Transportation agencies throughout the United States have been steadily designing and implementing a variety of alternative intersections and interchanges (AIIs) to address unique safety and operational problems. However, despite the growing number of AIIs, high-quality crash prediction models and crash modification factors (CMFs) used to quantify the benefits of these designs remain limited.

NCHRP Research Report 1195: Developing Crash Modification Factors for Alternative Intersections: A Guide, produced by TRB’s National Cooperative Highway Research Program, includes an abbreviated literature review focusing on the three types of AIIs studied and potential analysis methods, as well as a discussion of data collection and analytical approaches. It summarizes the results of CMF development and presents a case study on the use of safety performance functions (SPFs) in future analyses. The report also presents the movement-based SPF framework and results, documents a microsimulation case study, and offers suggestions for implementing the CMF findings and conducting future research.

The Highway Safety Manual (HSM; AASHTO, 2010) is described by the American Association of State Highway and Transportation Officials (AASHTO) as “the premier guidance document for incorporating quantitative safety analysis in the highway transportation project development process.” The current HSM (1st edition) was published in 2010, with an update expected to be released. Driver behaviors, including speeding, distraction, and drowsiness, are widely recognized as major contributors to roadway crashes, and the HSM devotes a full chapter to providing detailed background and the core elements on human factors that affect the interaction of drivers and roadways. However, the quantitative components of the HSM seldom make the contribution of driver behavior explicit.

NCHRP Web-Only Document 474: Incorporating Driver Behavior and Characteristics into Safety Prediction Methods, produced by TRB’s National Cooperative Highway Research Program, explores whether speed data and driver behavior-related elements documented in crash reports could be used as direct measures to improve crash prediction beyond existing HSM methods. The report suggests that current HSM Safety Performance Functions (SPFs) already capture much of the influence of driver behavior indirectly through roadway geometry, traffic exposure, and other infrastructure characteristics. Although the project did not find sufficient evidence to support the development of new behavior-based SPFs for the HSM, it advanced safety research and practice by developing the Probabilistic Induced Exposure (PIE) methodology and strengthening confidence in the current HSM methodology.