I read an article quite a while ago on an online science blog (I think it was a fairly popular one, although I’m not 100% sure - I know it wasn’t Orac’s or SBM.com) that outlined clearly and carefully an assortment of reasons why a treament might appear to work when in fact it does nothing. Does anyone know which article I’m talking about? Can anyone help me find it?
Thanks!
(alternatively, you could post reasons here. That would also be pretty neat.)
DR Steven Novella has done several posts on Neurologica Blog about this. I linked to one, but can’t find another that is more what you’re looking for.
I’d recommned Ben Goldacre’s “Bad Science” as a good source of information on this and related topics.
The two obvious ones that come to mind:
There’s Barry Beyerstein’s “Why Bogus Therapies Often Seem To Work” on Quackwatch.
Sometimes the results of testing says something like
“It works, but no better than, and equally as well as, a placebo”.
SO they are saying that a 5 cent sugar pill is just as good 
Another test is “is it better than FOOD ?”
A report like the following does tell you that some foods are at least as good as food.
Or then there are all the things they put into their marketing. I think my favorite is “Contains a clinically-studied active ingredient”.