On the use of confirmatory factor analysis in personality research

This paper is concerned with applications of confirmatory factor analysis (CFA) in personality research. It is argued that traditional utilizations of CFA capitalize excessively on restrictive assumptions and cannot be generally expected to be particularly informative if some of them are violated. Recent approaches to model evaluation, specifically alternative ways of fit-assessment as well as determination of the power of the model test, are then discussed and recommended for use in studies of personality. (C) 1998 Elsevier Science Ltd. All rights reserved.