In brief
- Thematic coding labels segments of data with codes; thematic analysis develops those codes into themes that answer your research question.
- Braun and Clarke's six phases run from familiarization to writing up, and you move back and forth between them.
- In reflexive thematic analysis, themes are patterns of shared meaning that the researcher actively constructs; they do not simply emerge.
- Inductive coding starts from the data; deductive coding starts from existing theory or questions.
- Software such as NVivo organizes coding but does not do the analysis for you.
What is thematic coding? ¶
A code is a label for a piece of data. A theme is a claim about what the data mean.
Thematic coding is the process of assigning short labels (codes) to segments of qualitative data, such as interview transcripts, that capture something relevant to your research question. Codes are then compared, grouped and refined into broader themes. Thematic coding is the core activity of thematic analysis, a method for identifying, analyzing and reporting patterns of meaning across a dataset.
Thematic analysis was set out as a distinct method by Virginia Braun and Victoria Clarke in their 2006 paper "Using thematic analysis in psychology", one of the most cited methods papers in the social sciences. It is flexible: it is not tied to one theory, and it works with interviews, focus groups, open survey responses, diaries and documents.
Thematic analysis in qualitative research: three main approaches ¶
Braun and Clarke now emphasize that "thematic analysis" covers a family of methods with different assumptions. They distinguish three broad types:
| Approach | How coding works | Examples |
|---|---|---|
| Coding reliability TA | A structured codebook applied by several coders; agreement measured (for example, Cohen's kappa) | Boyatzis; Guest and colleagues |
| Codebook TA | A codebook or framework developed early, but interpretation remains qualitative | Template analysis; framework analysis |
| Reflexive TA | Open, organic coding by the researcher; subjectivity treated as a resource; no inter-rater reliability | Braun and Clarke |
In reflexive thematic analysis, a theme is a pattern of shared meaning underpinned by a central organizing concept. A theme is not a topic heading such as "experiences of workload" that collects everything participants said about workload. Name the approach you used in your methods section and keep your procedures consistent with it. Mixing reflexive TA with inter-rater reliability scores is a common reviewer criticism.
Braun and Clarke's six phases of thematic analysis ¶
The phases were first published in 2006 and renamed slightly in Braun and Clarke's later work, including their 2021 book Thematic Analysis: A Practical Guide. The process is recursive: you will loop back to earlier phases.
| Phase | Current name (2006 name) | What you do |
|---|---|---|
| 1 | Familiarization with the data (Familiarizing yourself with your data) | Transcribe or check transcripts, read and reread, make notes on initial ideas |
| 2 | Coding (Generating initial codes) | Work systematically through the whole dataset, labeling anything relevant to the research question |
| 3 | Generating initial themes (Searching for themes) | Cluster codes that share a core idea into candidate themes; draft a thematic map |
| 4 | Developing and reviewing themes (Reviewing themes) | Check themes against coded extracts and the full dataset; merge, split or discard |
| 5 | Refining, defining and naming themes (Defining and naming themes) | Write a short definition of each theme's central concept and its boundaries; choose a clear name |
| 6 | Writing up (Producing the report) | Weave analytic narrative with vivid data extracts and connect to literature and the research question |
Keep a reflexive journal from phase 1 onward. A dated record of your decisions is the simplest way to show rigor to supervisors and peer reviewers.
Qualitative Data Analysis
Working on thematic analysis? Get guidance from a qualitative researcher who reviews your codes and themes with you.
Inductive vs deductive coding ¶
Inductive coding (bottom-up) builds codes from the data themselves, without a pre-set frame. It suits exploratory questions and is the usual starting point for inductive thematic analysis. Deductive coding (top-down) applies concepts from existing theory, previous research or your interview guide. Many projects combine the two: a few theory-driven codes plus new codes from the data.
Braun and Clarke also distinguish semantic codes, which capture what participants explicitly say, from latent codes, which interpret underlying assumptions or ideas. Analysis that stays entirely semantic risks becoming description rather than interpretation.
For example, in the night-shift study below, a deductive code drawn from the job demands-resources model might be "lack of recovery time", while an inductive code from the data might be "texting my partner like a flatmate". Both can sit in the same analysis, as long as you report where each came from.
Other types of coding in qualitative research
Saldaña's The Coding Manual for Qualitative Researchers describes many coding methods. Common first-cycle methods include descriptive coding (topics), in vivo coding (participants' own words), process coding (actions, using gerunds) and values coding. Pattern coding is a common second-cycle method that groups first-cycle codes.
Codes vs themes ¶
| Code | Theme | |
|---|---|---|
| Scope | One specific idea in a segment of data | A broader pattern across many codes and participants |
| Form | A short label, often a few words | A named central concept with a definition |
| Role | Building block | Analytic claim that answers the research question |
| Number | Dozens to hundreds | Usually two to six in a report |
| Example | "Answering emails at night" | "Work leaks into every hour" |
CODEBOOK
Project:
Approach: [reflexive TA / codebook TA / coding reliability TA]
Research question:
Version and date:
THEME: [name]
Central organizing concept: [one or two sentences]
Boundaries: [what this theme includes and excludes]
Code 1: [short label]
Definition: [what the code captures]
Type: [inductive / deductive; semantic / latent]
Include when:
Exclude when:
Example extract: "..." (P__, line __)
Analytic memo: [ideas, links to other codes]
Code 2: [short label]
Definition:
Type:
Include when:
Exclude when:
Example extract: "..." (P__, line __)
Analytic memo:
CHANGE LOG
[date] [code merged / split / renamed / dropped] [reason]
Note: in reflexive TA, treat this as a living record of your coding rather than a fixed instrument for measuring agreement.
Thematic analysis example: a coded excerpt ¶
Imagine an interview study with early-career nurses about working night shifts. Here is one short extract and how it might be coded.
Interview excerpt (Participant 7)
"When I started nights, I thought I'd get used to it in a month. I didn't. [1] I sleep in pieces, three hours here, two hours there. [2] My partner works days, so we basically text each other like flatmates. [3] And you can't really complain on the ward, because everyone's tired, and you don't want to be the one who can't cope."
Codes
- [1] Fragmented sleep (semantic); expectation of adaptation not met (latent)
- [2] Relationship reduced to logistics (latent); out of sync with partner (semantic)
- [3] Silencing own distress (latent); fear of appearing incompetent (latent); fatigue as shared norm (semantic)
Candidate theme after coding all transcripts
"Coping quietly": participants treated exhaustion as a private failing to be managed alone, because ward culture framed tiredness as normal and complaint as weakness. Supported by codes from 11 of 14 participants, including silencing own distress, fear of appearing incompetent and self-medicating with caffeine.
Notice that the theme makes an interpretive claim. A topic summary such as "sleep problems" would not.
Reflexivity in thematic analysis ¶
Reflexivity means examining how your own position, assumptions and experience shape what you notice and how you interpret it. In reflexive TA, this subjectivity is part of the method rather than a source of error to remove. In practice:
- write a short positionality statement before analysis begins;
- keep a reflexive journal of reactions, hunches and decisions during coding;
- discuss codes and themes with a supervisor or colleague to deepen interpretation, not to reach numerical agreement;
- report in your methods how your perspective may have influenced the analysis.
For a full methods chapter that explains your approach clearly, our research methodology examples show how to justify qualitative choices.
Thematic coding in NVivo: practical tips ¶
NVivo, from Lumivero, is widely used qualitative coding software. Since NVivo Release 1 (2020), what used to be called "nodes" are called codes.
- Import your data through the Import tab (Files). Transcripts with consistent speaker labels are easier to code; our transcription service can prepare them.
- Code as you read. Select text, then drag it onto a code or right-click and choose Code. Create codes freely in phase 2; tidy them later.
- Use memos and annotations for your reflexive notes and initial ideas, linked to the source.
- Turn on coding stripes to see which codes sit on each passage and to check consistency.
- Build candidate themes by grouping codes into a hierarchy, with the theme as the parent code.
- Use cases and classifications to store participant attributes, then run a Matrix Coding Query or Crosstab to compare groups.
ATLAS.ti, MAXQDA and the free tool Taguette offer similar functions. Auto-coding features can suggest topics, but they cannot interpret meaning. If you want an experienced qualitative researcher to review your codes and themes with you, see our qualitative data analysis support.
Sources and further reading
- Braun V, Clarke V. Using thematic analysis in psychology. Qualitative Research in Psychology 2006;3(2):77-101
- Braun V, Clarke V. One size fits all? What counts as quality practice in (reflexive) thematic analysis? Qualitative Research in Psychology 2021;18(3):328-352
- Thematic Analysis: resources from Virginia Braun and Victoria Clarke
- Byrne D. A worked example of Braun and Clarke's approach to reflexive thematic analysis. Quality & Quantity 2022;56:1391-1412