ERRORS AND BIAS IN ASSESSMENTS OF DANGER AND FREQUENCY OF TRAFFIC SITUATIONS

Six groups of nine drivers took part in a laboratory study which sought to investigate the ways in which judgements are made about the frequency of accidents and particular manoeuvres in various driving situations. For some groups a sorting task was interpolated between frequency judgement tasks, which required subjects to categorize driving errors either according to whether they were or were not dangerous, or did or did not commonly occur, in the contexts encountered earlier. Other groups did not have such interpolated tasks. The effects of the sorting exercise were to increase the subjective frequencies of accidents, but not subjective frequencies of particular manoeuvres in those same situations. These findings are discussed in terms of theories of heuristics and debiasing propounded by Fischoff, Kahneman, Tversky and others, and current research on drivers' assessments of risk and ability. © 1990 Taylor & Francis Group, LLC.