In 1971 and 1972 the Pennsylvania State Police collected special data on 15,415 automobile accidents in rural Pennsylvania. The data collected included both items routinely obtained by the police and supplemental items selected by the National Highway Traffic Safety Administration. Culpability for each accident was determined by the state police.
The data were analyzed to investigate the predictive ability of the following driver characteristics on culpability: age, sex, route familiarity, driving experience, and experience or familiarity with the accident vehicle. Iterative fitting methods of multivariate contingency table analysis were used to fit log-linear structural models to the data, to test the fit of these models, and to compare the predictive effect of these variables and their interactions. The results indicate that route familiarity is the strongest predictor, followed by age and driving experience. Driver sex and experience with the accident vehicle have only minimal predictive power. No interactions were significant: the route familiarity, age, and driving experience enter independently into the final model.
The findings suggest that drivers unfamiliar with their route may be more vulnerable to inadequate controls, poor highway design, information overload or other distractions at the time of the crash.
The pursuit of the culpable driver has long been with us. In 1927, the National Research Council sponsored a series of studies at Ohio State University on the psychology of the driver. These research efforts included accident investigations, comparisons of traffic law violators with control samples, laboratory experiments using driving simulators, and measures of vision and reaction time. As a result of these projects, Weiss and Lauer, in 1930(1) identified a small group of drivers who suffered more than their share of accidents.
Since that time research has continued to show that not all drivers run an equal risk of being involved in a traffic accident. Certain segments of the driving population are over-represented in collision occurrence statistics. Some authors attribute culpability for much of our accident losses to an easily identifiable group such as young drivers (2). Other approaches attribute the causes of accidents to driver characteristics found in most segments of the population (3) (4). The search for the culpable driver continues, and is of sufficient importance that the U.S. Department of Transportation's National Highway Traffic Safety Administration (NHTSA) has funded numerous human factors studies during the past decade. This paper compares some of the characteristics of the "culpable" driver with those of the "non-culpable" operator in a collection of Pennsylvania automobile accidents.