Re-investigation of the Mediating Effect of Brain Activities between Dispositional Factors and Hazard Recognition: A Multilevel Logistic Regression Approach

Understanding the relationships among dispositional factors, brain activities, and hazard recognition is important for developing auxiliary safety technologies and improving construction safety performance. However, current logistic statistical models have embedded potential biases into the interpretation of brain activities. The purpose of this study was to re-investigate the mediation effect of brain activities between dispositional factors (risk propensity, work experience, and injury history) and hazard recognition performance (HRP). An experiment was designed in which 71 construction workers each completed 120 hazard recognition trials (N = 8520), while the research team used an electroencephalogram recorder to record the brainwaves of the participants. The data were analyzed by the multilevel logistic regression method, and the results showed that the brain activities in FP1, FP2, and AF3 significantly mediated the relationship between dispositional factors and HRP. The findings theoretically supplement the understanding of brain activities in different hazard cognitive stages for hazard recognition. The results should contribute to the development of neuro-engineering technology and safety management strategies.