Thematic Analysis in NVivo: Braun & Clarke Step by Step

Why use NVivo for thematic analysis?

Thematic analysis (TA) is the most widely used method for analysing interviews in student research, and NVivo is one of the most widely licensed qualitative data analysis programs at universities in the UK, Ireland, Australia and elsewhere. Searching “thematic analysis NVivo” usually means: how do I translate Braun and Clarke’s six phases into concrete steps in the software?

NVivo helps with the organisational side: keeping all transcripts in one project, coding consistently, retrieving everything coded under one label, writing memos and comparing groups. It does not help with the interpretive side: deciding what matters, building themes with a central organising concept and writing an analytic story. Braun and Clarke (2022) are explicit that software is optional. Many excellent analyses are done on paper or in a word processor. Use NVivo if it makes your work easier, not because it makes your analysis look more rigorous.

The menu names below are described generally because they shift slightly between NVivo releases; the logic stays the same. Jackson and Bazeley (2019) give a detailed, screen-by-screen guide.

The six phases mapped to NVivo

This table links each phase of reflexive thematic analysis to typical NVivo functions.

Step by step: from transcripts to codes

1. Set up the project. Create a new project and import your anonymised transcripts (Word, PDF or text). Create a case for each participant and assign attributes (for example age group, degree programme, year of study) via a case classification. This lets you compare groups later.

2. Familiarise yourself. Read each transcript fully before coding. Use annotations for short comments on a passage and a memo per transcript for your overall impressions. In reflexive TA, also keep a reflexivity memo on your own assumptions.

3. Code. Select a passage and create a new code or drag it onto an existing one. Name codes so they capture meaning, not just topic: “paid work squeezes study time” is more useful than “work”. You can code one segment to several codes. Switch on coding stripes to see which codes you have applied in a document.

4. Refine the code list. After two or three transcripts, look at your code list. Merge duplicates, rename vague codes and write a short description for each code (in the code’s properties). Keep a dated memo noting what you changed and why.

From codes to themes: hierarchy, maps and memos

In phase 3 you look for patterns of shared meaning. In NVivo you can do this in two ways that work well together:

- Code hierarchy: create a parent code for each candidate theme (for example “Belonging is earned, not given”) and drag related codes under it as child codes. Keep a “miscellaneous” parent for codes that do not fit yet. - Maps: a mind map or concept map lets you sketch candidate themes and their relations visually before you restructure the hierarchy.

Write a theme memo for each candidate theme: what is its central organising concept, which codes belong to it, which do not, and what tensions you see. Braun and Clarke (2022) warn against “topic summaries” (themes like “experiences of work” that only collect everything said about a topic). A good theme makes a claim about the data.

Reviewing themes with queries

In phase 4 you check whether each theme is coherent and whether it fits the whole data set. Open a parent code and read all its segments in context. Then use queries carefully:

• Text search query: find passages with certain words that you may have missed (for example “belong”, “fit in”, “outsider”) and check whether they belong to a theme.
• Coding query: retrieve segments coded at one code AND another, or at a code for a specific group of cases.
• Matrix coding query: a table of themes by participant attributes (for example first-year vs final-year students). Use it to see where a theme appears, not to rank themes by frequency.
• Word frequency / word cloud: useful for familiarisation, but not an analysis in itself.

Exporting and writing up

For phase 6, export the coded segments for each theme (for example as a Word or PDF report) and your theme memos. Choose vivid, representative extracts and make sure your analytic commentary explains what each extract shows. Do not let quotations speak for themselves.

For your appendix, examiners often appreciate a code list with descriptions or a thematic map. A suitable methods sentence: “Data were managed in NVivo [version]. Following Braun and Clarke’s (2022) reflexive thematic analysis, transcripts were coded inductively; codes were organised into candidate themes in a code hierarchy and reviewed against the full data set. The software supported data management; all analytic decisions were made by the researcher.”

Alternatives to NVivo for thematic analysis

The same workflow works in other tools. In ATLAS.ti you work with quotations, codes, code groups and networks; in MAXQDA with the code system, memos, retrieved segments and the code matrix browser. Free options include Taguette and QualCoder, and Professor Digital’s free /codieren supports a codebook-style approach with category definitions, memos and export. A spreadsheet (one row per coded extract) or paper and highlighters also work well for small projects. Choose the tool that keeps you close to your data.