As the science of safety and the state of the practice of data-driven safety analysis have advanced and are integrated into transportation agency practices, users have occasionally encountered challenges understanding the results of predictions when comparing different models. The Highway Safety Manual (HSM) first edition, used broadly by agencies for decisions on planning, design, and operational safety, includes crash predictive methods for intersections on various rural and urban facility types. Subsequent NCHRP projects (17-58, 17-68) developed additional predictive methods that will be included in the HSM’s forthcoming second edition. These predictive tools guide agencies evaluating the safety performance of different traffic control configurations. 

At times, the HSM and other models yield "counterintuitive" (though not necessarily incorrect) results when comparing crash outcomes across a few different intersection traffic control types, particularly for fatal and injury crashes. While the updated HSM will improve upon some crash prediction models and safety performance functions (SPFs), it still lacks the necessary tools and adjustments for accurately predicting safety impacts when converting between facility types. Research is needed to investigate and understand differences in predictive safety outcomes across different traffic control types (stop, signalized, and roundabout) at intersections and ramp terminals.

This research will identify and develop a guide for how to use improved crash prediction models and other tools. This research will address inconsistencies in the HSM, particularly for intersections where unexpected or conflicting safety results have been observed when comparing results for different intersection control types. Expected outputs include validated crash modification factors (CMFs), optimized SPFs, and comprehensive guidance for intersection control evaluations (ICEs) to support informed decision-making by agencies.