Skiing vs cycling

[srw]

I have often pondered why seemingly reputable organisations come up with wrong conclusions. The nearest I can get to it is that it's perhaps a question of starting point and hence perspective. I don't believe they have bad statisticians, I believe their statisticians are working within too narrow constraints.

Looking at the membership list (under http://www.eurosafe.eu.com/csi/eurosafe2006.nsf/wwwVwContent/l2eurosafemembers.htm), I'm not sure there will be all that many statisticians working in the member organisations. They're mainly organisations (and some corporates) which are interested in individual safety, not population safety. Without wishing to caricature too much, the bias is towards Health & Safety rather than risk.

What would be really quite interesting would be if a number of individuals with a personal interest in statstical analysis of risk information and enough time on their hands to do the work (and enough money to afford the €240 per year membership fees) joined up to try and influence the direction of the organisation. At only 3 votes per institution and 1 vote per individual you wouldn't need many.

@slowmotion has a point, incidentally. It's this sort of activist organisation that European law-makers seem to lean on quite heavily. In my field it's activist academics and a relatively small handful of individuals who are willing to devote a lot of time to what can be a very dry subject. I can imagine this bunch being more influential than their numbers and status would suggest they should be.

 

[slowmotion]

It's this sort of activist organisation that European law-makers seem to lean on quite heavily. In my field it's activist academics and a relatively small handful of individuals who are willing to devote a lot of time to what can be a very dry subject. I can imagine this bunch being more influential than their numbers and status would suggest they should be.

I suspect that Greenpeace and WWF would fit this description superbly.

 

[slowmotion]

You are quicker than expected tonight.

 

[Fab Foodie]

But then again, if looking at data with an open mind, rather than looking at data from the narrow perspective of our prejudices, were easy, we'd all be Nobel material.

Food Scientists ....

 

[Wobblers]

I thought about going through your argument in detail but I think we both would agree, this thread does not need another sniping match. My opinion on your criticism is that it is just a tad too strident to be balanced.

I don't know much about this NGO but listing their sponsors is just revealing facts and I do not presume these are necessarily biased organisations like you do. I'm not going to go out on a limb to defend them but I am pointing out that their research is being quoted/pointed to by the national press and as you note, people who chose to delve into this information will very likely accept their conclusions and trust work sponsored by them. This clearly does have some effect in public consciousness.

What seems to be interesting though is that I think they claim their data is new as obtained from national institutions. That may be in the sense that they have a wide set of data pertaining not only to cycling but also to all other transport modes. It is not clear to me.

Don't doubt, I believe helmet usage for adults using current designs is more then likely as Adrian puts it a personal choice based upon emotions. For children or some people learning to ride for the first time who are more likely to have less acute accidents, I'm prepared to believe different. But this is not a point I want to argue, just to be clear about.

Anyway, I am genuinely interested and open minded to hear your or anyone else's considered (but please be a chilled out) opinion on what the motive of this NGO would be in making such claims as it has that I did highlight. I am trying hard to keep emotions out of my question. As I see it
1) They may be right and have new data that can back their claims up
2) They may be honestly mistaken - personally I do not think it is right to presume their statisticians are not up to scratch
3) They may be dishonestly mistaken - why would those sponsoring org.s be prepared to sponsor it
4) something else - in my book it is ok just to say I don't know

You misunderstand my argument. I am not saying that they are necessarily biased (any more than any of us are) but this is not, does not use or indeed refer to original research. This is a critical failing. The data is the least biased data available. Therefore this is the source that must be used, unless there is a convincing reason not to. It is simply stated that helmets are effective in casualty reduction, despite the relevant research being contradictory. Furthermore, the assertion that mandatory helmet wearing "has been proven to be effective" has not been shown to be true.

They do not claim to have any new data: it is in the main derived from the relevant national authorities, and is published on the net.

As swansonj has pointed out, the focus is entirely upon safety. This is too narrow, inasmuch that it ignores the detrimental effects that a mandatory helmet law has on cycling numbers. There has not been much work in the literature on this topic, but what little there is suggests that the health effects of less people cycling will increase mortality (mostly due to increased heart disease) over and above the numbers saved from the presumed injury reduction benefits. And remember that whole population studies have not shown any benefit to helmet wearing. The working approach thus is to discount unintended consequences.

So it's (4). They haven't examined the evidence in a rigorous way, for whatever reason. The result of this is questionable conclusions.

As I've said, this is intended to be a document to inform policy makers. Thus there is the absolute expectation that all relevant data and research will have been investigated - and controversial (I use this word in its scientific sense, that is, a conclusion that has not achieved consensus or is not adequately supported by data) conclusions and suppositions explicitly pointed out. This is not the case. This, then, falls below the standards that we have every right to expect and insist on. And if you think that this is too "strident", I think it fair to ask you to examine your own biases.

 

[thom]

You misunderstand my argument. I am not saying that they are necessarily biased (any more than any of us are) but this is not, does not use or indeed refer to original research. This is a critical failing. The data is the least biased data available. Therefore this is the source that must be used, unless there is a convincing reason not to. It is simply stated that helmets are effective in casualty reduction, despite the relevant research being contradictory. Furthermore, the assertion that mandatory helmet wearing "has been proven to be effective" has not been shown to be true.

They do not claim to have any new data: it is in the main derived from the relevant national authorities, and is published on the net.

As swansonj has pointed out, the focus is entirely upon safety. This is too narrow, inasmuch that it ignores the detrimental effects that a mandatory helmet law has on cycling numbers. There has not been much work in the literature on this topic, but what little there is suggests that the health effects of less people cycling will increase mortality (mostly due to increased heart disease) over and above the numbers saved from the presumed injury reduction benefits. And remember that whole population studies have not shown any benefit to helmet wearing. The working approach thus is to discount unintended consequences.

So it's (4). They haven't examined the evidence in a rigorous way, for whatever reason. The result of this is questionable conclusions.

As I've said, this is intended to be a document to inform policy makers. Thus there is the absolute expectation that all relevant data and research will have been investigated - and controversial (I use this word in its scientific sense, that is, a conclusion that has not achieved consensus or is not adequately supported by data) conclusions and suppositions explicitly pointed out. This is not the case. This, then, falls below the standards that we have every right to expect and insist on. And if you think that this is too "strident", I think it fair to ask you to examine your own biases.

McWobble, it was always transparent that Eurosafe are an NGO concerned with influencing policy. The fact that they influenced the Guardian article was always my point. I highlighted their claims not to support them, not to go into the quality of their research but to point out, they have some kind of influence and that their claims are in contradiction to the consensus within this forum. I personally believe they are writing in good faith in so far as I don't believe they are dishonest. Their reference may refer to articles that are not ultimately peer reviewed research at core. Or they might be - I don't think either of us can quite work that one out.
The point is, people looking at the bodies sponsors will see a diversity of institutions and ascribe some credibility to their policy documents. Separately, I don't think you would disagree with all the policies they propose.

@swansonj engaged with the question in my post and I thank him for that insight.

I don't like the stridency you have that takes this discussion back into the details of the helmet debate all the time.

I stated my position on that clearly - we all have some bias but I was honest enough to state the emotional content of it.

Let's be clear, the thing I object to in these debates is the inability to talk about anything regarding this subject without trying to place someone in a particular camp. That stinks.

 

[thom]

[QUOTE 2859494, member: 1314"]
thom came in with a post not thought through confirming her fixed viewpoints, and despite MacW's post above, has gone back to confirming MacW to be strident.
[/quote]
What do you think my viewpoint is ?

 

[thom]

Confused and lacking in understanding?

Answer me straightforwardly then, can you back that up and tell me what of you think I am lacking in understanding of here ?

 

[ufkacbln]

The big problem with all of this is the basis upon which questions are asked and the emphasis on research

The group selected is cyclists and research performed on the selected Cohort. This means that wider questions are never asked

When full A&E admissions are examined then cyclists become a tiny percentage of the head injuries that are seen


What would be very interesting would be to look at ALL head injuries and assess whether a helmet would have been beneficial..... Then and only then look at the cause

That would give a truer representation of the potential contribution of helmets

 

[thom]

The big problem with all of this is the basis upon which questions are asked and the emphasis on research

The group selected is cyclists and research performed on the selected Cohort. This means that wider questions are never asked

When full A&E admissions are examined then cyclists become a tiny percentage of the head injuries that are seen


What would be very interesting would be to look at ALL head injuries and assess whether a helmet would have been beneficial..... Then and only then look at the cause

That would give a truer representation of the potential contribution of helmets

In addition I think there can be a problem, as witnessed in this thread in various places, that the discussion quickly gets very heated and indeed positions are expressed in a strident fashion (I am not referring to you here). I think that is not constructive and for this reason I normally don't pay attention to these threads.

The reason I am more interested in this thread though, is the OP concerns itself with skiiing too, which I do occasionally. This is why I posted a link to an article in the Guardian and highlighted aspects of the references that is being linked to - there are organisations out there that are making claims to influence policy and I think they think they are being honest and thorough about their claims.

I agree with what you are saying about cycling though. The minor criticism I would say is that this thread is also about skiers; you might have asked for similar enquiry there too. It is not clear to me whether there is research that supports the efficacy of helmets at reducing injuries there either.

 

[steve52]

skiing v cycling.with or with snow? without and ill back cycling,

 

[ufkacbln]

This is entirely supposition and NOT evidence based

The logic would be that if we accept the performance design of cycle helmets they may prove to be even more effective in pedestrians or drivers as the impact speeds and hence energy are lower and within the design pararamters

 

[srw]

It is not clear to me whether there is research that supports the efficacy of helmets at reducing injuries there either.

Post #1 in the thread does, in its small way, attempt to provide at least a partial answer...

 

[thom]

Post #1 in the thread does, in its small way, attempt to provide at least a partial answer...

Sure, it seems so long ago… The reports I remember reading though seem to end up with a good few overlaying effects that muddy the waters - changes in behaviour, changes in equipment, age demographics, reporting etc.
As with cycling helmets, it still appears the helmet industry has work to do to prove scientifically their efficacy, while people like me make assumptions of their benefits.

 

[thom]

This is entirely supposition and NOT evidence based

The logic would be that if we accept the performance design of cycle helmets they may prove to be even more effective in pedestrians or drivers as the impact speeds and hence energy are lower and within the design pararamters

Logic that would inform policy/legislation would also take into account a cost-benefit analysis as well, not just the helmet performance itself. Just as people point to detrimental health effects of reduced participation in cycling where mandatory helmet laws exist, runners would similarly take issue were they compiled to wear something that provably prevented injuries of minuscule incidence.