# Lesson 1 — Foundations of Qualitative Data Analysis (v3 expanded)

*Companion-podcast transcript • Sarah & Kiffer*  
*~5,700 words • ~30 min audio*

---

**Sarah:** Welcome to Office Hours. I'm Sarah.

**Kiffer:** And I'm Kiffer. This is the first episode of a new arc on qualitative methods. It's a graduate-level course built around the Bernard, Wutich, and Ryan textbook, Analyzing Qualitative Data, Systematic Approaches, second edition. If you've been following the epidemiology track, this is the analytical companion that handles the data those courses politely refused to deal with.

**Sarah:** Interview transcripts, field notes, archival documents, focus-group recordings, free-list responses, the open-text comments at the back of every survey.

**Kiffer:** Exactly. The data your regression model can't ingest. And the working claim of this course is that there is a systematic, transparent, replicable way to analyze that material, and you should learn it the same way you learned to fit a logistic model. Not by vibe. By procedure.

**Sarah:** Before we get into the structure, give listeners an image of what qualitative data actually looks like.

**Kiffer:** The textbook opens with a scene that I keep coming back to. Imagine you're sitting at a kitchen table in Burnaby with an eighty-two-year-old man in long-term care, and he tells you that his loneliness is "the empty space all those people used to fill." That line is data. It has not been counted or scaled or coded yet. But it is empirical, it is patterned, and it carries information about what loneliness is for the person living it.

**Sarah:** And the problem of this course is how to get from a stack of statements like that one to defensible knowledge claims a public-health audience will actually believe.

**Kiffer:** Right. So this first lesson is foundational. It does four things. It gives you an operational definition of qualitative data analysis. It tells you why the line between qualitative and quantitative work is more porous than people pretend. It introduces four research goals that organize empirical work. And it walks through the five kinds of qualitative data and the three methodological commitments that anchor everything else.

**Sarah:** Okay, let's start with the definition. What is qualitative data analysis, operationally?

**Kiffer:** Bernard, Wutich, and Ryan open the textbook with a working definition. Qualitative data analysis is the search for patterns in non-numeric data and an explanation of why those patterns are there. There are three moving parts and each one does work.

**Sarah:** Search for patterns first.

**Kiffer:** Patterns are regularities. Co-occurrences, sequences, contrasts, gradients, absences. A theme that shows up in four of twenty transcripts is a pattern. A code that always appears immediately after another code is a pattern. A topic that no one mentions, even though you asked about it, is also a pattern. Sometimes a very informative one.

**Sarah:** And the second part, non-numeric data.

**Kiffer:** That's the easy part to say and the harder part to be precise about. A transcript is non-numeric. A photograph is non-numeric. A thirty-second clip of a focus group laughing in unison is non-numeric. So is the layout of a clinic waiting room. We'll be precise about the five kinds in a few minutes.

**Sarah:** And a piece of code, or a layout, or an objects-on-a-shelf inventory — those are all non-numeric in this sense.

**Kiffer:** That's right. The category is broader than people usually assume. And one of the things we'll keep coming back to is that "non-numeric" doesn't mean "cannot be counted." It means the data didn't arrive as numbers and the analytic move isn't built around variance and means.

**Sarah:** And the third part. Explanation.

**Kiffer:** This is the part that separates analysis from description. If you show me a pattern and stop there, that's description. It becomes analysis when you propose why. The textbook example I like: in the loneliness corpus we'll work with, the word "chair" shows up in eleven of twenty transcripts. Finding that is description. Proposing that chairs are the most stable physical traces of absent people in domestic space, and that's why participants reach for them, that's analysis.

**Sarah:** I notice the textbook doesn't open with philosophy. With ontology and epistemology and the interpretive turn.

**Kiffer:** They deliberately do not. Their stance is that qualitative analysis is best learned the way quantitative analysis is learned. By performing it on data, transparently, and being prepared to defend the moves you made. Doing over being.

**Sarah:** That stance reframes the whole intro for me. Because when you tell a quantitative-trained student "you're about to do qualitative work," they often hear "you're about to do something unrigorous." And the textbook is just refusing to let that framing stand.

**Kiffer:** Right. Their whole opening is a refusal of the framing. And they spend the next eight chapters demonstrating that disciplined qualitative work can be defended on the same kinds of methodological grounds as a regression analysis. Different grounds in detail. But same kind of defense.

**Sarah:** Let's stay on the meta question for a second. Because there's a longstanding fight about whether qualitative and quantitative are different substances. The textbook says they aren't.

**Kiffer:** The textbook is quite explicit that the line is more porous than the chapter headings around it suggest. Take a concrete example. The Canadian Community Health Survey is mostly pre-coded numeric items, but it also has a smaller set of open-text fields. Researchers routinely read those open-text fields, develop a coding scheme, count the resulting categories, and run chi-squared tests on the counts. They've just done qualitative analysis. They just didn't stop there.

**Sarah:** And the reverse happens too.

**Kiffer:** All the time. A grounded-theory study may begin with a frequency table of how many transcripts mention each emerging code. That's a quantitative move embedded inside a qualitative project.

**Sarah:** So if numbers versus words isn't the real distinction, what is?

**Kiffer:** The defensible distinction is about the type of question you're answering. Quantitative methods are typically the right choice when you want to estimate a population-level magnitude, test a pre-specified hypothesis, or measure an effect size. Qualitative methods are the right choice when you want to discover what something is, how people make sense of it, or how a process unfolds. Both can use counting. Both can use words. The deeper question is whether you're measuring a known phenomenon or characterizing an under-described one.

**Sarah:** That framing actually makes the whole "qual versus quant" war feel less existential. It's just two different jobs.

**Kiffer:** That's the textbook stance. Tools for different jobs. And the same study often uses both.

**Sarah:** And there's a third move the textbook makes, which is that mixed-methods work isn't a special hybrid. It's the normal case.

**Kiffer:** Almost any serious public-health study ends up using both, even when one is foregrounded. A trial of a behavioral intervention has a primary outcome that's quantitative — change in HbA1c, say — and a process evaluation that's qualitative. A surveillance system has count data and free-text comments from clinicians. The methods conversation should be about which tools answer which sub-question, not which camp you belong to.

**Sarah:** Okay, now the manifesto. Three commitments. Systematic, transparent, replicable. Walk me through each one operationally.

**Kiffer:** Systematic first. A systematic procedure has three features. It's specifiable, meaning you can write it down and another researcher can follow it. It's consistent, meaning the same rule is applied to every case. And it's iterative when needed but documented when revised. If you change your codebook halfway through coding, you note when and why, and you re-code the earlier transcripts under the new scheme.

**Sarah:** And the unsystematic version of the same activity reads as "we read the transcripts and noted what stood out."

**Kiffer:** Which is intuition, not analysis. The systematic version reads as "two analysts independently coded each transcript using a codebook developed from the first five transcripts; disagreements were resolved through discussion and the codebook was revised twice during the analysis, with revisions documented in the audit trail." That's the difference.

**Sarah:** Transparent next.

**Kiffer:** Transparency is what you owe your reader. Specifically four things. An explicit account of how you got the data, meaning sampling logic, recruitment, interview procedure. An explicit account of how you analyzed the data, meaning coding procedure, the codebook, how disagreements were handled. An explicit account of your positionality, who you are and what you brought to the interpretation. And an explicit account of the limitations, what your design and dataset cannot tell you.

**Sarah:** And those four belong in the methods section or appendix, not as an afterthought in the discussion.

**Kiffer:** That's the convention in contemporary public-health qualitative work, and it's the standard we hold capstones to.

**Sarah:** Then replicable, which is the contested one.

**Kiffer:** Right. Replicability in qualitative work does not mean two analysts working on the same dataset will produce identical interpretations. The textbook is clear that interpretation is partly perspectival. What replicability does mean is that two analysts following the same procedure would produce defensible interpretations and would identify similar patterns. The standard is "another competent researcher would arrive somewhere coherent with mine." Coherent, not identical.

**Sarah:** That's the part that feels least intuitive coming from a quantitative background. Because in regression world, replicability means bit-for-bit reproducibility of the same number.

**Kiffer:** And one of the unlearnings of this course is accepting that "coherent" is a defensible standard. Documentation of analytic choices is the operational goal, not identity of conclusions. There's a real intercoder-reliability check we'll do later in the term, where two analysts independently apply the codebook and we compute agreement. That's one operational test of replicability. But it isn't the only one, and for some interpretive methods it's actually the wrong test.

**Sarah:** Why does the textbook fight so hard for these three words?

**Kiffer:** Because the public-health audience for this work has been trained to ask methodological questions about quantitative studies. How was the sample drawn? What was the case definition? What's the confidence interval? Most of them have not been trained to ask the same kinds of questions about qualitative work. The Bernard, Wutich, and Ryan stance is that you should welcome those questions and have answers for them. Being systematic, transparent, and replicable isn't about philosophical purity. It's about making sure your qualitative work gets taken seriously by the audiences who decide policy.

**Sarah:** Some traditions push back on this. The fully constructivist position is that knowledge is co-constructed and pretending otherwise is dishonest.

**Kiffer:** And the textbook agrees with the co-construction part. They just reject the implication that systematic methods are inappropriate because of it. Their position, and the position of this course, is that transparency about subjectivity is the modern operational solution. You say what you brought, you make your moves visible, and you let the reader judge. You don't dissolve method into pure interpretation.

**Sarah:** Okay, let's talk about why a public-health researcher would do this work in the first place. Because that's a question that comes up in every first week of a qualitative methods course.

**Kiffer:** And the honest answer is that for many of the most important public-health questions, qualitative work is the only way in. Three examples from the lesson. Vaccine refusal. You can count refusals with surveys, you can identify predictors with logistic regression. But if you want to understand the specific arguments people give themselves and each other for refusing — the narratives, the framings, the felt experience of distrust — you need interviews. The qualitative literature on vaccine hesitancy is what produced the interventions the quantitative literature later tested.

**Sarah:** Second example, living with a chronic illness.

**Kiffer:** The phenomenology of type two diabetes or long COVID is not legible in administrative data. Patient-reported outcome measures give you a score. Qualitative work tells you what the score is a measure of. A quality-of-life score of nought point six two is a number. The interviews behind that scale are what give the number its content.

**Sarah:** And the third one is implementation science.

**Kiffer:** Implementation science is the study of why evidence-based programs work in trials but stall in real-world rollout. It's heavily qualitative. The numerical fact that a program failed is the starting point. The reasons are uncovered through interviews with implementers, observation of practice, document analysis of organizational policy. Why did the program fail is a qualitative question, even when the failure itself is measured quantitatively.

**Sarah:** I'd add a fourth example, which is health equity work generally. Because so much of equity research is about how people from marginalized groups experience and explain the structural conditions they're inside.

**Kiffer:** Yes. And that's where qualitative work isn't just one tool among others, it's often the only ethical way to do the work at all. A survey that asks racialized patients about their experience of healthcare with five Likert items, and then publishes the means, has flattened something that needed to be heard. Qualitative work lets the people who are subject to a system tell you what the system is doing to them in their own categories. That's a research-ethics argument as much as a methodological one.

**Sarah:** And it loops back to the textbook's emphasis on transparency about positionality. Because if you're doing equity work without naming who you are, you're hiding part of the analytic apparatus.

**Kiffer:** Exactly. Positionality is a methodological tool, not a confession.

**Sarah:** Okay, second big chunk. The four research goals. This is the framework that locates qualitative work in the broader landscape of empirical research.

**Kiffer:** Bernard, Wutich, and Ryan organize the whole enterprise of empirical research around four goals. Exploration, description, comparison, and the testing of theoretical models. Every study you've ever read can be located in one of these four, or more commonly in two or three of them at once. They are not a hierarchy. They are not phases of a single study. They are different jobs that empirical research can do, and each has its preferred methods.

**Sarah:** Start with exploration.

**Kiffer:** Exploration is what you do when you don't yet know enough about a phenomenon to make a hypothesis about it. The goal is to map the territory. To find out what is there, what the relevant categories are, what people consider important, what the underlying mechanisms might be. The textbook is explicit that qualitative work dominates exploration. Because quantitative methods generally require that you've already decided what the variables of interest are, and exploration is the work of deciding that.

**Sarah:** And most public-health questions actually begin in an exploratory phase, even if they later become hypothesis-testing.

**Kiffer:** A new pathogen emerges. A previously invisible population's experience comes onto the policy agenda. A digital harm like cyberbullying or AI-mediated relationships appears that no existing survey can ask about. The first scholars to study any of those are doing exploration. Their job is to give the rest of the field something to measure.

**Sarah:** Description next. And the textbook is emphatic that description is not a consolation prize.

**Kiffer:** That's right. Description gets treated, in some methods textbooks, as the work you do when you can't do anything "real." The textbook authors reject that emphatically. Many of the most influential studies in public health are descriptive. The Framingham Heart Study began as description. The BC Centre for Disease Control overdose mortality reports are description. The entire field of demography is description.

**Sarah:** And both qualitative and quantitative methods can do description, just of different aspects.

**Kiffer:** A quantitative survey can describe what percentage of adults in BC report being lonely in the past year. That's description by counting. A qualitative study can describe what loneliness feels like from the inside, what triggers it, what people do about it, and how they make sense of it. That's description by characterization. Both are valid. Often both are necessary.

**Sarah:** Third goal, comparison.

**Kiffer:** Comparison is what you do when you have characterized a phenomenon and now want to know how it varies across groups, settings, or conditions. The classical home of comparison in epidemiology is the case-control or cohort design — does the exposure differ between cases and non-cases, does the outcome differ between exposed and unexposed. Comparison is more often associated with quantitative work because of the statistical machinery, but qualitative comparison is real and rigorous.

**Sarah:** Like grounded theory's constant-comparative method.

**Kiffer:** Right. And qualitative comparative analysis, and matrix analysis. These are systematic ways to compare across cases — with words instead of variables — and to draw defensible inferences about why groups differ. The comparison is qualitative when the units being compared are texts or interpretive cases rather than rows in a spreadsheet, and when the conclusions are about patterns of meaning rather than effect sizes.

**Sarah:** Fourth goal, testing models. The one most introductory textbooks treat as the pinnacle.

**Kiffer:** You have a theory, you derive predictions, you check the predictions against data. That's the standard confirmatory logic. Qualitative work can do this too, though it's rarer and more contested. Analytic induction is a qualitative model-testing approach. You specify a hypothesis, you examine your cases, you find a case that doesn't fit, and you revise the hypothesis until it fits all cases. Qualitative comparative analysis tests Boolean propositions about combinations of conditions sufficient for an outcome. We'll come back to both later in the term.

**Sarah:** And a real study often hits more than one goal.

**Kiffer:** Almost no real study fits cleanly into one of the four. The original Cacioppo and Patrick work on loneliness explored what loneliness is, described its physiological correlates, compared lonely and non-lonely adults, and tested specific neurobiological models. Most of the capstone work in this course will be predominantly exploratory and descriptive, but you shouldn't avoid comparison or model-testing if your data support them.

**Sarah:** Worth flagging that the four goals aren't a developmental sequence either. It's not that you have to do exploration first, then description, then comparison, then model-testing.

**Kiffer:** Right. A field can do exploratory work for decades before any comparative work happens, or it can do all four in parallel. The goals are jobs, not stages. The reason to name them is so you can tell your reader which job your study is doing, not so you can sequence your career.

**Sarah:** Okay, the third big chunk. Five kinds of qualitative data.

**Kiffer:** Bernard, Wutich, and Ryan organize qualitative data into five kinds. Physical objects, still images, sounds, moving images, and texts. The categorization might feel pedantic until you realize that the kind of data you have shapes what analytic moves are available to you. A photograph and a transcript both look like qualitative data, but they are coded differently, sampled differently, and reported differently.

**Sarah:** Walk me through the five.

**Kiffer:** Physical objects first. Anthropologists call this material culture. The artefacts people make, use, exchange, and discard. In public health, that includes medication packaging, syringe-exchange kit contents, vaccine cards, mobility aids, the layout of a clinic waiting room, the contents of someone's medicine cabinet. Object-based analysis isn't common in mainstream public-health research but it's increasingly important in implementation science, environmental health, and Indigenous health research where physical context carries meaning that words don't.

**Sarah:** Second, still images.

**Kiffer:** Photographs, hand-drawn maps, satellite imagery, screenshots of social-media posts, anatomy diagrams in patient education materials, public-health campaign posters. Visual sociology and visual anthropology have their own techniques. The photovoice method, where participants take their own photographs and discuss them, is a staple of community-based participatory research.

**Sarah:** Third, sounds.

**Kiffer:** Recorded speech is the most common, but other audio is analytically tractable too. The sound of a clinic at peak hours, the music played in a hospice, the auditory environment of a school cafeteria. But the vast majority of qualitative analysis in health research starts with sound — an interview recording — and is converted to text — a transcript — before analysis. And that conversion step is itself an analytic act. We'll spend serious time on transcription conventions later in the course.

**Sarah:** Fourth, moving images. Video.

**Kiffer:** Video is sound plus image plus time. Clinical encounter recordings, simulated training scenarios, ethnographic field recordings, TikTok health-influencer content, telehealth call recordings. More time-consuming than audio or text, but it lets you attend to non-verbal communication, embodied action, and spatial arrangement. Conversation analysis has historically privileged video for exactly this reason.

**Sarah:** And fifth, texts. Which is most of what this course works on.

**Kiffer:** By far. Interview transcripts, focus-group transcripts, field notes, documents — policy papers, clinical guidelines, news articles — open-text survey responses, social-media corpora, patient-generated text like diaries and illness blogs, and free-list responses for cultural domain analysis. Texts dominate this course partly for practical reasons. They are cheap to store, easy to share, computationally tractable. And partly for a principled reason. Text is the form most amenable to the systematic-transparent-replicable analytic procedures the textbook advocates. The methods we'll learn will apply to images, video, and audio with adjustments, but the default unit of analysis is text.

**Sarah:** And the capstone dataset is text. Tell listeners about it.

**Kiffer:** The capstone is anchored by a set of twenty transcripts of semi-structured interviews about experiences of loneliness, conducted with adults in British Columbia. The twenty participants vary deliberately across age — eighteen to eighty-two — across gender, life-stage, immigration status, caregiving role, identity. Students, parents, retirees, widows, recent refugees, a man who came out at sixty after a long heterosexual marriage. The variation is engineered.

**Sarah:** And one important note on the dataset.

**Kiffer:** The twenty transcripts are fully synthetic composites developed for instructional use. They draw on themes documented in the published loneliness literature, but no individual transcript represents a real person. You treat them as you would any qualitative dataset for the purpose of learning the methods. The capstone paper just has to acknowledge in its methods section that the data are an instructional dataset.

**Sarah:** Why synthetic?

**Kiffer:** Because consenting twenty real people to have their interview transcripts used as a teaching dataset across multiple cohorts of graduate students is a research-ethics nightmare. Synthetic composites let us teach the methods on a corpus that's analytically coherent and pedagogically rich without putting any real participant in the room.

**Sarah:** Okay, fourth big chunk. The toolchain. R plus Taguette.

**Kiffer:** Same R environment students used in the epidemiology courses, but different packages. The text-analysis ecosystem in R is mature. The core stack we install in week one is tidyverse for general data wrangling, tidytext for text-as-data verbs, quanteda for industrial-strength text analysis, stringr for text manipulation, readtext for reading text corpora, igraph for network analysis, topicmodels for L D A topic modeling, and i r r for intercoder reliability statistics.

**Sarah:** And then Taguette as the second tool.

**Kiffer:** Taguette is a free, open-source qualitative coding application. It does what NVivo and ATLAS dot t i do — highlight passages, attach codes, build a codebook, export coded extracts — without the licence fee. It runs in your browser. The recommendation is to set up an account, create a project, upload a transcript, and just get familiar with the interface in week one. Configuring software the same week you're trying to learn content is a setup for misery.

**Sarah:** Both free, both open-source, both transferable beyond this course.

**Kiffer:** That's the design choice. We want students walking out with tools they can actually use in a postdoc or a public-health job without licensing barriers.

**Sarah:** And there's a working principle behind that choice. You want the analytic record to be inspectable.

**Kiffer:** Right. Free and open-source software produces files in formats other researchers can read, even years later. If your dissertation analysis is locked inside a proprietary file format whose company has been acquired or shut down, that's a transparency failure. Reproducibility in qualitative work depends partly on tool choice, and we're trying to teach the right defaults.

**Sarah:** Let's talk about the capstone itself, briefly, and the first milestone.

**Kiffer:** The capstone is a journal-article-format paper. Introduction, methods, findings, discussion, references. It reports a qualitative analysis of the loneliness dataset, or some subset of it that fits the student's specific research question. The methods section is going to be the heaviest in the paper, because that's where the three commitments are most visible.

**Sarah:** And every week of the course advances the capstone by one concrete milestone.

**Kiffer:** Right. The first milestone is small but real. Read the interview guide, read three transcripts of your choice, and write a five-hundred-word positionality memo. Plus install R and Taguette.

**Sarah:** What does the positionality memo contain?

**Kiffer:** Four things. Your own working definition of loneliness, before you read the transcripts. Your relevant social location — age, gender, family of origin, professional role, prior experience with loneliness. What you noticed in the three transcripts. And what you suspect you might over- or under-read, given who you are.

**Sarah:** That last one is the hardest.

**Kiffer:** It is. Because it asks you to name in advance the places where your own positioning might bias your interpretation. And students who come from a quantitative epidemiology background often haven't been trained to do that kind of explicit self-naming in a methods section. It feels strange the first time. But it's a core piece of the transparency commitment, and you only get good at it by doing it.

**Sarah:** Let me ask a meta question. For someone who's coming into this course from regression and surveillance and outbreak investigation, what's the thing they have to set aside?

**Kiffer:** The instinct to reach immediately for population-level claims. Qualitative work on twenty transcripts cannot tell you what percentage of British Columbians experience loneliness. It can tell you what kinds of loneliness exist, what they look like from the inside, what mechanisms might be at work. Letting go of the prevalence question while you're doing this work is the hardest unlearning.

**Sarah:** What's the second thing?

**Kiffer:** The wish for an unambiguous result. Qualitative findings come with interpretive ranges and competing readings that should be reported, not collapsed. A paper that says "we found three themes" and then leaves no room for the reading that emphasizes a different theme is doing rhetorical work that the data don't license.

**Sarah:** And what comes with from the quantitative training that actually helps?

**Kiffer:** A lot, honestly. The discipline of pre-specification — writing your analytic procedure before you run it — translates almost directly from quantitative epidemiology to qualitative analysis. The habit of declaring limitations explicitly. The methodological seriousness — treating your method as something you must defend — is exactly what Bernard, Wutich, and Ryan are asking for. The epidemiology track actually sets students up well to do rigorous qualitative work, as long as they're willing to expand what counts as rigor.

**Sarah:** Let me try to pull the lesson together. First takeaway. Qualitative data analysis is defined operationally — the search for patterns in non-numeric data and an explanation of why those patterns are there. Description without explanation isn't analysis.

**Kiffer:** Second. The qualitative-quantitative boundary is porous. The deeper distinction is between magnitude questions and characterization questions, not between numbers and words. Both methods can use counting. Both can use language.

**Sarah:** Third. There are four research goals — exploration, description, comparison, and testing models. Qualitative work dominates exploration and description, and contributes seriously to the other two.

**Kiffer:** Fourth. There are five kinds of qualitative data — physical objects, still images, sounds, moving images, and texts. Texts dominate health research for practical and principled reasons, and the capstone dataset is a text corpus of twenty synthetic loneliness transcripts.

**Sarah:** Fifth. Three methodological commitments anchor the course. Systematic — specifiable, consistent, documented. Transparent — data, procedure, positionality, limitations. Replicable — coherent with, not identical to, a second analyst's interpretation.

**Kiffer:** Sixth. The toolchain is R — tidyverse, quanteda, tidytext, igraph, i r r — plus Taguette. Both free, both open-source.

**Sarah:** Seventh. The first capstone milestone is a five-hundred-word positionality memo and a working toolchain setup. The positionality work is itself a piece of the transparency commitment, not a soft warmup.

**Kiffer:** And eighth. The point of all this discipline is not philosophical purity. It is so that the qualitative work you publish is taken seriously by the public-health audiences who decide policy. Rigor in qualitative work looks different from rigor in regression. But it is rigor, and it is teachable.

**Sarah:** And ninth, which is a meta-takeaway. The course will keep returning to this triangle of doing over being, procedure over intuition, and transparency over claim. Every method we cover is going to be evaluated on that triangle.

**Kiffer:** And the field's most respected qualitative work, the work that gets cited across disciplines, sits where all three corners are strong. Disciplined procedures, clearly described. Honest positionality. Findings that another competent analyst could reproduce coherently. That's the target.

**Sarah:** Anything you want listeners to carry forward into the next lesson?

**Kiffer:** One thing. The line at the kitchen table — the empty space all those people used to fill — is the kind of sentence this course is about. We treat it the way an epidemiologist treats a case. Carefully, systematically, with method. That's the whole posture.

**Sarah:** Next lesson, we move from definitions to design. Research questions, theory, and the literature. How to translate a public-health concern into a focused, defensible qualitative research question, and how to read the existing qualitative evidence well enough to position your own work in it.

**Kiffer:** Before the next session, install the toolchain, set up Taguette, read the interview guide and three transcripts, and draft the positionality memo. Bring the questions that didn't resolve.

**Sarah:** Thanks for listening. We'll see you in Lesson 2.

**Kiffer:** Take care of yourselves. See you in class.