Caer Abred

A more viable explanation: http://www.world-science.net/exclusives ... ncture.htm

I had read a while ago that akoopunkcha does have some effect, this demonstrated the how.

Praise adenosine

(As an aside, if something works how much does it matter how it's explained?)

treegod wrote:(As an aside, if something works how much does it matter how it's explained?)

Absolutely correct. Moast often effect is established before cause.

However it is very easy to assume effect and only a proper scientific studdy can remove the subjective nature of such perceptions in order to establish that the effect is real. For a start a placebo effect is not "real" beyond being placebo. The tendency of alternative practitioners to credit themselves with placebo effects is about as sensible as a car-owner boasting about the speed of their car having free-wheeled it downhill.

Full paper can be found HERE

Very interesting. The stats look good, methods seem good, and obviously using mice there's no placebo.....

BUT.... unless I'm being blind, they don't mention anywhere the most important thing - how many mice they used. Their sample size might be tiny, which would make their stats meaningless.

GRRRR bad science.

How many mice does an experiment need to make the stats meaningful?

I did read somewhere else that someone wasn't too impressed because what might be acupuncture points in humans may not be in mice. However, if there were enough mice to make the stats meaningful then I'd say they found the right point in the mice to make an effect.

Thanks for sharing the link Golden Eagle.

Sample size has to be large enough to ensure that the results were not just due to random chance.

In simple terms... if you toss a coin six times and it lands on heads six times in a row. Does that mean you can firmly conclude that coins always land on heads? Hell no.
You need to toss it over a hundred times to get the true, statistically significant 50/50 occurrence of heads/tails.

Another interesting response to the article HERE

Yes, I know about coin tossing etc. But I'm asking about mice, how many mice.

Specifically for this experiment. Is 10 for each group sufficient? 20? Or do you need 100+?

What's a good average of mice for getting meaningful statistics?

Once again, thanks for the link. Very interesting response.