Jessica Reed discusses the recent case of the woman who took 90 attempts to pass her theory test. Having been involved in this – albeit a few years ago now, I thought I’d comment.
Prior to the theory test, candidates were asked some Highway Code questions at the end of the practical test. Frankly, this was pretty feeble. Indeed, the preparation very much required candidates to learn for themselves. The practical driving instruction covered the rules in practice, but the theory behind it was not necessarily covered in sufficient detail – there simply wasn’t time. Clients were paying me to teach them to drive, not spend time sitting still going through the theory. During the years I worked as a driving instructor, I had one candidate fail on the Highway Code – and that took some doing. That said, I expected it. I’d questioned him enough during lessons to be aware that there was a weakness – he didn’t bother to read Driving or the Highway Code. As his instructor, there was only so much I could do. In the end, I let him do his own thing and he failed on that element of the test, which he found frustrating given that his driving ability was good. Still, he genned up for the retake.
Since then, the theory test has come about. Initially a paper test – similar to the one I took for my part 1 exam and subsequently a computerised one. Part of it is derived from Driving, so covers a range of driving related subjects, not just the Highway Code and some comes from the Highway Code itself. Most of it is common sense. Passing should be a fairly simple matter as none of it is particularly complex or difficult for someone of average intelligence.
The real bugbear is the Hazard perception test. It shows a series of video clips from the driver’s point of view. The candidate is required to click when they see the hazard. You would think that it is like a video game, but unfortunately it isn’t that well designed. In reality, you have a passive clip where the candidate is nothing more than an observer. There are time windows during which the designer of the test expects the candidate to click. The earlier they click the higher the score. The windows open and close based on time, nothing else, you don’t click on the hazard, which is the instinctive thing to do. You just click. Miss that time window and you fail – even if you click too soon. So, a really observant candidate can see the hazard developing and click, having correctly assessed it, but as the time window hasn’t opened, it doesn’t count and they fail.
You have to game it. If you click too much, the program assumes you are gaming it and you fail. So, you have to click multiple times, but not too quickly. A successful pass in the hazard perception test demonstrates that you have learned successfully to work the DSA’s risible test and has no bearing on one’s ability to successfully assess and react to real life hazards on the road. This is all the worse because people going through their part 1 of the driving instructor exams have to do this test and it is exactly the same as the one used for learner drivers. Yup, an experienced driver clicking correctly on the hazard when they see it, will fail. They have to behave as a learner is expected to behave according to the DSA. Crap doesn’t even begin to describe it.
That said, I’d have expected someone to have mastered it after 90 attempts. My own approach would be to gently suggest to this person that maybe her experience is telling her something here…
The lady who lives next door to me, was in her eighties when she passed her test, but she did not like driving with nobody else in the car so she stopped.
John Gibson
A lass who lives close to me had the problem of spotting the hazzards too quickly, despite having mild learning difficulties.
Some 8 years after my driving test I finally got round to doing my bike test and enroled on a course run by the local council, now ditched in favour of more speed cameras. Sunday morning we did an hour in and out of the cones and an hour on the road with an instructor. Thursdays was theory and film night.
We used Police Motorcycle Road Craft as our training bible (included in the £25 course fee). Best £25 I ever spent. Sailed through the old Part 1 and Part 2 test and would have sailed through theory as well if there had been such a test at that time.
The lessons learned in those 8 weeks have without doubt saved my life on several occasions.
One of my offspring has been through the driving test fairly recently. For the theory test, we were recommended to buy one of those computerised practice software packages that, particularly, take one through the hazard perception test. This all worked fine and was useful practice, which I suspect is necessary for most people. I also had a go or two, it being several decades since I took my test, as I wanted to see what the theory test was like.
On hazard perception, surely any problem with the timing window could be solved by either the actual theory test implementers and also by sufficient practice by learners.
The test itself could look at the total number of clicks and match that, and their timing, in a more sophisticated way than just having a simple time window. Given no reference information on the exact algorithm for scoring, they might already be doing that, and the ‘window description’ might be a simplified explanation.
And for learners, maybe they should just click like they should drive: having seen a hazard, it is not always necessary to take avoidance action; this is either at all or immediately. The avoidance action should be judged correctly and timed correctly; often, in actual driving, just taking the foot off the accelerator is sufficient (and who chooses to accelerate towards a hazard). So, in this part of the theory test, maybe what those with over-quick (rather than false alarming) responses should ‘learn’ (on those practice software packages) to click just a short period after their initial suspicion that there was a hazard: perceive, plan, react. Only the most serious (and mostly unexpected) hazards require actions such as emergency braking and/or swerving. IIRC, I did not have any of those in my software practice sessions.
Given that the substantive range of hazards only occur rarely in most driving tests, having some assessment on hazard perception as part of the theory test (so unavoidable) is actually a good idea.
Best regards
Nigel,
The problem isn’t the idea, it’s the execution. I’ll give you an example. During 2006 it was decided that all driving instructors had to take and pass the hazard perception test as a condition of their ongoing registration. A significant number failed. It became apparent that they were clicking too soon. The test itself was not modified to take into account the level of experience of the driver. Yet experienced driving instructors will inevitably spot a hazard before a novice will – yet the DSA failed to take this into account when making their demands. Consequently people were failing because they were too good at spotting hazards.
I recall at the time looking at DSA material that explained the time windows and recommended practising using the materials you describe. Unfortunately, there is nothing complex about the algorithm – a simple series of windows opens up. The earlier windows carry a higher score. It really is that simple and practising is the best way of upping your score. We coached people to click when they saw the hazard, then count to three before clicking again – it was that easy to game it.
This test could be improved overnight by a half decent games developer, but the DSA has gone for a cheap and nasty solution that measures people’s ability to learn how to pass this test and has nothing to do with their ability to recognise and respond appropriately to real life hazards on the road.
As an erstwhile driving instructor, I can get into a car with someone and tell within minutes what their hazard perception is like – you don’t need a great many to be able to tell that.
So, in conclusion, nice idea, badly executed.
As to the 90 attempts at passing. There should be a limit. Five fails and you are banned from taking it again for life.
After , say, five fails it is obvious the person is just an bloody idiot that will NOT improve even if they take the test 100 times.
Law of averages say you will pass one of these tests by just tickig random patterns, given enough goes at it. So any one passing after a certain amount of attempts has done so by pure fluke, and is no safer on the road than if they had failed.
I passed my test in January and, whilst there are things to avoid when doing the hazard perception (clicking too much, for example), I really don’t agree with your conclusion that we learn how to pass the test not analyze hazards. It’s mostly common sense but it’s a welcome addition to the theory test, in my opinion. It helps you look for telltale signs before an event happens. This is reinforced in the practical test, obviously
The new part of the paractical test, where they waste half of it seeing if you can follow directions at the expense of a manouvre, on the other hand, whilst making it easier for me to pass, was not a good addition. I only had to do 1 manouvre for god sake.
I’m coming at this from the position of an erstwhile driving instructor and current assessor. As a method of assessing competence it is woefully lacking. I took it and passed relatively easily back in 2006. I played with the practice DVD and learned to game it. At no time did it assess my ability to determine and respond to hazards. It was a game. I learned the rules and played accordingly. The objective was to get through a feeble attempt at a test, not to assess hazards – I could already do that to a far higher level than is supposedly measured by this test. Given the difficulty experienced by people who have held an ADI for decades and not necessarily IT literate, this tells you all you need to know about it, frankly. As a measure of an individual’s ability to assess hazards, it is pretty much useless.
Whilst I agree that in theory it would be possible to learn how to “game it”, it didn’t replace anything more worthy in the theory test and the majority of students don’t try to game it, myself and most of the others I talked to included. It’s pretty harmless
The removal of a manouvre in the practical on the other hand was and is a bad idea, in my opinion, even though it benefited me, like I said. What are your views on this?
Also, as a side point, as a Libertarian (which I’m guessing you too are), what are your views on speeding limits, speed cameras, etc. Considering your experience, background and, may I be so bold as to say, brains, I think it would make a really interesting blog subject
Timac.
I hope you’ll forgive me for jumping in on this one, but it may be a good idea to have a good look around the very excellent Safespeed site and forums.
Timac – as I said to Nigel earlier in this discussion, I’m happy with the concept of the hazard perception test, it’s the use of a cheap and nasty product that I object to. Assessment must always avoid disadvantaging the candidate. The HPT disadvantages anyone taking it who has prior driving experience as it is set up for the novice driver. I passed it by waiting for a second or two after observing the hazard, which is why I made the comment about it not measuring my hazard perception skills.
In order to make it effective and fair, two changes need to be made. Firstly, there should only be one time window that opens when the hazard first appears on the screen. This then starts a countdown. Secondly, the candidate should be required to place the cursor over the hazard to indicate what it is that has been observed. Then the measure is twofold – the candidate has correctly identified the hazard and we can tell how quickly.
Having done this, the scoring can be adjusted to reflect the candidate’s prior ability – a lower pass mark for the novice driver, a middling one for someone applying for a new vehicle group and a higher one for people taking their ADI part 1 exam.
I agree with you regarding the practical test. If I had my way it would reflect the ADI part 2 and take about an hour over varying road types and include all the relevant manoeuvres – left reverse, right reverse, parking and turn in the road.
As to your final paragraph, from a purist point of view, I would do away with all speed restrictions and cameras. However, given that we are where we are, I would compromise and keep restrictions in built up areas as these are about right for the conditions. Outside of that, I would do away with them. Speed itself is not dangerous, it is the context that matters and breaking an arbitrary limit does not make the driver dangerous, and being below it does not make him safe – it all depends on the variable factors such as traffic conditions, road surface, weather, ability and alertness of the driver and the condition and capability of the vehicle. Given this, speed cameras have nothing to do with safety and everything to do with revenue raising, so should be dispensed with in their entirety along with the offence of speeding. People should be prosecuted for driving too fast for the conditions – which is effectively covered by such offences as driving without due care and attention.