The infinite monkey cage

[Geriatrix]

Next on...
Cycle helmets and risk are on the agenda.

[MartinC]

Yes, heard this. Well some of it, I was driving so it didn't have my full attention. It seemed to be good in that it looked objectively at the risks in cycling and the health benefits accrued and concluded that you're far better off to cycle without a helmet than not cycle. But it did subscribe to the basic assupmtion that helmets substantively protected against injury.

[Mick F]

I listened as well, mind you I heard it before as it was a repeat.
Very interesting programme spoilt only by being a little too lighthearted and jokey.

PS:
just checked.
BBC R4 says it was first aired on the 18th Nov 2013.

I don't agree, because I've heard it before ............... or at least most of it.
I wonder whether they've just re-hashed it.

[TonyR]

I listened as well, mind you I heard it before as it was a repeat.
Very interesting programme spoilt only by being a little too lighthearted and jokey.

PS:
just checked.
BBC R4 says it was first aired on the 18th Nov 2013.

I don't agree, because I've heard it before ............... or at least most of it.
I wonder whether they've just re-hashed it.

There was something very similar recently on More or Less that it might have been confused with and that got a lambasting for, ironically, getting it wrong.
http://www.bbc.co.uk/news/magazine-11110665

[Mark1978]

As they said on QI recently, facts have a half life . That given a sufficient number of years, most which we know to be 'fact' now, will have been proven to be wrong.

[kwackers]

As they said on QI recently, facts have a half life . That given a sufficient number of years, most which we know to be 'fact' now, will have been proven to be wrong.

Some 'facts' have a half life measured in millennia judging by the number of false beliefs that refuse to die (and if anything gain momentum).

[CJ]

Well I think it was a very good programme, and that David Spiegelhalter, Winston Professor for the Public Understanding of Risk at Cambridge University, did a lot of good for the promotion of everyday transport cycling as a reasonably safe activity.

I especially liked when he described how he reached the studio on a Boris bike, not in a helmet (shock horror) and answered the hisses of some in the audience by explaining that the benefits of cycling, even bareheaded, outweigh the risks by a factor of 20. He added - and perhaps this is what some of you don't like to hear - that it very much depends on how one cycles, whether one pootles as he does, in common with most people in Holland where almost nobody wears a helmet, or ride in a sporty road-warrior style - for which not only helmets but also body armour should be worn!

I daresay he was playing to the gallery and deliberately exaggerating with that last remark, but there is mounting evidence of radically different risk levels in different types of cycling, and I'm sure he knows all about it, because it's his job to know such things.

The good news is this means the pootling commuters and tourists who insist they don't need helmets, and the sporty racers and mountain-bikers who insist they do, could both be right ...but only about their own kind of cycling. Where we go wrong is when a member of one group attempts to force his or her own cycling reality onto the other group.

[CJ]

As they said on QI recently, facts have a half life . That given a sufficient number of years, most which we know to be 'fact' now, will have been proven to be wrong.

That is mischievously unfair - or a way of excusing superstitious ignorance. It is true that the advances of scientific knowledge sometimes reveal a previous hypothesis, for example that the earth is flat, to be completely up the creek, but the reversal of that previous theory should not be used to undermine the well-founded current idea that the earth is an oblate spheroid.

Knowledge usually advances not by overturning previous beliefs, but by building upon them. All that "standing on the shoulders of giants to see a little further" stuff. And the old hypothesis often remains useful, as a good enough and more user-friendly tool for designing simple systems or when extreme precision is not required. For example: Newtonian physics remains perfecty adeqauate for most engineering problems. Only when they become very large, very fast, or require extreme precision must one first resort to more complex Einsteinian Relativity, or even the utter weirdness of Quantum Mechanics.

For the simple business of building a house, even the Flat Earth theory still works just fine!

[TonyR]

The good news is this means the pootling commuters and tourists who insist they don't need helmets, and the sporty racers and mountain-bikers who insist they do, could both be right ...but only about their own kind of cycling. Where we go wrong is when a member of one group attempts to force his or her own cycling reality onto the other group.

Actually since helmets became dominant and then mandatory in professional road races, cyclist death rates in competition have tripled.

[drossall]

It is true that the advances of scientific knowledge sometimes reveal a previous hypothesis, for example that the earth is flat, to be completely up the creek, but the reversal of that previous theory should not be used to undermine the well-founded current idea that the earth is an oblate spheroid.

I'm not sure that this entirely does justice to what happened in the philosophy of science from the late 1800s onwards.

It's much easier now to see Newton as an approximation than it was when it seemed that an accurate description of the world was nearly complete, and there were just the final pieces to put in place - at which point Einstein came along and knocked down the underlying assumptions of Newton. My understanding is that the effect of this kind of thing on scientists was a considerable demoralisation. We've recovered from it now, and most people believe again what those scientists had assumed - that science marches forward on an unstoppable wave of logic - even though that proved a rather shaky world view when it was actually put to the test.

Arguably, the idea of science producing proven facts is somewhat outmoded, in so far as it tends to involve an assumption that you haven't missed any key factors (like Newton did). Popper, for example, characterised science as producing provisional theories, which it then sought to disprove, and pointed out that proving those theories is logically impossible to achieve.

This isn't about arguing that the world is flat after all. It's about saying that, if your statistics on helmets show this or that result, they are only valid provided that someone later doesn't show that some completely unexpected factor is important - such as the time of day, or the temperature.

In practice, some helmet results have been shown to be affected by unanticipated factors, such as other safety legislation coming into force at the same time (so all accidents decreased, and hence helmets appeared more effective than they were), or the cyclists with and without helmets not being comparable in terms of the conditions in which they were riding.

[TonyR]

I'm not sure that this entirely does justice to what happened in the philosophy of science from the late 1800s onwards.

It's much easier now to see Newton as an approximation than it was when it seemed that an accurate description of the world was nearly complete, and there were just the final pieces to put in place - at which point Einstein came along and knocked down the underlying assumptions of Newton. My understanding is that the effect of this kind of thing on scientists was a considerable demoralisation. We've recovered from it now, and most people believe again what those scientists had assumed - that science marches forward on an unstoppable wave of logic - even though that proved a rather shaky world view when it was actually put to the test.

Newton was not "wrong". Einstein's theories reduce to Newton's theories for velocities small compared to the speed of light. And these sorts of events enthuse rather than demoralise scientists. See for example Stephen Hawking's recent comment that physics would have been much more interesting if they hadn't found the Higgs Boson - because that would have stimulated a whole new line of enquiry into why the theory was wrong and what theory might replace it. Relativity and quantum mechanics sparked a massive boom in scientific endeavour and interest.

Arguably, the idea of science producing proven facts is somewhat outmoded, in so far as it tends to involve an assumption that you haven't missed any key factors (like Newton did). Popper, for example, characterised science as producing provisional theories, which it then sought to disprove, and pointed out that proving those theories is logically impossible to achieve.

Science has for a long time been based on the "scientific method" which is that nothing is provable, only disprovable. Most reputable scientists work on the basis that current theories are a working hypothesis that currently best describes the world as we know it. In that context the emergence of new information or ideas that replace or change the current hypothesis is all part and parcel of the process.

This isn't about arguing that the world is flat after all. It's about saying that, if your statistics on helmets show this or that result, they are only valid provided that someone later doesn't show that some completely unexpected factor is important - such as the time of day, or the temperature.

In practice, some helmet results have been shown to be affected by unanticipated factors, such as other safety legislation coming into force at the same time (so all accidents decreased, and hence helmets appeared more effective than they were), or the cyclists with and without helmets not being comparable in terms of the conditions in which they were riding.

Mostly its about abysmally poor "science" by people, mainly medics, who have not been trained in the scientific method and therefore set out to prove their beliefs rather than to seek evidence to test them. As a result we get people who use the wrong measures[1] or have studies that are so riddled with un-addressed confounding factors that the conclusions are meaningless [2]. Unfortunately they appear to be predominantly, but not exclusively, on the pro-helmet side.

[1] e.g. those who claim helmet laws led to a drop in head injuries without considering whether they had also led to a drop in cycling. For example if head injuries dropped 20% but cycling had dropped 30% the law would have led to a 10% increase in the head injuries in those still cycling not a 20% reduction as claimed.
[2] e.g. The original Thompson, Rivara and Thompson study with its infamous 85% reduction claim that compared inner city kids riding on the road mainly without helmets or health insurance with suburban kids riding mainly in the parks with their parent and with helmets and assumed the difference in head injury numbers was due solely to the helmets. When re-analysed by others it was found the same data interpreted in the same way also "proved" that helmets prevent 75% of leg injuries indicating that the data probably indicated a 10% at most (and probably zero) prevention.

[pjclinch]

Mostly its about abysmally poor "science" by people, mainly medics, who have not been trained in the scientific method and therefore set out to prove their beliefs rather than to seek evidence to test them.

While peer review is a Good Thing in science publishing it's certainly not infallible. And of course it remains the case that the peers of medics who have not been trained in the scientific method are, errrr, medics who have not been trained in the scientific method. And the resulting questionable work is then hawked as "proven scientific fact" by well meaning campaigners who have typically not been trained in the scientific method.

Pete.

[CJ]

Ladies and gentlemen: we seem largely to be in agreement.

What a credit to CTC this forum is!

[BeeKeeper]

Better lock the thread and put it in the Too Good to Lose section then!

[drossall]

Newton was not "wrong".

I understand your point. My comments come out of a history & philosophy of science subsidiary to a physics degree - though a long time ago...

However, you're looking at it with a very modern perspective. The way it was put to me, in the 19C, scientists believed that they had it all sorted bar some loose ends. Discovering that precious, central theories were actually special cases, workable only at around 0.1c or below, was quite destructive to the confidence of science (for a time).

You're describing the pure Popperian scientific method. Not sure whether all scientists think about what they do in that way, though it seems to me to have a lot of merit. Kuhn put it another way. However, the popular view often seems to have gone back to Baconian induction.