Sampling

Conceptual Overview and Discussion

Sampling is a complex issue in case study research, because there are many variations of sampling strategies described in relevant literature. Researchers have generally agreed that the aims of the particular study should guide how cases are selected. Sampling in case study research is largely purposeful, that is, it includes the selection of information-rich cases for in-depth study. Information-rich cases are those from which the researcher can learn a great deal about issues of central importance to the purpose and investigated phenomena of the study. The case study approach offers flexibility in terms of the justification of sampling choice, the number of investigated cases, and sampling techniques.

A review of relevant literature puts forward different strategies for purposefully selecting information-rich case studies. For instance, Michael Quinn Patton identifies 18 different sampling strategies that may be employed in case study research—2 forms of random sampling (simple random samples and stratified and cluster samples) and 16 forms of purposeful sampling—and recommends that the selection of cases involves purposeful, not random, selection. The types of purposeful sampling identified are as follows: theoretical/theory-based/operational construct, convenience, extreme/deviant/outlier, intensity, maximum variation, homogeneous, typical, critical, snowball, criterion, confirming and disconfirming, stratified purposeful, opportunistic, purposeful random, politically important, and combination/mixed purpose. Sampling involves the initial selection of the case or cases and within-case sampling in terms of the informants, observations, documents, and so on.

The distinction among purposeful, selective, and theoretical sampling often lacks clarity in the literature. As a result, these terms are viewed synonymously and used interchangeably even though they are defined differently. Purposeful sampling is an umbrella concept that embraces the strategies of theoretical and selective sampling. Theoretical sampling has been defined in grounded theory terms as sampling on the basis of emerging concepts. Such an approach to case study sampling involves systematically examining and refining variations in emergent and grounded concepts. Thus, sampling can either be prespecified up front or evolve progressively once the field work begins. Theoretical sampling derived from grounded theory can be distinguished from selective sampling, which refers to a decision made before beginning the study to sample subjects according to a preconceived, initial set of criteria. Initial samples may be chosen at the early phases of the investigation (selective sampling), and then others can be selected [Page 838]according to categories emerging from the data (theoretical sampling).

The literature distinguishes between and purposeful and representative random sampling and suggests that in some case study research—for example, in evaluation studies—the credibility of systematic and randomly selected case examples is considerably greater than personal, ad hoc selection of cases. However, purposeful random sampling is used to enhance credibility; it is not a representative random sample for generalization to populations. Indeed, it has been argued that that random sampling of cases is neither necessary nor preferable. In qualitative sampling the focus is on selecting information-rich cases for in-depth study, to enhance the richness, validity, and depth of the information.

...