Take two: Sources and deterrents of score change in employment retesting

Retesting occurs at all stages of the candidate life cycle-from screening and selection through development and promotion-and therefore has numerous human resource implications. Existing frameworks of retest performance address statistical and conceptual issues related to retest score change; however, little is understood about what causes score gains due to repeat administrations and the effect this has on construct and criterion-related validity. This article reviews the literature on retest effects and introduces a number of factors, including a) construct-relevant change, b) construct-irrelevant change due to individual difference traits, states, and motivation, and c) methods and procedures that may deter construct-irrelevant contamination between tests. In cases where empirical evidence is lacking, propositions are presented to guide future research in evaluating the extent to which construct-irrelevant individual differences and motives contaminate retest performance. We also review ways organizations may limit construct-irrelevant score gains in retesting by attempting to standardize test administration and enhance test security. We end with a discussion of the framework's implications on retest score-based decisions, and methods to mitigate retest score contamination. It is hoped that this paper will facilitate a more comprehensive understanding of retest effects in organizational settings and a way forward for both research and practice.