Skip to main content icon/video/no-internet

Use of the case study in the natural and social sciences began in the early 20th century with the establishment of the concept of a case history in medicine. The case study is also used outside of the natural and social sciences in education, educational evaluation, and as a teaching method in business and law. In the contemporary social sciences, the case study can be a single case or multiple cases. One method of studying multiple cases is the cross-case design proposed by Matthew B. Miles and A. Michael Huberman. While Miles and Huberman are credited with stipulating the precise logics and mechanics of this approach, it has a long history and wide usage beyond their specific formulation and is a general qualitative research strategy. While the single case study emphasizes close inspection and description of one case, the cross-case is used to reinforce validity, support generalizability, and promote theoretical elaboration. In psychology, Sigmund Freud's theory of psychoanalysis is perhaps the most widely known example of cross-case synthesis and analysis. In sociology, a cross-case methodology is the logic inherent in Barney Glaser and Anselm Strauss's constant comparative method for the generation of grounded theory.

Conceptual Overview and Discussion

Miles and Huberman propose two interdependent strategies for cross-case analysis: variable-oriented analysis and cross-case analysis. They formulate the cross-case method as follows. First, the researcher conducts two or more descriptive case studies using the process known as thick description. Then the researcher examines a particular case for emergent patterns. The second step involves identifying the presence or absence of and alternatives to these patterns occurring in one or more other cases. Doing this analysis across a number of cases allows for the identification of similarities and differences across the cases and the identification of common themes.

Miles and Huberman recommend that researchers then subject the themes across all of the cases to variable-oriented analysis. First, standard variables across the cases and within the themes are identified. Each case is then written up and presented in a matrix, in as much detail as possible, with reference to these standard variables. Finally, a meta-matrix is created through synthesis or “stacking” of the variables across all of the cases. In this way the researcher retains both the detail of the individual cases and partial but overall explanation of a number of cases.

A good example of this method can be found in Molla S. Donaldson's study on the quality of healthcare in the United States. The results are based on thick description and the analysis of 45 healthcare delivery units, one of which was in Canada and one other in the United Kingdom. Interviews were conducted in seven cities in a variety of types of healthcare delivery locations such as hospitals, primary care centers, and family physicians, with large numbers of persons including some 270,000 patients and 15 medical teams. The protocol suggested by Miles and Huberman is followed closely: 14 variables are identified and then refined, through cross-case analysis, into eight themes. Generalizations about how healthcare delivery systems function and broader issues such as the sources of excellence and innovation among them as well as suggestions for future training of healthcare professionals and formulating healthcare policy are then made.

...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles

Sage Recommends

We found other relevant content for you on other Sage platforms.

Loading