This paper presents the results of a study to develop a quicker and improved programme for teaching hazard perception skills to drivers. The study's main aims were to: (1) determine the effects of a purpose-built training programme that could be conducted in a classroom rather than on the road; (2) develop a training programme to improve novice drivers, who are most likely to have an accident. The training course used the idea that hazard perception might best be conceived as a cognitive process, applying a model of how the traffic system and its elements interact. The training programme concentrated on forcing the novices to develop a more sophisticated mental model, by engaging them on anticipation tasks. To do this, sequences of road scenes were presented by video, and novices were asked to predict what would happen next. Using two parallel hazard perception tests, the abilities of novices could be assessed before and after this training. 16 novices participated in the training course, and, as a control, 14 similar novices were given pseudo-training. An analysis of covariance was conducted on the observations. The response latencies of the novice group, who received training in hazard perception, were improved by about 0.5 sec. After training, their performance level had become comparable to that of experienced drivers.
Abstract