Ethnography in Qualitative Research: Types & Steps

Last updated: November 8, 2026

Ethnography in qualitative research is the method you choose when the question is about how a group actually lives, works or makes sense of something, and when what people say in an interview room is unlikely to be the whole story. It started in anthropology, with long immersion in a community, and is now used in sociology, education, health research, organization studies, consumer research and HCI.

This guide is written for PhD students, research labs and qualitative research teams designing an ethnographic study in 2026. It covers the main types, the core methods, a step-by-step process, how to write fieldnotes, and where AI tools help and where they do not. It also includes an original Ethnographic Design Decision Matrix you can adapt for a proposal or ethics application.

Key takeaways

• Ethnography combines participant observation, informal and ethnographic interviews, documents and artifacts, recorded in fieldnotes over a sustained period.
• Main types include classic (realist), critical, focused, rapid, digital or online (including netnography), multi-sited, institutional and autoethnography.
• Your fieldnotes are the primary data. Write them the same day, separate description from interpretation, and keep an analytic memo trail.
• Rapid and focused ethnographies trade depth for speed. A 2018 systematic review in BMJ Quality & Safety flagged sampling, interpretation and field management as recurring challenges.
• AI can help with transcription, ethnographic interviews at scale and first-pass coding, but a January 2026 benchmark found the best LLM setups reached a mean F1 of only 0.41 when annotating ethnographic texts.

What is ethnography in qualitative research?

Ethnography is a qualitative research design in which the researcher studies a social group in its natural setting, over time, to understand its practices and meanings from the members' point of view. The word refers both to the fieldwork process and to the written account that comes out of it.

Three features separate ethnography from other qualitative designs:

• Presence in the setting. Data come from being where things happen, not only from people's reports about it. Malinowski's Argonauts of the Western Pacific (1922) set the template of long-term fieldwork.
• Multiple data sources. Observation, conversation, interviews, documents, objects, photos and, increasingly, digital traces are combined.
• Interpretive writing. The output is an account of what actions mean in context, what Geertz (1973) called thick description.

Hammersley and Atkinson's textbook Ethnography: Principles in Practice remains the standard reference for the design as a whole. For fieldwork technique, most methods courses still assign Spradley's The Ethnographic Interview (1979) and Participant Observation (1980), and Emerson, Fretz and Shaw's Writing Ethnographic Fieldnotes.

Types of ethnography

The type you choose decides how long you stay, how you enter the field and what counts as data. The table below summarizes the main variants.

Type	What it is	Typical fieldwork	Good fit when
Classic / realist	Long immersion; the researcher writes an authoritative account of a culture	Many months to years	A doctoral thesis on a community or organization
Critical	Ethnography that examines power, inequality and whose voice counts	Months to years	Research questions about marginalization or institutions
Focused	Short, intensive visits on one specific practice or activity (Knoblauch, 2005)	Days to weeks, often repeated	A defined practice in a familiar setting, such as a hospital handover
Rapid	Compressed fieldwork and analysis, often in teams, for timely findings	Days to a few months	Health services or policy evaluation with deadlines
Digital / online	Ethnography of and through digital media; includes netnography of online communities	Weeks to months, online and offline	Communities, platforms or practices that live partly online
Multi-sited	Following people, objects or ideas across several locations (Marcus, 1995)	Months, across sites	Supply chains, migration, global product use
Institutional	Tracing how texts and rules organize everyday work (Dorothy Smith)	Months	How policies shape frontline practice
Autoethnography	The researcher's own experience as the data, linked to culture	Variable, retrospective and ongoing	Experiences that are hard to access from outside

Rapid ethnography

Rapid ethnographies keep the attention to social theory, cultural meaning and the insider view, but run collection and analysis over a shorter period, as Cecilia Vindrola-Padros explained in a 2020 Cambridge University Press blog post. Her 2018 systematic review of rapid ethnographies in healthcare found recurring difficulties with sampling, interpreting findings and managing fieldwork. If you go rapid, plan for those three problems explicitly.

Digital ethnography and netnography

Pink and colleagues' Digital Ethnography: Principles and Practice (SAGE, 2016) proposes five principles: multiplicity, non-digital-centric-ness, openness, reflexivity and unorthodoxy. "Non-digital-centric" is the useful warning: study what people do with media, not the media in isolation. Netnography, developed by Robert Kozinets, adapts ethnography to online communities and cultures; Kozinets discusses how the method evolved in a 2020 interview with Henry Jenkins.

Core ethnographic methods

Participant observation

Participant observation means taking part in everyday activities while observing them. Gold's (1958) classic typology runs from complete participant through participant-as-observer and observer-as-participant to complete observer. State your role in the methods section and say how it changed over time; it affects what people showed you.

Ethnographic interviews

Ethnographic interviews range from informal conversations during fieldwork to scheduled interviews. Spradley distinguished descriptive questions ("Can you walk me through a typical shift?"), structural questions ("What kinds of patients are there?") and contrast questions ("What is the difference between a quiet night and a bad night?"). Good probing questions keep these grounded in specific episodes.

Documents, artifacts and digital traces

Rotas, forms, chat threads, photos, posters and objects show how a setting is organized. Treat them as data with their own context: who made them, for whom and why.

Fieldnotes and memos

Fieldnotes are the primary data of ethnography. Jottings in the field become full notes the same day. Keep interpretation separate from description, and write analytic memos as patterns appear.

How to do ethnography: 8 steps

1. Frame a question about practice or meaning. "How do night-shift nurses decide when to escalate?" suits ethnography; "How many nurses escalate?" does not.
2. Choose the type and the site. Use the decision matrix below. Justify why this site, and whether it is one site or several.
3. Negotiate access and ethics. Identify gatekeepers, plan consent for people you observe but do not interview, and decide how you will handle public versus private online spaces. See our guide to informed consent.
4. Enter the field and define your role. Record your position and assumptions from day one; this is where reflexivity starts.
5. Collect data in parallel streams. Observation, conversations, interviews and documents, logged with date, place and context.
6. Write fieldnotes daily and memo weekly. Use the checklist below.
7. Analyze iteratively. Code fieldnotes and transcripts, compare cases, look for negative cases and return to the field to test emerging interpretations.
8. Write the ethnography. Combine description, participants' voices and your interpretation, and show how you know what you claim.

The Ethnographic Design Decision Matrix

This is an original Qualitati planning tool. Score each row for your study, then read across to the design that best fits.

Decision factor	Points toward classic / critical	Points toward focused / rapid	Points toward digital / multi-sited
Time available for fieldwork	Six months or more	Under three months	Flexible, spread over time
Familiarity with the setting	Low; you need to learn the culture	High; the practice is defined	Mixed; the setting is distributed
Where the practice happens	One bounded place	One place, one activity	Online, or across several places
Who needs the findings, and when	Academic audience, thesis timeline	Decision-makers with a deadline	Mixed audiences
Team size	Usually a single researcher	Often a team, with shared synthesis	Single researcher or team across sites
Main risk to manage	Going native, scope creep	Shallow interpretation, thin sampling	Losing context, consent in online spaces

Fieldnote quality checklist

• Date, time, place and who was present are recorded for every entry.
• Concrete description comes before any interpretation, and the two are visibly separated.
• Direct quotes are marked as verbatim; paraphrases are marked as paraphrases.
• Your own role, feelings and influence on the scene are noted.
• Questions to follow up next time are listed at the end.
• Pseudonyms and identifiers are handled consistently with your data management plan.

Ethnography vs other qualitative designs

Ethnography overlaps with other designs, and examiners often ask why you did not choose one of them.

• Ethnography vs case study: a case study is bounded by the case and can use any methods; ethnography is defined by immersion and cultural interpretation. An ethnographic case study combines both.
• Ethnography vs interview study: interviews capture accounts; ethnography also captures action, context and the gap between what people say and do.
• Ethnography vs grounded theory: grounded theory is an analytic approach aimed at building theory; it can be used on ethnographic data.
• Ethnography vs diary study: diaries capture self-reported, in-the-moment experience when you cannot be present; they can complement fieldwork.

Can AI help with ethnography?

Bottom line: AI can speed up the parts of ethnography that are about handling text, but not the parts that depend on being there.

The clearest recent evidence comes from a January 2026 study by Goodall and colleagues that tested 7 large language models on 567 ethnographic excerpts. The best setups reached a mean F1 of 0.41. Models did better on concrete, explicitly described features and worse on interpretive ones, and different LLMs agreed with each other far less than two human coders did (κ = 0.23 vs κ = 0.57). We summarized the study in Can LLMs annotate ethnographic texts?

Ethnography is also being used to study AI itself. A February 2026 HCI paper by Szymanski and colleagues reports a 12-week ethnographic study of a team building a pedagogical chatbot.

A practical division of labor:

• Reasonable to delegate: transcription of recorded interviews, translation drafts, organizing fieldnotes, first-pass descriptive coding, searching across a large corpus.
• Keep human: access and relationships, observation, deciding what matters in a scene, interpretation, ethical judgment and the written account.

Where Qualitati fits

Qualitati is a European, budget-friendly qualitative research platform for universities and research companies, built with data privacy and GDPR as central priorities. It does not do participant observation, and it should not replace fieldwork. It can support the interview and analysis parts of an ethnographic project:

• Ethnographic interviews at a distance or at scale: AI-moderated interviews in text and voice, for example follow-up interviews with participants you met in the field, in any of 10 languages.
• Human-led interviews with support: Active Listener mode gives a human interviewer real-time prompts and section tracking during a conversation.
• Coding fieldnotes and transcripts: QDA Workspace supports AI-assisted inductive and deductive coding with a codebook you control.
• Cross-transcript synthesis: ThemeLens maps codes to research questions across up to 100 transcripts, with participant-anchored quotes you can check.

The free tier includes 30 monthly credits, with no credit card required. See pricing for paid plans.

Limitations and methodological concerns

• Time and access. Ethnography is slow and depends on gatekeepers. Rapid designs reduce time but, as the healthcare review found, create their own sampling and interpretation problems.
• Researcher influence. Your presence changes the setting. Report it rather than hide it.
• Ethics of observation. People in the background of a scene rarely sign consent forms. Online spaces blur public and private. Get ethics board guidance early.
• Generalization. Ethnography aims at transferable insight, not statistical generalization. Say so in the methods section.
• AI-assisted analysis. Evidence on LLMs coding ethnographic material is still weak. Treat AI codes as suggestions and keep an audit trail. Methodology note: these recommendations should be reviewed against your discipline's standards and your ethics board's requirements.

Who this is for, and when not to use it

Use ethnography when your question is about practices, routines, culture or meaning in context, and you can gain sustained access. Do not use it when you need prevalence estimates, when you cannot get access to the setting, or when your question is fully answerable through interviews or a survey.

Frequently asked questions

What is ethnography in simple terms?

Ethnography is a way of studying people by spending time with them in their own setting, watching and taking part in what they do, talking with them and writing detailed notes, so you understand their world from the inside.

What are the main types of ethnography?

Common types are classic or realist, critical, focused, rapid, digital or online (including netnography), multi-sited, institutional and autoethnography. They differ mainly in fieldwork length, setting and the role of the researcher.

How long does ethnographic fieldwork take?

Classic ethnography often takes many months to years. Focused and rapid ethnographies can run from days to a few months. Choose the length based on how unfamiliar the setting is and how much depth your question needs.

What is the difference between ethnography and participant observation?

Participant observation is a method; ethnography is the overall research design. Most ethnographies use participant observation alongside interviews, documents and fieldnotes.

Is ethnography qualitative or quantitative?

Ethnography is primarily qualitative. Some ethnographers add counts or surveys, but the core data are fieldnotes, conversations and observations interpreted in context.

Can AI do ethnography?

No. AI can help with transcription, interviews and first-pass coding, but a January 2026 benchmark found the best LLM setups reached a mean F1 of only 0.41 on ethnographic texts. Observation and interpretation remain the researcher's work.

Conclusion

Ethnography in qualitative research is still the most direct way to understand how a group lives and works from the inside. Pick the type that fits your time, access and question, write fieldnotes every day, and keep interpretation open to testing in the field. Use AI where it handles text well, and keep observation and interpretation with the researcher.

Planning follow-up interviews or analyzing fieldnotes? Start free with 30 credits, or read our guide to thick description next.

Last updated: November 8, 2026

Study findings are summarized from the public abstracts and pages linked above as of November 8, 2026. Qualitati is not affiliated with the authors.