Key points are not available for this paper at this time.
Abstract A method of simulating the effect of differential component weights on the predictive validity and adverse impact of a selection battery is described and illustrated. This simulation technique relies on previous selection research to generate descriptive statistics regarding the interrelations among test components and job performance. In addition, applicant data and projections of hiring needs are used to estimate hiring rates and adverse impact ratios using different weighting combinations. The result is an estimate of a selection battery's predictive validity under different subcomponent weighting schemes as well as a forecast of the number of members of each applicant subgroup that would be hired under each weighting scheme. The use of this method is illustrated through data from two situations involving selection for public sector jobs. Implications for the practice of personnel selection in the public and private sector are discussed. Suggestions for specific areas of future research are offered.
Doverspike et al. (Sun,) studied this question.