Analysis of Red Light Running Crashes Based on Quasi-Induced Exposure and Multiple Logistic Regression Method

Abstract

According to recent national statistics, red light running crashes represent a significant safety problem at signalized intersections. To examine the overall characteristics of red light running crashes, this study used the 1999 to 2001 Florida crash database to investigate the crash propensity related to traffic environments, driver characteristics, and vehicle types. The quasi-induced exposure concept and multiple logistic regression technique were used to perform this analysis. The results showed that traffic factors including number of lanes, crash time, weather, highway character, day of week, urban or rural location, speed limit, driver age, alcohol or drug use, physical defect, driver residence, and vehicle type were significantly associated with the risk of red light running crashes. Furthermore, it confirmed that there were significant interaction effects between the risk factors, including crash time and highway character, number of lanes and urban or rural location, weather condition and driver age, driver age and gender, alcohol or drug use and gender, and type of vehicle and gender.