Abstract
The purpose of this study was to examine the extent to which Maslach Burnout Inventory (MBI) related to a measure of similar construct. A cross-sectional research design was used to draw explicit conclusions regarding concurrent validity of the MBI. Subjects were residential workers at rehabilitation centers for people with mental retardation. Two hundred sixty-one subjects completed on all items of burnout in our investigation. The measurement model of MBI was the three-factor oblique model consisted of 18 items by Schaufeli and van Dierendonck (1993) . A second -order factor model regarding the revised version of Pines Burnout Measure, which was reported in our previous study, was used as a criterion standard. Confirmatory factor analysis was used to test correlations between latent variables, regarding the MBI's three factor and the criterion factor. The goodness-of-fit indices of the model were adequate. The factorial intercorrelations of Emotional Exhaustion (EE) toward the criterion factor was 0.863, Depersonalization (DP) 0.445, and Personal Accomplishment (PA) -0.111. The correlation between PA and the criterion was not statistically significant. Furthermore, PA showed correlations with neither EE nor DP. The correlation between PA and EE was posltive inconsistently in the expected direction defined by the previous literature. These findings tested by correction for attenuation of measurement error uslng structural equation modeling technique, supported that EE is considered as the more robust factor in burnout, and PA is independent of the construct in a cross-sectional study. The inadequacy of the three-aspect conceptualization of burnout was discussed.
