PREDICTING WHAT A DRIVER WILL DO: IMPLICATIONS OF THE THREAT-AVOIDANCE MODEL OF DRIVER BEHAVIOUR --ROAD USER BEHAVIOR. THEORY AND RESEARCH. PAPERS PRESENTED AT THE 2ND INTERNATIONAL CONFERENCE ON ROAD SAFETY HELD IN GRONINGEN, NETHERLANDS, AUGUST 1987
Richard A. Fuller · 1988
The threat-avoidance model of driver behaviour argues that within the context of the motivation for a particular journey, (usually but not inevitably a particular destination and within a particular period of time), driver behaviour is focussed on the avoidance of averse or potentially averse stimuli in the road-traffic environment. Because driving is a self-paced task, the majority of averse stimuli arise out of an interaction between what the driver does, and features of the environment. Safe, effective and efficient driving involves learning the relationship between antecedents, behaviour and consequences. At any moment in time this learning is represented in the driver partly in terms of his/her conditioning history, and partly as a symbolic representation of contingencies. Assuming normal functioning of the individual's information processing system it should thus be possible to predict driving behaviour, given a statement of his/her learning history, and a description of the effects of the consequences of alternative actions. Although these principles can be stated quite simply, in practice controls on driving behaviour may be remarkably complex. For example the rewards and punishments for any particular response may be multiple and not mutually exclusive: thus, the net gain of any action becomes extraordinarily difficult to predict. Furthermore because averse consequences to particular responses, given particular discriminative stimuli, are typically probabilistic in aggregated statistical terms, but always dichotomous in real phenomenal terms, there may be a range of probabilities where randomly varying state parameters determine the behavioural outcome. Examples of such parameters might be arousal level, impulsiveness, and distractability. The logic of traditional behavioural decision theory involving the simple weighting of gains by their probabilities (usually by multiplication), can only refer to optimal decision making over a series of decisions. Its psychological validity in discrete instances of behaviour, however, is questionable.(a) for the covering abstract of the conference see IRRD 815404.