A Heckman selection model for the safety analysis of signalized intersections

Abstract

Purpose: The objective of this paper is to provide a new method for estimating crash rate and severity simultaneously. Methods: This study explores a Heckman selection model of the crash rate and severity simultaneously at different levels and a two-step procedure is used to investigate the crash rate and severity levels. The first step uses a probit regression model to determine the sample selection process, and the second step develops a multiple regression model to simultaneously evaluate the crash rate and severity for slight injury/kill or serious injury (KSI), respectively. The model uses 555 observations from 262 signalized intersections in the Hong Kong metropolitan area, integrated with information on the traffic flow, geometric road design, road environment, traffic control and any crashes that occurred during two years. Results: The results of the proposed two-step Heckman selection model illustrate the necessity of different crash rates for different crash severity levels. Conclusions: A comparison with the existing approaches suggests that the Heckman selection model offers an efficient and convenient alternative method for evaluating the safety performance at signalized intersections.