Thematic analysis explained — the six-phase Braun and Clarke method

Thematic analysis is the most-used qualitative analysis method in UAE master’s and PhD dissertations. It’s appropriate, accessible, and supported by a clear framework — Braun and Clarke’s six-phase method (originally 2006, updated 2019 and 2022). But the popularity is also a weakness. Many dissertations claim to use thematic analysis without actually executing the six phases. The marker reads the methodology chapter, looks at the findings chapter, and can see the gap.

This is the working summary of how the six phases actually work, and what your methodology chapter and findings chapter need to demonstrate.

What Braun and Clarke’s thematic analysis is

Braun and Clarke distinguish between several types of thematic analysis. The original 2006 paper introduced what they now call reflexive thematic analysis — analysis grounded in the researcher’s active interpretation rather than purely in the data. The 2019 update emphasises that “themes don’t emerge from data” (a phrase undergraduate work uses lazily); themes are constructed by the researcher’s analytical work.

This matters for how you describe what you’re doing. Themes emerged from the data is the underwritten version. I generated themes through reflexive engagement with the data is the current Braun-and-Clarke version. The vocabulary signals whether you’ve read the recent updates.

The six phases

Phase 1: Familiarisation with the data

You read and re-read your data corpus (interview transcripts, focus group transcripts, observation notes, document data). You take initial notes — impressions, recurring topics, surprising moments, things you didn’t expect.

What this looks like in practice: a master’s-level qualitative dataset of 10–15 interviews requires reading each transcript twice before any coding begins. You should be able to describe the gist of each interview from memory after familiarisation.

In your methodology chapter, document this phase — how many readings, how long, what you noticed.

Phase 2: Generating initial codes

You work through the data systematically and apply codes to extracts. A code is a short label (1–4 words usually) that captures something the extract is about.

UAE student dissertations often use NVivo or Atlas.ti for this. Manual coding in Word with highlight colours and margin comments also works. The tool matters less than the systematic-ness of the coding.

Two coding approaches:

• Inductive coding — codes are derived from the data itself, bottom-up.
• Deductive coding — codes come from a pre-existing framework or theoretical model.

Most UAE dissertations use a hybrid — initial inductive coding refined against a theoretical lens. Name your approach in the methodology chapter.

A representative initial code list for a study of hybrid working might include codes like: concentration affordances at home, managerial visibility anxiety, peer connection loss, family negotiation, commute time reclaimed.

Phase 3: Generating initial themes

You group related codes into candidate themes. A theme is a higher-level concept that captures something meaningful across multiple data extracts.

Three-to-six initial themes is normal for a master’s-level dataset. The themes should be substantive — trust is not a theme; the renegotiation of trust between hybrid workers and their managers is.

This phase is often where dissertations stall. Students generate many codes, then can’t see how to group them. The fix is to keep grouping at the conceptual level rather than the topical level. Concentration affordances at home and peer connection loss are topically different but conceptually both about the trade-off between focus and connection in hybrid working.

Phase 4: Reviewing themes

Two stages of review:

• Review against coded extracts. Do the codes assigned to each theme actually support the theme as articulated? Are there codes that should move? Should the theme be split?
• Review against the dataset as a whole. Do the themes as a set capture the data’s main contours? Are there important patterns in the data that no theme captures?

This is the iterative phase. You refine, split, merge, or drop themes until the set genuinely reflects the data.

Phase 5: Defining and naming themes

For each theme, write a 3–5 sentence definition. What is this theme about? What is the central organising concept? What does it tell us about the research question?

Theme names should be specific and informative. Trust is a bad theme name. The renegotiation of trust between hybrid workers and their managers is a good theme name. Some Braun-and-Clarke updates encourage even more substantive theme names that capture the analytical claim — Hybrid working requires managers and employees to renegotiate the basis of trust from physical co-presence to documented output.

Phase 6: Producing the report

The findings chapter writes up the analysis, theme by theme. For each theme:

1. State the theme name and the central claim.
2. Provide 3–5 illustrative extracts from the data, with anonymous attribution.
3. Discuss what the extracts collectively tell us.
4. Connect the theme back to the research question and (when relevant) to the existing literature.

The most common findings-chapter mistake is letting extracts speak for themselves — quoting at length without analytical commentary. The extracts are evidence; your analytical claims are the argument. Both should appear, with analysis dominating.

What the methodology chapter needs to show

Beyond naming Braun and Clarke, the methodology chapter for a thematic analysis dissertation needs to specify:

• Which version of thematic analysis — reflexive thematic analysis is now the standard label; codebook thematic analysis is a distinct variant.
• Inductive, deductive, or hybrid coding approach.
• Number of coding passes through the data.
• Tool used for coding (NVivo, Atlas.ti, manual).
• Process for theme refinement — how the initial themes became the final themes.
• Reflexivity statement — how your positionality might have shaped what you noticed and how you interpreted it.

A methodology chapter that addresses these elements explicitly demonstrates familiarity with the method beyond textbook references.

When The Essay Atelier writes thematic analyses

For dissertation projects involving thematic analysis, the writer-editor pair works through Braun-and-Clarke’s six phases with you (or with your existing data if you’ve already coded). The methodology and findings chapters are written to match — what the methodology promises, the findings deliver.

If you have interview data you’ve collected but don’t know how to code, or coded data you don’t know how to write up, send the editors a paragraph describing where you are. We’ll tell you what would help most.