DB&T 2009 Presentation on 24.11.09 (Tony Machin)

Understanding the unique contribution of aversion to risk taking in predicting drivers’ self-reported speeding M. Anthony Machin Associate Professor University of Southern Queensland Presented at the 2009 Driver Behaviour and Training Conference, Amsterdam

Background to the study
We now know that a drivers’ attitude towards risk taking is one of the factors influencing safe driving behaviours.
Machin and De Souza (2004) found that aversion to risk-taking, aggression, and perceptions of management’s commitment to health and safety were significant predictors of unsafe behaviour in taxi drivers.

Aversion to risk taking and speeding
Machin and Sankey (2008) compared the predictive strength of aversion to risk taking with three other risk perception variables and five measures of personality.
We showed that drivers’ risk perceptions are an important predictor of speeding and unsafe driving.

Results from our previous research
Likelihood of an accident, driving efficacy, and aversion to risk taking were significant unique predictors of speeding accounting for 6%, 3%, and 15% of the variance respectively.
Excitement-seeking and altruism accounting for an additional 2% and 3% of the variance respectively.

Conclusions from our previous research
One of the difficulties in drawing conclusions from the previous study relates to the variables that were not included.
We recognised that drivers’ risk perceptions may also be related to their choice of coping strategies.

Appraisal of risk
The transactional model proposed by Matthews (2001) includes dispositional characteristics and coping strategies reflecting the various factors that combine to influence the appraisal of risk when driving.
Drivers who adopt more maladaptive coping mechanisms may be more likely to speed.

Drivers’ coping styles
Matthews et al. (1996) concluded that confrontive and emotion-focused coping were maladaptive coping styles associated with more negative outcomes.
Confrontive coping was linked to violations, errors, and greater speeding.

Figure 1 from Matthews (2001) Matthews, G. (2001). A transactional model of driver stress. In P. Hancock, & P. Desmond (Eds.), Human factors in transportation: Stress, workload, and fatigue (pp. 133-163). Majwah, NJ: Lawrence Erlbaum Associates.

Moderator effects
Various demographic characteristics are related to risky driving behaviour.
This study looked at the how well the set of predictors accounted for variance in speeding when evaluated separately for males and females, for younger and older drivers, and for more and less frequent drivers.

Aim of the current study
This study focused on the unique contribution of one measure of risk perceptions (aversion to risk taking) in the prediction of speeding whilst controlling for a range of other predictors of speeding.
It also examined whether this outcome depended on the age, gender, and the driver’s frequency of driving.

Method
N = 402 Australians over 17 years of age, were invited to participate through an email providing a link to a web-based survey.
Males = 20%
20.8% of the participants were young drivers, aged between 17 and 19.
73.8% drove every day

Study Variables
Driver Coping Questionnaire (DCQ; Matthews, et al., 1997) assessed Task-Focused Coping, Reappraisal, Avoidance Coping, Confrontive Coping and Emotion-Focused Coping ( α ’s = .83, .79, .70, .84 and .79).
Driver Stress Inventory (DSI; Matthews, et al., 1997) assessed Aggression, Hazard Monitoring, Thrill Seeking, Dislike of Driving, and Fatigue Proneness ( α ’s = .85, .78, .89, .85 and .80) .

Measures of Worry and Concern, Likelihood of Accident, Efficacy, and Aversion to Risk Taking were taken from Machin and Sankey (2008) ( α ’s = .92, NA, .88 and .78).
Speeding was assessed using Ulleberg and Rundmo’s (2003) scale ( α = .84).

Results – Standard regressions
When predicting Speeding
The overall model explained 50% of the variance in Speeding
Five variables added uniquely to the prediction of Speeding:
Aversion to Risk Taking ( sr 2 = .07)
Confrontive Coping ( sr 2 = .03)
Thrill Seeking ( sr 2 = .03)
Worry and Concern ( sr 2 = .01)
Likelihood of Accident ( sr 2 = .01)

Additional standard regression analyses were conducted for the following subgroups: drivers less than or equal to 20 years old ( N = 108), males ( N = 79), and drivers who are less frequent drivers ( N = 105)

For younger drivers
The overall model explaining 58% of the variance in Speeding
For males
For less frequent drivers

The unique contribution of Aversion to Risk Taking differed for these three subgroups
For younger drivers
Aversion to Risk Taking was still the strongest unique predictor ( sr 2 = .07)

For males
Aversion to Risk Taking was not a significant predictor ( sr 2 = .01)
For less frequent drivers
Aversion to Risk Taking was the second strongest unique predictor ( sr 2 = .04) after Thrill Seeking ( sr 2 = .06).

Results - Structural equation modeling
In order to examine the direct and indirect impact of the predictor variables on Speeding, a model was specified that allowed Thrill Seeking, Aggression and Confrontive Coping to be directly related to Speeding as well as indirectly via Worry and Concern and Aversion to Risk Taking. Note that Aggression was not a significant unique predictor of Speeding.

Structural equation model

This model is a good fit to the data.
χ 2 = 2.54 d f = 2, p = .28
CFI = 1.00, TLI = .99, RMSEA = .03
The combination of predictors is explaining 48% of the variance in Speeding, 11% of the variance in Worry and Concern and 17% of the variance in Aversion to Risk Taking.

Conclusions
There are several unique predictors of Speeding, including three risk perception variables (Worry and Concern, Likelihood of oneself having an accident, and Aversion to Risk Taking), one personality variable (Thrill Seeking), and one coping strategy (Confrontive Coping).

Note that these results show which variables can contribute uniquely to the prediction of Speeding after all of the other predictors have been controlled for.
The results suggest that at least three and perhaps as many as five predictors should be included in the conceptual model of predictors of Speeding.

What does the SEM show?
The prediction of speeding is best captured using a model that allows Thrill Seeking to influence Speeding both directly (as a positive predictor) and indirectly through its impact on Aversion to Risk Taking (negative predictor) and Worry and Concern (positive predictor).
The impact of Aggression on Speeding is entirely mediated through Aversion to Risk Taking and Worry and Concern.

The impact of Confrontive Coping on Speeding is both direct (as a positive predictor) and indirect through its impact on Worry and Concern (positive predictor).

What have we learned?
The current study extends the results of a previous study by Machin and Sankey (2008) by including a wider range of ages in the sample and also expanding the range of predictor variables to include drivers’ coping strategies.
We also included age, gender, and driving frequency as potential moderators of the importance of Aversion to Risk Taking.

Importance of Aversion to Risk Taking
We have developed a very strong conceptual model which explain a great deal of the variance in speeding.
The role of risk perceptions such as Aversion to Risk Taking is quite important across both younger and older drivers, but less important for drivers who drive less frequently and not important for male drivers.

In a nutshell
All drivers need to increase their self-awareness of the strong influence of their need for greater stimulation and expression of anger on their driving behaviour as well as the negative outcomes of dealing with driving situations through confrontive coping strategies.

Contact me if you have any questions
Associate Professor Tony Machin,
Department of Psychology,
University of Southern Queensland,
Toowoomba, 4350. Australia.
Telephone +61 7 46312587.
Fax +61 7 46312721.
Email: [email_address]

DB&T 2009 Presentation on 24.11.09 (Tony Machin)

by Tony Machin

Accessibility

Categories

Upload Details

Usage Rights

1 Embed 1

Statistics

DB&T 2009 Presentation on 24.11.09 (Tony Machin) Presentation Transcript