HAZARD PERCEPTION IN DRIVERS: A METHODOLOGY FOR TESTING AND TRAINING

The report is concerned with the development of a hazard perception test which aims to assess the ability of drivers to anticipate potential hazards as they unfold over time. Following the development of this test three major experiments assess the utility of the test by using a traditional expert-novice contrast and by examining the role of training. The first experiment was concerned with the development and validation of the test. In essence, this involved the testing of groups of expert and novice drivers, groups that would be expected to differ in terms of overall driving skill, and, consequently, also in their ability to perceive road hazards. The test revealed large, significant differences in the hazard perception abilities of the two groups, with virtually no overlap shown in the distributions of test scores. The Hazard Perception Test was thus able to correctly categorise every subject as either an expert or a novice. Information derived from this study lead to the production of a more refined form of the test, together with a second, parallel form, to enable before- and-after studies to be conducted. In experiment two, the new parallel forms of the test were subjected to validation by investigating the performance of groups of expert and novice drivers. In addition a group of experienced drivers were assessed. By matching the exposure of the experienced group with that of the expert group it was possible to examine the potential role of training (since the training received by the expert group was the major discriminating feature between the expert and experienced groups). The parallel forms of the test were highly correlated with each other and each test was able to distinguish between the experts and novices. Expert drivers performed more effectively than the experienced drivers, a result suggesting the influence of training on hazard perception. In experience three, a direct test of the role of training was carried out by examining a group of drivers undergoing ROSPA's advanced driving course. Relative to the control group the ROSPA group showed a significant improvement in their performance on the hazard perception test following training. The control group showed no significant change. Taken as a whole, the results indicate that the methodology described provides an appropriate and sensitive measure of hazard perception. It has been shown that, based on the performance on this test, there are large differences in the perceptual abilities of novice, experienced and expert drivers. It is also clear that training can exert a positive influence on hazard perception. (A)