Validity, reliability and predictive accuracy of the Driving Awareness Questionnaire

Abstract

Purpose. To examine the psychometric properties of the Driving Awareness Questionnaire (DriveAware) and to compare this rating of awareness of driving ability with the rating made in the context of an on-road assessment.

Method. A prospective design was employed to examine the psychometric properties of DriveAware using Rasch analysis. Participants (n = 91) with varying diagnoses were recruited from two driving rehabilitation centres in Sydney, Australia. They completed DriveAware in the context of a standard driving assessment. Awareness of driving ability determined by DriveAware was compared with awareness determined in the on-road assessment.

Results. Rasch analysis provided good evidence for construct validity and inter-rater reliability and some evidence for internal reliability of DriveAware. The item hierarchy was logical and goodness-of-fit statistics for four of five items were within an acceptable range. The test had a moderate reliability index (0.73). Measurement precision could be improved by including more items to assess drivers with little awareness. When compared with the on-road measure of awareness, the best DriveAware cut-off score yielded sensitivity of 84% and specificity of 94%.

Conclusion. This short and easy-to-administer test could provide a useful tool with sound psychometric properties to measure awareness of driving ability.

Keywords:

• Automobile driving
• awareness
• insight
• Rasch analysis