Can a video-based hazard perception test used for driver licensing predict crash involvement?

In 2008, the state of Queensland in Australia introduced a video-based hazard perception test as part of the licensing process for new drivers. A key validity check for such a test is whether scores are associated with crash involvement. We present data demonstrating that drivers who failed the hazard perception test (based on a ROC curve-derived pass mark) were 25% [95% confidence interval (CI) 6%, 48%] more likely to be involved in an active crash (defined as a crash occurring while the driver's vehicle was moving but they were not engaged in parking or reversing) during a one year period following the test (controlling for driving exposure, age, and sex). Failing drivers were also 17% (95% CI 6%, 29%) more likely to have been involved in active crashes prior to the test, in the period since obtaining their provisional license. These data support the proposal that the hazard perception test is a valid measure of crash-related driving performance.