Why advanced driving courses don’t always make you a safer driver
You are wasting your money if you think the average advanced driving course will make you a safer driver.
SORRY TO GO against the received wisdom here, but why don’t we step back from anecdotal stories or what “seems” right, and instead look at actual evidence and studies by specialist researchers who deal with facts and data. Here’s a few:
Although some portion of teenage crash involvements can be accounted for by poorer basic vehicle handling skills, the research suggests that it is young drivers’ immaturity and inexperience, and the resultant risk-taking, that contribute most to their increased crash risk.
—Teenage Driver Risks and Interventions, Scott V. Masten, January 2004.
Low risk on-road behaviour basically requires three things:
- the acquisition of necessary skills
- the ability to apply these skills efficiently and effectively when operating on the road and in traffic
- the willingness or motivation to apply these skills when operating on the road and in traffic
When trying to place a framework on novice driver safety, the interplay of the following global factors should receive consideration:
- the skilled performance literature indicates that novices perform significantly worse than their more experienced counterparts for a variety of reasons associated with the nature of information processing
- there is mounting evidence that the riskiness of novice driver driving cannot be fully explained by skill decrements and that there is a motivational component contributing to their over-representation in accident statistics.
….evidence in the area serves to further reinforce the view that novice driver safety has two, complementary aspects – the skills-based and the motivational and that safety improvements can potentially be derived from both.
–AN OVERVIEW OF NOVICE DRIVER PERFORMANCE ISSUES: A LITERATURE REVIEW: Alan Drummond
In a previous study, it is well documented that adolescents are more likely than adults to engage in risky behaviour (Arnett, 1992). Most evidence suggests that risk-taking is the most important major factor underlying the high crash rates among teens.
— The Relationships between Demographic Variables and Risk-Taking Behaviour among Young Motorcyclists Siti Maryam Md and Haslinda Abdullah
There is no sound evidence that either advanced or defensive driving courses reduce the crash involvement of experienced drivers who attend them
OK, I’ll stop now. There’s more excerpts at the bottom, and even more on Google. Feel free to do what I’ve done over the years and read some of the papers.
What is a “good” driver?
Before we go any further, let’s define a “good driver”, a term often used, rarely defined. It could mean a driver that, compared to the average, is:
- safer – less likely to crash
- quicker – can drive faster
- more specifically skilled – for example able to back a trailer, drive a 4WD offroad
- smoother and more sympathetic – less wear and tear on the car, a limousine sort of driver
- able to drive more economically
Some very safe drivers are very poor at backing trailers or controlling skids. Does that make them a bad driver? I would say no, but the prevailing wisdom looks at immediately impressive skills, not a track record of no accidents.
You have to match the training to the objective. The problem is when “good” or “better” is not defined, and you end up using the wrong training for the wrong purpose.
Now let’s look at driver training programmes, because ” advanced driver training” is a pretty broad term: I’m writing this from experience as a student having done many, many of these courses, and also as an instructor and designer of courses as diverse as 4WD, winching, towing and car control, plus work as a car demonstrator.
Types of driver training
- Race training – focuses on driving the car fast: skills such as racing lines, skid identification and recovery, car setup, trailbraking, heel’n’toe shifting. Much emphasis is placed on car control.
- Stunt skills – controlled skidding, understeeer and oversteer, drifting, handbrake turns, j-turns and the like. Another car-control focused course.
- Driving Experiences (short) – these are a few laps in some fast car or other. You will not learn very much, and the idea is to give you a thrill, a “say you’ve done it” moment. The instructor is there to stop you having an accident, and to ensure you have fun.
- Driving Experiences (manufacturer) – carefully not named ‘training’, these courses are primarily there for you to have fun and get a positive brand association, and never expose the car in a negative light! There is some element of training with the objectives of improving car control and safety. Usually half to a full day.
- Defensive driving – Wikipedia has a good definition – ” driving to save lives, time, and money, in spite of the conditions around you and the actions of others.” This is the “everyone else is an idiot” school of thought.
- Low Risk Driver Training – same objective as defensive, but a big, big change in philosophy where the onus is placed on the driver to take fewer risks regardless of right of way or legal right, yet still be assertive. In general terms, defensive is about imparting skills, low risk is about changing attitude and that is a very significant difference. Low risk is superseding defensive.
Very few courses are pure to the definitions above, for example defensive driving usually includes quite a bit of of car control, but nevertheless the categorisation works. There are also specific skills courses on towing, 4WD and the like but we’ll leave them out of it for the moment.
So before anyone says “advanced driver training” does or does not work, we need to define what we want to achieve from the training, and what that training is.
Let’s assume we want a safer driver. And here’s the contentious statement: what you don’t want is a course that involves car control or racing techniques, slaloms or any of that fun stuff you find on car control courses. While improved car control does have benefits for safety, it’s by no means the most important thing to focus on, and too much of a focus may in fact be counter-productive. Now to prove that point.
But surely car control courses work?
Try and find one, just one study that shows focusing on advanced car control skills makes road drivers safer. Send it to me if you do find it.
Why don’t car skills based training work?
The answer is in all the references in this article, but it can be summarised as follows:
- the basic mechanical skills of car control, sufficient to drive on public roads, are very easily acquired. There are 5 year-olds who are ace go-karters. A 17 year old just made it to Formula 1. The average driver has no problem controlling or manouvering a car without crashing. Maybe not very skillfully, but we’re talking safe not driving perfection.
- people crash mostly because they take risks, not because of lack of car control skill. Don’t we all drive very nicely when we’re being followed by a police car?
- so we know how to drive safely, we just don’t bother. That means safety is about motivating drivers to ensure that they actually drive as safetly as they know how.
- car control skills are useful, but cannot be learned in a day, need constant practice to keep current, and do not address the major causes of crashes
- the safety-related skills drivers lack tend to be observation, anticipation and risk assessment, not ability to hold a drift or manage skids
- pure car control skills do have a benefit, but it’s nowhere near the biggest safety factor, and increasing them may lead to false confidence and therefore be counter-productive
What sort of driver training lowers crash rates?
Training that addresses the reason people crash, which is their attitude to risk more than their skill, and their motivation to drive more safely.
This is the basis of low-risk driver training, where people are made aware of their propensity for risk, and their appetite for risk decreased.
Low-risk driver training:
- motivates the driver to drive as safely as they can, all the time
- changes the perception from accident to incident, ie “it was their fault’ or “it just happened” or “the road..the car…the everything” to “what could I have done, regardless of fault, or legal right of way”
- gives drivers skills in anticipation and observation so they can assess risk earlier
What about aviation?
Some people compare aviation training to driver training. Having qualiflied as a pilot myself, and then done a fair bit of instruction I can confidently say that pilot training is light years ahead of driver training, and far, far, more thorough.
However, pilot training does not involve spins or aerobatics, which is the equivalent of car control training. The closest you get is stalling, when you fly so slowly airflow over the wing breaks down and you fairly suddenly lose lift. The reason that’s done is so that pilots can recognise a stall before it happens (an “incipient stall”) and take avoiding action. The equivalent in cars would be recovering from a locked brake while braking. The difference is that with stalling it is presented as something to avoid, whereas skidding in car courses is more “if this happens, here’s how you recover” leading to unwarranted self-confidence. There is also considerable theory to go through in pilot training including stalling and aerodynamics, all part of a formal syllabus controlled by the relevant government authority so such exercises are carefully placed in context – unlike car driver training.
Pilot training is also to a competency. You keep going, no matter how long it takes, until you achieve the desired standard or give up. The instructor has to sign you off as competent, and that will only happen once they are confident you can demonstrate the skill over and over because it’s lives on the line.
I have spent many, many flights on specific skills before I could sign the student off as competent. In contrast, with car post-license training there’s a set number of attempts at a skill and that’s it, you move on regardless, and the student gets a certificate of attendance, not of competence.
Also, pilots are regularly re-tested and have to keep their skills current – for private pilots in Australia this is once every 24 months. When I was instructing we did minor checks every three months and full checks every year, which included stalls.
Nothing like that happens for private drivers, once you qualify that’s it, and there are no post-license training standards.
But here’s a maxim from the world of aviation that’s applicable to car driving: “above average pilots use their above average airmanship to avoid situations where their above average skill is never needed”.
Translated, the good pilots stay out of trouble and that’s why they’re good. A skilled car-control driver who relies on their car-control skills every day is a bad road driver.
But I car-controlled my way out of a situation!!!
Yes, we all did at some point. Well done you. But you know what? You utterly failed as a road driver if you needed to resort to that sort of skill. Refer to the maxim above.
A big part of low-risk is taking the attitude that accidents don’t exist, only incidents. An accident is an unavoidable situation, an incident is one you could have avoided, and almost all ‘accidents’ are incidents. Again, a look at aviation crash investigation is instructive. You don’t find air accident investigators writing reports that say “well, it just happened and nothing could be done about it”.
Yes, but what if you ARE in a bad situation?
If you are in a situation where you are sliding across the road, then yes, of course car control skills will be useful and could save lives. But:
- how do you acquire those skills? A one-day training course is not enough, you need a lot more time than that and constant, constant practice
- there is a danger, as the evidence shows, that drivers become over confident and think they can car-control themselves out of situations when in fact they cannot. Being able to recover a skid, once or twice, on one day, in one car, on one road when you’re prepared for it is quite different to recovering a different car, unexpectedly, in any given situation.
- On public roads there is often insufficient space to recover a skid. On a racetrack if the back end steps out you can forget about the apex and gather the car up….do that on a public road and you’re very likely into the opposing traffic.
Is racetrack and car control training really of no value on the road?
There’s no evidence I could find to say it is – but that’s because the studies look at safety. I believe car control training can certainly help make you a smoother driver, and even a more fuel-efficient one, and you could certainly call that “better”. Even for safety techniques like lifting your vision further down the road definitely helps, as does the correct steering technique, braking and so on. It’s all good stuff.
The problem, as the research indicates, is that this sort of training doesn’t address the true factors of most accidents, and doesn’t help very much with better being defined as “safer”.
Driver training has to focus on the biggest causes of crashes. For example, people crash because of car engine failure. But very, very few…so dealing with engine failure is not a priority. The research shows people crash mainly because of risk-taking behaviour which can be fixed by an attitude change, and the skills they need are mostly around observation and anticipation. Exactly how the steering wheel is held and their smoothness of braking are not as important.
There is also evidence from the insurance companies. Ask any one of them, and you will find the group of drivers least likely to crash are middle-aged women. The group most likely to crash are young men. Now we have to go anecdotal, but I think we can all agree that middle-aged women are not, as a group, particularly interested in cars compared to young men. And if you had a test of car control skill, then my experience is that the young men would come out well ahead of the middle-aged women.
Yet the women have fewer accidents. So what if they can’t brake hard into a corner…what about their driving record over the last five or ten years, not just on this artificial-situation day? Take ten blokes and ten women and you’ll find the women have fewer accidents, and the blokes will be more likely to blame everything but themselves.
So car control skills are totally useless?
No, just not as important as people think they are. The research tells us that skills do matter, because elderly drivers have a high accident rate. These drivers are not risk-takers, far from it, but their ability to process information and drive the car decreases. And sitting correctly behind the wheel, driving smoothly…it all helps for safety. It’s just not as important as your risk-taking attitude, and your ability to anticipate developing situations.
If you want another view on this subject read this from Paul Murrell. Feel free to comment with your views…
Even more evidence…
Here’s a few more excerpts from papers on the subject of driver training.
High teen crash risk is due to a number of factors, including (obviously) a fundamental lack of driving skill. However, contrary to what one might think, the evidence suggests that poor vehicle control skills account for only 10% of novice driver crashes; the remaining 90% is accounted for by factors such as inexperience, immaturity, inaccurate risk perception, overestimation of driving skills, and risk taking (Edwards, 2001).
Evidence shows that in the USA the highest skilled drivers (registered race and rally car drivers) have a much higher crash rate than the average driver (Naatanen and Summala, 1976).
… alleged benefits [of skidpan training] rest upon the assumption that a substantial proportion of crashes are attributable to a lack of vehicle-control skills: increased exposure to assorted manoeuvres on a skidpan will improve these skills and thus reduce accidents. However, again the evidence does not stand up to close examination: attendance at skid training programs has increased rather than reduced crash involvement (Langford, 2002, P.36).
An analysis of accident and conviction data for a two year post-training period showed no statistically significant differences between any of the groups, (ie trained vs. untrained young drivers)
Strang, Deutsch, James & Manders (1982) A comparison of on-road and off-road driver training. A comparison of novice driver training Road Safety & Traffic Authority.
Post Licence driver education and driver improvement programs have shown no differences between drivers who have completed the program and those who had not.
Circosta and Salotti (1990) A review & discussion of issues related to the training and education of novice drivers. Brisbane; Road Safety Division, QLD Transport.
Driver and rider training programs have suffered over recent years from the fact that many skilled and careful evaluations have shown little effect, none, or even a negative effect on accident rates when it has been employed.
Henderson (1991) Education, publicity and training in road safety; A literature review.
Racing drivers, young drivers and male drivers, the very groups with the highest levels of perceptual-motor skills and the greatest interest in driving, are the groups which have the higher than average crash involvement rates. This demonstrates that increased driving skill and knowledge are not the most important factors associated with avoiding traffic crashes. What is crucial is not how the driver can drive (driver performance) but how the driver does drive (driver behaviour).”
The clear failure of the skill model underlines the need to consider motivational models that incorporate the self-paced nature of the driving task.
Leonard Evans (1991) Traffic Safety & The Driver
In our view, the increased accident frequency of the racing (trained) drivers is not due to their superior driving but can be more likely contributed to a greater-than-average acceptance of risk, which induced them to pick up the activity of car racing to begin with.
Offers such as track days giving youngsters the chance to drive quickly in safe conditions would need to be considered very carefully and we strongly advise you need to further examine the available evidence in greater detail before proceeding with these ideas. Norwegian studies from 15-20 years ago (reported in Tronsmoen 2008) found an increased accident rate from people who had been on skid courses – probably because of over-confidence.
Courses have been trialled extensively in the U.S., taught by police or specific driving schools, however a number of studies (U.S., Norway) indicate that young drivers who joined these courses went on to have higher crash rates than those who did not (Jones, 1993; Glad, 1988). Williams and Ferguson (2004) attribute this to inspiring overconfidence and/or that tempt young people to try out the manoeuvres, while Horswill et al (2003) suggest that raising people’s estimations of their hazard detection skills through hazard perception improvement interventions may cause them to drive more riskily because they think they can handle the risk better.
Encouraging Road Safety Amongst Young Drivers: How can Social Marketing Help?
“There is little evidence that courses teaching advanced driving manoeuvres such as skid pan control improve driver safety, and they can produce adverse outcomes”
(Williams and Ferguson 2004, p.4) Wilde GS (1994) Target Risk
“These results suggest that a high level of driving skill is associated with a high crash risk. This apparent contradiction could be explained as follows: the belief of being more skilled than fellow drivers increases confidence in one’s abilities more than it increases actual abilities. A high confidence in one’s abilities could lead to an aggressive style of driving that could lead to more critical situations.” OECD (1990) Behavioural adaptations to change in the road transport system. Road Transport Research.
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.