Friday, June 02, 2017

Professor Logic, please call your office

The study described in this article purports to answer the question of whether red wine helps heart health by controlling for heart health.  So apparently when we control for whether or not our control sample has arteriosclerosis, high blood pressure, myocardial infarction and the like, we find that both moderate drinkers and teetotalers have about the same rates of arteriosclerosis, high blood pressure, myocardial infarction, and the like.

Well, yes, and I could have told you that before you ever selected a sample or hired any statisticians.  Scary thing is that this study made it past peer review, and even more scary, it's probably someone's Ph.D. dissertation that may land them in a professor's office.

No comments: