OBJECTIVE: To perform a cross-cultural adaptation of the Portuguese version of the Maslach Burnout Inventory for students (MBI-SS), and investigate its reliability, validity and cross-cultural invariance.
METHODS:The face validity involved the participation of a multidisciplinary team. Content validity was performed. The Portuguese version was completed in 2009, on the internet, by 958 Brazilian and 556 Portuguese university students from the urban area. Confi rmatory factor analysis was carried out using as fi t indices: the 2 /df, the Comparative Fit Index (CFI), the Goodness of Fit Index (GFI) and the Root Mean Square Error of Approximation (RMSEA). To verify the stability of the factor solution according to the original English version, crossvalidation was performed in 2/3 of the total sample and replicated in the remaining 1/3. Convergent validity was estimated by the average variance extracted and composite reliability. The discriminant validity was assessed, and the internal consistency was estimated by the Cronbach's alpha coeffi cient. Concurrent validity was estimated by the correlational analysis of the mean scores of the Portuguese version and the Copenhagen Burnout Inventory, and the divergent validity was compared to the Beck Depression Inventory. The invariance of the model between the Brazilian and the Portuguese samples was assessed.
RESULTS:The three-factor model of Exhaustion, Disengagement and Effi cacy showed good fi t ( 2 /df = 8.498, CFI = 0.916, GFI = 0.902, RMSEA = 0.086). The factor structure was stable (:
CONCLUSIONS:The Portuguese version of the Maslach Burnout Inventory for students showed adequate reliability and validity, but its factor structure was not invariant between the countries, indicating the absence of cross-cultural stability.