A case study dissertation studies one or more bounded instances — an organization, program, classroom, or community — in depth to understand a phenomenon in its real context. This guide covers single vs multiple case designs, building a case study protocol, data collection, and cross-case analysis.
A case study dissertation is a qualitative (or mixed-methods) design that investigates one or more bounded "cases" — an organization, program, classroom, community, or individual — in depth, using multiple sources of evidence to understand a phenomenon within its real-world context. It's a recognized methodology in its own right (most closely associated with Robert Yin's work), not just "an example used to illustrate a point."
The defining feature is the bounded case: a clearly defined unit of analysis with explicit limits in time, place, and scope. Your task is to describe the case richly, analyze it against your theoretical framework, and draw conclusions that are transferable (even though not statistically generalizable) to similar contexts.
| Design | When to Use | Trade-off |
|---|---|---|
| Single case | The case is unique, critical, or revelatory — access to a rare or extreme example | Deeper insight, but findings are harder to generalize beyond the case |
| Multiple case (2–4+ cases) | You want to compare across contexts and identify patterns that hold across sites | Stronger transferability via replication logic, but each case gets less depth |
Multiple case designs use replication logic, not sampling logic: each additional case is chosen to either confirm or extend the findings from prior cases (literal replication) or to deliberately test the boundaries of the theory (theoretical replication) — not to build a representative sample.
Before data collection begins, committees expect a documented case study protocol — the qualitative equivalent of a quantitative study's pre-registered procedure. It should specify:
Triangulate your data sources. A case study built on interviews alone is vulnerable to credibility challenges. Combine interviews with documents (policies, reports, emails), observations, and artifacts where possible — convergence across sources is what makes case study findings defensible.
Our specialists help with case selection, protocol design, within-case writeups, and cross-case analysis.
Before any comparison across cases, each case must be presented as a coherent, standalone narrative — sometimes called a "thick description." This section should let the reader understand the case on its own terms before you start drawing comparisons or applying theory. Organize it descriptively (chronologically, or by sub-theme) rather than jumping straight to your research questions.
If your design has multiple cases, the cross-case analysis is where the dissertation's real analytical contribution happens. Common techniques:
| Technique | What It Does |
|---|---|
| Pattern matching | Compares observed patterns in the case(s) against patterns predicted by theory |
| Explanation building | Iteratively builds an explanation for the phenomenon, case by case |
| Cross-case synthesis | Tabulates findings side by side to surface similarities and differences |
One well-justified case is sufficient if it's unique, critical, or revelatory. Most multiple-case dissertations use 2–4 cases — enough for cross-case comparison without diluting depth. More than four becomes difficult to manage with rich qualitative data.
Yes, when designed and documented properly. Rigor in case study research comes from triangulation, a documented protocol, and transparent analysis — not from sample size. Committees evaluate it against qualitative standards (credibility, transferability, dependability, confirmability), not quantitative ones.
Yes — many case study dissertations are mixed-methods, embedding survey data or institutional metrics within a primarily qualitative case design. See our mixed-methods guide for how the two approaches integrate.