Specific relationships between psychosocial job conditions and job-related stress: A three-level analytic approach

This cross-sectional questionnaire study presents a multi-level analysis on 2565 workers in 188 departments in 36 organizations in the Netherlands. A three-level model is used in which individual workers are nested within departments, which in turn are nested within organizations. Research questions concern (1) the amount and distribution of variance in job-related stress explained for the three levels in the study (individuals, departments, organizations), and (2) the specificity of relationships between psychosocial job demands and job-related stress in the three-level model. Well-being showed slightly more raw variance to be explained at supra-individual levels than strain. The full regression model explained about 35% of the total variance in both work-related strain and well-being. Psychosocial job conditions did not exceed the expected amount of 10 to 15% contribution to this explained variance. These results do not differ from comparable studies that do not use multi-level analysis. The variance distribution in the full model, however, showed unexplained variance to be located at the individual level for both strain and well-being, and at the departmental level only for well-being. This last finding shows a direction for possible improvement of work stress models. Specificity of relationships was also shown: psychological job demands were more strongly related to strain, whereas job content variables (i.e. job variety, job control) were more strongly related to well-being. Results also suggested that social support was more strongly associated with well-being than with strain. Well-being appeared to have a more widely varying range of predictors than strain.