Citation

Cambois MA, Fontaine H. Accid. Anal. Prev. 1982; 14(5): 387-396.

Copyright

(Copyright © 1982, Elsevier Publishing)

DOI

unavailable

PMID

unavailable

Abstract

Because it makes possible the development and evaluation of highway safety countermeasures, research on accident factors forms one of O.N.S.E.R.'s chief objectives. This research can be carried out by identifying classes of "driver-vehicle-road-environment" combinations which show especially high risk. The risk is estimated from the number of traffic-accident involvements in a given class divided by the corresponding exposure value for the same class. The measure of exposure used here is driving distance expressed in vehicle-kilometers travelled which can be drawn from several types of surveys that O.N.S.E.R. undertakes in addition to the "Traffic Safety Monitor". This latter is a periodical survey based on roadside observation of vehicles; the data obtained concern the environment, the weather at the observation post and vehicle types and speeds, as well as seat belt wearing. An exposure value survey was conducted at service stations during the year July 1973 to June 1974 which provided data on the vehicle and the driver. From these data and from the national file recording all injury-producing accidents, a study was carried out to determine the groups with the highest risk and to classify them. The aim was to rank the variables in terms of their impact upon risk, and for that purpose, a method was used which is similar to segmentation. The selection test was to choose the variable giving the greatest distance between the distribution of those involved in accidents and of exposed drivers. The present paper aims, on the one hand, at detailing the contents of several different traffic surveys giving an evaluation of vehicle-kilometers travelled and, on the other hand, at commenting on the results of the study of risk resulting from the comparison of various survey data.