Seriously? I only skimmed the link. I’m surprised they got away with that. Please accept my apologies.
When animals are mating they are doing science. I know it is science because there are scientists studying it.
Sunlight is art. I know it because photographers study natural lighting.
Did you just quote someone’s blog?
Probably he’s a scientist too. Because I know that Internet is built using science.
Chefs are scientists. I know they are because there are studies on how food (matter) reacts to temperature and other mechanical and chemical processes.
I just found this, looks like now they’re starting to uncover the mechanisms behind how acupuncture works. It’s not yin yang, or chi, or whatever, but I never claimed it is.
Yoga is science. There was this study of a neurologist who plugged these cables to a guy doing yoga and there was something on a computer. Yogins are scientists.
OK, let’s take a look at what this guy is saying:
I know, people will get all upset. They’ll throw out anecdotes, like people do with chiropractic, or [cherry pick] poorly designed studies to “prove” that acupuncture works
Yeah, I didn’t cherry pick any studies. I linked a metastudy, a study of studies.
The number of studies and the quality of those studies are important.
True. And I tend to trust Harvard more than these blogs.
It appears that the links mad masala provided all quote one person, Steve Novella.
He argued that the difference in the results between acupuncture and the sham needle was so small that it makes no difference (even though there was a statistically significant difference).
That sounds fishy to me, because for pharmaceuticals to be approved, they only need to work 5% more effectively than placebo. That’s because placebos can be really strong.
Full disclosure: I’m a statistician, not a doctor.
What are the chances of that metastudy to be right, while most studies and the scientific community seem to say the contrary?
That’s incorrect regarding the development of new pharmaceutical drugs. If it has a p.value of <0.05 that only indicates there is a significant difference between two groups and that MIGHT warrant further investigations, by itself it proves nothing except there is a difference between two groups NOT what caused the difference… Further investigations by pharmaceutical companies are complex and usually cost multi millions in further tissues culture, animal studies and human studies.
Chinese medicine on the other hand is not required to do the same stringent tests before they put their medicines and potions onto the market
You got pvalue and the difference in means mixed up.
When something is 5% more effective, that means the average impact of the treatment is 5% greater than the placebo. Says nothing about the pvalue.
You want to the pvalue to be lower and the difference in means to be higher.
The pvalue usually has to be 5% or lower, although I think they lowered it to 1% if you’re going to overturn existing findings.
OMG what ?
A p value of less than 5% indicates that the difference between groups could only happen 5 times out of 100 due to chance alone. Its nothing about being 5% more effective. It indicates there is a significant difference that could warrant further investigation nothing to do with something being 5% more effective. P values in social science are usually set much higher than the hard sciences.