Monthly frequencies of two-vehicle accidents occurring in Great Britain during the period 1970-1978 have been studied. The data were disaggregated by the types of vehicle involved, by built-up or non-built-up areas, and by road class. Thirty-six series of accident data with monthly mean frequencies from about 20 upwards were analysed by regression methods. Explanatory variables used included traffic volumes, temperature and rainfall, an index of petrol prices, and factors representing trend, seasonality, the presence of a fuel crisis, and a temporary speed limit. Clear relationships were found between the accident data and several of these variables. In particular, the relationships with traffic flows were clearly not of the simple proportional nature often assumed. Those with petrol prices were unexpectedly strong, and further studies will be made of these. The fuel crisis of 1973-74 had a marked effect on car accidents, but not on accidents involving two-wheelers. Temperature and rainfall were related to accidents in ways which might be expected, "better" weather tending to be associated with fewer accidents. (Author/publisher)
Abstract