Citation

Mountain L, Fawaz B, Jarrett D. Traffic Eng. Control 1995; 36(12): 673-679.

Affiliation

Univ of Liverpool, UK

Copyright

(Copyright © 1995, Hemming Group)

DOI

unavailable

PMID

unavailable

Abstract

The rare and random nature of accident occurrence leads to considerable difficulties for engineers in both the proper selection of candidate sites for remedial treatment and in the correct evaluation of the effectiveness of any safety measures implemented. An alternative to the current methods used is one in which the site selection and evaluation are based on better estimates of the long-term mean accident frequencies at candidate sites that can be achieved from accident counts alone. Another alternative is to base estimates on both sources of information by using an empirical Bayes method. The extent to which knowledge of the locations of minor junctions could improve the estimates of underlying mean accident frequencies is evaluated.