perspectives (pp~i 382-395) . Toronto: Al l y n & Bacon, Inc.) Violato, C. (1984). Effects of Canadianization of American-biased items on the WAIS and WAIS-R Information subtests. Canadian Journal of Behavioural Science, 16, 36-41. Vyas, H. (1983). Education and cognitive s t y l e s : A case study of Guj e r t i children i n B r i t a i n , Eastern United States and India. In C. Bagley and G. Verma (Eds.), M u l t i c u l t u r a l childhood (pp. 38-62). Hampshire, England: Grower Publishing. Wechsler, D. (1974). Manual for the Wechsler Intelligence  Scale for Children - Revised (WISC-R). New York: Psychological Corporation. Williams, R.L. (1970). From dehumanization to black i n t e l l e c t u a l genocide. A rejoinder. C l i n i c a l C hild  Psychology Newsletter, 9 (3), 6-7. Williams, R.L. (1972, September). The BITCH-100: A cu l t u r e - s p e c i f i c t e s t . Paper presented at the annual meeting of the American Psychological Association, Honolulu. Woodcock, R. & Johnson, B. (1977). Woodcock-Johnson psycho- educational battery. Boston: Teaching Resources Corporation. Wormeli, C T . (1984). Development of the B r i t i s h Columbia Quick Individual Educational Test. Unpublished doctoral d i s s e r t a t i o n , University of B r i t i s h Columbia. Wright, M.A. & Harris, M.W. (1972). The Canadian Lorge Thorndike Intelligence Tests. Technical Supplement. Toronto: Thomas Nelson. Zucker, S. (1985). The MSCA and the K-ABC with "high Risk" preschoolers. [Summary]. Proceedings of the 17th Annual  Convention of the National Association of School  Psychologists, 52-53. 224 APPENDIX A Instruments A - l . Parent Questionnaire p. 225 A-2. Teacher Questionnaire p. 229 A-3. Teacher Rating Scale p. 230 225 Code NumDer (For Office Use PARENT QUESTIONNAIRE Please answer the following questions about THIS CHILD , your family, and yourself. Some questions require you to PUT AN X beside the correct answer(s). Other questions require you to FILL IN the answer(s). Remember that the completion of this questionnaire Is voluntary and you may choose not to answer any particular question(s). PUT AN X 1. What language(s) do adults speak in the home? _______ Cantonese English ________ Punjabi other (specify: )• 2. What language^) does THIS CHILD speak in the home? _ _ _ _ _ Cantonese English Punjabi _ other (specify: ) 3. How often do adults speak English in the home? _______ always ______ three quarters of the time ______ half of the time _ _ _ _ _ one quarter of the time never 4. How often does THIS CHILD speak English in the home? always three quarters of the time half of the time _ one quarter of the time never FILL IN 5. Where was THIS CHILD born? city/town country 6. How many years has THIS CHILD lived in Vancouver? years 226 Including THIS CHILD, please list the first name of aU_ the children living in the home, their sex, and their age. NAME SEX (male/female) AGE The following questions seek information on THIS CHILD'S Mother (or female guardian) and Father (or male guardian). FILL IN 8. Where were you born? MOTHER FATHER town/city • country PUT AN X 9. How large was the place where you were born? MOTHER FATHER large city (over 500,000 people) small city small town (less than 20,000 people) farm or rural area 10. How long have you lived in Canada? MOTHER FATHER years 11. How long have you lived in Vancouver? MOTHER FATHER 227 12. Where were THIS CHILD'S grandmothers born? MOTHER'S side FATHER 'S side country 13. Where were THIS CHILD'S grandfathers born? MOTHER'S side FATHER 'S side _ . country PUT AN X 14. What is the highest level of formal education you have completed? FATHER _ _ _ _ _ _ no formal schooling some elementary schooling _ _ _ _ _ _ finished elementary school some high school finished high school _ _ _ _ _ _ some college or technical school _ _ _ _ _ _ finished university undergraduate degree _ _ _ _ _ _ some postgraduate training _ _ _ _ _ _ finished postgraduate degree FILL IN 15. In which country did you complete your highet level of education? MOTHER FATHER . country PUT AN X 16. What is your present employment status? FATHER employed full time employed part time retired student other (specify: ) MOTHER MOTHER 229 1 e a c u c i c u u c tr Child code # TEACHER QUESTIONNAIRE Please ask 's teacher the following questions: A. Read all statements, then put an X beside the description which the teacher feels best represents the student's facility in the following four areas: Understanding Spoken English 1. Understands no English. 2* Understands simple English. 3. Understands most ordinary conversation but has problems with unfamiliar subject matter. 4. Understands most of what goes on in classroom except for complex subject matter. 5. Understands everything that goes on in the classroom. Speaking English 1. Speaks no English. 2. ~ ~ Communicates basic needs but cannot sustain a conversation. 3. Can converse but still produces errors in pronunciation and structure. ~ ~* ~ Takes part in discussion with many structural and pronunciation errors. 5. Can take part in all discussions like a native speaker. Errors made are age acceptable. Reading English 1. Reads no English. 2. ~ _ Reads familiar simple statements. 3. Can read on his/her own simple stories at about grade i ievei. 4. ' Can read material at about a grade 2 level. 5. Can read at a level acceptable for a native speaker of the same grade level. Writing English (written language) 1. Writes no English. 2. Can take dictation on familiar words or generate simple sentences. 3. Can write a simple paragraph which might not always be comprehensible. 4. Can write longer paragraphs and show some creativity.. Work has errors but generally it is comprehensible. 5. Can write at a level acceptable for a native speaker of same age. B. What type and amount of remedial assistance is THIS CHILD receiving outside of his/her regular classroom? Type . Amount (hours per week) no remedial assistance English instruction (speaking and understanding) hrs/wk. academic instruction - reading hrs/wk. academic instruction - written language "hrs/wk. academic instruction - arithmetic hrs/wk. other (specify: ) hrs/wk. Thank you for your cooperation. 230 TEACHER RATING SCALE STUDENT'S NAME TEACHER NUMBER Using the rating scale provided, please put a single X within the parenthesis which best describes the ability level of the student whose name appears above. 1. What is this student's ability to master new material? ( ) ( ) ( ) ( ) ( Poor Below Average Above Superior Average Average 2. What is this student's ability to concentrate on a task? ( ) ( ) ( ) ( ) ( ) Poor Below Average Above Superior Average Average 3 . What is this student's ability to retain material taught? ( ) ( ) ( ) ( ) ( J Poor Below Average Above Superior Average Average 4. What is this student's ability to persevere at completing a task? ( ) ( ) ( ) ( ) ( ) Poor Below. Average Above Superior Average Average 5. What is this student's ability to plan and organize his/her time? ( ) ( ) ( ) ( ) ( ) Poor Below Average Above Superior Average Average THANK YOU for taking the time to complete this Rating Scale. Feel free to add any comments or qualifications below. 231 APPENDIX B Letters to Principals and Teachers B - l . P r i n c i p a l Information Letter p. 232 B-2. P r i n c i p a l Thank-You Letter p. 234 B-3. Teacher Thank-You Letter p. 235 THE UNIVERSITY OF BRITISH COLUMBIA FACULTY OF EDUCATION 2125 MAIN MALL UNIVERSITY CAMPUS VANCOUVER, B.C., CANADA V6T 1Z5 D e a r I am a d o c t o r a l s t u d e n t i n E d u c a t i o n a l P s y c h o l o g y a t t h e U n i v e r s i t y o f B r i t i s h C o l u m b i a . I am c o n d u c t i n g a s t u d y "An A n a l y s i s o f t h e V a l i d i t y o f t h e Kaufman A s s e s s m e n t B a t t e r y f o r C h i l d r e n ( K - A B C ) w i t h a S a m p l e o f C a n t o n e s e , E n g l i s h a n d P u n j a b i S p e a k i n g C a n a d i a n s " i n V a n c o u v e r . The p u r p o s e o f t h i s s t u d y i s t o i n v e s t i g a t e t h e p e r f o r m a n c e o f t h r e e l a n g u a g e g r o u p s ( v i z , C a n t o n e s e , E n g l i s h , a n d P u n j a b i ) o f V a n c o u v e r c h i l d r e n on two i n t e l l i g e n c e t e s t s ; n a m e l y , t h e W e c h s l e r I n t e l l i g e n c e S c a l e f o r C h i l d r e n -R e v i s e d ( W I S C - R ) a n d t h e n e w l y d e v e l o p e d Kaufman A s s e s s m e n t B a t t e r y f o r C h i l d r e n ( K - A B C ) . I h a v e b e e n g r a n t e d a c c e s s t o V a n c o u v e r S c h o o l s by t h e V a n c o u v e r S c h o o l B o a r d R e s e a r c h D i r e c t o r . The p r o j e c t i s f u n d e d by a M a j o r G r a n t f r o m t h e E d u c a t i o n a l R e s e a r c h I n s t i t u e o f B r i t i s h C o l u m b i a . The 210 (.70 i n e a c h l a n g u a g e g r o u p ) n o n i m m i g r a n t , t h i r d g r a d e r s a t t e n d i n g s c h o o l i n V a n c o u v e r a r e i n d i v i d u a l l y a d m i n i s t e r e d b o t h t h e WISC-R a n d t h e K-ABC on two s e p a r a t e , o n e h o u r t e s t i n g s e s s i o n s . A t p r e s e n t t e s t i n g i s u n d e r w a y a n d I h a v e s e c u r e d 70 p e r c e n t o f t h e s u b j e c t s f o r t h i s s t u d y . W h i l e a f e w C a n t o n e s e a n d E n g l i s h s p e a k i n g c h i l d r e n a r e s t i l l r e q u i r e d , t h e m a j o r i t y o f t h e s u b j e c t s n e e d e d t o c o m p l e t e t h i s s t u d y m u s t be P u n j a b i s p e a k -i n g . T h e r e f o r e , I am a p p r o a c h i n g a l l o f t h e p r i n c i p a l s i d e n t i f i e d a s h a v i n g t h e s e c h i l d r e n a t t e n d i n g t h e i r s c h o o l s t o a s k f o r t h e i r c o o p e r a t i o n i n t h e c o m p l e t i o n o f t h e s t u d y . I t i s a n t i c i p a t e d t h a t t h e t e s t i n g o f t h e s e c h i l d r e n w i l l be i n t h e m o n t h o f A p r i l . I w o u l d a p p r e c i a t e t h e o p p o r t u n i t y t o come t o y o u r s c h o o l a n d d i s c u s s t h e s t u d y w i t h y o u a n d t h e g r a d e t h r e e t e a c h e r s . T h i s s t u d y . h a s t h e p o t e n t i a l o f p r o v i d i n g s c h o o l p s y c h o l o g i s t s a n d e d u c a t o r s i n B . C . w i t h i n s i g h t i n t o t h e p o s s i b l e t e s t b i a s t h a t may o c c u r when a s s e s s i n g E n g l i s h a s a s e c o n d l a n g u a g e ( E S L ) a n d E n g l i s h a s a F i r s t l a n g u a g e ( E F L ) B . C . s t u d e n t s , p o s s i b l e e x p l a n a t i o n s f o r a n y d i f f e r e n c e s w h i c h m i g h t o c c u r among t h e g r o u p s , i n f o r m a t i o n o n how i n d i v i d u a l s a n d c u l t u r a l g r o u p s p r o c e s s c o g n i t i v e i n f o r m a t i o n ( v i z . s i m u l t a n t e o u s v e r s u s s u c c e s s i v e ) , a n d d i r e c t i o n f o r f u t u r e r e s e a r c h i n t h e a r e a o f a s s e s s m e n t a n d p r o g r a m d e v e l o p m e n t f o r ESL a n d EFL c h i l d r e n i n t h e p r o v i n c e . . / 2 236 APPENDIX C Parent Information Packages C - l . English Package p. 237 C-2. Cantonese Package p. 244 C-3. Punjabi Package p. 252 THE UNIVERSITY OF BRITISH COLUMBIA 23 7 F A C U L T Y O F E D U C A T I O N I U S M A I N M A L L U N I V E R S I T Y C A M P U S VANCOUVER, B.C.. CANADA V6T 1Z5 _ _s School has agreed to participate in a research projecTlnvolvfng the~use of intelligence tests for children who speak English as their first language as well as children who speak English as their second language. This project has been titled "An Analysis of the Validity of the Kaufman Assessment Battery for Children with a Sample of Cantonese, English, and Punjabi speaking Canadians". The project requires the cooperation of 210 children in Vancouver tc take a series of two tests, one of which is presently being used in the Vancouver school system. The second test is a recently developed instrument for assessing the ahievement and intelligence of English speaking and non-English speaking children. These two tests were developed for children in the United States and have never been checked for their applicability to children in Vancouver. The research project is being undertaken as a doctoral dissertation in the Department of Educational Psychology and Special Education at the University of British Columbia. It has been endorsed by the superintendent of this school district and by the principal of your school. [s name was randomly drawn as a possible participant in this research. If you and your child agree to participate, will be asked to take part in two testing sessions, each approximately one hour long. The testing will be done individually by a trained U.B.C. psychometrician and conducted in your child's school. This type of testing is common practice in schools and is usually experienced as interesting and enjoyable by the children involved. Your child's name will not appear on the test forms which will be returned to U.B .C. for scoring. The Vancouver School Board requested that these tests be kept in a confidential file to be used only if your child is referred to a school psychologist for an assessment. The purpose is not to check any one child's performance, but to determine the cultural fairness of the intelligence tests being used in Vancouver for three first language groups (viz. Cantonese, English, and Punjabi). Group profiles will in the future provide educators with possible instructional suggestions for improving the education of children in these groups. In turn, we request that you, as parents, complete the enclosed questionnaire form. The answers to the questions on this form will provide information on the similarities and differences among the groups of children in the study. For your convenience, if your child is reported to speak a first language other than English, a questionnaire in this language has also been enclosed. You can complete the questionnaire in the language of your choice. This questionnaire will be returned directly to U.B.C. and the information will be strictly confidential. 7 239 PARENT CONSENT FORM I consent to 's participation in the testing research study at School. I am aware that this will involve two testing sessions of approximately one hour each, and that the test will be returned anonymously to the University of British Columbia for scoring. I understand that the test results will be kept in a confidential file. Also, I understand that participation in this project is voluntary and it may be terminated at any time. In addition, I will complete the enclosed questionnaire and return it with this consent form. Please tear and send lower portion of this consent form. Thank you. CODE NUMBER I consent to have involved in the testing research study. Signature I am not willing to have research study. involved in the testing S igna tu re 240 Code Numoer For Office Use PARENT QUESTIONNAIRE Please answer the following questions about THIS CHILD , your family, and yourself. Some questions require you to PUT AN X beside "The correct answer(s). Other questions require you to FILL IN the answer(s). Remember that the completion of this questionnaire is voluntary and you may choose not to answer any particular question(s). PUT AN X 1. What language^) do adults speak in the home? _______ Cantonese English Punjabi other (specify: ) 2 . What language(s) does THIS CHILD speak in the home? _ _ _ _ _ Cantonese English _ Punjabi ______ other (specify: ) 3. How often do adults speak English in the home? always three quarters of the time _ _ _ _ _ half of the time one quarter of the time never 4. How often does THIS CHILD speak English in the home? always three quarters of the time ______ half of the time one quarter of the time never FILL IN 5. Where was THIS CHILD born? city/ town country 6. How many years has THIS CHILD lived in Vancouver^ years 7. Including THIS CHILD, please list the first name of all the children living in the home, their sex, and their age. 241 SEX (male/female) AGE NAME The following questions seek information on THIS CHILD'S Mother (or female guardian) and Father (or male guardian). FILL IN 8. Where were you born? MOTHER FATHER town/city - ' country PUT AN X 9. How large was the place where you were born? MOTHER FATHER large city (over 500,000 people) small city small town (less than 20,000 people) farm or rural area .10. How long have you lived in Canada? MOTHER FATHER years 11. How long have you lived in Vancouver? MOTHER FATHER 12. Where were THIS CHILD'S grandmothers born? MOTHER'S side FATHER 'S side country 13. Where were THIS CHILD'S grandfathers born? MOTHER'S side FATHER 'S side country PUT AN X 14. What is the highest level of formal education you have completed? MOTHER FATHER no formal schooling some elementary schooling finished elementary school some high school finished high school some college or technical school finished university undergraduate degree some postgraduate training finished postgraduate degree FILL IN 15. In which country did you complete your highet level of education? MOTHER FATHER country PUT AN X 16. What is your present employment status? MOTHER FATHER . employed full time employed part time retired student , other (specify: ) THE UNIVERSITY OF BRITISH COLUMBIA F A C U L T Y OF E D U C A T I O N 2125 M A I N M A L L UNIVERSITY C A M P U S VANCOUVER, B.C., CANADA V6T 1Z5 245 _ _ _ _ _ * 4 £ | ,1 ?_t :>f t& H J 4 - ^ . *itH&ftn&%*ltk.JS - * tf Hi ^ -if f i % & * * ^ * i - t l ^&t,h *i*t*Zrs\ w 4 H) 4 4 ^ T 4 ^ 14 ^ £ *s 2 249 250 THE UNIVERSITY OF BRITISH COLUMBIA FACULTY OF EDUCATION 2125 MAIN MALL q , UNIVERSITY CAMPUS ^ VANCOUVER, B.C.. CANADA V6T 1Z5 rTr "?n_f> "^sil of)_f) "3T-B* ) 253 - Z -ft ^ © v ^ r ^ r * l 3 - ^ ^TT f k r r 4 I j W r 5 T r W , ^ i ( ^ - ^ ^ 0 -PT£T -TTT^- _TT - £ W S^fif***7 -fe w ^ i i ' f W = ^ - a > o 3 » r - f k f W f -Jf-/ 9snr s £ r " ^ r r^r - 5 5 ^ - yjf ]^ zf- 3 " ^ J " ^ J J T / f e fWUtr ^ r r r - - c j _ft-*ft f Srftfsr ^f%prm / ^ - i S - t J «UMK -Jvn3=-T-fefc»*n- ^r&- s i r S T -5Ts£> ^ •pT***' S T ^ " | H ^ P A I V ^ / ^ & a V3"T J"/' v j 7 ~QTD TJ3~ ^SSi " 3 T 3 W ^ -&rk~d?F 255 -J?T _ f ^ > r f T <s'' f f e r feirr ^<s^d -cT^r -r _____ ^ f f X 9-53" - * £ T W -3"6T - _ T | /£r7^Tj ^ 3 * i & j fin. n & r •~TJT 17". 37? 3 $ ^ fU ^ fe /TJTFT fe M ^ ^ ^ $ - * r - f e ^ - ^ T ^ c f e ~• — \ 256 3*3 A^!a ?_Ut "=?3_? $5i 4^ zro fe-. -rvs> 4 -ap ^ v$ I- " g o v . U-Q ^ 8 3 ) ^ o ^ - " H f ? ^ 3 V — — — : — } — ^ a i ^ e r a ^-o ^a.V« - w a i W ] - I 5-— &3 6 ^ >iT?, {a 257 <- as W J L J , ' ) ^ -0(153 ^ J E R < J 0 S J _ . ^ T ' fe> a i ^ r ^hj n l>$" ^a-a fa ? \ (O- ^ 4 WW ^ ^ \ "5? - a i a s r ? * ^ Sura "Si \ . ^ x proa 5?£f S N 0 260 APPENDIX D Outline of K-ABC Training Workshop D-1. K-ABC Training Workshop p. 261 K-ABC Training Workshop 261 Day Content Time (minutes) 1 Introduction to Research Study 15 Overview of K-ABC's Development 30 Preview of K-ABC subtests 45 Break 15 Video of K-ABC's administration 90 Scoring of K-ABC items 30 Questions Assignment: Administer 1 K-ABC and bring to next session 2 Discuss problems in administering K-ABC 30 Review administration of K-ABC for 8 to 10 year olds 60 Break 15 Practice scoring K-ABC protocol 45 Questions Assignment: Administer and score at least 1 K-ABC and show me tne protocol. 262 APPENDIX E Item Changes E - l . K-ABC p. 263 E-2. WISC-R p. 264 263' K-ABC A r i t h m e t i c I terns  Standard V e r s i o n 28. "This big elephant ( p o i n t to l a r g e s t one) weighs 650 pounds, and t h i s small one ( p o i n t to s m a l l e s t elephant) weighs 550 pounds. How much more does the big one weigh than the small one?" 29. "How much do the big and small elephants weigh together? Remember, the big one weighs 650 pounds and the small one weighs 550 pounds." (Kaufman & Kaufman, 1983c, A r i t h m e t i c Subtest) M e t r i c V e r s i o n 28. T h i s b i g elephant ( p o i n t to l a r g e s t one) weighs 650 kilograms, and t h i s small one ( p o i n t to s m a l l e s t elephant) weighs 550 kilograms. How much more does the big one weigh than the small one? 29. How much do the big and small elephants weigh together? Remember, the big one weighs 650 kilograms and the small one weighs 550 kilograms. 264 WISC-R Standard V e r s i o n Information #20 "How many pounds make a ton?" #24 "How t a l l i s the average American man?" #27 "How f a r i s i t from New York to Los Angeles?" (Wechsler, 1974b, p. 68) S i m i l a r i t i e s #10 In what way are a pound and a yard a l i k e ? (Wechsler, 1974b, p. 74) Me t r i c V e r s i o n Information #20 How many kilograms make a tonne? #24 How t a l l i s the average Canadian man? #27 Question same as above - answer accepted i n me t r i c form. S i m i l a r i t i e s #10 Question read f i r s t as w r i t t e n above. If c h i l d does not give c o r r e c t answer - In what ways are a kilogram and a metre a l i k e ? 265 APPENDIX F Material contained within Tester's Package F - l . Testing Procedure Sheet p. 266 F-2. Cover Sheet p. 270 F-3. Request for Subject P a r t i c i p a t i o n p. 271 F-4. Parent Consent Form p. 272 F-5. Checklist p. 273 F-6. Teacher Questionnaire p. 274 F-7. Teacher Rating Scale p. 275 F-8. Parent Letter p. 276 266 Testing Procedure Sheet Testing Package Cover Sheet A. The school, p r i n c i p a l , teacher, and subjects are i d e n t i f i f e d on this page. B. Special Considerations outlined by the p r i n c i p a l and teacher are provided. Subject P a r t i c i p a t i o n Request Form To be read to each subject. Individual Subject Testing Packages contain: A. Subject's signed consent form, B. Examiner's Checklist with space at bottom for writing comments, C. Teacher Questionnaire, D. 1 K-ABC protocol,* E. 1 WISC-R protocol,* and F. Thank you l e t t e r to be given to c h i l d to take home. * Test protocols are numbered in the order they are to be administered. TESTING PROCEDURES Phone P r i n c i p a l A. Introduce yourself as the tester for UBC involved with JoAnne Gardner's study on testing Cantonese, 267 E n g l i s h , and Punjabi speaking t h i r d graders. I w i l l nave p r e v i o u s l y met with the p r i n c i p a l and teachers to d i s c u s s the study and to inform them of the procedures. B. Ask p r i n c i p a l and teacher when you can begin t e s t i n g . 2. A r r i v a l at School A. Introduce y o u r s e l f to the s e c r e t a r y , p r i n c i p a l and teacher. Ask them what procedures you are to f o l l o w . For example, — what days are not convenient f o r t e s t i n g , — how to remove c h i l d r e n from a c l a s s , and — where to t e s t . 3. T e s t i n g of C h i l d r e n A. Complete " C h e c k l i s t " . B. The WISC-R and K-ABC are to be administered i n the order s p e c i f i e d on the upper r i g h t hand corner of the p r o t o c o l s . C. Only one t e s t i s to be administered to each c h i l d per day. D. The second t e s t i s to be administered w i t h i n a week of the f i r s t . 268 E. Give the teacher as much c o n t r o l , as p o s s i b l e , over who i s to be removed from the c l a s s at a given time. F. Ask the c h i l d r e n not to t e l l h i s or her classmates the q u e s t i o n s asked. G. Return the completed p r o t o c o l s to me as soon as p o s s i b l e . There w i l l be a box i n the c l i n i c . H. When f i n i s h e d , thank the sc h o o l p e r s o n n e l . They w i l l a l s o r e c e i v e a l e t t e r from me. A d m i n i s t r a t i o n of K-ABC. A. Administer as i n s t r u c t e d i n workshop. B. Review s t a r t i n g and stopping procedures and teaching procedures. C. Make m e t r i c changes. D. Write down c h i l d ' s responses. E. Score items o n l y . F. S p a t i a l Memory subt e s t s c o r i n g sheets are in k i t as are p e n c i l s and e x t r a p r o t o c o l s . G. The k i t should be kept with the s e c r e t a r y i n the o f f i c e . H. You should have your own stopwatch. A d m i n i s t r a t i o n of WISC-R. A. Administer a l l 12 s u b t e s t s . 269 B. Make m e t r i c changes. C. Write down c h i l d ' s responses. D. Score items only. E. The k i t w i l l be kept with the s e c r e t a r y i n the o f f i c e. F. You should have your own stopwatch. 270 COVER SHEET Tester: School: P r i n c i p a l : Teacher: Subjects Birthdates WISC-R K-ABC Special Considerations: THE UNIVERSITY OF BRITISH COLUMBIA FACULTY OF EDUCATION 271 2125 MAIN MALL UNIVERSITY CAMPUS VANCOUVER, B.C., CANADA V6T 1Z5 REQUEST FOR SUBJECT PARTICIPATION Request for Subject Participation to be read to each subject individually prior to testing. , as you may know by now, you have been selected to take part in a research project to see how children in British Columbia answer questions on some tests. You were chosen partly because we need children your age, and partly because we need children who can speak (Cantonese/English/Punjabi). Altogether there will be 210 children in Vancouver doing the same tests that you will do. When we finish I will send these papers with your work to U.B.C. Your name will not be on them so nobody will know it was you - we only want to see how children answer the questions, okay? I want you to remember that these tests have nothing to do with your school work and will not count for your grades on your report card. Most children enjoy doing the tests and I am sure you will too. Before we start, I want you to know that you do not have to do this, but that your help is important for a lot of children in Vancouver. I would appreciate it if you would agree to work on these tests with me. Okay? 272 PARENT CONSENT FORM I consent to 's participation in the testing research study at School. I am aware that this will involve two testing sessions of approximately one hour each, and that the test will be returned anonymously to the University of British Columbia for scoring. I understand that the test results will be kept in a confidential file. Also, I understand that participation in this project is voluntary and it may be terminated at any time. In addition, I will complete the enclosed questionnaire and return it with this consent form. Please tear and send lower portion of this consent form. Thank you. CODE NUMBER I consent to have involved in the testing research study. Signature I am not willing to have research study. involved in the testing Signature 273 CODE NUMBER CHECKLIST 1. Did you read the Subject Participation Form? 2. Did you check the subject's birthdate? 3. Did you write the testing date on both protocols? 4. Did you write YOUR name on both protocols? 5. Did you write down the child's response to the rapport questions below? Rapport Questions a) What is your favourite T.V. show? b) If you could be any famous person, who would you like to be? (clarify i f unsure) c) What do you like most about (person's name)? (write adjectives in order) 6. Did teacher complete the Teacher Questionnaire and Teacher Rating Scale? 7. Did you complete the WISC-R? 8. Did you complete the K-ABC? 9. After completing both tests did you give the child the thank-you letter to take home to his/her parents? 10. Write below any PROBLEMS AND CONCERNS you have with the tests and/or procedures. 1 eacuei cuuc * Child code # TEACHER QUESTIONNAIRE 274 Please ask 's teacher the following questions: A. Read all statements, then put an X beside the description which the teacher feels best represents the student's facility in the following four areas: Understanding Spoken English 1. Understands no English. 2. Understands simple English. 3. ~ Understands most ordinary conversation but has problems with unfamiliar subject matter. 4. Understands most of what goes on in classroom except for complex subject matter. 5. Understands everything that goes on in the classroom. Speaking English Speaks no English. Communicates basic needs but cannot sustain a. conversation. Can converse but still produces errors in pronunciation and structure. Takes part in discussion with many structural and pronunciation errors. Can take part in all discussions like a native speaker. Errors made are age acceptable. Reading English 1. Reads no English. 2. Reads familiar simple statements. 3. Can read on his/her own simple stories at about grade 1 level. 4. ~ Can read material at about a grade 2 level. 5. Can read at a level acceptable for a native speaker of che same grade level. Writing English (written language) 1. Writes no English. 2. Can take dictation on familiar words or generate simple sentences. 3. Can write a simple paragraph which might not always be comprehensible. 4. Can write longer paragraphs and show some creativity. Work has errors but generally it is comprehensible. 5. Can write at a level acceptable for a native speaker of same age. B. What type and amount of remedial assistance is THIS CHILD receiving outside of his/her regular classroom? Type no remedial assistance English instruction (speaking and understanding) academic instruction - reading academic instruction - written language academic instruction - arithmetic ~ other (specify: ) i . 2. 3. 4. 5. Amount (hours per week) hrs/wk. hrs/wk. hrs/wk. ~hrs/wk. ~hrs/wk. Thank you for your cooperation. 275 TEACHER RATING SCALE STUDENT'S NAME TEACHER NUMBER Using the rating scale provided, ploase put a single X within the parenthesis which best describes the ability level of the student whose name appears above. 1. What is this student's ability to master new material? ( ) ( ) ( ) ( ) ( ) Poor Below Average Above Superior Average Average 2. What is this student's ability to concentrate on a task? ( ) ( ) ( ) ( ) ( ) Poor Below Average Above Superior Average Average 3. What is this student's ability to retain material taught? ( ) ( ) ( ) ( ) ( _) Poor Below Average Above Superior Average Average 4. What is this student's ability to persevere at completing a task? ( ) ( ) ( ) ( ) ( ) Poor Below Average Above Superior Average Average 5. What is this student's ability to plan and organize his/her time? { ) ( ) ( ) ( ) ( _) Poor Below Average Above Superior Average Average THANK YOU for taking the time to complete this Rating Scale. Feel free to add any comments or qualifications below. 277 APPENDIX G Test Order Eff e c t G-l . K-ABC p. 278 G-2. WISC-R p. 279 Table G-1 Test order effect on K-ABC subtests for each language group Group Source Multivariate Univariate F values F (df) prob (df) HM WD GC TR m SM PS FP AM RI RD RU Cantonese order .70 (13) .76 (1) .04 .78 .08 1.10 .01 .04 .81 .89 2.33* .02 .25 .32 1.58* within (56) (68) English order 1.49 (13) .15 (1) .22 .16 .01 1.39* .05 .11 .03 1.79* .26 1.8 .23 .20 2.6* within (56) (68) Punjabi order 1.83 (13) .06 (1) 2.39* 1.14 .82 .70 .17 1.64* .38 .45 .11 .12 .49 6.90* .07 within (56) (68) *£< .25 00 Table G-2 Test order effect on WISC-R subtests for each language group Group Source Multivariate Univariate F values F (df) prob (df) IN SI AR VO CO DS PC PA EO OA CD MZ Cantonese order 1.52 (12) .14 (1) 1.38* .03 6.40*2.01* .54 .54 .45 .12 .07 .41 2.02* .85 within (57) (68) English order 1.41 (12) .19 (1) 2.83* .97 1.36* .21 .01 .78 2.04* 1.01 .39 .74 .22 .34 within (57) (68) Punjabi order 1.03 (12) .44 (1) .00 .41 .02 .38 1.21 .03 .32 2.54* 3.52* 2.73* .94 1.04 within (57) (68) * £ < .25 APPENDIX H 280 Group Differences on K-ABC H-l. Analysis of variance p. 281 281 Table H-l Group differences on K-ABC Analysis of Variance K-ABCa df F r a t i o F pr o b a b i l i t y Subtestsa (2,207) Hand Movements .27 NS Number Recall 4.79 £ < .01 Word Order 4.99 E < .01 Gestalt Closure 8.55 £ < .001 Triangle 31.53 E < .001 Matrix Analogies 7.89 £ < .001 Spatial Memory 12.59 E < .001 Photo Series 16.10 E < .001 Faces & Places 18.46 £ < .001 Arithmetic 9.70 E < .001 Riddles 42.08 E < .001 Reading/Decoding .93 NS Reading/Understanding 8.43 E < .001 ilesb (2,207) Sequential 4.21 E < .05 Simultaneous 31.08 E < .001 Mental Processing 22.78 E < .001 Achievement 20.18 E < .001 Prorated Achievement 17.59 E < .001 Nonverbal 21.74 E < .001 5* Bonferroni method: /£ = .05/13 = .004. b Bonferroni method: /£ = .05/6 = .008. 282 APPENDIX I K-ABC Subtest Intercorrelations I - l . Cantonese p. 283 1-2. English p. 284 1-3. Punjabi p. 285 Table 1-1 Correlations among K-ABC subtests f o r the Cantonese 5 Subtests Subtests Hand Gestalt Number Triangles Wbrd Matrix Spatial photo Eaces A r i t h - Riddles Reading/ Movements Closure Recall order Analogies Msnory series Places metic Decoding Mental processing Gestalt Closure -.08 Number Recall .29 .16 Triangles .04 .30 .05 Ward Order .13 -.11 .49 .13 Matrix Analogies .02 .09 .04 .46 .19 Spatial Memory .37 .00 .27 .20 .07 Photo Series .12 .04 -.02 .31 -.01 Achievement Faces & places .03 .08 .03 .30 .18 .20 -.07 .37 Arithmetic .22 .05 .23 .09 .32 .26 .23 .29 .38 Riddles .07 .06 .23 .24 .18 .20 .08 .25 .60 .47 Reading/Deocding .15 -.12 .30 .07 .52 .24 -.05 .23 .60 .45 .47 Reading/Under-standing .10 .07 .15 .23 .21 .30 .04 .20 .57 .58 .62 Note: Correlations above .200 are s i g n i f i c a n t at p_ < .05 and those above .280 are sig n i f i c a n t at ja < .01. a n = 70. TBble 1-2 Correlations among K-ABC subtests f o r the E n g l i s h 3 Subtests Subtests Hand Gestalt Number Triangles Word Matrix s p a t i a l photo Faces A r i t h - Riddles Reading/ Movements Closure Recall Order Analogies Manory Series Places metic Decoding Mental processing Gestalt Closure .14 Number Recall .22 .03 Triangles .24 .35 .14 Word order .16 .24 .46 .11 Matrix Analogies .28 .13 .24 .41 .20 Spatial Manory .31 .23 .03 .32 .19 .27 Photo Series .27 .20 .05 .46 .12 .45 .29 Achievement Faces & Places .29 .35 .21 .31 .28 .32 .23 .31 Arithmetic .25 .34 .06 .40 .32 .33 .34 .40 .58 Riddles .13 .12 .14 .07 .29 .21 .04 .13 .54 .44 Reading/Decoding .25 .26 .19 .22 .39 .24 .36 .25 .56 .42 .37 Reading/Under-standing .25 .18 .28 .30 .30 .38 .33 .34 .72 .52 .47 Note; Correlations above .200 are s i g n i f i c a n t at p_ < .05 and those above .280 are s i g n i f i c a n t at £ < .01. a n = 70. correlations among K-ABC subtests fear the Punjabi 5 Subtests Subtests/Scales Hand Gestalt Number Triangles Word Matrix Spatial photo Faces A r i t h - Riddles Reading/ Movements Closure Recall Order Analogies Memory Series Places metic Decoding Mental processing Gestalt Closure .06 Number Recall .05 -.09 Triangles .36 .29 .07 Word Order .10 -.05 .42 .17 Matrix Analogies .20 .38 -.03 .38 .17 Spatial Memory .16 .25 .21 .28 .28 Photo series .05 .26 -.11 .32 .02 Achievement Faces & places .01 .23 .02 .07 .13 .23 .07 .09 Arithmetic • 26 .14 .17 .26 .20 .40 .38 .30 .14 Riddles .22 .30 -.02 .24 .12 .41 .28 .23 .45 .33 Readinc^Deccding .10 -.03 .25 .33 .34 .06 .12 .01 .33 .27 .36 •Reading/under-standing .42 .20 .19 .38 .25 .26 .42 .18 .32 .58 .46 Note: Correlations above .200 are s i g n i f i c a n t a t p_ < .05 and those above .280 are sig n i f i c a n t at p_ < .01. a n = 70. to 00 Ln 286 APPENDIX J Scree Tests Mental Processing Composite J - l . Cantonese p. 287 J-2. English p. 288 J-3. Punjabi p. 289 Mental Processing & Achievement J-4. Cantonese p. 290 J-5. English p. 291 J-6. Punjabi p. 292 CO u c •H O PM cu 3 CO >' c CU oo •H W F a c t o r O r d e r 1 2 3 4 5 6 7 8 E i g e n v a l u e s 2 . 1 1 . 5 1 . 2 1 . 0 . 6 7 . 6 0 . 4 8 . 3 4 F i g u r e J - l . S c r e e t e s t on t h e 8 M e n t a l P r o c e s s i n g s u b t e s t s f o r t h e C a n t o n e s e . F i g u r e J-2. S c r e e t e s t on t h e 8 M e n t a l P r o c e s s i n g s u b t e s t f o r t h e E n g l i s h . F i g u r e J-3. S c r e e t e s t on t h e 8 M e n t a l P r o c e s s ! s u b t e s t s f o r t h e P u n j a b i . 290 l . o ; F a c t o r O r d e r E i g e n v a l u e s 3 . 9 1 . 7 1 . 6 1 . 2 1 . 0 .72 . 6 7 . 6 2 . 4 8 . 3 9 .31 . 2 5 .17 5 8 10 11 12 13 F i g u r e J - 4 . S c r e e t e s t o f t h e 13 M e n t a l P r o c e s s i n g & A c h i e v m e n t s u b t e s t s f o r t h e C a n t o n e s e 291 o 3 <0 > c Hi F a c t o r 1 2 3 4 5 O r d e r 1 * J 4 5 E i g e n v a l u e s 4 . 6 1 . 5 1 . 2 . 9 9 . 9 0 .74 7 8 9 10 . 6 5 . 5 4 . 5 2 .48 11 12 13 .41 . 2 9 .21 F i g u r e J - 5 . S c r e e t e s t on t h e 13 M e n t a l P r o c e s s i n g & A c h i e v e m e n t s u b t e s t s f o r t h e E n g l i s h . 292 o eu <u 3 rH > c <u 00 •H W i.o -F a c t o r 1 8 10 O r d e r E i g e n v a l u e s 3 . 8 1 . 7 1 . 3 1.1 . 8 7 . 8 4 . 6 8 . 6 2 . 5 4 . 4 9 11 12 13 . 4 6 .31 . 2 6 F i g u r e J - 6 . S c r e e t e s t on t h e 13 M e n t a l P r o c e s s i n g & A c h i e v e m e n t s u b t e s t s f o r t h e P u n j a b i . 293 APPENDIX K Factor Analysis for Cantonese K - l . P r i n c i p a l components - 2 factor p. 294 K-2. Unweighted least squares - 3 factor p. 295 K-3. P r i n c i p a l components and unweighted least squares - 4 factor p. 296 294 Table K-l Factor leadings for two factor and three factor principal components analysis with  a varimax rotation for the Cantonese Mental Processing Composite Subtest S e q u e n t i a l Hand Movements Number Recall Word Order Simultaneous Gestalt Closure Triangles Matrix Analogies Spatial Memory Photo Series Variances Two factors 1 68 79 64 -09 10 08 58 11 Three factors 1.84 2 1 2 3 -03 25 -16 72 -02 83 04 22 02 83 13 01 44 07 51 -20 83 03 82 17 71 11 74 05 23 10 07 80-61 -27 47 54 1.81 1.54 1.75 1.56 Note: Decminals have been emitted. Factor loadings > .350 are underlined a n = 70. 295 Table K-2 Factor leadings for the three factor unweighted least squares analysis with a varimax rotation for the Mental Processing subtests for the Cantonese Subtests 1 Factors 2 3 Hand Movements 19 -06 48 Number Recall 98 01 22 Word Order 48 12 05 G e s t a l t Closure 08 01 22 Triangles 02 87 15 Matrix Analogies 06 54 04 Spatial Memory 09 08 74 Photo Series -10 35 33 Variances 1.26 1.26 .97 Note; Decimals have been emitted. Factor loadings > .350 are underlined 296 Table K-3 Factor loadings for the four factor, unweighted least squares analysis and principal components analysis with a varirtax rotation for a l l the subtests of the K-ABC for the Cantonese • 1 2 Factors 3 4 — Analyses Subtests PC UWLS PC UWLS PC UWLS PC UWLS Hand Movements -01 -15 -06 21 -02 07 85 71 Number Recall -14 13 80 81 00 06 02 29 Word Order 03 02 77 77 15 29 07 03 Gestalt Closure 71 71 -13 10 18 -12 -03 -12 Triangles 37 80 10 -05 07 19 59 15 Matrix Analogies 73 56 10 03 15 32 19 04 Spatial Memory 53 13 50 06 -02 -06 21 84 Photo Series 63 27 -08 -38 00 43 14 45 Faces and Places 15 16 00 -08 84 82 -11 -07 Arithmetic 37 04 31 20 17 64 49 32 Riddles 37 14 -01 07 67 75 25 06 Reading/Decoding -18 -10 35 37 64 80 31 03 Reading/Understanding 20 12 27 10 44 82 61 01 Variances 2.33 1.67 1.82 1.66 1.89 3.37 1.98 1.65 Note: Decimals have been emitted. Factor loadings > .350 are underlined APPENDIX L 297 Group Differences on WISC-R L - l . Analysis of variance p. 298 298 Table L - l Group differences on the WISC-R Analysis of Variance WISC-R df F r a t i o F pro b a b i l i t y Subtests a (2,207) Information 27.88 £ < .001 S i m i l a r i t i e s 25. 24 £ < .001 Arithmetic 1. 70 NS Vocabulary 26.18 £ < .001 Comprehens ion 15.98 £ < .001 Dig i t Span . 66 NS Picture Completion 15.66 £ < .001 Picture Arrangement 8.04 £ < . 001 Block Design 35.60 £ < .001 Object Assembly 20.89 £ < .001 Coding 10.43 P < .001 Mazes 12.97 £ < .001 £ < .001 Scales* 3 (2,207) Verbal IQ 33.98 £ < .001 Performance IQ 37.00 £ < .001 F u l l Scale IQ 44.94 £ < .001 a Bonferroni method: /p_ = .05/12 = .004. b Bonferroni method: /p = .05/3 = .02. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0054558/manifest

Comment

Related Items