Nobody. The problem is that if you conduct 20 trials of the efficacy of rabbit feet, you'd expect 1 trial to show a significant effect. (if you're interpreting

Nobody. The problem is that if you conduct 20 trials of the efficacy of rabbit feet, you'd expect 1 trial to show a significant effect. (if you're interpreting P-values the way many people do, incorrectly.)