Predicting the Outcomes of Single- and Multistage Selections When Applicant Pools Are Small and Heterogeneous

Abstract

This article presents analytic methods and accompanying computational tools for estimating the expected quality and diversity outcome of general multistage selections when the applicant pool is a finite, typically small sample from a mixture of majority and minority applicant populations. The new methods generalize the analytic estimation of the expected quality of small heterogeneous applicant pool selections from simple, single-stage selection situations to the more general, multistage selection context. In addition, the new methods also compute the sampling variability of the selection quality outcome for these selection decisions.