Evaluating the Safety Effects of Signal Improvements

Abstract

Many fatal and injury crashes on roadways occur at intersections. Traffic signals are often installed or modified to reduce these severe crashes. Whether the traffic signal improvements help reduce severe crashes can be evaluated with crash modification factors (CMFs). Recent research showed that traditional safety evaluation methods were inadequate for developing CMFs. Bayesian statistical methods have been used in traffic safety studies to analyze the effectiveness of safety improvements more accurately. The hierarchical Bayesian method is an advanced statistical technique that can account for the shortcomings of traditional methods and more fully reflect the effectiveness of safety improvements. This study used a hierarchical Bayesian model to analyze the effectiveness of new traffic signal installations and modifications to existing traffic signals from permissive-only phasing to protected–permissive left-turn phasing through the development of CMFs for multiple scenarios. A benefit-to-cost (B-C) analysis was performed for each improvement to determine how long it would take to recover the cost of installation. The results showed a slight increase in overall and nonsevere crashes and a decrease in severe crashes for both new signal installations and modifications to existing signals. The B-C analysis indicated that a benefit came from both improvements and that new signal installation costs could be recovered in approximately 5 years and the cost of the installation of a left-turn signal modification could be recovered in approximately 9 weeks.