How to Report Statistical Analysis in the Methods Section

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

Quick Answer

• Report the statistical software, version, and specific tests used so readers can verify your analysis path.
• Describe each test alongside the variable types and research question it addresses, using standard phrasing like "differences were assessed using."
• State your significance threshold and how you handled missing data, but do not report results or p-values in the methods section.

At a Glance

Why the Methods Section Must Describe Statistical Analysis

The methods section of a lab report or research paper serves one primary function: it tells the reader exactly what you did so another researcher could repeat the work. Statistical analysis is a core part of that description because the choice of test determines whether your conclusions are valid. A reader cannot judge the strength of your findings without knowing how you processed the data.

Statistical reporting in the methods section is not about showing results. It is about showing the analytical path. When you describe the tests you used, the software that ran them, and the assumptions you checked, you give the reader the tools to evaluate whether your approach was appropriate. This transparency is a central expectation in scientific publishing. The EQUATOR Network maintains reporting guidelines that help authors structure their methods so that statistical choices are clear and complete.

For students writing lab reports, the methods section is often the most difficult part to write because it requires precision without interpretation. You must state what you did, not why the results matter. The statistical analysis subsection is a formal record of your analytical decisions. It should be written in past tense, passive voice, and without reference to specific outcomes.

Core Principles of Statistical Reporting

Match the Test to the Research Question

Every statistical test answers a specific type of question. Before you can write about a test, you need to know what question your data can answer. The main distinctions are:

• Comparing groups: t tests, ANOVA, Mann-Whitney U, Kruskal-Wallis
• Examining relationships: Pearson correlation, Spearman correlation
• Predicting outcomes: linear regression, logistic regression
• Testing distributions: chi-square test, Fisher exact test
• Analyzing survival: Kaplan-Meier, Cox regression

Your methods section should state the research question that motivated each test. For example, if you compared growth rates between two treatment groups, you would report that differences in mean growth were assessed using an independent samples t test. This phrasing connects the test to the question it answers.

Describe Variable Types

Statistical tests are chosen based on whether your variables are continuous, categorical, ordinal, or count data. Your methods section should identify the variable types you collected. This information helps the reader understand why you selected particular tests.

For continuous variables like weight, height, or temperature, you typically report means and standard deviations. For categorical variables like sex or treatment group, you report frequencies and percentages. For ordinal variables like pain scores or disease severity ratings, you report medians and interquartile ranges. State these choices explicitly in your methods.

State the Software and Version

Statistical software packages produce different outputs and sometimes use different algorithms for the same test. Reporting the software name and version is standard practice. Common examples include:

• R version 4.3.1
• SPSS version 28
• SAS version 9.4
• GraphPad Prism version 10
• Stata version 18

If you used a specific package or library within R, such as the tidyverse or lme4, mention that as well. This level of detail allows another researcher to reproduce your exact analytical environment.

Report the Significance Threshold

The conventional threshold for statistical significance is P < 0.05, but you should state this explicitly instead of assuming the reader knows it. If you used a different threshold, such as P < 0.01 for multiple comparisons, explain why. Also state whether your tests were one tailed or two tailed. Two-tailed tests are more conservative and are the default in most research contexts.

Describe Missing Data Handling

Missing data is a reality in most research projects. Your methods section should state how you handled incomplete observations. Common approaches include:

• Excluding cases pairwise (using only complete data for each test)
• Excluding cases listwise (removing any case with any missing value)
• Imputing missing values using mean substitution or multiple imputation

Each approach has tradeoffs, and the reader needs to know which one you used to evaluate the robustness of your findings.

How to Structure the Statistical Analysis Subsection

Placement Within the Methods Section

The statistical analysis subsection typically appears at the end of the methods section, after you have described the study design, participants, and measurements. This placement allows the reader to understand the data collection process before learning how the data were analyzed.

Standard Paragraph Structure

A well-written statistical analysis subsection follows a logical order:

1. State the software and version used
2. Describe how continuous and categorical variables were summarized
3. List the specific tests used for each comparison or relationship
4. State the significance threshold
5. Describe how missing data were handled
6. Mention any adjustments for multiple comparisons

Here is an example of a complete statistical analysis paragraph:

"Statistical analyses were performed using R version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria). Continuous variables were summarized as mean (SD) if normally distributed and median (IQR) if not. Categorical variables were summarized as frequencies and percentages. Normality was assessed using the Shapiro-Wilk test. Differences in normally distributed continuous variables between two groups were evaluated using independent samples t tests. Differences in non-normally distributed variables were evaluated using Mann-Whitney U tests. Associations between categorical variables were assessed using chi-square tests or Fisher exact tests when expected cell counts were low. Statistical significance was set at P < 0.05, and all tests were two tailed. Missing data were excluded pairwise from each analysis."

Use Past Tense and Passive Voice

The methods section describes actions you already completed. Write in past tense: "were performed," "was assessed," "were summarized." Use passive voice to keep the focus on the actions instead of the researcher: "Data were analyzed" instead of "I analyzed the data."

Common Statistical Tests and How to Describe Them

Comparing Two Groups

When you compare two independent groups on a continuous outcome, you typically use a t test or a Mann-Whitney U test. The t test assumes the data are normally distributed and have equal variances. The Mann-Whitney U test is the nonparametric alternative and does not require normality.

Example phrasing: "Differences in body weight between the treatment and control groups were assessed using an independent samples t test."

If you checked the normality assumption and found the data were not normal, you would write: "Differences in body weight between groups were assessed using the Mann-Whitney U test."

Comparing Three or More Groups

When you compare three or more groups, you use one-way ANOVA for normally distributed data or the Kruskal-Wallis test for non-normal data. If ANOVA shows a significant difference, you typically follow up with post hoc tests to identify which groups differ.

Example phrasing: "Differences in growth rate among the four dietary treatment groups were evaluated using one-way ANOVA followed by Tukey post hoc comparisons."

Examining Relationships Between Variables

When you examine whether two continuous variables are related, you use Pearson correlation for normally distributed data or Spearman correlation for non-normal data. The correlation coefficient ranges from -1 to 1, but you do not report the value in the methods section.

Example phrasing: "The relationship between temperature and germination rate was assessed using Pearson correlation."

Predicting Outcomes

When you predict a continuous outcome from one or more predictor variables, you use linear regression. When you predict a binary outcome, you use logistic regression. Your methods section should state which variables were entered into the model and how you selected them.

Example phrasing: "Multiple linear regression was used to examine the association between soil pH, moisture content, and plant height. Predictors were entered simultaneously into the model."

Analyzing Categorical Data

When you examine whether two categorical variables are independent, you use a chi-square test. If any expected cell count is below 5, you use the Fisher exact test instead.

Example phrasing: "The association between treatment group and survival status was assessed using the chi-square test."

Practical Steps for Writing Your Statistical Analysis Section

Step 1: List Every Test You Performed

Before writing, make a complete list of every statistical test you ran during your analysis. This includes descriptive statistics, assumption checks, and inferential tests. Review your analysis output or code to ensure you do not miss anything.

Step 2: Match Each Test to a Research Question

For each test on your list, write down the research question it answers. This step ensures that every test you report has a clear purpose. If a test does not answer a research question, consider whether it belongs in the methods section at all.

Step 3: Write the Software Statement First

Start your statistical analysis subsection with the software statement. This establishes the analytical environment before you describe specific tests.

Step 4: Describe Descriptive Statistics

State how you summarized your variables. This is typically one sentence covering continuous and categorical variables.

Step 5: Describe Inferential Tests in Order

List your inferential tests in the order they appear in your results section. This makes it easy for the reader to follow along when they reach the results.

Step 6: State the Significance Threshold and Missing Data Handling

End with the significance threshold and how you handled missing data. These details apply to all tests, so they belong at the end of the subsection.

Step 7: Review Against Reporting Guidelines

Check your methods section against relevant reporting guidelines from the EQUATOR Network. These guidelines provide checklists for different study types and can help you identify missing information.

Common Failure Patterns in Statistical Reporting

Reporting Results in the Methods Section

The most common mistake students make is including p-values or effect sizes in the methods section. These belong in the results section. The methods section should only describe what tests were used, not what they found.

Omitting the Software Version

Writing "analyses were performed using SPSS" without the version number is incomplete. Software updates can change algorithms and output formats, so the version matters for reproducibility.

Failing to State the Significance Threshold

Some students assume P < 0.05 is universal and do not state it. While this threshold is common, it is not universal, and the reader should not have to guess.

Not Describing Missing Data Handling

If you excluded any data points, you must say so. Omitting this information makes it impossible for the reader to assess whether your results are biased by missing data.

Using Vague Language

Phrases like "appropriate statistical tests were used" or "data were analyzed statistically" are too vague. The reader needs specific test names to evaluate your approach.

Forgetting Assumption Checks

If you tested for normality, homogeneity of variance, or other assumptions, mention these checks in your methods. This shows the reader that you verified the conditions required for your chosen tests.

Records and Measurements to Keep During Analysis

Maintain an Analysis Log

Keep a running log of every analysis you perform, including the date, the test used, the variables included, and the software settings. This log will make writing the methods section much easier and will also help you respond to reviewer or instructor questions.

Save Your Code or Syntax

If you use R, SPSS, SAS, or Stata, save your code or syntax files. These files are the most precise record of your analysis. You can reference them in your methods section or supplementary materials.

Document Data Cleaning Decisions

Record every decision you made about excluding outliers, transforming variables, or handling missing data. These decisions affect your results and must be reported in the methods section.

Track Software Versions

Note the exact version of your statistical software and any packages or libraries you used. This information is part of your methods and should be recorded at the time of analysis.

Quality Controls for Statistical Reporting

Have a Second Person Review Your Methods

Ask a classmate, colleague, or supervisor to read your statistical analysis subsection and try to follow your analytical path. If they cannot determine what tests you used and why, your description is incomplete.

Compare Your Methods to Your Results

Read your results section and check that every test mentioned in the results appears in the methods section. Conversely, every test in the methods section should appear in the results. Mismatches indicate an error in one section or the other.

Use Reporting Guidelines as a Checklist

The EQUATOR Network provides reporting guidelines for many study types. These guidelines include specific items related to statistical methods. Use them as a checklist before submitting your work.

Verify Your Software Output

Before writing your methods, verify that your software output matches the tests you intend to describe. It is easy to misremember which test you ran, especially if you performed many analyses.

Limitations of the Methods Section

The Methods Section Does Not Justify Test Choice

The methods section describes what you did, but it does not explain why you chose one test over another. If your test choice requires justification, such as using a nonparametric test because data were not normal, you should state the assumption check that led to that decision. However, a full discussion of statistical theory belongs in the discussion section or a statistics textbook.

The Methods Section Cannot Fix Poor Study Design

A well-written methods section cannot compensate for a poorly designed study. If your sample size was too small, your groups were not comparable, or your measurements were unreliable, no amount of statistical description will make the results valid. The methods section is a record of what you did, not a justification for doing it.

Reporting Guidelines Are Not Universal

Different journals and instructors may have different expectations for statistical reporting. The EQUATOR Network guidelines are widely used, but they are not mandatory for every publication venue. Check the specific requirements of your target journal or your instructor's rubric.

Safety and Ethical Context

Data Handling and Reproducibility

Accurate statistical reporting is part of responsible research conduct. The Committee on Publication Ethics core practices emphasize the importance of transparent reporting and data integrity. When you describe your statistical methods accurately, you support the reproducibility of your work and allow others to verify your findings.

Authorship and Accountability

Every author of a research paper is accountable for the accuracy of the methods section. The Committee on Publication Ethics core practices state that all authors should have access to the data and the opportunity to review the analysis. If you are a co-author, you should understand the statistical methods well enough to describe them accurately.

Data Management Plans

Funding agencies increasingly require data management plans that describe how data will be stored, shared, and preserved. The National Institutes of Health Data Management and Sharing Policy outlines expectations for data sharing that apply to NIH-funded research. Your statistical methods should be detailed enough that someone with access to your data could reproduce your analysis.

Researcher Identity and Record Keeping

Maintaining a consistent researcher identity is important for tracking your publications and research outputs. ORCID provides a persistent identifier that links your research activities across platforms. When you publish research with detailed statistical methods, your ORCID record helps others find and verify your work.

Professional Escalation Criteria

When to Seek Help From a Statistician

You should consult a statistician or your instructor if you are unsure which test is appropriate for your data. Signs that you need help include:

• You cannot determine whether your data are normally distributed
• You have multiple predictor variables and are unsure how to model them
• You have repeated measures or nested data structures
• You are unsure whether to use parametric or nonparametric tests
• Your data violate the assumptions of the test you planned to use

When to Revise Your Methods Section

You should revise your methods section if a reviewer or instructor cannot follow your analytical path. Specific triggers for revision include:

• A reviewer asks what software you used
• A reviewer questions your choice of test
• A reviewer notes that your methods do not match your results
• A reviewer asks how you handled missing data

When to Consult Institutional Resources

If you are conducting research at a university or research institution, your institution may have a statistical consulting service. These services can help you choose appropriate tests and describe them correctly. Check with your department or the National Institutes of Health Grants and Funding resources for guidance on research rigor and reproducibility.

Frequently Asked Questions

What is the difference between the methods section and the results section for statistics?

The methods section describes what statistical tests you used and how you applied them. The results section reports what the tests found, including p-values, effect sizes, and confidence intervals. Never report specific statistical findings in the methods section.

How much detail should I include about the statistical software?

Include the software name, version number, and manufacturer if relevant. For example, "SPSS version 28 (IBM Corp., Armonk, NY)" or "R version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria)." If you used specific packages or libraries, mention those as well.

Do I need to report every statistical test I ran?

Yes, you should report every test that contributes to your results. If you ran exploratory analyses that did not answer your main research questions, you may mention them briefly or exclude them. However, any test that appears in your results section must be described in your methods section.

What if I used a test that is not commonly known?

Describe the test by name and provide a citation if the test is not widely used. You may also briefly describe what the test does and why you chose it. The goal is to give the reader enough information to understand your analytical approach.

Should I report the significance threshold in the methods section?

Yes, always state your significance threshold explicitly. The conventional threshold is P < 0.05, but you should state it instead of assuming the reader knows it. If you used a different threshold or adjusted for multiple comparisons, explain that as well.

How do I describe missing data in the methods section?

State how many observations were missing and how you handled them. Common approaches include pairwise exclusion, listwise exclusion, or imputation. For example, "Missing values were excluded pairwise from each analysis" or "Missing values were imputed using mean substitution."

Can I use the same statistical analysis description for every lab report?

You can use a template, but you must adapt it to the specific tests you used in each study. Copying a generic description without adjusting the test names and variables is a common error that makes the methods section inaccurate.

What should I do if I am unsure which statistical test to use?

Consult your instructor, a statistician, or a statistical textbook before you run the analysis. Choosing the wrong test can invalidate your results. The Research Methods Resources from the National Library of Medicine provide background on research methods that can help you make informed decisions.

Using the Evidence

Related Articles

• Statistical Analysis Methods
• Statistical Analysis
• Statistical Methods
• Statistical Analysis Biostatistics
• Statistical Data Analysis

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.