The use of individually administered cognitive abilities tests with academically

Abstract

The purpose of this study was to gather information about criterion-related and construct test validity as applied to a referral or "at-risk" sample. Two individually administered cognitive abilities tests, the Wechsler Intelligence Scales for Children-Third Edition (WISC-III) and the Woodcock-Johnson Psycho-Educational Battery-Revised Tests of Cognitive Ability (WJ-R Cog), frequently used in the special education placement decision process have recently been revised. Before test generalizations obtained from the standardization sample can be presumed to apply to a referral or "at-risk" sample, research is needed to explore the applicability of these assumptions to this population. No previous research has used a random sample of children referred for special education evaluation who are not yet placed into a program to explore the construct and concurrent validity of both complete test batteries. A random sample of 183 students referred for possible special education placement was used. Ages ranged from 6 to 16 years. Students were classified into three age groups, four ethnic groups, and six possible special education classifications, including a non-placed category. Data were analyzed using a variety of techniques including: exploratory factor analysis, confirmatory factor analysis, stepwise multiple regression, and multiple discriminant function analysis. Findings of this study support the four factor WISC-III model and the seven factor WJ-R Cog model. Study results indicate a concurrent verbal component (approximately 50%) comprised of the WISC-III VC factor and the WJ-R Cog Gc, Gs, and Ga factors. This verbal component was the best composite predictor variable of academic achievement. Although "at-risk" students scored lower on these verbal component factors, BWL, or broad written language, was the best single discriminator variable in both a six and two group analyses. Special education classifications were examined for specific category profiles and the ACID/ACIDS profile was investigated. This study initiated information about the underlying test constructs and how these constructs relate to each other when given to an "at-risk" referral sample. It is hoped that this study will stimulate further research in this area.