TRAFFIC ACCIDENTS FREQUENCY AND THE HUMAN FACTOR

This paper attempts to investigate the effect of the human factor on the frequency of traffic accidents made over a specified period of time. A disaggregate right-truncated Poisson regression model was formulated to model the frequency of a driver's traffic accidents over a five year period. Traffic accidents in this paper are viewed as discrete random events resulting from a number of causal factors. Estimation results based on disaggregate data set underscore the appropriateness of the probabilistic approach and suggest that both the driver's socioeconomic characteristics and daily usage of vehicle play a dominant role in traffic accident occurrencs. (A)