Following the methodology of dialectology, this study was designed to investigate certain linguistic features of careful speech in a limited number of relatively homogeneous counties, each represented by an informant belonging to a certain age and social class, i.e., a 1975 high school graduate entering the job world immediately after graduation.

The research procedure consisted of five major steps: (1) the analysis of language textbooks on the current state-adopted list, (2) a survey of teacher attitudes toward items of usage garnered from the textbook analysis, (3) interviews with five selected informants, (4) coding of the transcripts of the interviews in relation to the items on the postal survey and (5) analysis of the findings.

The pretext of the interview with the five informants was the relevance of high school education to career preparation. The interview was conducted in a setting simulating a formal job interview so that the informant would employ careful or controlled usage suitable to the interview situation. The tapes of the interviews were transcribed, and the transcripts were analyzed and coded in relation to items of usage garnered from the textbook analysis.

A detailed description of the linguistic characteristics of each of the five informants served as the basis for a general description of the predominant linguistic characteristics of the careful speech of the five informants. In this study, the predominant nonstandard usages of careful speech were the introductory word-singular-verb-plural subject pattern as in "There is too many facts"; nonstandard variants of common irregular verbs such as "I seen"; and the use of the pleonastic subject such as "My father he died."

The findings appeared to indicate that the informants classified as U, upward, in social mobility exercised strong conscious control of their linguistic behavior and that the speech of the upwardly mobile informants was more stilted and confined than that of the other informants. The researcher concluded that many linguistic features that serve as social markers are, in varying degrees, part of the linguistic behavior of speakers on all social levels; however, the difference appears to be quantitative rather than qualitative.