Case Study Research Design

Understand what a case study is and when to use one, define what counts as a ‘case’, work through the single/multiple × holistic/embedded typology, choose a case selection strategy, and answer the common critiques of the method.

1 Introduction

Some research questions cannot be answered by averaging over thousands of cases. When you want to understand how a complex process unfolds in its real-world setting — and where the phenomenon is hard to separate from its surroundings — the case study comes into its own. This page sets out what a case study is, when to choose one, how to design it, and how to defend it against its critics.

2 What is a Case Study?

A widely used definition (Robert K. Yin) captures the essentials:

“An empirical inquiry that investigates a contemporary phenomenon within its real-life context; when the boundaries between phenomenon and context are not clearly evident; and in which multiple sources of evidence are used.”

Three features stand out:

• it studies a contemporary phenomenon;
• it takes context seriously rather than stripping it away;
• it relies on the triangulation of multiple sources of evidence.

2.1 Case studies among other strategies

Yin’s comparison locates the case study relative to neighbouring strategies:

Strategy	Form of research question	Requires control of behavioural events?	Focuses on contemporary events?
Experiment	How, why?	Yes	Yes
Survey	Who, what, where, how many, how much?	No	Yes
Archival analysis	Who, what, where, how many, how much?	No	Yes / No
History	How, why?	No	No
Case study	How, why?	No	Yes

Case studies share the how/why focus of experiments and history, but — unlike experiments — do not require control over events, and — unlike history — concern the present.

3 When to Use Case Studies

Choose a case-oriented approach when you need:

• depth in:
  • exploratory how-questions — understanding complex social phenomena;
  • why-questions — identifying the exact causal process;
  • evaluative questions;
• to work where pre-existing knowledge is scarce;
• to handle a complex object of research with blurry boundaries between the phenomenon and its context;
• to cope with a scarcity of empirical information and cases.

The function of a case study also shifts with the role of theory: it can be exploratory (plausibility probe, hypothesis-generating), explanatory (hypothesis-testing), or descriptive (idiographic). And how it is used depends on the disciplinary angle — Henry Mintzberg’s close studies of what managers actually do (1970s–80s) look different from a case study in political science, even though both bear the same label.

4 What is a “Case”?

Defining the case is half the design:

A case is a spatially delimited phenomenon (a unit) observed at a single point in time or over some period. It comprises the type of phenomenon that an inference attempts to explain.

Practical implications:

• there are different levels of analysis (an individual, a team, a firm, a country);
• the unit of analysis must align with your research questions and theoretical interests;
• you must establish clear boundaries for what is inside vs. outside the case.

5 Context is Central

Because a case study refuses to strip away context, you must think carefully about how context operates. There are three types of contextual effect:

• context as cause — the setting itself drives the outcome;
• context as barrier — the setting constrains what can happen;
• context changes meaning — the same act means different things in different settings.

Context also operates on multiple levels — micro, meso and macro.

6 A Typology of Case Studies

Two dimensions cross-classify case study designs:

1. How many cases? — single-case vs. multiple-case designs.
2. How many units of analysis? — holistic (a single unit) vs. embedded (multiple units within the case).

	Single-case	Multiple-case
Holistic (one unit)	One case analysed as a whole	Several whole cases compared
Embedded (multiple units)	One case with sub-units analysed inside it	Several cases, each with embedded sub-units

7 Designing a Case Study

The fundamental elements of a case study design form a chain:

Research question → Propositions / hypotheses → Units of analysis → Link between data and propositions → Interpretation criteria

The middle links are what give a case study analytical force: without explicit propositions, defined units, and stated criteria for interpreting evidence, a case study risks becoming mere description.

7.1 Case selection strategies

Whether single- or multiple-case, you must pick cases deliberately. Common strategies:

• typical cases (representative of a broader class);
• critical cases (where a theory should hold if it holds anywhere);
• extreme or unique cases;
• revelatory cases (offering access to a previously hidden phenomenon);
• longitudinal cases (the same case studied over time).

ImportantJustification is key

There is no random sampling to fall back on. Whatever strategy you choose, the justification of your case selection — why this case, given this question — is what makes the design defensible.

8 Answering the Critics

Case studies attract recurring critiques. Know them, and the standard responses:

1. Lack of scientific rigor — addressed by an explicit design chain and documented procedures.
2. Limited generalizability — case studies aim at analytic (theoretical) generalisation to propositions, not statistical generalisation to populations.
3. Researcher bias — mitigated by triangulation and transparent interpretation criteria.
4. Fails to determine causation — within-case process tracing can establish causal mechanisms that correlational designs cannot.
5. Difficult to replicate — addressed by thorough documentation of the design and evidence base.

Note🎓 MA / PhD

The generalisation critique repays careful thought. Case studies trade statistical generalisation (sample → population) for analytic generalisation (case → theory): the question is not “how many cases does this represent?” but “what does this case let us infer about the proposition?”. This connects to process tracing as a within-case method for establishing causal mechanisms, and to the logic of critical and deviant case selection, where a single well-chosen case can decisively support or undermine a theoretical claim. For causal questions specifically, contrast the case study’s mechanism-centred logic with the variance-centred logic of DAG-based reasoning.

9 Summary

• A case study investigates a contemporary phenomenon in its real-life context, where phenomenon and context blur, using multiple sources of evidence.
• Use it for how/why questions, scarce prior knowledge, complex boundaries, and few available cases; its function depends on the role of theory and the discipline.
• Define the case as a spatially delimited unit with clear boundaries, aligned to the research question; take context (cause, barrier, meaning-changer) seriously.
• Designs cross-classify as single/multiple × holistic/embedded; build the RQ → propositions → units → data-link → interpretation chain and justify your case selection.
• The standard critiques (rigor, generalizability, bias, causation, replicability) have standard, design-based answers.

10 What to Read Next

• Conducting Research Interviews — a frequent source of evidence within case studies.
• Qualitative vs Quantitative Research — the broader landscape in which case studies sit.

11 Further Reading

• Yin, R. K. (2018). Case Study Research and Applications: Design and Methods (6th ed.). SAGE.
• Gerring, J. (2017). Case Study Research: Principles and Practices (2nd ed.). Cambridge University Press.
• Flyvbjerg, B. (2006). “Five Misunderstandings About Case-Study Research.” Qualitative Inquiry, 12(2), 219–245.