Predicting job success with the assessment center : Validity or artifact ?

Abstract

At the outset the question was asked whether the results of Assessment Centres are really as reliable in predicting future job performance as recent research indicates. Critical evaluation of earlier studies of the predictive validity of Assessment Centres seems to suggest the possibility that a considerable proportion of the observed connection between Assessment Centres predictors and subjective indicators of job performance could be the result of criteria contamination. This alternative interpretation of the validity of Assessment Centres would mean that Assessment Centres do not, as originally intended, measure the characteristics which are required for success in a particular job. Rather such characteristics as the ability to 'fit in' with the particular elements of an organization's culture, and the skill to present ones own abilities are brought to the fore. Because on both sides the same characteristics, equally irrelevant to later performance, are emphasised, one would expect a corresponding level of validity to be observable. However, the use of an objective criterion of job performance as in this study should mean the presence of a subjective and irrelevant element on the criteria side will have no tendency to exaggerate predictive validity. As can be shown, even when objective measurements of performance are used the Assessment Centre in this study has a predictive validity, which in terms of quality is comparable with other studies, which use less objective job performance criteria.