Explore "The Sociological Imagination" by C. Wright Mills on Textopian. Read with annotations and highlights.
The specific methods-as distinct from the philosophy-of empiricism are clearly suitable and convenient for work on many problems, and I do not see how anyone could reasonably object to such use of them. We can of course, by suitable abstraction, be exact about anything. Nothing is inherently immune to measurement.
If the problems upon which one is at work are readily amenable to statistical procedures, one should always try to use them. If, for example, in working out a theory of elites, we need to know the social origins of a group of generals, naturally we try to find out the proportions coming from various social strata. If we need to know the extent to which the real income of white-collar people has gone up or down since 1900, we run a time-series of income by occupation, controlled in terms of some price index. No one, however, need accept such procedures, when generalized, as the only procedure available. Certainly no one need accept this model as a total canon. It is not the only empirical manner.