Good morning everyone. I'm feeling a bit perplexed. I've spent the last few weeks reading different dissertations that were similar in structure.
Current Set-Up
I proposed a multicase study in which I would research language ideologies held by individuals overseeing a specific program in four different school districts. The initial plan called for a semi-structured interview followed by a more open-ended interview. The goal would then be to draw cross-case conclusions and hopefully inform my initial theory.
The feedback that I was given was that case studies needed to be more in-depth and that this set-up didn't necessarily lend itself to a case study.
Rather than re-writing my entire chapter 3, I'm looking to see if anyone could share any insights on how I can modify the design (ie: do I limit the study to two districts and interview 2 individuals at each with a slightly different level of oversight of the program) to allow me push through this last obstacle in my proposal.
Any assistance would be greatly appreciated.
Jesus Navarrete
I didn't really understand what exactly you are studying. In any way, you need to define on what basis you have chosen those four districts? If your districts show specific crisscrossing characteristics and that's the reason you chose those, you can call it a case study, because those four are then sort of considered "the case". Otherwise, I believe, it could be just a comparative study.
I would tend to agree that two interviews in each of four sites is not enough for a thorough case study. The best book that I know on comparative case studies is Yin, Case Study Research and Applications.
Thank you both for your comments.
Hayk Smbatyan - The program being studied is a mandatory program across all public schools in Texas that contain a particular subgroup. In this particular "case" the chosen schools would all be charter schools that offer a variation of the program. So in this instance, they would all be Texas Charter Schools that implement a variation of the required program. For transparency sake, it's programs that target ELs, bilingual and/or ESL programs.
David L Morgan - I've been reading Yin quite a bit. The part where I'm a bit stuck is on whether it's a single case (looking at the programs themselves) or if each site could be considered a case.
Here's a tidbit of what I had in my Chapter 3.
In a multiple-case design, sampling logic, an approach in which the resulting data from the sample of potential respondents are assumed to be representative of the entire pool, is not used. Sampling logic relies on “statistical generalization,” which is influenced by empirical data collected about a sample that has been processed by quantitative formulas to determine the confidence in which the generalizations can be made. Replication logic, on the other hand, allows for more compelling findings should the cases turn out similar results or conclusions. Yin (2003b) recommends five, six, or more replications of cases if a higher degree of certainty is preferred. The cases are not “sampling units,” but should be considered “like multiple experiments” (Yin, 2003b, p. 32). “Analytic generalization” occurs when a previously developed theory is utilized to compare the results of a different case study, and they show support for the same theory. This is when replication can be claimed.
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology