A Bayesian spatial random parameters Tobit model for analyzingcrash rates on roadway segments
This study develops a Bayesian spatial random parameters Tobit model to analyze crash rates on roadsegments, in which both spatial correlation between adjacent sites and unobserved heterogeneity acrossobservations are accounted for. The crash-rate data for a three-year period on road segments within aroad network in Florida, are collected to compare the performance of the proposed model with that ofa (fixed parameters) Tobit model and a spatial (fixed parameters) Tobit model in the Bayesian context.Significant spatial effect is found in both spatial models and the results of Deviance Information Criteria(DIC) show that the inclusion of spatial correlation in the Tobit regression considerably improves modelfit, which indicates the reasonableness of considering cross-segment spatial correlation. The spatial ran-dom parameters Tobit regression has lower DIC value than does the spatial Tobit regression, suggestingthat accommodating the unobserved heterogeneity is able to further improve model fit when the spa-tial correlation has been considered. Moreover, the random parameters Tobit model provides a morecomprehensive understanding of the effect of speed limit on crash rates than does its fixed parameterscounterpart, which suggests that it could be considered as a good alternative for crash rate analysis.