In this study the author assesses the accuracy and completeness with which Police Accident Reports(PARs) are completed by state and local police. Focusing on data elements commonly used in studies carried out by the National Highway Traffic Safety Administration (NHTSA), the author compared data from PARs to data collected in independent reinvestigations of the same crashes by NHTSA's Crashworthiness Data System (CDS) researchers. A total of 939 crashes in 1991 from four CDS study sites were investigated. In all 24 data elements were studied. For four data elements, very low discrepancy rates were found. For sixteen data elements, discrepancy rates varied by jurisdiction -- high in some, low in others. For four data elements, uniformly high discrepancy rates were found. These four data elements were manner of collision, vehicle identification number, occupant protection system use, and deployment of air bags (however the latter was treated as a separate data element in only one jurisdiction). The study results will be of interest to analysts relying on PAR-reported data. In addition, the author gives various recommendations for the improvement of PAR data quality.

  • Corporate Authors:

    Applied System Technologies, Incorporated

    6110 Executive Boulevard, Suite 900
    Rockville, MD  United States  20852

    National Highway Traffic Safety Administration

    1200 New Jersey Avenue, SE
    Washington, DC  United States  20590
  • Authors:
    • Greenberg, L
  • Publication Date: 1996-10


  • English

Media Info

  • Features: Appendices;
  • Pagination: 93 p.

Subject/Index Terms

Filing Info

  • Accession Number: 00732403
  • Record Type: Publication
  • Report/Paper Numbers: HS-808 487, Final Contractor Report
  • Contract Numbers: DTNH22-94-D-07087
  • Created Date: Feb 18 1997 12:00AM