PSY1023 Chap.8 Reflexive Thematic Analysis
Reflexive Thematic Analysis
Reflexive thematic analysis is the analysis method behind the third assessment task in this unit, the group Qualitative Results Report. Taught across the Week 7 seminar and the Week 9 tutorial, it follows Braun and Clarke's (2006) guide to finding and reporting patterns of meaning in qualitative data.
Their paper frames the method around six phases: familiarising yourself with your data, generating initial codes, searching for themes, reviewing themes, defining and naming themes, and producing the report. Seminar slides relabel the third phase as generating initial themes, a newer name for the same activity.
Week 9 then turns the method into practice, with a tutorial built around coding interview material and documenting each step of the analysis.
What makes the approach reflexive is the stance it asks you to take. Patterns do not sit in transcripts waiting for someone to collect them; the analyst builds each theme by deciding what matters for the research question.
Braun and Clarke make this explicit in the last item of their 15-point quality checklist, which asks that the researcher be positioned as active in the research process. So your write-up should explain how your group's assumptions and experiences shaped the reading rather than pretending they were switched off.
The paper also warns against common pitfalls, such as stringing quotes together without interpretation or letting themes overlap until they blur into one another.
Two distinctions carry most of the weight in the analysis rows of the marking guide. The first separates codes from themes: a code labels a small, meaningful piece of data, while a theme is a broader pattern organised around a single central concept.
The second separates semantic codes, which stay with what participants said, from latent codes, which interpret the ideas underneath.
In the unit's handout these appear as two ends of a spectrum rather than two boxes, so expect some codes to sit in the blurry middle.
For AT3 your group generates between three and five themes, with four recommended, inside a report capped at 2,000 words with roughly 1,000 of those words for the analysis.
Every theme also appears in Appendix D, a table that links each theme to its codes and each code to its supporting quotes. The paid pages walk through the phase cycle as a diagram, compare codes and themes line by line, code an original extract into a fully defined theme, and practise the review decisions that turn a rough thematic map into a report-ready set.
What this chapter covers
- 01
Braun and Clarke's six phases and why the cycle is recursive
- 02
Codes versus themes as units of analysis
- 03
Semantic and latent coding on one spectrum
- 04
Writing a theme definition and a claim-style theme name
- 05
Phase 4 review at Level 1 and Level 2
- 06
Recording themes, codes and quotes in the Appendix D table
Worked example · free
Clustering codes into a candidate theme
- 2Cluster the codes that share a unifying feature: checking contributions, redoing a section late and not wanting to look bossy all concern carrying responsibility without authority. Splitting tasks by deadline is a logistics code and stays out for now.
- 2Name the theme as a short claim rather than a topic label, for example 'Holding the group together without permission to lead', so the name already says something about the research question.
- 1State the boundary: the theme covers informal leadership burdens; the code about quiet members being left out may belong to a separate theme on unequal voice, so park it in a miscellaneous pile rather than forcing it in.
- 1Review it twice: read every collated extract to check the pattern is coherent, then re-read the whole data set to see whether the theme fits and whether any missed extracts belong to it.
Key terms
- Data Set
- The portion of all collected material that a particular analysis actually draws on, as distinct from the whole data corpus gathered for the project.
- Code
- A short label attached to the smallest piece of data that can be read in a meaningful way against the research question.
- Candidate Theme
- A provisional grouping of codes and extracts produced while searching for themes, still open to being merged, split or discarded during review.
- Semantic Code
- A label that captures the surface or explicit content of what a participant said, taken largely at face value.
- Latent Code
- An interpretive label capturing the underlying assumptions or ideas that make a participant's explicit words make sense.
- Thematic Map
- A visual layout of candidate themes, sub-themes and their relationships that is checked against the whole data set during review.
Reflexive Thematic Analysis FAQ
Is the third phase called searching for themes or generating initial themes?
Both labels describe the same work of collating codes into candidate themes. Braun and Clarke's 2006 paper uses searching for themes, while the unit's Week 7 seminar uses generating initial themes; either is acceptable if you cite the method correctly.
How many themes should our group report in AT3?
The assessment template asks for between three and five themes and recommends four. With a 2,000-word ceiling and about 1,000 words for analysis, more themes would leave each one too thin to show depth and supporting quotes.
Can a theme be named after one of our interview questions?
No. Themes should describe patterns across the whole data set that answer the research question, and Braun and Clarke list using interview questions as themes among the main pitfalls of the method, because it means little analysis has been done.
What goes in Appendix D?
Appendix D is the completed reflexive thematic analysis table with three columns for themes, codes and quotes. It lists every code feeding each theme and every quote behind each code, while the report body uses only selected illustrative quotes.
Do latent codes have to be right about what participants really meant?
Latent codes are interpretations, so they need to be plausible and traceable to the words that prompted them rather than provably correct. Point to the phrasing that supports each reading and keep alternative interpretations in view during review.
Assessment move
Treat this topic as a skill you practise on small pieces of text rather than a list of phases to memorise. Start by writing the six phase names from memory, then check them against Braun and Clarke's 2006 wording and note where the seminar's labels differ.
Next, take a few lines from your own group's edited transcripts and code them twice: once semantically, staying close to the words, and once latently, asking what assumptions make those words sensible. Compare your two lists with a teammate and argue about any code that sits in the middle of the spectrum, because explaining that judgement is exactly what the analytic process row of the marking guide rewards.
When your group sorts codes into candidate themes, use a table or sticky notes so codes can move freely, and keep a miscellaneous pile instead of forcing awkward codes into a theme. For every candidate theme, write a definition of a couple of sentences that states its central organising concept and what it leaves out.
Then run both review levels: read the collated extracts for coherence, and re-read the full data set to check fit and catch missed extracts. Finally, draft your theme names as short claims, test each one against the question of whether a reader could tell at a glance what it argues, and fill in Appendix D as you go rather than at the end.
Working through Reflexive Thematic Analysis in PSY1023? Sia is AskSia’s AI Psychology tutor — ask any PSY1023 Reflexive Thematic Analysis question and get a clear, step-by-step explanation grounded in how PSY1023 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
