THE CONCEPT OF DYNAMIC CRITERIA: A CRITICAL REANALYSIS

Abstract

Discussions of “the criterion problem” stress the assumedly frequent incidence of dynamic criteria. However, different concepts of dynamic criteria have not been distinguished. Three views of dynamic criteria are clarified in this paper. These are dynamic criteria conceptualized as (a) changes in group average performance over time, (b) changes in validity over time, and (c) changes in the rank‐ordering of scores on the criterion over time. The evidence cited for each concept of dynamic criteria is critically analyzed and submitted to significance tests. The results of analyses of 735 r 's and 532 paired r 's across time from studies pertinent to dynamic criteria are reported. It is concluded that dynamic criteria are rare phenomena, with the significant changes found in key studies explainable by methodological artifacts. Though many sources suggest that dynamic criteria have dire consequences for the overall selection process, we argue that the phenomena have been over‐emphasized in the literature, and that practitioners' concern should focus instead on removing sources of criterion unreliability.