The CycleChat Helmet Debate Thread

[srw]

The statistical evidence you rely on is incomplete and flawed.

Yup. But if it's treated in the right way it becomes far better than anything an anecdote can provide. What's more, the statistics can tell you just how incomplete and flawed your statistical evidence is - and good practice is to explain that alongside your explanation of what the statistics say.

The amusing thing is that if treated in the right way anecdotes can mount up to statistical evidence. As someone incredibly intelligent said on page one, in post 4...

Can I be the first to point out that although anecdote is of interest psychologically it can't tell you anything about either component of the helmet debate as @Shaun has outlined them? For that you do need evidence, collected reasonably systematically and examined dispassionately.

 

[srw]

You what???

You do some sums.
The sums tell you an answer.
The sums don't tell you a perfect answer, but a likely answer.
When you tell someone your answer, you should also tell them how likely the answer is to be right or wrong.
An answer which is extremely likely to be right is a very useful answer. An answer which is a bit less likely to be right might also be quite useful.
If you're basing your answer on one story, or on five stories, or on a hundred stories then your answer is useless, unless you've made sure that your stories are the right kind of stories on which to do sums.

 

[swansonj]

OK, it has some value but, in the context of this thread, the oft repeated claim that it definitely saved his life has little to no value at all.

Ah. Light dawns. As so often, different use of terms has led to an impression of disagreement.

To me, the assertion "the helmet saved my life" is not part of a case report, has little or no evidential value, and contributes volume and heat but no substance or light to this thread.

On the other hand, a description of an incident, complete with any available information on which bits of the body hit what in what order at what angle and what speed, and particularly if including photos of the helmet, is what I'm calling a "case report" and is actually quite useful in developing our understanding - partly because of its rarity.

 

[swansonj]

Yup. But if it's treated in the right way it becomes far better than anything an anecdote can provide. What's more, the statistics can tell you just how incomplete and flawed your statistical evidence is - and good practice is to explain that alongside your explanation of what the statistics say.

The amusing thing is that if treated in the right way anecdotes can mount up to statistical evidence. As someone incredibly intelligent said on page one, in post 4...

You do some sums.
The sums tell you an answer.
The sums don't tell you a perfect answer, but a likely answer.
When you tell someone your answer, you should also tell them how likely the answer is to be right or wrong.
An answer which is extremely likely to be right is a very useful answer. An answer which is a bit less likely to be right might also be quite useful.
If you're basing your answer on one story, or on five stories, or on a hundred stories then your answer is useless, unless you've made sure that your stories are the right kind of stories on which to do sums.

First base to get past is to persuade people that systematic statistical analysis of data is necessary.

Second base is then to persuade them that not all supposedly systematic and statistical data is equally good. And that's not just about the statistical significance. Non-statistical factors like bias, confounding, measurement error etc need assessing. (The main reason we give very low weight to Thompson/ Riverra is not, after all, because it is statistically uncertain, but because it is fatally flawed by a confounding factor.)

Which is why I prefer to see a continuum from random individual "anecdotal" evidence through to better quality, more systematic evidence - but a continuum not a dichotomy.

I think our difference is one of emphasis only not substance.

 

[mjr]

I go on tour for one weekend and you all just keep on posting and some of you very voluminously. Knock it off!

This is the first time I've seen anyone unfortunate enough to argue for helmets and then within a day or two have an accident where they have hit their head and the helmet has done some service*.

Well, it seems that helmet users are much more frequently head-banging, so it was ever likely to happen at some point.

Helmets really are fantastic self-advertising products. It seems they destroy themselves very photogenically in minor crashes, giving their buyers a real "you made the right decision" positive feedback so they rush out and waste another £10 to £350 and it's very difficult to dissuade them of that, to point out that the likely alternative was not to suffer a head injury and possibly even no crash at all.

Can we think of other products that con their users into buying them again and again because of how smart they make them feel...?

There is simply no way to know whether the helmet saved me from further injury, whether cuts and bruises or something more serious, i very much suspect it probably did and I certainly wouldnt want my un protected head to have hit the tarmac with the same force.

Except it wouldn't have been the same force - it would be about half a pound of helmet lighter, which may even be enough of a reduction that your neck may have been able to keep your head from the ground, doing a job that necks have evolved to do.

Also, the edge of most helmets is not part of the tested area in any of the current standards. It seems highly unlikely that the manufacturer designed that bit for anything more than looks because making such an area protective would be unnecessary cost which would leave them open to allegations of dereliction of fiduciary duty from their shareholders.

I am still very much pro-choice and don't think my habits when it comes to wearing a helmet will change, always when on the road bike and sometimes when on the hybrid.

Given the government's statements that the low rate of helmet use is a strong reason they haven't introduced compulsion, it's not possible to be "very much pro-choice" and use a helmet. If you use a helmet, you may be theoretically pro-choice, but your actions are taken as support for compulsion.

I think the main thing I've noticed in this thread is... helmet wearers don't half fall off a lot.

I caught myself doing it too. When I used helmets, I crashed more. I think the last time I used a helmet, I crashed... and I realised it was because I'd used the dratted thing instead of actually attempting to address the reasonably-forseeable cause of a crash. Fortunately, the bike only suffered minor damage.

Since stopping using a helmet, I don't think I've ended up on the floor yet. The bike has done once, but I'd jumped off early

To be fair I think it might have been mjray ... who pointed it out.

Not me. I'm a statistician-turned-software-developer. The finer points of tax decisions often surprise me.

Indeed. However I bet there's nobody out there that wears a helmet in case a deer jumps on their head.
There may be many of us though that wear one in case an unknown event happens that leads to a head impact.

It's absurd, though. You use a helmet in case of an unknown event while cycling but don't use one in case of an about-as-unlikely unknown event while walking - one of those decisions must be wrong and I don't think it's the walking one...

That denial would indeed by absurd. I cannot recall anyone saying they do not need to take care because they are wearing a helmet either. Absurd also.

They don't often say it, but as in my experience, I used a helmet instead of taking sufficient care and crashed as a result. So I think risk compensation is a reality and not absurd at all.

Anyway for all those having trouble with the Deer story here's a little reminder of the definition of the word "accident" you know, the things we sometimes have on bikes.

And yet, the deer collision meets neither of the definitions! Deer in a park that's home to 500 deer isn't unexpected (note the photographer expecting them) and it definitely had an apparent cause!

 

[Big Andy]

Tomorrow when in front of a PC rather than on my phone I will respond fully to your comments. Just to be going on with though, are you prepared to accept that someone wearing a helmet could have an accident that has nothing to do with the fact they are wearing a helmet?

Are you prepared to accept that the force of impact if I wasnt wearing a helmet could still have been enough to lead to a head injury possibly serious?

 

[mjr]

Tomorrow when in front of a PC rather than on my phone I will respond fully to your comments. Just to be going on with though, are you prepared to accept that someone wearing a helmet could have an accident that has nothing to do with the fact they are wearing a helmet?

Of course, by definition of "accident". Are you prepared to accept that helmet users on this forum report far more crashes than non-users? And that the same phenomenon seems to occur on most forums? And that there are various mechanisms (including but not only risk compensation, hearing obstructed by wind noise over straps and decision-making impaired by brain heating) which might explain why helmet-users crashing more may be the reality?

Are you prepared to accept that the force of impact if I wasnt wearing a helmet could still have been enough to lead to a head injury possibly serious?

In the particular crash reported earlier? I accept it could - but are you prepared to accept that, given that the crash put a small crack in an edge section of the hat outside the tested area, it seems rather unlikely that it offered sufficient dissipation to make the difference between serious injury and no apparent injury?

 

[Justinslow]

Are you prepared to accept that helmet users on this forum report far more crashes than non-users?

Maybe they are embarrassed to come out and say they had an accident without a helmet and it hurt their head a fair bit!
After all they could have reduced the "hurt" by errrrrr wearing a helmet

 

[velovoice]

Maybe they are embarrassed to come out and say they had an accident without a helmet and it hurt their head a fair bit!
After all they could have reduced the "hurt" by errrrrr wearing a helmet

Quite a few have mentioned their helmet-less crashes.

 

[srw]

First base to get past is to persuade people that systematic statistical analysis of data is necessary.

Second base is then to persuade them that not all supposedly systematic and statistical data is equally good. And that's not just about the statistical significance. Non-statistical factors like bias, confounding, measurement error etc need assessing. (The main reason we give very low weight to Thompson/ Riverra is not, after all, because it is statistically uncertain, but because it is fatally flawed by a confounding factor.)

Which is why I prefer to see a continuum from random individual "anecdotal" evidence through to better quality, more systematic evidence - but a continuum not a dichotomy.

I think our difference is one of emphasis only not substance.

I agree, as perhaps post 4 (requoted a few posts back) illustrates - even if I (and others) tend to diminish the importance of anecdotal evidence because it needs so much doing to it to make sense of it.

I tend, naively, to assume that the population at large (as represented by posters in this thread) have got their minds to the home run that says that good quality statistical evidence is the gold standard for assessing population-level risk data and applying it to individual choices. In fact, as the last few pages illustrate, the population hasn't even worked out how to pick up its bat, let alone where the batting plate is.

Which is very sad indeed.

 

[srw]

Maybe they are embarrassed to come out and say they had an accident without a helmet and it hurt their head a fair bit!
After all they could have reduced the "hurt" by errrrrr wearing a helmet

Quite a few have mentioned their helmet-less crashes.

It was a helmetless crash involving a nasty blow to the head that someone else took on a ride I was leading that finally persuaded me to apply the science to my own situation. Admittedly it took another year or two, but that's confirmation bias for you.

 

[srw]

Ah. Light dawns. As so often, different use of terms has led to an impression of disagreement.

To me, the assertion "the helmet saved my life" is not part of a case report, has little or no evidential value, and contributes volume and heat but no substance or light to this thread.

On the other hand, a description of an incident, complete with any available information on which bits of the body hit what in what order at what angle and what speed, and particularly if including photos of the helmet, is what I'm calling a "case report" and is actually quite useful in developing our understanding - partly because of its rarity.

It's also (perhaps) labouring the point that "evidence" means different things to different people. And observing that almost by definition it's vanishingly unlikely that a reliable report of "which bits of the body hit what in what order at what angle and what speed" is ever going to be available.

 

[Deleted member 26715]

How's about a "test" where we assess speeding, but exclude BMWs, Audis, Sports cars, Hatchbacks, family cars, petrol driven cars, diesel cars, hybrid cars, and all cars with doors as they are all irrelevant to the test?


Thus proving that all speeding is performed by Renault Twizzy owners

Why would yu want to do that & what relevance to a falling hammer is a speeding car, please at least keep it on topic

 

[Deleted member 26715]

Can we think of other products that con their users into buying them again and again because of how smart they make them feel...?

Cigarettes or the new electronic dummies

 

[Mugshot]

Maybe they are embarrassed to come out and say they had an accident without a helmet and it hurt their head a fair bit!
After all they could have reduced the "hurt" by errrrrr wearing a helmet

I had an off a couple of months ago.
For no good reason whatsoever I rode into a drainage ditch. I emerged rather muddy and bedraggled but otherwise unscathed. The only thing that had changed on that ride to any other was that I was, for the first time ever, wearing a cycling cap which had arrived in the post that day. I put it down to a lack of attention caused by the invulnerability I was feeling due to the cap. You will I'm sure be pleased to know that I did not hit my head at all (well It probably grazed a few blades of grass) and unlike most of the rest of my body there was not even a fleck of mud on my brand new cap.