ABSTRACT
Most social researchers employ probability concepts in their research either explicitly or implicitly whether or not they are aware of it. The chief notable exceptions are those who are proponents of what has come to be called "analytic induction," a rationale that explicitly excludes certain probability formulations. Much of social research is informed by the strategy of searching for reliable empirical generalizations which, because of their empirical constancy, must be included as propositions within any theory that might later be formed to deal with the phenomena. In this strategy the test of significance is often viewed as a means of determining which generalizations have that compelling empirical property. The chapter distinguishes three major uses of probability in science: in intrinsic statistical hypotheses; in auxiliary statistical hypotheses, including the theory of errors of observation, sampling from finite populations, and randomized experimentation; and thirdly in inductive probability policies, particularly the theory of tests of significance.
