Cognitive Decision Theory

Abstract

That scientists make decisions no one doubts. They allocate research funds, in effect deciding which problems are most worth attacking, they decide how large a sample to observe, which experimental methods to use, and even which outlying observations to reject or discount. These are all garden-variety decisions of a well understood sort, even though some of the utilities that enter might be deemed ‘epistemic’. It is thought, however, that other decisions lurk back of all these which are far more germane to the scientific enterprise. I mean, of course, decisions to accept or reject hypotheses. These decisions are held by Levi (1967) to be purely cognitive in the sense that all the pertinent utilities are epistemic. Levi’s program is to develop an appropriate cognitive decision theory which will enable scientists to compute and compare the expected epistemic utilities associated with the acceptance of different hypotheses.