I am hearing the song “Secretly Minnesotan” by Tullycraft.
For this journal entry, I am reviewing the “Writing Guides: Case Studies” site from Colorado State University. The site, in general, was helpful in refining some of the vocabulary that I have to describe the effectiveness of a case study.
According to this site, case studies are created to answer qualitative “how” or “why” questions. The term “case study” is used interchangeably with ethnography, field study and participant observation. Case studies are based on observation in real-life environments rather than in a controlled lab environment. (Overview).
There are 4 different types of case studies:
- Illustrative: Have the goal of giving readers examples of a particular phenomenon and provide a common vocabulary for discussing the phenomenon.
- Exploratory: Quick, small case studies executed to refine guiding questions for a larger-scale case study. Findings should not be generalized.
- Cumulative: Collects and analyzes information from several related case studies to eliminate time needed to engage in several new studies.
- Critical Instance: Examines a unique scenario with one of two goals: to establish the how and why in a particular scenario or to provide a counterexample to a commonly-held belief.
(Types of Case Studies)
Naturally, case studies are guided by questions from the researcher. Questions usually tie to a theoretical approach. Three acceptable theoretical approaches are the following:
- Individual Theories: focuses on individual development of subject.
- Organizational Theories: focuses on the impact of formalized, organized social structures on overall organizational performance.
- Social Theories: focuses on the cultural impact on a group.
(Identifying a Theoretical Perspective)
Once the guiding questions have actually been defined, it is time to design the study. Before the case study is actually started, it is important to identify:
- relevant data,
- which data to actually collect,
- and how the data will be analyzed.
In addition, items such as how the data will be linked to the initial assumptions (propositions) of the study and criteria for evaluation should be considered. (Designing a Case Study)
This site recommends a multi-modal (gathering data from a variety of sources) approach to collecting data for the study. Different methods mentioned include:
- interviews
- protocol analyses
- field studies
- participant-observations
- conversations iwth students
- discrete observations of actual composition
- writing samples
- school records
- classroom observations
- faculty and student interviews
- self reports from subjects
(Method: Single or Multi-modal?)
Interestingly, the site recommends that the pool of participants remain relatively small. And, it suggests that a “cross section” of representatives across a student body is not necessary. Part of the case study should be dedicated to describing the “case histories” of the study participants. One question that I would ask myself here is how much “case history” should I present? I would assume that only the “relevant” parts… but, then that raises the question: what method do I use to determine which parts of the history are relevant? (Participant Selection)
According to the site, there are 6 different types of data that can be gathered for a case study. These include:
- documents,
- archival records,
- interviews,
- direct observation,
- participant observation, and
- artifacts.
The article again reinforces that a variety of types of data be collected to have the study operate in a “corroborating mode”. Throughout the course of the study, the researcher should continue to question and analyze the data. This could lead to the identification of new patterns and questions that could be pursued in a later study. (Data Collection)
This is a difference from the article that I read from George Mason University on teacher research (see: The Question Cycle). In the outline presented here, the questions or patterns that come out of observation turn into new questions that might drive another study, but it does not impact the current study. From my perspective, this seems to be a sounder methodological approach.
Once you have collected data, you need methods of analyzing it. Roughly speaking, there are two general categories of analysis:
- Holistic: When you reflect on the entire scenario and draw conclusions.
- Coding: When you look at instances of different pieces of evidence that serve as indicators of whether or not the guiding question is right or wrong. Selecting aspects of the data to serve as coding indicators needs to be carefully done so as to ensure that it actually fits the guiding question. It may be the case that new variables emerge and are identified in the coding process that may not have been planned for in the beginning. It is up to the coders as to whether or not to use those variables.
(Data Analysis)
After data collection and analysis are completed, then it is time to write up the case study. The site points out that the case studies are basically a story that steps through the research process. What is the most challenging aspect of this, they note, is knowing which aspects of the research process need to be included in the case study writeup. They specifically mention including:
“a detailed explanation of the researchers theoretical positions, of how those theories drove the inquiry or led to the guiding research questions, of the participants backgrounds, of the processes of data collection, of the training and limitations of the coders, along with a strong attempt to make a connection between the data and the conclusions…”
As a closing, the case study should identify new questions or variables that emerged from the study. Implications for the study should be drawn. Suggestions for possible other studies should be made. (Composing the Case Study Report)
One of the most interesting points of the site (to me) was the discussion of reliability and validity. The site readily admits that it is problematic to try to categorize a case study as valid and reliable. They spend time looking at different types of research reliability (equivalance relability, stability reliability, internal consistency, interrater reliability) and validity (face validity, criterion related validity, construct validity, content validity). Further, they also question the idea that case studies are generalizable. (Concerns about Generalizability) That was refreshing for me. They did, however, list methods of making studies more “confirmable”. These include:
- prolong process of data gathering,
- employ process of triangulation (there’s that concept again),
- talk to your subjects about findings,
- colect referential material and
- get feedback from peers along the way.
(Steps to Improve Validity and Reliability)
This site was particularly helpful in shaping my overall understanding of the elements of a case study. This site does leave some holes (triangulation, methods for selecting the “relevant” case history facts).
Case Studies. (2009, Februar 16). Colorado State University Writing Guides. Accessed February 18, 2009, from http://writing.colostate.edu/guides/research/casestudy/index.cfm.