Effects of State-Specific SPFs, AADT Estimations, and Overdispersion Parameters on Crash Predictions Using SafetyAnalyst

SafetyAnalyst performs network screening using empirical Bayes approach which requires safety performance functions (SPFs) in addition to roadway characteristics, traffic, and crash data. Simple SPFs (i.e. crashes are predicted as a function of traffic alone) are used within SafetyAnalyst. These simple SPFs were generated using CA, MN, NC, OH, and WA data. Within the software, these default SPFs are calibrated to represent the agency’s data. How well these calibrated default SPFs fit another state’s data is a question yet to be answered. This research aims at comparing the fit of Georgia specific SPFs and the default SPFs calibrated to Georgia data. As hypothesized, it was found that state specific SPFs fit the data well, however, the accuracy of traffic data significantly affected the fit of SPFs. SPFs developed using longer segments and with actual traffic data have lower overdispersion parameter with minimal amount of unaccountable dispersion than the SPFs developed using segments with both actual and estimated traffic data.

Language

  • English

Media Info

  • Media Type: Digital/other
  • Features: Figures; References; Tables;
  • Pagination: 22p
  • Monograph Title: TRB 91st Annual Meeting Compendium of Papers DVD

Subject/Index Terms

Filing Info

  • Accession Number: 01372423
  • Record Type: Publication
  • Report/Paper Numbers: 12-4332
  • Files: TRIS, TRB
  • Created Date: Jun 14 2012 10:55AM