Case Studies and Theory Development in the Social Sciences - Alexander L. George [16]
MODELING AND ASSESSING COMPLEX CAUSAL RELATIONS
A final advantage of case studies is their ability to accommodate complex causal relations such as equifinality, complex interactions effects, and path dependency.44 This advantage is relative rather than absolute. Case studies can allow for equifinality, but to do so they produce generalizations that are narrower or more contingent. We find great value in such middle-range theories, but others may prefer theories that are more general even if this necessarily means they are more vague or more prone to counterexamples. Case studies also require substantial process-tracing evidence to document complex interactions. Analogously, statistical methods can model several kinds of interactions effects, but only at the cost of requiring a large sample size, and models of nonlinear interactions rapidly become complex and difficult to interpret. New statistical methods may be able to improve upon the statistical treatment of equifinality and interactions effects.45
Trade-offs, Limitations, and Potential Pitfalls of Case Studies
It is important to distinguish among the recurrent trade-offs, inherent limits, and examples of poor implementation of case study methods and not to misinterpret these aspects through the prism of statistical methods, as has been done in the past. Recurrent trade-offs include the problem of case selection; the trade-off between parsimony and richness; and the related tension between achieving high internal validity and good historical explanations of particular cases versus making generalizations that apply to broad populations. The inherent limitations include a relative inability to render judgments on the frequency or representativeness of particular cases and a weak capability for estimating the average “causal effect” of variables for a sample. Potential limitations can include indeterminacy and lack of independence of cases.
CASE SELECTION BIAS
One of the most common critiques of case study methods is that they are particularly prone to versions of “selection bias” that concern statistical researchers.46 Selection biases are indeed a potentially severe problem in case study research, but not in the same ways as in statistical research.
Selection bias, in statistical terminology, “is commonly understood as occurring when some form of selection process in either the design of the study or the real-world phenomena under investigation results in inferences that suffer from systematic error.”47 Such biases can occur when cases or subjects are self-selected or when the researcher unwittingly selects cases that represent a truncated sample along the dependent variable of the relevant population of cases.48 If for some reason a statistical researcher has unwittingly truncated the sample of cases to be studied to include only those whose dependent variable is above or below an extreme value, then an estimate of the regression slope for this truncated sample will be biased toward zero. In other words, in statistical studies selection bias always understates the strength of the relationship between the independent and dependent variables. This is why statistical researchers are admonished not to select cases on the dependent variable.49
In contrast, case study researchers sometimes deliberately choose cases that share a particular outcome. Practitioners and analysts of case study methods have argued that selection on the dependent variable should not be rejected out of hand. Selection