Motorcycle accident prediction model for junctions on urban roads in Malaysia

This paper presents a prediction model for motorcycle accidents at junctions on urban roads in Malaysia. The models were developed using the Generalized Linear Modeling approach. The final model reveals that motorcycle accidents are proportional to the rate of traffic flow. The estimates indicate that an increase in non-motorcycles and motorcycles entering a highway interchang is associated with an increase in motorcycle accidents. Non-motorcycle flow on major roads had the highest effect on the probability of accidents involving motorcycles. Approach speed, lane width, number of lanes, junction control, shoulder width, and land use were also found to be significant in explaining motorcycle accidents. The final model should allow traffic engineers to establish appropriate highway junction treatment that is specifically designed for motorcycle lane facilities.