Introduction

Talib Rothengatter; Raphael D. Huguenin

FAIRY TALES

Accidents occur, like encounters with fairies or werewolfs, to the weary traveller, but accidents or encounters with fairies or werewolves, are random events. The behaviour of the traveller, and his mental state, are factors that influence the likelihood of occurrence. How important these factors are, is matter of debate. One extreme view is the sustainable transport philosophy maintaining that sustainable road design would virtually eliminate accidents and would ensure that those accidents that would occur will not inflict serious injury or death. No such claim is made about fairies or werewolves, but it is likely that in an orderly world they too will behave orderly. On the other hand of the spectrum remains the assertion made that 85% of the accidents occurring can be attributed to human error. Obviously, when humans are prone to imagine fairies or werewolfs, they are prone to other types of error as well.

In 1994, a special issue on traffic psychology appeared, based on presentations at the 23rd International Congress of Applied Psychology. In this issue, we contended that traffic psychology appeared a promising new area of research (Rothengatter, 1997). Research into road user behaviour and transport safety had been dominated for decades by other professions. Consequently, theoretical concepts and models dominant in mainstream psychology were not considered in the domain of road user behaviour. Application of these concepts and models would enhance our understanding of road user behaviour and would help to create the new possibilities of influencing road user behaviour.

MODELS AND THEORIES

Elsewhere, we argued that traffic psychology would benefit from developing a general domain model or theory (Huguenin, 1997). Although, a plethora of models and theories has been developed (see Ranney, 1994, for an overview), most of these are too specific to serve as a general research framework. As a result, research results often are reported without reference to any theoretical framework. We contended that a general model or theory would have to incorporate at least a component related to the disposition of the driver; a component related to social processes, cognition and motor skills, and a component related to situational characteristics (Huguenin, 1997).

Two concepts appear central in the earlier discussion on models of road user behaviour: risk and motivation (e.g. Huguenin, 1988: Rothengatter, 1988, 1990; Summala, 1988). Road user behaviour was conceptualised as adaptive to perceived risk. Road users were perceived as risk adaptors. If they perceived the risk to be high, they would adjust their behaviour to lower their risk. The debate concerned the parameters (zero risk or target risk). In Summala’s zero-risk model, road users act if risk was perceived above zero; in Wilde’s target risk model (1994) road users act if risk was above target risk, but also act if risk was below target risk. That road users regulated their behaviour based on perceived risk was hardly in question; road users were risk adaptors. The risk models have been criticized on various grounds (see Michon, 1989) such as the presence of a “monitor” or “comparator” in the models, the lack of definition of the risk construct (individual or aggregate) and the discrepancy between “perceived” (subjective) risk and “actual” (objective) risk. Experimental studies meanwhile have demonstrated that drivers also display adaptive behaviour when the perceived risk is low (e.g. Van der Hulst, Rothengatter & Meijman, 1998; Cnossen, Rothengatter & Meijman, 2000). Instead of adapting their behaviour to perceived risk, drivers were found to adapt their strategic and tactical behaviour to experienced task demand, the amount of effort required to maintain adequate performance. This task demand was found to fluctuate as a result of both task characteristics such as complexity or preview and state characteristics such as time-on-task or fatigue. Fuller (2000) introduced a task capability model in which drivers match performance with task demands. Although this model needs further elaboration (it incorporates a task difficulty model that again uses a comparator, this time comparing perceived task difficulty and acceptable task difficulty), it is clearly a step forward from the earlier risk models as the components of the model do allow empirical verification.

ATTITUDES AND DRIVER PERFORMANCE

Risk and performance models have one element in common. They all acknowledge that the quality of task performance of drivers is dependent on their aspiration level, be it in terms of perceived risk, comfort, perceived task control, workload or whatever the construct may be, and that this aspiration level is under motivational control. Summala (1988) explicitly acknowledges the importance of “external motives” in determining the outcome of the monitoring system in his model.

Attitude theory incorporates these motivational aspects as attitudes are conceptualised as psychological tendencies to evaluate aspects positively or negatively including affect and mood. Motivational processes are considered crucial to attitude formation and change (see Eagly & Chaiken, 1993). The Theory of Reasoned Action and its successor, the Theory of Planned Behavior, have been used extensively to study road user behaviour (e.g. Parker, Manstead, Stradling & Reason, 1992) albeit mostly in relation to law violations such as speeding, drink-driving and seat belt use. More recently, these theories have been applied to study aggressive road user behaviour (e.g. Parker, Lajunen & Stradling, 1998). In these theories, the motivational processes are conceptualised as a product of the estimated likelihood of the outcome of behaviour and the evaluation of that outcome. Strong, consistent positive relationships between these “motives” and intentions to display specific behaviour have been established. However, three observations must be made. The first is that these relationships have been established between attitudes and intentions, or at best, between attitudes and reported behaviour. Studies that demonstrate relationships between attitudes and actual, observed road user behaviour are rare, and these studies usually measure attitudes after the behaviour is observed, hence their predictive value can be questioned. The second observation is that the results of attitude studies have not been integrated in performance-oriented models, even though there is theoretically every possibility to do so. Finally, if attitude studies have contributed to our understanding of the motives road users have to display deviant and dangerous behaviour, they have contributed very little to measure to remedy this behaviour. This is not due to attitude theory. In other areas of applied research based on attitude theory, interventions have been developed on the basis of attitude theory, as for example in smoking and other health-related behaviour. It is a major shortcoming that needs to be addressed.

INDIVIDUAL DIFFERENCES

Accident-prone drivers may not exist, or if they exist may not be found, (see Haight, this volume), but individual differences are a subject of substantial research. Sensation-seeking is consistently associated with risky driving (Jonah, 1997). Walton (1999) compared truck drivers’ self-assessment and assessment of others on measures such as speed, safety and consideration with the averages of independently-recorded objective indicators in attempt to determine the direction of bias. He found that drivers do not demonstrate ‘self-enhancement’ indicative of driver overconfidence and conclude that the self-enhancement bias found operates as a negative-other rationalisation. McKenna (1993) found illusion of control but not optimism bias to affect drivers’ expectancy of accident involvement. In addition to these and similar cognitive social psychology constructs, differences in accident involvement of younger versus older, female versus male, and experienced versus inexperienced drivers have attracted considerable research efforts. Groeger (2000) provides an overview of this research. It is clear that people differ in their cognitions regarding driving and differ in their driving style. How these differences determine accident-involvement is as yet largely unclear. Accident-prone driver remain elusive.

ACCIDENT-COUNTERMEASURES

The reason for carrying out applied research is not primarily to contribute to the scientific body of knowledge or to develop new models or theories; it is to contribute to solving the problem to which the research is applied. In this traffic psychology falls short. Risk theories have done little do reduce the risk of driving. Performance models may in other areas of task performance have had an impact on task design, not so in transportation. Road design still is primarily a task of road engineers, car design primarily a task of vehicle engineers. The traditional realm of traffic psychology – road safety campaigns, driver selection, training and rehabilitation – cannot claim an impressive impact on road safety.

There is one exception. The introduction of information technology in the transport system has created new opportunities for traffic psychology. In first instance, traffic psychology has concentrated on the question how the changes in task environment that resulted from the introduction of...