How to Write the Methodology Section of a Research Paper
Most methods sections are written as a diary of what happened. They should be written as a set of instructions. Here's the structure reviewers expect, the checklist to find before you draft, and the difference between a paragraph that passes and one that gets sent back.
The one job this section has
Every other section is about what you found and what it means. The methods section is about whether anyone should believe you. The ICMJE recommendations put it plainly: identify the methods, apparatus and procedures in enough detail that others could reproduce your results, and be clear about how and why the study was done a particular way.
That gives you a usable test for every sentence you write. Ask: could a competent stranger in my field repeat this step without emailing me? If the answer is no, the sentence is too vague. "Participants completed a short survey" fails. "Participants completed the 10-item Perceived Stress Scale online via Qualtrics, in a single session lasting approximately 12 minutes" passes.
Methods or methodology?
Strictly speaking, your methods are what you did and your methodology is the reasoning that makes those methods appropriate for your question. Journals mostly use "Methods"; taught masters and doctoral programmes often ask for "Methodology" and expect a paragraph of justification with it. Use whichever heading your guidelines specify — but include the why either way. A design choice with no stated rationale reads like an accident.
The four blocks
Almost every methods section, in any discipline, is built from four blocks in this order. APA's Journal Article Reporting Standards (JARS-Quant) ask for essentially the same set: the research design, participant characteristics, sample size, materials and instruments, outcome measures, and data-gathering procedures.
- Design. Name the design in the first sentence — cross-sectional survey, randomised controlled trial, semi-structured interview study, retrospective cohort. Then say why it fits the question.
- Participants, sample or materials. Who or what was studied, where they came from, and the eligibility and exclusion criteria. Report how the sample size was decided — a power calculation, saturation, or a practical constraint you state honestly.
- Procedure. What happened, in order, with enough detail to repeat it. Include instruments and their validated versions, apparatus with manufacturer, timing, and randomisation or blinding if relevant.
- Analysis. How the data were prepared and analysed: statistical tests and the software and version used, or the coding approach and analytic framework for qualitative work. Say how you handled missing data and outliers — reviewers notice when nobody mentions it.
Find your reporting checklist before you draft
This is the step that saves the most time, and almost nobody does it first. For most study designs there is already a published checklist of everything the methods must contain, indexed by the EQUATOR Network: CONSORT for randomised trials, STROBE for observational studies in epidemiology, PRISMA 2020 for systematic reviews, and COREQ for interview and focus-group research. Psychology and the social sciences have APA's JARS, in quantitative, qualitative and mixed-methods versions.
Open the relevant checklist, draft against its items, and you will not get a revision request that begins "the authors do not report whether…". Many journals now require the completed checklist at submission anyway.
Qualitative work needs one extra move
If your data are interviews, observations or documents, a list of procedures is not enough. JARS-Qual asks researchers to describe how their own backgrounds and perspectives shaped the research — reflexivity — and to show the structured ways they tracked it, such as memos, field notes, journals or bracketing. Report your analytic approach by name (thematic analysis, framework analysis, grounded theory), say who coded, and say what happened when coders disagreed.
Ethics, and the sentence reviewers look for
If humans or animals were involved, the methods section needs an explicit approval statement. The ICMJE recommendations ask you to confirm that the study was approved by an independent local, regional or national review body — an ethics committee or institutional review board — and to state how informed consent was obtained.
The study was approved by the University of X Research Ethics Committee (reference 2025/0417). All participants gave written informed consent before taking part.
Declare your AI use here
If an AI tool did any of the research work — screening abstracts, coding responses, generating or analysing data — that belongs in the methods. The ICMJE recommendations ask authors who used AI technology to describe it in sufficient detail to enable replication of the approach, including the tool, the version, and the prompts where applicable. Treat it exactly like any other instrument: name it, version it, and say what it was used for. Related but separate: how you cite an AI tool in the reference list, and whether the use counts as assistive writing or research conduct.
Every scale, protocol, software package and prior study needs a reference. Drop the details into our free APA citation generator and copy a correct reference and in-text citation for each one.
Tense, voice and length
Use the past tense — you are describing something already done. APA's verb-tense guidance puts the description of your procedure in the past tense or present perfect, and asks you to stay consistent across adjacent paragraphs; the present tense belongs with your conclusions later on.
On voice: passive constructions are traditional here ("samples were incubated") and many lab manuals still require them, but APA actively prefers the active voice and first person for describing your own actions ("we incubated the samples"). Follow your target journal or module brief, and be consistent. On length, there is no magic word count — write to the reproducibility test, then move overflow detail to an appendix or supplementary file.
Before and after
Here is the same procedure, written twice.
Before: We surveyed a number of students about their stress levels using a well-known questionnaire, and the data were analysed statistically.
After: We recruited 214 undergraduates (aged 18–24) from two UK universities by course-mailing list between March and May 2025. Participants completed the 10-item Perceived Stress Scale online in Qualtrics; the target sample size was set by an a priori power analysis for a medium effect (f² = 0.15, α = .05, power = .80). Scores were compared across year groups using one-way ANOVA in R 4.4.1, with Tukey HSD post hoc tests.
The second version is not longer for the sake of it. Every added clause answers a question a reviewer would otherwise have to ask.
Write it up with sources you can verify
CiteGood drafts from real papers and formats every citation for you — so the instruments and protocols you cite in your methods actually exist.
Try CiteGood free →Frequently asked questions
What is the difference between methods and methodology?
Strictly, methods are what you did (the procedures, instruments and analyses) and methodology is the reasoning behind why those methods suit your question. Many journals and departments use the two words interchangeably as a section heading, so check your own guidelines — but a strong section always covers both the what and the why.
How long should the methodology section be?
There is no fixed word count. The test is functional, not numeric: the ICMJE recommendations say the methods should be detailed enough that others with access to the data could reproduce the results. Write to that standard, then cut to fit your journal's or department's word limit, moving overflow detail into supplementary material or an appendix.
Should the methodology be written in past tense?
Yes — you are describing something you already did, so APA's verb-tense guidance puts the description of your procedure in the past tense (or present perfect). Keep the tense consistent within and across adjacent paragraphs; save the present tense for conclusions and implications in the discussion.
Do I need to mention ethics approval in the methods section?
If your study involved human participants or animals, yes. The ICMJE recommendations ask authors to state that the research was approved by an independent review body such as an ethics committee or institutional review board, and to say how informed consent was obtained. Name the committee and quote the approval reference number.
Do I have to declare that I used AI in my methodology?
If AI did any of the research work — screening, coding, analysis, image generation — then yes. The ICMJE recommendations ask authors who used AI technology to describe its use in the methods in enough detail to enable replication, including the tool, the version and the prompts where applicable. Using AI as an assistive writing tool is usually disclosed separately, and AI cannot be listed as an author.