STRESS AND TURNOVER INTENTION - A COMPARATIVE-STUDY AMONG NURSES

A three-stage linear model of turnover with role ambiguity, role conflict, and role overload as antecedents and stress as an intervening variable was constructed to guide this research. In addition, the moderator effects of external and internal opportunities, social support, and personal experience on the stress-turnover linkage were examined. Two samples of data were collected from nurses working in both general hospitals (N = 689) and specialized hospitals (N = 441) in the Greater Montreal area, through a field survey. Structural modelling using the LISREL technique was employed to examine the fit between the proposed model and the data. In addition, data from American nurses collected by Bedeian and Armenakis were reanalyzed using the proposed model to test for generalizability of our findings across different national groups. The proposed model was found to exhibit a better fit for both the Canadian and American data. All the hypothesized role stressors were significant predictors of stress. While the stressors failed to predict turnover intention consistently across samples, stress in all samples yielded significant predictions of turnover intention. However, none of the proposed moderating variables showed significant effects on turnover. The implications of these findings for future research on turnover is discussed.