Confirmatory factor analysis and a short measure of five-factor model of personality

Trait theories of personality have frequently been reliant on exploratory factor analysis (EFA) not only for the development of questionnaires but for the development of the theory underlying these questionnaires. While EFA lead to the development of the Five-Factor model of personality, the most dominate trait based model in psychology today, the advancement of CFA (confirmatory factor analysis) has resulted in criticism of this model. This is because measures of the Five-Factor model have consistently resulted in a poor fit between the data and the underlying theory. This research explores methods of reconciling CFA and the Five-Factor model of personality where the NEO-FFI, a short measure of the Five-Factors is explored under CFA conditions. Five Hundred and twenty three tertiary education students completed the NEO-FFI and there responses were explored via CFA. A number of methods for improving model fit were compared and contrasted in order to determine the best approach. Results suggested that item parcels preformed better than any other method where model fit, reliabilities, average factor loadings and correlations between factors were considered. Regardless of the method used, correlations between factors were consistently large suggesting theory needs to address the distinctness and thus the “basicness” of the Five-Factor model of personality. It is suggested that the various attempts at reconciling CFA and the Five-Factor of personality should be treated with caution given theoretical concerns raised in relation to the Five-Factor model of personality