From types to typological thinking: A reply to asendorpf

We continue to disagree with Asendorpf (2006) on the best way to analyse Q-sort data and on our priorities for personality research. We believe on statistical grounds that the large first factor found in inverse factor analyses of raw CAQ items tells us much about response norms, but little or nothing about individual differences. These emerge more clearly in analyses of standardised items, which show the familiar dimensions of the Five-Factor Model. Based on our research on types and the mixed results reported by other researchers, we do not believe that replicable empirical types are likely to be found, and suggest that a more profitable line of research would focus on the heuristics of types and the configural interpretation of traits.

Duke Scholars

Author Paul T. Costa Jr. Medicine, Geriatrics and Palliative Care