A Case for the Case Study

Joe R. Feagin, Anthony M. Orum, Gideon Sjoberg, Joe R. Feagin, Anthony M. Orum, Gideon Sjoberg

• Publisher: The University of North Carolina Press
• Year: 2016
• Print ISBN: 9780807843215, 9780807819739
• eBook ISBN: 9781469621401
• Topic: Social Sciences
• Subtopic: Sociology
• Index: Social Sciences

Since the end of World War II, social science research has become increasingly quantitative in nature. A Case for the Case Study provides a rationale for an alternative to quantitative research: the close investigation of single instances of social phenomena.

The first section of the book contains an overview of the central methodological issues involved in the use of the case study method. Then, well-known scholars describe how they undertook case study research in order to understand changes in church involvement, city life, gender roles, white-collar crimes, family structure, homelessness, and other types of social experience. Each contributor confronts several key questions: What does the case study tell us that other approaches cannot? To what extent can one generalize from the study of a single case or of a highly limited set of cases? Does case study work provide the basis for postulating broad principles of social structure and behavior? The answers vary, but the consensus is that the opportunity to examine certain kinds of social phenomena in depth enables social scientists to advance greatly our empirical understanding of social life.

The contributors are Leon Anderson, Howard M. Bahr, Theodore Caplow, Joe R. Feagin, Gilbert Geis, Gerald Handel, Anthonly M. Orum, Andree F. Sjoberg, Gideon Sjoberg, David A. Snow, Ted R. Vaughan, R. Stephen Warner, Christine L. Williams, and Norma Williams.

1 The Case Study Approach in Social Research

Basic Methodological Issues

Gideon Sjoberg, Norma Williams, Ted R. Vaughan, and Andrée F. Sjoberg

Our primary goal is to restore the role of the case study as a major methodological tool in social science inquiry—both as a supplement to the natural science model and as a distinctive means of providing valid social knowledge. At one time the case study was a taken-for-granted mode of carrying out research. However, in recent decades it has been pushed aside, or at least sharply downgraded, in sociology (as well as in other social sciences). Thus the case study method receives only limited attention in current textbooks on social research (e.g., Babbie 1989; Bailey 1987; Phillips 1985; Dooley 1990; Denzin 1989; Schwartz and Jacobs 1979). Both qualitative and quantitative researchers tend to shun the basic issues involved in the use of the case study method in social inquiry. Yet exceptions to this generalization can be cited (cf. Yin 1984; Whyte 1984). Perhaps the major monograph addressing the case approach in recent years is that by Ragin (1987), and the subject receives attention in the edited book by Kohn (1989). Still, our approach differs from that of other social scientists who deal with this problem area.

Although our analysis is by no means inclusive, we nonetheless address the salient aspects of the use of the case study in social research. We not only delineate fundamental methodological issues but also outline the manner in which these can be resolved. Our emphasis is on the use, rather than the construction, of case studies (though the two merge in actual research practice).

Our central thesis is that the case study not only serves as a strategic supplement to the natural science model but is an essential feature of sociological inquiry in its own right. We formulate our rather complex argument in the following manner. We first set forth our basic orientation toward methodological inquiry. Here we explicate the premises underlying the natural science model as well as our alternative formulation. This sets the stage for considering the concept of the case study, a subject of much controversy. We then discuss some of the historical issues involved in the use of the case study method, in the process spelling out how the case study approach has been crucial to the growth of grand theory and the construction of classic works in the social sciences. Our reasoning leads to the query: Why has the case approach fallen into disfavor in recent decades? By answering this question, we can better examine the dominant natural science model of inquiry and detail how and why its adherents must rely, in a variety of ways, on the case study approach to supplement their quantitative research procedures. Most important still, we contend that the case study method is essential if social science is to grapple with major social issues on both the historical and the contemporary scenes. Such matters lie beyond the grasp of the natural science model. We conclude by discussing our general orientation toward those methodological issues that demand careful exploration if the case study method is to advance effectively in social science. Throughout, our analysis rests on a broad cross-cultural approach toward the use of the case study method in social scientific inquiry.

Ours is an alternative to the natural science model on the one hand and the historicist approach on the other. As we elaborate below, the natural science model, in its present form, seeks to establish “universal laws” such as those developed in certain natural sciences. The historicist approach, in its extreme form, denies that social science can establish cross-cultural generalizations, contending instead that each sociocultural order contains its own unique patterns of development. We align ourselves with scholars who are committed to the advancement of social science not by emulating natural science but by formulating a methodology for social inquiry that has an integrity of its own. In practice we reject the basic presuppositions of the natural science model but realize that particular logical procedures can be adapted for our purposes.

The Nature of Methodology

To understand the issues that have swirled about the case study approach, we must place our analysis within a larger methodological context. Nowadays methodology is frequently equated with the resolution of technical problems, especially in the quantitative arena. This is reflected in the items that appear in Sociological Methodology, a publication of the American Sociological Association.

Our use of methodology harks back to a somewhat earlier era. It does not focus on the refinement of particular research procedures but, rather, involves the analysis of the intersection (and interaction) between theory and research methods and data (e.g., Galtung 1967; Gibbs 1972; Sjoberg and Nett 1968). In recent years, theorists have gone their own way and have paid scant attention to research procedures and data analysis (see such texts as Turner 1986 and Wallace and Wolf 1986); on the other hand, specialists on research methods typically set theory aside in favor of refining, codifying, and standardizing the techniques of data collection and analysis. This division of labor inhibits a constructive examination of the issues involved in carrying out social research. The case study, with both its strengths and its weaknesses, must be considered within a broad rather than a narrow framework.

Adopting a more general conception of methodology, we find that mainstream sociology—with its many shifting currents—has since World War II been dominated by the natural science model (e.g., Sewell 1987; Wilner 1985). This mode of inquiry in sociology resulted from the confluence of a variety of intellectual and social conditions. Although adherents of the natural science model often differ among themselves, they nonetheless share (as we detail below) certain basic tenets.

A major intellectual influence in shaping the natural science model in sociology has been the philosophers (originally mathematicians and natural scientists) identified with the Vienna Circle (e.g., Ayer 1959; Polkinghorne 1983). This school of thought, often termed logical positivism or logical empiricism, held sway in philosophy departments in the 1950s and onward (Rorty 1982, chapter 12) and structured intellectual discourse, not just in philosophy but in the social sciences more generally.

The Vienna Circle has not been alone in fostering the natural science model. Another source of this intellectual heritage has been the British empiricist tradition. One wing can be traced to John Stuart Mill and his analysis of the experimental design. Interwoven with (but at times separate from) the experimental tradition has been the statistical approach, which had its forerunners in the likes of Pearson, Galton, Fisher, and others. (A statistical tradition also developed in continental Europe, but this has not been as influential as the British heritage.)

An important spin-off of the Vienna Circle was the proponents of the logico-deductive mode of inquiry (e.g., Hempel 1965). Their works gave rise to a widespread championing of the logico-deductive method by such sociologists as Blalock (1969), Gibbs (1972), Blau (1977), Homans (1982), and Turner (1986). Even Merton (1957, 85–101), in an early essay, identified himself with this logical mode of analysis (although it exhibits little, if any, relationship to the functionalist analysis he has championed).

That social theory is to be equated with the logico-deductive format is expressed by, for instance, Homans:

A theory of a phenomenon is an explanation of the phenomenon. …

An explanation consists of a set of at least three propositions. … To be allowed to take part in an explanation, a proposition must meet at least two conditions. It must state a relationship between at least two variables and it must begin to state what that relationship is. … It must say, for instance, that if x is present, y is also present, or that if x increases in numerical value, y also increases. (1982, 285)

One of the propositions in the set (usually of a low order of generality), which Homans termed an empirical proposition, is that to be explained. Of the other two propositions, one states the condition (or conditions) under which the general proposition can be applied.

Quantitative data are used to test propositions. In some instances these data are derived from surveys or other kinds of “official statistics” (e.g., data on the money supply); in other instances they are the products of social experiments.

Although many advocates of the natural science model are committed to the logico-deductive format, others utilize what can loosely be described as the inductive approach. Skinner (1982), the behavioral psychologist on whom Homans drew for his assumptions about human nature, was a proponent of this orientation. In sociology a host of experimentalists and, especially, survey researchers—with their elaborate statistical procedures such as path analysis and regression analysis—utilize inductive logic.

To enhance our understanding of this model, we counterpose it to a version of the sociology-of-knowledge approach toward methodology (cf. Sjoberg and Nett 1968).¹ This latter orientation recognizes that social research is first and foremost a social enterprise. We view theory as a social process constructed by theorists and research as a social process carried out by researchers, both processes taking place within a social context.

For us, social theory—which empirically reflects what sociologists do and not what they say—involves a process of interpreting (and constructing) social reality (see, e.g., Vaughan and Sjoberg 1986). It has three basic aspects or components: domain assumptions, logical forms, and assessment procedures.

Gouldner (1970), for all the criticisms leveled against his work The Coming Crisis of Western Sociology, has nonetheless made common currency of the notion of “domain assumptions”; this in turn is congruent with Alexander’s (1982) concept of “presuppositions.” If we examine sociological theory and the research, we find that most of the major debates revolve around the question of the proper assumptions to be adopted by researchers. Some of these are highly abstract; others can be debated with respect to their empirical plausibility (though they cannot be tested empirically in any strict sense).

Although the “idealist-materialist” controversy involves abstract notions about the nature of human nature and of social reality, other less abstract assumptions shape the course of the research process (at times in concrete ways). We phrase these in the form of questions (though still others could be added): (1) What is the proper unit of analysis (a central issue for the case study method)? (2) Is the researcher committed to the study of the unique or of the general or of some variation thereof? (3) Are human nature and social reality consistent or are they rent by inherent tensions or contradictions? (4) Are appearances to be taken as given or is there a reality underlying appearances? (5) Are human nature and social reality fixed and well ordered or are they processual—that is, in a state of becoming? (6) What is the nature of rationality? (7) What is the relationship between the social researcher and his or her subject matter? This last query gives rise to subquestions such as: Should one take a top-down or a bottom-up perspective in carrying out research?

As indicated above, the proper unit of analysis is a major issue. Debates about this matter are typically combined, in a variety of ways, with disagreements over other assumptions. Tilly (1984), after taking note of various possible units of analysis, considered the study of the unique versus the general. He then stated: “A purely individualizing comparison treats each case as unique. … A pure universalizing comparison … identifies common properties among all instances of a phenomenon” (81). Among the scholars in the former category is the sociologist Bendix. Tilly contrasts him with researchers such as Gerhard Lenski and Barrington Moore, who are concerned with universalizing comparisons. Social scientists oriented toward the natural sciences model are committed to isolating universals—and so are certain social scientists (like ourselves) who do not align themselves with that mode of inquiry. Nevertheless, we would be remiss if we did not recognize that investigating the unique is a time-honored aspect of one facet of social research.

In ideal type terms, the social scientist’s approach to the unique necessarily involves discovery of the categories developed by the system in question; social scientists then use these theoretical categories to elaborate on, and thereby provide a broader understanding of, the system (e.g., the individual or larger unit) under study. This abstract argument comes into focus when we consider social scientists’ efforts to understand the Soviet Union before the Gorbachev “revolution.” A source of running debate was the question of whether one could apply Western concepts—for example, “political pluralism,” as formulated by U.S. political scientists—to the analysis of the Soviet system (e.g., Solomon 1983). In seeking to interpret the “command economy” (that managed from above by the governmental bureaucratic apparatus), some social scientists have contended that concepts associated with the free market are inappropriate and that one must work within the framework defined by the command economy (e.g., Birman 1989,144—45). Moreover, we strongly suspect that the facile effort to explain significant changes in Eastern Europe and the Soviet Union in 1989 in free-market terms will, in retrospect, be viewed with disfavor, for this orientation likely reflects the ethnocentrism of Western scholars and especially politicians and newspaper reporters. For one thing, disagreement exists among scholars in Western Europe and the United States over the nature of the market, especially one dominated by large-scale bureaucratic organizations. Nowadays the contemporary “market mechanism” bears little relationship to the neoclassical economists’ formulation of this concept.

The unit of analysis is linked not only with relativism versus nonrelativism (i.e., the unique versus the general) but also with other domain assumptions. Whether one takes appearances as real or looks behind them for the basic reality has been a chief source of controversy in social science. Marx and Freud took the lead in stripping the “officially defined” reality of its veneer and insisted that the basic reality lies behind appearances. Freud, in his early work on the nature of human nature, thus stressed the centrality of the “unconscious.”

Although we shall not explicitly examine each of these domain assumptions as they relate to the case study approach, the interplay of domain assumptions and case studies emerges in various guises in the succeeding analysis. We must recognize that the choice of assumptions can and does shape research in concrete ways. Freud, for example, having postulated that the basic psychological reality is the unconscious, then proceeded to devise methods—such as the analysis of dreams and linguistic slips—that would permit him to understand the essential nature of human nature, which he approached primarily through case studies that he investigated either directly or indirectly by drawing on secondary sources. (We indicate below how the advocates of the natural science model have forged a link between their domain assumptions and their research procedures.)

In addition to domain assumptions, another facet of social theory is the particular logical forms a researcher employs. The natural science model has been associated with the logico-dedu...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Contents
• Preface
• Introduction: The Nature of the Case Study
• 1 The Case Study Approach in Social Research: Basic Methodological Issues
• 2 Middletown As an Urban Case Study
• 3 A Tale of Two Cases
• 4 Researching the Homeless: The Characteristic Features and Virtues of the Case Study
• 5 Oenology: The Making of New Wine
• 6 The Case Study Method in Sociological Criminology
• 7 Case Studies and the Sociology of Gender
• 8 Case Study in Family Research
• Conclusion: The Present Crisis in U.S. Sociology
• Contributors
• Index