Background

Road trauma is recognized as a serious problem both in Australia and internationally. Young drivers are over-represented in crashes among all classes of road user. A range of factors may potentially contribute to road crashes. Risky driving has been identified as an important contributor to road crashes, although its role is not comprehensively understood.

The focus of the present research will be on four risky driving-related behaviours: speeding, drink-driving, driving while fatigued, and not wearing seat belts. The successful manipulation of risky driving requires a good understanding of contributing factors.

However, until recently no research has directly examined the differences between risky driving behaviours, in terms of precipitating factors. Our preliminary research (Fernandes & Job, 2003; Fernandes, Job & Hatfield, 2004) indicated that different factors predict different risky driving behaviours.

For example, speeding was predicted by the authority rebellion scale, while drink driving was predicted by sensation seeking and illusory invulnerability. The Health Belief Model (HBM; Janz & Becker, 1984) is used an initial framework, and additional factors that appear to be important are also considered.

The HBM proposes that a person’s decision whether or not to follow a health-related behaviour is determined by four dimensions: the perceived susceptibility to, and perceived severity of, the consequences of a risky behaviour, as well as the perceived benefits, and perceived costs, of the alternative safety behaviour (Becker, 1974).

The dependence on self-report measures to provide information about attitudes is a limitation of current research in this area. The present research will attempt to redress this issue, through the use of the Implicit Association Test (IAT; Greenwald, McGhee, and Schwartz, 1998). The IAT represents a hidden measure of attitudes.

It measures differential associations of two target concepts with evaluations. It provides a measure of performance difference (response time) that indicates whether an individual’s attitude to a particular risky driving behaviour (e.g. speeding) is positive or negative.

This study examined a range of demographic factors, personality factors, attitudes and beliefs in the prediction of speeding, drink-driving, not wearing seat belts, and driving while fatigued, for young drivers aged 16-25 years, by the use of a Risky Driving Questionnaire.

Study 1 - examined whether different beliefs and attitudes predict different risky driving behaviours among students.

Study 2 - examined whether the results from the Study 1 student population sample generalize to a general population sample, and compared findings from the metropolitan Sydney sample with findings from a rural NSW sample.

Study 3 - introduced four risky driving-related versions of the IAT versions to the field of road safety, and aimed to validate the attitudinal scales within the Risky Driving Questionnaire with the use of the IAT. Study 3 examined the relationship between the IAT and behavioural intentions, for each of the four risky driving behaviours. Additionally, Study 3 examined the test-retest reliability of the Risky Driving Questionnaire.

Method

For each of the four risky driving behaviours examined, the Risky Driving Questionnaire assessed demographic variables (such as age and gender), road-unrelated illusory invulnerability, road-related general illusory invulnerability, road-related specific illusory invulnerability, general and specific perceived susceptibility, general and specific perceived severity, perceived costs, perceived benefits, peer influence, behavioural intentions, authority-rebellion, time urgency, sensation seeking, driver anger, socially desirable responding, and road crash and infringement history.

In Study 1, 215 first-year School of Psychology students from the University of NSW (who were required to be 25 years of age or less, and to hold a current NSW drivers license) completed the Risky Driving Questionnaire individually.

In Study 2, participants were 587 drivers from metropolitan Sydney and 422 drivers from rural NSW who were required to hold a current NSW drivers license. Participants (age range of 16-25 years) were recruited outside RTA Motor Registries.

In Study 3, participants were 135 drivers from metropolitan Sydney who were required to hold a current NSW drivers license. All participants had previously participated in Study 2 and had given consent to be contacted via telephone regarding participation in a follow-up study. A meeting place and time was arranged with each participant.

 Participants completed the same Risky Driving Questionnaire version that they had completed in Study 2, as well as the risky driving-related IAT version relating to the same risky driving behaviour.

Results And Discussion

For Study 1 and Study 2, hierarchical regression models for each of the four driving behaviours were compared. A gradation of factors was proposed, arranged in order of how stable each set of factors were as an intrinsic characteristic of the person.

Accordingly, three regression models were considered for each behaviour: the ‘Demographics only’ model, followed by the ‘Demographics plus personality factors’ model, and finally the ‘Demographics plus personality factors plus beliefs’ model. Study 3 assessed test-retest reliability of the Risky Driving Questionnaire, and the relationship between implicit attitudes and Risky Driving Questionnaire measures employing Pearson’s correlation coefficients, for each of the four risky driving behaviours.

The possibility of different factors predicting different risky driving behaviours Results indicated that different factors predict different risky driving behaviours, supporting our preliminary research in this field (Fernandes & Job, 2003; Fernandes et al., 2004). The e results indicate the importance of considering separate underlying mechanisms for individual risky driving behaviour.

As such, each behaviour should be investigated individually in search of different causal factors. In addition, practical attempts to curb one particular behaviour cannot be assumed from the apparent success of another. For all three samples (student population, metropolitan Sydney population, and rural NSW population) those factors specific to a particular risky driving significantly predicted that behaviour, for each of the three samples.

It appears that, for these behaviours, the more approving attitudes people hold (specific to each risky driving behaviour), the more they will engage in that particular behaviour.

This finding suggests that specific attitudes are consistent and robust predictors of risky driving, and supports recent research illustrating the importance of investigating those attitudes and beliefs specific to individual risky driving behaviours, rather than general road safety attitudes and beliefs (Fernandes & Job, 2003; Fernandes et al., 2004; Iversen, 2004).