Case Study Reports as Part of Larger, Mixed Methods Studies

Your completed case study may include data from other methods (e.g., surveys or quantitative analysis of archival data such as health status indicators). In particular, Chapter 2 pointed to the possibility that within a single case might exist embedded units of analysis, which might have been the subject of data collection through these other methods (see Chapter 2, Figure 2.3). In this situation, the case study encompasses the other methods, and your completed case study report would incorporate the reporting of the data from these other methods (e.g., see Chapter 4, BOX 18).

A totally different situation occurs when your case study has been deliberately designed to be part of a larger, mixed methods study (Yin, 2006b). In this situa­tion, the larger study encompasses the case study. The larger study will contain your completed case study but also should report separately the findings about the data from the other methods. The larger study’s overall report would then be based on the pattern of evidence from both the case study and the other methods.

This mixed methods situation deserves a bit more attention so that you will understand its implications for your case study, even though you might not compose your case study report any differently than if it had been a “stand­alone” report. At least three different rationales might have motivated the larger study to use mixed methods.

First, the larger study may have called for mixed methods simply to deter­mine whether converging evidence (triangulation) might be obtained even though different methods had been used (Datta, 1997). In this scenario, your case study would have shared the same initial research questions as those driving the other methods, but you would likely have conducted, analyzed, and reported your case study independently. Part of the larger study’s assess­ment would then be to compare the case study results with those based on the other methods.

Second, the larger study may have been based on a survey or quantitative analysis of archival data—for example, a study of households’ financial situ­ations under different income tax conditions. The larger study might then have wanted case studies to illustrate, in greater depth, the experiences of individual families. In this scenario, the questions for your case study might only be surfaced after the survey or archival data had been analyzed, and the selection of cases might come from the pool of those surveyed or contained within the archival records. The main implications for your case study effort are that both its timing and direction may depend on the progress and find­ings of the other inquiries.

Third, the larger study might knowingly have called for case studies to elu­cidate some underlying process and used another method (such as a survey) to define the prevalence or frequency of such processes. In this scenario of complementarity as opposed to convergence, the case study questions are likely to be closely coordinated with those of the other methods, and the com­plementary inquiries can occur simultaneously or sequentially. However, the initial analysis and reports from each inquiry should be conducted indepen­dently (even though the final analysis may merge findings from all the dif­ferent methods). BOX 42 contains two examples of larger studies done under this third scenario.

These three different situations show how your case study and its reporting may have to be coordinated within some broader context. Beware that when your case study is not independent, you may have to coordinate deadlines and technical directions, and your case study report may not proceed as you might have expected initially. Also assess carefully your willingness and ability to be part of a larger team before making any commitments.

Box 42

Integrating Case Study and Survey Evidence:
Complementarity of Findings

Multimethod studies can pose complementary questions that are to be addressed by different methods. Most commonly, case studies are used to gain insight into causal processes, whereas surveys provide an indication of the prevalence of a phe­nomenon. Two studies illustrate this combination.

The first was a study of educational projects funded by the U.S. Department of Education (Berman & McLaughlin, 1974-1978). The study combined case studies of 29 projects with a survey of 293 projects, revealing invaluable information on the implementation process and its outcomes. The second study (Yin, 1981c) combined case studies of 19 sites with a survey of 90 other sites. The findings contributed to the life cycle of technological innovations in local public services.

Source: Yin K Robert (2008), Case Study Research Designs and Methods, SAGE Publications, Inc; 4th edition.