Several studies challenge the three-dimensional structure of the Maslach Burnout Inventory-Human Services Survey (MBI-HSS), citing alternative measurement models including bifactor models. While bifactor models have merit, if data sampling violates assumptions of Stochastic Measurement Theory (SMT) the bifactor model requires modification prior to application. The present study compared five alternative MBI-HSS factor models using both Confirmatory Factor Analysis (CFA) and Exploratory Structural Equation Modeling (ESEM). Data from a cross-sectional survey of United Kingdom (UK) social workers were examined (N = 1257), with validation analyses conducted in an independent sample (N = 162). Bifactor models, re-specified to account for SMT, provided good fit. However, improved fit was observed for a bifactor-ESEM specification, in both test (χ 2 = 1,112.93, df = 149, p <.001, CFI =.969, RMSEA =.072, 90% CI [.068,076]) and validation (χ 2 = 227.89, df = 149, p <.001, CFI =.978, RMSEA =.057, 90% CI [.042,072]) samples. The results confirm the MBI-HSS possesses a bifactor structure in UK social workers when SMT is considered, and that bifactor-ESEM may provide a better framework to examine MBI-HSS.
- social work