General Job Stress: A Unidimensional Measure and Its Non‐linear Relations with Outcome Variables

Abstract

This article aims to examine the non‐linear relations between a general measure of job stress [Stress in General (SIG)] and two outcome variables: intentions to quit and job satisfaction. In so doing, we also re‐examine the factor structure of the SIG and determine that, as a two‐factor scale, it obscures non‐linear relations with outcomes. Thus, in this research, we not only test for non‐linear relations between stress and outcome variables but also present an updated version of the SIG scale. Using two distinct samples of working adults (sample 1, N = 589; sample 2, N = 4322), results indicate that a more parsimonious eight‐item SIG has better model‐data fit than the 15‐item two‐factor SIG and that the eight‐item SIG has non‐linear relations with job satisfaction and intentions to quit. Specifically, the revised SIG has an inverted curvilinear J‐shaped relation with job satisfaction such that job satisfaction drops precipitously after a certain level of stress; the SIG has a J‐shaped curvilinear relation with intentions to quit such that turnover intentions increase exponentially after a certain level of stress. Copyright © 2011 John Wiley & Sons, Ltd.