ABSTRACT
Social science theorems were derived largely unencumbered by statistical data. Probably the greater part of social science theorizing is kept apart from any attempts of adducing statistical evidence. Whether, in view of all the threats, to draw or not to draw an inference is a practical question transcending the statistical data proper. It is a commonplace to point this out, but one should realize that a theoretical inference in a piece of social science research is much less of a risk than making a wrong decision, for instance, in a vaccine test. It has become fashionable to chide social scientists for drawing inferences from data which, on analysis, prove to be not significant by customary statistical standards. In the social sciences no greater need than the development of theoretical insights guided by empirical data. By comparison, the social sciences are at a stage where for decades to come the formation of even tentative theoretical structures will be at a premium.
