Practical considerations in case study research: the relationship between methodology and process.

This paper explores the extent to which the design of a research study is influenced by pragmatic as well as theoretical considerations by describing the authors' experience of conducting multiple case studies in the context of a policy-orientated research project. The case studies in question formed part of a collaborative project exploring the development of innovative roles in nursing and the professions allied to medicine. The researchers adopted Yin's approach to the conduct of multiple case studies to explore issues relating to the effectiveness of new roles, their resource implications and educational issues surrounding their development. They address the fact that methodological decisions have practical implications and also explore the effects of practical and ethical issues on research design. Three purposively selected acute National Health Service Trust hospitals in England formed the case study sites for the work recorded in this paper. In each Trust, three nursing roles were selected, making nine cases for study. A variety of data collection strategies were used including semi-structured interviews with post-holders and other stakeholders (n = 51), non-participant observation of meetings, collection of audit, financial and statistical data and review of relevant documentation. This paper describes some of the dilemmas which arose during the course of the study, the rationale for methodological decisions taken to resolve these dilemmas and the steps taken to enhance rigour. In particular, the authors discuss the problems associated with obtaining informed consent in the context of qualitative research and with member validation of transcripts in a context where confidentiality was crucial. The decision to present findings thematically rather than as individual cases is justified both in terms of the nature and purpose of the research and in relation to the particular importance of anonymity and confidentiality in this study. Finally, the degree of reciprocity between researcher and researched is discussed.