Reviewers often say that a paper stands or falls on its methodology section. That is because the methodology is the only part of the paper that lets a reader judge whether the results can be trusted. A striking finding means little if the reader cannot tell how the data were collected, who was studied, and how the analysis was done. A clear methodology section does two things at once: it convinces reviewers that the study was done properly, and it gives other researchers enough detail to repeat the work.
This guide explains what belongs in a methodology section, how to organise it, and how much detail to give. It covers research design, sampling, data collection, and analysis, with the common mistakes that lead to rejection or major revision. It follows naturally from our guides on writing a strong introduction and writing a literature review, and it prepares the ground for the results and discussion sections that come next.
The methodology section answers one question: how did you get your results? Everything in it should serve that question. A reader who finishes the section should know what kind of study was done, what or who was studied, how data were gathered, how they were analysed, and what steps were taken to keep the work honest and accurate. If a detail does not help the reader judge or repeat the study, it probably does not belong here.
There is a simple test for completeness. Imagine a competent researcher in your field trying to repeat your study using only your methodology section. Could they do it? If they would have to guess at the sample size, the instrument, the dose, the time period, or the statistical test, the section is incomplete. You do not need to describe every trivial step, but every decision that could change the outcome must be stated.
At the same time, the methodology is not a laboratory notebook. Reviewers do not need the brand of every generic item, a day-by-day diary, or a defence of textbook procedures. The art of the section is selection: report the decisions that matter, cite standard methods instead of re-explaining them, and spend your words where your study departs from the standard or where a reader could reasonably disagree with your choice.
Most methodology sections follow the same logical order, because it mirrors the order in which the study itself was done. Journals vary in their exact headings, so check the target journal's author guidelines, but the components below appear in almost every empirical paper.
Start by naming the design in plain terms: experimental, quasi-experimental, cross-sectional survey, cohort study, case-control study, case study, or qualitative design such as interviews or focus groups. State the design in the first sentence of the section if you can, because it frames everything that follows. Then explain why this design fits the research question. A design choice always involves trade-offs — a survey can reach many people but cannot prove cause and effect — and naming the trade-off shows reviewers you chose deliberately.
Describe where the study took place and who or what was studied. For human or animal research, give the population, the eligibility criteria, and how participants or animals were recruited or selected. Report the sample size and, where relevant, how it was decided — a power calculation for an experiment, a saturation argument for interviews, or a census of an available group. If some participants dropped out or were excluded, say how many and why; unexplained losses are one of the first things reviewers notice, because they can bias the results.
Sampling language matters. Random sampling, convenience sampling, purposive sampling, and snowball sampling mean different things, and each carries different limits on how far the findings can be generalised. Name the method you actually used, not the one that sounds best, and state the limitation honestly. A small convenience sample described honestly is far stronger than a convenience sample dressed up as representative.
Describe the tools you used to gather data: questionnaires, interview guides, laboratory assays, clinical measures, or equipment. For a published questionnaire or scale, name it, cite its source, and report its reliability in your sample where that is standard in your field. For a questionnaire you designed yourself, explain how it was developed and whether it was pilot-tested. For laboratory or field work, name the key materials and settings — the assay, the dose, the equipment settings — that another researcher would need to repeat the procedure. Standard, widely known equipment can be named briefly; unusual or modified equipment needs more detail.
Walk the reader through what actually happened, in chronological order: recruitment, consent, the intervention or observation, follow-up, and the time period covered. Give dates or durations where they matter, and state who collected the data and under what conditions. Be specific about anything that could affect the data — for example, whether interviews were recorded and transcribed, whether observers were blinded to group allocation, or whether measurements were taken at a standardised time of day.
State how the data were analysed, in enough detail for a reader to judge the analysis and repeat it. For quantitative work, name the statistical tests, the software and version, the significance level, and how missing data and outliers were handled. For qualitative work, name the analytic approach — thematic analysis, content analysis, grounded theory, or another named approach — and describe the coding process, including whether more than one coder was involved and how disagreements were resolved. In both cases, the analysis must match the research question and the design.
Close the methodology with ethics approval and quality measures. State the ethics committee or institutional board that approved the study and the approval reference where the journal requires it, and confirm that informed consent was obtained from participants or, for animal work, that welfare standards were followed. Also report the steps taken to protect quality: blinding, randomisation, calibration, pilot testing, or double data entry. These details are short to write and carry real weight with reviewers. Our guide on research ethics, consent, and animal welfare explains these requirements in more detail.
The shape of the methodology changes with the design, and reviewers judge each type by its own standards. Quantitative methodology is judged on sampling, measurement, control of bias, and statistical appropriateness. Qualitative methodology is judged on the richness of the data, the transparency of the coding, and the credibility of the interpretation — not on sample size, which is a quantitative concern applied to the wrong design. Mixed-methods studies must explain both strands and, crucially, how they were integrated: the point of mixing methods is that the two strands speak to each other, so a mixed-methods section that reports two separate studies without integration has missed its own rationale.
Whatever the design, keep the methodology consistent with the rest of the paper: the design named here must match the abstract and the research question, and the analysis described here must be the analysis whose results appear in the results section.
The first mistake is vagueness: writing "data were analysed statistically" instead of naming the test, or "a questionnaire was used" instead of naming and describing it. Vague methodology reads as if the author is hiding something, even when the real cause is haste. The second mistake is the opposite — drowning the reader in irrelevant detail while omitting the decisions that matter.
The third mistake is describing what should have happened rather than what actually happened. If recruitment fell short, if a measure was changed mid-study, or if some data were lost, say so. Reviewers understand that real studies hit problems; what they do not forgive is a methodology that describes an ideal study that never took place, because the results then cannot be trusted. The fourth mistake is tense and voice confusion: methodology is conventionally written in the past tense, because the study is complete, and in a direct style that states who did what. Overuse of the passive voice — "it was decided that" — hides responsibility and makes the section harder to follow.
Finally, many authors forget that methodology is also where the study's limitations begin to show. You do not need a full limitations discussion here — that belongs in the discussion — but a methodology that honestly notes its own constraints, such as a small sample or a single study site, prepares the reader for a balanced discussion later and builds credibility. Our guide on why research papers get rejected shows how methodology problems lead to desk rejection, and our guide on how to publish a research paper places the methodology in the full submission process.
The quickest way to see what a good methodology section looks like is to take one paragraph apart. Consider this invented example: a study surveys final-year veterinary students about their confidence in handling farm animals. A strong methods paragraph might read as follows. The study used a cross-sectional survey design. All final-year students at two veterinary schools were invited, and 118 of the 142 invited students returned completed questionnaires, a response rate of 83 percent. The questionnaire had 24 items covering handling confidence, prior farm experience, and training needs, and it was piloted with 12 students from the year below before use. Responses were summarised with descriptive statistics, and differences between the two schools were tested with a chi-square test.
Notice what each sentence does, and why none of them can be dropped. The first sentence names the design. The second gives the population, the sampling frame, and the achieved sample with its response rate, so the reader can judge how far the findings travel. The third describes the instrument and the pilot test, which is the evidence that the questions were understood. The fourth states the analysis in one line. A reviewer reading this paragraph knows exactly what was done, and a researcher in another country could repeat the survey from it. The paragraph also justifies the sample honestly: it does not claim the two schools represent all veterinary students, and the response rate is reported rather than hidden.
You can reuse this sentence order for almost any design: design first, then who was studied and how they were chosen, then what was measured and how it was tested, then how the data were analysed. For an experiment, the middle sentences become the intervention, the allocation, and the blinding. For interview research, they become the recruitment, the interview guide, and the coding approach. Whatever your design, finish by applying the repeatability test: could a competent researcher redo the study from your words alone? If not, the missing detail is usually the sample, the instrument, or the analysis.
| Component | What to report | Reviewer check |
|---|---|---|
| Research design | Design name and why it fits the question | Does the design answer the question asked? |
| Participants and sampling | Population, criteria, method, sample size, dropouts | Can the findings be generalised, and how far? |
| Materials and instruments | Tools, scales, assays, settings, pilot testing | Are the measures valid and reliable? |
| Procedure | Steps in order, timing, blinding, conditions | Could another researcher repeat the study? |
| Analysis | Tests or coding approach, software, missing data | Does the analysis match the design and question? |
| Ethics and quality | Approval, consent, welfare, quality checks | Was the study done honestly and safely? |
These authoritative external resources explain the standards and guidance mentioned in this article:
How long should a methodology section be?
There is no fixed length; it depends on the study's complexity. A straightforward survey study may need 800 to 1,200 words, while a complex experiment or mixed-methods study may need more. The test is completeness: can a competent researcher repeat the study from your description?
Should the methodology be written in past or present tense?
Past tense, because the study is complete. Present tense is used only for general truths, such as a statement about what a standard instrument measures, and for the description of a method that is still in use.
Do I need to justify every methodological choice?
Justify the choices a reader could reasonably question — the design, the sample size, and any departure from standard practice. Standard, widely accepted procedures can be named and cited without a long defence.
What if something went wrong during data collection?
Report it honestly. State what happened, how it was handled, and how it may affect the results. Reviewers respect transparency; a methodology that describes an ideal study that never took place destroys trust in the results.
Can I reuse my methodology from a previous paper?
Yes, where the method is genuinely the same, but adapt it to the current study and cite the earlier description rather than copying it silently. Check the journal's policy on text recycling, and always update sample sizes, settings, and dates.
Where do I describe ethics approval?
In the methodology, usually at the end of the section: the approving committee or board, the approval reference where required, and confirmation of informed consent or animal welfare standards. Some journals also ask for an ethics statement in the manuscript declarations.