The Sociological Imagination by C. Wright Mills

I am not of course attempting to summarize the results of all the work of abstracted empiricists, but only to make clear the general character of their style of work and some of its assumptions. Accredited studies in this style now tend regularly to fall into a more or less standard pattern. In practice the new school usually takes as the basic source of its ‘data' the more or less set interview with a series of individuals selected by a sampling procedure. Their answers are classified and, for convenience, punched on Hollerith cards which are then used to make statistical runs by means of which relations are sought. Undoubtedly this fact, and the consequent ease with which the procedure is learned by any fairly intelligent person, accounts for much of its appeal. The results are normally put in the form of statistical assertions: on the simplest level, these specific results are assertions of proportion; on more complicated levels, the answers to various questions are combined in often elaborate cross-classifications, which are then, in various ways, collapsed to form scales. There are several complicated ways of manipulating such data, but these need not concern us here, for regardless of the degree of complication, they are still manipulations of the sort of material indicated.

Apart from advertising and media research, perhaps ‘public opinion' is the subject-matter of most work in this style, although no idea which re-states the problems of public opinion and communications as a field of intelligible study has been associated with it. The framework of such studies has been the simple classification of questions: who says what to whom in which media and with what results? The going definitions of the key terms are as follows: