How to write a methodology section

The reproducibility standard

The test for every sentence you include or cut: could a competent researcher in your field, given this section, repeat the study and check the result? That standard decides the eternal "how much detail?" question. Standard techniques get a citation instead of a recipe ("thematic analysis followed Braun and Clarke, 2006"); anything you adapted, built, or chose between gets described and justified.

Methods are written in past tense ("participants completed…") because you are reporting what was done — except in proposals and preregistrations, where future tense is correct. Passive vs active voice is a venue convention; modern style guides increasingly accept "we recruited…".

Quantitative studies: the five standard subsections

Most empirical quantitative papers use the same skeleton, and reviewers navigate by it:

• Design — the study type (between-subjects experiment, cohort study, cross-sectional survey) and the key variables, stated up front.
• Participants — how many, recruited how, inclusion/exclusion criteria, and the sample-size justification (report the power analysis: effect size assumed, alpha, power, resulting n).
• Materials / measures — every instrument with its citation and, for scales, reliability in your sample (e.g. Cronbach’s alpha), not just in the original validation.
• Procedure — what participants actually experienced, in order, with timing; randomization and blinding described precisely.
• Analysis plan — which tests answer which hypotheses, the software, alpha level, and how you handled missing data, outliers, and assumption checks.

• Which statistical test should I use?
• Cronbach’s alpha calculator

Qualitative studies: rigor has its own vocabulary

Qualitative methods sections carry different obligations: name the tradition (thematic analysis, grounded theory, IPA, ethnography) with its citation and follow its actual steps; describe sampling logic (purposive on what criteria?) and the saturation or information-power argument for the sample size; document the interview guide, recording, and transcription conventions; and lay out the coding process — who coded, how disagreements were resolved, how codes became themes.

Add a positionality/reflexivity statement where the tradition expects it, and report against the field’s checklist (COREQ for interviews and focus groups, SRQR more broadly). Journals increasingly require the completed checklist at submission.

• How to analyze interview data
• Reporting guidelines reference (COREQ, SRQR, CONSORT…)

Ethics, data, and transparency statements

Human-participant research needs the ethics sentence: approving body, protocol/reference number, and consent procedure. Increasingly expected alongside it: a data-availability statement, preregistration reference if one exists, and disclosure of any AI assistance where the venue requires it. These are cheap to include and expensive to be asked for in review.

If personal or sensitive data are involved, say where and how they were processed and stored — for European research, GDPR-compliant handling (and EU processing for services like transcription) is exactly the kind of detail ethics committees and reviewers now probe.

The reviewer objections you can preempt

Most methods criticism falls into a few repeatable patterns — audit your draft against them:

• "Why this sample size?" — no power analysis or saturation argument reported.
• "Justify this choice" — a non-obvious design decision (measure, threshold, analysis) presented without rationale or citation.
• "Could not reproduce" — missing parameters: stimulus timing, exclusion counts, software versions, prompt wording.
• "Analysis does not match hypotheses" — tests appear in Results that were never announced in Methods.
• "Validity unaddressed" — no word on reliability, manipulation checks, or the design’s known threats.

Where Kahubi fits

Kahubi keeps methods and manuscript in one place: statistics flows record what was actually run on your CSVs and write the analysis plan and results in APA form; the interview flow documents transcription (EU-processed) and coding with an audit trail; and the agent drafts the methods section from what was genuinely done in the project — in your voice, in a LaTeX or rich-text draft that compiles to PDF.