Exploring Issues of Personality Measurement and Structure Through the Development of a Short Form of the Eysenck Personality Profiler

Abstract

In this article, we develop a revised short form of the original Eysenck Personality Profiler (EPP; H. J. Eysenck & Wilson, 1991). In addition, we address topics of broad theoretical importance such as the recurrent empirical finding of correlations between conceptually orthogonal personality dimensions and the possibility that gender differences in these dimensions are partly spurious. In Study 1 (N = 227), we demonstrate that the existing short form of the EPP (EPP-SF; H. J. Eysenck, Wilson, & Jackson, 1996) provides a poor fit to the data and we develop a revised well-fitting version. In Study 2, we retest this version on an independent new sample (N = 3,374) where it is again found to fit the data well. We show that most of the structural and measurement parameters of the revised EPP-SF are invariant across genders. Structured means analysis indicated a significant gender difference in Psychoticism, with men scoring higher than women, but no differences in Extraversion or Neuroticism. Our discussion focuses on issues concerning personality measurement and structure, including an examination of the role of confirmatory factor analysis in personality research.