Estimation of safety performance functions for urban intersections using various functional forms of the negative binomial regression model and a generalized Poisson regression model

Intersections are established dangerous entities of a highway system due to the challenging and unsafe roadway environment they characterized for drivers other road users. In efforts improve safety, an enormous interest has been shown in developing statistical models intersection crash prediction explanation. The selection adequate form model is great importance accurate estimation frequency correct identification contributing factors. Using six-year data, infrastructure geometric design traffic flow data urban intersections, we applied three different functional forms negative binomial (i.e., NB-1, NB-2, NB-P) generalized Poisson (GP) develop safety performance functions (SPF) by severity signalized unsignalized intersections. This paper presents relationships found between explanatory variables expected frequency. It reports comparison total, injury & fatal, property damage only crashes order obtain ones with maximum accuracy. was based on goodness fit measures. fitted showed that several related significantly influence Further, measures revealed NB-P outperformed most levels For GP best performing model. When NB were compared, performed better than traditional NB-1 and, more specifically, NB-2 models. conclusion, our findings suggest potential improvement accuracy SPFs intersections applying