Factorial Validity of the Personality Adjective Checklist in a Dutch-Speaking Sample

Abstract

We examined the factorial structure of the Dutch version of the Personality Adjective Checklist (PACL–D) in a Belgian sample of 3,012 community-dwelling adults. Exploratory factor analyses revealed a 5-factor structure (Neurotic, Aggressive/Dominant, Introverted vs. Extraverted, Conscientious, and Cooperative), that showed considerable overlap with 3 of the Big Five factors (i.e., Neuroticism, Extraversion, and Conscientiousness). Moreover, the 5-factor structure closely resembled the structure found in the original American PACL and was equivalent across gender and age.

Notes

A study by Vanroy (2008) on the validation of the PACL showed no influence of the method used (online vs. paper-and-pencil), so we decided to combine these two samples.

Although confirmatory factor analysis is usually considered most appropriate to test a priori specified models, Church and Burke (1994) claimed that this technique is best suited to analyze simple structures, but is less appropriate for complex models containing lots of meaningful secondary loadings or cross-loadings. Because this is mostly the case when dealing with personality data (see also McCrae et al., 1996), we believed that Procrustes rotation better served our purpose, as this technique allows taking into account meaningful cross-loadings as well.

We would like to thank the authors for providing us the full factor loading matrix.

For reasons of comparability, and in response to the editor's comment, we also conducted the current congruence analysis using the same estimation, extraction, and rotation strategies as the target matrix of Strack and Lorr (1990; i.e., a principal components analysis on the Pearson correlation matrix with orthogonal rotation). This analysis yielded the following coefficients of congruency:.80,.87,.79,.80, and.71 for Aggressive/ Dominant, Neurotic, Conscientious, Detached/ Introverted, and Surgent/ Extraverted, respectively.