Citation

Datta TK, Opiela KS, Smith RJ. Transp. Res. Rec. 1977; 643: 12-18.

Copyright

(Copyright © 1977, Transportation Research Board, National Research Council, National Academy of Sciences USA, Publisher SAGE Publishing)

DOI

unavailable

PMID

unavailable

Abstract

A Highway Safety Program Standard (Identification and Surveillance of High Accident Locations) requires each state and local community to have an established procedure for the identification of high accident locations. The standard, however, specifies neither definite criteria nor procedures for the identification of such locations. Consequently, communities throughout the country use a variety of identification methods with varying degrees of success and accuracy. These range from the accident-frequency method to the accident-rate method and various combinations of them. The result of these procedures is the identification and selection of the most critical accident locations. A methodology for the analysis of large numbers of locations has been developed and implemented in Oakland County, Michigan, as part of a countywide comprehensive traffic engineering project. The methodology uses both accident-frequency and accident-rate data for each intersection and highway link to identify the most critical locations. The procedure stratifies the data from a number of intersections (or links) and assigns each location to a cell within a matrix that considers accident frequency on the horizontal axis and accident rate on the vertical axis. The locations contained in the cell corresponding to the highest frequency and the highest rate are identified as the most critical locations. Locations with a high frequency and a low rate or a high rate and a low frequency are considered less critical. A computer program was developed that determines the rate and frequency for all highway locations (intersections or links) being analyzed, assigns each location to the appropriate cell in the rate and frequency matrix, and then prepares reports indicating the locations contained in each cell and the pertinent data for each location. The rate and frequency analysis procedure was tested by using countywide accident data, as well as data from smaller political jurisdictions, and was an effective and valuable traffic-engineering tool.