Research Context & Design. Baxter and Jack (2008) summarized Yin's discussion of when case study design may be a useful option:
"(a) the focus of the study is to answer “how” and “why” questions; (b) you cannot manipulate the behaviour of those involved in the study; (c) you want to cover contextual conditions because you believe they are relevant to the phenomenon under study; or (d) the boundaries are not clear between the phenomenon and context."The applicability of case study methodology in these contexts has, as Tellis pointed out, meant that "The effects of community-based prevention programs have been widely investigated using case methodology." (Tellis, July 1997) Yin (1989) discussed methods to ensure case study quality for construct validity, internal validity, external validity, and reliability (40-45). Baxter and Jack (2008) summarized key foundations for research quality:
"As a basic foundation to achieve this, novice researchers have a responsibility to ensure that: (a) the case study research question is clearly written, propositions (if appropriate to the case study type) are provided, and the question is substantiated; (b) case study design is appropriate for the research question; (c) purposeful sampling strategies appropriate for case study have been applied; (d) data are collected and managed systematically; and (e) the data are analyzed correctly." (556)Case studies are designed accordingly. Yin (1989) described a 2X2 matrix, where case studies may be single or multiple with holistic or embedded units of analysis. (46) He further described three approaches or purposes for a case study: 1) explanatory, 2) exploratory, and 3) descriptive. (13) Care must also be taken to determine the unit(s) of analysis and bind the case. (Baxter and Jack 2008). To this end, Tellis (September 1997) outlined the "procedures recommended in the literature, followed by a discussion of the application of those procedures in the proposed study."
Data Collection & Sources of Evidence. Case study methodology also allows for inclusion of a variety of data sources. Yin (1989) listed six sources of evidence: "documentation, archival records, interviews, direct observations, participant-observation, and physical artifacts" (85). As Tellis (September 1997) discussed,
"The rationale for using multiple sources of data is the triangulation of evidence. Triangulation increases the reliability of the data and the process of gathering it. In the context of data collection, triangulation serves to corroborate the data gathered from other sources."However, the advantage is not without risks. Baxter and Jack (2008) points out that one common danger is "the collection of overwhelming amounts of data that require management and analysis." To cope with large amounts of qualitative data, Tellis (September 1997) also discussed three principles for collecting and using data: "1. Use multiple sources of data, 2. Create a case study database, 3. Maintain a chain of evidence."
Yin (1989) highlighted the importance of training and preparation for conducting case studies. Investigators must have highly developed communication skills, be adaptive, understand a multitude of related issues, and maintain openness without bias to preconceived notions. The research team must also be briefed in areas relevant to their participation, including: field procedures, case study methodology, tasks, interview skills, data management and study purposes. (61-83)
Data Analysis & Program Implications. Yin (1989) emphasized the importance of developing an analytic strategy. This can be especially important since, as Baxter and Jack (2008) pointed out, collection and initial analysis may occur concurrently. Tellis (July 1997) summarized four key principles for analysis:
- "Show that the analysis relied on all the relevant evidence
- Include all major rival interpretations in the analysis
- Address the most significant aspect of the case study
- Use the researcher's prior, expert knowledge to further the analysis"
propositions, explanation building, time-series analysis, logic models, and cross-case
synthesis." (Baxter and Jack 2008) Yin (1989) calls "relying on theoretical propositions" "The first and more preferred strategy" (106). Theoretical foundations typically shape the data collection and are therefore especially relevant for analysis. Pattern matching is strongly related to theoretical foundations, including using rival explanations or explanation-building to analyze patterns. (ibid, 109-115)
This reliance on theoretical propositions extends throughout the logic of the case study methodology. As Tellis (July 1997) concluded, "generalization of results, from either single or multiple designs, is made to theory and not to populations."
