Abstract:
Social variables, rooted in the perceived social order, have appeared in various societies over the centuries. In this century, a vast array of techniques have been generated and assembled for the construction and testing of educational and psychological variables. A ubiquitous index used as a criterion to establish the quality of test items will be shown to contain a weakness which produces a talent bias which disadvantages doubly already dis- advantaged groups. In addition to evidence of bias in the index, the factor which together enabled the realisation of this bias after more than a half century of its use and the continued resistance to this evidence are studied.