grad-case-study

Original：🇺🇸 English

Translated

Apply case study research design (Yin) to investigate how and why questions within real-life contexts using single or multiple case designs and triangulation. Use this skill when the user needs to choose between single and multiple case designs, apply holistic versus embedded units of analysis, triangulate evidence from multiple sources, or when they ask 'how should I structure a case study', 'single or multiple cases', or 'how do I ensure validity in case research'.

3installs

Sourceasgard-ai-platform/skills

Added on2026-04-15

NPX Install

npx skill4agent add asgard-ai-platform/skills grad-case-study

SKILL.md Content

View Translation Comparison →

Case Study Research (Yin)

Overview

Case study research is an empirical inquiry that investigates a contemporary phenomenon in depth and within its real-life context, especially when the boundaries between phenomenon and context are not clearly evident. Yin's framework provides systematic design choices — single vs. multiple cases, holistic vs. embedded analysis — and emphasizes triangulation to strengthen construct validity.

When to Use

Research questions are "how" or "why" questions about contemporary events
The researcher has little or no control over behavioral events
Contextual conditions are highly pertinent to the phenomenon of study
Boundaries between phenomenon and context are blurred

When NOT to Use

When frequency or incidence data is needed (use survey or experiment)
When context is irrelevant and can be controlled (use experiment)
When the goal is statistical generalization to a population

Assumptions

IRON LAW: Case study answers HOW and WHY questions in context — if you
need frequency or incidence data, use a survey or experiment instead.
Applying case study to "how many" or "how much" questions misuses the
methodology.

Key assumptions:

The case is a bounded system — define temporal, spatial, and conceptual boundaries
Multiple sources of evidence are essential for construct validity
Multiple cases follow replication logic (not sampling logic) — each case is an experiment, not a survey respondent
A case study protocol and database enhance reliability

Methodology

Step 1: Design the Case Study

Define the research question (how/why). Select the case type using Yin's 2x2 matrix:

	Single Case	Multiple Case
Holistic (single unit)	Critical, unique, or revelatory case	Literal or theoretical replication
Embedded (multiple units)	Multiple units within one case	Multiple units across cases

Develop propositions to guide data collection.

Step 2: Collect Evidence from Multiple Sources

Gather data from at least three of six source types: documents, archival records, interviews, direct observation, participant observation, physical artifacts. Maintain a chain of evidence linking questions to data to conclusions.

Step 3: Apply Triangulation

Triangulation Type	Description
Data	Multiple data sources converge on the same finding
Investigator	Multiple researchers independently analyze the same data
Theory	Multiple theoretical perspectives applied to the same data
Methodological	Multiple methods (qual + quant) address the same question

Step 4: Analyze and Report

Use pattern matching, explanation building, time-series analysis, or cross-case synthesis. Report the chain of evidence transparently.

Output Format

markdown

## Case Study Analysis: [Context]

### Research Question
- Question: [the how/why question]
- Case type: [single/multiple] x [holistic/embedded]
- Unit of analysis: [what constitutes the "case"]

### Case Selection Rationale
| Case | Rationale | Expected Pattern |
|------|-----------|-----------------|
| [name] | [why selected] | [literal/theoretical replication] |

### Evidence Matrix
| Source Type | Data Collected | Key Findings |
|------------|---------------|--------------|
| [documents/interviews/etc.] | [description] | [finding] |

### Triangulation Results
- Convergent findings: [where sources agree]
- Divergent findings: [where sources disagree]
- Explanation: [how divergence is resolved]

### Pattern Matching
- Predicted pattern: [from propositions]
- Observed pattern: [from evidence]
- Match assessment: [strong/moderate/weak]

### Conclusions
1. [Key finding with chain of evidence]
2. [Analytical generalization — how findings extend theory]

Gotchas

Case studies generalize to THEORY (analytical generalization), not to populations (statistical generalization)
A single-case design requires explicit justification: critical, extreme, unique, revelatory, or longitudinal
Replication logic in multiple cases means each case must independently confirm or disconfirm a proposition
The chain of evidence must be traceable from question to protocol to database to report
Do NOT confuse case study with case history or case report — Yin's case study is a research strategy with formal design
Rival explanations must be addressed explicitly, not just dismissed

References

Yin, R. K. (2018). Case Study Research and Applications: Design and Methods (6th ed.). Sage.
Stake, R. E. (1995). The Art of Case Study Research. Sage.
Eisenhardt, K. M. (1989). Building theories from case study research. Academy of Management Review, 14(4), 532-550.