I believe you meant p < 0.05.
Thanks,
@the_doc. I should check for typos before posting, but often forget!
Anyway, the notion that 1 in 20 studies with p = 0.05 will be wrong is spooky enough; however, I recently watched an even spookier video about the misleading use of the p-statistic. I'll look it up and post the link when I get a chance.
I appreciate you posting that video. I'd seen it but had forgotten it. It makes excellent points, and in reality, there are other biases that come into play, with studies being done but not published if the data don't come up with a sexy conclusion.
I was on the Statistics technical network for a Fortune 100 corporation, and I still feel incompetent at judging stats without experts reviewing my work. A week from Tuesday, I'm going to a meeting regarding issuance of a state document based on statistical evaluation that I think requires more professional input. The fact is, too often, stats are misunderstood and misused. The point about statistical power is very important!
Thanks again for making these points. It's a good sign that this was retracted, indicating the self-policing was successful. It doesn't put any hole in evolutionary theory.