An investigation of differential prediction with the WISC-R

Abstract

Test bias continues to be a topic of particular concern in determining the eligibility of minority students for placement in special education programs. Although critics of IQ testing have made some concessions concerning the internal structure of IQ tests, the differential prediction of academic achievement for minority group members remains a critical issue. Investigations of the correlation between IQ tests and standardized achievement tests have typically not produced evidence supporting the differential validity hypothesis. However, critics have claimed that standardized achievement and IQ tests are alternate samples of the same behavioral domain and have called for the validation of IQ with classroom achievement. The relationship between IQ, California Achievement Test scores, and grade point average was examined for a sample of 168 (83 black and 85 white) elementary and middle school students. An examination of the correlation coefficients and regression lines demonstrated little evidence of differential prediction. A significant relation was obtained between Full Scale IQ and grade point average for both black and white students. Overall, the data suggest that the WISC-R IQ is an equally valid predictor of academic achievement for the two groups.