The stress process in occupational settings : the role of psychosocial factors
This thesis is concerned with the direct and indirect mechanisms through which psychosocial stressors affect well-being, within the framework of the Michigan Model (House, 1981). An additional aim is to examine the stress process during workplace transitions. Empirical studies were conducted in three occupational settings: a cross-sectional pilot study with a sample of white-collar employees of several companies (N=144), and two longitudinal studies with samples drawn from a single retail chain, the "manager study" (N=261) and the "relocation study" (N=175). Although statistical control for individual differences (including response style) attenuated associations between work characteristics and strain outcomes, in all three samples work characteristics jointly contributed to the explanation of job satisfaction and psychological distress. More specifically, perceptions of control and social support were associated with job satisfaction, while work demands played a significant role in explaining symptom report. Furthermore, symptom levels predicted work absence in a combined sample of retail employees (N=221). In addition to direct effects, mediation and moderation patterns were examined. Results suggested that work perceptions partially mediated relations between personality (specifically locus of control beliefs and neuroticism) and strain. A similar pattern of mediation was apparent for social support. However, evidence for moderation of the stress process by individual and work characteristics was equivocal. Cross-sectional analyses using the manager sample data revealed significant interactions of control and Type A behaviour with work characteristics, but these interactions were not consistent in form with theory and past research. In contrast, in the longitudinal relocation study interactions were consistent with expectations: the negative impact of change was buffered by social support from senior colleagues and perceptions of control. Longitudinal analyses also demonstrated negative effects of workplace change; decreasing job satisfaction was observed during organisational restructuring, and transfer to a new branch was associated with increased psychological distress. Furthermore latent variable models revealed that changes in support, role ambiguity, and control over time were predictive of changes in job satisfaction. Overall, the present studies emphasise the need to examine simultaneously the joint influences of individual and work characteristics in occupational stress research. In this way the direct and indirect mechanisms through which psychosocial factors influence strain may be more fully understood, and strain-reducing interventions devised.