A constant goal of sociologists and other social science researchers is to determine objective, or relatively objective, ways of measuring something that is inherently very far from "objective," namely human (individual or social) attitudes and actions. One such technique is content analysis, or the use of statistical methods to analyse the contents of documents usually written documents of one sort or another, though transcripts of speech, television shows, and other non paper "documents" can also be subjected to content analysis. In the simplest instance, content analysis may be performed by counting and tabulating the occurence of particular words in a document. If, for example, a given document has frequent references to "God," "Christ," and "salvation," as compared to other documents, we might reasonably infer that the document is somehow concerned with Christian religion. If it also has frequent references to "Virgin Mary," this would be some evidence of a Catholic orientation; references to "born again," in contrast, would indicate an evangelical Protestant orientation. Such an analysis does not tell us everything the document may in fact be a passionate attack on the beliefs associated with these words, but it is strong evidence that the document's writer or intended readers were concerned with these beliefs.
As another illustration of content analysis, we may actually perform a simple content analysis, taken here from Bailey (1987: 300 302). Here are three definitions of content analysis from standard works.
Content analysis is a research technique for the
objective, systematic, and quantitative description
of the manifest content of communication.
Content analysis is any research technique for making
references by systematically and objectively identifying
specified characteristics within text.
In the subsequent discussion, we propose to use...