Accident statistics are considered high when thought of in terms of the social and economic costs that they incur. However, with road accident rates in the UK of approximately 0.1 accidents per driver per year, researchers need extremely large sample sizes in order to carry out reliable research into accident distributions, causality and prevention. In addition to this, self-reporting of accidents is subject to memory problems and reporting bias, while insurance and police records are limited to more serious accidents. A recent study conducted by the authors used a two phase approach. A large sample of company vehicle drivers participated in a questionnaire study, following which a subset of these subjects completed a diary study, in which they recorded details about their driving on a journey by journey basis for a two week period. As part of this study, they were asked to report any near-accidents they experienced. A comparison of the real and near-accidents recorded during the course of these studies is reported here. Near-accident rates correlated well with real accident rates, and while there were some differences in the distribution of real and near-accident types, it is concluded that this methodology could prove useful for future research studies. For the covering abstract see ITRD E113725 (C 22328 CD-ROM).
Abstract