Round Table On The Case-Study Method Of Sociological Research
Stuart A. Queen
University of Kansas
The round-table session on "The Case-Study Method of Sociological Research" consisted in a discussion of eight questions which had been mimeographed and distributed beforehand. Copies were also in the hands of all persons attending the session. There were no prepared statements, the whole discussion being informal and impromptu.
The questions on which the discussion was based were these: What is the case-study method of sociological research? To what types of sociological problems is it applicable? What kinds of data may it be expected to yield? What methods may be employed in assembling data? How may the reliability of data be tested? What concepts are most helpful for organizing and interpreting data from case studies? How can 'generalizations be derived from such materials? What kinds of persons may legitimately employ the case-study method of sociological research?
Inasmuch as no minutes were kept nor was there any attempt to bring matters to a vote, it is impossible to say what were the results of this session. Each participant doubtless took away a different set of conclusions. What follows represents, therefore, the impressions of the chairman and might well be supplemented by similar statements from others who took part in the discussion.
The case-study method of sociological research is to be distinguished from the practice of social case work on the one hand and from the statistical method of research on the other. It is intimately related to both, but different
(226) from either. Social case work has as its objective direct service to particular persons. Sociological research has no reference to any particular persons or types of service. It is a search for general principles. Now while the statistical technique depends upon the reduction of data to quantitative terms, in order to yield totals, averages, and correlations, the case study technique seeks data in terms of processes, which for the most part cannot be stated numerically. The statistician selects certain specific factors involved in social situations and manipulates them so as to discover the relations between the several variables. The "case student" examines single situations, persons, groups, or institutions as complex wholes in order to identify types and processes. Now the two techniques are not incompatible; indeed, both may be employed in the same project. Preliminary statistical studies may guide the investigator in the selection of cases for detailed examination and may suggest factors worthy of special attention. Also some of the results of case studies may be summarized in statistical form.
For the most part the data from case studies appear in the form of "running accounts," narratives of events, and descriptions of personalities and situations. However, it is often convenient to present particular factors or aspects of the case on charts or schedules. But questionnaires and "tests" are not essential, nor do their results alone constitute case studies. Most of the data, and perhaps the most reliable data, will come from rather informal interviews. Other important sources will be records of social agencies, autobiographies, diaries, and letters. This means that the facts will appear in varied forms and sequences. It is here that the case-study method seems to differ most sharply from the statistical. The latter involves the preparation of schedules with fixed categories into which the data must fall. The former develops its categories inductively from the data as they reveal themselves more or less informally.
The reliability of such data as may be secured through interviews and personal documents is, of course, always open to question. Fortunately it is often possible to verify statements from independent sources, and sometimes the nature of the statement itself is such as to indicate the credibility of the narrator. Moreover, it is important to remember that even lies, legends, and delusions are facts when properly classified.
One of the most difficult problems connected with the use of the case method of research is, How can generalizations be derived from such complex bodies of data? Statistics may be used, but the major portion of the data cannot be reduced to statistical form. Moreover, the identification of types of personalities, groups, institutions, and situations seems to require only the most elementary mathematics, though it does demand very careful classification. Also, the discovery of processes appears to require relatively little in the way of a quantitative procedure. Perhaps its simplest statement will be in terms of typical sequences of events or situations, but this involves no recognition of the basic assumption of continuity and what may be called social
(227) "dynamics." Hence we are faced with a very real problem of bow to translate what looks like "another moving little tale" into data for sociological generalization.
Who is competent to employ such a method of research? On the one hand it may be held that prying into the intimate affairs of men should be restricted to those who have both discretion and skill. The very making of a study may complicate a serious situation. The possible reaction against indiscriminate delving into human relations may later handicap more competent researchers. On the other hand it is held that there is nothing esoteric about the making of case studies. Anyone may bring in data which may prove to have real value. The ordinary student is not likely to do much damage in securing such facts as are accessible to him at all. Certainly the majority of social workers are not in a position to do research work; they are too busy; they are primarily responsible for serving their clients rather than treating them as specimens; frequently they lack training in the social sciences. Yet they have direct and legitimate access to the most intimate facts about human conduct, facts which are essential to the researcher. It may be that the best results can be secured through the collaboration of social worker and sociologist.