Research Methods for Everyone, Part 5: Case Studies, Fieldwork, and Research That Acts

In this post, the fifth in a six-part series on Amanda M. Rosen's Great Courses series Effective Research Methods for Any Project, I follow lectures fourteen through sixteen into the tools of depth: case studies, field research and the interpretivist philosophy behind it, and the applied research traditions that set out not just to describe the world but to change it.^[1] Part four covered experiments, surveys, and polls, the tools of breadth. For those of us in the humanities, depth is home territory, and watching a political scientist map it with a methodologist's eye is instructive. I will argue that each depth tradition makes a trade (generalizability for mechanism, the single account for plural meanings, detachment for effectiveness), and that the difference between rigor and malpractice lies in declaring the trade.

The case study and its temptations #

Rosen opens with a case that launched a research program: Kitty Genovese, stabbed to death outside her New York apartment in 1964 while, the story goes, nearly forty neighbors watched and did nothing. That story generated decades of research on the bystander effect. The story is also, Rosen notes, mostly false. Later investigation found fewer witnesses, most of whom saw nothing, and some who called out or called for help. The murder was real; the forty silent watchers were "mostly urban legend."^[2] The case makes a fitting overture because it shows both faces of the case study at once: a single vivid case can open an entire field of inquiry, and a single vivid case, once it takes on a life of its own, can mislead an entire public for decades.

A case study is an analysis of one or a small number of cases aimed at in-depth, multifaceted understanding in real-life context, using multiple sources of evidence. Its structural predicament, in a phrase Rosen returns to, is that it has many more variables than cases, so it trades away generalizability (external validity) for depth. What keeps it rigorous is systematic data collection, triangulation across sources, and above all disciplined case selection. Here she leans on Robert Yin's five categories: 1) critical cases, which any adequate theory must explain ("if you have a theory explaining revolutions, you need to be able to explain the French Revolution"); 2) extreme or deviant cases, outliers like the diagnostic puzzles that structured the television show House; 3) representative cases, the typical instances findable by z-scores near zero; 4) revelatory cases, newly accessible for the first time, like a 2015 study of the Kenya-Netherlands rose supply chain; and 5) longitudinal cases, studied at two or more points in time.^[3]

Case studies also serve purposes short of full theory-testing. They generate hypotheses that larger studies can then examine, and they work as plausibility probes, pilot tests of an argument on an initial case before committing to a bigger design. Rosen's examples of the case study in the wild are refreshingly ordinary. A medical teaching case from Johns Hopkins describes a 77-year-old dementia patient whose agitation and abuse toward his wife resolved within a month once depression was diagnosed and treated, a single case that teaches clinicians to look for the treatable condition hiding inside the untreatable one. A 2010 study followed electronic health record adoption across five sites of the UK's National Health Service and traced the delays to conflicts between central government and local implementers. Neither generalizes cleanly; both teach something no regression could.

Comparative case studies extend the logic, choosing most-similar or most-different systems to mimic experimental control through case choice. The traps are all traps of selection. Choose Fiji, Samoa, the Maldives, and the Seychelles and you can "prove" warm countries never fight wars, so long as you never glance at Vietnam, Algeria, or Iraq. Rosen's sternest rule concerns circularity: you cannot build a theory from North Korea's nuclear program and then test the theory on North Korea. A theory tested on the case that generated it "is built to pass the test." That single sentence would improve half the qualitative dissertations ever defended.

The lecture also revisits the Perestroika movement from a new angle. In 2000, an anonymous "Mr. Perestroika" email attacked political science's marginalization of qualitative and area studies work, tying it to the field's failure to anticipate the Soviet collapse. The methodological monoculture, the argument runs, had optimized itself away from exactly the deep case knowledge that mattered. Case studies are not quantitative research's poor cousin; they answer questions breadth cannot reach.

Interpretivism, or the philosophy under the fieldwork #

Lecture fifteen begins with the course's most consequential fork in the road. Positivism asserts that there is an objective, observable reality we can study, the assumption beneath everything from part one's scientific method onward. Interpretivists, Rosen explains, "argue that there is no single objective reality," that human meaning comes in multiple layers and multiple truths, and that the researcher's task is interpretation rather than measurement. Her example is unsparing: after a sexual encounter, one party may sincerely understand consent to have been given while the other experienced fear of saying no. "The objective truth is that a sexual encounter occurred," she says, "but that hardly captures the complete story." Choosing interpretivism, she adds, is "more like choosing a philosophy than a design."^[4] What I respect in her treatment is that she declines to caricature either side. She does not withdraw the positivist toolkit of the earlier lectures; she places it, revealing it as one philosophical option (one cosmovision among others, in my vocabulary) rather than the definition of rigor itself.

Field research is where interpretivism does its best work. The field is simply wherever the data lives, Bali or your backyard or your office. Ethnography, dominant in anthropology, embeds the researcher in a community for participant observation, and its patron saint is Clifford Geertz. Rosen retells the famous story: at an illegal Balinese cockfight raided by the authorities, Geertz and his wife ran with the villagers and joined the collective pretense of having been elsewhere, and that shared flight transformed him from tolerated outsider to accepted presence. He earned the access behind his "thick description" in that moment of participation.^[5]

The craft details are the lecture's quiet gift. Structured interviews are functionally surveys; unstructured interviews hand control to the subject; the semi-structured interview is the workhorse. Recording improves accuracy but risks reactivity. Snowball sampling reaches hidden populations but delivers your subject's friends and allies, "so you might need several snowballs." And fieldwork is hostage to the field. Rosen planned a May 2010 London trip to interview policymakers, only to have a general election called for her arrival week, which produced the first hung Parliament in over thirty years and a week of staring at her inbox. "It was mostly a wasted trip." Every fieldworker owns a version of that story, and pretending otherwise is how methods sections lie.

Research that acts #

Lecture sixteen turns from understanding the world to changing it, and it completes the basic-versus-applied distinction from the course's first lecture. Evaluation research assesses the need for, design, progress, and outcomes of a particular program. It comes in three moods: needs assessment before a program exists, formative assessment while it runs, and summative evaluation after, which Rosen describes as "essentially giving the program a grade." She is frank about the politics. Evaluators rarely choose their own topics, and "jobs may depend on whether or not you recommend the initiation or termination of a program." Rigor here is not academic scruple; it is a bulwark for the people whose livelihoods ride on the verdict.^[6]

Action research goes further, finding collaborative solutions to real-world problems with a community, whose members join the research team itself. It is cyclical rather than linear, "a democratized type of research," central to education, healthcare, and human rights work, and Rosen distills Ernest Stringer's framework to three words, "Look, think, act," repeated until the problem yields.^[7] She is also honest about the tensions, noting that "democracy in research isn't always a good thing. There can be too many cooks in the kitchen."

The applied family also includes market and product research, and Rosen's treatment of the focus group doubles as consumer education. A focus group typically gathers ten to twelve people, is nowhere near representative, and lives or dies on the skill of its moderator, since one dominant voice can swamp the room and the fear of social repercussions can silence honest reactions. Online focus groups lower those inhibitions and widen the geography. She notes wryly that when textbook publishers demonstrate e-textbooks to professors over a catered lunch, the focus group is simultaneously research on her and a sales pitch to her; research and commerce are here flip sides of the same coin, and pretending otherwise is its own methodological error. Doing applied work well means the same disciplines as ever: define the problem, set key indicators, establish baselines, and set targets you can actually measure, whether the goal is weight loss or a functioning after-school program.

Then comes the course's most charming moment. Rosen goes meta: this very course, she reveals, is a product of action research. The Great Courses identified a gap in its catalog, consulted its customers about whether they wanted research methods training, and built the course in response, with feedback loops closing the cycle. The content itself was not democratic ("As the expert, I made the decision about what content to cover"), but the project was collaborative in conception, and even its form reshaped her teaching away from visuals, since some customers listen audio-only. The reveal is a small masterclass in reflexivity, the researcher turning her instruments on her own practice.

What depth is for #

Reading these three lectures as a unit clarifies what the depth traditions actually offer, and it returns me to my thesis. The case study trades generalizability for mechanism, for seeing how causes work rather than merely that they correlate. Interpretivism trades the single authoritative account for fidelity to the plurality of meanings people actually live within. Action research trades detachment for effectiveness, accepting a stake in the outcome as the price of mattering to the community studied. Each trade is legitimate, and the researcher must declare each one. The failures Rosen catalogs (the cherry-picked island comparison, the theory tested on its own origin, the fieldwork that becomes advocacy without saying so) are all failures to name the trade being made.

Humanists should feel both flattered and challenged here. Flattered, because she treats the interpretive traditions we work in as serious methodology rather than soft alternatives; challenged, because Rosen's standards (purposive selection you can defend, variation in your variables, honesty about access and failure) apply to an archive or a text corpus exactly as they apply to a Balinese village. Depth is not an exemption from rigor. Depth is a different rigor.

In the final post, I will take up what to do with data once you have it, from descriptive statistics through regression, the craft of qualitative coding, and why presenting your findings is part of the method, not an afterthought.

Notes #

---

1. Amanda M. Rosen, Effective Research Methods for Any Project (Chantilly, VA: The Teaching Company/The Great Courses, 2019). This post draws on Lecture 14, "Research by Case Study," Lecture 15, "Interpretivism and Field Research," and Lecture 16, "Applied, Evaluative, and Action Research." ↩︎

2. Rosen, Effective Research Methods, Lecture 14. On the debunking, see Kevin Cook, Kitty Genovese: The Murder, the Bystanders, the Crime That Changed America (New York: W. W. Norton, 2014). ↩︎

3. Robert K. Yin, Case Study Research: Design and Methods, 5th ed. (Thousand Oaks, CA: Sage, 2014), discussed in Rosen, Lecture 14. ↩︎

4. Rosen, Effective Research Methods, Lecture 15. ↩︎

5. Clifford Geertz, "Deep Play: Notes on the Balinese Cockfight," in The Interpretation of Cultures (New York: Basic Books, 1973). ↩︎

6. Rosen, Effective Research Methods, Lecture 16. ↩︎

7. Ernest T. Stringer, Action Research, 4th ed. (Thousand Oaks, CA: Sage, 2014), discussed in Rosen, Lecture 16. ↩︎