Are meta-analyses that just average all the results from all the studies less reliable than picking a paper at random? I'd expect that to be the case if the bo

Are meta-analyses that just average all the results from all the studies less reliable than picking a paper at random? I'd expect that to be the case if the bogus studies report the most dramatic results. A single bogus study out of 10 can skew the whole meta-analysis, but if you picked randomly you'd be 90% likely to get an honest one.