On his fourth point, Flyvbjerg contends that this criticism is true for all methods of social research, and thus does not apply specifically to case study. This is an argument that I have a lot of sympathy for; all too often, the motivation for undertaking research is to prove that the researcher’s hypothesis, opinion or gut feeling is correct.
On his fifth point, which seems to overlap with his second, Flyvbjerg argues that:
The problems in summarizing case studies, however, are due more often to the properties of the reality studied than to the case study as a research method. Often it is not desirable to summarize and generalize case studies. Good studies should be read as narratives in their entirety. (2004, p. 402)
This, to me, seems to be asking a lot, both of one’s readers and one’s critics. One has to doubt, therefore, if Flyvbjerg and his (real or imagined) critics could ever be reconciled; they would appear to be taking widely diverging perspectives on what is of most important and significance in social research.
Concern over the status of case study has been the stimulus for a series of conferences and edited texts during the last three decades (e.g. Feagin, Orum and Sjoberg 1991; Gomm, Mammersily and Foster 2000a; Ragin and Becker 1992). This move to not just defending, but arguing for, case study as a useful and important research design has also been reflected in its recent growth in popularity among social researchers (David 2006).
Feagin, Orum and Sjoberg (1991) start from the premise that ‘social scientists need to make much more use of the case study approach to studying social life’ (p. vii). They then outline how issues of reliability and validity can be handled (by conducting team research and triangulating data sources), and assert the strengths of case study in providing a close reading set within its context.
Ragin (1992a) opens a second edited text – in an examination of ‘what is a case?’ – by arguing that the response to criticisms of case study is implicit in its practice:
Implicit in most social scientific notions of case analysis is the idea that the objects of investigation are similar enough and separate enough to permit treating them as comparable instances of the same general phenomenon… Social scientists who conduct case studies argue that their cases are typical or exemplary or extreme or theoretically decisive in some other way. Thus, even in case-study research the principle of repetition is often implicated in statements concerning the relation between the chosen case and other cases. (pp. 1–2)
In other words, as Flyvbjerg (2004) also argues, and as most social researchers making use of case study recognise, it is the selection of the case or cases to study that is paramount. Even if the case chosen for study may not always be exemplary, the reasons for its selection must be explained and justified in order for the completed case study to be of interest and use to other researchers (the practicalities of this are discussed in the section on Sampling and Selection Issues in Chapter 8).
Gomm, Hammersley and Foster (2000b) argue that ‘while some case study research may be able to avoid ‘the problem of generalization’ because the case(s) studied have sufficient intrinsic relevance, this is not true of most of it’ (p. 102, emphasis in original). They then articulate effective strategies for drawing general conclusions, involving theoretical inference and empirical generalisation, either to a larger population or within cases.
On the issue of theory, Hammersley, Gomm and Foster (2000) note that two main rationales are advanced by case study practitioners: that theoretical/relations may be perceived directly within the case (cf. Mitchell, 1983, argument), or through the application of some form of comparative method; with the former, in their view, necessarily relying upon the latter. They conclude that a number of problems remain:
Some of these are practical in character, for example the fact that effective use of comparative analysis probably requires the investigation of a relatively large number of cases. Others are more fundamental, for example to do with whether we can reasonably assume deterministic laws of human behaviour. (p. 252)
But they seem confident that these problems may be resolved with some more work. Previously, Hammersley (1992) had offered another possible way forward for case study, in suggesting that:
‘case study’ be defined as one case selection strategy among others; the others being experiment and survey… Compared with the survey, there is a trade-off between generalisability of findings to finite populations on the one hand and detail and accuracy of information on the other. Compared with the experiment, case study involves a trade-off between control of variables and level of reactivity. (pp. 184, 196)
However, while this appears to be an admirably clear and reasonable – and arguably undeniable – rationalisation, it doesn’t seem to take the debate any further forward. In a somewhat similar vein, Verschuren (2003) argues that single case studies should be downgraded and termed ‘case research’; however, simple changes of label, which will not, in any event, be universally agreed and acted upon, do not really address the underlying issues.
George and Bennett (2005) offer a robust defence and elucidation of the case study method from the perspective of political science. They argue that:
Case studies are generally strong precisely where statistical methods and formal models are weak. We identify four strong advantages of case study methods that make them valuable in testing hypotheses and particularly useful for theory development: their potential for achieving high conceptual validity; their strong procedures for fostering new hypotheses; their value as a useful means to closely examine the hypothesized role of causal mechanisms in the context of individual cases; and their capacity for addressing causal complexity. (p. 19)
In a balanced way, they do then identify trade-offs, limitations and potential pitfalls:
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 indeterminancy and lack of independence of cases. (p. 22)
In other words, the potential trade-offs, limitations and pitfalls appear to be as significant as the strengths of case study. The same, however, could be said of other research designs; and, as researchers, we have to make the best of the designs we have.
The debate over case study as a research design continues, of course, and you will find detailed reprises of, and responses to, it throughout this book.
Strengths and Desirable Qualities of Case Study
These debates enable us to construct a shortlist of the strengths and weaknesses of case study as a research design (see Box 3.2). To strengths and weaknesses, I have added a third shortlist of desirable qualities: that is, qualities which case study research – like other forms of research – should strive to achieve, but may in practice fall short of. In this section we will discuss the strengths and desirable qualities, while the following section will focus – at somewhat greater length – on both the perceived weaknesses and the responses that have been made to them.
Box 3.2 Strengths, Weaknesses and Desirable Qualities of Case Studies
Strengths
in-depth/detailed/particular
holistic
typical/exemplary
critical/extreme
feasible/bounded