You don't need an expensive bike helmet to ride safely---BHSI LabTests Finds no difference between expensive and cheap helmets.

In article ,
mike wrote:

In article ,
says...
On Sun, 11 Apr 2010 10:44:09 -0700, SMS
wrote:

On 11/04/10 9:28 AM, Tim McNamara wrote:
In ,
wrote:

Those damn statstically sound case studies. They always interfere with
junk science.

Dear Steven,

Er, last week, you wrote:

" . . . so the 63-88% range given by Rivara and Thompson is almost
certainly a bit lower than the actual reduction."
http://groups.google.com/group/rec.b...d1e21f302db047

So is it "almost certainly" greater than 88%?

No - it is "almost certainly" greater than 63% (at some unspecified
confidence level).

That's not how Steven wrote it.

It is called statistics - and it doesn't always lie.

Statistics lie when not enough information is presented to properly
evaluate whether the numbers mean what they purport to mean.

In article ,
SMS wrote:

On 11/04/10 4:36 PM, mike wrote:

snip

No - it is "almost certainly" greater than 63% (at some unspecified
confidence level).

It is called statistics - and it doesn't always lie.

This is true. You have to look at the big picture. Case studies
report on the actual cases, but sometimes you have to read between
the lines. Obviously those individuals whose helmets prevented and
reduced injuries to the extent that treatment by medical
professionals was unnecessary are not included in the results.

It's sad to see people so invested in junk science. Every time you
see someone use the "foam hat" schtick, it's an excellent indication
that they're about to lie again. They have utter contempt for the
concepts of facts, logic, science, and statistics because in
contradicts what they want, but know isn't, true.

Thanks for the irony supplement, Steven.

On 11/04/10 6:16 PM, * Still Just Me * wrote:

And therein lies the dilemma that those opposed to helmets don't like
to acknowledge. Even with an accurate random population survey,
analyzing whether or not an accident would have produced a head injury
that was prevented by a helmet is troublesome.

Because it's the perfect dilemma for them. There can be no double-blind
study like that. Case studies of injury levels of helmeted versus
non-helmeted cyclists are problematic for them because so few helmets
cyclists end up being included in ER case studies. Population studies
are problematic for them because population studies show decreases in
injury and fatality levels as helmet usage increases. That's why they
are so vested in promoting the idea, unsupported by any evidence of
course, that cycling levels go down as helmet usage goes up.

On 11/04/10 5:47 PM, AMuzi wrote:

I know a few right wingers, none of whom wears a helmet.

That's the best explanation I've seen of what happened to the brains of
Sean Hannity, Ann Coulter, Glenn Beck, Sarah Palin, etc.

On 11/04/10 7:42 PM, wrote:

How about the FACT that people who are not injured generally don't go
to the hospital? If they don't go to the hospital they are out of the
count - right?

That's a big problem with the case studies. They can compare helmeted
versus unhelmeted cyclists that have injuries serious enough to warrant
medical treatment, but they leave out all the cyclists where the helmets
prevented any injury or mitigated the injuries to a level where no
hospital treatment was sought by the cyclists.

As we've seen, it's a good idea to seek medical treatment even if you
feel okay after a head-impact crash.

The old addage, figures don't lie, but liars figure can come into play
here.

The best way to know if there's lying going on is to look at any
"references" that those that write long diatribes tend to present as the
basis for their position. You can be certain that if anything came from
the "BICYCLE HELMET RESEARCH FOUNDATION" which in fact does no research,
that it's untrue.

On Sun, 11 Apr 2010 22:42:02 -0400, wrote:

On Sun, 11 Apr 2010 18:41:31 -0700 (PDT), Frank Krygowski
wrote:

On Apr 11, 9:16*pm, * Still Just Me *
wrote:
On Sun, 11 Apr 2010 17:10:48 -0700, SMS
wrote:



This is true. You have to look at the big picture. Case studies report
on the actual cases, but sometimes you have to read between the lines.
Obviously those individuals whose helmets prevented and reduced injuries
to the extent that treatment by medical professionals was unnecessary
are not included in the results.

And therein lies the dilemma that those opposed to helmets don't like
to acknowledge. Even with an accurate random population survey,
analyzing whether or not an accident would have produced a head injury
that was prevented by a helmet is troublesome.

sigh This is an example of fantasy that just won't go away!

Among the papers I've cited over 30 times is Scuffham et. al.,
"Trends in Cycling Injuries in New Zealand Under Voluntary Helmet
Use," 1997, Accident Analysis and Prevention, Vol 29, No 1. I got my
copy directly from Scuffham, but you can get one from the library.

Once again: Scuffham and his fellow pro-helmet researchers tracked
down essentially every hospital record of a hospitalized cyclist in
the entire country for well over 10 years. That included a time
period when helmet wearing rose from near zero to (for kids) about
90%. Almost all of that jump occurred in a period of less than three
years.

Their plan was simple: From the hospital records, determine the
percentage hospitalized due to head injury. Of course, they
anticipated seeing that percentage drop sharply when helmet use rose
sharply.

Why would it drop sharply? Read the following paragraph slowly,
TWICE. Then think about it:

_If_ helmets kept people from showing up in the hospital, it would be
detected by the reduction in percentage hospitalized due to head
injury. Seriously - Isn't that clear?

If that's not clear, read it again. Ask for help. I can give
numerical examples.

But back to the study: unfortunately for helmet promotion, they found
no such thing. They go into great detail in the paper about the many
math tricks they tried to find evidence of those missing head
injuries. No matter. Nothing they did could find any evidence of
cyclists protected from hospitalization by their helmets.

Again, for emphasis: The helmets kept NOBODY out of the hospital.

If you don't believe this interpretation of the paper - that is, if
you pretend I'm misrepresenting what they did and learned, PLEASE go
to your librarian and ask to get a copy. Read it, look at the very
clear graphs, and come back here and discuss it.

Or at _least_ drop that Scharfian line of "reasoning," claiming
phantom, undetectable benefits. They don't exist.

- Frank Krygowski

How about the FACT that people who are not injured generally don't go
to the hospital? If they don't go to the hospital they are out of the
count - right?

The old addage, figures don't lie, but liars figure can come into play
here.

Dear SJM,

Frank's point is that if helmets prevented injury in crashes that
previously sent them to the hospital, the hospitalization rate would
drop.

It didn't.

Maybe this example will help you follow Frank's point.

Let's say that a country's hospital system sees 10,000 chest-trauma
patients per year for 10 years years from all causes.

A law is passed that requires all bar patrons to wear kevlar vests in
hopes of reducing the number of chest stab-wounds suffered by people
in bars.

Observation shows a massive increase in kevlar vests in bars.

But the hospital system stubbornly records the same 10,000
chest-trauma patients per year.

The kevlar vests failed to produce the predicted effect.

You can replace the vests with the helmets in the New Zealand study.

You can also replace the vests with an amazingly long list of medical
treatments that really seemed like sure-fire cures, both in theory and
in small case-control studies.

Cheers,

Carl Fogel

SMS wrote:
On 11/04/10 5:47 PM, AMuzi wrote:

I know a few right wingers, none of whom wears a helmet.

That's the best explanation I've seen of what happened to the brains
of Sean Hannity, Ann Coulter, Glenn Beck, Sarah Palin, etc.

That's tellin' 'em!

LOL

On Apr 11, 10:42*pm, wrote:
On Sun, 11 Apr 2010 18:41:31 -0700 (PDT), Frank Krygowski



wrote:
On Apr 11, 9:16*pm, * Still Just Me *
wrote:
On Sun, 11 Apr 2010 17:10:48 -0700, SMS
wrote:

This is true. You have to look at the big picture. Case studies report
on the actual cases, but sometimes you have to read between the lines..
Obviously those individuals whose helmets prevented and reduced injuries
to the extent that treatment by medical professionals was unnecessary
are not included in the results.

And therein lies the dilemma that those opposed to helmets don't like
to acknowledge. Even with an accurate random population survey,
analyzing whether or not an accident would have produced a head injury
that was prevented by a helmet is troublesome.

sigh *This is an example of fantasy that just won't go away!

Among the papers I've cited over 30 times is *Scuffham et. al.,
"Trends in Cycling Injuries in New Zealand Under Voluntary Helmet
Use," 1997, Accident Analysis and Prevention, Vol 29, No 1. *I got my
copy directly from Scuffham, but you can get one from the library.

Once again: Scuffham and his fellow pro-helmet researchers tracked
down essentially every hospital record of a hospitalized cyclist in
the entire country for well over 10 years. *That included a time
period when helmet wearing rose from near zero to (for kids) about
90%. *Almost all of that jump occurred in a period of less than three
years.

Their plan was simple: *From the hospital records, determine the
percentage hospitalized due to head injury. *Of course, they
anticipated seeing that percentage drop sharply when helmet use rose
sharply.

Why would it drop sharply? *Read the following paragraph slowly,
TWICE. *Then think about it:

_If_ helmets kept people from showing up in the hospital, it would be
detected by the reduction in percentage hospitalized due to head
injury. *Seriously - Isn't that clear?

If that's not clear, read it again. *Ask for help. *I can give
numerical examples.

But back to the study: *unfortunately for helmet promotion, they found
no such thing. *They go into great detail in the paper about the many
math tricks they tried to find evidence of those missing head
injuries. *No matter. *Nothing they did could find any evidence of
cyclists protected from hospitalization by their helmets.

Again, for emphasis: *The helmets kept NOBODY out of the hospital.

If you don't believe this interpretation of the paper - that is, if
you pretend I'm misrepresenting what they did and learned, PLEASE go
to your librarian and ask to get a copy. *Read it, look at the very
clear graphs, and come back here and discuss it.

Or at _least_ drop that Scharfian line of "reasoning," claiming
phantom, undetectable benefits. *They don't exist.

- Frank Krygowski

*How about the FACT that people who are not injured generally don't go
to the hospital? If they don't go to the hospital they are out of the
count - right?

sigh Wrong, if the count is done properly.

Obviously, some people need detailed explanations. Let me try again.

Let's pretend helmets work, and that they prevent 75% of hospital
admissions for head injuries. Let's see how that would affect
hospital data.

Let's take two different groups of cyclists: The first group is 1000
cyclists in a place where helmets are (almost) never used - like New
Zealand in 1985. And let's be sure to pick 1000 cyclists who are in
crashes that cause hospitalization. Of those 1000 cyclists, roughly
60% are hospitalized due to head injuries - that is, 600 due to head
injuries, 400 due to other injuries. (That was the New Zealand data
for kids in 1985.)

Now let's take 1000 cyclists in a place where 90% of the cyclists wear
helmets. That would be New Zealand kids in 1992. Let's assume
they're in identical accidents. For 400 kids soon to be hospitalized
for NON-head injuries, the helmets don't matter. But for the 600 kids
that _would_ have been head injured, the helmets supposedly matter;
they supposedly prevent 75% of the head injuries. Of the 600 kids,
450 are saved! Wow! Only 150 get head injured! Got that?

So instead of 1000 kids headed for hospital, only 550 kids go. 450
protected kids are "missing." Can they be detected?

If you've got the mathematical competence of a high-school senior in
Algebra II, yes, you can detect them. You can compute the percentage
of kids in hospital for head injuries in the two cases.

For 1985, you get 600 out of 1000 in for head injury. Yep, that's
60%. For 1992, in our hypothetical "protected" case, you get only 150
out of 550 in for head injury. That's only 27%. The benefit of
helmets is detectable, because the missing "protected" cyclists affect
the hospital data. A much smaller percentage are in the hospital
because of their head injuries.

The problem for helmet promoters is this: The 1992 data still shows
about 60% of the kids in the hospital were there because of head
injury. The expected benefit was just not there. There were
apparently no cyclists who were "saved" from hospitalization. And
BTW, the results were the same for all three age groups studied -
kids, teens and adults.

Is the math clear?

There's more I can say, by way of detail. The details have been given
before, but I'll give them again if you like. But I think I'd better
stick to basics for now, and cover the fine points if needed.

- Frank Krygowski

In article ,
mike wrote:

In article ,
says...
On Sun, 11 Apr 2010 10:44:09 -0700, SMS
wrote:

On 11/04/10 9:28 AM, Tim McNamara wrote:
In ,
wrote:

Those damn statstically sound case studies. They always interfere with
junk science.

Dear Steven,

Er, last week, you wrote:

" . . . so the 63-88% range given by Rivara and Thompson is almost
certainly a bit lower than the actual reduction."
http://groups.google.com/group/rec.b...d1e21f302db047

So is it "almost certainly" greater than 88%?

No - it is "almost certainly" greater than 63% (at some unspecified
confidence level).

It is called statistics - and it doesn't always lie.

Statistics never lie. People lie.

--
Michael Press

On Apr 10, 10:11*pm, Frank Krygowski wrote:
On Apr 10, 9:59*pm, SMS wrote:

They tested six helmets, priced between $10 and $207, and found no
difference in simulations of real-world impacts.

"http://www.bhsi.org/testbycost.htm"

"http://www.hometownlife.com/article/20100408/NEWS10/4080579/1027/You+...."

Wow. *According to that article, "Medical research shows that bike
helmets can prevent 85 percent of cyclists' head injuries."

Isn't that astonishing?

I also like this more recent study:
http://www.ncbi.nlm.nih.gov/pubmed/19944823
where the abstract gives their results as:
"RESULTS: Of 324 patients with bicycle-related head injuries, 90
(27.8%) had severe head injuries. Boys compared with girls had a
higher proportion of severe head injuries (34.1% vs 23.4%; P = .048).
Children aged 5 to 9 years had a higher proportion of severe head
injuries compared with ages 10 to 14 years (65.2% vs 6.4%; P = .043).
Bicycles without reflectors had a higher proportion of severe head
injuries compared to bicycles with reflectors (69.0% vs 5.7%; P = .
004). Bicyclists carrying goods (such as backpacks or weighted toward
the road) and speeding were associated with severe head injury (P .
05). Collisions with vehicles of a larger size resulted in a higher
rate of severe head injury compared with collisions with pedestrians
(76.9% vs 3.6%; P = .043)."

Note the apparent effectiveness of having a reflector on the bike -
the chance of a severe head injury dropped from 69% to 5.7%, so the
presence of the reflector was over 90% effective in preventing severe
head injury. More effective than a helmet, much less expensive, and
far more convenient to use (just install once and leave it on the
bike).

Once again showing the problems associated with taking case-control
studies (and their inherent self-selection bias) at face value.