Self-reports are used for many studies of individual accident liability and assessments of countermeasures, but there is good evidence that they are rarely veridical. This paper presents a study which explored several variables relevant to recall, especially for near-accidents. The severity of individual incidents was examined, together with possible driver culpability. The potential effects of driving experience were also investigated. Thus the study compared two methods of obtaining self-reports, rather than comparing memories with objective data. 77 male and 71 female drivers were tested, consisting of 90 young novices, 21 older novices, and 37 older experienced drivers. It was found that the change in reporting style produced by the tests affected descriptions of near-accidents very differently from estimates of numbers of journeys, distance travelled, and time spent driving. Many drivers seemed to forget incidents in which they had been involved.
Abstract