As has already been stressed, when considering definitions of case study, some authors view case study as wholly or primarily qualitative in nature (and the classifications by Punch, 2005, of ‘approaches to qualitative research design’, and Burns, 2000, of ‘qualitative methods’, summarised in Box 3.1, provide examples of this tendency). When looking at the origins and history of case study in Chapter 2, it also appeared that there had been much debate regarding the relationship between case study and statistical (i.e. quantitative) methods. It is important, therefore, to be clear about the position of case study with respect to the qualitative/quantitative debate.
As you are probably aware, much energy and paper has been expended on arguing the relative merits of qualitative and quantitative forms of research, and a great deal of this has not been particularly productive. Often, opponents in the debate have employed stereotypical or misleading representations of each other’s assumptions and practices so as to be better able to critique and knock them down. Indeed, it is not uncommon to hear or read the view expressed that qualitative and quantitative research are based on fundamentally different views of the world and how it can, or should, be studied. If so, it would be very difficult – and, some would argue, completely out of the question – to employ both qualitative and quantitative methods in the same research study.
Thus, quantitative researchers may align themselves with the scientific method, portraying themselves as searchers for the objective truth about the world (social or physical) and how it works, and dismiss qualitative research as subjective, small-scale and lacking in rigour. Qualitative researchers, for their part, may highlight the richness and depth of their data and analysis, while accusing quantitative research of superficiality, arguing that everything – and certainly not the social world – cannot be reduced to mere numbers.
Often, of course, these preferences are linked to the skill sets of the researchers concerned (e.g. some people feel at home with large data sets and multivariate analyses; others are much more comfortable talking to or observing other people in natural settings). They may also reflect their underlying world-views of knowledge and how it can be accessed or developed (i.e. their ontologies and epistemologies).
Nowadays, however, more pragmatically orientated researchers (of which, as you may have surmised, I am one) have distanced themselves from these debates, arguing that mixed methods research, making use of both qualitative and quantitative methods as appropriate, has much more to offer (e.g. Bryman 2004; Scott 2007; Tashakkori and Teddlie 1998). Indeed, I would go further in arguing that all researchers (or, at least, social researchers) should be able to use, and interpret, both qualitative and quantitative methods, at least to some level of understanding.
The same argument may be applied to case study research, which, though it is sometimes claimed by qualitative researchers (probably because of its typically small-scale, detailed focus), may make use of quantitative techniques as well as, or instead of, qualitative methods. It is both possible and acceptable, therefore, to pursue case study research using quantitative data and methods. Thus, for example, the Sage Handbook of Case-based Methods (Byrne and Ragin 2009) contains five chapters on quantitative approaches to case-based method and five on qualitative approaches.
Ragin (2000) himself has developed another way forward, through his application of fuzzy set methods to case study research. Fuzzy set methods are an approach which accepts that the researcher may not be able to categorise particular cases as belonging wholly to particular groups (e.g. if the case is a nation, and we are examining democracy, is the case wholly democratic?), but can give them an indicative percentage membership figure (e.g. 70% democratic). This then allows an approach to the cross-case analysis of multiple case studies which is both qualitative and quantitative.
Debates Regarding Case Study
Criticisms of, and debates about, case study are of long standing (see the discussion in Chapter 2 on The Origins and History of Case Study; for more recent, but now also historic, accounts, see: Atkinson and Delamont 1985; Mitchell 1983; Platt 1992). Hence, those in favour of, or using, case study as a research design often feel obliged to defend their choice. Here, for example, Mitchell provides a firm defence of case study, resting on the rigour with which it is carried out:
[C]ase studies of whatever form are a reliable and respectable procedure of social analysis and… much criticism of their reliability and validity has been based on a misconception of the basis upon which the analyst may justifiably extrapolate from an individual case study to the social process in general… The validity of the extrapolation depends not on the typicality or representativeness of the case but upon the cogency of the theoretical reasoning. (Mitchell 1983, p. 207)
Note that Mitchell refers directly to the issues of both extrapolation (more commonly termed generalisation) and theoretical reasoning. As we will see, these come up repeatedly in discussions of case study.
Proponents of case study also often try to anticipate their critics by identifying and listing (at least some of) the criticisms that have been made of case study research, and then articulating responses to these. There is often a feeling of ‘setting up paper tigers’ about these exercises, with the supposed criticisms presented in an unduly favourable way which lends itself to a pointed riposte; one can, at times, be overly defensive.
One recent analysis along these lines, provided by Flyvbjerg (2004), is worth considering in some detail as an exemplar of its kind:
[T]he problems with the conventional wisdom about case-study research can be summarized in five misunderstandings or oversimplifications about the nature of such research:
1 General, theoretical (context-independent) knowledge is more valuable than concrete, practical (context-dependent) knowledge.
2 One cannot generalize on the basis of an individual case: therefore, the case study cannot contribute to scientific development.
3 The case study is most useful for generating hypotheses, that is, in the first stage of a total research process, while other methods are more suitable for hypothesis testing and theory-building.
4 The case study contains a bias towards verification, that is, a tendency to confirm the researcher’s preconceived notions.
5 It is often difficult to summarize and develop general propositions and theories on the basis of specific case studies.
These five misunderstandings indicate that it is theory, reliability and validity that are at issue; in other words, the very status of the case study as a scientific method. (Flyvbjerg, 2004, p. 391; see also Ruddin 2006)
Flyvbjerg then proceeds to challenge each of these ‘misunderstandings’ in turn. On his first point, he argues that in social research ‘we have only specific cases and context-dependent knowledge’ (2004, p. 392). We may accept that this statement is literally true: all research, as we have argued, involves the study of cases, while few social researchers would argue nowadays that they were seeking ‘objective’ knowledge. Yet, Flyvbjerg’s counter-point imposes a particular view of ‘reality’, denying the perspective of others who are searching for universal laws or explanations.
On his second point, Flyvbjerg contends that it depends upon the case and how it is chosen, an argument that, in Flyvbjerg’s view, also disposes of the third misunderstanding, with attention given to the selection of extreme, critical or even paradigmatic cases for study. This may seem a rather trite response. Is it really always possible to select cases that are particularly pertinent for study, and can one always know that this is the case before studying them?
In this respect, however, Flyvbjerg would appear to have the support of both Yin (2003), who offers a series of rationales for studying a single case,