A Fuzzy-set Approach to Re-Exploring Work-related Stress-outcome Relationships: Implications for Research Methods, Theory and Practice

Abstract

This study employs fuzzy-set Qualitative Comparative Analysis to re-examine the more traditional correlational and regression models used to understand work stress and outcome relationships. Indeed, theoretical frameworks have been constructed underlying this premise and so have interventions assumed such principles. This study argues that the relationship between perception of stressors is not necessarily continuously linear and hence assessing non-linear but also more complex configurations becomes pivotal. While it is acknowledged that theory is important to simplify the explanation underlying the phenomenon of interest, likewise theory should also acknowledge new patterns that hide ‘inside’ the data and which reveal new pockets of information that explain the phenomenon better. Therefore, in line with calls for researchers to explore further new configurational patterns in the data and thus help build and refine theory, we examine potential complex associations between work-related stress and salient outcomes namely mental health, physical symptoms, presenteeism and intention to quit using fuzzy-set Qualitative Comparative Analysis instead of the more traditional correlational approach. Fuzzy-set methodology involves a number of parameters that include non-linearity but also equifinality and asymmetric associations. Reponses were obtained from 509 employees who volunteered to fill in a work-related stress questionnaire as part of a wider campaign related to the management of psychosocial hazards at work and the data was subject to fuzzy-set configuration analysis between stressors and outcomes. Fuzzy-set analyses revealed that specific points of high or low levels of work stress did not necessarily concur with more common regression results but revealed interesting pattern associations between work related stress and outcomes. The results imply that by adopting traditional methods, the relations observed would be either over- or under- estimated and researchers are urged to explore new research methodological approaches such as fuzzy-set methodology in the analyses of their data that may reveal new patterns and hence better inform both theory and practice.