Improving Assessment Centers by the Trimodal Concept of Personnel Assessment

Abstract

Lance (2008) attacks the ‘‘assessment center validity paradox,’’ arguing that construct validity is not as low as a number of researchers seem to have agreed about. But he neglects the second part of the supposed paradox, predictive validity. Assessment center predictive validity is not as high as is circulated by many researchers—and all the more by most of the users. This seems to me the even more important deficit. My suggestion is to solve the theoretical problem of construct validity as well as the pragmatic problem of predictive validity by (re)conceiving the assessment center as a truly multimodal procedure. Assessment centers used to be valuable diagnostic tools—at times when they were run as procedures integrating a broad methodological diversity. One of the icons of assessment center research, the extensive ATT that is, there is a rather continuous decline of assessment center validity over the past 40 years. There are several possible explanations for this decline. My suggestion is that assessment centers often perform poorly because too simplistic methods are employed. Their attractiveness for managers and practitioners in personnel departments is connected with concentrating on ‘‘exercises’’such as a group discussion, roleplay, and presentation, which allow for behavioral observations and a lively personal impression formation but are essentially nonpsychometric tools. Correspondence concerning this article should be addressed to Heinz Schuler. E-mail: schuler@unihohenheim.de Address: Department of Psychology (540F), University of Hohenheim, D-70593 Stuttgart, Germany Heinz Schuler, Department of Psychology (540F), University of Hohenheim Industrial and Organizational Psychology, 1 (2008), 128–130. Copyright a 2008 Society for Industrial and Organizational Psychology. 1754-9426/08