Book summary “Case Study Research”: summary

Introduction

You often face complex social and behavioral research problems that require deep empirical investigation. Choosing a suitable method proves difficult when you need to answer how or why questions about contemporary events outside your control. You need an approach that handles situations with many more variables of interest than data points.

Case study research is a rigorous empirical inquiry that investigates contemporary phenomena within real-world contexts using multiple sources of evidence and replication logic. Robert K. Yin provides a comprehensive guide that spans every phase of your investigation. The book outlines how you define research questions, design your studies, collect evidence, analyze findings, and compose reports.

You will find answers to several critical questions as you read:

• When should you choose a case study over other research methods?
• How do you construct a valid and reliable research design?
• What protocols and sources of evidence must you deploy in the field?
• How do you analyze complex data without statistical distortion?

1. Getting Started With Case Studies

You should use a case study when addressing how or why questions about contemporary events outside your control.

Defining the Method

Case study research is the preferred method when addressing how or why questions, when the researcher has little control over behavioral events, and when the focus is on contemporary phenomena. These three conditions dictate the choice of research method by comparing case studies against experiments, surveys, histories, and archival analyses. Experiments allow behavioral control, histories focus on entirely non-contemporary events, and surveys focus on prevalence rather than operational links over time.

A case study investigates a contemporary phenomenon in its real-world context, especially when boundaries between phenomenon and context are not clearly evident. This twofold definition encompasses both the real-world contextual scope of the study and the technical characteristics required to handle situations where there are more variables of interest than data points.

The choice of research method depends primarily on the type of research question, your control over behavioral events, and the focus on contemporary versus historical events. How and why questions favor case studies, experiments, or histories because they require tracing operational links over time. What questions can be exploratory, or about prevalence.

Addressing Concerns

Rigorous case study research requires actively addressing five traditional concerns: rigor, confusion with teaching cases, generalization, management of effort, and comparative advantage.

Critics have historically called case studies soft, but conducting them properly means countering these concerns through disciplined design, avoiding pedagogical teaching cases, understanding analytic generalization, and properly scoping the research effort.

Research case studies must strictly adhere to fair and rigorous presentation of empirical data. Teaching cases used in business, law, or medicine are often deliberately altered or simplified to establish a framework for student discussion and debate. In contrast, research case studies require rigorous procedures, fair reporting of all evidence, and accurate rendition of actual events without alteration.

The core of the scientific method in case studies is the strategy of specifying and empirically testing plausible rival hypotheses. As Donald Campbell argues in the foreword, case studies mirror experimental isolation by controlling for specifically stated rival explanations through extended networks of implications and ramification extinction.

2. Designing Case Study Research

A rigorous research design acts as your logical bridge connecting data collection and conclusions to initial study questions.

Every empirical study contains an implicit or explicit research design. You must articulate a framework that strengthens the design and lays the groundwork for analytic generalizations.

• Articulate a theory about what is being studied and what is to be learned.
• Define the case or unit of analysis to be studied.
• Set limits or bounds to the case.
• Examine the quality of the emerging design against construct validity, internal validity, external validity, and reliability.

Components and Theory

A rigorous case study design requires five specific components that connect initial questions to final conclusions and guide data analysis. These essential components prevent the evidence from failing to address the initial research questions.

Correctly defining and bounding the unit of analysis prevents the study from becoming overly broad and clarifies what the case actually represents. You can study an individual, small group, organization, program, or event while establishing spatial and temporal boundaries.

Tracy Kidder's book The Soul of a New Machine illustrates the unit of analysis challenge by describing the development of a minicomputer by Data General's engineering team, leaving open whether the case is the technology or group dynamics.

Developing preliminary theoretical propositions prior to data collection distinguishes case studies from other qualitative methods and guides both data collection and generalization. Theoretical propositions act as a hypothetical story explaining why events occur, forming the foundation for later analysis.

Diane Vaughan used a case study of the space shuttle Challenger disaster to explicate the sociology of mistake, showing how organizational social structure transformed deviance into routine behavior.

Quality Tests and Designs

Case study research quality is judged using four established logical tests: construct validity, internal validity, external validity, and reliability. These tests require operational measures for concepts, establishing causal relationships for explanatory studies, defining the domain for generalization, and ensuring procedures can be repeated.

Single-case designs are justified under specific conditions analogous to a single experiment. You can choose a critical test of an existing theory, an unusual clinical circumstance, an everyday common situation, a revelatory observation, or a longitudinal study tracking changes over time.

Gross, Bernstein, and Giacquinta used a single school with a prior history of innovation as a critical case to test prevailing theories about barriers to organizational innovation.

Multiple-case studies must follow a replication logic rather than a sampling logic, predicting either similar or contrasting results for predictable reasons. Each case serves as a literal replication predicting similar results or a theoretical replication predicting contrasting results for anticipatable reasons.

Peter Szanton reviewed eight unsuccessful attempts by university groups to help city governments as literal replications, followed by contrasting groups of cases to illustrate theoretical replication.

Case studies can be structured as holistic designs examining a global phenomenon or embedded designs incorporating subunits of analysis. Embedded designs help prevent abstract drift but require you to return attention to the larger unit of analysis.

Seymour Martin Lipset, Martin Trow, and James Coleman used an embedded single-case design in Union Democracy to study the International Typographical Union across multiple levels from the organization as a whole down to individual members.

3. Preparing to Collect Evidence

Case study research requires exceptional intellectual and interpersonal skills rather than simple technical routines. Unlike experiments or surveys where data collection can be routinized and delegated to assistants, case studies involve continuous interaction between theory and emerging data. You must navigate procedural uncertainties and make delicate judgments without falling into biased procedures.

You must master specific investigator skills, protect human subjects, and establish protocols before entering the field.

Skills and Ethics

As a case study researcher, you exercise discretionary behavior throughout the inquiry. You must practice specific investigator abilities to maintain high standards:

• ask good questions and interpret the answers fairly;
• be a good listener not trapped by existing ideologies or preconceptions;
• stay adaptive to see newly encountered situations as opportunities;
• have a firm grasp of the issues being studied;
• avoid biases by being sensitive to contrary evidence and maintaining high ethical standards.

Case study researchers must obtain formal approval from an institutional review board to protect human subjects before collecting data. Because case studies investigate contemporary phenomena in real-world contexts involving human affairs, you are obligated to practice high ethical standards akin to medical research. Guidelines from the National Research Council 2003 cover human subjects protection across 23-28 pages. You must follow ethical rules strictly:

• gain informed consent from all persons participating in the study;
• protect participants from any harm and avoid deception;
• protect privacy and confidentiality;
• take special precautions to protect vulnerable groups like children;
• select participants equitably.

Training for case study research must treat every team member as a senior researcher through group collaboration and protocol mastery. When multiple researchers conduct a case study, training focuses on group discussions of readings, theoretical issues, and methods rather than didactic instruction. This training process helps uncover design flaws, time constraints, or incompatibilities among team members early.

Peter Blau initially intended to study a single organization in his study of government agencies, but switched to comparing two agencies when access to a private firm failed. Fifty years later, Peter Blau's 1955 study The Dynamics of Bureaucracy is still valued. Training sessions involve specific tasks:

• review assigned readings and background materials related to protocol sections;
• lead group discussions to clarify protocol questions and build shared understanding;
• uncover potential design flaws, time limits, or team incompatibilities.

Protocols and Pilot Studies

A comprehensive case study protocol containing four specific sections is essential to guide data collection and increase research reliability. The protocol serves as a major tool to guide data collection for a single case, keeping you targeted and anticipating reporting formats. It differs from survey questionnaires by being primarily directed at you as a set of operational queries. You structure the protocol into four distinct parts:

• Section A: Create an overview including objectives, auspices, issues, and relevant readings;
• Section B: Outline data collection procedures including human subjects protection and logistics;
• Section C: Define substantive data collection questions and likely sources of evidence;
• Section D: Provide a guide for the case study report including outline, format, and bibliography.

Case study protocols must distinguish among 5 levels of questions to avoid confusion between inquiry goals and interview prompts. Articulating Level 2 questions clearly helps maintain a working knowledge of the entire inquiry without letting data collection distort the unit of analysis. The 5 levels of potentially relevant questions in case study research cover distinct analytical tiers:

• Level 1: Ask specific questions of individual interviewees;
• Level 2: Answer protocol questions for the individual case during the study;
• Level 3: Analyze patterns of findings across multiple cases;
• Level 4: Examine information for the entire study including broader literature;
• Level 5: Address normative policy recommendations and conclusions.

Rigorous screening procedures ensure that candidate cases are viable and properly selected before formal data collection begins. Depending on the size of the candidate pool, you use either a one-phase qualitative query or a two-phase quantitative and qualitative reduction strategy. Pools of 12 or fewer candidate cases warrant a one-phase screening, versus larger pools requiring two stages. Marwell's study of urban neighborhood revitalization used demographic data to reduce an initial pool of 59 neighborhoods down to 14 candidates, and then applied four criteria to select 2 finalists. You follow specific screening steps based on pool size:

• define operational criteria for qualified case candidates;
• for small candidate pools, use knowledgeable informants and limited documentation;
• for large candidate pools, use archival quantitative data first to reduce candidates to 12 or fewer before detailed querying.

A pilot case study acts as a formative laboratory to help you refine data collection plans, develop lines of inquiry, and test field procedures. Unlike a formal pretest dress rehearsal, a pilot case is flexible and can have a broader scope. A study of technological innovations in local services utilized 7 pilot cases across different technologies and cities, visiting four local ones first and three in other cities second, to improve conceptualization. You execute a pilot case using practical guidelines:

• select pilot cases based on convenience, access, and geographic proximity;
• adapt parts of the protocol to suit informants' needs in exchange for feedback;
• test formal field procedures and logistical variations on a trial basis;
• write clear memos on lessons learned regarding research design and field procedures.

4. Collecting Case Study Evidence

You must rely on six primary sources of evidence while enforcing data triangulation, databases, and chains of evidence.

Four Principles

Data triangulation ensures that findings are supported by more than a single source of evidence. This approach resembles navigation where different reference points intersect. It distinguishes a rigorous case study from a simple interview study or historical narrative.

Basu, Dirsmith, and Gupta used field observations with diaries, 55 interviews, historical accounts, public records, administrators' personal files, and news articles to triangulate findings on the U.S. Government Accountability Office.

Commingling narrative text with author interpretations in the final report prevents critical readers from inspecting raw data. You must create a formal, orderly case study database that is distinct from the final report to preserve raw data and increase reliability. A separate database should include field notes, documents, tabular materials, and new narrative compilations arranged in computer files and physical folders.

Maintaining a chain of evidence allows any external reader to trace the derivation of data from initial research questions to ultimate conclusions in either direction. This forensic-like principle ensures that no original evidence is lost through carelessness or bias. A reader should be able to move from report conclusions back to cited sources, verify that sources match collection circumstances, and link those circumstances to protocol questions and original study questions.

Six Sources

Case study evidence draws from six primary sources. Each source requires distinct data collection procedures and offers complementary strengths.

• Documentation
• Archival records
• Interviews
• Direct observation
• Participant-observation
• Physical artifacts

Because no single source holds a complete advantage, combining them allows you to address a broader range of historical and behavioral issues. Documents serve primarily to corroborate and augment evidence from other sources rather than provide literal, unmitigated truth. Because documents are written for specific audiences and purposes, you must act as a vicarious observer by identifying those objectives to avoid being misled. Systematic searches should include internet resources, local libraries, and organizational files.

R.N. Jacobs studied the Rodney King crisis by drawing from two different newspapers. He produced 357 articles from a major daily and 137 from an African American community weekly, illustrating how different understandings of the crisis were constructed during a time when 58 people were killed, 2,000 were injured, and 11,000 were arrested.

Case study interviews resemble guided conversations rather than structured queries. You must operate on two levels simultaneously by following your own line of inquiry according to the protocol while asking open-ended and non-threatening questions. Interviews can range from prolonged sessions lasting 2 or more hours to shorter focused interviews taking 1 hour.

William Foote Whyte used a key informant named Doc in his study Street Corner Society, gaining deep insights and access to other sources.

Observations and participant-observation place you in real-world settings to capture behavioral conditions and insider perspectives. Participant-observation allows you to assume functional roles within a situation, granting access to otherwise inaccessible events. Herbert Gans studied Italian-Americans in The Urban Villagers by becoming an actual resident in the neighborhood along with his spouse, utilizing six approaches where participation proved most productive, 20 years after the publication of Whyte's Street Corner Society.

5. Analyzing and Reporting Findings

You analyze case studies using specific strategies like pattern matching and logic models before structuring final reports.

Unlike statistical analysis, case study research lacks fixed formulas or cookbook recipes. Because software tools like CAQDAS only assist in coding and counting text rather than performing final interpretations, you must rely on your own rigorous empirical thinking, clear protocols, and consideration of alternative interpretations.

Researchers must establish an overall analytic strategy before collecting data to prevent case studies from stalling during the analysis stage.

• Start with questions from your case study protocol rather than with the data.
• Start with a small question first, then identify your evidence that addresses the question.
• Draw a tentative conclusion based on the weight of the evidence, asking how to display it for reader verification.
• Continue to a larger question and repeat the procedure until addressing the main research questions.

Gregory Hooks used two cases—the aeronautics industry and the microelectronics industry—to show how the U.S. defense department implemented industrial planning and supported early development, disproving the idea that only foreign governments like Japan used centralized planning.

Analytic Techniques

Pattern matching compares an empirically based pattern found in the case study with a predicted pattern defined prior to data collection. Bradshaw studied the closure of a U.S. military base in a modestly sized California community by examining pre- and post-outcomes across sectors like housing sales, employment, and retail markets, showing outcomes were less severe than anticipated. Explanation building is an iterative process aimed at stipulating a set of causal links and developing a final explanation about an explanatory case study. Edgar Schein studied a major computer maker that grew for thirty years before failing, explaining how the firm lacked a missing gene prioritizing business and marketing operations over technical excellence.

Logic models stipulate and operationalize a complex chain of sequential cause-and-effect events over an extended period. Yin evaluated community planning and prevention initiatives across eight case studies using a program-level logic model depicting the rationale underlying public programs aimed at reducing HIV/AIDS incidence, covering 65 state and local health departments funded across the United States in the HIV/AIDS prevention program. Cross-case synthesis applies to multiple-case studies by treating each individual case study as a separate study to generate robust generalizations. COSMOS Corporation analyzed 14 organizational centers, splitting them into 7 program-supported intervention centers and 7 comparison centers, using word tables to evaluate co-location arrangements.

Composing Reports

Having draft case study reports reviewed by study informants and participants is a crucial validating procedure that enhances construct validity. Alkin, Daillak, and White had informants from school districts review draft case studies and sometimes published the informants' insightful responses directly in their book, a practice mirrored in 17 case studies sponsored by the U.S. Office of Technology Assessment, where each case was reviewed by 20 to 40 or more outside reviewers.

• Share an early recognizable version of the draft with case informants and participants.
• Collect comments and feedback using an open-ended questionnaire or formal review process.
• Resolve factual discrepancies through a search for further evidence.
• Apply requested disguises or anonymity measures after facts are settled if required.

10 Key Ideas

1. Case study research is the preferred method when addressing how or why questions about contemporary events outside the researcher's control.
2. A rigorous case study design requires five specific components—including theoretical propositions and a defined unit of analysis—that connect initial questions to final conclusions and guide data analysis.
3. Case study research quality is judged using four established logical tests: construct validity, internal validity, external validity, and reliability.
4. A comprehensive case study protocol containing four specific sections is essential to guide data collection and increase research reliability.
5. Data triangulation ensures that findings are supported by multiple primary sources of evidence rather than a single source.
6. Creating a formal, orderly case study database that is distinct from the final report preserves raw data and increases reliability.
7. Researchers must establish an overall analytic strategy before collecting data to prevent case studies from stalling during the analysis stage.
8. Pattern matching compares an empirically based pattern found in the case study with a predicted pattern defined prior to data collection.
9. Having draft case study reports reviewed by study informants and participants is a crucial validating procedure that enhances construct validity.
10. Exemplary case studies require rigorous adherence to fair reporting of all evidence and accurate rendition of actual events without alteration.