Derek C
May 26th 10, 09:30 AM
On May 26, 5:38*am, JMS > wrote:
> On Tue, 25 May 2010 21:00:59 +0100, "DavidR" >
> wrote:
>
>
>
>
>
> >"Derek C" > wrote
>
> >> As one of the control groups were cyclists who had *not* suffered head
> >> injuries, but had suffered other injuries including leg injuries,
> >> maybe this is the cause of this misconception. Wearing cycle helmets
> >> would/should have made no difference to this relatively small selected
> >> group. This is one of the problems with hospital based studies, where
> >> the numbers are generally not big enough to be statistically
> >> significant. However every hospital study I have read from a variety
> >> of countries all suggest a reduction in death and serious brain injury
> >> when helmets are worn, so this must be a real effect.
>
> >> BHRF are just playing with statistics to prove their dubious point!
>
> >I am not a statistican but there is a flaw I would have thought should be
> >obvious.
>
> >The group that did not have head injuries cannot possibly have qualified as
> >a control group. They merely divided up the entire set of injuries into
> >those that didn't have head injuries and those that *included* head
> >injuries.
>
> >We don't get to know the severity of the injuries that accompanied the head
> >injuries so it gives absolutely no clue about the type of accident the two
> >groups had; for all we know one group had trips & falls and the other had
> >serious accidents. (Anyone know more?) How anybody could imagine the former
> >gives any lessons about the latter jusr seems incredibly blinkered.
>
> >So a 95% confidence that it is 85% ********.
>
> Back to my reason for making the post:
>
> are you the only one who has spotted this - or has there been any
> serious review of the paper *- pulling it to bits and negating its
> findings?
>
> -- * * *
>
As I have only read the abstract of this paper, I Googled it to try to
get the full version. Guess what the first link was? Yes it was
www.chapmancentral.co.uk !
Derek C
> On Tue, 25 May 2010 21:00:59 +0100, "DavidR" >
> wrote:
>
>
>
>
>
> >"Derek C" > wrote
>
> >> As one of the control groups were cyclists who had *not* suffered head
> >> injuries, but had suffered other injuries including leg injuries,
> >> maybe this is the cause of this misconception. Wearing cycle helmets
> >> would/should have made no difference to this relatively small selected
> >> group. This is one of the problems with hospital based studies, where
> >> the numbers are generally not big enough to be statistically
> >> significant. However every hospital study I have read from a variety
> >> of countries all suggest a reduction in death and serious brain injury
> >> when helmets are worn, so this must be a real effect.
>
> >> BHRF are just playing with statistics to prove their dubious point!
>
> >I am not a statistican but there is a flaw I would have thought should be
> >obvious.
>
> >The group that did not have head injuries cannot possibly have qualified as
> >a control group. They merely divided up the entire set of injuries into
> >those that didn't have head injuries and those that *included* head
> >injuries.
>
> >We don't get to know the severity of the injuries that accompanied the head
> >injuries so it gives absolutely no clue about the type of accident the two
> >groups had; for all we know one group had trips & falls and the other had
> >serious accidents. (Anyone know more?) How anybody could imagine the former
> >gives any lessons about the latter jusr seems incredibly blinkered.
>
> >So a 95% confidence that it is 85% ********.
>
> Back to my reason for making the post:
>
> are you the only one who has spotted this - or has there been any
> serious review of the paper *- pulling it to bits and negating its
> findings?
>
> -- * * *
>
As I have only read the abstract of this paper, I Googled it to try to
get the full version. Guess what the first link was? Yes it was
www.chapmancentral.co.uk !
Derek C