Case Studies and Theory Development in the Social Sciences - Alexander L. George [94]
This alternative approach focuses not on the analysis of variables across cases, but on the causal path in a single case. Within-case analysis may be used also in conjunction with studies making cross-case comparisons or for the development of typological theories. Indeed, our position is that within-case analysis is essential to such studies and can significantly ameliorate the limitations of Mill’s methods.
Chapter 9 considers the congruence procedure, which can be employed either in a single case study or for each case in a comparative study. Unlike the greater flexibility of the controlled comparison method, it requires a theory that predicts outcomes on the basis of specific initial conditions. Depending on how developed a theory is, its predictions may be abundant and precise, or scarce and highly general. Working with the preexisting theory, the researcher establishes the value of independent and dependent variables in the case at hand, and then compares the observed value of the dependent variable with that predicted by the theory, given the observed independent variables. If the outcome of the dependent variable is consistent with the theory’s prediction, then the possibility of a causal relationship is strengthened.
The congruence and process-tracing methods for making causal inferences provide alternatives to controlled comparison, and therefore constitute the basis for a different type of comparative method. The results of individual case studies, each of which employs within-case analysis, can be compared by drawing them together within a common theoretical framework without having to find two or more cases that are similar in every respect but one. The process-tracing method is discussed in detail in Chapter 10.
Chapter 9
The Congruence Method
The congruence method occupies a special place in our conception of how a single case or a small number of cases can be used for theory development. As we noted in Chapter 8, the method of controlled comparison requires the investigator to find two cases similar in every respect but one. Since this requirement is difficult to meet, an alternative approach is often needed—one that does not attempt, as a controlled comparison does, to achieve the functional equivalent of an experiment. The alternative we propose is the within-case method of causal interpretation, which may include congruence, process-tracing, or both, and which does not operate according to the structure or causal logic of experiments. This chapter discusses the congruence method, and we turn to process-tracing in Chapter 10.
The essential characteristic of the congruence method is that the investigator begins with a theory and then attempts to assess its ability to explain or predict the outcome in a particular case. The theory posits a relation between variance in the independent variable and variance in the dependent variable; it can be deductive or take the form of an empirical generalization. The analyst first ascertains the value of the independent variable in the case at hand and then asks what prediction or expectation about the outcome of the dependent variable should follow from the theory. If the outcome of the case is consistent with the theory’s prediction, the analyst can entertain the possibility that a causal relationship may exist. Of course, the finding of mere consistency between a theory’s predictions and case outcomes may not be significant, and in this chapter we discuss several questions that can guide researchers as they assess the significance of preliminary findings.376
The congruence method has several attractive features. The investigator does not have to trace the causal process that leads from the independent variable to the case outcome; so the method does not require a great deal of data about the case being studied. Because the congruence method does not use process-tracing, it does not require a search