This paper investigates the inclusion of the spatial effects in accident prediction models. Two types of spatial modeling techniques, namely, the Gaussian conditional auto-regressive (CAR) and the multiple membership (MM)models, were compared to the traditional Poisson log-normal (PLN) model. A variation of the MM model (Extended MM or EMM) was also investigated to study the effect of clustering segments within the same corridor on spatial correlation. Full Bayes estimation was used by means of the Markov ChainMonte Carlo methodology to estimate the parameters. The study made use of281 urban road segments in Vancouver, British Columbia. Various traffic and geometric variables were included in the accident prediction models. The models were compared in terms of their goodness of fit and inference. For the dataset under consideration, the results showed that Average Annual Daily Traffic (AADT), business land use, the number of lanes between signals and the density of unsignalized intersections have significant positiveimpact on the number of accidents. The fitted CAR and MM models demonstrated significant estimates for both heterogeneity and spatial correlation parameters. Furthermore, a significant portion of the total variability wasexplained by the spatial correlation. A significant correlation was also found between the heterogeneity and spatial effects. This may be due to neighboring road segments typically having similar environmental and geographic characteristics, thereby forming a cluster with similar accident occurrence. It turned out that corridor variation was a major component of total variability and that the spatial effects have been considerably alleviated by clustering segments within the same corridor. The best fit model wasEMM followed by CAR.
Abstract