The formatting is a bit wonky converting from word into padlet but hopefully it is easy enough to follow. I look forward to discussing this chapter with you two in person next weekend!

I.               General Approach to Designing Case Studies

a.     There are five components of case study design:

                                               i.     Study questions- Case study is a robust approach when asking how or why questions that are context bound.

                                             ii.     Study propositions- Statements about what you theoretically think may be going on related to your how and why questions.  These statements can focus your attention on what you may want to explore or examine and where to begin looking for evidence. 

                                            iii.     The “case”- “The desired case should be a real-world phenomena that has some concrete manifestation” (p. 31).  The case can be anywhere from a single person to multiple organizations or governments, and you can conduct single-case or multiple-case studies.  No matter what the case it must be bound in some way that distinguishes who falls within the case and who is outside of the case.  The case should also be time bound geographically and within a time-period.

                                            iv.     Linking data to propositions- “The actual analyses will require that you combine or assemble your case study data as a direct reflection of your study propositions” (p. 33).  During this part of the design, you try to anticipate and plan for collecting just the right amount of data that will allow you to connect the data to those initial theoretical ideas you may have had about your how and why questions.  It is important even when linking the data to propositions to be open to considering the data outside of some of those propositions, which helps build the strength of the findings as discussed in the next section. 

                                             v.     Criteria for interpreting the strength of a case study’s findings- Rather than relying on p values for statistical analysis, case study considers the identification and addressing of explanations that offer alternative views or “rival explanations” to your findings as essential to demonstrating the strength of your findings.  You do not want to wait until you are analyzing the data to address these rival explanations.  Instead, when designing the case study, you want to anticipate as many rival explanations as possible and plan for how to address these because “The more rivals that have been addressed and rejected, the stronger will be your findings” (p. 34). 

 

II.             The Role of Theory in Research Designs

a.     When developing the theoretical framework to guide a case study you should first consider your initial theoretical notions, your propositions, what are you already thinking about who or where you might want to research and how you will go about doing that research.  It is also helpful at first to review completed case studies related to your topic and consider the theoretical frameworks that have guided those scholars.

b.     Multiple types of theories can inform the theoretical framework of your case study research such as individual theories, group theories, organizational theories, and social justice theories.  It is important to consider your desired outcome and audience for your research to try to avoid a significant mismatch in theoretical grounding and outcome. 

c.     Unlike quantitative methods that rely on statistical generalization, case study relies on analytic generalization.  When generalizing analytically, you are looking at the case study as a whole and how it might generalize to other concrete situations, not generalizing about the individual cases.  You can do analytic generalization whether doing a single-case study or a multiple-case study but either way the “challenge with making analytic generalizations involves understanding that the generalization is not statistical (or numeric) and that you will be making an argumentative claim” (p. 41).  The initial theoretical propositions that you plan to tie to the data will be the guiding framework for analytic generalization from your findings. 

 

III.           Criteria for Judging the Quality of Research Designs

a.     Construct validity- This is a challenge and a common critique of case study due to the subjectivity of researchers and the possibility of using data to confirm your initial propositions rather than present rival explanations.  Providing solid definitions of the specific concepts within the case and providing operational measures that match those concepts support you to establish construct validity.  Additionally, like with other forms of qualitative research multiple sources of evidence help with construct validity. 

b.     Internal validity- This can be a concern if you are trying to prove or show a direct causal relationship in an explanatory case study.  There is concern about case study researchers making inferences rather than drawing theoretical framed conclusions.  Yin suggests anticipating answers to the following questions in your design; is the inference correct?  Have all the rival explanations and possibilities been considered?  Is the evidence convergent?  Does it appear to be airtight?  Using specific tactics for analytic generalization can be helpful with internal validity, and those tactics are discussed in later chapters of the book. 

c.     External validity- This deals with whether the findings of any particular case study are generalizable to other contexts and situations.  The specific research question and the form of research it warrants can be used to plan for external validity.  For example, descriptive case studies are a good match for finding out about your how questions whereas explanatory case studies are a better fit for exploring your why questions.  Thus, matching the questions and providing justification for the specific method that matches the questions can be helpful in establishing external validity. 

d.     Reliability- While the ability to repeat a study entirely does not necessarily apply to case study there is still a need to document the procedures you plan to use and then do eventually follow in as much detail as possible to address reliability.  According to Yin, “the general way to approach the reliability problem is to make as many procedures as explicit as possible and to conduct research as if someone were looking over your shoulder” (p. 46). 

 

IV.           Case Study Research Designs

a.     The selection of your case will depend on your questions and your theoretical framework.

b.     Single-case

                                               i.     There are five rationales for using single-case study.  These rationales are based on what type of case you feel is represented; a critical, unusual, common, revelatory, or longitudinal case.

1.     A critical case is critical to the theory or theoretical propositions you have outlined and can be used to find out if there are in fact specific circumstances in which your propositions are true.

2.     An unusual case would be one that deviates from the theoretical norms or everyday circumstances.

3.     A common case would capture the circumstances and conditions that are typical for your theoretical stance.

4.     A revelatory case is a case that has not had much chance to be studied before so you can provide completely new insight into the theoretical propositions.

5.     A longitudinal case would be one you could follow to examine theoretical propositions over time.

                                             ii.     Holistic vs. Embedded            

1.     Each of these methods has their benefits and drawbacks.  With holistic it is easier to keep your focus on the big picture of the case itself but problems can arise if your focus shifts too much or in ways you do not give proper attention.  Embedded single-cases are where some subunits within the overall case are explored and this can help you to maintain the focus of your research as they look at potential smaller pieces but focusing on smaller pieces within the overall case can cause problems if those smaller pieces are not then brought back into the frame of the case overall. 

c.     Multiple-case

                                               i.     A multiple-case study can allow for a richer analysis of data and is generally seen as providing more robust findings but also takes significantly more resources and time to complete compared to a single-case study.

                                             ii.     Rather than relying on sampling logic like survey studies do, multiple-case study relies on replication logic where each individual case is meant to replicate in some way the findings from the other cases (whether to confirm or contradict).

                                            iii.     It is important to be conscious of the feedback that you are receiving throughout a multiple-case research design to note whether or not some redesign may be necessary.  Otherwise, you are jeopardizing the validity of your study.  Remember that an important way to strengthen the findings in case study is to address rival explanations so in multiple-case design you want to consider those individual cases within your overall design that may provide rival explanations that you can then address.

                                            iv.     Holistic vs. Embedded.  As in single-case design, a holistic approach would deal with each of the cases in the multiple-case design as a whole unit whereas embedded would require you to examine the subunits within each case of the multiple-cases and then tie that back to the bigger picture of the overall case.  In an embedded multiple-case design you may want to conduct a survey or some other quantitative analysis of each case within the multiple-cases to reveal and explore subunits within those cases.

 

V.             Modest Advice in Selecting Case Study Designs

a.     Multiple case studies are preferred over single case studies whenever possible.  While multiple-case studies require more time and resources, the analytic conclusions “will be more powerful than those coming from a single-case” (p. 61).  Even if you can only do a “two-case” case study that is preferable to a single case study so that should be your minimum goal.

b.     While too much modification can take you off track from your research focus decreasing the quality of your case study some modification is OK and even sometimes unavoidable when presented with the information from data collection process.  “The point is that the needed adaptiveness should not lessen the rigor with which case study procedures are followed” (p. 63).  

c.     There are times when case study may take on the form of mixed methods, but the two do not rely on each other—mixed methods do not have to include case study and case study does not need to occur in a mixed method format.  The two examples of given of these intersecting are case study within a survey and a survey within a case study. 

How does the "current" mentorship model impact the development of the teacher candidate and cooperating teacher agency and efficacy?
2. Prepositions:

Theoretical Lens: The concepts of efficacy and agency through Social Cognitive Theory (Bandura) will guide the exploration of this research question. 

An effective mentorship model could support the development of efficacy and agency for both teacher candidate and cooperating teacher.
3. Case:
Single-study case and case we are investigating is the co-teaching model

Next Steps:
4. Data Collection
  We will be conducting a 30-minute semi-structured interview protocol.Please take a look at the first round of data collection tab for more information
5. Linking data to propositions:

Once the first round of data collection is completed we will link data to our prepositions
6. Criteria for interpreting the strength of a case study’s findings