Abstract: Drivers' emotional states impair their cognitive abilities, such as driving decisions and concentrations. This increases the likelihood of traffic crashes. This study examines the effect of emotional attributes such as depression, anxiety, and anger on self-reported risky driving behavior based on data from 1,139 bus drivers from a bus company. From self-reported data, risky driving behaviors are classified into three distinct categories: low-risk, moderate-risk, and high-risk. A Bayesian Belief Network (BBN) is adopted to examine the relationships among random variables to represent causation and underlying determinants (i.e., causal probabilities) for different conditions of driving behavior. BBN predicted that high-risk drivers have a greater probability of suffering from severe depression, anxiety, and anger. Moreover, sensitivity analysis showed that depression, anxiety, anger, age group, gender, education level, insomnia, job satisfaction, driving experience, daily driving duration, and safety climate perception significantly influence risky driving behavior. As self-reported driving behavior may give bias results, to validate further, results from non-parametric tests revealed a significant difference in depression, anxiety, and anger scores between the drivers involved in self-reported crashes and those who were not involved. Emotional attributes are often difficult to distinguish. Thus, Spearman's correlation test was performed, and results showed that emotional attributes are positively correlated. The study concludes that these three emotional attributes impair bus drivers' driving behavior indicating the necessity of monitoring them continuously.
Shoumic Shahid Chowdhury, Xuesong Wang*, Mohammed Quddus, Siyang Zhang, Moinul Hossain. Examining the Effect of Emotional Attributes on Self-Reported Driving Behaviors: A Study on Bus Drivers from Suzhou, China. Transportation Research Board 101st Annual Meeting, Washington D.C., USA, 2022. 1.9-13.