Correction of validity coefficients for direct restriction in range occasioned by univariate selection

Frequently, in evaluating a selection instrument, criterion data are available only for those applicants selected. Validities computed on the selectees are reduced by restriction in range and do not reflect the actual utility of the predictors. K. Pearson's (1903) correction for direct restriction in range uses the restricted validity (computed in the selectee group) and the predictor standard deviations in the unrestricted and the restricted groups. Often, the applicant group predictor standard deviation is not available. However, the practitioner may know, or be able to estimate, the selection ratio. An equation and a table for estimating the unrestricted validity from the restricted validity and the selection ratio are provided. (5 ref)