Skip to main content

Method guide

Reflexive Thematic Analysis: A Practical Guide

Reflexive thematic analysis (reflexive TA) is Virginia Braun and Victoria Clarke's approach to thematic analysis. Their 2006 paper set out six phases for doing thematic analysis; in 2019 they named their approach "reflexive" to separate it from other versions that share the label but not the underlying values. It is now one of the most cited qualitative methods in psychology, health, education and the social sciences.

In reflexive TA, themes are not found in the data. The researcher builds them, through sustained and interpretive engagement with the dataset, and the researcher's perspective is treated as a resource rather than a bias to be removed. This guide covers what that means in practice: how reflexive TA differs from other types of thematic analysis, the six phases, what counts as a theme, how to handle reflexivity and sample size, and the reporting mistakes examiners and reviewers raise most often.

What makes thematic analysis "reflexive"?

Braun and Clarke describe thematic analysis as a family of methods rather than one method. Reflexive TA sits at the fully qualitative end of that family. Its defining commitments are that coding is open and organic, with no fixed codebook; that themes are actively generated by the researcher; and that quality comes from reflexivity, depth of engagement and coherence, not from agreement between coders.

That last point is the one that most often trips researchers up. In reflexive TA there is no single correct coding of the data to agree on, so measures of inter-coder reliability such as Cohen's kappa do not fit the approach. A second researcher can still be useful, for example to question your interpretations and widen your reading of the data, but the aim is richer analysis, not consensus.

Reflexive TA vs codebook and coding-reliability TA

Braun and Clarke distinguish three broad types of thematic analysis. Naming which one you are using, and why, is the first thing an examiner will look for in your methods chapter.

  • Reflexive TA: coding is flexible and evolves as you analyse; themes are interpretive stories built around a central idea; the researcher's subjectivity is part of the analysis; no inter-coder reliability.
  • Codebook TA (for example template analysis or framework analysis): a structured codebook, often partly developed in advance, organises the analysis; suits team projects, applied research and tight timelines; sits between the other two.
  • Coding-reliability TA: themes are often defined early, a fixed codebook is applied by several coders, and agreement between them is measured; it reflects a more positivist idea of reliability.

The six phases of reflexive thematic analysis

Braun and Clarke's later writing (including their 2022 book, Thematic Analysis: A Practical Guide) renames some of the original 2006 phases, to make clear that themes are developed rather than searched for. The phases are recursive: expect to move back and forth between them.

  1. 1Familiarisation with the dataset — read and re-read the data, listen to recordings, and make notes on initial ideas and on your own reactions to what you read.
  2. 2Coding — work systematically through the whole dataset, labelling anything relevant to your question. Codes can be semantic (what is said) or latent (the assumptions and meanings beneath it), and they will change as you go.
  3. 3Generating initial themes — cluster codes that seem to share a core idea into candidate themes. These are provisional: you are building possible stories, not settling on results.
  4. 4Developing and reviewing themes — test each candidate theme against its coded extracts and then against the full dataset. Expect to merge, split, rework or discard themes.
  5. 5Refining, defining and naming themes — write a short definition of each theme that captures its central organising concept, its boundaries and how it relates to the other themes. A good name says something about the meaning, not just the topic.
  6. 6Writing up — weave analytic narrative and data extracts together, and connect the analysis to your research question and the literature. Writing is part of the analysis, not a report of it.

What counts as a theme (and what doesn't)

In reflexive TA a theme is a pattern of shared meaning, held together by a central organising concept. It tells an interpretive story about the data in relation to your research question.

The most common problem Braun and Clarke report in published studies is what they call a topic summary: a "theme" that simply collects everything participants said about a topic, often mirroring an interview question (for example "Experiences of remote work"). A topic summary describes what was discussed; a theme says what it means. A useful test is whether your theme could be named before you had looked at the data. If it could, it is probably a topic.

Language matters here too. Writing that themes "emerged" from the data implies they were waiting to be discovered, which contradicts reflexive TA's view that the researcher constructs them. Write that you developed, generated or constructed them.

Choosing your orientation to the data

Reflexive TA can be used in many theoretical frameworks, so you need to say where your analysis sits. Braun and Clarke describe several choices, each best thought of as a continuum rather than an either/or:

  • Inductive to deductive: are codes and themes driven mainly by the data, or by existing theory and concepts?
  • Semantic to latent: are you staying close to what participants explicitly said, or interpreting the ideas and assumptions underneath?
  • Experiential to critical: are you trying to capture participants' own perspectives and meanings, or interrogating how language and meaning construct the topic?
  • Realist or essentialist to constructionist: what do you assume about the relationship between language, experience and reality?

Reflexivity in practice

Reflexivity means examining how your own position, assumptions, theoretical commitments and choices shape the analysis. In reflexive TA it is not an optional extra; it is how quality is demonstrated.

In practice that usually means keeping a reflexive journal from the start of the project (what surprised you, what you expected to find, where your own experience pulled your reading), writing analytic memos as themes develop, and including a short positionality statement in your methods chapter. The point is not to confess bias. It is to show the reader how your perspective informed the analysis, so they can judge it.

How many interviews do you need?

There is no fixed number. Braun and Clarke (2021) argue that data saturation, the idea that you stop when no new information appears, fits poorly with reflexive TA, because meaning is generated by the researcher rather than sitting in the data waiting to run out.

Instead, justify your sample by the richness of the data, the scope of your question, the depth of interpretation you are aiming for and the practical limits of your project. Malterud, Siersma and Guassora's (2016) concept of "information power" is one widely used way to frame this argument. Whatever you decide, explain the reasoning rather than citing a number.

Mistakes examiners and reviewers flag

  • Calling the analysis "Braun and Clarke's thematic analysis" while using a codebook and reporting inter-coder reliability, which belong to other types of TA.
  • Themes that are topic summaries, often one per interview question.
  • Saying themes "emerged", or presenting the analysis as if it were objective.
  • Citing only the 2006 paper and ignoring the authors' later clarifications of reflexive TA.
  • Claiming data saturation as the reason for stopping recruitment without explaining what it means in a reflexive approach.
  • No account of the researcher's position or of how it shaped the analysis.
  • Too many themes, or themes with little analytic narrative between long quotations.

Reporting reflexive TA: checklists and guidelines

Braun and Clarke (2021) published twenty questions for evaluating thematic analysis manuscripts, and in 2024 introduced the Reflexive Thematic Analysis Reporting Guidelines (RTARG), listed by the EQUATOR Network. Both are designed for reflexive TA specifically. Generic qualitative checklists can push you towards practices, such as reporting inter-coder agreement, that do not fit the approach. If your target journal requires a checklist, it is worth explaining where the reflexive TA guidance differs.

Using software and AI in reflexive TA

Software can help you organise codes, extracts and memos, but it cannot do reflexive TA for you, because the analysis rests on your interpretation. Using AI in reflexive TA is debated. If you use it, keep it to supporting tasks such as surfacing passages that may be relevant to a code, make every coding and theme decision yourself, keep a record of what the AI suggested and what you did with it, and disclose it in your methods section. Check your institution's policy first.

QualIntel OS is built around that boundary. AI surfaces candidate evidence and you accept or reject each suggestion; nothing is coded automatically. Every decision is logged in an audit trail, and the tool generates an AI disclosure statement for your methods chapter.

Key references

  • Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101.
  • Braun, V., & Clarke, V. (2019). Reflecting on reflexive thematic analysis. Qualitative Research in Sport, Exercise and Health, 11(4), 589–597.
  • Braun, V., & Clarke, V. (2021). One size fits all? What counts as quality practice in (reflexive) thematic analysis? Qualitative Research in Psychology, 18(3), 328–352.
  • Braun, V., & Clarke, V. (2021). To saturate or not to saturate? Questioning data saturation as a useful concept for thematic analysis and sample-size rationales. Qualitative Research in Sport, Exercise and Health, 13(2), 201–216.
  • Braun, V., & Clarke, V. (2022). Thematic analysis: A practical guide. SAGE.
  • Braun, V., & Clarke, V. (2024). Supporting best practice in reflexive thematic analysis reporting in Palliative Medicine: A review of published research and introduction to the Reflexive Thematic Analysis Reporting Guidelines (RTARG). Palliative Medicine, 38(6), 608–616.
  • Byrne, D. (2022). A worked example of Braun and Clarke's approach to reflexive thematic analysis. Quality & Quantity, 56, 1391–1412.
  • Malterud, K., Siersma, V. D., & Guassora, A. D. (2016). Sample size in qualitative interview studies: Guided by information power. Qualitative Health Research, 26(13), 1753–1760.

Frequently asked questions

What is reflexive thematic analysis?

Reflexive thematic analysis is Braun and Clarke's approach to thematic analysis. The researcher develops themes, patterns of shared meaning built around a central idea, through open, evolving coding and interpretation. The researcher's perspective is treated as part of the analysis, and quality is shown through reflexivity and depth of engagement rather than agreement between coders.

What is the difference between thematic analysis and reflexive thematic analysis?

Thematic analysis is an umbrella term for several methods. Reflexive thematic analysis is one of them, developed by Braun and Clarke. Unlike codebook or coding-reliability thematic analysis, it uses no fixed codebook, does not measure inter-coder reliability, and treats themes as interpretations the researcher constructs rather than patterns waiting to be found.

What are the six phases of reflexive thematic analysis?

In Braun and Clarke's later writing the phases are: (1) familiarisation with the dataset, (2) coding, (3) generating initial themes, (4) developing and reviewing themes, (5) refining, defining and naming themes, and (6) writing up. The phases are recursive, so you move back and forth between them.

Do you need inter-rater reliability in reflexive thematic analysis?

No. Reflexive thematic analysis assumes there is no single correct way to code the data, so inter-rater reliability measures such as Cohen's kappa do not fit it. A second coder can still help, by questioning interpretations and widening your reading of the data, but the aim is richer analysis, not agreement.

How many interviews do I need for reflexive thematic analysis?

There is no set number. Braun and Clarke argue that data saturation fits poorly with reflexive thematic analysis. Justify your sample by the richness of your data, the scope of your question and the depth of interpretation you need; the concept of information power (Malterud et al., 2016) is a common way to frame that argument.

Can I use AI for reflexive thematic analysis?

It is debated. If you use AI, keep it to supporting tasks such as surfacing passages that may be relevant, make every coding and theme decision yourself, record what the AI suggested and what you did with it, and disclose its use in your methods section. Check your institution's AI policy first.

Related guides

Run your thematic analysis in QualIntel OS

AI surfaces candidate evidence, you confirm every decision, and the audit trail writes itself. One project free, no card.

Start for free