Quo vadis taxonomies of consistent behavioural patterns: time to change horses?

For over a century, factor analysis (and its variants) has been the dominant statistical methodology in differential psychology, in particular in the development of taxonomies of consistent behavioural patterns (CBP) whether normative or pathological. Factor analysis suffers from considerable methodological problems such as arbitrariness of factor number, rotation indeterminacy, factor indeterminacy, lack of model robustness, universality (non-specificity). Over the years considerable effort has been applied to ameliorate many of these problems. This review examines the most recent literature addressing these methodological issues. The use of factor analysis appears to still be guided more by tradition (“if I have only a hammer, then everything is a nail”), art, and heuristics than by rigor. New methods are suggested including time series, computational, constructivist and complex systems approaches.