How can wearing a helmet offer no protection from injury?

[ufkacbln]

[QUOTE 2158852, member: 45"]The claim was that it's more dangerous for me walking than cycling. As soon as we move past quoting statistics to applying them -"you're more likely to...." or ".....is more dangerous than...." rather than "50% of hospital admissions..." then you've moved into talking about risk.

As to the claim, I'm not over 65, nor drunk. Is that the case?

Stopping at "n% of A&E admissions involve a head injury" is no use whatsoever if we're looking at reducing head injury, or risk of head injury.

Should I go on a stop smoking course?[/quote]

Reread the posts?

 

[mcshroom]

[QUOTE 2158958, member: 45"]I asked you a question...



...and by doing so am inviting you to consider more than an outlying statistic. Why are you so averse to doing so? Any chance of an answer.

I don't know whether the numbers are statistically significant. Neither do you. We know that they may be. I'm quite happy to look into it and continue to learn. And more importantly to move closer to being able to give accurate information on the subject.

And the bottom line, as I've already said, is that once you go past quoting a general figure, you're into the realms of individual risk assessment. And then the risk assessment processes come into play. I do know that I've got a better chance of crossing a road safely than a drunk has. And so do you.[/quote]

I think you'll find my request for evidence came first (and is still unanswered), and secondly you are as free to study the information as anyone else, why should I have to spoon feed all the numbers to you. I have claimed that a dataset shows what I have said, you question the dataset so you prove that your question is significant.

 

[mcshroom]

You keep up this line about individual risk assessment, yet all you seem to want to do is discount the data you have as it doesn't suit. I think we are showing you're claimed neutrality on the issue is false. The data is statistically valid. If you have evidence that the numbers are not, as I have asked for over a number of pages, show it to us, I'm sure we'd like to see it. All I can take from your reluctance to actually back up anything is that you have no evidence to back it up.

 

[Miquel In De Rain]

 

[mcshroom]

I doubt this will be released in 3D any time soon MIDR

 

[mcshroom]

I am happy to consider the numbers in more detail, but until you provide some evidence that being young/old/drunk makes a statistically significant difference to the overall dataset then I am going to look at that objectively, not subjectively cherry pick to suit my own ends as I don't believe that the individual cases you suggest are statistically significant.

[QUOTE 2158975, member: 45"]I
Again, there's only one of you. I'll ask you to then please clarify your opinion on my position.....
[/quote]

There's only one of me, there's only one of you (I think), between us we make at least two

 

[mcshroom]

[QUOTE 2159001, member: 45"]Again, you're using this cherry picking distraction as a get-out. I asked you a simple question, which was not about numbers. There's no harm in you giving your opinion.

And then I'd ask that you explain how that in bold above is valid for you without backup. You're claiming that I need to provide evidence on a view that the stats are inaccurate when I'm not holding that view. What's good for the goose....
[/quote]
I'm not the one claiming a flaw in the data. Evidence please, or you can stick to cherry picking (it seems to have decended to mud slinging in the vain hope that some will stick now, but I'll stick with cherry picking as it is quite accurate)

The difference being that I'm not using "we" instead of "I" to shore up some kind of power in numbers. It might make you feel better, but it's not necessary.

I cannot show it. All I can do is facilitate you to show it yourself, which you have done quite succinctly over the last few pages, so it was a joint effort.

 

[mcshroom]

And my answer - not to a level where they would be statistically significant.

If you are not cherry picking, why do you mention being drunk when the stats are given, and why only for pedestrians?

 

[mcshroom]

See you are sticking to the smoke and mirrors again

 

[steve52]

[QUOTE 2158849, member: 45"]You could always try one of these......

[/quote]
thats a hell of a fair ground ride

 

[mcshroom]

What leads me to that conclusion. The risk of being injured in either activity is approximately 1 injury per 1000 people every month, with an error band of 0.5 injuries per 1000 people per month.

To move them by a statistically significant amount you need to move the datum value (1) by the error value (0.5), Therefore after weighting for the number of people injured per age group, or by being drunk, or whatever other spurious subdivision you decide to throw in, these subdivisions would have to be great enough in rate and frequency to half the rate of head injury in the control group.

As subdividing the total subset up into smaller segments, means that the control group always has a greater weighting, then this would mean that you would not just have to have an incidence rate of double the mean, but a value high enough to double the risk when weighted against the size of the subset.

To make the jump to whether a helmet would be beneficial in these cases, you then need to only account for the number of injuries caused to areas that a helmet may be effective against.

Unless and until there is evidence that there is such a disproportionate risk to the other subsets that it would distort the statistical viability of the overall dataset, then there is no reason to believe that the subsets you described are statistically significant, therefore no reason not to use the overall dataset for comparison.

Again if you happen to have data to cover this then by all means post it.

This is an evaluation of the data, not an interrogation, therefore I stated quite simply that it was my belief.

Now if you believe that there are enough issues with the data from these subsets then you are free to believe so, but you haven't said that it is your belief or opinion. You have claimed that the case cannot be made and that the numbers given are a work of deception. In the case of alcohol you have stated that drinking (or even more accurately alcoholism) has an effect on whether you obtain a head injury (our little exchange about Fish and Chips).

You have presented absolutely no evidence for this or where the deception lies however (despite being asked to do so on numerous occasions).

My issue is that you have branded a researched and evidenced position deception, without giving any evidence why.

 

[mcshroom]

The reason i keep referring to the validity of the overall dataset is that as i have said, I do not believe that there is a statistically significant difference between the relative risks of any particular groups and I am trying to explain this to you in order to show that the overall risk reported is valid for assessing the risks in the individual cases you keep mentioning.

While you are trying to help me to explore all your smaller subsets, I am trying to help you understand why the stats are still valid for those subsets, and the reasons why they should be valid.

 

[mcshroom]

What leads me to that conclusion. The risk of being injured in either activity is approximately 1 injury per 1000 people every month, with an error band of 0.5 injuries per 1000 people per month.

To move them by a statistically significant amount you need to move the datum value (1) by the error value (0.5), Therefore after weighting for the number of people injured per age group, or by being drunk, or whatever other subdivision you decide to throw in, these subdivisions would have to be great enough in rate and frequency to half the rate of head injury in the control group.

As subdividing the total subset up into smaller segments, means that the control group always has a greater weighting, then this would mean that you would not just have to have an incidence rate of double the mean, but a value high enough to double the risk when weighted against the size of the subset.

To make the jump to whether a helmet would be beneficial in these cases, you then need to only account for the number of injuries caused to areas that a helmet may be effective against, making the relevant injury rate lower than 1 (+/- 0.5). .

Unless and until there is evidence that there is such a disproportionate risk to the other subsets that it would distort the statistical viability of the overall dataset, then there is no reason to believe that the subsets you described are statistically significant, therefore no reason not to use the overall dataset for comparison in those cases.

 

[mcshroom]

[QUOTE 2159263, member: 45"]You really, really need to have a rethink.

You're using a statistical value covering the generic activity of cycling to calculate the risk of all of the specific activities and scenarios within this group, some of which are very different, and concluding that the risk of all of the activities are similar.

You're using the wrong statistic.[/quote]

I am using a dataset that is valid, as I have explained twice - unless you have evidence to the contrary that these activities are so dangerous that they fall outside the error band of the data.

To do otherwise is to ignore the data and have a guess on your perception of the risk, which is often a long way off the actual risk for an activity.

 

[lukesdad]

No trouble following, only I asked the same question, which you dismissed with

Early/Late, depends on how you view the evidence before you.

You dismissed a question that you later go onto ask, for your benefit, yet you still cling to your outdated notion that you & only you are correct

You do understand the difference between opinion, advice and evidence do you not ? Im hardly ever correct, fortunately I recognise this fact now where is this outdated notion you refer to ? All Ive asked for is for some more reliable evidence than Ive been able to gather on helmet use in certain groups. In reply all ive heard is a load of bull about any group but the ones ive mentioned, as Adrian has mentioned its probably of little significance, if so why do you and your pals KEEP BANGING ON ABOUT IT ? I don't care what you do in relation to helmets, as you probably dont care about my choice either. Where I do draw the line is when blanket evidence is used to cover all groups.

You don't even know or have been interested to ask if, where or when I use a helmet have you ? ( The avatar picture is an event which required helmet use by the way just in case it clouded your judgement )

In fact your just another helmet evangelist spouting the party line