Road traffic crashes are count (discrete) in nature. When modeling discrete data for characteristics and prediction of events, it is appropriate using the Poisson Regression Model. However, the condition that the mean and variance of the Poisson are equal, poses a great constraint, hence necessitating the use of the Generalized Poisson Regression (GPR) and the Negative Binomial Regression (NBR) models, which do not require these constraints that the mean and the variance be equal, as proxies. Data on Road traffic crashes from the Anambra State Command of the Federal Road Safety Commission (FRSC), Nigeria were analyzed using these three methods, the results from the two proxies are compared using the Akaike Information Criterion (AIC) with GPR showing an AIC value of 3508.595 and the NBR showing an AIC value of 2742. Having shown a smaller AIC value, the NBR was considered a better model when analyzing road traffic crashes in Anambra State, Nigeria.
Keywords: Akaike Information Criterion, Crashes, Discrete, Over-dispersion, Road Traffic
