The Transport Research Laboratory (TRL) is conducting the Review of the Practical Driving Test project on behalf of the Road Safety Division of the Department of the Environment, Transport and the Regions (DETR). One of the project's tasks is to investigate the reliability of the current UK driving test. This paper presents some initial findings from a study of test-retest reliability, and discusses their implications. It first identifies some factors that tend to make driving tests unreliable. A study of test-retest reliability was conducted at 20 driving test centres in November and December 1998, and it was found that the practical driving test had rather low test-retest reliability. This does not imply that the test fails to induce good training and practice, or that the test is conducted incorrectly; nevertheless, higher test reliability is still desirable for reasons of efficiency and perceived fairness. Potential sources of unreliability include: (1) failing some candidates of moderate ability; (2) using only one error as a failure criterion; and (3) other factors associated with variations in test events and conditions. A simplified model of the test can give some useful insights into its reliability, and into its ability to identify and fail candidates with unacceptable underlying error rates.
Samenvatting