A national postal questionnaire survey of motorcyclists has been carried out to provide information about themselves, their bikes, their riding and their recent accident involvement. Generalised linear modelling techniques have been used to develop a model based on 5336 riders for the prediction of an individual's annual accident frequency (all accidents on the main bike ridden, including minor spills and excluding fatalities). The resulting relationship shows that accident liability depends on both the age and riding experience of the rider, older and more experienced riders being the safest. It increases with increasing mileage ridden (though less than proportionately), and is lowest for those who ride mainly during the summer and for those who do at least about half their riding outside built-up areas, on large bikes. The main purpose of riding the bike, the rider's sex, and the amount of car driving done were also found to influence accident liability. The report summarises the data obtained from the survey and describes the analysis technique and procedures. It then goes on to present the relationship derived, to demonstrate the predictive effects of the explanatory variables and to discuss the implications of the results. (A) The full analysis of the survey datais presented in TRRL report CR 146 (IRRD 832454)
Abstract