A random parameter bivariate probit model for injury severities of riders and pillion passengers in motorcycle crashes

This study proposes a random parameter bivariate probit model to analyze risk factors on the crash injury severity of both motorcycle riders and passengers in a single modeling framework. The proposed model can not only account for the underlying correlation of common factors affecting the rider and its pillion passenger simultaneously, but also can capture the unobserved heterogeneity across crash samples. The case analysis is based on 3665 motorcycle-carrying-passenger crashes in Hunan province of China. Model comparisons show that the proposed random parameter bivariate probit model outperforms two conventional models in the goodness-of-fit. The results of parameter estimations show that, age and gender differences in passengers pose significant effects on injury severity of the rider in crashes. Specifically, when carrying a vulnerable passenger including women, children and elders, the rider is less likely to sustain severe injuries. But for injury severity of the passenger, these vulnerable passengers are more likely to suffer from severe injuries. Apart form age and gender attributes, factors including collision objects, helmet use, drunk riding, night without lights, peak periods, high-speed roads have significant effects on rider injury and/or passenger injury. Relevant suggestions to alleviate the injury severity for motorcycle-carrying-passengers crashes are recommended.