Sensitivity analysis of unsafe behaviors in the spinning and weaving factories: Exploring the association with burnout and resilience using Bayesian networks

Job burnout and resilience skills are factors that can affect safety performance in the workplace. However, the contribution of these variables to unsafe behaviors through various paths has not been determined. This study aimed to investigate the association of three burnout dimensions and resilience with safety compliance and safety performance using Bayesian network modeling. This research was performed with cross-sectional design. Participants were 200 employees working in some spinning and weaving factories. Participants provided responses to printed survey items during work rest periods. The survey comprised a demographic information section, validated Persian versions of the Connor–Davidson resilience scale, the Maslach burnout questionnaire, and the safety behavior assessment. The Bayesian network was analyzed using version 2.3 of the GeNIe academic software. At the high state with a probability of 100% for each of the three burnout variables: depersonalization, emotional exhaustion, personal accomplishment, and (poor) resilience, the probability of poor safety compliance increased by 16%, 16%, 7%, and 24% and the probability of poor safety participation rose by 6%, 12%, 29%, and 17%, respectively. All variables with a probability of 100% also elevated the likelihood of diminished safety compliance and reduced safety participation by 51% and 34%, respectively. Each of the three dimensions of burnout can be associated with changes in resilience, safety compliance, and safety participation. Resilience plays a significant role in mediating the association between burnout dimensions and unsafe behaviors.