Premium Writing, Design & Advisory · Dubai, UAE

ACADEMIC SUPPORT · GCC

Reflexive Thematic Analysis or a Codebook? Choose Before You Code

Both approaches can organise interview data, but they answer different methodological questions. Make the choice explicit before you promise a codebook, inter-coder agreement or themes that will “emerge” without interpretation.

Anonymized interview transcripts and color-coded notes arranged into a qualitative theme map
QUICK ANSWER

Which thematic-analysis approach fits your study?

Choose a reflexive approach when your aim is to interpret patterns of shared meaning and your researcher position is part of the analysis. Choose a codebook approach when a defined coding framework is useful for comparing material consistently across cases, people or time points. A coding-reliability design is different again: it treats agreement between independent coders as a quality concern. None of these is inherently superior. Your research question, philosophy, team arrangement and planned claims must fit the chosen approach.

Virginia Braun and Victoria Clarke’s thematic-analysis resource explains that thematic analysis is a family of approaches. Their FAQ distinguishes reflexive coding from using a fixed codebook to direct the analysis. A published worked example of reflexive thematic analysis clarifies why treating all forms of thematic analysis as one method creates reporting problems. Check your supervisor’s or journal’s methodological expectations as well as the method literature.

METHOD DECISION

Compare the promise you are making about the data

Researchers often write “Braun and Clarke thematic analysis” in a proposal while also planning a predetermined codebook and a Cohen’s kappa score. That combination needs examination. The issue is not whether a spreadsheet of labels is allowed; it is whether those labels direct coding and whether agreement is being used as evidence that one interpretation is objectively correct.

Decision Reflexive thematic analysis Codebook or framework approach
Analytic aim Develop an interpretive account of shared meaning in relation to a question. Organise material through a documented coding framework for structured comparison.
How codes develop Open and revisable as familiarity and interpretation grow. Defined and revised through an explicit framework-development process.
Role of multiple analysts Different readings can enrich reflexive discussion; consensus is not the quality target. Team application, discussion and framework consistency may be important to the design.
What to report Researcher positioning, interpretive choices, theme development and a persuasive analytic account. Framework origin, definitions, revisions, application and how comparisons answer the question.

This table simplifies a varied methodological field. In particular, a codebook approach does not automatically require a numerical reliability statistic. Name the specific approach you actually use and cite sources appropriate to it. Braun and Clarke’s evaluation questions help spot combinations of claims and procedures that do not fit.

WORKED EXAMPLE

Follow one fictional interview excerpt through the decision

Illustrative data, not a real participant quotation: A postgraduate studying remote-team onboarding records an interviewee saying, “I got the checklist on day one, but I still didn’t know whom to ask when a customer issue appeared.” The research question asks how new employees experience support during their first month.

Reflexive reading

The researcher might code the contrast between formal onboarding and practical access to help. Across interviews, they might develop a theme about “support on paper, uncertainty in action”—if the wider dataset and interpretation support it. They record why that reading was made and what alternative interpretations were considered.

Framework reading

A team might define categories such as formal resources, help-seeking routes and unresolved task barriers, with inclusion rules and an example for each. They can then compare how those categories appear across roles or sites. The framework may need revision after pilot coding.

What neither proves

A single sentence does not establish the prevalence of an experience, the quality of an organisation’s onboarding or a finished theme. Software can store codes and retrieve excerpts; it cannot decide what the findings mean.

The practical output differs. A reflexive account should explain an interpretive pattern with selected extracts and analysis. A framework account should show how the categories were constructed and applied, then explain what the case comparisons reveal. A table of repeated topics alone is not a compelling findings chapter.

BEFORE CODING

Write a one-page method decision before opening NVivo or Excel

  1. State the question: what experience, process or comparison is the dataset meant to illuminate?
  2. Name the approach: reflexive, codebook/framework or coding reliability, with a source that actually supports that procedure.
  3. Explain your assumptions: are themes understood as interpretive outputs, structured summaries or something else?
  4. Describe the team: who will code, discuss, revise and make final analytic decisions?
  5. Set the evidence trail: keep interview-context notes, coding decisions, analytic memos, framework versions where relevant and a clear chain from extracts to claims.
  6. Plan the write-up: decide how examples and interpretation will answer the question rather than merely list codes.

For reflexive analysis, the six phases described in the worked-example paper are recursive; returning to earlier coding or theme development is normal. For a structured codebook, a pilot on a small, varied sample can reveal ambiguous definitions before the full dataset is coded. These are different quality practices, so do not combine them by habit.

QUALITY CHECK

Make the methodology and results tell the same story

Review the proposal, ethics documents, analysis notes and final chapter together. If your methods section promises open interpretive coding but your results use a fixed list of interview questions as “themes,” explain the change or revisit the analysis. If the methods section promises a shared framework, show when the definitions were agreed and how revisions were handled. Do not claim inter-coder reliability if no independently coded sample and appropriate calculation exist; do not add a reliability statistic simply because a software menu offers one.

Consider three common weak statements. “Themes emerged from the data” hides the researcher’s decisions. “NVivo analysed the interviews” mistakes a tool for an analyst. “Two coders agreed, therefore the interpretation is valid” makes a claim that may not fit a reflexive approach. More transparent alternatives identify who interpreted the material, what procedure was used and how the illustrative extracts support each claim. The Braun and Clarke quality questions and authors’ FAQ are useful checks before submission.

READER QUESTIONS

Frequently asked questions

Can I use a codebook with reflexive thematic analysis?

Braun and Clarke distinguish a list of codes used to think with from a fixed codebook used to direct systematic coding. If the framework governs coding and consistency is the core aim, describe that approach accurately rather than labelling it reflexive by default. See the authors’ explanation.

Do I need two independent coders?

That depends on the approach and the project. Reflexive work can benefit from collaborative interpretation without treating agreement as a quality score. A structured coding-reliability design may require independent coding and an agreement procedure. Do not import a generic rule from another discipline without checking your method and institution.

Is NVivo required?

No. Qualitative software helps organise, retrieve and annotate data. It does not supply an epistemology, determine good themes or replace researcher judgement. Select a tool according to dataset size, team workflow, institutional access and data-security requirements.

Can one interview excerpt establish a theme?

No single illustrative extract establishes a pattern across a dataset. It may illuminate an analytic claim, but the claim needs a defensible relationship to the wider material and the research question. Frequency by itself is also not the only measure of importance in interpretive analysis.

NEXT READING

Build a defensible analysis brief

Before commissioning assistance, share your approved question, sample, interview guide, anonymised extracts, ethics limits and supervisor’s methodological guidance. Ask for a short rationale explaining the selected approach and a small worked coding demonstration. That is more useful than requesting “themes from transcripts” without specifying the intended analytic logic. Keep authorship, interpretation and final academic responsibility with the researcher.

The Braun and Clarke resource site, quality questions and worked example support the methodological distinctions in this guide. Tool selection is a separate decision from method fit.

NEXT STEP

Make your qualitative analysis coherent

Labeeb can help you review the analysis plan, coding workflow and presentation of findings for methodological clarity while you retain ownership of the research, decisions and academic work. Bring your institution’s rules and a small anonymised sample to a consultation so the support can be scoped responsibly.