Evaluation and enhancement of accident prediction models and accident modification factors of rural intersections

According to the Traffic Safety Facts 2001, there were 6,322,795 traffic crashes in 2001 in the United States. Of these, highway accidents accounted for almost 87 percent of traffic facilities, or about 37,000 deaths annually. It is imperative that transportation researchers and academicians understand the complex relationships between highway safety and roadway design. Despite decades of research on the subject, major research gaps still remain regarding state-of-the-knowledge in highway safety—and without exception the relationships between roadway design elements and their impacts on highway safety. The primary reasons for the remaining knowledge gaps are: (1) the observational setting characterized by lack of controls under which academicians and researchers examine highway safety; (2) the small magnitude of engineering countermeasure effects; (3) the lack of uniformity in application of analytical methods; (4) the random nature and small magnitudes of crash rates; (5) biases resulting from non-random samples, regression to the mean, and omitted variables; and (6) the lack of available tools for enabling proactive safety planning (instead of reactive safety planning implemented routinely throughout the United States).