Basic driving behaviour (BDB) can be defined in terms of laws or legal norms, social norms and personal norms. A questionnaire survey was carried out on a sample of 1002 drivers, comprising questions relating to speeding, territorial violation, alcohol, dangerous and risky overtaking, right of way and seat belt. `Illegal driving' was defined on the basis of prosecutions for traffic violations, and `dangerous driving' on the basis of accidents the driver was involved in. A qualitative discriminant analysis and the usual squared test were used. The alcohol factor emerges as highly prominent in the discriminant analysis: anything but absolute abstinence before driving constitutes the best predictor of accident risk among drivers never convicted. The analysis shows that: (1) driving performance as calculated on the basis of the two discriminant functions is closely related to the number of offences and accidents admitted by drivers; (2) the percentage of drivers correctly classified by BDB is 66% in the case of accident performance and 43% in the case of offence performance; (3) BDB does have significant power for predicting involvement in accidents, comparable to that given by criteria based on complex observation techniques. From a quantitative standpoint, BDB is as effective as criteria normally employed for predicting accidents. It is also a valuable research tool.
Abstract