Key research findings

  • Historically, research in regard to driver licensing tests has yielded disappointing results. Driving tests that have been used vary in reliability and validity, many have poor psychometric properties, are relatively undemanding, capable of screening out only the very incompetent, and have little ability to predict who will be involved in subsequent crashes.
  • The basic road tests that have been evaluated have generally not been found to be associated with safety benefits.1
  • Safety benefits of new tests that have been developed to accompany GDL programs are not well established.
  • The tests that have been developed to assess progress through GDL systems in Australia, Canada, and New Zealand have been carefully and thoughtfully designed, and they have considerable theoretical and logical appeal.
  • Most of the research on the new GDL-related tests has dealt with predictive validity, the extent to which tests distinguish between safe and unsafe drivers. There is mixed evidence in this regard2 but positive evidence that some of the tests can identify safer drivers.
  • Those passing the New South Wales hazard perception test on the first try have been found to have fewer subsequent crashes than those who do not, and the same is the case for the New South Wales exit test.3
  • Vanlaar et al. (2009), in a meta-analysis of the effects of GDL features on fatal crashes in Canada, found that, at least for 19 year olds, the relative fatality risk decreases in jurisdictions that require an exit test to graduate from the intermediate stage.

1MacDonald, 1987; Mayhew et al., 2001

2Senserrick and Williams, 2014

3Roads and Traffic Authority of NSW, 2008