Investigating Drivers’ Hazard Perception Tests and Training

Wetton, Mark Alexander (2013). Investigating Drivers’ Hazard Perception Tests and Training PhD Thesis, School of Psychology, The University of Queensland.

Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials)
Name Description MIMEType Size Downloads
s4007325_phd_submission.pdf Thesis (fulltext) application/pdf 1.99MB 4
Author Wetton, Mark Alexander
Thesis Title Investigating Drivers’ Hazard Perception Tests and Training
School, Centre or Institute School of Psychology
Institution The University of Queensland
Publication date 2013
Thesis type PhD Thesis
Supervisor Mark Horswill
Total pages 111
Language eng
Subjects 1701 Psychology
1799 Other Psychology and Cognitive Sciences
Formatted abstract
An ongoing focus of the Australian government is to increase road safety through the development of improved driver assessment and training methods. Research indicates that drivers’ hazard perception ability is one of the only skills specific to driving to be linked to crash risk. Therefore, hazard perception tests and training interventions hold promise as ways to improve road safety. Five studies, spanning three journal articles, were conducted to investigate new measures of hazard perception for research, to develop a hazard perception test for driver licencing, and to evaluate components of an existing hazard perception training package.

The aim of the first paper was to evaluate two new instruments to measure drivers’ hazard perception. The Australian Capital Territory Hazard Perception Test (ACT Hazard Perception Test) involved drivers watching video clips of driving situations and predicting when a potential traffic conflict was likely to occur via a touchscreen. The Australian Capital Territory Hazard Change Detection Task (ACT Hazard Change Detection Task) used a flicker paradigm that involved drivers viewing two static images of the same road scene in alternation, separated by a mask, where they were required to detect the single difference (which was the presence or absence of a road hazard) between the images as fast as possible. The ACT Hazard Perception Test scores could differentiate between young novice and mid-age experienced drivers (mapping onto crash risk differences) and were correlated with established hazard perception tests. For drivers aged 65 and over, test performance correlated with vision tests known to predict crash risk in older drivers (contrast sensitivity and useful field of view) and also declined with age. The findings support the validity of this test. In contrast, young novice drivers responded quicker than mid-age experienced drivers in the ACT Hazard Change Detection Task and this test failed to correlate with other hazard perception measures, suggesting that it may not be appropriate for these age groups. However for the drivers aged 65 and over, the ACT Hazard Change Detection Task did correlate with the ACT Hazard Perception Test, and useful field of view, and test performance did decline with age. This suggests that this test may be useful, but only for older drivers.

The aim of the second paper was to develop and validate a new video-based hazard perception test for driver licensing in Queensland, Australia (the test used a similar method to the ACT Hazard Perception Test). A set of five principles for effective hazard perception tests was proposed and used to guide test development. The first study in this paper established that an instruction video devised for the test was understandable even by individuals with a low level of English, which was considered to be an important equity consideration for a licensing assessment. The second study in this paper found that the pool of video clips created for the test could discriminate young novice drivers from mid-age experienced drivers, which was considered an indication of test validity. There was no evidence for test score differences in gender, income, or educational level, indicating that the test could be considered equitable on these variables. All Queensland drivers on a probationary licence must currently pass this test to be awarded their full driver licence.

The aim of the third paper was to determine whether two different hazard perception training interventions (self-generated commentaries, and what happens next? exercises) contributed to the effectiveness of a multi-component hazard perception training intervention. Novice drivers experienced one of five intervention conditions: full training package, expert commentary alone, hybrid commentary (self-generated and expert commentary combined), what happens next? alone, or placebo control. Hazard perception skill was measured using a video-based test prior to the intervention, immediately after the intervention, and after a delay of approximately one week. All four training interventions resulted in significant improvements in hazard perception test response time immediately after training, compared to the placebo group. Only the full training intervention resulted in a significant improvement in response times after one week. The addition of the self- generated commentary exercise did not yield a greater training effect than the expert commentary exercise alone. The what happens next? condition yielded a smaller training effect than the commentary interventions. The results from the third paper provide support for the efficacy of hazard perception training as an alternative to traditional advanced driver training approaches as a means of improving novice driver safety.

Overall, the research described in this thesis has provided further insight into how hazard perception tests should be designed for both research and licensing purposes, as well as indicating how hazard perception training could be improved. These findings are hoped to contribute to greater road safety both domestically and internationally. 
Keyword Hazard perception
Driving safety
Driver training
Driver licensing
Driver assessment
Test validation
Driver behaviour
Change detection
Automobile driving -- Psychological aspects

Citation counts: Google Scholar Search Google Scholar
Created: Tue, 29 Apr 2014, 12:35:55 EST by Mark Wetton on behalf of Scholarly Communication and Digitisation Service