Modelling negative affectivity and job stress: a contingency-based approach

Understanding the role of negative affectivity (NA) in relations between job stressors and resultant strains has been a major source of research interest in the organizational literature for almost two decades. In this study, we propose a contingency approach, whereby the role of NA depends on the nature of the strain construct under investigation. Specifically, we predicted that perceived job stressors would fully mediate the relationship between NA and turnover intentions, while job stressors would only partially mediate the relationship between NA and job and life satisfaction, and job stressors would be unrelated to depression once NA was taken into account. The relative fit of these four models is tested utilizing data from two divergent samples (sales representatives and managers) employing both self- and informant reports of strain measures where feasible. Hypotheses regarding the specified nature of the NA --> stressor --> strain relationship generally were supported and replicated across samples. In addition, an alternate model specifying that NA moderates the relationship between stressors and strains was not supported for any strain variable. Copyright (C) 2004 John Wiley Sons, Ltd.