Does high scatter affect the predictive validity of WAIS-III IQs?

Research paper by Joseph J JJ Ryan, David S DS Kreiner, D Bradley DB Burton

Indexed on: 14 Feb '03Published on: 14 Feb '03Published in: Applied neuropsychology


We tested the assumption that high amounts of intersubtest scatter on the Wechsler Adult Intelligence Scale-Third Edition (WAIS-III) subtest profiles compromise predictive validity of the IQs for predicting Wechsler Memory Scale-Third Edition (WMS-III) indexes. Data from a sample of 80 male Veteran's Administration medical center patients were analyzed, half with high intersubtest scatter and half with low scatter. The 2 groups were matched on Full Scale IQ. Correlations of WAIS-III Full Scale IQ with WMS-III indexes were not significantly different between the 2 groups. Further, the regression equations for predicting WMS-III indexes did not depend on the amount of scatter. The results suggest that, when differences in IQ are controlled, the validity of WAIS-III scores in predicting memory performance does not depend on the amount of intersubtest scatter. Further research is needed with samples from different populations using a variety of criterion variables.