How to Report Qualitative Data Analysis?

What Should the Analysis Section Explain?

• Analytic approach and rationale
• Data preparation
• How codes were developed
• How categories or themes were developed
• Inductive, deductive, or combined logic
• Researcher roles
• Use of software
• Reflexivity
• Strategies supporting analytic quality

Choosing an Analytic Approach

The choice should fit the research question, qualitative design, epistemological position, data type, and intended form of knowledge. Common approaches include thematic analysis, content analysis, grounded theory, phenomenological analysis, discourse analysis, and narrative analysis.

Thematic Analysis

Thematic analysis identifies and develops patterns of meaning across a dataset. It is not one fixed procedure; structured, coding-reliability, and reflexive versions may differ substantially.

Familiarisation → Initial coding → Pattern development → Candidate themes → Review → Definition and naming → Analytic narrative

Reflexive Thematic Analysis

In reflexive thematic analysis, the researcher actively interprets the data. Themes are developed through engagement with the dataset rather than treated as objects waiting to be discovered. Inter-coder agreement is therefore not automatically required.

Content Analysis

Content analysis may be conventional/inductive, directed/deductive, or summative. The chosen variant and its logic should be stated clearly.

Inductive, Deductive, and Hybrid Coding

Inductive: codes are developed mainly from the data.

Deductive: analysis is guided by theory, a framework, research questions, or a coding scheme.

Hybrid: predefined and emergent codes are combined; their roles should be made explicit.

Codes, Categories, and Themes

Code: an analytic label applied to a meaningful segment of data.

Category: a higher-order grouping of related codes.

Theme: an analytic pattern that says something meaningful about the research question.

Codebooks and Theme Development

A codebook may include code names, definitions, inclusion/exclusion rules, examples, and parent categories. Themes should move beyond topic labels and articulate a meaningful pattern in the data.

Grounded Theory Analysis

Grounded theory often integrates data collection and analysis. Depending on the tradition, analysis may involve initial coding, focused coding, constant comparison, category development, theoretical sampling, memo writing, and theoretical integration.

Open, axial, and selective coding are associated with particular grounded theory traditions and are not compulsory in every grounded theory study.

Phenomenological Analysis

Phenomenology is not a single analytic technique. Researchers may use descriptive phenomenology, interpretive phenomenology, IPA, or another clearly specified approach, with reporting consistent with its philosophical and analytic commitments.

Discourse, Narrative, and Document Analysis

Discourse analysis: examines how language constructs meanings and social realities.

Narrative analysis: examines how experience is storied, sequenced, and linked to identity and context.

Document analysis: should explain document sources, selection criteria, context, authenticity, and analytic procedure.

Transcription and Data Preparation

State whether transcription was verbatim, cleaned, or conversation-sensitive. If automated transcription was used, report whether transcripts were checked by a human researcher.

AI-Assisted Transcription or Coding

Report the tool, its purpose, human verification, privacy safeguards, and who made analytic decisions. AI assistance does not remove the researcher's methodological or ethical responsibility.

Qualitative Analysis Software

MAXQDA, NVivo, and ATLAS.ti can support organisation, coding, retrieval, comparison, and visualisation. They do not themselves perform the scholarly interpretation.

Multiple Researchers and Inter-Coder Agreement

Describe who coded which data, whether a shared codebook was developed, how disagreements were discussed, and how final analytic decisions were made. Agreement coefficients are appropriate in some structured approaches but not mandatory across all qualitative methodologies.

Reflexivity

Reflexivity considers how the researcher's position, prior experience, assumptions, and relationships may shape knowledge production. Insider status may provide contextual insight while also creating assumptions or power dynamics that should be addressed transparently.

Saturation

If saturation is used, explain how it was defined, what analytic unit it referred to, whether new interviews continued to add information, and when the decision was made. Saturation is not required in every qualitative tradition.

Negative Cases and Triangulation

Negative or deviant cases should not simply be removed; they may clarify theme boundaries and support alternative explanations.

Triangulation may involve data, researchers, methods, or theory, but should be methodologically justified rather than added mechanically.

Member Checking and Audit Trail

Member checking may involve summary confirmation, transcript review, discussion of preliminary interpretations, or theme discussion, but it is not obligatory in every methodology.

An audit trail may include coding versions, theme revisions, memos, sampling decisions, and analytic meeting notes.

Using Participant Quotations

Quotations support analytic claims but do not replace analysis. They should be relevant, non-repetitive, and appropriately anonymised.

Theme → Analytic explanation → Subtheme where needed → Illustrative quotation → Analytic interpretation

Quantifying Qualitative Findings

Frequencies may be useful in some content analyses, but in interpretive work thematic importance should not be reduced to how many participants mentioned something.

Suggested Methods Flow

Analytic approach → Methodological grounding → Data preparation → Coding → Theme/category development → Researcher roles → Software → Quality/reflexivity

Example Thematic Analysis Paragraph

Example: Interviews were transcribed verbatim and reread to support familiarisation with the dataset. Data were examined using an inductive thematic analysis aligned with the research question. Initial codes were developed from meaningful data segments and grouped into candidate themes. Themes were reviewed against the full dataset and research question, refined, and named. Analytic memos were maintained throughout the process and decisions concerning theme development were documented.

Common Mistakes

• Stating only that “content analysis was conducted”
• Treating thematic and content analysis as synonyms
• Failing to specify methodological grounding
• Reporting codes as themes
• Using topic headings as themes
• Using quotations instead of analysis
• Judging theme importance only by frequency
• Making the researcher's role invisible
• Implying software “did the analysis”
• Calculating kappa in every qualitative study
• Assuming member checking is always required
• Claiming saturation without explanation
• Excluding deviant data
• Stopping grounded theory at coding without theoretical development
• Calling any thematic analysis phenomenological
• Failing to report AI-assisted analysis

Before Submission, Check

• Is the analytic approach named clearly?
• Does it fit the question and design?
• Are transcription and preparation described?
• Is code development explained?
• Is theme/category development explained?
• Is the inductive/deductive logic clear?
• Are researcher roles described?
• Is software use characterised accurately?
• Is reflexivity addressed appropriately?
• Is saturation explained if used?
• Do quality strategies fit the methodology?
• Do quotations support rather than replace analysis?
• Are deviant cases considered where relevant?
• Is participant confidentiality protected?