Interpolis Car Experiment: research findings summarized The Dutch insurance company Interpolis asked SWOV to investigate the effect of three training courses that focus on the road safety of young drivers. These training courses had similar learning goals, but used different didactic tools. The training courses were the existing training of driving skills called TRIALS and two virtual trainings that had been developed by Interpolis, one using an iPad and the other using an Oculus Rift. The effects of the trainings were determined using questionnaires and a hazard perception test. TRIALS and the virtual iPad training were found to have a positive effect on two aspects associated with crash involvement in the research group. Furthermore, TRIALS was found to have a positive effect on self-reported risk behaviour. TRIALS also had some other effects, one of which may be unfavourable for road safety and offers the possibility to further optimize TRIALS. Below the background of the study and the results are summarized. Young drivers have a relatively high crash rate Young novice drivers have a considerably increased crash rate compared to that of more experienced drivers. Drivers between the ages of 18 to 24 years constitute only 8% of the license owners, but they are involved as a driver in 23% of the crashes. There are two reasons for this high crash rate. In the first place, the lack of driving experience plays a role. The second reason is that certain areas in the brain - particularly areas having to do with self-assessment and impulse control - are not yet fully developed. As a result, young people sometimes take risks that are not justified, especially in combination with their limited driving experience. Interpolis investigates ways to improve driving skills Their higher crash rate is the reason that young drivers pay a higher premium, sometimes as much as five times higher than an experienced driver. Interpolis wanted to examine whether it is possible to reduce the damage behaviour of this group by offering them specific youth-oriented training courses. If this is the case, it may be possible to offer them a discount on the premium. Three types of training have been examined in this study: 1. a driving skills training in exercise grounds (TRIALS); 2. a virtual driving skills training with an Oculus Rift; 3. a virtual driving skills training using an iPad. It was investigated whether drivers who had followed these trainings showed any improvement on different aspects that are related with crash involvement. Extensive study among 18 to 24 year-olds The study focused on young drivers aged 18 to 24 in possession of the B-licence. Participants who had been recruited through various channels could register online. After they had registered, they received the first online questionnaire. Next, participants were randomly divided into four groups: the TRIALS group, the Oculus Rift group, the iPad group and the control group. The control group did not participate in any of the trainings, but did fill in all questionnaires. On 11 and 12 July 2015, the Interpolis Car Experiment was held. TRIALS was carried out in the grounds of Brabanthallen in 's-Hertogenbosch and the two virtual variants took place at Interpolis in Tilburg. Immediately after the training all participants filled in a second online questionnaire. In the same weekend, the participants in the control group also received the second online questionnaire. A number of participants (121) in the three training groups were asked to also take a hazard perception test. Three weeks after the Interpolis Car Experiment all participants received the last questionnaire. Only the data of young drivers who had completed all three questionnaires (345) were used for the analysis. Several aspects related to crash involvement investigated The variables that have been examined are related with self-reported crash involvement and can, according to previous research, be influenced by training. These variables are: - knowledge: the number of correct answers in a knowledge test; - confidence: the degree to which one feels safe in traffic and positively assesses the own driving skills; - self-assessment: the degree to which one positively assesses one’s own skills at specific traffic tasks: technical driving skills and avoiding risks; - Locus of Control: the degree to which one thinks that one’s own behaviour contributes to the occurrence of crashes; - risk perception: the degree to which one considers a number of behaviours to be hazardous; - risk behaviour: the frequency in which a number of behaviours is displayed; - hazard perception: the degree to which one can detect relevant hazards on film images. After the training sessions, the participants were also asked about their intention to behave more safely, the degree to which they believe they have learned certain things during training, and the degree to which one believes to be capable of executing specific actions in traffic. Internal Locus of Control and hazard perception related to crash involvement The present study indicates that internal Locus of Control and risk perception are related with the self-reported crash involvement in the research group. Participants who report having been involved in a crash are found to believe less strongly that their own behaviour influences the occurrence of traffic crashes. Crash-involved drivers are also less likely to believe that certain traffic behaviour (e.g. speeding or keeping too little distance) is dangerous. Therefore we specifically look at the changes in these two variables to assess the trainings. TRIALS and the iPad training increase risk awareness After the trainings participants showed the following differences compared with the before situation: - Both TRIALS participants and the iPad training participants have a higher score on the two aspects identified in the research group as being related to crash involvement (Locus of Control and risk perception). They believe that the causes of traffic crashes can be influenced by the driver him- or herself self, rather than these causes being determined by chance. This is a positive road safety effect. Furthermore, they assess various behaviours as being more dangerous. This is also a positive road safety effect. - Three weeks after the training,TRIALS participants report risky behaviour slightly less frequently than they did before the training. - TRIALS participants have a higher score on a knowledge test. We do not know the effects on road safety, but knowledge is supposed to be conditional to being able or willing to perform certain behaviour. - After the training, the Oculus Rift participants have less self-confidence than before the training. This is considered to be a positive road safety effect is: too much self-confidence is thought to be related with crash involvement. - TRIALS participants estimated their own technical driving skills ánd their risk avoidance skills to have increased. Research has shown that overestimation of one’s own technical driving skills is associated with a higher crash rate. However, the same study also concludes that a strong belief in one’s own risk-avoidance skills can compensate for this. Participants in the TRIALS training are enthusiastic and more daring Participants say that they have learned a lot. In the TRIALS group 86% indicate having become more aware of the traffic hazards; in the two virtual training groups this is 25%. The majority in the TRIALS group also say that they will drive more cautiously from now on (77%). Participants in the TRIALS group report having learned what to do in dangerous situations (94%), and also indicate that they are no longer hesitant to drive if conditions are unfavourable (80%). This means that confidence in the own skills has increased. However, this can also be undesirable. If the increase in confidence also leads to an increase in risk-taking while the skills have not yet improved substantially after the short training, this may increase rather than decrease damages. Research shows that it is better to teach young people that they have little control in adverse conditions. Then they will be more inclined to avoid such conditions, for instance by reducing speed, or by not getting behind the wheel at all. This information can be included in an adapted training (see recommendations on the following page). SWOV recommends monitoring damages and optimizing training Based on the findings of this study SWOV recommends the following: - TRIALS is better able to influence aspects associated with crash involvement than the two virtual trainings that were examined. However, effects are also found on the sense of confidence in unfavourable conditions and the assessment of the own technical driving skills. This may increase risks. SWOV recommends monitoring the damage history of TRIALS participants so that it can be established whether they actually cause less damage. - If Interpolis were to decide to actively stimulate participation in TRIALS, optimizing the programme is recommended. Especially those components of the programme that deal with vehicle control, such as practicing swerving and making an emergency stop, should focus on experiencing lack of control and not on the acquisition of skills. This should ensure that young people - with their still limited skills - take fewer risks, especially in dangerous conditions. - It is recommended to supplement the driving skills training with a hazard perception training (possibly as part of an iPad game). Research has shown that in many crashes young drivers were surprised by the situation. They did not notice the other road user, or did not expect that, for example, a particular cyclist could cross. In none of the three training courses that were evaluated, did participants learn to anticipate other road users. A In a hazard perception training this aspect is addressed. Research has shown that such training clearly improves the viewing habits. Including a hazard perception training in a driving skills training, may have positive effects. - The study indicates that the virtual trainings also produced effects; these effects, however, were small and disappeared rapidly. It is advisable to do further research into the possibilities of strengthening the effects and making them more lasting.