Experiments at Ann Arbor indicate that rat life expectancy increases with a diet of ecstasy. Raffish rats live as long as eight years, the tests show, while puritan rats, on love-free and otherwise grim and grubby diets, succumb at five. The autobiographical parallel is tempting, but though “love” is said to be a good metabolizer, and though bacchanals have been recommended for the circulation, whether my two-and-fourscore years have been sustained by my devotion to both, or whether I am still here in spite of them, is only a lay opinion-and a “philosophy of life.” The facts are rat facts. -Igor Stravinsky
Extrapolation from rat to man, or from game simulations to the marketplace or the summit conference, is done by the student of men and groups of men, not because he prefers to, but because he must. For how can the scientist inflict on Mr. Stravinsky or on U.S. Steel the rigid controls and intense, sustained observation necessary to obtain useful data? On the other hand, how can he be sure that his carefully designed laboratory experiments are valid tests of hypotheses about the natural world? This dilemma is explored in the present paper, and a framework for testing hypotheses is presented.