How to Describe Experimental Design in Your Methods Section

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-08-27

Quick Answer

• Describe experimental design by naming the independent and dependent variables, then state how subjects were assigned to conditions and how many replicates each condition received.
• State the randomization method, blinding procedures, and inclusion or exclusion criteria so another researcher can repeat the exact protocol.
• A complete design description does not guarantee reproducibility because unpublished details about equipment calibration, reagent lots, and environmental conditions also affect outcomes.

At a Glance

Why Experimental Design Descriptions Fail

Students and early-career researchers often write methods sections that list materials and procedures but omit the structural decisions that determine whether the data can be interpreted. A methods section that says "we treated cells with compound X and measured viability" does not tell a reader how many independent cultures were used, how the treatment concentrations were assigned, or whether the person measuring viability knew which wells received the compound. Each of those omissions weakens the reproducibility of the work.

The National Library of Medicine Research Methods Resources gateway provides access to biomedical texts that explain how study design choices affect the validity of conclusions. These resources emphasize that a study design is a plan for answering a research question while minimizing bias and error. When you describe that plan, you are not simply listing steps. You are showing the logic that connects your question to your data.

The EQUATOR Network maintains reporting guidelines for many study types, including randomized trials, observational studies, and diagnostic accuracy studies. These guidelines exist because incomplete reporting of design elements is a persistent problem across disciplines. Using a reporting guideline as a checklist when you write your methods section helps you identify which design features a reader needs to evaluate your work.

Core Principles of Design Description

Identify the Research Question and Unit of Analysis

Before you can describe your experimental design, you must know what you are studying and at what level you are making observations. The unit of analysis is the entity on which independent measurements are made. In an animal study, the unit might be an individual animal. In a cell biology experiment, the unit might be an individual culture flask or well. In a chemistry experiment, the unit might be an individual reaction vessel.

State the unit of analysis explicitly in your methods section. If you measured three technical replicates from each of four biological replicates, your unit of analysis is the biological replicate, and your sample size is four, not twelve. Describing this distinction prevents readers from overestimating your statistical power.

Define Variables Operationally

An operational definition explains exactly how you measured or manipulated a variable. For example, "cell viability" is not a complete variable description. A complete description states the assay used, the instrument settings, the wavelength read, and the calculation applied to raw data. Similarly, "temperature" is not a complete description of an independent variable in a chemistry experiment. You must state the set point, the tolerance range, how temperature was monitored, and where the probe was placed relative to the reaction.

The National Library of Medicine Research Methods Resources gateway includes texts that discuss how operational definitions reduce ambiguity in research. When you provide operational definitions, you allow another researcher to reproduce your measurement conditions instead of approximate them.

Distinguish Independent, Dependent, and Controlled Variables

The independent variable is the factor you manipulate. The dependent variable is the outcome you measure. Controlled variables are factors you hold constant to prevent them from confounding your results. In your methods section, name each category explicitly.

For a biology experiment testing the effect of a growth factor on cell proliferation, the independent variable is the growth factor concentration, the dependent variable is the proliferation rate, and controlled variables include culture medium composition, passage number, seeding density, and incubation time. For a chemistry experiment testing the effect of catalyst loading on reaction yield, the independent variable is catalyst mass, the dependent variable is product yield, and controlled variables include solvent volume, temperature, stirring rate, and reaction time.

Practical Workflow for Writing the Design Description

Step 1: Choose a Reporting Guideline

The EQUATOR Network provides a searchable library of reporting guidelines organized by study type. For a randomized controlled trial, use the CONSORT statement. For observational studies, use STROBE. For animal research, use ARRIVE. For in vitro studies, consult the relevant discipline-specific guidance.

Selecting a guideline before you write gives you a checklist of items that must appear in your methods section. Work through the checklist systematically and address each item, even if the answer is "not applicable." If an item does not apply, state that explicitly instead of omitting it silently.

Step 2: Describe Subject or Sample Selection

State where your subjects or samples came from, how they were selected, and what criteria were used to include or exclude them. For animal studies, include species, strain, sex, age, weight range, and housing conditions. For cell studies, include cell line, passage number, authentication status, and mycoplasma testing. For chemical studies, include reagent source, purity grade, and lot numbers.

Describe the sample size and how it was determined. If you performed a power analysis, state the expected effect size, alpha level, and power that informed the calculation. If you did not perform a power analysis, say so and explain the practical constraints that determined your sample size.

Step 3: Describe Randomization

Randomization assigns subjects to treatment groups in a way that reduces selection bias. Describe the specific method you used. Common approaches include simple randomization with a random number generator, block randomization to balance group sizes, and stratified randomization to balance important covariates across groups.

State who performed the randomization and when it occurred relative to treatment administration. If you used a random number table, a computer algorithm, or a web-based service, name the tool. If you used block sizes, state the block size and whether it was fixed or varied.

Step 4: Describe Blinding

Blinding prevents knowledge of group assignment from influencing behavior or measurement. Describe who was blinded and when. In a single-blind study, participants or subjects do not know their group assignment. In a double-blind study, both the participants and the investigators who interact with them are blinded. In some studies, only the outcome assessors or data analysts are blinded.

State how blinding was maintained. For example, you might describe using coded labels that concealed treatment identity until data analysis was complete. If blinding was not possible, state that limitation and explain why.

Step 5: Describe Replication

Replication means running the experiment on multiple independent units. Distinguish between technical replicates, which are repeated measurements on the same sample, and biological replicates, which are independent samples that capture natural variation.

State the number of replicates per condition and whether those replicates were independent. For example, "each treatment condition was tested in six independent cultures" is a complete replication statement. "Each sample was measured in triplicate" describes technical replication and does not tell the reader how many independent samples existed.

Step 6: Describe the Statistical Analysis

Name the statistical tests you used and link each test to the specific comparison it addressed. State the software package and version. State the significance threshold you used and whether you adjusted for multiple comparisons.

Describe how you handled missing data, outliers, and violations of statistical assumptions. If you excluded any data points, state the criteria for exclusion and how many points were excluded.

Options and Tradeoffs in Design Choices

Complete Randomization versus Blocking

Complete randomization assigns every subject to a group independently. This approach is simple and valid when your subjects are relatively homogeneous. Blocking groups subjects into blocks based on a characteristic that might affect the outcome, such as litter, batch, or baseline weight, and then randomizes within each block. Blocking reduces the impact of known sources of variation but requires you to identify and measure the blocking variable before the experiment begins.

Technical versus Biological Replication

Technical replication measures the precision of your assay. Biological replication measures the variability in the system you are studying. Increasing technical replication improves the precision of individual measurements but does not increase the generalizability of your conclusions. Increasing biological replication improves your ability to detect true effects but costs more time and resources. Your methods section should state which type of replication you used and why that choice was appropriate for your question.

Blinding Feasibility

Blinding is not always possible. In some chemistry experiments, the color or physical properties of a reaction mixture reveal the treatment condition. In some animal studies, the route of administration differs between treatment groups. When blinding is impossible, describe the limitation and state what steps you took to reduce bias, such as having a separate investigator perform outcome measurements.

Observations and Measurements to Record

Pre-Experimental Records

Record the date of the experiment, the identity of the person who performed it, and the version of the protocol used. Record the source and lot numbers of all reagents, the calibration status of all instruments, and the environmental conditions in the laboratory.

During-Experiment Records

Record the actual conditions that occurred during the experiment, beyond the intended conditions. If an incubator temperature drifted, record the drift. If a reagent was prepared on a different day than planned, record that change. These records become part of your methods description because they document what actually happened.

Post-Experiment Records

Record the raw data, the analysis scripts or calculation spreadsheets, and the version of the analysis that produced the final results. The NIH Data Management and Sharing Policy describes expectations for managing and sharing research data. Even if your project is not NIH-funded, following similar practices for data organization and documentation strengthens the reproducibility of your work.

Common Failure Patterns in Design Descriptions

Failure to Distinguish Replicate Types

A common error is writing "experiments were performed in triplicate" without clarifying whether the three runs were independent experiments or three measurements from one experiment. This ambiguity prevents readers from assessing the statistical basis of your conclusions.

Omitting Randomization Details

Writing "subjects were randomly assigned" without describing the randomization method leaves the reader unable to judge whether the assignment was truly random and whether the method was appropriate for the study design.

Describing Blinding Vague

Writing "the study was blinded" without stating who was blinded and when creates confusion. A study can be blinded for participants but not for outcome assessors, and that distinction matters for interpreting the results.

Reporting Only Final Sample Sizes

If animals or samples were excluded from analysis, state the number excluded and the reasons. Reporting only the final sample size hides attrition and makes it impossible for readers to assess whether exclusion criteria introduced bias.

Mixing Methods and Results

Some students describe outcomes in the methods section, such as "the treatment increased cell viability." Outcomes belong in the results section. The methods section should describe what you did, not what you found.

Quality Controls and Reproducibility Measures

Use Reporting Guidelines as Checklists

The EQUATOR Network reporting guidelines function as quality checklists for your methods section. Before you submit a manuscript or report, compare your methods section against the relevant guideline and confirm that each required item is present.

Register Your Study When Appropriate

For clinical trials and some other study types, prospective registration is expected. The Committee on Publication Ethics Core Practices describes ethical expectations for research conduct, including transparency about study methods. Registration creates a public record of your design before data collection begins, which helps prevent selective reporting.

Maintain a Laboratory Notebook

A dated laboratory notebook that records design decisions, protocol changes, and raw observations provides the source material for your methods section. When you write the methods section, you are translating the notebook record into a format that other researchers can follow.

Limitations of Design Descriptions

A methods section cannot capture every detail of an experiment. Even a thorough design description leaves out tacit knowledge about how the experiment was actually performed. Two researchers following the same written protocol may produce different results because of differences in technique, equipment, or environmental conditions.

The NIH Grants and Funding resources describe how funding applications are evaluated, including the importance of rigorous experimental design. Reviewers assess whether the proposed methods are adequate to answer the research question. A clear design description helps reviewers understand the scientific rationale and feasibility of the work.

Another limitation is that design descriptions are written after the fact. Memory is imperfect, and records may be incomplete. Writing the design description immediately after data collection, while the details are fresh, produces a more accurate account than writing it weeks later.

Safety and Regulatory Context

Animal Research

If your experiment involves animals, your methods section must describe the ethical approval obtained and the institutional animal care and use committee that reviewed the protocol. State the housing conditions, the number of animals used, and the steps taken to minimize pain and distress. The ARRIVE guidelines, available through the EQUATOR Network, provide a checklist for reporting animal research.

Human Subjects Research

If your experiment involves human subjects, your methods section must describe the institutional review board approval, the informed consent process, and the measures taken to protect participant confidentiality. The Committee on Publication Ethics Core Practices describes expectations for ethical conduct in research involving human participants.

Chemical Safety

For chemistry experiments, your methods section should describe the safety precautions taken, including personal protective equipment, ventilation, and waste disposal procedures. If you used hazardous chemicals, state the quantities and the containment measures.

Professional Escalation Criteria

When to Consult a Biostatistician

Consult a biostatistician before you finalize your experimental design if you are uncertain about sample size calculations, randomization schemes, or the appropriate statistical test for your data. A statistician can also help you describe your design accurately in the methods section.

When to Consult Your Institution's Research Office

Consult your institutional research office if you are uncertain about regulatory requirements, such as animal care approvals, human subjects approvals, or biosafety permits. These requirements vary by institution and jurisdiction, and your methods section must accurately reflect the approvals you obtained.

When to Consult a Reporting Guideline Expert

If you are submitting to a journal that requires a specific reporting guideline, consult the guideline before you write. The EQUATOR Network provides access to the guidelines and related resources. If you are uncertain which guideline applies to your study type, contact the journal or the EQUATOR Network for guidance.

Templates for Common Experiment Types

Template for a Cell Biology Experiment

"Human lung epithelial cells (A549, passage 8) were cultured in DMEM supplemented with 10% fetal bovine serum at 37 degrees Celsius and 5% carbon dioxide. Cells were seeded at 5,000 cells per well in 96-well plates and allowed to attach for 24 hours. The independent variable was compound X concentration at 0, 1, 10, and 100 micromolar. Each concentration was tested in six independent wells, and the entire plate was replicated three times on separate days. Wells were assigned to treatment conditions using a random number generator. The person measuring cell viability was blinded to treatment assignment. Viability was measured using the MTT assay at 570 nanometers. Data were analyzed by one-way ANOVA followed by Dunnett's test for comparison to the control group."

Template for a Chemistry Experiment

"The independent variable was catalyst loading at 0.5, 1.0, and 2.0 mole percent. The dependent variable was the yield of product after 2 hours of reaction time. Each catalyst loading was tested in three independent reaction vessels. Reaction vessels were assigned to catalyst loadings using a random number table. The reaction temperature was maintained at 60 degrees Celsius with a tolerance of plus or minus 1 degree, monitored by a thermocouple placed in the reaction mixture. Product yield was determined by gas chromatography with an internal standard. Data were analyzed by linear regression of yield against catalyst loading."

Template for an Animal Study

"Male Sprague Dawley rats aged 8 weeks were housed in groups of three under a 12-hour light-dark cycle with free access to food and water. After a 7-day acclimation period, rats were randomly assigned to receive either vehicle or compound Y at 10 milligrams per kilogram by oral gavage. Randomization was performed using a computer-generated random number sequence with block sizes of four. The investigator administering the treatment and the investigator measuring outcomes were blinded to group assignment. The primary outcome was body weight change over 14 days. Sample size was determined by a power analysis assuming a 20 percent difference between groups with 80 percent power and an alpha of 0.05."

Records and Measurements for Design Verification

What to Keep in Your Records

Keep the original protocol document, the randomization sequence, the blinding codes, the raw data files, and the analysis scripts. Keep records of any protocol deviations and the dates they occurred. The NIH Data Management and Sharing Policy describes expectations for data management plans, including how data will be preserved and shared.

How to Verify Your Design Description

After writing your methods section, compare it against your laboratory notebook and your raw data. Confirm that the number of replicates stated in the methods matches the number of data points in your raw data. Confirm that the randomization method described matches the records of group assignment. Confirm that the blinding description matches who actually knew the group assignments during the experiment.

Frequently Asked Questions

How do I describe randomization in a methods section?

State the specific method you used, such as a random number generator, a random number table, or block randomization. Name the tool or software used to generate the random sequence and state who performed the randomization. If you used block randomization, state the block size.

What is the difference between technical and biological replicates?

Technical replicates are repeated measurements of the same sample and assess assay precision. Biological replicates are independent samples that capture natural variation in the system. Your methods section should state which type you used and how many of each you performed.

How do I describe blinding when it is not possible?

State that blinding was not possible and explain why. Describe the steps you took to reduce bias, such as having a separate investigator measure outcomes or using coded samples that concealed treatment identity during measurement.

What should I include in the sample size description?

State the number of subjects or samples per group, the total sample size, and how the sample size was determined. If you performed a power analysis, state the effect size, alpha level, and power used. If you did not, state the practical constraints that determined the sample size.

How do I describe the statistical analysis in the methods section?

Name the statistical tests you used and link each test to the specific comparison it addressed. State the software package and version, the significance threshold, and any adjustments for multiple comparisons. Describe how missing data and outliers were handled.

Should I include exclusion criteria in the methods section?

Yes. State the criteria you used to include or exclude subjects or samples and the number excluded for each reason. Reporting only the final sample size without explaining attrition prevents readers from assessing whether exclusion introduced bias.

How do I know which reporting guideline applies to my study?

The EQUATOR Network provides a searchable library of reporting guidelines organized by study type. Select the guideline that matches your study design, such as CONSORT for randomized trials, STROBE for observational studies, or ARRIVE for animal research.

What should I do if my methods section is too long?

Focus on the design elements that affect interpretation of the results. Omit routine procedural details that are standard for your field and can be cited to a published protocol. Keep the randomization method, blinding procedures, replication counts, and statistical analysis in the main text.

Using the Evidence

Related Articles

• Replication Study Design for Bioequivalence: A Guide
• Pseudoreplication in Experiments: What It Is and How to Avoid It
• Replication Study Design: A Guide to Choosing the Right Approach
• Replication in Experimental Design: Why It Matters and How to Do It Right
• Mixed Methods Study Design: Integrating Quantitative and Qualitative Approaches

References and Further Reading

• Research Methods Resources. National Library of Medicine.
• EQUATOR Network. EQUATOR Network.
• Core Practices. Committee on Publication Ethics.
• NIH Grants and Funding. National Institutes of Health.
• ORCID for Researchers. ORCID.
• Data Management and Sharing Policy. National Institutes of Health.
• NCBI Literature Resources. National Center for Biotechnology Information.
• PubMed. National Library of Medicine.

This article is educational and does not replace institutional policy, professional advice, or applicable safety and regulatory requirements.