Cross-sample and cross-language invariance of the Five-Dimensional Personality Test (5DPT): exploratory and confirmatory analyses using Dutch, American and Italian samples

ABSTRACT Demonstrating factor invariance across population characteristics is an important issue in personality assessment, particularly if considered from a semi‐idiographic point of view. Using two Dutch samples of general population subjects, we conducted both exploratory and multigroup confirmatory factor analyses to investigate whether the clinically relevant and theory‐informed Five‐Dimensional Personality Test (5DPT) dimensions—Insensitivity, Extraversion, Neuroticism, Orderliness and Absorption—remain stable across subsamples that differ in terms of sex, age, educational level or the subjects' position on each of the 5DPT dimensions themselves. Using samples of American and Italian respondents, we performed similar analyses, testing the cross‐language invariance of the 5DPT factors. To prevent workspace problems, the confirmatory analyses were executed on item parcels instead of individual items. The inter‐factor correlations in these analyses were allowed to deviate from zero. Whether calculating the Tucker coefficients of factor similarity (exploratory analyses) or evaluating model fit by the cut‐off criteria recommended by Hu and Bentler (confirmatory analyses), the results attested the invariance of the 5DPT factors, both across the eight sample parameters and across language. Copyright © 2012 John Wiley & Sons, Ltd.