Newly qualified drivers (especially those who are young) have a much higher collision risk that is the case for drivers who have greater experience (Wells, Tong, Sexton, Grayson & Jones, 2008; Forsyth, Maycock & Sexton, 1995; Maycock, Lockwood & Lester, 1991). The importance of driving experience presents an opportunity for driver testing; if testing can be undertaken in such a way as to increase and improve pre-licence on-road experience, it may improve safety for drivers when they begin driving unsupervised. Evidence from Sweden and Australia (Gregersen, Berg, Engström, Nolén, Nyberg & Rimmö, 2000; ScottParker, Watson, King, Hyde & Bates, 2012) has suggested that increasing the amount of prelicence supervised practice undertaken by learner drivers can reduce collision risk in early driving. Research in GB has suggested that the type of pre-licence experience can also have safety implications; Sexton and Grayson (2010) showed that drivers who had experience of driving in busy town centres and in the rain for at least two hours when learning were safer in post-test driving than those who lacked such experience. The DVSA proposed a new practical driving test for trialling in late 2014. The main changes introduced by the proposed test are a longer independent driving section (20 minutes, versus the current 10 minutes) supported by a satnav, and changes to the manoeuvres undertaken. The intention is that these changes will make it easier for the test to encompass a wider range of busier and faster roads. This is because the modified manoeuvres can be done in a wider range of locations than the old ones, and the use of a satnav will enable greater route variability. Since the contents of the test are known to have an influence on the training people undertake to prepare for it, it is hoped that the changes will promote a wider pre-test learning experience, and that this will have an impact on safety outcomes. The project reported here examined the impact of the proposed new driving test on the preparation learner drivers undertake for the test, and on their post-test driving. The project also sought to understand how learners, approved driving instructors (ADIs) and supervising drivers such as parents respond to the revised test, in terms of their opinions, and in terms of learners’ test performance. Learner drivers who were training for their first driving test were recruited to the study through their ADIs from 32 test centres across GB. They were pseudo-randomised to take either the existing test, or the revised test, for licence acquisition. The first learner recruited by each ADI was randomly allocated to one of the two tests, and each subsequent learner from each ADI was allocated to alternating test types. ADIs were informed the test type to which each of their learners in the trial was assigned, and were asked to instruct them accordingly. When learners passed their test, they were sent a survey (through an online link) which asked about their experiences in learning to drive (the types of roads they had driven on, amount of training and so on). They were then sent another survey, six months from their test pass date, which asked about their first six months of driving (again road types, types of driving, and also any collisions or near-collisions). Some learners and ADIs, and also some supervising drivers (e.g. family members providing private practice) were interviewed or attended focus groups. Some test passers were also sampled from non-trial test centres (so-called national comparison group learners) in an attempt to control for potential training bias of which the project team became aware during the project; feedback had suggested that some ADIs were training all their candidates according to the requirements of the revised test. Findings: Views and opinions of the revised test: Views and opinions on the revised test were almost all positive from those ADIs, test passers, and supervising drivers interviewed. It was perceived as relevant to ‘real world driving’ and while interviewees accepted that it could not teach everything required to keep newly qualified drivers safe, it was perceived as providing a good basis for independent driving and decision making. Impact of the revised test on learning to drive: When compared directly with drivers who took the existing test, those who took the revised test showed only very modest changes in terms of their learning experience. There were no differences in the amount of driving undertaken, or in the types of roads driven on. The only clear statistically significant difference was that test passers in the revised test group undertook more training with their ADIs while using a satnav, and slightly more private practice using a satnav. When the two main trial groups were compared with those in the national comparison sample, some other differences were found. Specifically, those in both the trial groups undertook more training on country road and dual carriageways. If we assume that this is the result of some training bias bought about through the revised test ‘spilling over’ into the existing test group, then we might conclude that the revised test has the potential to alter the types of roads on which people learn to drive. When compared with existing test passers, those who passed the revised test also showed some differences in their insight and confidence. Specifically, those in the revised test group felt they needed more improvement on a range of driving skills, but were more confident that they would be a safe driver. This finding can be interpreted as showing that the revised test seems to confer some insight in test passers as to their relative inexperience, potentially leading to some slight increase in confidence that they are going to drive well, and safely, post-test. Finally, revised test passers were more confident that they could drive safely using a satnav post-test, but not that they could drive safely using mobile devices for any other reason. Impact of the revised test on test performance: The revised test appears to have had no noticeable impact on test difficulty, whether measured by self-reported number of attempts before passing, or DL25 minor faults. Impact of the revised test on post-test driving: There were no differences between the revised and existing test participants in attitudes, confidence and driving style at six months post-test. The only driving variable that remained at six months post-test between the groups was that revised test passers had undertaken slightly more driving while using a satnav. Thus their confidence at test pass seems to have translated into their behaviour. Given the potential for training bias, the national comparison sample was included in all the analyses looking at post-test collisions (and near-collisions). This analysis showed that there were no differences between any of the groups on the proportion of test passers reporting at least one collision in their first six months of driving, or the number of collisions per 1,000 miles driven. There was a slight difference in reported near-collisions in the first six months of driving, with participants in both of the trial groups having slight fewer of these than participants in the national comparison sample. Other findings of note: As with previous work of this type, when all test passers who reporting driving post-test were considered, having a lower age at test pass, and greater exposure to driving (driving more miles and driving more frequently), were both found to be associated with higher collision risk. Several other variables also had an impact on collision risk, namely: * Having access to a vehicle owned by parents, relatives or friends (during the post-test period), higher confidence, reported likelihood of avoiding risky driving situations, and time spent with a driving instructor on country roads or driving independently while learning were all associated with lower collision risk post-test. * Time spent driving in busy town centres, frequency of driving for work, and being named on a ‘telematics’ insurance policy were all associated with higher collision risk post-test. Note that only single variables were included in the collision modelling, and therefore there is no way of knowing why some of these variables were associated with changes in collision risk. For example it is possible that those driving for work also drive more in town centres (where there are more opportunities for collisions to occur) and that this is what underlies the increased risk in both cases. On the measure of collision involvement used (self-reported and predominantly damage only collisions) there was no statistically significant difference between male and female test passers. Finally, it is worth noting that the proportion of test passers reporting a collision of some kind within their first six months of driving has fallen considerably since the findings of the Cohort II study (Wells, Tong, Sexton, Grayson & Jones, 2008). Approximately one in five drivers reported a collision in the Cohort II dataset, while the corresponding proportion in this survey was around one in eleven. Future research should focus on understanding the reasons for this apparent increase in safety, as well as on understanding the reasons for the other variables associated with collision risk in this dataset. Summary of findings related to the revised test: The revised practical driving test shows promise. It is well received by test takers, ADIs and supervising drivers. Although there is no evidence of a direct change on collisions in the first six months of driving post-test, its positive effects on insight and confidence, and its potential impact on the types of training undertaken when learning to drive are encouraging. (Author/publisher)
Abstract