Improving Traffic Law Compliance: Attitudes and Behaviour

Abstract

Three driving violations, speeding, drink-driving and failure to wear a seat-belt are major contributors to New Zealand road crash statistics and are the main focus of New Zealand's traffic policing enforcement and media campaigns. Leivesley (1987) has proposed that traffic policing changes behaviour as a response to the threat or presence of enforcement, and by improving attitudes and social norms towards safer driving. Research shows that enforcement initiatives, especially when combined with publicity, have a demonstrated but limited effect on driver behaviour. Psychological theories of attitudes provide a structure to the potential influence of attitudes and social norms to driver behaviour. The research however is less clear about the effect of enforcement in changing attitudes and there is no clear relationship between attitude change and on-road behaviour change. The aim of this thesis is to examine further the correspondence between the psychological variables outlined in attitude models on behaviour, and to follow the effect of a new road safety intervention and its effect on behaviour over time. Two studies were conducted with respect to three driving behaviours, speeding, drink-driving and seat-belt wearing. Study 1 used a questionnaire based upon the Theory of Planned Behaviour (Ajzen, 1991) and an extension to the model (Manstead and Parker, 1995) to examine the ability of the model to predict intention and reported behaviour. An additional measure of reported behaviour was included where friends rated the participant's driving behaviour. Results indicated that the full model could predict (depending on the driving behaviour) between 46-61% of the variance in intention and 57-88% of the variance in reported behaviour. The two measures of reported behaviour were highly correlated. Study 2 consisted of five surveys conducted over a 10 month period, examining the effect upon behaviour of the introduction of a new road safety campaign, called the Supplementary Road Safety Package (SRSP). Drivers were randomly stopped by a police officer and asked to participate in a road-side survey. The same questionnaire from Study I was used to assess the Theory of Planned Behaviour components and measures of on-road behaviour were also recorded. The full model, which included a past behaviour measure, accounted for 61-87% of the variance in intention to speed, drink-drive or wear a seat-belt, and between 56-70% of the variance in reported speeding behaviour. The model however accounted for substantially less variance (10-24%) in speeding behaviour measured on the road. The TPB model appears to be insufficient at predicting intention and behaviour, and may need to be modified to better describe the processes at work for different driving behaviours. There was evidence of behaviour change on the road (speeding, drink-driving and seat-belt wearing) associated with the SRSP campaign. Public awareness of the road safety campaigns significantly increased, but there was no significant or consistent change in any of the model components except for one belief measure, the risk of apprehension. The effect of the SRSP campaign therefore appears to be largely through Leivesley's (1987) first mechanism of behaviour change. Reasons for the lack of change observed in the TPB model component and the implications for road safety initiatives are discussed.