Robert K. Yin’s case study research methodology has introduced case study as a systematic approach to social science inquiry.
Curated by Business Science Daily — peer-reviewed sources, human-verified.
Learn more
About Our Curation Process
Business Science Daily curates academic research in business and management. Each featured study is selected from reputable, peer-reviewed journals, institutional repositories, or working papers (e.g., Elsevier, Sage, NBER, SSRN).
Articles are carefully summarized to ensure clarity and accuracy, with direct citations or links to original sources. Our process emphasizes transparency, academic integrity, and accessibility for a broader audience.
Learn more in our Editorial Standards & AI Policy.
The sixth edition of his book, Case Study Research and Applications, provides a complete framework for designing and conducting case studies. Yin addresses the method’s core challenges: defining the case, selecting cases (through replication rather than sampling logic), collecting evidence from multiple sources, and analyzing data through techniques like pattern matching and explanation building.
Four quality tests including construct validity, internal validity, external validity, and reliability, provide concrete standards.
| Question | Answer |
|---|---|
| What is case study research methodology? | Case study research methodology is an empirical method that investigates a contemporary phenomenon in depth and within its real-world context, especially when boundaries between the phenomenon and context are unclear. It relies on theoretical propositions to guide design and multiple sources of evidence that converge through triangulation. |
| When should you use a case study research design? | Use a case study when your research questions are “how” or “why” questions, when you have little or no control over behavioural events, and when your focus is on a contemporary (not entirely historical) phenomenon. |
| Can you generalise from a single case study? | Yes, but through analytic generalisation to theory, not statistical generalisation to populations. Case studies, like experiments, are generalisable to theoretical propositions, not to populations or universes. |
| How many cases do you need for a case study? | The number depends on your design. Single-case studies can be valid for critical, unusual, common, revelatory, or longitudinal cases. Multiple-case studies should follow replication logic rather than sampling logic. Even two cases can be stronger than one. |
| What are the four tests for case study quality? | Construct validity (correct operational measures), internal validity (causal relationships, for explanatory studies), external validity (generalisability), and reliability (repeatability). Each has specific tactics like using multiple sources of evidence, pattern matching, and maintaining a chain of evidence. |
| What is the difference between case study and qualitative research? | The relationship is debated and largely unresolved. Case study research can include quantitative evidence and follows its own logic of design. Some see it as a type of qualitative research; others see it as a separate method with its own procedures. |
| What are the main sources of evidence in case study research? | The six main sources are documentation, archival records, interviews, direct observations, participant-observation, and physical artifacts. Good case studies use multiple sources to triangulate findings. |
One case cannot possibly represent a population. Therefore, one case cannot produce knowledge worth trusting. Therefore, case studies are at best exploratory and at worst anecdotal.
The problem, as Robert K. Yin began arguing in the early 1980s and has now refined across six editions, was that the entire premise was wrong. Case studies do not seek to represent populations. They seek to build theory.
Yin argues that what makes case study hard is the same thing that makes it essential: it investigates contemporary phenomena in depth and within their real-world contexts, especially when the boundaries between phenomenon and context are unclear. In other words, a case study is what you do when the phenomenon cannot be separated from its environment.
An experiment is designed to isolate a phenomenon from its context. That is its strength and its limitation. The researcher controls the environment precisely so that only the manipulated variables can explain the outcome. The problem is that the real world is not a laboratory. Organizations do not operate in isolation. Policies do not unfold in controlled conditions. Decisions do not occur in vacuums.
The Replication Fallacy
Consider a frequent criticism of case study research: you cannot generalize from a single case.
Yin’s response to how can the researcher generalize from a single experiment is that he does not. You generalize from a series of experiments, each of which replicates findings under different conditions, he says. Generalization in experimental research is never about a single study. It is about the accumulation of evidence across multiple studies, each building on the last.
Case studies should work the same way. The logic is replication, not sampling.
The design of multiple-case studies follows an analogous logic. Each case must be carefully selected so that the individual case studies either predict similar results (a literal replication) or predict contrasting results but for anticipatable reasons (a theoretical replication).
This is a profound shift. It rejects the sampling logic that underpins survey research and instead embraces the replication logic that underpins experiments. The goal is not to select a representative sample of cases. The goal is to select cases that will test and extend the theoretical framework.
The implications are not merely methodological. They are epistemological. Case study research is not trying to do what surveys do. It is trying to do something different: to build, test, and refine theory.
If the criticism is that case studies cannot represent populations, the answer is that they were never meant to.
— Robert K. Yin (2018, p. 21)
A case study is an empirical method that investigates a contemporary phenomenon (the ‘case’) in depth and within its real-world context, especially when the boundaries between phenomenon and context may not be clearly evident.
A case study:
- Copes with the technically distinctive situation in which there will be many more variables of interest than data points;
- Benefits from the prior development of theoretical propositions to guide design, data collection, and analysis;
- Relies on multiple sources of evidence, with data needing to converge in a triangulating fashion.
Key insight: Yin’s deepest contribution was philosophical, not methodological. He understood that the hierarchy of methods was not a description of reality but a preference masquerading as one. The case study is not a method of last resort – it is the method of first resort when studying complex, contextualised phenomena.
Multiple-case studies should follow replication logic, not sampling logic. Cases are selected to predict similar results (literal replication) or contrasting results for anticipatable reasons (theoretical replication).
Single-case (holistic)
A single case studied as a whole. Appropriate for critical, unusual, common, revelatory, or longitudinal cases.
Single-case (embedded)
A single case with subunits of analysis. The case is studied holistically, but attention is also given to subunits.
Multiple-case (holistic)
Multiple cases studied holistically. Follows replication logic across cases.
Multiple-case (embedded)
Multiple cases, each with subunits. Allows for within-case and cross-case analysis.
Five rationales for single-case studies
- Critical case – tests a well-formulated theory.
- Extreme or unusual case – deviates from theoretical norms.
- Common case – captures everyday circumstances.
- Revelatory case – observes a phenomenon previously inaccessible.
- Longitudinal case – studies the same case at multiple points in time.
Tactics: Multiple sources of evidence, key informant review, chain of evidence.
Tactics: Pattern matching, explanation building, addressing rival explanations, logic models.
Tactics: Theory in single-case studies, replication logic in multiple-case studies.
Tactics: Case study protocol, case study database, chain of evidence.
Analytic generalisation is the logic whereby case study findings can apply to situations beyond the original case study, based on the relevance of similar theoretical concepts or principles.
Statistical generalisation involves making inferences about a population on the basis of empirical data collected from a sample – not relevant for case study research.
The goal is to expand and generalise theories, not to extrapolate probabilities.
Letters, emails, agendas, administrative documents, news clippings.
Census data, service records, organisational records, survey data produced by others.
Open-ended, focused conversations. Prolonged, shorter, or survey-type interviews.
Formal or casual observations of meetings, activities, workspaces, environments.
Active involvement in the case being studied – a special mode of observation.
Technological devices, tools, works of art, cultural artifacts.
- Principle 1: Use multiple sources of evidence (triangulation).
- Principle 2: Create a case study database.
- Principle 3: Maintain a chain of evidence.
- Principle 4: Exercise care when using data from social media sources.
Five analytic techniques
- Pattern matching – comparing an empirically based pattern with a predicted one.
- Explanation building – building an explanation about the case through iterative refinement.
- Time-series analysis – tracing changes over time.
- Logic models – stipulating a complex chain of cause-effect-cause-effect patterns.
- Cross-case synthesis – compiling data for multiple cases and observing patterns.
No formal procedures exist for rigorously testing rival explanations. Researchers currently exercise complete discretion over identifying rivals, seeking evidence, and determining when a rival has been ruled out.
Initial step: A 4-point scale indicating the degree of presence of rival considerations in a case study.
The holistic feature of the case being studied remains a core feature. The goal is to understand what the case is, how it works, and how it interacts with its real-world context. Yet variables remain important.
Challenge: How to maintain holistic orientation while still appreciating variables and creating meaningful typologies.
The entire issue of whether case study research is automatically subsumed under qualitative research, or whether it is a separate method, requires further explication.
- One perspective: Case study as one of five major types of qualitative research.
- Opposing perspective: Case study research follows its own customized research procedures, separate from qualitative research.
Exemplary case study characteristics
- Significant – unusual cases, nationally important issues.
- Complete – clear boundaries, exhaustive evidence collection, no artifactual conditions.
- Alternative perspectives – considers rival propositions.
- Sufficient evidence – enables independent reader judgment.
- Engaging composition – seduces the reader’s eye.
The Quality Framework
Yin has built a comprehensive methodological framework with explicit standards for quality. Four tests define the framework:
Construct validity asks whether the researcher has identified the correct operational measures for the concepts being studied.
Internal validity—relevant only to explanatory studies—asks whether the evidence supports a causal relationship rather than a spurious one.
External validity asks whether the findings can be generalised to other situations.
Reliability asks whether the study could be repeated with the same results.
Each test has specific tactics. Construct validity is strengthened by using multiple sources of evidence, by having key informants review draft reports, and by maintaining a clear chain of evidence. Internal validity is addressed through pattern matching, explanation building, and the explicit consideration of rival explanations. External validity is achieved through theoretical propositions in single-case studies and replication logic in multiple-case studies. Reliability requires a case study protocol and a formal database.
The framework is striking for its specificity. In a method often dismissed as subjective, Yin has provided objective standards. In a field often characterized by vague assurances of quality, he has provided clear, operationalizable tactics.
The Database Principle
In experimental research, the data are clearly separated from the analysis. Other researchers can inspect the raw data, run their own analyses, and test the original conclusions. The same is true of survey research, where the dataset can be shared independently of the final report.
Case study research has historically been different. The evidence tends to be narrative, embedded within the final report, inseparable from the author’s interpretation. A critical reader has no recourse: the data and the analysis are one and the same.
Yin insists on a different approach. Every case study should create a formal database, separate from the final report, containing all the evidence: field notes, documents, tabular materials, and narrative compilations. The final report draws from this database but does not replace it.
The principle serves multiple purposes. It increases reliability by allowing others to inspect the raw evidence. It forces the researcher to be systematic about data storage and retrieval. It prevents the common problem of case studies drifting into selective reporting. And it creates a foundation for genuine transparency.
Other related sources:
Research Handbook of Academic Mental Health Edwards et al. (2024)
The Craft of Research | Booth et al. (5th ed., 2024)
Research Design (6th Edition) — Creswell & Creswell
Writing for Social Scientists: H. S. Becker (Third Edition)
Research Methods Saunders, Lewis & Thornhill (9th ed.)