ABSTRACT
In Chapter 4, I discussed the logic of hypothesis testing and some philosophy of science. In particular, I contrasted the philosophy of Karl Popper with that based on the Bayesian approach. Popperians, apparently, spend their lives trying to find experiments that will refute their own theories, gladly giving them up when an odd experiment produces an inconsistent result. If scientists like these exist, I have never met one. Bayesians, on the other hand, build up beliefs in theories or lose faith in them in a gradual manner. They deal in uncertainties. Not only are they uncertain about their theories and hypotheses, holding these with a degree of belief, or subjective probability, but they are also uncertain about their experiments. Experimental findings also hold with some probability, due to uncertainty in methodology and the possibility of error - human or otherwise - in experimental design and data analysis. Scientists that I know are much more like these imaginary Bayesian creatures.
