Assessment Centers More Valid than Cognitive Tests

For many years meta-analyses have reported that cognitive ability tests were more predictively valid than assessment centers (r = .51 vs r = .37). A recent refined meta-analysis by Sackett, Shewach, & Keiser (2017) shows that assessment centers are more valid (r = .44) than cognitive ability tests (r = .22). This conclusion results from comparison of “head-to-head” studies in which both methods were used to predict the same broad criterion, namely job performance, the outcome of interest in applied studies in operational organizational situations. Past meta-analyses of non-overlapping studies have aggregated validities from studies of (a) cognitive ability tests in relation to a wide variety of very narrow criteria vs (b) assessment centers in relation to all criteria. This new “head-to-head” method of comparison has the advantages of using the same sample, corrections for range restriction and measurement error, as well as the same criterion most relevant to assessment center application, that is job performance.  Sackett et al (2017) conclude: “The present findings lead to a conclusion that, for predicting the broad criteria of interest to AC users, ACs commonly match or exceed the validity achieved using an ability test (p. 1445).”


(Sackett, P.R., Shewach, O.R., & Keiser, H.N. (2017). Assessment centers versus cognitive ability tests: Challenging the conventional wisdom on criterion-related validity. Journal of Applied Psychology, 102, 1435-1447.)