Skip to main content

Complete guide

Thematic Analysis: A Complete Guide

Thematic analysis is a method for identifying, analysing, and reporting patterns of meaning (themes) across a qualitative dataset. It is the most widely used qualitative analysis method in the social, health, and management sciences — flexible across research questions, theoretical frameworks, and data types, from interviews to focus groups to open-ended survey responses.

This guide covers what thematic analysis is, the six phases of doing it, the main types you can choose between, and how to keep an AI-assisted thematic analysis defensible to examiners and journals.

What is thematic analysis?

Thematic analysis is the process of coding qualitative data and developing those codes into broader themes that answer your research question. A theme captures something important about the data in relation to your question and represents a patterned response or meaning across the dataset.

Unlike methods tied to a single epistemology, thematic analysis is theoretically flexible: it can be used within realist, constructionist, or critical framings. That flexibility is its strength and its risk — without a clear account of how themes were developed, a thematic analysis can look unsystematic to an examiner.

The six phases of thematic analysis

Braun and Clarke's (2006) six-phase framework remains the standard reference for doing thematic analysis. The phases are recursive, not strictly linear — you move back and forth as your analysis develops.

  1. 1Familiarisation — immerse yourself in the data: read and re-read transcripts, note initial ideas.
  2. 2Generating initial codes — systematically code interesting features across the entire dataset, collating data relevant to each code.
  3. 3Searching for themes — collate codes into potential themes, gathering all data relevant to each candidate theme.
  4. 4Reviewing themes — check the themes work against the coded extracts and the full dataset; refine, merge, or split.
  5. 5Defining and naming themes — refine the specifics of each theme and the overall story; generate clear definitions and names.
  6. 6Producing the report — select compelling extracts, relate the analysis back to your research question and literature, and write up.

Types of thematic analysis

"Thematic analysis" is an umbrella term. The three main variants differ in how coding is done and what counts as rigour:

  • Reflexive thematic analysis (Braun & Clarke) — themes are actively developed by the researcher as interpretations; reflexivity, not inter-coder reliability, is the rigour marker. The most widely used variant.
  • Codebook thematic analysis — a structured codebook (often partly developed in advance) is applied to the data; suits team research and more deductive questions.
  • Coding-reliability thematic analysis — multiple coders apply a fixed codebook and inter-coder agreement (e.g. Cohen's kappa) is reported; suits positivist framings.

Inductive vs deductive thematic analysis

Inductive thematic analysis develops codes and themes from the data itself, without trying to fit them into a pre-existing frame — useful when little is known about a topic. Deductive (theoretical) thematic analysis approaches the data with pre-defined codes drawn from theory or prior research. Most real analyses are a blend: you may start with a few a priori codes and remain open to emergent ones.

Keeping AI-assisted thematic analysis defensible

AI tools can speed up coding by surfacing candidate evidence — but a thematic analysis is only defensible if you can show how themes were developed and that you, not a tool, made the interpretive decisions. The safe pattern: AI surfaces candidates, you accept or reject each one, and every decision is recorded in an audit trail you can hand to an examiner.

QualIntel OS is built for exactly this: methodology-aware modes for reflexive, codebook, and coding-reliability thematic analysis, an audit trail that documents how each theme was built, and an auto-generated AI disclosure statement for your methods chapter.

Frequently asked questions

What is thematic analysis in simple terms?

Thematic analysis is a way of finding and reporting patterns of meaning (themes) across qualitative data such as interview transcripts. You code the data, group the codes into themes, and use those themes to answer your research question. It is the most common qualitative analysis method across the social and health sciences.

What are the six steps of thematic analysis?

Braun and Clarke's six phases are: (1) familiarisation with the data, (2) generating initial codes, (3) searching for themes, (4) reviewing themes, (5) defining and naming themes, and (6) producing the report. The phases are recursive — you move back and forth rather than strictly forward.

What is the difference between inductive and deductive thematic analysis?

Inductive thematic analysis builds codes and themes from the data without a pre-existing framework, suiting exploratory questions. Deductive (theoretical) thematic analysis applies codes drawn from theory or prior research. Many studies blend the two — starting with some a priori codes while staying open to emergent ones.

Can I use AI for thematic analysis?

Yes, with disclosure and on the condition that you remain the analyst. AI can surface candidate evidence and help organise codes, but the interpretation must be yours. Keep an audit trail of every AI suggestion and your accept/reject decision so you can show how themes were developed. QualIntel OS does this automatically and generates a disclosure statement. Confirm your institution's AI policy.

Which type of thematic analysis should I use?

Reflexive thematic analysis suits interpretive, qualitative questions and is the most widely used. Codebook thematic analysis suits team-based or more deductive work. Coding-reliability thematic analysis suits positivist framings where inter-coder agreement is expected. Choose based on your epistemology and research question, not convenience.

Related guides

Run your thematic analysis in QualIntel OS

AI surfaces candidate evidence, you confirm every decision, and the audit trail writes itself. One project free, no card.

Start for free