What if everything we know is wrong - or at least 90 percent of what we think we know.
There was a very interesting topic covered in The Health Report on July 28th concerning the reliability of published research data.
Professor John Ioannidis, Chairman of the Department of Hygiene and Epidemiology University of Ioannina Greece and Tufts University Boston, claims that 90 percent of research findings end up being false. There are a number of reasons given in an article he authored as well as the show transcpript and podcast. I highly recommend both.
His major point appears to boil down to the fact that p > 0.5, while generally recognized as being significant, is susceptible to false positives when huge amounts of data are being trolled for any possible associations. Only 10 percent of these associations pan out following further investigation.
Another way to view the same thing is if 10 groups study an association and only one group finds a significant association, that is the paper which gets published.
