Chapter 9 - Case Study research
Multiple case study - Literal Replication
Used to anticipate similar findings Simplest design, with two cases and predicting similar results between them
Case study research involves
Used to explore events, construct theories, develop models, create treatments, and understand the behavior of individuals. Answers the "how" and "why" a phenomenon exists Observe rather than manipulate
Multiple case study - Theoretical Replication
Used to predict contrasting results. More complexity in the study
Case study research is qualitative or quantitative?
Uses multiple sources of data observation AND interview info = mixed method
Common data sources - Participant observations
researcher may play a role in the context in addition to investigator (counselor, teacher)
Steps for designing and conducting case study research 5. Report the Findings
A. Know the target audience - who is the report geared towards? B. Determine format and structure C. Have others do a final review of the report
Steps for designing and conducting case study research 3. Prepare and Collect the data
A. Maintaining flexibility of a study B. Develop a case protocol to address validity and reliability C. Train investigators in proper data collection protocols D. Properly collect and store data E. Conduct a pilot study F. Gain approval of study G. Collect final data
Yin (2009) suggests three principles of data collection to maximize the use of data:
A. Triangulation - use variety of sources to corroborate and validate findings B. Establish a case study database C. Create a chain of evidence
Steps for designing and conducting case study research Identify the Case(s)
Accessibility and appropriateness Unique versus common cases Single versus multiple cases
External validity
allows for replication of findings to additional cases
Types of case studies
Context-oriented (Stake, 1995) - Intrinsic - Instrumental - Collective Purpose-Oriented (Yin, 2009) -Exploratory - Descriptive - Explanatory
Analytical Strategies
- Start with data most relevant to the research question/propositions -Use a case description to create a structure of the study -Use a mixed methods approach -Explore rival explanations; look for contradictory evidence
Common data sources for case study research
1. Direct observations 2. Participant-observation 3. Interviews 4. Documentation 5. Archival records 6. Physical artifacts
Steps for designing and conducting case study research
1. Establish research questions and propositions 2. Identify the case(s) 3. Prepare and collect the data 4. Evaluate and analyze the data 5. Report the findings (Final review)
Steps for designing and conducting case study research Identify the Case(s) - Single vs. Multiple cases
A single case takes place within a single context(setting) regardless of #of participants. Multiple cases include a different context for each case. Used when replication is necessary to create a more robust study
Steps for designing and conducting case study research 1. Types of Research Questions
1. Outcome questions show extent that treatment was effective or how client changed 2. Theory building q's help create/validate theoretical 3. Pragmatic q's for practical info of a case, procedures, techniques, guidelines 4. Experiential or narrative q's explore the clients experiences through their own perspectives
Specific Analytic Techniques
1. Pattern matching - compare observed and predicted patterns to determine where patterns overlap to improve internal validity 2. Explanation building - explain causal relationships; for explanatory case studies 3. Time-series analysis - Track changes over time and compare expected patterns to actual ones 4. Logic Models - measure complex chain of events where one stage leads to the next; causal relationships and patterns are observed over sequence of events to determine if they match predicted patterns 5. Cross-case synthesis - looks across cases to identify patterns
The four types of designs for case study research
1. Single-case holistic design - single unit within a single case 2. Single-case embedded design - One context, multiple units 3. Multiple-case holistic design - singular unit within each of the cases. 4. Multiple-case embedded design - several units within multiple cases
Types of case studies: Context-Oriented Collective case study
A comparison between multiple cases to identify similarities and differences between them
Types of case studies: Purpose-Oriented Descriptive
A description of an issue within its current context
Types of case studies: Purpose-Oriented Explanatory
An explanation of an issue in more detail within its natural context; helps to clarify causal links
Types of case studies: Purpose-Oriented Exploratory
An exploration of an issue in depth within an actual context
Formats; Question and Answer
Asks and answers the questions for each case in the study. Multiple cases each case is done separately for each case using same format of q's
Steps for designing and conducting case study research 4. Evaluate and Analyze Data
Begin by evaluating your data using a broad analysis strategy and then selecting an appropriate analysis technique.
Structure of report; Theory Building
Build logic for creating a theory
Structure of report; Suspense
Conclusions presented first with the processes to follow
Types of case studies: Context-Oriented Instrumental case study
Emphasized the specific issue rather than the individual case
Types of case studies: Context-Oriented Intrinsic case study
Focus is on the unique case rather than the broader issue
Format; Cross-Case
For multiple cases, includes only the compiled cross-case data
Steps for designing and conducting case study research 1. Research Questions and Propositions
Four types of questions will help in case study design - Outcome questions - Theory building - Pragmatic q's - Experiential or narrative q's Propositions help identify what specific details you might study to answer the questions (Yin 2009)
Structure of report; Unsequenced
No particular order as long as all parts are included
Formats; Single Case Narrative
Provides a single description of the case study
Structure of report; Comparative
Repeat the same case study with differing perspectives
Structure of report; Chronological
Report the events of the case in sequential order
Formats; Multiple Case Narrative
Single narrative section/description of each case study and a final section of cross case findings of all cases
Steps for designing and conducting case study research Identify the Case(s) - Unique vs. common cases
Sometimes an average case may not provide the most useful info, but rather a unique case may provide more clarity towards the phenomenon in question
Holistic approach - for single or multiple cases
Study of a singular unit (participant/group/entity) as the selected case. One participant, one case
Embedded approach - for single or multiple cases
Study of several units (participants, groups, or entities) with the intention to evaluate them as part of a larger case within a single context. Many participants, one context. The intention is to study them as part of a larger case rather than individual cases.
Structure of report; Linear Analytic
Typically used in journal article writing; linear sequence of topics (e.g. literature review, methodology, findings)
Steps for designing and conducting case study research Identify the Case(s) - Accessibility and appropriateness
Will the case provide data that can help to answer the research question(s)? Will you have access to the case to collect enough data? Is the case appropriate for the study so that relevant data can be collected?
Steps for designing and conducting case study research Identify the Case(s) - Questions to ask yourself
Will the case provide data that can help to answer the research question(s)? •Will you have access to the case to collect enough data? •Is the case appropriate for the study so that relevant data can be collected? •Will your case represent a unique or common situation? •Will you use a single case or multiple cases? •Will you use a holistic or embedded approach within your design?
Internal validity
a form of validity for explanatory case studies where inferences can be made about causal relationships
Common data sources - Interviews
direct feedback from the participants and evidence to corroborate other data sources
Construct validity
ensures that the study measures what it is designed to measure
Case studies are valuable to
study a current phenomenon relating to participants within their current real life context.
Common data sources - Physical artifacts
tangible evidence relevant to the case such as cultural objects, artwork, photos
Pilot Study
test run for observing or interacting with participants to identity potential obstacles and allow researcher to practice
Common data sources - Archival records
those not readily available as documents (treatment records, previous collected survey data)
Common data sources - Direct observations
viewing case participants within their natural contexts (home, classroom, office)
Reliability
when the study can be replicated and similar results obtained
Common data sources - Documentation
written evidence to corroborate or complement other data (letters, email, calendars, case notes)
Limitations of Case Study Research
• Difficult to provide rigor • May not be seen as valuable as other research designs • Generalizability
Steps for designing and conducting case study research 5. Report the Findings; Structure
•Linear Analytic •Comparative •Chronological •Theory Building •"Suspense" •Unsequenced
Steps for designing and conducting case study research 5. Report the Findings; Format
•Single case narrative •Multiple case narrative •Question and answer format •Cross case