First, on a recent case of forged credentials and misrepresented data at Duke University.
Secondly, a study of the placebo effect that seems to indicate that anti-depressant medication may be nothing more than a very expensive investment in the placebo effect.
- The main piece about looking at all of the research over the years and finding a stunning lack of significance.
- The journalist reflects on the story.
As a statistician, the ability to get a drug approved if two studies show an effect and ten studies show none is ridiculous. This is not good scientifically or statistically.