Methods Section of a Research Paper, With Methodology Examples

The methods section of a research paper is the account of how the study was done. A useful section names the people or materials, the steps, and the analysis, with enough detail for a reader in the field to judge the work. A methodology example that only says a survey was conducted does not do that work.

The methodology examples in this guide were written for the guide. None of them is a methods section from a published paper. Copy the shape, then use the design, the people, and the steps from your own draft. Where a real methods section would include a count, a dose, or a measurement, these models leave the number out. There is no study behind them.

When a sentence already says the right procedure but sounds stiff, Rephrase AI can rephrase that sentence. Rephrase AI is designed to preserve the meaning. Compare the new sentence with the procedure in your notes before you keep the sentence. Your instructor's rules on AI use still apply.

1. What the methods section of a research paper is for

The George Mason University Writing Center describes the method section in an IMRaD report as a step-by-step account of how the empirical research was conducted, so the research is transparent and replicable. IMRaD means introduction, method, results, and discussion. The section should give another scholar enough relevant information to understand the process, compare the findings with similar studies, and replicate the research if replication is needed. Writers typically introduce the data and explain how the data were collected and analyzed, and they justify those choices. George Mason's handout was last updated on May 13, 2019.

PLOS describes the same job as the information a reader needs in order to judge a study's validity. The section should explain how the experiments were designed, how the results were analyzed, and which methods answered the research question. The reader in view is someone trained in the field. A lay reader does not need every basic term defined. PLOS also treats the section as the place to show constraints and choices that were not ideal. Hiding those choices makes the study harder to trust.

APA Style Journal Article Reporting Standards tell journal authors, reviewers, and editors what a manuscript should report. APA also says educators and students can use the standards as a teaching tool. Use JARS–Quant when the data are numerical or the findings are reported through statistical analyses. Use JARS–Qual when the data are natural language and expression. Use JARS–Mixed when the study uses both. APA says the quantitative standards are meant to increase reproducibility, and the qualitative standards are meant to increase methodological integrity. The JARS page was last updated in October 2026.

The methods section is not the introduction, and the methods section is not the results. The UW–Madison Writer's Handbook says the introduction can name the general design and can justify that design when alternatives exist. The methods section then answers three questions. How did you study the problem? What did you use? How did you proceed? The same handbook says to keep findings out of the procedure. For the one or two sentences that name the method inside a summary, see abstract examples. For the opening that can name the design before the full account, see introduction paragraph examples.

A humanities essay often has no methods section. This guide is for empirical papers, lab reports, surveys, interview studies, and reviews that have a procedure to report.

2. How long a methods section should be

Use the limit in the assignment or the journal. The guides checked for this article do not share one word count.

Nature's formatting guide adds two space rules that change how the section reads. A method that is already published should not be copied at length. A reference can stand in for that published method, and any new addition or variation should be stated. The Methods section also cannot contain figures or tables. Essential display items go in Extended Data or, exceptionally, in Supplementary Information.

PLOS ONE draws the citation line differently. Authors may cite articles where established materials, methods, and protocols are described in detail. The submission should still include enough information to be understood without those references. New methods need the specific protocol in the paper. PLOS also encourages a public protocol, for example on protocols.io, linked from the methods section with a DOI.

The PLOS author guide on writing a methods section suggests a practical way to choose the level of detail. Imagine writing for your future self, at a different institution, trying to redo the same experiment without the old logins, the old notebooks, or the old reagents. Name the cell line, antibody, software, or reagent a reader would need, including a Research Resource ID when one exists. Name the questionnaire version, the stimulus, who was excluded, and any change made to the protocol during the work. If a word limit cuts the section short, PLOS says the extra searchable detail can go in supplemental materials. Being more thorough than the minimum is acceptable.

A class lab report is often shorter than a Nature Article methods section. A qualitative method section is often longer than a quantitative one, for the reason in the APA table. Neither pattern replaces the limit on the assignment in front of you.

3. How to structure the methods section

Start with the three UW–Madison questions, then add the items your study type requires.

Order the steps by time, or group them by type of procedure and keep time order inside each group. Use past tense, because the work is finished. When the study measured something, put the amount in the section. Concentrations, volumes, times, and temperatures belong there when they shaped the result. Leave the routine arithmetic of a common test out. Keep the finding itself for the results.

George Mason says method sections commonly use past tense for that same reason. Passive voice is often appropriate when the procedure matters more than the person who ran the procedure. "The seedlings were weighed at the same hour" keeps the scale in view. "We weighed the seedlings at the same hour" is also clear. Use the voice your field expects.

George Mason lists three moves that commonly appear. The moves are not a required sequence, and a writer can move back and forth among them. Qualitative interview papers often spend longer on participants and coding. A chemistry paper may spend longer on equipment.

A quantitative study

The JARS–Quant report in American Psychologist in 2018 sets the quantitative standards. APA Style's JARS pages still describe those 2018 standards as the current quantitative standards. Table 1 lists information recommended for manuscripts that report new data collections. The note under the table says the author selects the items that apply to the particular study. For the method, the table asks for the items below when they apply.

• Inclusion and exclusion criteria, including any restriction based on demographic characteristics.
• Major demographic characteristics, plus topic-specific characteristics that matter for this study. Animal research includes genus, species, strain or stock, the supplier, and details such as sex, age, weight, and physiological condition.
• How participants were selected, what share of the people approached actually participated, and whether people or units such as schools selected themselves into the study.
• Settings, locations, and dates of data collection.
• Agreements and payments made to participants.
• Institutional review board agreements, ethical standards met, and safety monitoring.
• Intended sample size, achieved sample size when the achieved size differs, and how the size was set, including a power analysis or another precision method when one was used.
• Primary and secondary measures and covariates, including measures that were collected but left out of the report.
• How the data were collected, and how the quality of the measurements was protected.
• Whether conditions were manipulated or only observed, and the name of the design.
• Planned checks on the data, including rules for excluding participants after collection, missing data, outliers, and transformations.
• The analytic strategy for primary hypotheses, secondary hypotheses, and exploratory hypotheses.

The 2018 report puts data checks and the analytic strategy in the method for a specific reason. Rules for changing a data set should be set before the analysis. If a change happens after analysis has started, the report should say so and give the reason. The report also says to keep the unmodified data available for verification, and to describe results from both the modified data and the unmodified data. One of those two descriptions can sit in supplemental materials.

A qualitative interview study

JARS–Qual Table 1 asks the method to summarize the research design, including the data-collection strategies and the data-analytic strategies. Name the approach to inquiry when that approach makes the design clearer. Give the reason this design was selected.

The table says a method section can be chronological or narrative. Another investigator should be able to follow the description. The standard does not require that investigator to reach the same conclusions. The description should support conclusions with a similar degree of methodological integrity. Qualitative work is often iterative rather than linear, so the collection section and the analysis section may be combined.

The same table asks for the researchers' prior understanding of the topic and how that understanding was managed, the number of participants or documents, how people were recruited, any pay or incentive, and the consent process. There is no agreed minimum number of participants. The paper should say why the number fits the design, and why collection stopped. Saturation is one stopping reason the table names. The table also asks for the form of the data, how interviews were recorded and transcribed, whether codes were set before the analysis or emerged during the analysis, the unit that was coded, and the software if software was used.

Materials and methods in a lab report

Many journals use the heading Materials and Methods for this same job. PLOS ONE says the middle of a research article can rename Materials and Methods, Results, and Discussion, and can put those parts in any order. The heading is a label. The job is still the procedure.

Nature asks for short bold headings that name the methods used, and encourages separate subsections for statistics, reagents, and animal models. A published method is cited, and the variation you actually used is stated.

PLOS ONE asks for a statistical analysis part that names the software and the version, the procedures required to reproduce the analysis, and a repository identifier for any code. The same guidelines ask authors to define the significance threshold, to say how sample size was determined when a calculation was used, and to say how outliers and missing data were handled. A t-test should be identified as one-tailed or two-tailed, and as the kind of t-test that was run. Those details belong in the section when the study used them. A class lab that only compared two groups still needs the comparison named, even when the journal list is longer than the lab.

A systematic review

A review has a methods section too. The PRISMA 2020 statement is a reporting guideline written first for systematic reviews of health interventions. The statement says the checklist items also apply to reviews of other interventions, such as social or educational ones, and that many items apply to reviews with other aims, such as prevalence. The explanation and elaboration states the methods items in plain checklist language.

• Specify the inclusion and exclusion criteria, and how studies were grouped for the synthesis.
• Specify every database, register, website, organization, reference list, and other source, and the date each source was last searched or consulted.
• Present the full search strategy for every database, register, and website, including filters and limits. One sample strategy is not enough.
• Specify how a record was judged against the criteria, how many reviewers screened each record and each retrieved report, whether those reviewers worked independently, and which automation tools were used if any were used.
• Specify how data were collected from the reports, including how many reviewers collected data and whether they worked independently.
• Describe how the synthesis was planned, including which studies were eligible for each synthesis and which model was used if a meta-analysis was performed.

The statement says a journal may limit words and sections. Detailed methods can then sit in a supplement, if the main report points to that file. The suggested heading is not a rule about placement. The information has to be reported. Complete reporting lets a reader judge whether the methods were appropriate, and therefore how far to trust the findings.

Where a limit of the method belongs

A limit of the procedure belongs next to the procedure. PLOS treats an unstated constraint as a threat to credibility. Say that the questionnaire was given in one course, that the plants grew in one room, or that authors of missing papers were not contacted.

A limit on how far the findings travel often belongs later. JARS–Qual places the study's strengths, limitations, and limits of transferability in the discussion. The conclusion examples guide shows how to state that kind of limit in the ending without an apology and without new evidence. The methods section does not have to repeat the ending. The methods section does have to show the constraint that produced the limit.

If a published protocol is cited, the reference list still needs a normal entry. See references example for the shape of an APA 7 entry.

4. Methodology examples

Each passage below was written for this guide. None of them reports a real lab, a real survey, a real set of interviews, or a real review. A published methods section would use the counts and settings from the study. These models do not invent those numbers.

Phrases for participants

These lines were written for this guide. Replace the people with the people in your study.

• Students enrolled in the course were invited by a message on the course site, and participation was voluntary.
• People were eligible if they had taken the bus at least once in the previous month.
• Interviews were limited to staff who answer customer messages as part of the job.
• The seedlings came from one seed packet and were grown in the same room.

Phrases for the procedure

• Each person completed the questionnaire once, on paper, during the class meeting.
• The consent page came first. The questions began only after that page was signed.
• Seedlings on the longer light cycle and seedlings on the shorter cycle were watered on the same days.
• Two reviewers screened titles independently, then screened the full text of each record that remained.

Phrases for materials

• The questionnaire was a written list of closed questions about the weekly grocery trip.
• Both groups used the same lamp model, with a timer set to a different daily cycle.
• Interviews were recorded and transcribed in full before any coding began.
• The search covered two databases, and the exact search lines are listed with the date of the last search.

Phrases for data analysis

• Answers were compared across the two course sections. The comparison was planned before the forms were opened.
• Transcripts were read in full, then coded for repeated actions. The codes were not written before that first reading.
• Height was the measurement. The analysis compared the two light cycles and did not add a third comparison after the plants were measured.
• Records that met the eligibility rules were grouped by study design before any summary was written.

Phrases for ethics

• The campus review board approved the study before anyone was recruited.
• Each person read a consent page that named the topic, the time, and the right to stop.
• No name was written on the questionnaire answers.
• Recordings were stored where only the research team could open them, and the recordings were deleted after the transcripts were checked.

Phrases for limitations of the method

• The questionnaire was given in one course, so the answers describe that course.
• The light comparison used one room, and the room was not a greenhouse.
• Interviews stopped when new conversations repeated actions already coded. That stop is a judgment, not a census of every employee.
• The review searched two databases and did not contact authors for papers the search missed.

If a line is accurate but stiff, rephrase these phrases in your own words with the public paraphrasing tool. Keep the people, the steps, the materials, and any numbers from your notes. The tool changes wording. The tool does not supply a method.

Lab experiment

Basil seedlings were grown under two daily light cycles in the same room. Both groups used the same soil, the same pots, and the same watering days. One group received a longer light cycle. The other group received a shorter cycle. Height was measured at the end of the growth period, at the same hour for every pot. The comparison of the two cycles was planned before the seeds were planted. No third treatment was added after the plants were measured.

Why this example works. The materials are named. The steps are in past tense and in time order. The planned measurement is in the section, and the finding is not. A real lab report would add the hours of light, the temperature, and the heights from the data sheet. This model leaves those values out.

Survey study

Students in two course sections were invited by a message on the course site. Participation was voluntary, and no grade depended on participation. Students enrolled in the section that week were eligible. Each person completed one written questionnaire about how the section treats non-urgent messages, on paper, during the class meeting. A consent line came before the questions. Names were not written on the answer pages. The planned comparison was between the section with a written reply rule and the section without one. Answers were compared only on that planned contrast.

Why this example works. The passage covers eligibility, a voluntary sample, the setting, the material, consent, and the analysis that was planned before the forms were opened. A real paper would add how many students were invited, how many finished the form, the dates, and the name of the review board. This model has no sample, so the model does not invent one.

Qualitative interview study

Staff who answer customer messages were invited by email. Each person who agreed sat for one interview about what that person does when a message looks urgent and the work is not. Interviews were recorded and transcribed in full before any coding. Transcripts were read once without codes, then coded for repeated actions. The codes were not written before that first reading. Collection stopped when new interviews repeated actions already coded. That stop was a judgment about repetition, and the judgment is stated here so a reader can see the judgment. The interviewer had worked in a support role before the study. That history was written down before the interviews and was checked against the codes.

Why this example works. The passage names recruitment, the form of the data, recording and transcription, codes that emerged rather than codes set in advance, a reason collection stopped, and how the interviewer's earlier job was handled. A real paper would add the number of people, the length of the interviews, the site, and the consent process approved for that project.

Systematic review

Studies of daytime bus frequency and grocery trips were eligible. Papers that measured only commuting by car were excluded. Two databases were searched, and the date of the last search is stated with each database name. The full search lines for both databases are in the appendix. Two reviewers screened titles independently, then screened the full text of each record that remained. Disagreements were resolved by discussion. Data were taken from each included paper with the same form, covering the design, the setting, and the outcome the paper measured. Papers were grouped by design before any summary was written. Authors were not contacted for missing data, and that limit is stated here.

Why this example works. Eligibility, sources, search dates, the full strategy, independent screening, data collection, and the grouping plan are all named. A real review would name the databases and print the search lines. This model has no appendix and no database names, so a reader cannot mistake the model for a finished review.

A weak methods paragraph and a revision

Weak version

A survey was conducted to explore communication. Various methods were used. The results were interesting and are discussed later.

Revision

Students in two course sections completed one written questionnaire about how the section treats non-urgent messages. One section used a written reply rule. The other section did not. The comparison of those two sections was planned before the forms were opened.

Why the revision works. The weak version names a topic, hides the procedure, and points ahead to the results. The revision names the people, the material, the contrast, and the plan.

A last check before you submit

• The section reports the procedure. A finding stays in the results.
• A reader in the field could follow the steps without your lab notebook or your interview files.
• The analysis was named before the data were opened, or any later change is labeled as a later change and given a reason.
• Consent and review-board approval are stated when people were involved, using the names and the consent form your board actually used.
• A cited protocol includes the variation you used. The paper can still be understood if the reader does not open that citation, when the journal requires that.
• A rephrased sentence still matches the notes. People, steps, materials, and numbers survive the wording change.

FAQ

What goes in the methods section of a research paper?

How long should a methods section be?

What is a methodology example?

What is the difference between materials and methods and a methods section?

Should results appear in the methods section?

Does a methods section need an ethics statement?

Can a methods section use passive voice?