How to Write a Methodology Section for a Research Paper

A methodology section explains how you conducted your research and why you chose that approach. To write one, describe your research design, participants or data sources, data collection procedures, analysis methods, and relevant ethical safeguards. Give readers enough detail to evaluate the study and, where appropriate, reproduce its analysis. This emphasis on both explanation and transparency is central to the International Committee of Medical Journal Editors’ guidance on reporting methods.

Consider the difference between these two statements:

“We collected survey data and analyzed the results.”

“We administered an online questionnaire to doctoral researchers and used Spearman’s rank correlation to examine the relationship between database-use frequency and self-reported search confidence.”

The second statement still needs more detail, but it gives the reader something concrete to assess: a population, a collection method, two variables, and an analytical approach.

That is the goal of a strong methodology section. It should make your research decisions visible—not simply attach technical labels to them.

This guide explains how to structure the section, what to include, how requirements change across research types, and how to turn research records into a clear, defensible account.

What Is the Difference Between Methods and Methodology?

A useful distinction for writing is that methods describe the procedures, while methodology explains the reasoning behind the approach.

An interview is a method. The justification for using interviews to investigate how people interpret an experience belongs to the methodological explanation.

For example:

Method: “We conducted semi-structured interviews with laboratory managers.”

Methodological rationale: “Semi-structured interviews allowed us to explore how managers interpreted recruitment challenges while asking comparable core questions across laboratories.”

Your section should connect these two levels. The University of Manchester’s Academic Phrasebank treats describing procedures and explaining why particular methods were selected as complementary functions of methods writing.

For your manuscript, use the heading required by the journal or institution, whether that is Methods, Materials and Methods, or Methodology. Focus less on the label and more on whether the section explains the decisions readers need to understand.

What Should a Methodology Section Include?

Use the following as a drafting outline, not a universal set of mandatory subheadings.

Before drafting paragraphs, fill in these questions using your protocol, recruitment records, questionnaires, laboratory notes, search logs, and analysis scripts.

Leave a visible placeholder wherever information is missing. A gap you can investigate is more useful than a polished sentence you cannot verify.

How to Write a Methodology Section in 7 Steps

1. Check the Journal’s Requirements and Relevant Reporting Guidelines

Begin with the instructions for your target journal or institution. Identify the required headings, word limits, supplementary-material rules, and any reporting checklist that applies.

In health research, common examples include CONSORT 2025 for randomized trials, STROBE for cohort, case-control, and cross-sectional studies, PRISMA 2020 for systematic reviews, and SRQR or COREQ for relevant qualitative research. The EQUATOR Network organizes reporting guidelines by study type.

Outside health research, look for the reporting standards used in your own discipline rather than assuming that a biomedical checklist is automatically appropriate.

Importantly, a reporting guideline is not a substitute for a research design. PRISMA, for example, explicitly distinguishes guidance for reporting a review from guidance for conducting or evaluating its methodology.

Use the appropriate checklist to identify information you need to document—not to add procedures you never performed.

2. Introduce the Research Design and Explain Its Fit

Open the section with a direct statement of your study’s approach and purpose.

An adaptable opening is:

“This study used a [specific research design] to investigate [research question or objective] in [population, setting, or dataset].”

Then explain why that design was suitable.

Avoid stopping at broad labels such as “quantitative research” or “qualitative research.” Add the decisions that distinguish your study.

For example, the following hypothetical opening is more informative:

“This study used a cross-sectional online survey to examine the association between academic-database use and self-reported literature-search confidence among doctoral researchers at one university.”

The sentence tells readers what was measured, who was studied, and the type of relationship being investigated.

Next, state the rationale in plain language:

“The survey provided a way to collect comparable reports of search behavior and confidence across researchers within a defined recruitment period.”

A useful editing test is to ask: Could this explanation be pasted into almost any paper? If it could, make it more specific to your research question.

3. Describe the Sample or Data Sources

Explain where your evidence came from and how it entered the study.

For participant-based research, describe the source population, recruitment route, eligibility criteria, and selection process. For secondary-data research, identify the dataset and explain how you selected the relevant records. Keep the distinction between the population of interest and the sample actually available clear. STROBE emphasizes transparent reporting of participant selection and study size.

Compare these hypothetical statements:

Too vague: “Participants were selected randomly.”

More informative: “An invitation was distributed through the university’s doctoral-researcher mailing list, and eligible researchers volunteered to participate.”

The second version describes what happened without giving the recruitment process a stronger label than it deserves.

Explain the sample-size rationale as well. When you performed a power or precision calculation before collecting data, report its assumptions. When the sample was determined by a fixed dataset, available population, or recruitment window, state that instead. STROBE advises authors to explain how study size was determined rather than manufacture a retrospective justification.

For your draft, separate two questions: “Why did we aim for this sample?” and “How many observations ultimately entered each analysis?” They may require different explanations.

4. Explain the Materials and Data Collection Procedures

Describe what was measured or recorded and how the process worked.

Rather than writing “a questionnaire was used,” explain the questionnaire’s content, response formats, scoring rules, and administration. Rather than writing “samples were tested,” identify the relevant equipment, preparation steps, settings, and measurement conditions. Cite established procedures and explain modifications instead of relying on a citation to carry the entire description. PLOS One’s submission guidance emphasizes sufficient methodological detail to support reproducibility.

A hypothetical survey description might read:

“Database-use frequency was measured as the number of days participants had searched an academic database during the previous seven days, from 0 to 7. Search confidence was measured using a single item ranging from 0, ‘not at all confident,’ to 10, ‘extremely confident.’”

This identifies the variables more precisely than “research habits and confidence were assessed.”

For a chronological procedure, write the sequence first: invitation, consent, instructions, task, measurement, and follow-up. Then add the details that could change how another researcher would implement it.

Put lengthy instruments or protocols in supplementary materials when permitted, but keep the essential explanation in the main paper.

5. Describe the Analysis, Not Just the Software

“Data were analyzed using statistical software” names a tool, not an analytical method.

For quantitative research, identify the statistical procedures, variables or comparisons, treatment of missing data, and relevant assumptions. Explain how uncertainty was assessed, and distinguish planned analyses from exploratory ones. ICMJE also recommends reporting software versions and avoiding reliance on significance tests alone when effect estimates and their precision are relevant.

A hypothetical analytical statement might be:

“We estimated the association between database-use frequency and search confidence using Spearman’s rank correlation. Respondents missing either variable were excluded from this analysis. Confidence intervals were calculated using respondent-level bootstrap resampling.”

For an actual paper, add the implementation details and justification relevant to your study.

For qualitative research, name the analytical approach and explain how you moved from the source material to interpretations. For example, the Framework Method involves organizing and comparing material within and across cases using an analytical framework and matrix; simply saying “themes were identified” would omit much of that process.

As a practical check, remove the software name from your paragraph. Can the reader still understand what analysis you performed? If not, the explanation needs work.

6. Report Quality Checks, Researcher Roles, and Ethics

Replace general assurances with descriptions of actual safeguards.

Compare:

“The data were checked carefully.”

“A second researcher checked extracted values against the source documents, and discrepancies were resolved by returning to the original records.”

The second statement is useful only when that checking procedure genuinely occurred. Do not borrow it because it sounds rigorous.

For qualitative work, explain relevant researcher roles and how interpretive decisions were examined. Reflexivity means considering how researchers’ positions and assumptions influence the research process; an audit trail or reflective notes may help make those decisions visible. Choose practices that fit the analytical approach rather than treating every quality technique as compulsory.

Where applicable, report the documented ethics-review status, reviewing body, consent arrangements, and confidentiality safeguards. Requirements depend on the study and setting; use the decision and wording that actually apply, rather than assuming either approval or exemption. PLOS provides specific reporting requirements for research involving human participants.

Never invent approval details, consent procedures, validation checks, or researcher involvement to complete a template.

7. Check the Section Against What Actually Happened

Treat the final draft as an account of the completed work, not a cleaned-up version of the original plan.

Compare the text against your research records. Check recruitment dates, inclusion criteria, instrument versions, variable names, exclusions, and analysis settings. Identify procedures that changed and explain important departures from the protocol.

Transparent reporting should let readers distinguish what was planned, what was done, what was found, and what those findings mean—a distinction emphasized by STROBE.

A useful final exercise is to trace each major result backward:

Finding → analysis → measurement → data source → research question

Every link should be understandable from the manuscript and its supporting materials.

Do not repair a mismatch by describing an analysis you did not perform. Resolve it by checking the research, correcting the writing, or explicitly explaining the change.

How Methodology Writing Changes by Research Type

The steps above provide a starting structure, but some research designs need additional explanations.

Mixed-Methods Research: Explain Where the Components Connect

A mixed-methods methodology should explain more than the presence of both interviews and numerical data. Describe their sequence, their relationship, and how they were integrated. Integration may occur through sampling, data collection, analysis, or interpretation—for example, when survey findings guide the selection of interview participants.

A hypothetical explanation might read:

“Survey responses informed the selection of participants for follow-up interviews. We selected respondents with contrasting confidence scores to explore experiences that could help explain the quantitative patterns.”

Then explain how the interview findings and survey results were brought together, including how you handled disagreements between them.

Systematic Reviews: Document the Route From Search to Synthesis

For a systematic review, explain how studies were identified, selected, appraised, and synthesized. Report the information sources and search dates, full search strategies, eligibility criteria, screening and extraction procedures, risk-of-bias assessment, and synthesis methods. Describe reviewer involvement and any automation accurately. PRISMA 2020 includes detailed reporting expectations for these processes.

Do not use “the review followed PRISMA” as a replacement for describing the work.

A useful drafting question is: Could someone understand why a study entered the review and how its findings contributed to the synthesis?

Where meta-analysis was conducted, explain the effect measure, statistical model, and analytical decisions. Where it was not, describe the synthesis approach you actually used.

Computational Research: Make the Experiment Traceable

For computational and machine-learning research, document the dataset versions, preprocessing, training and evaluation splits, model configuration, parameter-selection process, baselines, and evaluation metrics. Include relevant software environments and computational resources. The NeurIPS Paper Checklist explicitly addresses these aspects of reproducibility and experimental reporting.

For your own draft, pay particular attention to decisions hidden in code. A preprocessing rule, filtering step, or evaluation setting can be central to the experiment even when it occupies only one line of a script.

Theoretical Research: Explain the Reasoning Framework

A theoretical paper may need assumptions, definitions, derivations, and proofs rather than participant-recruitment subsections. For example, NeurIPS asks authors reporting theoretical results to state their assumptions and provide complete proofs or appropriate supporting material.

Do not force an empirical-study template onto a paper whose contribution rests on formal reasoning.

Methodology Section Example

The following is a fictional quantitative example, provided to illustrate writing rather than prescribe a research design. It is not a report of a real study. Bracketed information must be replaced with verified details.

Design and Recruitment

This study used a cross-sectional online survey to examine the association between academic-database use and self-reported literature-search confidence among doctoral researchers at one university. All doctoral researchers on the institution’s research-student mailing list were invited during a four-week recruitment period. Participation was voluntary. Eligibility required current doctoral enrollment and completion of a literature search within the previous three months. Study size was determined by the available population and recruitment period rather than a prospective power calculation.

Measures and Procedure

Participants reported the number of days they had used an academic database during the previous seven days, from 0 to 7. Search confidence was assessed using a single item ranging from 0, “not at all confident,” to 10, “extremely confident.” The questionnaire also recorded discipline and year of doctoral study. Participants completed the questionnaire online after reviewing the study information and providing electronic consent.

Analysis

We summarized database-use frequency and confidence scores using medians and interquartile ranges. Their association was estimated using Spearman’s rank correlation, with a 95% percentile bootstrap confidence interval calculated from 5,000 respondent-level resamples. Respondents missing either primary variable were excluded from the correlation analysis. Available responses to other items were retained for descriptive summaries. No missing values were imputed. Analyses were conducted using [software, version, and relevant packages].

Ethics

[Insert the documented ethics-review decision, reviewing body, reference number where applicable, and relevant data-protection arrangements.]

Notice what this example does not claim. It does not describe volunteer recruitment as random sampling, a single confidence question as a validated multi-item scale, or an association as a causal effect.

The design also has limitations worth discussing: it uses one institution, voluntary participation, and self-reported measures. Clear methodology writing makes those choices visible rather than disguising them.

A Research Methodology Template You Can Adapt

Use this template to organize verified information. Delete sections that do not apply, and expand those that are central to your study.

Research design and rationale
This study used [specific design] to investigate [research question]. The approach was selected because [reason connected to the evidence needed].

Setting and data sources
The research was conducted in [setting and period]. Participants, records, materials, or studies were selected from [source] using [selection process]. Eligibility criteria were [criteria]. Study size was determined by [actual rationale].

Data collection
Data were collected using [instrument, equipment, or procedure]. The process involved [sequence]. Key variables or concepts were defined as [definitions]. Relevant materials are provided in [verified location].

Analysis
Data were analyzed using [specific approach]. The analysis involved [steps, models, comparisons, or interpretive procedures]. Missing information, exclusions, and relevant assumptions were handled through [actual procedures].

Quality and ethics
Research quality was addressed through [procedures actually used]. The documented ethics-review status was [status and authority]. Relevant consent and confidentiality arrangements were [details].

Transparency
The protocol, materials, data, or code are available through [location and access conditions]. Important changes from the original plan were [changes and reasons, where applicable].

The template is a starting point. Each bracket represents a research decision that requires evidence—not an invitation to generate a plausible answer.

Common Methodology Writing Mistakes

Using labels instead of explanations. “Mixed methods,” “thematic analysis,” and “experimental design” should lead into a description, not replace one. Explain what those choices meant in your study.

Making unsupported quality claims. Replace “the instrument was highly reliable” with the relevant evidence or assessment procedure. Do not describe pilot testing as comprehensive validation.

Writing an idealized account. A methodology section should not quietly add independent reviewers, randomization, preregistration, or other safeguards that were absent from the actual research.

Blurring procedures and findings. Explain how an outcome was measured or analyzed in the methodology; present its observed value in the results. Keep the reader’s path from procedure to finding clear.

Frequently Asked Questions

How Long Should a Methodology Section Be?

Let the journal’s requirements and the complexity of the study determine the length. New, unfamiliar, or substantially modified methods generally need more explanation than established procedures. The University of Manchester notes that methods sections tend to be more extensive when readers need additional explanation of the approach.

A useful target is not a fixed percentage of the paper. It is enough detail to understand the consequential decisions without unnecessary repetition.

Should a Methodology Section Be Written in the Past Tense?

Use the past tense for completed procedures: “We recruited participants,” “Samples were stored,” or “We fitted the model.” Manchester’s methods-writing guidance notes that the simple past is common in this section.

For a proposal, make planned work explicit: “We will recruit.” When converting a proposal into a completed paper, verify the procedures as well as changing the tense.

Do I Need Citations in the Methodology Section?

Yes, when you use established instruments, analytical methods, protocols, or procedures developed by others. Cite the source and describe the details or modifications necessary to understand your implementation. A reference should support the explanation, not replace it.

Can AI Help Write a Methodology Section?

Use AI to help organize verified notes, revise unclear explanations, or generate questions about missing information. Do not treat generated procedural details as evidence that those procedures occurred.

Human authors remain responsible for the accuracy of AI-assisted writing. ICMJE recommends reviewing AI-generated content carefully and disclosing relevant use according to its guidance and the journal’s requirements.

A useful drafting instruction is:

“Draft from the research records provided. Mark missing information explicitly. Do not invent participants, procedures, citations, approvals, or analyses.”

When AI was part of the research itself—not merely the writing—describe its role and relevant implementation details sufficiently for readers to understand the approach.

Final Thoughts

Writing a methodology section is an exercise in making research decisions understandable.

Start with the question. Explain the design. Identify the evidence. Describe how it was collected and analyzed. Make the safeguards, constraints, and changes visible.

Before submitting, ask one final question:

Can a reader follow the path from my research question to my findings without guessing what I did?

The strongest methodology section is not the one with the most technical vocabulary. It is the one that gives a clear, accurate account of the work behind the paper.