The wintertime accident risks of drivers and the factors affecting the risks were analysed using statistical accident models. The evaluation method was based on reliability theory and on survival modelling. The data consisted of two parts: responses to a postal questionnaire, addressed to 10,000 vehicle owners, about driving and accidents during wintertime in the years 1991-1993; detailed records of fatal accidents, in the years 1987-1991, generated by in-depth accident investigation teams and compiled by the Motor Insurers' Road Safety Committee (VALT). Replies to the postal questionnaires were received from 5,881 vehicle owners, giving a response rate of 59%. The replies included 296 self-reported accident involved drivers. The VALT data on fatal wintertime accidents contained 658 drivers involved in fatal accidents. The analyses of the two data sources confirmed that driving conditions and kilometreage driven during the wintertime contribute to amount of accidents. The best explained variables in both survival and risk models of wintertime driving were driver characteristics (age, driver behaviour, kilometreage driven, speed, and use of safety-belt), drivers' state (driving under the influence of alcohol) and vehicle characteristics (vehicle weight and condition of tyres). Young and inexperienced on the one hand, or old (and experienced) drivers, on the other hand, had the highest fatal accident risks. There were no direct sex-related differences in accident risks between female and male drivers. Drivers using non-studded winter tyres had a somewhat greater accident risk than drivers using studded tyres, but the difference was not statistically significant. Survival modelling with two distinct data sources, a postal survey data and an in-depth accident data, were used in the analysis. Modelling methods were also analysed by simulation and it was concluded that survival modelling promises to be a useful tool for safety analysis but the method needs further development.
Abstract