Evaluation of Accident Analysis Methodology

Details

• Creators:
  Pendleton, Olga J.
• Corporate Creators:
  Texas A&M University System. Texas Transportation Institute
• Contributors:
  Griffith, Michael S.
• Corporate Contributors:
  United States. Federal Highway Administration. Office of Safety and Traffic Operations R&D
• Subject/TRT Terms:
  Bayes' Theorem Crash Analysis Raised Road Markings Signalization Traffic Safety
• Publication/ Report Number:
  FHWA-RD-96-039
• Resource Type:
  Tech Report
• Right Statement:
  Public Domain
• Geographical Coverage:
  United States
• Edition:
  Final Report
• Contracting Officer:
  Griffith, Michael S.
• Corporate Publisher:
  United States. Federal Highway Administration. Office of Safety and Traffic Operations
• Abstract:
  This study compared both traditional and empirical Bayes (EB) methodologies for ranking high hazard locations and evaluating highway safety treatments for five case studies. The case studies were either based on data collected for previous studies or data retrieved from the Highway Safety Information System (HSIS). These included two high hazard location ranking studies for New York and Illinois and three treatment evaluation studies: Michigan raised pavement markers, Michigan change in posted speeds, and Minnesota traffic signal installation. In addition, an extended empirical Bayes methodology was developed that allows for the use of explanatory variables (covariates) in a regression model along with a computer program for its implementation. The explanatory variables used in these case studies were roadway characteristics such as curvature, pavement width, shoulders, divided highway, and initial posted speed. The regression methodology also allowed for the examination of a non-linear relationship between vehicle-miles traveled (VMT) and accidents. There was strong evidence that such a non-linear relationship exists, casting doubt on the validity of accident rate-based methods. Although there was no significant difference among the methodologies, there was a significant increase in accidents for roads with initial posted speeds of <30 mi/h (48.3 km/h) in Michigan (23 to 27 percent) where speed limits were raised and a significant reduction in accidents (25 to 30 percent) in Minnesota after the installation of new traffic signals. Guidelines are presented for determining when the various methodologies are warranted and appropriate.
• Format:
  PDF
• Funding:
  DTFH61-92-R-00029
• Collection(s):
  Federal Highway Administration
• Main Document Checksum:
  urn:sha-512:a1b636742c0f8216840f889514aa5d6f0b6238448eec451ebc9f64fc1d2fbaf9a1615bae067706111fe1691ee8f94ffcf202347df2e615744e57c70b9c884b0e
• Download URL:
  https://rosap.ntl.bts.gov/view/dot/88409/dot_88409_DS1.pdf
• File Type:
  [PDF - 4.73 MB ]