Analysis of Factors Affecting the Severity of Motorcycle Casualties in Phnom Penh Using a Bayesian Approach

抄録

The number of motorcyclist fatalities in Cambodia accounted for almost 70% in 2010 and 2011. The contributing factors to the severity of motorcycle casualties should be identified to provide worthwhile information for planning or policy making to reduce the severity of motorcyclist crashes. The aim of this research study is to analyze the factors affecting the severity of motorcycle casualties in Phnom Penh, the capital city of Cambodia, using a Bayesian approach to apply the ordered probit (OP) model, which is called the Bayesian OP (BOP) model. The advantage of the BOP approach is to allow the researchers to use prior information about the explanatory variables in fitting the models. Unlike OP, BOP produces reasonable estimated coefficients compared with OP when the sample size is small. This study found that male drivers, middle-aged groups (25 to 59 years), speeding, nighttime, peak hour, weekends, heavy truck crash opponent, crashing alone, and head-on collisions are more likely to result in higher levels of injury severity relative to its reference case.