Case Study Research
Enter a case in depth. Follow decisions, practices, and events through their real-world setting, and build an account that connects what happened with the evidence explaining it.
What you will learn
- Define a case, its context, boundaries, and units of analysis.
- Distinguish intrinsic and instrumental purposes from exploratory, descriptive, and explanatory aims.
- Choose single-case or multiple-case designs with holistic or embedded units.
- Collect and analyse evidence within and across cases.
- Test interpretations, write case assertions, and explain the reach of the findings.
Define: Work through Sections 1–4 to identify the case and design.
Investigate: Use Sections 5–9 to plan the evidence and develop the analysis.
Assess and write: Apply Sections 10–12, then extend your reading in Section 13.
Begin with a case worth understanding
Case study research investigates a bounded case in depth, attending to the setting, relationships, and processes that give it meaning. A case might be a person, organisation, programme, decision, event, community, or implementation process.
The researcher defines what constitutes the case and explains why studying it can answer the question. A case can sit at an individual, organisational, or societal level. It can contain many participants and draw on qualitative, quantitative, or combined evidence.
Follow the relationship between phenomenon and setting
Consider a newsroom introducing a correction protocol. A policy document can tell you the formal procedure. To understand its use, investigate editing routines, deadlines, staff roles, technical permissions, and decisions about what readers should see. These connections make context part of the explanation.
Yin’s approach gives particular attention to contemporary phenomena in real-world context. Other case study traditions also examine historical cases or place different emphasis on interpretation, comparison, and theory. State the tradition guiding your design.
How did staff at Newsroom A translate a new correction protocol into everyday publishing decisions during its first twelve weeks?
This question invites investigation of an unfolding process. It can connect formal intentions, observed practices, staff accounts, and publication records.
Understand what the design contributes
A case study can illuminate a puzzling decision, reveal how a programme operates, develop a proposition, or refine an explanation. It can preserve sequences and relationships that disappear when evidence is detached from its setting.
Clarify what you want the study to accomplish
Purpose guides the choice of case, evidence, and interpretation. Distinguish your reason for studying a case from the kind of account you intend to produce.
Stake: intrinsic and instrumental interests
Intrinsic interest
The particular case deserves understanding in its own right. You might investigate how one community radio station sustained a service through a local disruption, attending closely to its history and relationships.
Instrumental interest
The case helps illuminate a broader issue. The newsroom correction study could investigate how formal accountability rules become everyday routines. One case or several cases may serve that purpose.
Stake emphasises the particularity and complexity of the case. A collective study extends the inquiry across several cases. These interests can overlap, and an intrinsic interest can concern an ordinary case whose circumstances warrant close attention.
Specify the analytical aim
On a narrow screen, scroll the table sideways.
| Aim | What it seeks | A newsroom question |
|---|---|---|
| Exploratory | Identify issues, concepts, or promising lines of inquiry. | What difficulties arise as staff begin using the protocol? |
| Descriptive | Develop a detailed, organised account of the case. | How are corrections identified, approved, published, and recorded? |
| Explanatory | Investigate processes and conditions that account for events. | Why do public correction labels appear promptly in some episodes and later in others? |
Existing scholarship can inform all three aims. A descriptive study still chooses what to describe and how to organise it. An exploratory study can begin with sensitising concepts and revise them as the inquiry develops.
Set boundaries and identify the units
Boundaries specify the case’s scope: the activity, location, organisation, participants, and period under study. They give a focused inquiry enough room to examine the context that matters.
For the teaching example, the case is the implementation of a correction protocol at Newsroom A during its first twelve weeks. The main unit is that bounded implementation process. If the design separately analyses the reporting and copy-editing desks, these become embedded units within the case.
Editorial history, staffing, deadlines, platform features, and professional expectations.
Correction-protocol implementation at Newsroom A · Weeks 1–12.
Write inclusion rules
- Include: decisions, communications, and records directly concerning factual corrections during the twelve-week period.
- Distinguish: factual corrections, stylistic revisions, and other updates through a stated classification rule.
- Use as context: earlier policies and accounts that help explain the starting situation.
- Document changes: record any extension of the boundary and explain how it affects the question and analysis.
A site is a location where research occurs. One site may contain several cases, such as separate programmes, if the design defines each programme as a case. Conversely, one case may span several sites, such as a programme delivered across campuses.
Select a case and a design deliberately
Case selection should explain what the chosen case makes possible to investigate. Access and resources matter, alongside the intellectual reason for choosing this case rather than another.
Give the selection a rationale
An ordinary or typical case can illuminate a familiar process, with typicality supported by relevant information. An unusual or extreme case can make a process especially visible. A deviant case challenges an expected pattern. A critical case tests a carefully specified theoretical expectation. A revelatory case opens access to a phenomenon that has been difficult to investigate.
A longitudinal design follows a case over time. For the newsroom study, observing early implementation and later adjustments can reveal how responsibilities develop. Explain the timing and what access permits you to observe.
Use Yin’s design map
One case as a whole
Analyse Newsroom A’s implementation as one integrated process.
One case with subunits
Analyse the reporting and copy-editing desks separately, then connect them with Newsroom A’s implementation.
Several whole cases
Analyse implementation at Newsrooms A and B, developing an account of each before comparing them.
Several cases with subunits
Analyse comparable desks within each newsroom, then compare cases through a common framework.
Plan the logic of comparison
In Yin’s replication logic, literal replication selects cases expected to show similar patterns; theoretical replication selects cases expected to differ for specified reasons. Record the expectation before examining the outcomes, while remaining open to evidence that revises it.
Gerring’s account foregrounds how intensive study of a unit can illuminate a wider class of units. This orientation helps clarify the comparison you intend to make, while other traditions may place the particular case itself at the centre.
Build a protocol before entering the field
A case study protocol records the study’s purpose, fieldwork procedures, questions, and analytical plan. It helps researchers collect evidence consistently while documenting necessary changes.
Connect five design components
The following planning frame develops the five components discussed in the lecture through Yin’s approach.
On a narrow screen, scroll the table sideways.
| Component | Decision to make | Illustration |
|---|---|---|
| Research question | What does the study seek to understand? | How is the correction protocol translated into daily publishing decisions? |
| Propositions or guiding issues | Which ideas focus attention, where appropriate? | Clear responsibility may help staff produce traceable public corrections. |
| Case and units | What is bounded, and what receives analysis? | One implementation over twelve weeks; two desk-level embedded units. |
| Logic connecting evidence and ideas | How will the observations inform the explanation? | Compare policy, observed decisions, accounts, and publication sequences. |
| Interpretive criteria | What would support, complicate, or challenge a claim? | Examine responsibility, staffing, software changes, and correction complexity against the episode records. |
A proposition is a tentative statement directing the inquiry. Exploratory and some interpretive designs may instead use guiding questions or issues. Explain the choice through your framework; preserve room for unexpected findings.
Prepare the fieldwork tools
Develop an interview guide, observation prompts, document-request list, source register, and plan for accessing relevant records. Pilot the tools. Check whether the planned sources can actually reveal the decisions or processes the question concerns.
Question and purpose: [the issue, intended contribution, and framework].
Case and selection: [boundaries, rationale, sites, and embedded units].
Evidence: [sources, access, selection within the case, and missing material].
Fieldwork: [contacts, consent, guides, recording, and secure storage].
Analysis: [chronology, coding, comparison, propositions, and rivals].
Reporting: [case description, analytical claims, source references, and scope].
Keep a dated record of revisions. If an unexpected process becomes central, explain why it changed the evidence plan and how that change affects the final account.
Gather sources with different strengths
Multiple sources can illuminate different parts of a case: formal rules, recorded events, participants’ meanings, observed practices, and the material arrangements that make action possible.
On a narrow screen, scroll the table sideways.
| Source | What it may contribute | A check to make |
|---|---|---|
| Documents | Correction policies, emails, meeting notes, and training material. | Check authorship, date, purpose, version, and intended audience. |
| Archival records | Publication histories, correction registers, staffing records, and logged timestamps. | Establish what was recorded, how consistently, and what gaps remain. |
| Interviews | Accounts of decisions, responsibilities, pressures, and disagreements. | Distinguish recollection, interpretation, and directly witnessed events. |
| Direct observation | Editing meetings, correction decisions, and work across a shift. | Record timing, setting, actions, and the researcher’s observational position. |
| Participant observation | Experience of routines while occupying an agreed role in the setting. | Explain the role, access, influence, and reflexive decisions. |
| Physical or digital artefacts | Templates, interface fields, workflow boards, devices, and versioned screenshots. | Explain how the artefact was used and preserve the relevant version. |
Audio, video, photographs, and screen recordings can support these categories. Select the sources needed for the question. Quantitative records can also contribute: for example, counts of eligible corrections, availability of public labels, and time between verified updates and labels.
Create a case study database
Give each item a source ID. Record its date, origin, location, access conditions, case, embedded unit where relevant, and an analytical note. Keep raw material distinguishable from summaries and interpretations. Link transcript passages to recordings and observation notes to the relevant episodes.
Examine convergence and disagreement
Triangulation compares evidence across sources, methods, perspectives, or researchers. A policy may assign responsibility clearly while staff accounts reveal different local understandings. The discrepancy can identify an important part of the implementation process.
Build an analysis within the case
Within-case analysis connects events, meanings, and conditions into an account of the particular case. Begin by understanding its chronology and organisation, then develop interpretations that remain linked to identifiable evidence.
Establish the chronology
Map the protocol’s introduction, training, early decisions, disputes, revisions, and later episodes. Use more than one relevant source where possible to establish dates and sequences. Mark uncertain timing explicitly.
Develop categories and direct interpretations
Code passages and episodes relevant to the question: responsibility, approval, workload, technical access, or reader visibility. Through categorical aggregation, bring together instances that help develop an issue or pattern. Through direct interpretation, examine what a particular episode reveals within the case.
For example, repeated exchanges about approval may develop into an analytical category of “uncertain authorisation.” Explain how the exchanges support that reading. Preserve differences between roles and phases instead of reducing every passage to the same theme.
Match patterns and build explanations
Pattern matching compares an observed pattern with a specified expectation. Explanation building develops and revises an account as evidence accumulates. If labels remain delayed after responsibility is clarified, examine additional conditions such as staff availability or correction complexity.
Process tracing investigates evidence about the steps through which a proposed mechanism operated. Ask who recognised the error, who could change the text, who authorised the label, and what happened between these actions. Collier’s discussion develops this form of inquiry.
- Locate the sourceObservation note O02, linked to a dated editing episode.
- Record the observationThe reporter updates the text while label approval remains unresolved.
- Develop the interpretationThe workflow separates correction of content from public acknowledgement.
- Test the case assertionCompare other episodes, accounts, and policy versions before stating the claim.
Work through the newsroom example
A worked evidence record shows how case study reasoning develops. The fictional packet below contains selected episodes from the teaching case; a complete study would examine its wider case database.
Inspect the source record
On a narrow screen, scroll the table sideways.
| Week and source ID | What the record contains | Analytical relevance |
|---|---|---|
| Week 1 · D01 | A protocol says reporters initiate corrections and editors approve them; responsibility for adding a public label remains unspecified. | Establishes the formal starting arrangement. |
| Week 3 · O02 | An observed reporter changes incorrect text while staff discuss who may add the label. | Shows how responsibility becomes an issue in practice. |
| Week 3 · L03 | The version log records a text correction at 09:20 and the public label at 10:05. | Establishes a 45-minute interval for this episode. |
| Week 4 · I04 | A reporter describes uncertainty about approval of public correction labels. | Provides a participant’s interpretation to compare with the episode. |
| Week 6 · D05 | A clarification assigns the copy editor responsibility for adding the label after approval. | Documents a change in the formal workflow. |
| Week 7 · L06 | A different article is corrected at 11:10 and labelled at 11:12. | Shows a shorter interval in a later episode. |
| Week 8 · O07 | Another label is delayed during a busy shift when the responsible editor is unavailable. | Complicates an explanation based on role clarity alone. |
Develop an explanation and its rivals
The initial proposition concerns clear responsibility. D01, O02, I04, and D05 help examine that proposition. L03 and L06 provide episode-level timing. O07 suggests that an assigned role may still depend on someone being available to perform it.
Consider alternative explanations. A software change might have made labels easier to add; inspect release records, permissions, and interfaces. Later corrections might have been simpler; compare their type and approval requirements. Workload may interact with responsibility; examine staff availability across comparable episodes.
Define the outcome carefully. Public labelling, correction of the underlying error, and audience trust are different outcomes. The packet directly concerns the correction workflow and visible labelling. Audience responses would require appropriate additional evidence.
Read a model case assertion with its evidence
In the selected fictional episodes from Newsroom A, public acknowledgement of corrections depended partly on coordination between reporting and copy-editing roles. The initial protocol left label ownership unspecified (D01), and an observed editing episode exposed uncertainty about authorisation (O02), also described in a reporter’s account (I04). The formal clarification in Week 6 assigned responsibility to the copy editor (D05). A later log showed a two-minute correction-to-label interval (L06), compared with 45 minutes in the earlier recorded episode (L03). However, the Week 8 observation documented a further delay when the responsible editor was unavailable (O07). These episodes support an explanation involving both responsibility and staffing conditions. A full-case assessment would examine the remaining episodes, software changes, and differences in correction complexity before judging how consistently that explanation holds.
Move from within-case analysis to cross-case synthesis
Cross-case analysis examines similarities, contrasts, and conditions across cases while preserving an account of each case. A shared framework makes comparisons systematic and keeps differences visible.
First prepare a case description, chronology, and evidence-based assertions for each case. Then compare them through dimensions relevant to the question. Record absent or inaccessible evidence as gaps rather than treating it as evidence that a process did not occur.
Use a comparison matrix
The table extends the fictional exercise with an imagined Newsroom B. Its entries illustrate analytical comparison rather than report fieldwork findings.
On a narrow screen, scroll the table sideways.
| Dimension | Newsroom A | Imagined Newsroom B |
|---|---|---|
| Formal responsibility | Label ownership is clarified during implementation. | Label ownership is clear from the start. |
| Staff availability | Some shifts leave the responsible editor unavailable. | Peak shifts also create gaps in editorial cover. |
| Pattern requiring explanation | Early uncertainty and later staffing gaps both feature in delayed episodes. | Delays occur during staffing gaps despite clear formal responsibility. |
| Comparative implication | Examine the interaction between responsibility and capacity. | Test whether capacity helps explain delays where role ambiguity is less plausible. |
This comparison refines the question: under what conditions does a formal responsibility become an action someone can reliably perform? It also directs further evidence collection. Examine handovers, backup arrangements, correction complexity, and technical permissions in both cases.
Look for differences as well as shared themes
A common theme can take different forms across cases. One newsroom may coordinate through a formal register; another may rely on a senior editor’s informal judgment. Describe how each arrangement works before considering whether a wider proposition captures both.
Strengthen quality, reflexivity, and ethical practice
A convincing case study makes its selections, evidence, reasoning, and limits visible. Quality criteria should fit the study’s philosophical position and analytical purpose.
Use design checks that serve the question
On a narrow screen, scroll the table sideways.
| Concern | What to establish | A practical action |
|---|---|---|
| Construct validity | The evidence adequately represents the concepts being studied. | Define “public correction” and distinguish labels, text changes, and other updates. |
| Internal validity | An explanatory claim addresses plausible rival accounts. | Examine the proposed mechanism alongside workload, software, and episode complexity. |
| External validity | The reach of the explanation or theoretical claim is justified. | State the conditions under which the proposed process may apply and use a reasoned comparison. |
| Reliability | The procedures and evidence handling can be followed and reviewed. | Maintain a protocol, source database, version history, and chain of evidence. |
Internal-validity checks are especially relevant to explanatory work. Interpretive traditions may place greater emphasis on credibility, contextual depth, reflexivity, and the adequacy of the account. Explain which criteria you adopt and how the study addresses them.
Record your position and access
Reflexivity examines how the researcher’s role, relationships, assumptions, and decisions shape the inquiry. A former journalist may recognise an editorial routine quickly while also taking parts of it for granted. Record those expectations and seek perspectives that complicate them.
Describe who authorised access, which shifts or records were available, and whose participation was difficult to secure. Participants may help clarify a factual chronology or respond to an interpretation. Explain how their feedback was used and retain relevant disagreements.
Plan consent, confidentiality, and reporting
Agree on observation and recording arrangements, protect sensitive records, and explain how findings may be reported. Organisational access and individual consent address different relationships. Staff should understand what participation involves and how their accounts will be handled.
A protocol supports consistency and review. In a changing social setting, it also allows readers to distinguish procedural changes from changes in the case itself.
Write the case and explain its contribution
A case study report combines a clear account of the setting with an argument grounded in evidence. Give readers enough detail to understand the process and assess how you reached the claims.
Organise the report around the question
- Orient the reader: introduce the case, issue, setting, and reason for investigation.
- Explain the design: state the framework, boundaries, case selection, evidence, procedures, and ethical decisions.
- Describe the case: provide the chronology, roles, and conditions needed to understand the analysis.
- Develop the key issues: connect excerpts, observations, and records with interpretive claims, including difficult evidence.
- Compare where relevant: present each case before developing cross-case propositions.
- State the contribution: explain what the case clarifies, what remains uncertain, and the reach of the account.
A short vignette can draw the reader into an observed moment. Identify its evidential basis and signal any reconstruction. For a methods-focused report, a direct opening with the research problem may be more useful. Choose the structure through the purpose and publication context.
Explain how the findings travel
Analytic generalisation
Relate the evidence to a concept, proposition, or explanation, specifying its conditions and alternatives.
Transferability
Describe the case and context sufficiently for a reasoned assessment of relevance to another setting.
Naturalistic generalisation
Offer a detailed account through which readers can connect the case with their own knowledge and experience.
A strategically selected case can refine or challenge a theoretical claim. Flyvbjerg develops the contribution of case knowledge and the value of careful case selection. Explain the logical argument rather than treating a case count as the warrant.
Use the case to examine values
Cases can also inform normative inquiry. Ask what “accountability” should require, whose interests a correction process serves, and how competing obligations are handled. Thacher’s account develops this contribution. Make the empirical evidence and the reasoning about values explicit, showing how each informs the judgment.
Use the lecture as a case study workshop
Watch Dr. Uma Shankar Pandey’s lecture with a case-definition sheet beside you. Pause at the links below to turn each concept into a decision for your own study.
If the player does not load, watch the lecture on YouTube. Links mark approximate starting points from the supplied transcript.
Attempt each question before opening the response
1. Do twenty interviews necessarily constitute twenty cases?
The case count follows the defined unit. Twenty staff interviews may provide evidence about one newsroom implementation. If individual career trajectories are separately bounded and analysed as cases, the design would be different. State the case before counting participants.
2. Can one case serve an instrumental purpose?
Yes. One case can illuminate a broader issue, such as the translation of accountability rules into routine practice. The selection rationale and evidence should explain what the case helps reveal about that issue.
3. What makes a design embedded?
The design specifies subunits for separate analysis within the bounded case. Reporting and copy-editing desks could be embedded units. The final account must connect their analysis back to the case-level question.
4. What would weaken the role-clarity explanation in the worked example?
For example, evidence that label ownership was already clearly understood in early delayed episodes would challenge the initial reading. A coinciding software change could also provide a rival account of later timing. Inspect relevant records and participants’ practices to assess these possibilities.
5. How would you investigate the workload explanation?
Compare episodes with different staffing and workload conditions while attending to correction type, approval requirements, and technical access. Use schedules, observations, and episode records. Examine both prompt and delayed labels, including cases that contradict the expected pattern.
6. What should come before a cross-case thematic table?
Develop an account of each case: its context, chronology, evidence, and principal assertions. This preserves the meaning of each entry and helps ensure that a shared heading refers to comparable processes.
7. How can a case study contribute beyond its immediate setting?
It can develop or challenge a proposition, specify a process and its conditions, or provide contextual detail for assessing relevance elsewhere. Explain the inferential route and its scope. A claim about population frequencies requires an appropriate population design.
Your independent learning project
Design a case study of a newsroom decision, campus campaign, community media initiative, public communication programme, or platform-related controversy.
- Write a focused question and purpose statement.
- Define the case, boundaries, and units.
- Justify selection and the number of cases.
- Identify three complementary evidence sources and access requirements.
- Draft a guiding proposition or issue and one rival interpretation.
- Prepare a source register, chronology template, and analytical memo.
- Outline the final report and state how its contribution will be assessed.
Completion check: Ask a peer to identify your case, reconstruct your evidence plan, and explain what would lead you to revise a claim.
Glossary and further learning
Keep these terms beside your protocol. Begin further reading with the approach closest to your question and use the open articles to compare design choices.
- Case
- The bounded entity or process investigated in depth.
- Context
- The circumstances, relationships, and practices relevant to understanding the case.
- Unit of analysis
- The entity about which a particular level of analysis makes claims.
- Embedded unit
- A specified subunit analysed within a case and related back to the case-level inquiry.
- Intrinsic interest
- An interest in understanding the particular case for its own sake.
- Instrumental interest
- An interest in using a case to illuminate an issue or broader question.
- Proposition
- A tentative statement that helps focus the evidence and analysis where appropriate.
- Triangulation
- Comparison across relevant sources, methods, perspectives, or researchers to examine an interpretation.
- Pattern matching
- Comparison of an observed pattern with a specified expectation.
- Process tracing
- Investigation of diagnostic evidence about the sequence and operation of a proposed process or mechanism.
- Case assertion
- An interpretive claim about the case supported by a transparent evidence trail.
- Chain of evidence
- Traceable links connecting the question, source material, analysis, and conclusions.
- Analytic generalisation
- Reasoned movement from case evidence to concepts, propositions, or explanations with stated conditions.
References and learning resources
- Pandey, U. S. How to do case study research. USP Lectures [Video].
https://www.youtube.com/watch?v=LvNfSLSv2EYThe companion lecture introduces definitions, purposes, design, evidence, interpretation, and reporting. - Yin, R. K. (2018). Case study research and applications: Design and methods (6th ed.). SAGE.
Publisher’s book page · Companion learning resourcesDevelops design components, evidence, analytic strategies, and replication logic. The book carries a 2018 copyright; publisher listings give its 2017 release. - Stake, R. E. (1995). The art of case study research. SAGE.
Author’s university publication recordA foundation for understanding particularity, interpretation, the researcher’s role, triangulation, and reporting. - Baxter, P., & Jack, S. (2008). Qualitative case study methodology: Study design and implementation for novice researchers. The Qualitative Report, 13(4), 544–559.
Read the open-access articleA practical route into questions, propositions, case boundaries, design types, and sources. - Crowe, S., Cresswell, K., Robertson, A., Huby, G., Avery, A., & Sheikh, A. (2011). The case study approach. BMC Medical Research Methodology, 11, Article 100.
Read the open-access methods articleConnects design, fieldwork, analysis, and reporting through published health-service cases. - Gerring, J. (2004). What is a case study and what is it good for? American Political Science Review, 98(2), 341–354.
Publisher’s article recordExamines intensive study, case definition, and the relationship between single-unit and cross-unit research. - Flyvbjerg, B. (2006). Five misunderstandings about case-study research. Qualitative Inquiry, 12(2), 219–245.
Publisher’s article recordDevelops the contribution of case knowledge, case selection, and theoretical learning. - Collier, D. (2011). Understanding process tracing. PS: Political Science & Politics, 44(4), 823–830.
Read the author-hosted article with addendum [PDF]Further guidance on within-case evidence and the assessment of explanatory processes. - Ragin, C. C. (2008). Redesigning social inquiry: Fuzzy sets and beyond. University of Chicago Press.
Publisher’s book pageA further reading for set-theoretic comparison and QCA. - Thacher, D. (2006). The normative case study. American Journal of Sociology, 111(6), 1631–1676.
Read the author-hosted article [PDF]Examines how cases can inform judgments about public values and obligations.
Choose the next layer of evidence
A case study becomes stronger when each method has a clear job. Ethnography can deepen attention to routines and relationships; Content Analysis can organise patterns in the media material gathered within the case. Bring those findings back to the bounded case and the question that gives it coherence.