Written for replication, read for credibility
The stated purpose of a methods section is reproducibility: a competent researcher in your field should be able to run your study from your description alone. That is the standard to write against, and it is a useful one because it is testable. Hand the draft to a colleague and ask which step they would have to guess at.
In practice, most readers are not replicating anything. They are deciding whether to believe your results, and the methods section is where they form that judgment. Vague sampling, an unnamed instrument, or an analysis chosen after the fact all register as reasons for caution, whatever the results say.
These two audiences pull in the same direction, which is convenient. Detail sufficient for replication is also detail sufficient for trust. Where they conflict — long procedural material that helps a replicator but slows a reader — the appendix or a supplementary file resolves it.
Justification is the part most drafts skip
The difference between an adequate methods section and a strong one is whether choices are explained. “Semi-structured interviews were conducted” tells a reader what happened. “Semi-structured interviews were chosen because the study needed comparability across participants while leaving room for practices the literature had not anticipated” tells them why, and pre-empts the obvious reviewer question.
Justify the decisions where a reasonable researcher could have chosen otherwise: the design, the sampling strategy, the instrument, the analysis, and the sample size. You do not need to defend routine steps such as randomizing trial order. A useful filter is to ask whether a reviewer might write “why this and not the alternative?” in the margin.
Justification does not mean claiming your method was optimal. Constraints are legitimate reasons, stated plainly: “a cross-sectional design was used because the cohort was only accessible during a single term” is honest and complete. The corresponding limitation then goes in the discussion, where a reader is already expecting it.
The standard order, and why it runs that way
Most methods sections move from design, to participants or materials, to procedure, to analysis, to ethics. The order is not arbitrary: it goes from the general frame to specific operations, so each subsection makes sense in light of the one before it. Readers scan for the subheadings, so use them even when the journal does not require them.
Field conventions vary in the details. Laboratory papers often separate Materials from Apparatus; qualitative work adds sections on researcher positionality and on trustworthiness; mixed-methods papers need an explicit statement of how the strands connect and in what sequence. Look at two recent papers in your target journal and follow their subheadings rather than inventing your own.
Keep results out of this section entirely. A sentence such as “participants performed better after training” belongs in the results, even when it appears while describing the procedure. The boundary is strict for a reason: it lets a reader evaluate the method before knowing what it produced.
- Design — the approach and its fit with the question.
- Participants or materials — who or what, how selected, how many.
- Procedure — chronological, from the participant’s perspective.
- Analysis — the tests or coding, plus software and assumptions.
- Ethics — approval, consent, data protection, registration.
How much detail is enough
The threshold is functional, not a word count: include anything that would change the result if it were done differently. Room lighting rarely matters; the interval between training and testing usually does. When you are unsure, ask what you would want to know if you were reading someone else’s version of your study.
Standardized procedures can be cited rather than described. If you used a published protocol or a validated scale, one citation plus the details you changed is enough, and it is more useful than a paraphrase of the original paper. Reserve full description for anything you designed yourself.
Length problems in this section are usually structural. Move interview schedules, full item lists, and detailed coding frames to appendices or supplementary material, and keep the running text at the level a reader can follow in a single pass. Word limits are a reason to relocate detail, not to omit it.
Qualitative methods carry the same obligations
Qualitative sections sometimes get written as if the method were self-evident, with a sentence naming thematic analysis and nothing else. The same requirements apply: how participants were selected and why those people, how much data was collected, how coding proceeded, who did it, and how the analysis moved from codes to themes. Naming the specific tradition you are following, and citing it, does a lot of this work.
Trustworthiness needs its own treatment, since the checks differ from quantitative ones. Describe what you actually did — member checking, an audit trail, independent coding of a subset, negative case analysis — rather than asserting rigor in general terms. Where you claim saturation, say what told you it had been reached.
Positionality belongs here in many traditions, briefly and concretely. State the relationship between you and the setting and how it shaped access and interpretation. Two or three sentences that name a specific effect are worth more than a paragraph of general reflexive language.
Tense, voice, and person
Write the methods in past tense throughout, because the study is finished. The most common tense error is text carried over from a proposal, where the same material was necessarily in the future: “participants will complete two sessions” has to become “participants completed two sessions”. Search for “will” before submitting.
The active-versus-passive question is genuinely field-dependent. Many journals now prefer “we randomized participants to two conditions” for its clarity about who did what; others still expect the impersonal “participants were randomized”. Either is acceptable prose. Pick the one your target journal uses and apply it consistently rather than alternating.
Avoid the hybrid that appears in a lot of drafts: passive constructions with a hidden agent, such as “it was decided that the second measure would be dropped”. Decided by whom, and when? Naming the actor and the timing turns an evasive sentence into a methodological fact.
