The primary objectives of this study were to examine characteristics of drivers involved in fatal accidents and to determine if those drivers could be distinguished from California's general driving population on the basis of prior driving record. A sample of drivers involved in 1970-71 fatal accidents was analyzed and compared to a sample of drivers from the general driving population during the same time period. Drivers who had been drinking prior to the accident, who were considered at-fault for the accident, or whose accident occurred at night were found to have worse prior driving records than other fatal accident-involved drivers. The results also indicated that, as a whole, drivers involved in fatal accidents had worse violation and/or accident records, as well as different demographic and license characteristics than drivers in the general population. The classification functions derived to predict fatal accidents, however, did not differ greatly from regression equations that have been constructed to predict total accidents. It was therefore concluded that prediction systems keyed to total accidents will, to a large extent, also identify high-risk fatal accident drivers.
Abstract