Thematic Analysis
Read closely. Code with a question in view. Develop themes that explain patterns of meaning, then write an argument that readers can trace back to the material.
What you will learn
- Distinguish codes, topics, and themes.
- Choose a coherent approach and explain its assumptions.
- Work recursively through six phases of reflexive thematic analysis.
- Develop, review, define, and map candidate themes.
- Write a supported analysis and assess its quality.
Orient yourself: Use Sections 1–4 to choose the question and approach.
Work with the material: Follow the practice extracts through Sections 5–9. Keep your own notes beside the examples.
Write and reflect: Read Sections 10–11, watch the lecture, and complete the workshop in Section 12.
Understand the analytical task
Thematic analysis develops patterns of meaning across qualitative material to address a research question. Researchers build an interpretation through familiarisation, coding, theme development, and repeated engagement with the evidence.
The material may include interviews, focus groups, diaries, open survey responses, documents, or media texts. A study might investigate how journalists describe verification, how audiences understand a programme, or how a set of campaign materials represents responsibility.
The lecture presents thematic analysis as an accessible route into qualitative analysis. Its appeal comes from a combination of systematic work and interpretive flexibility. The researcher must still explain what the data represent, how the analysis was conducted, and why the resulting account is persuasive.
Distinguish three levels of analytical work
A code
A label for an analytically relevant feature: “Losing one’s voice through polishing.”
A topic
A subject discussed in the material: “AI and writing.” It tells the reader the broad area under discussion.
A theme
A developed pattern of meaning: “Making fluent text feel one’s own.” It proposes an interpretation to examine.
In reflexive thematic analysis, a theme has a central organising concept: the idea that connects its different observations. “Benefits”, “Problems”, and “Suggestions” may organise an interview guide. A thematic analysis asks what those answers mean together in relation to the question.
Choose the method through the intended answer
Thematic analysis suits questions about patterned experiences, meanings, representations, and practices. A question about the detailed sequence of conversational turns may call for conversation analysis; a question about the form and unfolding of a life story may call for narrative analysis. Decide what the analysis needs to preserve and explain.
Locate your approach within the thematic analysis family
Thematic analysis includes approaches with different assumptions about meaning, coding, and quality. State which approach you use and follow its logic consistently.
Braun and Clarke distinguish coding-reliability, codebook, and reflexive approaches. The table offers a starting comparison; individual methods within each group have their own procedures.
On a narrow screen, scroll the table sideways.
| Approach | Coding and theme development | Typical quality emphasis |
|---|---|---|
| Coding-reliability approaches | A structured coding frame supports consistent application, often by independent coders. | Clarity of categories, coder training, and an appropriate assessment of coding agreement. |
| Codebook approaches | A shared coding framework organises analysis and can evolve during the project. | A transparent framework, useful comparisons, and procedures suited to the particular method. |
| Reflexive thematic analysis | Codes and themes develop through the researcher’s situated interpretation and engagement with the material. | Reflexivity, conceptual coherence, analytical depth, and a convincing relationship between evidence and claims. |
Understand the role of the analyst in reflexive work
In reflexive thematic analysis, the researcher actively develops the themes. Their questions, theoretical commitments, experience, and attention shape the interpretation. These influences become matters for reflection and explanation.
Collaboration can enrich that work. Analysts may discuss different readings, revisit assumptions, and consider what another perspective makes visible. A team should explain what these conversations contributed to its analysis.
This chapter follows the reflexive approach and draws on Braun and Clarke’s 2006 paper alongside their later clarifications. Their six phases provide a guide to analytical work within that approach.
Frame a question and define the material
A focused question gives the analysis direction. A clear account of the material and the researcher’s perspective explains the basis on which an answer can be developed.
How do journalism students understand authorship when using AI writing tools for coursework?
The inquiry concerns meanings of authorship in students’ accounts. It can examine voice, understanding, responsibility, task completion, and the role of other people without deciding their significance in advance.
Bound the dataset
A data item might be one interview. A data extract is a selected portion of an item. The dataset is the material selected for the particular analysis; a larger research corpus may support several analyses. Record what was included, why it fits the question, and what context is available.
Selection should support a meaningful inquiry. Justify the experiences sought, the range of situations, and the depth of material. In reflexive thematic analysis, explain adequacy through the question, scope, richness, and analytical possibilities of the dataset. Treat saturation claims according to the assumptions of the chosen method.
Make two analytical orientations clear
Inductive or deductive emphasis
An inductive emphasis develops codes through engagement with the content. A deductive emphasis uses existing concepts or theory to direct attention. Both involve interpretation and assumptions.
Semantic or latent emphasis
A semantic emphasis works with explicit meanings. A latent emphasis develops an interpretation of assumptions or ideas that organise what is said. Explain how your claims are supported.
These orientations address different decisions and can be combined coherently. An analysis may begin close to participants’ wording and develop a more interpretive account of what ownership means. A theory-informed study may also examine explicit descriptions. State the actual emphasis and show it in the analysis.
For the worked example, we interpret the accounts as situated statements made in a university context. We begin with the wording of the extracts, then explore how voice, understanding, and dialogue enter accounts of authorship. The concepts guide attention while remaining open to revision.
Map the six phases
The six phases organise the work of learning the material, coding, developing themes, and writing an interpretation. Movement between phases allows the analysis to deepen.
- Become familiarRead, revisit, and make notes about the material and its context.
- Develop codesLabel features relevant to the question across the dataset.
- Generate candidate themesExplore meaningful connections among codes and their extracts.
- Review and reviseExamine the candidate themes against their evidence and the wider dataset.
- Define and nameClarify each theme’s concept, boundaries, and contribution.
- Write the argumentDevelop the interpretation through analytical narrative, extracts, and relevant literature.
Keep a record of thinking as well as sorting
A useful working record links source IDs, extracts, codes, candidate themes, and memos. An analytic memo records an interpretation, its supporting material, an alternative reading, and the next question to pursue.
Keep versions of the thematic map or theme definitions. Later, these records help you explain how a broad idea was divided, refined, or abandoned. They can support a transparent account of your reasoning without turning the phases into a rigid checklist.
I noticed ______ in ______. One possible interpretation is ______ because ______. A different reading would be ______. To examine this, I will revisit ______ and compare it with ______. My own assumptions may be influencing ______.
The following sections demonstrate each phase through a small practice set. In a full study, the work would involve sustained engagement with the complete dataset and its context.
Become familiar with the data
Familiarisation develops a sense of individual accounts and the dataset as a whole. Read for detail, tone, context, tensions, and the questions that the material opens.
Where recordings are available and relevant, listen alongside reading. Check transcripts, mark uncertainty, and retain meaningful context. Decide which speech features your analysis needs. If translation is involved, document choices that affect the interpretation.
Read the practice extracts before coding them
The six extracts below are invented comments from fictional student interviews. They provide a compact exercise in analytical decisions; they are not a completed empirical dataset.
On a narrow screen, scroll the table sideways.
| Source ID | Fictional extract | A question for close reading |
|---|---|---|
| P01 | “I ask it for a first sentence. Once I see that, I can write the rest.” | What role does the first sentence play in beginning the work? |
| P02 | “It polished my paragraph so much that I could no longer hear myself in it.” | How does the speaker connect polished language with personal voice? |
| P03 | “If I cannot explain a sentence in class, I take it out.” | What makes a sentence acceptable for the speaker to retain? |
| P04 | “We compare the output in our group and point out what sounds wrong.” | How does discussion enter the evaluation of the output? |
| P05 | “I show my tutor the prompt and the changes. Then we discuss why I kept some parts.” | What becomes visible and discussable through showing the process? |
| P06 | “I submit the wording as it comes if it answers the task.” | What criterion makes the wording usable in this account? |
Write an initial familiarisation note
An early note
“The accounts describe several relationships with generated wording: a starting aid, a threat to voice, something to defend, something to discuss, or a ready answer. I should examine what each speaker treats as a sufficient basis for using the text. P06 may complicate an interpretation centred on rewriting or personal voice.”
This note opens possibilities. It does not settle the thematic structure. Read the set again and ask how the extracts differ, what the interview question might have invited, and what surrounding material would be needed to interpret them well.
Develop initial codes
Coding attaches concise labels to features that matter to the question. In reflexive thematic analysis, code labels develop as your understanding of the material grows.
Choose meaningful segments while retaining enough surrounding material to understand them. One extract can receive several codes. Revisit labels that become too broad, duplicate another idea, or conceal an important difference. Work across the dataset so that vivid passages sit within a wider analytical engagement.
On a narrow screen, scroll the table sideways.
| Extract | Possible initial code | What the code directs attention to |
|---|---|---|
| P01 · A first sentence enables writing | Using generated text to cross the starting point. | The output’s role in initiating the student’s own activity. |
| P02 · A polished paragraph loses voice | Polish displacing recognisable voice. | A tension between fluency and a sense of ownership. |
| P03 · Removing a sentence one cannot explain | Understanding as a condition of retaining text. | The speaker’s criterion for accepting wording. |
| P04 · Comparing output in a group | Testing output through peer discussion. | Collective evaluation of generated wording. |
| P05 · Showing prompts and changes to a tutor | Making writing decisions open to discussion. | Visibility of the process and dialogue about choices. |
| P06 · Submitting wording that meets the task | Task fit as sufficient grounds for use. | A pragmatic criterion that may differ from voice or rewriting. |
Explore semantic and latent readings
“If I cannot explain a sentence in class, I take it out.”
P03 · Fictional practice extract
A semantic code might be “Removing sentences one cannot explain.” A more latent reading might be “Claiming authorship through the ability to account for the text.” The second proposes a relationship between understanding and ownership that needs examination across the wider material.
These readings offer different emphases. Explain the interpretation in relation to the question, context, and other extracts. A latent reading becomes persuasive through the argument and evidence supporting it.
Keep the coding thorough and adaptable
An inductive emphasis involves close attention to the material while allowing relevant concepts to develop. Line-by-line attention can help, but the appropriate segment and detail depend on the question and data. A deductive emphasis also requires careful reading and room to notice material that challenges the starting concepts.
Generate candidate themes
Theme development asks how coded observations connect around an idea that helps answer the research question. Begin to construct an analytical account across the material.
Bring relevant coded extracts together, then examine their relationship. P02 connects fluency with a loss of recognisable voice. P03 connects acceptable wording with the ability to explain it. Together, they suggest that generated text may need to become personally recognisable and answerable before it feels usable as one’s own work.
P03: “If I cannot explain a sentence … I take it out.”
Central idea: ownership is worked through recognition and the ability to account for the wording.
Develop a second candidate through a different connection
P04 and P05 direct attention towards discussion with peers and a tutor. A candidate theme, “Making judgement through dialogue,” could investigate how decisions about output become discussable and are worked through with other people.
The first candidate centres on personal ownership of wording; the second centres on how evaluation is carried out through interaction. Their relationship can be explored while their central ideas remain distinct.
Use a topic label as a starting point
A folder labelled “Writing quality” can organise early material. To develop it analytically, ask what “quality” means here and why that matters. P02 suggests a tension: polished language can weaken a speaker’s recognition of their own voice. This relationship gives the analysis something to explain.
Try writing each candidate as a sentence: “This theme shows that ______.” If the sentence only names an interview question or lists a subject, return to the coded extracts and develop the meaning further.
P01 may contribute context about beginning to write without fitting either candidate strongly at this stage. Keep it accessible. A focused report can select themes that address the question while documenting the analytical decisions behind that selection.
Review themes against extracts and the whole dataset
Review develops a stronger fit between the candidate themes, their supporting material, and the overall interpretation. It can involve combining, dividing, narrowing, or discarding candidates.
Review at two connected levels
Within each candidate
Read its collated extracts together. Identify the connecting idea, important variation, and material that fits poorly.
Across the dataset
Return to original items and the wider set. Examine missing context, overlooked passages, competing patterns, and the contribution to the research question.
In the practice set, P06 treats task fit as sufficient for using unchanged wording. This complicates a claim that all students require rewriting, personal voice, or the ability to defend every sentence. The developing account needs room for different criteria of acceptable use.
A review memo
“The first theme should concern a particular way of establishing ownership through voice and understanding. P06 uses a different criterion: whether the wording answers the task. I will narrow the theme’s claim and examine whether fuller accounts show how these criteria vary by situation.”
Use a thematic map to think
A map can expose overlap, weak connections, or a theme that is doing too much. Change it as the analysis develops. A common research question can connect the themes without requiring one overarching theme to contain every finding.
Define and name the themes
A theme definition explains its central idea, the meanings it connects, and the boundaries of the claim. A useful name helps the reader grasp that contribution.
Candidate 1 · Making fluent text feel one’s own
Central organising concept: Ownership is worked through personal recognition and the ability to account for wording.
Analytical contribution: P02 and P03 suggest that polished language becomes problematic when the writer cannot recognise their voice or explain what they retain.
Boundary: The theme concerns a particular account of ownership. P06 introduces a different basis for acceptable use: task fit.
Candidate 2 · Making judgement through dialogue
Central organising concept: Evaluation takes shape through discussion of output and writing decisions.
Analytical contribution: P04 and P05 describe different interactions through which generated text is examined: peer comparison and a discussion with a tutor.
Boundary: These extracts concern the work of discussing choices. They provide limited evidence about how the final coursework was evaluated.
Use subthemes when they earn their place
A subtheme can develop a particular facet while sharing the parent theme’s central idea. “Hearing one’s voice” and “Being able to explain the text” might become subthemes of the first theme if the fuller material supports sufficiently developed analysis of each.
A smaller project may communicate more clearly through two well-developed themes with variation discussed within them. Choose the structure through the material, the question, and the space available. Define what each level contributes.
Test the name by writing the argument
“Student experiences” gives little direction. “Making fluent text feel one’s own” identifies a process to explain. A phrase taken from a participant can also work, provided the accompanying analysis clarifies the pattern it represents.
This theme explains ______. Its central idea is ______. It brings together ______, supported by ______. Variation within it includes ______. It differs from ______ because ______. Its boundary or complicating evidence is ______. It answers the research question by ______.
Write an analytical narrative
Writing develops the analysis through a connected argument. Each theme explains a pattern, shows its evidence, examines variation, and contributes to answering the research question.
Open a findings section with a brief overview of the thematic structure. For each theme, establish the central idea, introduce contextualised extracts, and explain how their details support the interpretation. Keep returning to the question.
“Students discussed AI writing. P02 mentioned voice, while P03 mentioned explaining sentences.”
This records the subjects discussed. The next paragraph develops the relationship between them.
A model analytical paragraph
In these fictional extracts, ownership is described through recognition and the ability to account for a text. P02’s comment that the polished paragraph no longer sounded like them connects fluency with a loss of personal voice. P03 uses a different test, removing a sentence when they cannot explain it in class. Read together, the accounts suggest that usable wording involves more than a polished result: the writer seeks a recognisable or explainable relationship with it. P06’s willingness to submit unchanged wording when it meets the task complicates the reach of this interpretation. The proposed theme therefore captures one way of understanding authorship within the practice material.
Connect an extract with its analytical role
Choose extracts for the work they perform in the argument. Explain a key phrase, contrast two accounts, or examine an exception. Include enough context for the reader to follow the interpretation and use pseudonyms or identifiers consistently.
Quotations can support a claim, while the surrounding narrative develops its meaning. A report with many extracts still needs a clear argument. A report making broad claims needs evidence that lets the reader assess them.
Explain the method through the work actually done
We used reflexive thematic analysis to investigate ______. Our analysis was informed by ______ and emphasised ______ meanings. We became familiar with ______, coded across the dataset, and developed candidate themes through ______. We reviewed these against ______ and refined the definitions through ______. Reflexive memos examined ______. Discussions within the team contributed ______. The final account presents ______ within the scope of ______.
Use counts only for a clear purpose, with the counted unit and denominator stated. The prominence of an idea can depend on the interview prompts, the depth of discussion, and analytical relevance. Explain the reach of claims about this dataset and its possible relevance elsewhere.
When literature enters the findings or discussion, show how it helps interpret, extend, or question the account. Keep your own analytical contribution visible.
Practise reflexivity and evaluate quality
A strong reflexive thematic analysis makes its perspective and reasoning visible. Readers can assess the coherence of the method, the depth of engagement, and the relationship between claims and evidence.
Trace a researcher assumption into a decision
A reflexive memo for the worked example
“As a teacher, I may read rewriting as a sign of responsible authorship. That expectation draws me towards P02 and P03. P06 uses task fit as the criterion for acceptable wording. I need to examine this account without reducing it to a moral judgement, and reconsider how my preferred definition of ownership shapes the proposed themes.”
A positionality statement can identify relevant roles and experiences. Reflexive work then follows their consequences through reading, coding, theme development, and reporting. Record where a decision changed because an assumption became visible.
Use quality questions that fit the approach
- Coherence: Do the question, perspective, analytical procedures, and claims fit together?
- Engagement: Has the whole dataset received sustained attention, including material that complicates the preferred account?
- Theme development: Does each theme have a clear central idea and a useful analytical contribution?
- Evidence: Can the reader follow the interpretation through contextualised extracts and the surrounding argument?
- Distinction and connection: Are the themes sufficiently distinct, with their relationships explained?
- Reflexive openness: Does the report explain the researcher’s role and consequential decisions?
Braun and Clarke’s quality paper and Reflexive Thematic Analysis Reporting Guidelines help develop this evaluation. Use the guidance thoughtfully in relation to the study’s purpose.
Protect context and participants
Plan consent for participation, recording, quotation, and any sharing. In a teacher–student study, consider how assessment relationships may shape participation and the accounts offered. Review distinctive details and searchable quotations when protecting identities.
Participant feedback can clarify a passage, add another perspective, or contribute to a participatory design. Explain its purpose and treatment in the analysis. The choice should follow the methodology and the research relationships.
Use tools to support the analytical record
A spreadsheet can link source IDs, extracts, codes, and memos. Qualitative analysis software can help retrieve and organise material. Choose a tool you can use securely and explain the analytical decisions it supports.
A useful record makes the reasoning traceable while preserving the interpretive character of the work. Allow time for revisiting material and writing; both can change the thematic account.
Learn with the lecture and analysis workshop
Use the lecture to revisit the chapter’s main decisions. Pause to compare your own codes, candidate themes, and definitions with the worked example.
Open the lecture on YouTube. These approximate starting points follow the supplied transcript. The chapter develops the lecture’s reflexive approach with a connected practice example.
Attempt each task before opening the response
Activity 1 · Move from a topic to a candidate theme
Prompt: Your folder is labelled “AI writing quality”. How could P02 and P03 help develop it analytically?
Suggested response: Ask how quality relates to ownership. P02 connects polished language with loss of voice; P03 uses understanding as a criterion for retaining a sentence. A candidate such as “Making fluent text feel one’s own” connects the observations around an interpretive idea, which needs further review.
Activity 2 · Examine a competing interpretation
Prompt: P03 removes sentences they cannot explain in class. Could this concern assessment as well as ownership?
Suggested response: Yes. The reference to class may concern accountability, confidence, assessment, or several related meanings. Return to the surrounding account and compare other relevant material before deciding how strongly to develop a particular reading.
Activity 3 · Use a contrasting account
Prompt: What does P06 do to a claim that students require personal voice before accepting AI-generated wording?
Suggested response: It narrows that claim. P06 describes task fit as sufficient. Examine whether the contrast represents a different criterion, a particular situation, or a wider pattern in fuller material. Preserve the variation in the analysis.
Activity 4 · Choose a coherent quality procedure
Prompt: A researcher uses reflexive thematic analysis and asks a colleague to discuss alternative readings. How should this be reported?
Suggested response: Explain what the discussion contributed: questioning assumptions, considering variation, or refining the theme’s concept. The purpose is to deepen the interpretation. Describe that work and its consequences for the analysis.
Your independent analysis workshop
Use the practice set or an ethically appropriate dataset available to you. Produce a short analysis with a visible record of decisions.
- State one question, the dataset boundaries, and your interpretive emphasis.
- Write a familiarisation memo, including one expectation you bring to the topic.
- Code across the selected material and retain source IDs and context.
- Develop two candidate themes, each with a central organising concept.
- Review the candidates against their extracts and the full practice set.
- Draw a small thematic map and write a definition for each theme.
- Write 600–800 words of analysis, balancing argument and contextualised evidence.
- Add a short reflexive account of a decision you revised and the limits of the interpretation.
Completion check: A reader can follow each theme’s idea through its evidence and understand how the set of themes answers the question.
Glossary and further reading
Use these terms to describe the work precisely. The readings develop the six phases, methodological choices, and evaluation of reflexive thematic analysis.
- Data item
- One unit of source material, such as an interview or document.
- Data extract
- A selected passage or portion of a data item.
- Dataset
- The material selected for a particular analysis.
- Code
- A concise label for an analytically relevant feature of the material.
- Topic summary
- An account organised around a subject discussed in the data.
- Theme
- In reflexive thematic analysis, a pattern of shared meaning organised around a central idea.
- Central concept
- The organising idea that connects a theme’s different observations and gives it analytical coherence.
- Semantic emphasis
- Attention to explicit meanings in the material.
- Latent emphasis
- Interpretation of ideas or assumptions that organise the explicit content.
- Reflexivity
- Examination of how the researcher’s position, assumptions, and decisions shape the analysis.
- Analytic memo
- A record of developing interpretations, questions, comparisons, and consequential decisions.
- Thematic map
- A visual arrangement used to explore or communicate themes and their relationships.
Sources and guided reading
- Dr. USP. How to Do Thematic Analysis [Video]. YouTube.The accompanying lecture and transcript introduce coding, theme development, semantic and latent analysis, and the reflexive approach.
- Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101. Author repository record.The foundational six-phase account; read it alongside the authors’ subsequent clarifications.
- Braun, V., & Clarke, V. (2019). Reflecting on reflexive thematic analysis. Qualitative Research in Sport, Exercise and Health, 11(4), 589–597. https://doi.org/10.1080/2159676X.2019.1628806Clarifies the development of the approach and common conceptual mismatches.
- Braun, V., & Clarke, V. (2021). One size fits all? What counts as quality practice in (reflexive) thematic analysis? Qualitative Research in Psychology, 18(3), 328–352. https://doi.org/10.1080/14780887.2020.1769238Use to align the procedures, assumptions, and quality claims of the analysis.
- Braun, V., & Clarke, V. (2022). Thematic analysis: A practical guide. SAGE.An extended guide to designing, conducting, and communicating reflexive thematic analysis.
- Braun, V., & Clarke, V. Doing Reflexive TA. Thematic Analysis.Author guidance on the phases and recursive analytical work.
- Braun, V., & Clarke, V. Understanding TA. See also the FAQs and editor and reviewer guidance.Compare thematic analysis approaches and review assumptions about codes, themes, interpretation, and quality.
- Braun, V., & Clarke, V. (2024). Supporting best practice in reflexive thematic analysis reporting in Palliative Medicine: A review of published research and introduction to the Reflexive Thematic Analysis Reporting Guidelines (RTARG). Palliative Medicine, 38(6), 608–616. https://doi.org/10.1177/02692163241234800Guidance developed in health research that helps frame coherent and reflexively open reporting.
Test the strength of your theme
Choose one candidate from your notebook and place its name, definition, two extracts, and a competing reading on a single page. Ask whether the central idea remains clear throughout. Then use Content Analysis to compare how a different analytical purpose could organise the same material, or return to Research Methodology to connect the analysis with the design of the study.