Open AccessDimensionality of the 9‐item Utrecht Work Engagement Scale revisited: A Bayesian structural equation modeling approach
Author(s) -
Fong Ted C. T.,
Ho Rainbow T. H.
Publication year - 2015
Publication title -
journal of occupational health
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.664
H-Index - 59
ISSN - 1348-9585
DOI - 10.1539/joh.15-0057-oa
Subject(s) - structural equation modeling , confirmatory factor analysis , curse of dimensionality , bayesian information criterion , factor analysis , residual , work engagement , scale (ratio) , bayesian probability , statistics , deviance information criterion , prior probability , mathematics , computer science , bayesian inference , work (physics) , algorithm , engineering , mechanical engineering , physics , quantum mechanics
Dimensionality of the 9‐item Utrecht Work Engagement Scale revisited: A Bayesian structural equation modeling approach: Ted C. T. F ong , et al . Centre on Behavioral Health, The University of Hong Kong, Hong KongObjective The aim of this study was to reexamine the dimensionality of the widely used 9‐item Utrecht Work Engagement Scale using the maximum likelihood (ML) approach and Bayesian structural equation modeling (BSEM) approach. Methods Three measurement models (1‐factor, 3‐factor, and bi‐factor models) were evaluated in two split samples of 1,112 health‐care workers using confirmatory factor analysis and BSEM, which specified small‐variance informative priors for cross‐loadings and residual covariances. Model fit and comparisons were evaluated by posterior predictive p‐value (PPP), deviance information criterion, and Bayesian information criterion (BIC). Results None of the three ML‐based models showed an adequate fit to the data. The use of informative priors for cross‐loadings did not improve the PPP for the models. The 1‐factor BSEM model with approximately zero residual covariances displayed a good fit (PPP>0.10) to both samples and a substantially lower BIC than its 3‐factor and bi‐factor counterparts. Conclusions The BSEM results demonstrate empirical support for the 1‐factor model as a parsimonious and reasonable representation of work engagement.