Cohort II: A Study of Learner and New Drivers Volume 1 – Main Report

  • Published: TRL, Road Safety Research Report No. 81 2008, May
  • Authors: P wells, S Tong, B Sexton, G Grayson, E Jones
  • Date Added: 03 Apr 2012
  • Last Update: 12 Jan 2018
  • Format: pdf


‘Cohort II’ was a major six-year study, funded by the Department for Transport, providing an up-to-date picture of how ‘cohorts’ of learner drivers in Great Britain undertake driver training and testing, and of their subsequent experiences as new drivers. It builds upon and further develops the evidence base from the smaller Cohort I study in 1988 – 89.

The aims of the study were:

  • to investigate how people learn to drive, including the number of hours of tuition and practice, and to compare this to outcomes from the theory and practical driving tests;
  • to assess the impact of changes to the testing regime, speci?cally the hazard perception test which was introduced during the period of study;
  • to explore new drivers’ experiences and attitudes to driving; and
  • to identify their level of accident involvement over time.


Every three months, from November 2001 to August 2005, a random sample of 8,000 practical test candidates was drawn by the Driving Standards Agency (DSA) from candidates in a given week (this was approximately one-third of those taking their test in that week). For the purposes of the study, the resulting 16 cohorts were labelled A to P. Postal questionnaires were sent to these candidates and, if they passed their test, follow-up questionnaires were sent at specific points in their driving career. For all questionnaires, reminders were sent if a response was not received within two weeks.

The sample initially comprised 42,851 learner drivers, however not all of these passed their practical tests to be involved in the subsequent surveys of new drivers. The sample of new drivers in Cohort II varied from over 10,000 at six months after the practical test to just fewer than 2,000 at three years after taking the test.

Key Findings:

  • People who pass the test at a young age tend, initially, to drive less safely than others. This effect is strongest soon after the test, and declines during the first three years of driving.

  • The pattern of results is consistent with the notion that residual effects of starting to drive young become diluted by other influences as time progresses.

  • It shows that there is not something persistently different about those who start to drive very young – in terms of driving safety, they become like other drivers within a year or two.

  • However, within this period they do have an excess accident liability, and this result re-emphasises the importance of finding ways of targeting safety interventions at very young drivers.

  • The introduction o the hazard perception component in the theory test appears to have been associated with some reduction in subsequent accident liability, although the size of the estimated effect varies with the type of accident.