73-76 Virkler Feature (WEB)

73 WHILE REVIEWING VARIOUS high-crash-location (HCL) analysis guides, the authors identified some common misconceptions and areas needing further elaboration. In the companion feature “Improving the Identification, Analysis and Correction of High-Crash Locations,” the authors provide an overview of the HCL process and describe five important concepts for understanding and analyzing HCLs.1 These concepts relate to: • Accounting for crash severity; • Misapplying the normal distribution in describing expected crash rates at sites with a relatively low number of expected crashes; • Describing statistical significance; • Addressing the problem of “regression-to-the-mean”; and • Applying benefit/cost analysis to compare alternatives. The companion feature addresses the second and third concepts in a manner suitable for a broad range of transportation engineers. This feature describes these two concepts (misapplying the normal distribution and describing statistical significance) in more detail. It is intended for traffic-engineering professionals with a background in statistics or a desire to understand the concepts in more detail. Applications of the normal distribution to HCL studies are described below, followed by information describing the statistical significance of results one might generate in an HCL study.