Test item bias in the WISC--R.

Abstract
Test item bias in the WISC and WISC-R Information subtest has been a concern for many Canadian psychologists for over three decades. Several investigators have attempted to identify and evaluate the source of this bias. The method of choice in these studies has been rank order comparisons of pass-rates between the standardized and modified items. It is suggested that this methodology is incomplete to establish the presence or absence of item bias. In addition to observing the rank order to pass rates, it is recommended that a chi square approach to used based on the proportion of responses within ability levels to evaluate test item bias. The advantages of this strategy are demonstrated with data from 342 psychiatric inpatients and outpatients. The results demonstrate the power of this methodology over simple rank order comparisons.

This publication has 2 references indexed in Scilit: