Help needed: Why a high-viz vest might not work

**Jantar** · 16th March 2011, 21:03

Originally Posted by FJRider

...

In the right hands ... The same set of statistics, can prove opposite points of view ...

A popular misconception. Statistics cannot prove anything; they can only indicate a probability.

**steve_t** · 16th March 2011, 21:10

"Aw, people can come up with statistics to prove anything, Kent. Forfty percent of all people know that"

**FJRider** · 16th March 2011, 21:12

Originally Posted by Jantar

A popular misconception. Statistics cannot prove anything; they can only indicate a probability.

If SOMEBODY believes it ... case proved in some quarters ...

Too many times people get into strife for lack of indication ... probability of possible direction is as often as not ... just guesswork ...

Hence my lack of support/belief ... of statistics being worthwhile ...

**MSTRS:Me** · 17th March 2011, 08:21

Originally Posted by FJRider

Hence my lack of support/belief ... of statistics being worthwhile ...

Can't accept that.
Statistics can be very worthwhile.
The problem comes when the data is incomplete, or particular facts are ignored or manipulated to achieve a pre-determined result.

**Bald Eagle** · 17th March 2011, 08:31

Originally Posted by MSTRS

Can't accept that.
Statistics can be very worthwhile.
The problem comes when the data is incomplete, or particular facts are ignored or manipulated to achieve a pre-determined result.

Well ACC and the Minister would certainly agree with that.

**Swoop** · 17th March 2011, 13:13

Aha!
Solution found. If the gummint (bound to happen under labour as they like to fuck with personal choice) make fluoro vests' mandatory.

**Hellzie** · 17th March 2011, 13:21

http://www.bmj.com/content/328/7444/857.full

Haven't read the whole thing but seems to have some interesting points and references to other useful data at the end.

**bogan** · 17th March 2011, 13:29

Originally Posted by Hellzie

http://www.bmj.com/content/328/7444/857.full

Haven't read the whole thing but seems to have some interesting points and references to other useful data at the end.

Don't make too much of a habit of it!

already posted in this thread you see

**Hellzie** · 17th March 2011, 14:21

Originally Posted by bogan

Don't make too much of a habit of it!

already posted in this thread you see

Ooooops, I did actually read that post, but it's all too hard to click on links when on my iphone. That'll learn me for using an iphone.

**Smifffy** · 17th March 2011, 18:49

Originally Posted by steve_t

How can you guarantee the result of an experiment that tests 'nothing more than a theory'?

I reckon if we look at the statistics that show the number of accidents per 1000 people who currently wear hi vis vs the number of accidents per 1000 bikers who don't wear hi vis, we could find that wearing hi vis is statistically more dangerous

The number of accidents as a percentage of riders is already so small and random that any difference between two randomly selected samples of 1000 would not be statistically significant.

The biggest problem with statistics is that they are quoted by bureaucrats, politicians and salesmen without any understanding and a barrow to push.

**steve_t** · 17th March 2011, 19:00

Originally Posted by Smifffy

The number of accidents as a percentage of riders is already so small and random that any difference between two randomly selected samples of 1000 would not be statistically significant.

Exactly

Also, per 1000 doesn't necessarily mean go out and find 1000 of each and poll them

Poll everyone!

LOL

**FJRider** · 17th March 2011, 21:33

Originally Posted by MSTRS

Can't accept that.
Statistics can be very worthwhile.
The problem comes when the data is incomplete, or particular facts are ignored or manipulated to achieve a pre-determined result.

I think you better learn to accept it ...

The "problem" is ... those reading those statistic's ... in a position of Authority to make decisions on those statistic's ... do not know (or wont say they dont know) if the data is incomplete ... facts are ignored ... or manipulated ...

**Smifffy** · 17th March 2011, 21:40

Originally Posted by FJRider

I think you better learn to accept it ...

The "problem" is ... those reading those statistic's ... in a position of Authority to make decisions on those statistic's ... do not know (or wont say they dont know) if the data is incomplete ... facts are ignored ... or manipulated ...

The more I think on it, the more I think it's the media that does it. Some twonk does a 6 month polytech course looks good in a cheap suit/dress andgoes out to tell the world that "studies show that...." by extrapolating a year 9 classroom survey to an entire population. Then they get their agency to arrange an interview/grilling with the relevant minister, or a bashing session with the opposition spokesperson.

Then the sheeple that vote along party lines are polarised according to which gloryhound accepted the interview, and start flooding the talkback lines.

**FJRider** · 17th March 2011, 21:50

Originally Posted by Smifffy

The more I think on it, the more I think it's the media that does it. Some twonk does a 6 month polytech course looks good in a cheap suit/dress andgoes out to tell the world that "studies show that...." by extrapolating a year 9 classroom survey to an entire population. Then they get their agency to arrange an interview/grilling with the relevant minister, or a bashing session with the opposition spokesperson.

Then the sheeple that vote along party lines are polarised according to which gloryhound accepted the interview, and start flooding the talkback lines.

WHO CARES WHO (or how many) "DOES IT" ... Belief and faith in statistics could be an embarrassment to you ... YOUR choice to believe ...

**Jantar** · 17th March 2011, 22:20

Originally Posted by FJRider

If SOMEBODY believes it ... case proved in some quarters ...

No. Belief is not proof. Any student who has taken even a basic level of statical studies knows that statistics merely allows a way to present data to predict a possible and/or likely scenario. Statistics cannot be used as a proof. As Albert Einstein is quoted as saying, "If your experiment needs statistics, you ought to have done a better experiment"

Originally Posted by FJRider

Too many times people get into strife for lack of indication ... probability of possible direction is as often as not ... just guesswork ...

Hence my lack of support/belief ... of statistics being worthwhile ...

Yes, making predictions from statistical data is just guesswork, but it can be calculated guesswork, rather than random guesswork. The trick is in understanding just what the statistics tell you.