Wishcasting

In my sadly misspent youth, I worked for a couple of companies that did "contract research" for the big pharmaceutical companies. I wasn't involved in the clinical trials, just the statistical analysis, cranking the data through SAS and plotting points for the statisticians and tech writers. All pharmaceutical research in the U.S. has to be double-blind, meaning that neither the subjects or the doctors know which participants are getting which drugs. This eliminates both the placebo effect and the experimenter effect.

William M. Briggs, a statistician, ties Wishcasting, the experimenter effect, pharmaceutical research, global warming, and politics together. It is a brief and lucid explanation and well worth reading.