Construct validation of assessment center final dimension ratings: Definition and measurement issues
Existing evidence concerning assessment center construct validation suggests that assessment center final dimension ratings do not exhibit adequate discriminant validity. This evidence is considered in terms of several possible types of evidence suggested by Ghiselli, Campbell, and Zedeck (1981) and with respect to three possible explanations for lack of construct validity offered by Cronbach and Meehl (1955). A review of the literature leads to the conclusion that ad hoc definition of performance models, limitations of research and analytic methodology, and concentration on legal validity versus scientific validation are at least as plausible as explanations for findings as are problems with the measurement method itself. Assessment center ratings may be measuring constructs which have meaning to assessors and subject matter experts, but which have not been empirically identified by designers and evaluators of the method. Suggestions for further research are offered.
Jones, Robert G. "Construct validation of assessment center final dimension ratings: Definition and measurement issues." Human Resource Management Review 2, no. 3 (1992): 195-220.
Human Resource Management Review