Case studies are everywhere in business, education, health, policy and engineering theses, partly because they fit a student timeline and partly because organisations are often willing to be studied. The difference between a strong and a weak case study is almost always in the design decisions made before any data is collected.
This post covers what counts as a case study, the main types, how to select and bound a case, which data to collect, how to analyse it, and how to structure the write-up.
What a Case Study Is
A case study is an in-depth investigation of a contemporary phenomenon within its real-life context, especially when the boundary between phenomenon and context is not clear (Yin, 2018). The "case" can be a person, an organisation, a project, a policy, an event or a community. What makes it a case study is not the size of the sample but the depth, the use of several sources of evidence, and the attention to context.
Types of Case Study
| Type | Purpose | Example question |
|---|---|---|
| Exploratory | Investigate something little understood, often before a larger study | How do small farms decide whether to adopt drone monitoring? |
| Descriptive | Give a detailed account of a phenomenon in context | How is a hospital's new triage system used on night shifts? |
| Explanatory | Explain how and why an outcome occurred | Why did the merger of two regional banks fail to deliver its savings? |
| Intrinsic | The case itself is of interest | How did one city achieve zero homelessness? |
| Instrumental | The case illuminates a wider issue | What does one firm's four-day week reveal about productivity measurement? |
| Multiple / comparative | Compare several cases to find patterns | How do three universities implement the same AI policy differently? |
Write the case study chapter from your sources
ThesisAI drafts complete academic documents with methodology, findings and discussion, and checks every citation against the paper it came from.
Начать писатьSelecting the Case
The case must be chosen for a reason that supports the claim you want to make. "My employer agreed" is how you gained access, not why the case is informative.
- Typical case: representative of many others, so findings are likely to apply more widely.
- Extreme or unusual case: shows a phenomenon in its clearest form.
- Critical case: if a theory fails here, where conditions are most favourable, it will fail elsewhere.
- Deviant case: an exception to an established pattern, useful for refining theory.
- Revelatory case: a situation previously inaccessible to researchers.
Bounding the Case
State what is inside the case and what is context, in time, place and scope. "The digital transformation of Company X" is unbounded. "The rollout of a new CRM system in Company X's sales department between January and December 2025" is a case. Drift beyond the boundary is the most common reason case studies grow unmanageable.
Data Sources
Triangulation across sources is what gives a case study its strength. Aim for at least three of the following:
- Interviews with people in different roles.
- Documents: reports, minutes, emails, policies, strategy papers.
- Archival records: performance data, budgets, service records.
- Direct observation of meetings, work or events.
- Surveys of a wider group within the case.
- Physical artefacts such as products, tools or software.
Our guide to data collection methods covers each in detail.
Analysing Case Study Data
- Build a case database. Keep all notes, transcripts and documents organised and dated, so a reader could trace any claim to its evidence.
- Code the data. Thematic analysis is the usual starting point for qualitative material.
- Match patterns. Compare what you found with what your theory predicted.
- Build an explanation. For explanatory cases, assemble a causal story and test it against rival explanations.
- Construct a timeline where sequence matters.
- Compare across cases in multiple-case designs: analyse each case on its own first, then look for patterns across them.
Structure of the Write-up
- Introduction and problem
- Literature review and theoretical lens
- Methodology: why a case study, case selection, boundary, sources, analysis, ethics
- Case description: the context a reader needs, kept short
- Findings, organised by theme or research question, not by source
- Discussion against the theory
- Conclusion, with the limits on generalisation stated
FAQs About Case Study Research
Is a case study qualitative or quantitative?
Usually qualitative, but it can use quantitative data or both. The design is defined by depth and context, not by the kind of data. See qualitative vs quantitative research.
How many cases do I need?
One is enough if it is well chosen and justified. Multiple-case designs typically use two to four cases in a thesis. More than that and depth suffers.
Can you generalise from a case study?
Not statistically, to a population. You can generalise analytically, to a theory, and you can describe the conditions under which your findings are likely to hold elsewhere.
Do I need to anonymise the organisation?
Usually yes, unless the organisation has agreed in writing to be named. Even then, protect individual participants. Our post on ethical considerations in research covers this.
Pick a case for what it can teach you, draw a firm boundary around it, collect evidence from several sources and analyse it against a theory. That combination is what turns a case description into a case study an examiner will credit.