ABSTRACT
So far, all we have said about an experiment is that it should be efficient and well thought out before execution. Since most people will buy this concept, where does statistical experimental design have any advantage over the basic, “normal” way experimenters have been doing business for centuries? Let’s look at a long established method that is recommended in physics books (1) and is usually called the “classical, one-factor-at-a-time” (1-F.A.A.T.) technique. We will investigate a 1-F.A.A.T. to see if it does get the required information with the least expenditure of resources.
