PSYCHOSOCIAL WORK-ENVIRONMENT AND HEALTH IN UNITED-STATES METROPOLITAN-AREAS - A TEST OF THE DEMAND-CONTROL AND DEMAND-CONTROL-SUPPORT MODELS

The authors use confirmatory factor analysis to investigate the psycho social dimensions of work environments relevant to health outcomes, in a representative sample of five U.S. metropolitan areas. Through an a ggregated inference system, scales from Schwartz and associates' job s coring system and from the Dictionary of Occupational Titles (DOT) wer e employed to examine two alternative models: the demand-control model of Karasek and Theorell and Johnson's demand-control-support model. C onfirmatory factor analysis was used to test the two models. The two m ultidimensional models yielded better fits than an unstructured model. After allowing for the measurement error variance due to the method o f assessment (Schwartz and associates' system or DOT), both models yie lded acceptable goodness-of-fit indices, but the fit of the demand-con trol-support model was significantly better. Overall these results ind icate that the dimensions of Control (substantive complexity of work, skill discretion, decision authority), Demands (physical exertion, phy sical demands and hazards), and Social Support (coworker and superviso r social supports) provide an acceptable account of the psychosocial d imensions of work associated with health outcomes.