Evaluation of Four Network Screening Performance Measures

2016Thompson, StuartPDFTech ReportFHWA-SA-16-103DTFH61-10-D-00022-T-13005Network analysis (Planning)Performance measurementHighway safetyData analysisUnited StatesUnited States. Federal Highway Administration. Office of SafetyGross, Frank B.Harmon, TimAlbee, MattHimes, ScottSrinivasan, RaghavanCarter, DanielDugas, MichaelUnited States. Federal Highway Administration. Office of SafetyVHB/Vanasse Hangen Brustlin, Inc.Thompson, StuartFederal Highway AdministrationHigh quality data and reliable analytical methods are the foundation of data-driven decision-making. The Reliability of Safety Management Methods series includes five information guides that identify opportunities to employ more reliable methods to support decisions throughout the roadway safety management process. Four of the guides focus on specific components of the roadway safety management process: network screening, diagnosis, countermeasure selection, and safety effectiveness evaluation. The fifth guide focuses on the systemic approach to safety management, which describes a complimentary approach to the methods described in the network screening, diagnosis, and countermeasure selection guides. The purpose of the Reliability of Safety Management Methods series is to demonstrate the value of more reliable methods in these activities, and demonstrate limitations of traditional (less reliable) methods. The Evaluation of Four Network Screening Performance Measures is a supplement to the network screening guide, and also serves as an example of diagnosis and countermeasure selection. The objectives of this guide are to 1) raise awareness of more reliable network screening performance measures, and 2) demonstrate the value of more reliable performance measures using real-world data. The target audience includes data analysts, program managers, and project managers involved in projects that impact highway safety. The primary question related to network screening is \u201cwhich performance measure is most likely to produce a list of sites with the greatest potential for improvement and subsequently result in the greatest benefit and most cost-effective safety improvements?\u201d To answer this question, the study follows the safety management process from network screening through economic analysis, using statewide intersection data from New Hampshire. The dataset includes fatal and injury crash data, traffic data, and roadway data for years 2010 through 2014. The research team identified ranked lists of sites based on four network screening performance measures, performed a safety diagnosis on the top 35 intersections, developed potential strategies to target the underlying safety issues, and conducted an economic analysis for each intersection improvement package. This guide compares the overall economic benefit and overall benefit-cost ratio for each of the four measures. The EB excess expected measure and the EB expected measure produced the list of sites with the highest overall economic benefit and the highest return on investment, respectively. The crash rate measure produced the lowest overall economic benefit and the lowest return on investment. Readers will understand the value of and be prepared to select more reliable performance measures in network screening to account for potential bias, obtain more reliable results, and achieve decisions that are more effective.84