An examination of the five-factor model of normal personality variation with reliable component analysis

Abstract

Evidence has been presented in recent years which suggests that five personality domains are needed to describe normal personality variations. The domains of the five-factor model (FFM) have been termed Neuroticism, Extraversion, Openness to Experience, Agreeableness, and Conscientiousness. Nearly all of the evidence in support of the FFM is dependent on the choice of analytic method. The method selected by the vast majority of researchers has been principal component analysis followed by varimax rotation. The present study examined the correlations between the facet scales of the five domains of the NEO Personality Inventory, Revised Edition [NEO-PI-R; Costa & McCrae, 1992b Professional Manual: revised NEO Personality Inventory (NEO-PI-R) and NEO Five-Factor Inventory (NEO-FFI)] with reliable component analysis. Reliable component analysis is a multivariate method which defines orthogonal composites with maximum reliability. When five reliable components were varimax-rotated, the FFM was recovered. However, rotating three components supported an alternative personality model, that of Eysenck, despite the fact that the analyses were conducted on measures designed to assess the FFM.