This essay by Pomeroy reminds us of the value of Stan Young’s work. Why all these epidemiology studies are so unreliable when they trumpet assertions based on small associations.
One of Jim Enstrom’s persecutors, and a prominent junk scientist for the American Cancer Society, is quoted here, admitting that small associations and false positive claims are too common in this world of publication bias. But the problem is that one of Thun’s biggest crusades, on second hand smoke,s is exhibit one for cheating in epidemiology with cancer claims that came from small association results in observational studies.
Troubling to see the hypocricy and “lying for justice.”
I also remind you of John Ioannides’ similar debunking of bad research disussed here at JS.
Archive on epidemiology:
Archive on Ioannides: