Driver Training

Driver Training

How Effective?

Evidence of effectiveness

  • All programmes should be evaluated against their aim. Most of the time the outcome measure is therefore safety. That is, what evidence is there that driver training (or education) can improve driver safety? In some instances, there may be an alternative aim, such as training to pass the driving test or to discourage reoffending behaviour.

  • Pre-driver

    • A recent review of pre-driver training and education found no evidence to support the presumption that such programmes directly improve driver safety (Kinnear et al., 2013). While there was some evidence of short-term attitudinal change, the vast majority of programmes are not evaluated sufficiently, or at all, to determine if this is a consistent finding.

    • Peck (2010) reanalysed data from historical pre-driver education studies and suggests that road safety education could have a small effect at reducing crash risk in countries adopting GDL. The requirements of GDL would negate the risk of early licensure for those receiving the pre-driver intervention.

  • Learning to drive

    • The vast majority of drivers who pass the driving test have some form of driver training prior to taking the test (Wells et al., 2008). Driving instruction in this sense seems to be useful in helping to develop the skills necessary to pass the practical driving test. However it should be noted that there has been little formal assessment of the effectiveness of driving instruction in GB either at preparing learner drivers for the practical test, or in terms of its impact on safety.

    • There is some evidence that hazard perception skill may be important in safety terms. Wells et al. (2008) and Boufous et al. (2011) have both shown that better performance on the hazard perception test is associated with a lower risk of crashing when driving post-licence. However it is not clear if this is due to training, or simply due to the fact that the hazard perception test measures an important competency for safety, therefore delaying access to driving to those drivers who lack the required level of skill.

    • There is not yet a consensus for how to best train, or test, higher order cognitive skills such as hazard perception. Studies have shown for example that hazard perception can be trained by having drivers undertake on-road training (McKenna & Crick, 1994), by listening to a commentary drive accompanying video (e.g. McKenna & Crick, 1997), and by interacting with hazards in a simulator (Wang et al., 2010).

  • Post-licence

    • New drivers are at the highest risk of being crash involved when they first receive their full drivers licence – higher than at any other time in their driving career. Their crash risk drops sharply with on-road experience, and with increasing age.

    • Several reviews and meta-analyses of driver training and education targeted at improving the safety of new drivers have been conducted. The findings are consistent. All of the reviews have concluded that driver education and training has little or no reliable direct effect on road safety in terms of reductions in collision risk for new drivers (e.g. Brown, Groeger, & Biehl, 1987; Christie, 2001; Clinton & Lonero, 2006; Helman et al., 2010; Kinnear et al., 2013; Mayhew, Simpson & Robinson, 2002; Mayhew, Simpson, Williams, & Ferguson, 1998; Roberts & Kwan, 2001; Vernick, Li, Ogaitis, Mackenzie, Baker & Gielen, 1999).

    • Since these reviews, similar findings continue to emerge. For example in China, Senserrick et al. (2012) carried out a randomised controlled trial of an intervention for recently licenced drivers. The intervention was a DVD-based education program on novice-specific risks, and six hours of in-vehicle training. The intervention group showed no reduction in crashes relative to the control group, and showed trends (although non-significant) for having greater exposure to risk, and inflated perceptions of their ability.

  • In the absence of strong evidence of effectiveness, the most important advice for those seeking to administer training and education interventions for pre-drivers, learner drivers, and novice drivers is to ensure that such interventions are based on formal theory and knowledge from relevant domains such as psychology and the other behavioural sciences, and to ensure that such interventions are properly evaluated using scientifically robust designs (e.g. Helman et al., 2010; Kinnear et al., 2013.; Lonero & Mayhew, 2010; McKenna, 2010; Thomas et al., 2012).

  • Driver improvement

    • A systematic review of post-licence driver training (Ker et al., 2003) looking largely at remedial education courses in the United States, came to the same conclusion as novice driver training reviews; there is no strong evidence that such interventions reduce crashes, and only very weak evidence that they reduce re-offending. There are some promising findings from recent research (e.g. af Wåhlberg, 2011), but as yet firm evidence showing consistent positive benefits of such training are lacking.

  • Certificate of Professional Competence

    • This is a requirement for professional drivers, but there is limited critical analysis or evaluation of this or other HGV training programmes (Murphy & Leach, 2013). These authors also note that there is scepticism in the industry that the EU Directive will have any impact on road safety.

    • Drivers do seem aware of the need for CPC training to address fatigue and stress (two key risk factors in vocational driving), and think that better training would be of benefit (Murphy & Leach, 2013).

    • More evaluation work is needed to establish the actual effects of Driver CPC on safety.

Why does driver training not demonstrate safety benefits?

  • One reason for the lack of demonstrable safety benefits could be that so few programmes are evaluated. It is possible that some aspects of a programme are effective but are unnoticed as no evaluation is performed, or where one is, a scientifically valid methodology is not employed to draw firm conclusions.

  • McKenna (2010) has provided a list of potential reasons why driver education has not produced any clear benefits:

    • Many programmes are not designed on a theoretically sound basis.

    • Many drivers are aware of the risks many programmes seek to inform them about and increasing knowledge does not necessarily change behaviour.

    • Programmes are often short (e.g. half a day) and cannot compete with more day-to-day motivations that impact on driver behaviour.

    • For some the programme may have the opposite effect to that which is desired. For example, some may take pleasure from increased risk, or the programme may increase confidence without increasing skill.

    • Communicating what normally happens (e.g. young driver behaves recklessly) may normalise that behaviour and lead to drivers acting as others expect them to.

    • Pre-driver programmes may encourage earlier licensure, which may inadvertently expose drivers to risk earlier than might otherwise have occurred.

    • Economic conditions often dictate the value placed on evaluation (i.e. a weak methodology is employed to complete it cheaply, or it is not funded at all).

    • Unclear aims. If the general aim is to improve safety then the programme must be evaluated against crashes and injuries. More specific and defined aims are often necessary.

New approaches to driver training and education

  • Despite the general conclusion that driver training and education remain unproven as safety interventions, research work is beginning to reveal some areas of promise. It is important that such new techniques are properly evaluated so that a robust evidence base can be built to guide their content and implementation in the future.

  • More recent approaches to driver training and education have focused on what are termed ‘higher order cognitive skills’ (e.g. perception, motivation and insight) rather than traditional control skills. One such study tested two groups of young, inexperienced drivers (Isler, Starkey and Sheppard, 2011). One group received higher-order driving skill training and the other received car control skill training. The participants who received higher-order driving skill training showed a statistically significant improvement in relation to visual search, improvement in hazard perception, safer attitudes to close following and to dangerous overtaking, and a decrease in driving related confidence. The participants who received vehicle handling skill training showed significant improvements in relation to their on-road direction control and speed choice. However, this group showed no improvement in hazard perception, attitudes to risky driving or driver confidence. Such training shows promise and potential for further research and evaluation in relation to behaviours related to safety.

  • Washington, Cole and Herbel (2010) reviewed changes to post-licence training in some European countries noting that a focus on teaching drivers about self-assessment and anticipation of risk, as opposed to teaching drivers how to master driving at the limits of tire adhesion is demonstrating some promise. Such programs focus on factors such as self-actualisation and driving discipline, rather than low-level mastery of driving skills.

  • Molesworth and Prabhakharan (2012) have similarly sought to apply experiential training, trialled within the aviation industry, to improve drivers’ speed management. This training method is founded on the principles of cognitive engagement where individual’s self-beliefs and skills are directly challenged. Though limited, this series of studies complement a general movement away from direct vehicle control skills and towards insight, life skills, and higher order cognitive skills.

  • One study found that ‘brain training’ software improved performance in lab tests (the useful field of view test, on which the training is modelled) but did not transfer to driving outcomes consistently (Dobres et al., 2013).

  • Another study has sought to test e-learning as an alternative to classroom-based education schemes or fines-only for offending young drivers (af Wåhlberg, 2011). It found some evidence for a reduction in offending in those taking e-learning compared with the other groups, and also a possible reduction in self-reported crashes (although a regression to the mean could not be ruled out). af Wåhlberg is careful to point out that such findings, while promising, are not necessarily generalizable to all drivers, since the focus of the study is on offenders.

  • De Groot et al. (2012) showed that reducing tyre-grip in driving simulator training sessions (using young and inexperienced drivers) led to later reductions in speed in rural road driving in the simulator when tyre grip was normal, both in an immediate transfer session after the training, and a later transfer session the next day. However, historical studies of real world skid pan training have previously been associated with increased crash risk due to increased driver confidence (see Helman et al., 2010 for a discussion).

  • Mynttinen et al. (2010) examined accident data and traffic offences from Finnish and Austrian 2nd-phase driver education systems (mandatory training and education after passing the initial driving test). Self-report attitudes were also assessed. Completers of the 2nd phase education were compared with non-completers (in Finland, those who had not yet completed their course, in Austria those who had not had the opportunity to be involved in the system). Contradictory results were found, with Finnish 2nd phase education showing no impact on safety, while Austrian completers showed a safety benefit. For both groups, self-report measures showed almost no differences. The authors conclude that more work is needed to establish the best content and approach for 2nd phase driver education.

Redefining the role for driver training and education

  • While driver training and education has failed to demonstrate effectiveness for improving safety directly, researchers nevertheless acknowledge that training and education play an extremely important role in developing cultural values, beliefs, skills and legitimising safety relevant legislation such as speed enforcement and graduated driver licensing.

  • Driver training and education to date (because it has been asked to) has focused on preventing crashes and casualties and has therefore been evaluated with this as an outcome measure. Researchers have questioned however whether it is realistic to expect a (typically) short educational programme such as a half day workshop or track session to have any direct effect on crashes, especially given that the research base overall shows that even much larger programmes often have no effect.

  • Another focus that has been suggested for driver training and education is the improvement of specific attitudes to legislation and enforcement activities that we know are related to crash risk (e.g. McKenna, 2010). For example, a driver programme could set out to improve and legitimise young peoples’ attitudes to enforcement (by law or by parents) or to a graduated driver licensing system (that restricts their exposure). If the aim of programme is to improve attitudes to risk factors known to be related to safety (enforcement and reduced exposure) then there is a plausible mechanism for improving safety and also manageable aims for evaluation – the improvement of attitudes to enforcement and the licensing system, not crashes and casualties.

  • Kinnear et al. (2013) summarise that driver training and education should not be expected to improve safety on its own. Ideally, driver training and education should occur within a holistic system such as graduated driver licensing and a developmental curriculum, providing support and legitimacy for the things that do reduce risk, for example enforcement and reduced exposure. The views expressed within Kinnear et al. (2013) are an extension of those put forward by McKenna (2010).

  • The value of training and education is likely to be achieved in collaboration with other measures, as is believed to be the case in other areas of road safety. For example the annual drink drive campaign is believed to be effective because it is delivered in combination with associated enforcement. Similarly, there is a case that education about seat-belt wearing helped legitimise the legislation that was introduced in the UK in 1983 (see e.g. Helman et al., 2010).

  • Lonero and Mayhew (2010) suggest that another way in which driver education needs to be ‘done differently’ is in its use of theory (see also McKenna, 2010). According to Lonero and Mayhew theory in driver education “…is still weak and shows little improvement. Driver education delivery is highly fragmented, and both consolidation and further fragmentation appear to be taking place simultaneously. Driver education needs to be more firmly based in sound research and theory concerning young drivers and, at the same time, in the principles of effective behavior change. It needs better management of the linkage of driver education with parental and community influences, graduated licensing, and other behavioral influences such as incentives and cultural factors.” (Lonero and Mayhew, 2010, p41). This is another example of researchers calling for a more holistic and comprehensive approach, that moves away from the notion (largely discredited by the evidence) that driver training and education can achieve safety benefits without being set within a wider context of road safety (see also Helman et al., 2010; Kinnear et al., 2013).


  • A well designed resource and evaluation are crucial to know whether a programme is improving the attitudes, behaviours and skills being targeted, and whether the resource needs improving or whether it needs scrapping because it is doing harm (e.g. attitudes are getting worse or it is leading to increased exposure to risk).

  • The value to society of driver training and education is probably not in direct prevention of crashes and casualties, but in the legitimising of other activities, and improving safety culture, so that it can provide the mechanisms that do reduce the risk of crashes and casualties. Evaluation in this context is manageable (i.e. the aims are more easily tested) and critical to determine that maximum benefit is being derived from programme resources.

  • The DfT/RoSPA E-valu-it tool ( is a useful resource for those road safety professionals who are looking for support in their evaluation activities.

  • Date Added: 03 Apr 2012, 08:09 AM
  • Last Update: 01 Feb 2014, 11:46 AM