Writing the Methods Section (With Examples)

The Methods section (also called “Materials and Methods”) is the part of a research manuscript most directly graded on a single question: could someone else, working only from what you wrote, actually do what you did? Every other section of a paper is judged on clarity and argument. The Methods section is judged on whether it functions as a set of instructions.

This guide gives a working structure, a worked example, a reproducibility checklist drawn from the major reporting standards (ICMJE, ARRIVE, CONSORT, STROBE), and answers to the specific questions researchers search for when drafting this section.

What the Methods section has to do

The International Committee of Medical Journal Editors (ICMJE) states the standard plainly in its Recommendations for the Conduct, Reporting, Editing, and Publication of Scholarly Work in Medical Journals: “The Methods section should aim to be sufficiently detailed such that others with access to the data would be able to reproduce the results.” That single sentence is the test to write against for every paragraph in the section, regardless of discipline or journal.

ICMJE breaks the requirement into three concrete obligations:

Reproducibility of procedure. Report equipment manufacturers’ names, cite references for established methods rather than re-describing them from scratch, and describe new or modified methods in enough detail that a reader could carry them out.
Ethics oversight. Include “a statement indicating that the research was approved by an independent local, regional or national review body (e.g., ethics committee, institutional review board),” and address compliance with the Declaration of Helsinki where applicable.
Statistical transparency. Describe statistical methods “with enough detail to enable a knowledgeable reader with access to the original data to judge its appropriateness for the study and to verify the reported results” — including software and version numbers, which analyses were prespecified versus exploratory, and reporting confidence intervals rather than relying on p-values alone.

As of the current ICMJE Recommendations, a fourth obligation applies where relevant: if AI technology was used in conducting the study, its use must be described “in the methods section in sufficient detail to enable replication of the approach, including the tool used, version, and prompts where applicable.”

Journals vary in house style, but this underlying reproducibility standard is close to universal across ICMJE-following biomedical journals and is echoed, in field-specific form, by the major reporting guidelines indexed by the EQUATOR Network — CONSORT for randomized trials, STROBE for observational studies, and ARRIVE for animal research among them.

Standard structure of a Methods section

Subsection	What it answers	Typically included
Study design	What kind of study is this?	Design type (RCT, cohort, cross-sectional, qualitative, etc.), setting, and time period
Participants / materials	Who or what was studied?	Eligibility/inclusion-exclusion criteria, recruitment source, sample size and how it was determined, animal strain/cell line/reagent sourcing
Ethics and consent	Was this conducted appropriately?	IRB/ethics committee approval and protocol number, informed consent process, animal welfare approval
Procedures / interventions	What was actually done?	Step-by-step procedure, equipment with manufacturer and model, randomization and blinding approach, timeline
Outcome measures	What was measured?	Primary and secondary outcomes, instruments used, validity/reliability of measures
Statistical analysis	How was the data analyzed?	Statistical tests, software and version, significance threshold, handling of missing data, prespecified vs. exploratory analyses

Not every subsection applies to every study — a qualitative study will not have a “statistical analysis” subsection in the quantitative sense, and a secondary-data-analysis paper will not have a “recruitment” subsection. Use the subsections that match what was actually done, in the order a reader would need them to follow the logic of the study, and use bolded or italicized run-in headers for each if the journal’s house style allows it.

Worked example (illustrative composite)

The paragraph below is a composite example written for this guide to illustrate the level of detail a reproducible Methods section requires. It does not describe a real study, dataset, or institution.

Participants. Adults aged 18–65 with a confirmed diagnosis of generalized anxiety disorder (GAD-7 score ≥10) were recruited from three outpatient clinics between March 2024 and September 2024. Exclusion criteria were current psychotropic medication use, a lifetime diagnosis of psychotic disorder, or inability to complete study questionnaires in English. The study was approved by the institutional review board (protocol #2024-0142), and all participants provided written informed consent.

Procedure. Participants were randomly allocated 1:1 to the intervention or waitlist-control arm using a computer-generated randomization sequence (block size of 4, stratified by clinic site) implemented via REDCap’s randomization module. Allocation was concealed from the assessing clinician until after baseline assessment. The intervention consisted of eight weekly 50-minute sessions delivered by licensed clinicians trained to a fidelity checklist (available in Supplementary Material).

Statistical analysis. The primary outcome (change in GAD-7 score from baseline to week 8) was analyzed using a linear mixed-effects model with participant as a random effect, using R (version 4.3.1, lme4 package version 1.1-35). Analyses followed the intention-to-treat principle. A two-sided alpha of .05 was used; effect sizes are reported with 95% confidence intervals. The statistical analysis plan was prespecified and registered prior to unblinding (see Preregistration).

Notice what makes this reproducible rather than merely descriptive: a named randomization method and block size, a named and versioned software tool, a specific alpha threshold, and an explicit statement of which analysis was prespecified. A version of this paragraph that said “participants were randomly assigned to groups and outcomes were analyzed statistically” would satisfy no journal’s methods requirement and would not let another team replicate the procedure.

Reproducibility checklist

Before submission, check the Methods section against this list. It draws on ICMJE’s general requirements plus the discipline-specific reporting guidelines that apply to a given study type.

Every method has a citation or a full description. Established methods are referenced, not re-derived; new or modified methods are described step by step.
Every instrument, reagent, or piece of equipment is identified. Manufacturer, model, and, where relevant to reproducibility, catalog or lot number.
Software is named with a version number. “Analyzed in R” is not reproducible; “R version 4.3.1 with the lme4 package (v1.1-35)” is.
Sample size derivation is stated or referenced, not just the final N. See the companion guide on justifying sample size in a manuscript for how to report a power calculation.
Ethics approval and consent are explicit, including the approving body and a protocol/reference number where the journal requires one.
Randomization and blinding methods are named, not just asserted, if the design uses them.
Statistical tests, thresholds, and software are all specified, and prespecified analyses are distinguished from exploratory ones.
A data and code availability statement is present or cross-referenced — see data availability statement for what that requires.
The applicable reporting guideline was followed, if the study type has one — see the table below.

Reporting guidelines by study type

Journals increasingly require authors to complete and submit the relevant checklist alongside the manuscript. Using the checklist while drafting, rather than retrofitting it before submission, produces a more complete Methods section with less rewriting.

Study type	Guideline	What it covers
Randomized controlled trial	CONSORT 2010	Trial design, randomization, blinding, outcome definitions, statistical methods
Observational study (cohort, case-control, cross-sectional)	STROBE	22-item checklist covering design, setting, participants, variables, and bias
Animal research	ARRIVE 2.0	The “Essential 10”: study design, sample size, inclusion/exclusion criteria, randomization, blinding, outcome measures, statistical methods, experimental animals, experimental procedures, and results
Systematic review / meta-analysis	PRISMA 2020	Search strategy, study selection, data extraction, synthesis methods
Qualitative research (interviews, focus groups)	COREQ	Research team, study design, analysis, and reporting for interview-based studies

All of these are indexed at the EQUATOR Network, which functions as the umbrella registry for reporting-guideline standards across study types. NIH’s rigor and reproducibility policy (NOT-OD-15-103, in effect for applications from the January 2016 due date onward) separately requires an authentication plan for key biological or chemical resources — cell lines, antibodies, and other biologics central to the proposed work — reflected in a Methods section as explicit reagent identification rather than generic descriptions like “using standard antibodies.”

Common reasons a Methods section fails peer review

Circular description instead of instruction. “Data were analyzed appropriately” or “standard procedures were followed” tells a reviewer nothing they can check.
Missing software versions. Statistical packages change default behavior between versions; an unversioned citation cannot be replicated exactly.
Results embedded in Methods. The Methods section describes what was done, not what was found — see the companion guide on writing the Results section for where that boundary sits.
No statement of what was prespecified. Reviewers and readers cannot distinguish a confirmatory finding from a post-hoc one without this.
Omitted or vague ethics approval. A missing IRB/ethics statement is one of the most common desk-rejection triggers at journals that require it.
Ignoring the applicable reporting guideline. Journals that require CONSORT, STROBE, or ARRIVE compliance will return a manuscript that omits checklist items, even if the underlying science is sound.

Frequently asked questions

What is an example of a good Methods section?

A good Methods section names its design, states eligibility criteria, describes procedures step by step with manufacturer/version information for any equipment or software used, states ethics approval, and specifies the statistical analysis plan including which comparisons were prespecified. See the worked example above for a paragraph-level illustration of the level of specificity involved.

How long should a Methods section be?

There is no fixed word count — length is set by what a reader needs to reproduce the study, not by a target. Journals with strict word limits typically allow supplementary material (a protocol document, a fidelity checklist, extended statistical code) to carry procedural detail that would otherwise crowd the main text; cross-reference it explicitly rather than omitting it.

Should Methods be written in past tense?

Yes. Methods describes what was done, so it is written in the past tense, typically in passive voice (“participants were randomized”) though many journals now also accept active voice (“we randomized participants”). Check the target journal’s author guidelines — this is one of the more common house-style variations between journals.

What is the difference between Methods and Materials and Methods?

They are the same section under different names. “Materials and Methods” is more common in the physical and life sciences, where reagents, specimens, and equipment need explicit listing; “Methods” is more common in social science and clinical journals organized primarily around procedure. Follow the target journal’s own section-naming convention.

Can the Methods section cite a previous paper instead of re-describing a procedure?

Yes, and ICMJE explicitly recommends it for established methods: cite a reference for the original method and describe only what was changed or added. Re-describing a well-established, unmodified procedure from scratch adds length without adding reproducibility.

Related CASRAI guides

Last verified: August 16, 2026, against the ICMJE Recommendations (icmje.org) and the ARRIVE 2.0 guidelines (arriveguidelines.org). ICMJE and reporting-guideline wording is periodically revised; re-check before treating exact quoted language as current beyond approximately 12 months from this date.

Writing the Methods Section So It Can Be Reproduced

Ask about Writing the Methods Section So It Can Be Reproduced