Does concurrent validity really estimate predictive validity in psychological testing? Two local studies

Abstract

Concurrent validity designs are used widely in applied psychology as proxy estimates of predictive validity in operational settings, although few primary (local) studies have investigated the generalizability of concurrent validity coefficients empirically. The present study compared the same assessment tool and performance criterion between incumbent and applicant consumer samples from two large financial institutions (N = 2942 and N = 2880), without the common issue of range restriction in the applicant groups. The results found no significant differences in the observed validity coefficients between the groups, despite evidence of impression management in the applicant samples. In addition, range restriction corrections in the concurrent samples would have likely overestimated the predictive validities. Practical implications are briefly discussed.