Effects of traffic signal installation on intersection crashes

Traffic signals have been considered a way to improve traffic safety and traffic operations at intersections. However, the impacts of intersection signalization on crashes are complicated and have not been fully investigated. These include the impacts on number of crashes and crash rates for all crashes, type of crashes, and crash severity. This research performed a study of 502 newly signalized intersections in Florida to evaluate the effects of signalization on crashes at intersections. Average annual number of crashes and crash rates were compared to explore the effects of signalization on intersection crashes. Paired t-test was employed to determine if there was a statistically significant difference between the before and after periods. Results indicated that all crashes, rear-end crashes and all crashes would increase for number of crashes and crash rates after signal installation. Angle and left-turn crashes would decrease in reference to number of crashes and crash rates. Right turn crash rate would decrease, and number of sideswipe crashes would increase. Regarding crash severity, number of no-injury crashes and no-injury crash rate would increase, fatal crash rate would decrease, and number of injury crashes would increase. Regarding land use, results indicated that crash rates would decrease in rural areas, and number of crashes and crash rate would increase in urban areas. The research also fitted Poisson and Negative Binomial (NB) distributions to observed crash distributions. Results indicated that NB distribution was the best fitted distribution for crash data.