The Dimensionality of the 19-Item Copenhagen Burnout Inventory Adapted for Students: A Bifactor Penalised SEM Analysis

Research on the validity of the multidimensional Copenhagen Burnout Inventory (CBI) has predominantly relied on confirmatory factor analysis (CFA), despite concerns regarding inflated factor correlations, local misspecification, and limited subscale distinctiveness. This study evaluated the dimensionality of the 19-item CBI using an adapted student version and cross-sectional data from 1938 South African undergraduates. Three-factor, two-factor, and second-order CFA models were compared with a bifactor penalised structural equation model (PSEM) employing alignment loss function (ALF) priors, auxiliary indices, and criterion validity regression analyses involving study demands and resources. While the three-factor and second-order CFA models demonstrated acceptable global fit, the PSEM local misfit analysis revealed inflated loadings and factor correlations, cross-loadings, and item dependencies that were not sufficiently captured by the global fit indices. The retained bifactor PSEM model demonstrated an excellent fit and indicated a criterion-valid, congeneric general burnout factor. Personal and peer burnout were identified as interpretable residual factors; only peer burnout demonstrated criterion validity, and study burnout was largely accounted for by the general factor. Examples of how to interpret bifactor general-plus-specific burnout scores on the CBI for intervention purposes are provided.