Principled failure to detect a five-factor structure in a canonical Big Five measure

A standard claim in the personality literature is that the “Big Five” dimensions of personality are universal, in the sense that they show up across virtually all human populations, and naturally fall out of PCA/EFA analyses of lexical and questionnaire data. How- ever, it is not clear whether the repeated emergence of five factors genuinely reflects a fundamental insight into the causal structure or human personality, or simply recapitulates structure intentionally or unintentionally built into standard measures of personality. Here I show that, in the absence of top-down selection effects, PCA of indi- vidual personality items shows no systematic tendency to produce a five-factor solution—even when the items are drawn from a broadband five-factor inventory (the NEO-PI-R).