Reflexive Thematic Analysis (RTA)
A method for identifying meaningful themes in qualitative data such as open-ended survey responses or interviews. Reflexive Thematic Analysis (RTA), proposed by Virginia Braun and Victoria Clarke, is distinctive in that it actively treats the researcher's subjectivity not as a bias to be eliminated, but as a resource for analysis.
What this page covers
- Basic terminology and the six phases of RTA
- A seminar exercise plan (an RTA experience using survey data from the sound art work Marine)
- Downloadable exercise worksheet
What is RTA?
Core ideas
- Themes are not "naturally emerging from the data"; they are constructed by the researcher through analysis.
- Coding is not a mechanical classification but an interpretive act.
- The researcher's expertise, interests, experience, and positionality are explicitly considered in how they shape the analysis (i.e., reflexivity).
Key terms
Theme
A recurring pattern across the data that holds research significance. Not merely a summary of a topic, but a conceptual unit that captures meaning or argument behind the data.
Example: From artwork survey responses, construct a theme "immersive experience through fusion of bodily sensation and sound."
Coding
Assigning short labels that summarize content or meaning to individual units of data (a sentence or segment of speech). It is an interpretive act in which the researcher decides what is interesting and what stands out.
Example: For the response "a texture I want to keep touching forever," assign the code 〈sustained tactile pleasure〉.
Code
The label itself, produced through coding. The smallest unit of analysis and the material from which themes are generated.
Example: 〈healing〉 〈relaxing〉 〈comfortable〉 are separate codes that may later support a single theme.
Reflexivity
The stance of explicitly recognizing how the researcher's own positionality (discipline, values, prior knowledge, relation to the work) influences the analysis, and describing that as part of the analysis itself.
Example: Noting "I am a sound artist who designs tactile interfaces, viewing this work from that standpoint" helps explain why one attended to 〈the fusion of touch and sound〉.
Inductive / Deductive
- Inductive: Set aside existing theories and frameworks; build codes and themes while staying close to the data.
- Deductive: Read the data through the lens of existing theories or hypotheses; explore how specific concepts appear in the data.
In RTA these are treated not as exclusive options but as a spectrum in which one chooses where to place emphasis.
Semantic / Latent
- Semantic: The surface-level meaning explicitly stated by the respondent.
- Latent: Underlying assumptions, values, contexts, and other implicit meanings.
Example: "It felt like being underwater" can be read semantically as 〈metaphor of underwater sensation〉, and latently as 〈desire to escape the everyday〉.
The six phases of RTA
Braun & Clarke organize the RTA process into the following six phases. Note that this is not a linear procedure — it is a recursive process in which one moves back and forth.
| Phase | Name | Content |
|---|---|---|
| 1 | Familiarisation | Read the data repeatedly to grasp the whole; note what stands out. No classification yet. |
| 2 | Coding | Assign short labels (codes) to segments of the data. |
| 3 | Generating initial themes | Group related codes and propose provisional themes. |
| 4 | Developing and reviewing themes | Check how well provisional themes fit the data; merge, split, or discard. |
| 5 | Refining, defining and naming themes | Clearly define what each theme represents and finalize its name. |
| 6 | Writing up | Write up the analysis, weaving data excerpts with thematic interpretation. |
Differences from other qualitative analysis methods
There are several schools of thematic analysis: coding reliability approaches that emphasize inter-coder agreement, codebook approaches that prepare codes in advance, and the reflexive (RTA) approach used in this exercise. RTA is distinct in that it harnesses the researcher's subjectivity rather than seeking objectivity.
Other methods that build concepts from data — Grounded Theory Approach (GTA), SCAT, KJ method — each have different purposes (theory generation, conceptual organization, ideation) and premises. RTA aims at the construction of themes itself, and does not extend to theory generation.
Seminar exercise: an RTA experience using the Marine survey
Purpose
Using survey data from experiencers of the sound art work Marine, students experience the full RTA process firsthand. By assigning codes, formulating themes, and discussing them with each other, the exercise aims to convey three things experientially:
- Coding is an interpretive act, not a mechanical classification.
- Themes do not "naturally emerge from the data" — they are constructed by the researcher.
- The researcher's expertise and positionality (reflexivity) are reflected in the analysis.
Materials
About the work Marine
When the experiencer touches the dome-shaped interface, ocean-related sound (waves, dolphin calls, bubbles) and vibration are emitted. Multiple domes of different stiffness are provided, and the experiencer explores the relationship between sound and touch through how they touch and press the surfaces.
Data
Open-ended responses from 24 experiencers, structured around three questions:
- Q1. Other impressions you felt about the texture of the domes
- Q2. Impressions of the sound and vibration when touching the domes
- Q3. Impressions of the work as a whole, including the sound
Missing data
Different respondents answered different questions (some answered all three, some only Q3). This pattern of missing data is realistic for actual research and provides good practice material for coding.
Worksheet (download)
Download the Excel file below to use for the exercise. The worksheet contains four sheets:
- How to proceed — Instructions for the exercise
- Coding worksheet — Working sheet where you write codes (yellow cells)
- Theme organization sheet — For grouping codes and proposing provisional themes
- Source data (reference) — Original list of open-ended responses
Download the RTA exercise worksheet (.xlsx)
Note
The worksheet itself is in Japanese. Use it as a template structure; you can translate the column headers as needed.
Schedule (90-minute version)
| Time | Phase | Content |
|---|---|---|
| 0:00–0:10 | Introduction | Purpose of the exercise and terminology review |
| 0:10–0:20 | Phase 1: Familiarisation | Read all data; mark what stands out |
| 0:20–0:40 | Phase 2: Coding | Individual work: assign codes to each response |
| 0:40–0:55 | Phase 3: Generating initial themes | Share and group codes; propose provisional themes |
| 0:55–1:10 | Phase 4: Reviewing and naming themes | Organize and name themes |
| 1:10–1:25 | Phase 5: Discussion of reflexivity | Discuss why you attended to certain themes; share positionality |
| 1:25–1:30 | Wrap-up | Instructor comments and reflection |
Phase details
Phase 1: Familiarisation (10 min)
Purpose: Get a feel for the data as a whole before beginning analysis.
- Read through all open-ended response columns in the worksheet.
- Use sticky notes or memos to mark words, expressions, or respondents that catch your attention.
- Don't try to classify at this stage — just leave traces of where you find things "interesting" or "striking."
Phase 2: Coding (individual work, 20 min)
Purpose: Assign each response a short label (code). Codes should not be too close to the wording of the response, nor too abstract — they should add one layer of interpretation.
- In the "Code" column, write short phrases (around 5–15 characters) that capture the feature of each response.
- A response can have multiple codes. Enclosing codes in 〈 〉 makes them easier to spot later.
- If undecided, write multiple candidates on sticky notes and pick later.
Good vs. less good code examples
Good:
- 〈sustained tactile pleasure〉
- 〈ocean metaphor as escape from the everyday〉
- 〈resonance with biological vibration〉
Less good:
- 〈feels nice〉 〈ocean〉 — too surface-level
- 〈the artist's intent〉 — interpretive leap
Phase 3: Generating initial themes (15 min)
Purpose: Group codes and propose provisional themes from clusters of related codes.
- Show codes to each other; look for commonalities and relationships.
- Group similar or related codes. Large paper or sticky notes help with discussion.
- Give each group a "provisional theme name." It doesn't need to be final yet.
- If a code doesn't fit any theme, leave it as "unclassified" rather than forcing it.
Phase 4: Reviewing and naming themes (15 min)
Purpose: Share the formulated themes and clean up relationships and naming.
- Each person shares their work — 2–3 minutes per theme works well.
- Check for themes that are similar, opposed, or one containing another.
- Refine theme names into the form "phenomenon + meaning."
- Example: 〈fusion of touch and sound〉 → 〈the work experience as dissolution of sensory boundaries〉
- Converge to roughly 3–6 themes for manageability.
Phase 5: Discussion of reflexivity (15 min)
Purpose: Discuss why you attended to certain themes and how your expertise, interests, and experience shaped that.
- Each person spends one minute on "why did this word resonate so strongly with me?"
- Confirm that, for the same data, different people focus on different things.
- Imagine: "If someone from a completely different field analyzed this, what themes might emerge?"
- Confirm that, in RTA, this kind of self-awareness about positionality itself supports analytic quality.
Extracted themes (example)
〈Multisensory healing through light, sound, and touch〉
Responses such as "healing," "comfortable," "I want to keep touching it" are treated not merely as affective reactions, but as multisensory experiences that arise through the cooperation of light, sound, and touch. The jellyfish-like cuteness of the form is also a component of this multisensory healing scene. The fact that one experiencer spoke of the need for this in today's stressful society and its significance for people who cannot care for themselves connects to this meaning-making.
〈Immersion in "touching the living" through device shape, vibration, and softness〉
The dome shape, internal vibration, and soft elastic material — these material conditions produce, for the experiencer, the sensation of "touching a living thing." Responses such as "an organic nuance like a creature's soft skin, fetal movement, or heartbeat" and "like a sea creature — a jellyfish" show that the work appears not as mere object but as a kind of life-bearing other.
Furthermore, operational frictions — hardness to push, force that doesn't transmit as expected — also connect to this theme. As one experiencer found it "interesting" that "more force ends up being applied than expected," the gap between intention and response becomes a moment in which the device emerges as an other with its own response-logic, not a tool that simply complies. The life-likeness statically evoked by material texture and the otherness dynamically raised by response gaps overlap, producing a structure in which the experiencer becomes immersed in "touching the living."
〈Immersion into the "ocean world" through acoustic space〉
Responses such as "felt like being underwater," "the sand sounds made me feel the sea," "surrounded by sound, feeling the ocean with my whole body" are read not as visual or linguistic representations of the sea, but as a transition into an imagined "ocean world" that the acoustic space itself constructs. Wave sounds, dolphin calls, bubble sounds, and the spatial arrangement that surrounds the experiencer together imply a way of experience in which another world rises up.
〈Instinctive appeal through elastic material and vibrational feedback〉
The primal, instinctive reaction to the very structure in which touching an elastic material immediately returns vibration and sound. Implies the appeal of "touch → respond" responsiveness that operates without going through conceptual interpretation.
Discussion prompts
- Where do the extracted themes and the students' themes overlap?
- What viewpoints appear in the students' themes that do not appear in the extracted themes? And vice versa?
- Are there points that, from a maker's perspective, may have been overlooked?
- For responses that read negatively (such as operational frictions), how should the decision to include them in a theme or treat them separately change depending on the research purpose?
- If the research purpose were "to elucidate the affective value of haptic interfaces," how should the structure of themes change?
Discussion tips
Keep in mind that "there is no right answer"
In RTA, themes are constructed; there is no single correct classification.
Don't code too quickly
Spending enough time with the data in Phase 1 determines the quality of later analysis. Rushing into classification leads to shallow pattern recognition.
Value negative reactions too
Negative remarks such as "hard to press" or "weak vibration" are important data that constitute the work experience. Collecting only praise doesn't make analysis.
References
- Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101.
- Braun, V., & Clarke, V. (2022). Thematic Analysis: A Practical Guide. SAGE Publications.
- Braun, V., & Clarke, V. (2019). Reflecting on reflexive thematic analysis. Qualitative Research in Sport, Exercise and Health, 11(4), 589–597.
- Oka, M., Ishikawa, J., Kamihoshi, H., Matsumoto, M., Takahashi, S., & Ito, M. (2022). Overview and analytical process of thematic analysis, focusing on reflexive thematic analysis by Virginia Braun and Victoria Clarke (in Japanese). Journal of Japanese Society of Nursing Research, 45(2), 145–158. https://doi.org/10.15065/jjsnr.20211222158