The NEO-PI-R: Factor structure and gender invariance from exploratory structural equation modelling analyses in a high-stakes setting

This study presents new analyses of NEO Personality Inventory–Revised (NEO-PI-R) responses collected from a large British sample in a high-stakes setting. The authors show the appropriateness of the five-factor model underpinning these responses in a variety of new ways. Using the recently developed exploratory structural equation modeling (ESEM) technique, the authors show that model fits improve markedly over conventional confirmatory factor analyses (CFA) of the same data set, but that (a) factor interpretations do not change under ESEM analyses, (b) ESEM factor scores, just like CFA factors scores, correlate at near unity with sums of observed scores, (c) NEO-PI-R facets under ESEM analyses are invariant across gender, and (d) ESEM highlights the inappropriateness of alpha and beta as a higher order representation of NEO-PI-R facets, whereas a CFA approach might lead researchers to believe in the appropriateness of these higher order factors. These results, coupled with the existing validity evidence for the NEO-PI-R, suggest that the five-factor structure is the most parsimonious structure for summarizing NEO-PI-R responses from high-stakes settings in the United Kingdom