Modeling Motorcycle Road Accidents with Traffic Offenses at Several Potential Locations using Negative Binomial Regression model in Malaysia

摘要

The major contributor to the total number of road accidents in Malaysia is the motorcycle road accidents. In this study, the relationship between motorcyclist road traffic offenses and the number of accidents at several places will be determined using the statistical generalized linear model. Negative binomial regression analysis was used as an alternative model other than Poisson. The model was validated using the Pearson chi-square tests, the method of deviance, the Akaike information criterion, and the Bayesian information criterion. The goodness of fit test shows that the negative binomial model is the best model, and it helps with overcoming the overdispersion problem, resulting in better estimation. The results also indicate that "turning dangerously" was the most common traffic offense of motorcyclists. The most likely location for motorcycle accidents to happen is in a "residential area".