Hazard perception is considered to be an essential part of the driving task. In young novice drivers, however, hazard perception still is poorly developed. In its research programme for the period 2007-2010, SWOV therefore wishes to investigate the possibilities to measure and train hazard perception. There also was the policy intention to already begin testing hazard perception in the driving examination for passenger cars (the Dutch B license) in 2008. This report discusses what hazard perception is and what is already known from international studies about the testability and trainability of hazard perception. In addition, SWOV developed two types of tests for the Dutch theoretical examination; they are described in this report. The validity of these tests has been investigated. SWOV also investigated the trainability of hazard perception in the Dutch driver training. The results are presented in this report. Although testing and training of hazard perception is new for the Netherlands, it has a long tradition from the international perspective. Australia, Great Britain and New Zealand have already incorporated the testing of hazard perception in the driving examination. A distinction needs to be made between the testing of hazard detection and the testing of the reaction to hazards. The majority of tests that have been developed only test hazard detection. Most of the international studies show that older, more experienced drivers do better on hazard perception tests than young novice drivers. International studies also indicate that subjects with a high crash rate have lower score on hazard perception tests than subjects with a low crash rate. However, there are a small number of studies that showed no connection between the test scores and age and experience. The two tests that have been developed by SWOV are a so-called film test and a photo test. The film test aims at measuring hazard detection and the photo test is intended to measure the reaction to hazards. In the film test subjects are shown animations in which the picture is frozen for a few seconds a number of times. In these breaks the subjects have to use a mouse to click the spots and the directions they specifically want to keep an eye on because of potential hazards. In the photo test subjects are shown photographs that are taken from the driver's position. They have to indicate whether, based on the hazard that may or may not be present, they would brake (acute danger), release the accelerator (latent danger), or do nothing at all (no danger). It is relevant that both types of test will eventually be part of the theoretical examination, because hazard perception cannot be measured well with only one test. The present study investigated for both tests if there was a connection between the crash rate and the test scores of young novice drivers. This appeared to be the case for both tests. On both tests young novice drivers with a high crash rate perform worse than young novice drivers with a low crash rate. The performance on the photo test is better with increasing age and driving experience, but this is not the case for the film test. More explanation and practise is probably required to make this test as it is intended. Contrary to what could be shown in various international studies, the Dutch study does not show that following a special training results in higher test scores. It was demonstrated, however, that both tests indicate that hazard perception has improved in the last part of the driver training, independent of whether the special training has been followed or not. The performance on the photo test probably gets better as a result of the regular driving lessons in traffic during the last period of the driver training. The so-called test effect (a higher test score as a result of having made the test before) did not occur for the photo test. For the film test, however, the test effect could not be ruled out, because technical problems prevented a number of subjects from participating in the film test. Therefore it is not possible to say whether the film test was made better on the post-test than on the pre-test because it had already been made before, or because of improvement due to regular driving lessons. There may have been different causes for the ineffectiveness of the special training. Possible causes are: insufficient driving experience to have an educational effect, insufficient teaching experience of one of the teachers, or insufficient motivation of the subjects. The norms for both the photo test and the film test still need to be established (What is the lowest possible passing-score?). However, the film test first needs to be improved before the norm can be established. At the end of this report a method for assessing the photo test is discussed. This method was proposed by the testing and assessment company Cito. Also, recommendations are made about for follow-up studies to improve the film test and the trainability of hazard perception.
Abstract