Further Investigation of the Dimensionality of the Questionnaire for Eudaimonic Well-Being

The dimensionality of the Questionnaire for Eudaimonic Well-Being (QEWB) has been a topic of debate and divergent findings in the literature up to date. This study investigated the factor structure and measurement invariance of the QEWB in four culturally diverse South African samples using confirmatory factor analysis (CFA), bifactor CFA, exploratory structural equation modelling (ESEM), and bifactor ESEM. Three student samples completed the English (n = 326), Afrikaans (n = 478), or Setswana (n = 260) version of the QEWB. An adult sample (n = 262) completed the English version. The one-factor structure revealed poor fit for the student samples. Although the four-factor models generally showed slightly better fit than the three-factor models, the latter was preferred for parsimony. The bifactor ESEM model displayed good fit for the student samples, with the general factor and some specific factors attaining sufficient reliability scores, pointing to the potential use of the scale in these samples. Configural invariance between the student samples was supported, but not metric nor scalar invariance. For the adult sample, none of the models displayed good fit and the use of the QEWB in this sample is not recommended. The results point towards the existence of a global eudaimonic well-being factor and, at the same time, the interrelatedness of facets of eudaimonic well-being. It suggests that eudaimonic well-being may be represented by the same items across the three student groups. The influence of developmental phase on the manifestation and measurement of eudaimonic well-being should be explored in future.

Automated summary

This study investigated the factor structure and measurement invariance of the QEWB in four culturally diverse South African samples. The one-factor structure did not fit well for the student samples, while the four-factor models showed slightly better fit. The bifactor ESEM model displayed good fit for the student samples. Configural invariance between the student samples was supported, but not metric nor scalar invariance. The use of the QEWB in the adult sample was not recommended. The results suggest the existence of a global eudaimonic well-being factor and the interrelatedness of facets of eudaimonic well-being. The influence of developmental phase on eudaimonic well-being should be explored in future.