Psychometric evaluation of a contextualized version of the German BFI-2 for the digital world across self- and other-ratings

While the Big Five Inventory-2 offers efficient assessment of personality domains and facets, continuous psychometric evaluation remains crucial to ensure valid personality assessment. The present study examined whether the dimensional structure and measurement properties of the German BFI-2 remain invariant across offline and online contexts and methods (self- and other-ratings). Data from 403 self-raters and 151 informants were analyzed using confirmatory factor analysis, exploratory structural equation modeling, and measurement invariance testing. Reliability was generally good, and the Big Five structure was largely replicated, though several items showed context- or method-specific deviations, particularly within Agreeableness and Negative Emotionality. Partial scalar invariance was achieved at facet level, although model fit was poor for other-ratings, while only metric invariance emerged at domain level. Exploratory analyses revealed that correlations between self-ratings across contexts were highest, and that self-other agreement was stronger online than offline. Regression models predicting internet use showed that traits often exhibited opposite effects across contexts, with online traits explaining incremental variance, although overall explained variance was small. Findings highlight that both context and method affect the properties of the BFI-2, underscoring the importance of psychometric evaluation across diverse conditions.