About the Event
In recent years increased attention has been given to predicting the effects of roadway improvements on traffic safety. Tools have been developed in cooperation with the Federal Highway Administration and the Transportation Research Board that attempt to predict crash experience; these tools require estimates of crash modification factors (CMF) to produce predicted reduction benefits. The tools make use of empirical Bayes statistics, which currently require that crash data be overdispersed. This research illustrates an alternative method for estimating CMFs that can be applied whether the crash data are overdispersed or not. The method combines the hierarchical Bayes model with a model that allows for temporal changes in the covariates. The method was used to compute estimates for the CMFs associated with signalizing or changing the left-turn phasing of sets of non-rural intersections. This research was especially interested in the effect on left-turn crashes. Marginally significant results included: signalization does not produce a definite effect on major left-turn crashes; phase-changes on the major approaches from permitted/protected to protected phasing decrease major approach left-turn crashes; and phase-changes on the minor approaches from permitted to permitted/protected do not significantly affect the minor approach left-turn crashes.