Relation of job stressors to affective, health, and performance outcomes: A comparison of multiple data sources.

It is widely accepted that job conditions are a causal factor in stress outcomes for employees. This conclusion, however, is based almost entirely on single data source, self-report studies, which demonstrate correlations between environmental perceptions and stress outcomes. This study collected stressor data from two sources, the job incumbent and her supervisor. Convergent and discriminant validities were found for four stressors (autonomy, workload, number of hours worked, and number of people worked for) but not for three others (role ambiguity, constraints, and interpersonal conflict). Correlations were found between perception of stressors and outcomes, the latter including both affective and symptoms. Smaller correlations were found between supervisor reports of stressors and outcomes, the latter including both affective and symptoms. Alternative causal models relevant to these results are discussed. The need for causal research including experimental designs, longitudinal designs, and multiple data sources are also discussed.