Posted: Aug 28, 2013 6:20 pm
by Transilvanian
Parapsyologists are trying to proove there claims with super-signifficant meta-analyses like this: http://forum.szkeptikus.hu/download/fil ... &mode=view

4-choice
709 from 2124(33,4%)
P=2,26ยท10-18

Are m-a usually credible? Do you know cases where m-a had shown results like this, but later it was prooved that is false?

Just an example. A big meta-a. like this(there were 2124 trials from 56 studies) with similar good results from something what later was totally disprooved? From medicine or somewhere else? Thank you VERY MUCH.