8.14 In this chapter case study is defined as an in-depth, possibly longer term investigation of a single or very limited number of people, event, context, organisation or policy. A case study might be used when seeking to understand a significant or novel situation and to provide particularly rich data. Although the conduct of a case study can sometimes appear to be similar to that of a process evaluation, including in the generation of research questions and choice of methods (discussed below), there are key differences between the two which will affect how they are conducted and how the data generated can be understood.
8.15 Case studies will tend to be more localised or context specific than a process evaluation. That is to say they may look at a small-scale policy or project that is happening in only one, or a very small number of areas, and with limited numbers of participants. The policy or event being studied may even be a one-off situation such as the impact of the 1980 Cuban expulsion of workers into Miami on the labour market.3 Alternatively, a case study approach may be used to investigate a larger scale policy but the case study itself would tend to focus on the experience of the policy for a limited number of people or in a limited number of locations which are of particular interest to policy makers. This type of case study may be used on its own simply to provide data about the people or areas of interest or may contribute to a larger process evaluation by providing this more in-depth account as part of a wider analysis of the overall implementation and delivery of a policy. Whatever the context, case studies are likely to be used when what is required is a very detailed, in-depth understanding that is holistic, comprehensive and contextualised.
8.16 Case studies will tend to use a variety of research methods and triangulation (see Table 8.B) to develop a clear, well reasoned and comprehensive understanding of the situation, project or people being studied. This can provide very useful learning for analysts and policy makers to identify why something happened or did not happen, the mechanics of how a policy works, how people worked together or how behaviour was influenced, and very in-depth information about how a policy has been working in practice. As a result, this can help generate hypotheses and templates for wider roll out of a promising policy or suggest ways of working that might work in similar circumstances. However, because of their focus on a limited number of examples, unique situations, or small-scale projects, case study data should not simply be generalised to a context beyond that being studied and it is important that their results are reported and used with this understanding.
8.17 In some situations a case study approach can be used to assess both the implementation and the results of policy in a particular area, generating quantitative data to support an evaluation of the policy's impact. In this situation, if the data is suitably robust, there is sufficient sample size and an appropriate comparison group to assess what would have happened in the absence of the policy (the "counterfactual"), then an impact evaluation may be possible within the case study. Guidance on impact evaluation can be found in Chapter 9. However, there is likely to be a lower possibility of generalising findings from impact evaluations which have been conducted within a small case study than in evaluations which have a broader scope.
_________________________________________________________________________
3 The Impact of the Mariel Boatlift on the Miami Labor Market, Carol, 1990, Industrial and Labor Relations Review Vol. 43, No. 2 (Jan., 1990), pp. 245-257