PREDICTING ACCIDENT TRENDS AND TRAFFIC IMPROVEMENT

A PILOT STUDY IS PRESENTED ON ACCIDENT-TREND ANALYSIS. A METHOD IS PROPOSED WHEREBY ACCUMULATED ACCIDENT DATA MAY BE ASSEMBLED TO BE OF VALUE IN EVALUATING ANY SET OF CONDITIONS OR PROGRAM DESIGNED TO REDUCE TRAFFIC ACCIDENTS. LINEAR, LOGARITHMIC AND RECIPROCAL TIME SERIES TRENDS WERE CALCULATED FOR EACH OF TWO AREAS OF STUDY. THE PREDICTION OF TRAFFIC DEATH RATES SHOULD NOT EXCEED A PERIOD OF MORE THAN 10 YEARS FOR PRACTICAL PURPOSES. A RECIPROCAL TREND LINE WAS DEVELOPED BY THE METHOD OF LEAST SQUARES FOR THE SAME TWO GROUPS. A LOGARITHMIC CURVILINEAR RELATIONSHIP WAS CONSIDERED WHEREBY AN IRREDUCIBLE MINIMUM WAS DETERMINED. THE USE OF ONE OF MORE OF THESE OR SIMILAR TECHNIQUES OF EVALUATING TRAFFIC DEATH RATE TRENDS MIGHT BE USED BY ENFORCEMENT AGENCIES, TRANSPORTATION OR INSURANCE COMPANIES, AND HIGHWAY ENGINEERS IN FURTHER APPRAISING PHASES OF THEIR RESPECTIVE PROGRAMS.

Media Info

  • Monograph Title: Investigating and forecasting traffic accidents
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 00220317
  • Record Type: Publication
  • Files: TRIS, TRB
  • Created Date: Apr 21 1994 12:00AM