Empirical Tests are only
Auxiliary Devices
Abstract
We test a theory empirically by deducing from it a 'prediction' and comparing it with observation. In this paper, I argue tests are but auxiliary devices. Though definitely useful, they are dispensable. The reason for this is, there is a more important method of evaluating hypotheses, the method of progressive evaluation (MPE). To explain this point, I use a simple example from cryptanalysis. This example provides us with a model of an empirical investigation in compact form, making it much easier to see the rationality of the method used.
Full text available from author by email.
