03 · What You Need to Know
How Single-Case Studies and Single-Case Experiments Differ Methodologically
What is a single-case study?
A single-case study is an intensive investigation of one bounded case, such as an individual, classroom, program, organization, or institution.
The researcher examines the case in relation to a particular phenomenon and its context. Evidence may come from interviews, observations, documents, records, assessments, or other relevant sources.
For example, a researcher might investigate how a university implemented an individualized support program for one student with complex learning needs.
The investigation could examine institutional decisions, teacher experiences, accessibility arrangements, and the student's participation.
Case study methodology is not necessarily qualitative. Quantitative records and mixed evidence may also contribute to understanding the case.
Importantly, the case study does not become an experiment simply because an intervention occurred during the period being investigated.
What is a single-case experimental design?
A single-case experimental design, commonly abbreviated SCED, uses systematic manipulation of an intervention and repeated outcome measurement to evaluate a functional relationship between the intervention and the outcome.
The researcher establishes planned experimental conditions, often including baseline and intervention phases, and examines whether outcome patterns change in correspondence with intervention manipulation.
For example, researchers may repeatedly measure independent task completion before introducing a prompting intervention and then examine performance under the intervention.
Stronger experimental arrangements incorporate additional contrasts, such as staggered intervention introduction across participants or repeated intervention withdrawal and reintroduction where appropriate.
Single-case experimental research is also frequently called single-subject experimental research, although its experimental units are not always individual people.
How do their main characteristics compare?
| Characteristic |
Single-case study |
Single-case experimental design |
| Primary purpose |
Understand a bounded case or phenomenon. |
Evaluate the effect of a manipulated intervention. |
| Research question |
Often asks how or why something occurs. |
Often asks whether an intervention changes a measurable outcome. |
| Intervention manipulation |
Not required. |
Essential to the experimental logic. |
| Repeated outcome measurement |
Optional, depending on the question. |
Fundamental. |
| Experimental phases |
Not inherently required. |
Planned conditions or intervention changes are required. |
| Evidence |
May include interviews, documents, observations, and numerical data. |
Emphasizes repeated outcome observations and planned experimental contrasts. |
| Typical inference |
Contextual description, interpretation, or explanation. |
Evidence about a functional intervention-outcome relationship. |
| Broader applicability |
Depends on analytic generalization and contextual relevance. |
Depends on experimental replication and relevance across cases or settings. |
The approaches are not arranged in a simple hierarchy of research quality. Their appropriateness depends on the question being asked.
Does the word single mean that only one participant is allowed?
No. In case study methodology, a single-case design ordinarily means that one bounded case is the principal unit of analysis. That case may contain multiple people.
For example, a university can be one institutional case even when dozens of faculty members and students contribute evidence.
In single-case experimental research, an investigation may include several participants, each measured repeatedly across experimental conditions.
Multiple participants are often useful because they provide opportunities to replicate an intervention effect.
Consequently, counting participants is not a reliable way to distinguish the two approaches.
How do their research questions differ?
A case study might ask, "How did teachers and students experience the implementation of an individualized digital learning program?"
This question requires understanding experiences, institutional arrangements, and contextual processes.
A single-case experiment might ask, "Does introducing a structured prompting intervention increase independently completed learning tasks?"
This question concerns an observable intervention effect.
The distinction is not that one approach studies people while the other studies numbers. Rather, each approach is organized around a different inferential objective.
A contextual explanation may be rigorous without an experimental manipulation. Likewise, a well-controlled experiment may estimate an intervention effect without explaining every aspect of the participant's experience.
Why is intervention manipulation the defining distinction?
In a case study, the researcher may investigate an intervention that an institution or practitioner has already implemented.
The researcher does not necessarily determine when the intervention begins, which conditions participants experience, or how the intervention changes over time.
In a single-case experiment, those changes form part of the research design.
For example, the researcher may stagger intervention introduction across participants or systematically alternate conditions according to a planned schedule.
This manipulation creates opportunities to compare outcomes under different experimental conditions.
However, manipulation alone is not sufficient. The conditions must provide credible evidence for distinguishing intervention effects from maturation, historical events, and other alternative explanations.
Does collecting repeated measurements make a case study experimental?
No. Repeated measurement provides information about how an outcome changes, but it does not automatically identify what caused that change.
Suppose a researcher records a student's weekly engagement for twelve weeks while documenting a new instructional program.
The observations may reveal a useful trajectory. However, if the program begins at the same time as other changes and the design contains no credible experimental contrast, the observed improvement cannot confidently be attributed to the program.
Repeated measurement is necessary for many single-case experimental designs, but it is not sufficient for experimental control.
The distinction becomes clearer when considering the role of repeated measurement in single-case experimental research.
Is a simple AB design a genuine experiment?
An AB design measures an outcome during a baseline phase, A, and an intervention phase, B.
It provides a structured description of change associated with intervention introduction.
However, a simple AB design ordinarily offers limited evidence of experimental control because the intervention is introduced only once.
A coincident event, an existing trend, or measurement reactivity may explain the change.
Researchers should therefore distinguish a basic AB phase comparison from stronger single-case experimental arrangements involving additional demonstrations of effect.
More detailed distinctions among AB, ABA, and ABAB arrangements are relevant when selecting the experimental structure.
How do multiple-baseline designs establish stronger experimental contrasts?
A multiple-baseline design introduces an intervention at different times across participants, behaviors, or settings.
Repeated outcomes are observed before and after intervention introduction.
If outcomes change systematically following the staggered introductions rather than simultaneously across all tiers, the pattern may strengthen evidence of a functional relationship.
For example, three students may begin an intervention at different prespecified sessions while baseline measurements continue for those who have not yet received it.
This structure provides additional opportunities to distinguish the intervention from common time-related influences.
Nevertheless, its strength depends on sufficiently informative baselines, appropriate tier independence, intervention fidelity, and replication.
The design is discussed further in when multiple-baseline designs are preferable.
Can a case study make causal claims?
Case studies can contribute to causal explanation, particularly through process tracing, temporal evidence, and the examination of competing explanations.
However, their causal logic differs from the experimental manipulation used in single-case experimental designs.
For example, a detailed institutional case study may examine whether particular decisions contributed to the failure of an educational initiative.
Documents, interviews, and chronological evidence may support a plausible causal account.
That does not mean the researcher experimentally demonstrated the effect of one decision by manipulating it across controlled conditions.
Researchers should therefore avoid claiming that case studies can never address causation. The more accurate distinction concerns the evidence and identification strategy supporting the causal claim.
Does a single-case experiment establish causation automatically?
No. A study described as experimental can still have weak experimental control.
For example, an intervention introduced after an unstable baseline may coincide with an existing improvement trend.
Insufficient phase observations, inconsistent intervention implementation, or changes in outcome measurement may also undermine inference.
Researchers should evaluate whether the design provides adequate demonstrations of effect and whether alternative explanations remain plausible.
Relevant methodological standards, including those used by the What Works Clearinghouse, distinguish design adequacy from the strength of the observed evidence.
The label experimental is not a substitute for demonstrating experimental control.
How do the two approaches use qualitative evidence?
Qualitative evidence is often central to case study research because the objective may concern experience, meaning, implementation, or contextual processes.
Interviews and documents can help explain why participants acted as they did and how institutional conditions shaped events.
Single-case experiments may also use qualitative information to understand acceptability, implementation barriers, or participant experiences.
However, the primary experimental demonstration generally depends on systematically collected outcome measurements and planned condition changes.
Qualitative information can enrich interpretation without replacing the evidence required for experimental control.
How do their approaches to generalization differ?
Case study research may support analytic generalization by connecting findings to theoretical propositions or explanatory mechanisms.
For example, a carefully investigated university case may illuminate how institutional support affects the implementation of an accessibility policy.
Such generalization depends on the argument and evidence, not on assuming that one institution statistically represents all institutions.
Single-case experimental research may strengthen external validity through replication across participants, behaviors, settings, and circumstances.
However, demonstrating an intervention effect in one case does not automatically establish a population-average effect.
Both approaches require researchers to specify the scope of their conclusions and avoid unsupported extrapolation.
Can both approaches be combined?
Yes, when the integration serves a clear research objective.
For example, a researcher may conduct a multiple-baseline experiment to evaluate whether an intervention changes student performance and use interviews to investigate how participants experience its implementation.
The experimental component addresses the intervention-outcome relationship. The contextual component may explain acceptability, implementation conditions, or perceived mechanisms.
However, the resulting investigation should identify each methodological component and explain how the evidence is integrated.
Adding interviews does not transform weak experimental control into strong causal evidence, just as adding a numerical outcome does not automatically make a case study experimental.
Which approach should researchers choose?
Choose a case study when the primary question concerns understanding a bounded case in context and the evidence required is descriptive, interpretive, or explanatory.
Choose a single-case experimental design when the primary question concerns the effect of a manipulable intervention and repeated outcome measurement can support credible experimental contrasts.
For example, investigating why an individualized support program succeeded in one university may call for a case study. Testing whether a specific prompting procedure changes task completion may call for a single-case experiment.
When both questions matter, an integrated design may be justified.
The decision should follow the inferential objective, not the number of participants or the perceived prestige of the word experimental.
