Mixed Methods in Case Study Research

Conceptual Overview and Discussion

Case study research lends itself particularly well to mixed methods research, as myriad approaches to research design, analysis, and interpretation are possible. Historically, mixed methods research ended the paradigm wars and the “incompatibility wars” because it combined the quantitative and [Page 562]qualitative research methods. From the 19th century to the 1950s, the purist approach to research utilized one single data source or several data sources within either a quantitative or qualitative research paradigm. This method of research has been labeled monomethod or monostrand due to its limited research approach. From the 1960s to the 1980s, monomethod research designs gave way to multimethod (or multistrand) approaches that ensured any of the four types of triangulation: data, investigator, theory, and methodological. From the 1990s to the present day, mixed model studies have come to the forefront; in these, quantitative or qualitative can be used as the focus of the inquiry, at the data collection stage, and at the data analysis and interpretation stage, or a combination of quantitative and qualitative methods at each of the aforementioned phases of the research study.

Mixed methods research works particularly well for case study research as it allows the researcher to take the rich empirical data yielded from case studies and apply either quantitative or qualitative methods or quantitative and qualitative methods to the data. In this manner, qualitative data can be quantitized or quantitative data can be qualitized to extract meaning from the data sets that might otherwise be hidden. For example, a case study researcher might transform interview data on how the roles of teachers have changed with technology infusion into a display comparing who has and who has not changed with a number of technology infusion inhibitors such as infrastructure, monies, and administrative support. A Fisher's exact probability test could be applied to see if there is a statistically significant difference between a teacher infusing technology and an inhibiting factor. Conversely, a quantitative researcher might administer a 50-item questionnaire related to the causes of juvenile delinquency from which the researcher can cluster techniques to form groups on which statistical analysis can be performed. These same data can be analyzed qualitatively to form another data set that will describe separate narrative profiles of a juvenile delinquent. The combination of qualitative and quantitative techniques enhances legitimation as the qualitative analyses involve descriptive precision and the quantitative analyses ensure numerical precision.

Application

Two seminal exemplars of mixed methods research are Jennifer Greene, Valerie Caracelli, and Wendy Graham's 1989 conceptual framework for mixed method evaluation designs, and David Fetterman's 2000 evaluation of the Stanford University Teacher Education Program (STEP).

Greene and her Cornell University colleagues examined 57 empirical mixed methods studies to develop a sound conceptual framework upon which myriad subsequent studies have been based. They surveyed the professional literature for studies conducted between 1980 and 1988 and reduced the initial number to 18 published evaluation studies, 17 evaluation reports, and 22 evaluation papers. Their analysis revealed five different purposes for mixed methods research and seven related design characteristics.

...