Helmet propaganda debunked
At Thu, 23 Jun 2005 03:02:06 GMT, message
was posted by
(Bill Z.), including some, all or none of
The words I snipped (about an independent issue) are not relevant.
The fact that Rodgers found no benefit is not relevant, and the fact
that he found a small but significant increase in fatalities is? Of
course, Bill, if you say so.
claimed there was a statistically significant *increase* and I basically
asked if you believed it and what confidence level *you* would consider
adequate to make that claim.
No, Bill, I said that Rodgers, the only study I can think of which
comes close to the standards of statistical significance you appear to
require, found no benefit for injuries and a small but significant
increase in injuries. As usual you want to turn that round and start
from the assumption that there is benefit, and that the study can't
disprove it, but as usual that is arse about face because as usual in
broadly based studies the finding is: no benefit.
As it happens the 95% confidence intervals in Rodgers are such that
your supposed zero increase is outside them - which is why Rodgers
said the increase was significant, I guess.
You are of course aware that the confidence intervals in pro-helmet
studies are typically very large indeed due to the tiny sample sizes.
A confidence interval half the size of the stated figure is not at all
uncommon, and I seem to recall one study with a 95% confidence
interval stretching from 6% to over 90%.
But I forgot: you are not looking at the facts, you are looking for
support for a preconceived position.
As stated: what you said was a partial quote, you evaded the fact that
the HI figure (with an even smaller SD than the fatality figure, due
to larger numbers) was zero within the limits of experimental accuracy
- actually the risk was slightly but not significantly increased.
You appear to be the one in denial here.
As usual, you are handwaving. What level of statistical significance
do you think is significant? I'll note too that you never seem to
quantify what you are talking about.
Because the question as put is irrelevant. The level of significance
in whole population studies is considerably better than in the tiny
pro-helmet studies, and your question is founded on the idea that a
finding of zero benefit is really a finding of benefit as long as some
arbitrarily chosen level of benefit never claimed by any pro-helmet
study lies somewhere within the 95% confidence interval. The reality
is, making injuries 10% worse is just as likely as making them 10%
better, form the data, so asking what level of benefit might not be
denied by zero pus or minus x% is simply begging the question.
As I asked before, what sample size do you consider to be adequate?
All you have to do is to produce a number.
As stated, rendering complex issues into simple headline grabbing
numbers is the source of the problem in the first place. Quite why
you think that this question is relevant is beyond me anyway: not one
pro-helmet study of which I am aware has a population base as big as
the New Zealand population, and the only study which includes anything
like the numbers you appear to want is Rodgers, showing no benefit.
OK, you could probably also count the CPSC study showing a 40%
increase in risk as helmet use rose from under 20% to 50%.
I wonder why it is that you will only accept a finding of no benefit
when stated in terms which assume the opposite? No, on second
thoughts perhaps I can work that out.
Well, did you email him to ask? You guys try to pass off statistical
noise as facts all the time.
Why should I? His figures are published. Look at his tables 4 and 5.
Why should I need to email Rodgers?
So he'll get a chance to laugh at you.
In a way that he would not laugh at your attempts to claim that a
small but significant increase in fatality, consistently over 15
years, is really evidence that helmets work as advertised?
rest of Guy's rant snipped out of boredom
As ever, Bill switches to "laa laa I'm not listening" mode and
pretends that "rant" means disagreeing, with the support of evidence.
As ever, nobody is fooled.
May contain traces of irony. Contents liable to settle after posting.
85% of helmet statistics are made up, 69% of them at CHS, Puget Sound
|Thread||Thread Starter||Forum||Replies||Last Post|
|Helmet propaganda debunked||[email protected]||Racing||17||April 27th 05 04:34 PM|
|Ontario Helmet Law being pushed through||Chris B.||General||1379||February 9th 05 04:10 PM|
|published helmet research - not troll||Frank Krygowski||Social Issues||1716||October 24th 04 06:39 AM|
|Reports from Sweden||Garry Jones||Social Issues||14||October 14th 03 05:23 PM|
|Helmet Advice||DDEckerslyke||Social Issues||17||September 2nd 03 11:10 PM|