Citation

Carroll PS. Accid. Anal. Prev. 1973; 5(2): 81-94.

Copyright

(Copyright © 1973, Elsevier Publishing)

DOI

unavailable

PMID

unavailable

Abstract

Analysis of driving exposure and accident rates will be important in the future for evaluations of highway safety countermeasures. Each should be uniquely classified according to variables of the highway transportation system, i.e. driver-vehicle-road-environment combinations, in order to allow analyses which deal with the most significant differences in exposure and accident rate. Concepts of driving exposure are discussed, and a working definition is presented.A pilot survey of exposure was conducted, based on a random national sample of 7145 licensed drivers. Data was analyzed using the Automatic Interaction Detector (AID) algorithm, and hierarchies were produced of best predictors of accidents, exposure, and accident rate. The best predictor of exposure was whether or not a person drives on the job. When variables are limited to those appearing on accident reports, Driver Sex and Type of Vehicle were the best predictors, followed by Percent Driving on Streets, Model Year of Vehicle, and Driver Age. The best predictor of accidents was Driver Age, followed by Driver Sex and other variables in different order than the exposure hierarchy.