Single-Case Study — Single-Case Study Design
- Introduced by
- Robert K. Yin; Robert E. Stake
- Year
- 1984 (Yin's seminal protocol); 1995 (Stake's art-of-case-study framework)
- Type
- Qualitative research method
- Family
- Case Study
What does it do?
A single-case study is a qualitative research design that investigates one bounded instance — an organization, program, event, individual, or community — in its real-world context through multiple converging sources of evidence.
Developed into a rigorous social-science method chiefly by Robert Yin and Robert Stake, it is especially powerful when the case is unique, extreme, critical, or revelatory, and when the research question begins with 'how' or 'why' rather than 'how many.'
The intuition behind the method
Think of a journalist who spends six months inside a single hospital ward to understand how nurses make life-and-death decisions under resource constraints. No survey average could capture the institutional rules, informal workarounds, team dynamics, and individual moral reasoning that shape those decisions in that specific ward at that specific time. A single-case study works the same way: the researcher immerses in one carefully chosen case, gathers evidence from every available angle — interviews, documents, observations — and then builds a coherent account of how and why things happen the way they do. The goal is not to generalize statistically to a population but to generalize analytically to theory: to show that this case confirms, challenges, or extends what we thought we knew.
History
Case study research has been practised in clinical medicine, law, and social work for well over a century, but it lacked a consolidated social-science methodology until Robert Yin published the first edition of Case Study Research: Design and Methods in 1984, providing a systematic protocol-driven framework that treated the case study as a rigorous empirical method rather than an exploratory fallback. In parallel, Robert Stake developed an interpretivist perspective — articulated in The Art of Case Study Research (1995) — that emphasized the intrinsic value of understanding the particular case on its own terms rather than for what it might generalize to. These two lineages, Yin's more positivist and Stake's more constructivist, continue to shape how single-case studies are designed and evaluated today.
When to use it
Choose a single-case study when you have a 'how' or 'why' question about a contemporary phenomenon in its real-world context that you cannot manipulate experimentally. It is the right design when the case is unique, extreme, critical, or revelatory — when the specific instance itself is the intellectual contribution. It suits research where you need to understand context-dependent causal mechanisms, trace organizational or policy processes in depth, or test theoretical propositions against real-world complexity. Do not use it when the research goal is statistical generalization, when you need population-level prevalence estimates, or when the phenomenon can be isolated from context in a controlled experiment.
Limitations and considerations
- Statistical generalization to a population is not possible; a single case cannot support probability-based inference about how common a phenomenon is.
- Dependent on access — if the case site or key informants become unavailable, the study cannot be completed or replicated in the same way.
- The researcher's prolonged involvement with one site creates risks of bias, over-identification with participants, and selective attention to confirming evidence.
- Time and resource intensive: thorough single-case research typically involves months of fieldwork and hundreds of pages of raw data.
Source conditions do not replace independent scientific review or validation by an analysis engine.
Strengths
- Preserves the holistic complexity of a real-world phenomenon — context, process, and outcome are studied together rather than isolated.
- Particularly powerful for exploratory and explanatory questions ('how' and 'why') that surveys and experiments cannot answer.
- Multiple converging data sources strengthen the credibility of findings beyond what any single source could provide.
- Enables in-depth analysis of unique, rare, or revelatory cases that would be lost in cross-case aggregation.
- Findings generalize analytically to theory — a single critical or extreme case can confirm or disconfirm a theoretical proposition as decisively as a larger sample can test a hypothesis.
Applications
- Investigating how a single organization implements a new policy or technology — tracing the process, the obstacles, and the adaptations over time.
- Studying an unusual clinical case that challenges existing diagnostic categories or treatment protocols in medicine or psychology.
- Examining the development of an innovative educational program at one school to understand why it succeeded or failed in that specific context.
- Analysing a landmark legal case or policy decision to explain the institutional, political, and social factors that shaped its outcome.
- Exploring how a community responds to a sudden crisis — a natural disaster, a plant closure — by drawing on documents, interviews, and observation collected within that community.
- Evaluating a pilot project or social intervention at a single site before deciding whether to scale it to other contexts.
How it works
How it works
The researcher articulates a focused 'how' or 'why' research question and justifies the selection of a single case on one of five grounds: (1) the case is critical — it tests a well-developed theory; (2) it is unusual or extreme — it represents a condition rare enough to be studied on its own merits; (3) it is common — it exemplifies a frequently occurring situation that has not been examined in depth; (4) it is revelatory — the researcher has access to a site previously inaccessible to inquiry; or (5) it is longitudinal — the same case is studied at two or more points in time. The unit of analysis (the 'case') and its boundaries must be explicitly stated before data collection begins.
A formal case study protocol is prepared: it specifies the research questions, the propositions or conceptual framework guiding data collection, the data sources to be used, the logic linking data to propositions, and the criteria for interpreting findings. Yin stresses that the protocol serves the same methodological function as a structured instrument in quantitative research — it makes the study replicable and transparent. The design also decides whether the case will be treated holistically (a single unit of analysis) or embedded (multiple sub-units analysed within the case).
Evidence is gathered from multiple sources simultaneously. The six canonical sources identified by Yin are: documents (reports, memos, policies), archival records (statistics, organizational charts), interviews (open-ended or focused), direct observation (field visits), participant observation (the researcher takes an active role), and physical artifacts (tools, technology, art objects). Using at least three sources is strongly recommended; relying on a single source — for instance, interviews alone — undermines the evidential strength of the case.
Triangulation is the deliberate cross-checking of evidence across sources, methods, investigators, or theories to confirm or disconfirm emerging findings. Convergent evidence from independent sources strengthens the credibility of a finding; divergent evidence signals that the phenomenon is more complex than any single source reveals and prompts deeper analysis. Triangulation replaces the statistical logic of replication in quantitative research: rather than repeating a measurement, the single-case researcher converges on the same fact from multiple directions.
Analysis follows one of four main strategies: (1) pattern-matching — comparing observed patterns with theoretically predicted ones; (2) explanation-building — iteratively constructing a causal account of the case; (3) time-series analysis — tracing how conditions change over time within the case; (4) cross-case synthesis — used when the single case is compared with existing cases in the literature. In all strategies, a chain of evidence must be maintained so that an external reader can trace the path from raw data through to conclusions without gaps.
The case study report tells the story of the case in a form appropriate to the intended audience — a linear-analytic structure (background → findings → conclusions) for academic audiences, or a narrative vignette structure for practitioner audiences. The report must explicitly address each research question, present the evidence that supports each conclusion, acknowledge rival explanations that were considered and ruled out, and discuss the scope and limits of any theoretical generalization claimed.
Common pitfalls
- Selecting the case for convenience rather than for a principled reason — a case chosen solely because access is easy may not be critical, extreme, revelatory, or theoretically informative, and the justification for a single-case design collapses.
- Relying on a single data source — interviews alone, or documents alone, cannot support triangulation; without convergent evidence the findings rest on one type of account that may be partial or biased.
- Neglecting rival explanations — a credible case study must consider and explicitly rule out alternative accounts of the findings; omitting this step leaves the conclusions open to easy challenge.
- Conflating the case with the unit of analysis — the 'case' is the bounded context (the school, the project, the organisation) while the 'unit of analysis' is what is actually being studied within it (teachers' decision-making, a team's workflow); confusing them leads to unfocused data collection and incoherent analysis.
- Overgeneralizing — claiming that what was found in one case is universally true without limiting the theoretical claim to the conditions represented by that case.
Frequently asked questions
What makes a single-case study different from a multiple-case study?
In a single-case study, one bounded instance is investigated in depth; in a multiple-case study, the same protocol is applied to two or more cases and findings are compared across them. Multiple cases provide stronger grounds for theoretical generalization because convergent findings across independent cases are more convincing than findings from one case alone. However, a single case is justified when it is genuinely unique, extreme, critical, or revelatory — conditions where replication across cases is neither possible nor the point.
Is a single-case study the same as a qualitative case study?
Not exactly. Case studies can use qualitative data, quantitative data, or both. What defines the method is the focus on a bounded instance in its real-world context, not the type of data collected. A single-case study that traces an organization's quarterly performance over five years using archival records is still a case study even though the data are numeric. That said, most published single-case studies are primarily qualitative because the 'how' and 'why' questions they address call for interpretive rather than statistical analysis.
How do I establish rigor in a single-case study?
Yin's framework maps four quality criteria onto the design: construct validity (using multiple data sources and having key informants review draft reports), internal validity (applying pattern-matching or explanation-building to rule out rival interpretations), external validity (articulating the theoretical domain to which findings generalize and why), and reliability (maintaining a case study protocol and a case study database so another researcher could repeat the data collection and reach the same findings).
Can I generalize from a single case?
Not statistically — but analytically, yes. A single well-chosen critical case can confirm or disconfirm a theoretical proposition in the same way that a single experiment can. Yin calls this 'analytic generalization': the researcher generalizes findings to theory, not to a population. The claim is not 'this is how all hospitals work' but 'this case provides evidence that challenges / supports theory X under conditions Y and Z.'
When should I choose a case study over ethnography or phenomenology?
Choose a case study when the primary goal is to understand a bounded real-world instance — an event, program, decision, or organization — and you want to explain how and why it unfolded as it did using multiple data sources. Choose ethnography when the goal is to describe the shared culture, meanings, and practices of a group over a prolonged period of immersion. Choose phenomenology when the goal is to reveal the essential lived-experience structure of a phenomenon as reported by participants. Case study is the most flexible of the three: it imposes fewer restrictions on data type and analytic tradition.
Related methods
Sources
Yin, R. K. (2018). Case Study Research and Applications: Design and Methods (6th ed.). Sage.
ISBN: 978-1506336169Copied.Stake, R. E. (1995). The Art of Case Study Research. Sage.
ISBN: 978-0803957671Copied.