ABSTRACT
In any test of significance there is a testing of some hypothesis about a universe from which the set of observations may be considered a random sample. At present the sociologist must face the fact that the postulated, hypothetical, infinite universe of possibilities, concerning which he tests hypotheses to establish the “significance” of his results, is merely a logical structure, for which he can offer no real counterpart in his research situation. Since the construct of a hypothetical universe of possibilities has proved useful and fruitful in other fields of research, this chapter looks at its application in one of them, that of agricultural experimentation. Generalizing to a hypothetical universe provides the sociologist with hypothetical “universals,” which are not “universal” in an absolutist sense, however. They differ from the universals similarly arrived at by chemists, physicists, and others in a greater degree of complexity in the specification of the “similar conditions”.
