Conceptual Argument

Overview and Discussion

A case can be considered an experiment, capable of producing theoretical propositions that allow predictions to be made about the circumstances under which similar or contrary findings might be expected. Case studies excel at answering “how” or “why” questions about contemporary reallife phenomena, where the researcher has no control over events and the relationships are complex and unfold over time. Case studies are frequently used to generate theory when little is known about a phenomenon, or findings are contradictory or in conflict with common sense, but they can also reveal limitations of, or extend, existing theories, and can provide insight about typical, extreme, or deviant phenomena.

According to Robert Yin, who has written extensively on case study research methodology, ruling out threats to internal and external validity to produce credible findings requires rigorous case study research design. The research design establishes the roadmap for how the study will be conducted, so that causal inferences can be made. It covers what questions to study, what data to collect, and how to analyze results. To make a convincing conceptual argument using a case study, the case(s) must be selected purposely to generate maximum insight into the phenomenon of interest. Data collection should be guided from the outset by tentative hypotheses or propositions that identify plausible, potentially significant relationships and constructs.

To enhance the reliability of the study findings, a research protocol must be established in advance that sets out, for each of the questions to be addressed, what procedures will be followed for data collection, analysis, and record-keeping. It must be flexible enough to permit the investigation of unexpected findings. The procedures outlined must be followed throughout the study. A database of researcher notes, documents, tables, and transcripts should be created. As well, to make replication possible, a separate evidentiary base that links the study questions and conclusions should be maintained. Evidence must be kept separate from interpretations.

To collect high-quality data, it is imperative that data collection be guided by preliminary theory or, if no theory exists, by a clear explanation of what is to be explored and why, and how the success of the study will be demonstrated. Data should be collected from multiple theoretical perspectives and data sources, using a variety of methods, and, where possible, multiple researchers. This approach to data collection is known as triangulation. Robert Yin states that triangulation provides “between methods” validation, which increases internal validity because measures are independent, and “within methods” validation, which increases reliability by providing assurances of consistency. Obtaining converging evidence via triangulation aids in theory development. Collecting both qualitative and quantitative data can help clarify relationships between variables, overcome researcher biases and blind spots, and prevent unsubstanti-ated theorizing. Data collection should continue until new data yield few additional insights.

...