A system dynamics approach for modeling construction workers' safety attitudes and behaviors.

Construction accidents are caused by an unsafe act (i.e., a person's behavior or activity that deviates from normal accepted safe procedure) and/or an unsafe condition (i.e., a hazard or an unsafe mechanical or physical environment). While there has been dramatic improvement in creating safer construction environments, relatively little is known regarding the elimination of construction workers' unsafe acts. To address this deficiency, this paper aims to develop a system dynamics (SD)-based model of construction workers' mental processes that can help analyze the feedback mechanisms and the resultant dynamics regarding the workers' safety attitudes and safe behaviors. The developed model is applied to examine the effectiveness of three safety improvement policies: incentives for safe behaviors, and increased levels of communication and immersion in accidents. Application of the model verifies the strong potential of the developed model to provide a better understanding of how to eliminate unsafe acts, and to function as a robust test-bed to assess the effectiveness of safety programs or training sessions before their implementation.