To write a case study, you investigate one or a few bounded cases in depth, in their real-life context, using several sources of evidence, and report what they reveal about your research question. It is a research design, not just a story.
The term is also used for teaching and clinical case reports, which follow simpler templates. This page focuses on the research case study used in master's and doctoral work, drawing mainly on the widely cited frameworks of Robert Yin and Robert Stake.
The Case Study in Five Decisions
- Question: usually "how" or "why" about a contemporary phenomenon.
- Design: single or multiple cases, holistic or embedded.
- Boundary: what is inside the case and what is context.
- Evidence: several sources, such as interviews, documents and observation.
- Analysis: a stated strategy, such as pattern matching or cross-case synthesis.
Choosing a Design
Yin classifies designs by the number of cases and the number of units of analysis in each.
| Design | Cases | Units of analysis | Suits |
|---|---|---|---|
| Single holistic | One | The case as a whole | Critical, unusual or revelatory cases |
| Single embedded | One | Several subunits | One organization with meaningful internal variation |
| Multiple holistic | Two or more | One per case | Comparing cases using replication logic |
| Multiple embedded | Two or more | Several per case | Within-case and cross-case comparison |
Yin justifies a single case when it is critical for testing a theory, unusual or extreme, common or typical, revelatory, or longitudinal. Multiple cases are chosen by replication logic: each is expected to produce similar results, or contrasting results for predictable reasons. They are not a statistical sample.
Stake offers a complementary view, distinguishing intrinsic cases studied for their own interest, instrumental cases studied to understand a wider issue, and collective studies of several cases.
Bounding the Case
Define clearly what belongs to the case. Boundaries are usually set by time, place, people or organization, and by the theoretical focus. A case of "nurse-led discharge on Ward 7 between January and June 2025" is far easier to study and defend than "discharge practice in the hospital".
Collecting and Triangulating Evidence
Case studies draw strength from converging evidence. Common sources include interviews, documents and records, direct observation, participant observation, archival data and physical artifacts.
- Use at least two or three sources so findings can be corroborated.
- Write a case study protocol listing questions, sources and procedures.
- Keep a case study database separate from your report, so others could trace your evidence.
- Maintain a chain of evidence from question to data to conclusion.
Analyzing Case Study Data
| Strategy | How it works |
|---|---|
| Pattern matching | Compares observed patterns with those predicted by theory or rival explanations |
| Explanation building | Develops a causal account iteratively as evidence is examined |
| Time-series analysis | Traces change across a defined period |
| Logic models | Tests a chain of expected events against what occurred |
| Cross-case synthesis | Compares findings across cases in multiple case designs |
Qualitative coding, such as thematic analysis, is often used within these strategies. Name your approach and show how you handled rival explanations.
Structuring the Written Case Study
| Section | Content |
|---|---|
| Introduction | Problem, question, purpose and significance |
| Literature and theory | Prior research and any propositions guiding the study |
| Methodology | Design, case selection, boundary, sources, protocol, analysis, ethics |
| Case description | Rich context so readers can judge transferability |
| Findings | Evidence organized by theme, proposition or case |
| Discussion | Interpretation, links to theory, rival explanations, limitations |
| Conclusion | Answer to the question and implications |
Multiple case reports often present each case separately before a cross-case chapter. Follow your program's template where one exists.
Writing It Step by Step
- Frame a "how" or "why" question and any initial propositions.
- Select and justify the case or cases.
- Draw the boundary and record it in your protocol.
- Collect evidence from several sources and log it in a database.
- Analyze using a named strategy and test rival explanations.
- Write a detailed case description, then the findings.
- Link findings back to theory and state the limits of generalization.
Example Excerpt
The case was bounded as the introduction of bedside shift handover on one 28-bed surgical ward over six months. Evidence came from 14 staff interviews, 40 hours of handover observation and ward incident records. Pattern matching tested the proposition, drawn from normalization process theory, that staff who helped design the handover script would adopt it more consistently. Observation and interviews converged: adoption was highest on the shifts led by design-group members, while incident records showed no change, suggesting that adoption alone did not alter reported events within the period.
The excerpt names the boundary, sources, theory and strategy, and shows how evidence converges and where it does not. It is illustrative only.
Generalizing from Cases
Case studies do not generalize statistically to populations. Yin describes analytic generalization, in which findings extend or test theory; qualitative researchers also speak of transferability, supported by thick description that lets readers judge fit with their own setting. State which you claim.
Quality Checklist
- Question suited to case study, usually how or why.
- Case selection justified and boundary stated.
- Several evidence sources, triangulated.
- Protocol and database kept; chain of evidence clear.
- Analysis strategy named and rival explanations considered.
- Construct validity, internal validity where relevant, external validity and reliability addressed, or qualitative equivalents.
- Ethics, consent and anonymity of the case handled.
How IvyDrafts Helps with Case Studies
Send your research question, details of the case, your program's template and any data or transcripts you have collected. A graduate methods writer returns a custom design section, an analysis of your own data, or a full case study for study and reference.
We never invent interviews, observations or organizational data. Free revisions within the scope of your original request are included.
Build a Case Study That Stands up to Review
Send your research question, case details and data or brief. Your writer returns a custom design section, analysis or full case study to study and adapt.
Order a Case StudyFree revisions · Full refund if late · Written from scratch for your order
Frequently Asked Questions
An in-depth study of one or a few bounded cases in their real-world context, using several sources of evidence to answer a how or why question.
One can be justified if it is critical, unusual, typical, revelatory or longitudinal. Several cases strengthen findings through replication.
Usually, but not always. Case studies can include quantitative data, and many use mixed evidence.
Not statistically. They generalize analytically to theory, or are judged for transferability through rich description.
A clinical case report describes one patient's presentation and care. A research case study is a designed investigation with a protocol, several sources and a stated analysis.
Robert Yin and Robert Stake are most widely cited, with Sharan Merriam also common in education research.