may be defined by role or function
may be shaped by organizational or institutional arrangements. (p. 254)
This set of characteristics seems both broad-ranging and rather repetitive.
Of course, there are many other social research methods texts which we might also examine, but this small sample already illustrates the range of alternative perspectives that are taken towards case study. Judging by these texts – summarised in Box 3.1 – we could view case study as a method, approach, style, strategy or design. And, whichever (and however many) of these it might be, it can be conceived in relation to a wide, but differing, range of other social research methods, approaches, styles, strategies or designs. These could include action research, comparative studies, cross-sectional studies, ethnography, experiments, grounded theory, historical studies, internet-based studies, interviews, longitudinal studies and surveys (these relations are considered further in the next section).
Box 3.1 Alternative Research Classifications of Case Study
Punch (2005): ‘approaches to qualitative research design’
Case studies
Ethnography
Grounded theory
Action research
Burns (2000): ‘qualitative methods’
Ethnographic research
Unstructured interviewing
Action research
Case studies
Historical research
Bryman (2004): ‘research designs’
Experimental design
Cross-sectional design
Longitudinal design(s)
Case study design
Comparative design
Cohen Manion and Morrison (2007): ‘styles of educational research’
Naturalistic and ethnographic research
Historical and documentary research
Surveys, longitudinal, cross-sectional and trend studies
Internet-based research and computer usage
Case studies
Ex post facto research
Experiments, quasi-experiments, single-case research and meta-analysis
Action research
But can we not be clearer about what case study is and isn’t? In a carefully considered assessment, Verschuren (2003) sets out to ‘clarify some ambiguities and misconceptions as to case study as a research methodology, and to define it more clearly as a research strategy’ (p. 122). He notes differences among practitioners with respect to: ‘(a) the empirical object of a case study and the way we look at it; (b) the research methods that are used; and (c) the adequacy of the results to be obtained’ (p. 122, emphasis in original). He argues that ‘the main characteristic of a case study as a way of doing research is that it is a holistic rather than a reductionistic approach’ (p. 128), with implications for both the object of observation and the methods used for generating research material.
We should not be surprised, of course, to find that case study has been interpreted in a variety of ways (see also the discussion of What is a Case Study? in Chapter 2). Similarly, we may readily accept that the terms method, approach, style, strategy and design – as they are used in the discussion of research – share overlapping meanings. Nevertheless, it will be helpful, for the discussion in the remainder of this chapter and throughout the book, to pin things down a little more firmly.
My own perspective, then, is that it is most sensible to view case study as a research design. As such, it represents a way of pursuing a particular research project or projects. Within this research design, as within others, particular methods may then be adopted in order to progress the research.
Case Study as a Research Design
Where does case study sit in comparison with other research designs? We have already seen (in Box 3.1) that research designs, and the place of case study within them, have been variously categorised by different authors. There is, then, no single, definitive and widely accepted classification of research designs.
We will examine in more detail, however, one of the classifications already mentioned, that proposed by Bryman (2004). Bryman’s classification has, to my mind, the considerable advantages of being both straightforward and neatly structured. He begins by explaining what he means by research design:
A research design provides a framework for the collection and analysis of data. A choice of research design reflects decisions about the priority being given to a range of dimensions of the research process. These include the importance attached to: expressing causal connections between variables; generalizing to larger groups… understanding behavior… in its specific social context; having a temporal appreciation of social phenomena. (p. 27, emphasis in original)
Five research designs are then identified: experimental (the manipulation of one variable to measure its impact on another), cross-sectional (‘the collection of data on more than one case and at a single point in time’ [p. 41, emphasis in original]), longitudinal (the repeated collection of data on at least two different occasions), case study (which Bryman defines simply as ‘the detailed and intensive analysis of a single case’ [p. 48]) and comparative (‘the study using more or less identical methods of two contrasting cases’ [p. 53]).
In Bryman’s terms, then, case study as a research design is less concerned with establishing causal connections, generalisation or temporal connections; it focuses instead on ‘understanding behavior in its specific social context’. Or, to put it the other way around, if you are engaging in a research project with the intention of understanding some aspect of behaviour in its social setting, case study is an obvious and appropriate research design to employ.
Interestingly, case study would also be a possible approach (note that I am now using the term ‘approach’ rather than ‘design’ to indicate that an alternative overall design to case study is being employed) for at least three of the other four research designs suggested by Bryman. Indeed, this is clearly suggested by his use of the word ‘case’ in the sense of ‘item’ or ‘individual’ in the quotations given. Thus, a comparative design could proceed using a multiple case study approach, as might a cross-sectional design (though here a survey is probably a more obvious approach). A longitudinal design might involve returning to the same case study (or case studies) on a number of occasions.
The only one of Bryman’s five designs where case study does not suggest itself as an approach is the experimental design, though even here the manipulation of the variable(s) involved could take place within cases (see the discussion of experiment as a research design that can be combined with case study in Chapter 6). What this discussion illustrates, of course, is the care needed in talking about, and distinguishing between, case study and cases. As was noted in Chapter 2, all research involves the study of cases in some form; case study is a research design which focuses on the detailed study of a particular case or cases.
The Qualitative/Quantitative Debate
[D]ifferent conceptions of the term ‘case’ are central to