Purpose: This study illustrates complementary variable- and person-centered approaches allowing for a more complete investigation of the dimensionality of psychometric constructs. Psychometric measures often assess conceptually related facets of global overarching constructs based on the implicit or explicit assumption that these overarching constructs exist as global entities including conceptually related specificities mapped by the facets. Proper variable- and person-centered methodologies are required to adequately reflect the dimensionality of these constructs. Design/Methodology/Approach: We illustrate these approaches using employees’ (N = 1077) ratings of their psychological wellbeing at work. Findings: The results supported the added value of the variable-centered approach proposed here, showing that employees’ ratings of their own wellbeing simultaneously reflect a global overarching wellbeing construct, together with a variety of specific wellbeing dimensions. Similarly, the results show that anchoring person-centered analyses into these variable-centered results helps to achieve a more precise depiction of employees’ wellbeing profiles. Implications: The variable-centered bifactor exploratory structural equation modeling (ESEM) framework provides a way to fully explore these sources of psychometric multidimensionality. Similarly, whenever constructs are characterized by the co-existence of overarching constructs with specific dimensions, it becomes important to properly disaggregate these two components in person-centered analyses. In this context, person-centered analyses need to be clearly anchored in the results of preliminary variable-centered analyses. Originality/Value: Substantively, this study proposes an improved representation of employees’ wellbeing at work. Methodologically, this study aims to pedagogically illustrate the application of recent methodological innovations to organizational researchers.
- latent profiles
- factor mixture