Lab Report Statistical Reporting

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

Quick Answer

• Report every statistical test with its exact p-value, test statistic, and degrees of freedom so readers can verify your conclusions independently.
• Choose error bars that match your experimental design, using standard deviation for descriptive data and standard error or confidence intervals for inferential claims.
• Match your reporting style to the journal or assignment guidelines you are following, since biology, chemistry, and psychology labs each have distinct conventions.

Why Statistical Reporting Matters in Lab Reports

Statistical reporting is the bridge between raw measurements and scientific conclusions. When you record that a treatment increased plant height, you must also show the evidence that this increase was not due to random variation. A lab report that omits this evidence forces readers to trust your conclusion without verification, which undermines the entire purpose of scientific writing.

The Committee on Publication Ethics identifies data integrity and accurate reporting as core responsibilities for researchers. Their Core Practices emphasize that authors must present their work honestly and transparently, which includes reporting statistical methods and results without omission or misrepresentation. This standard applies also to published papers but also to the lab reports you write as a student, since these documents train you in the habits that professional researchers use throughout their careers.

Statistical reporting errors are not minor formatting issues. A missing p-value, an incorrect decimal place, or a misleading error bar can change the interpretation of your results. In some cases, these errors can lead readers to believe that a non-significant result is significant, or that a significant result is not. The practical consequence is that your conclusions lose credibility, and in professional settings, flawed statistical reporting can contribute to irreproducible research.

The EQUATOR Network maintains a comprehensive collection of reporting guidelines for health research, including specific checklists for randomized trials, observational studies, and diagnostic accuracy studies. While these guidelines target published research, they offer a useful framework for understanding what complete statistical reporting looks like. The underlying principle is that your report should contain enough information for another researcher to understand exactly what you did and why your conclusions follow from your data.

At a Glance: Common Statistical Reporting Errors

Core Principles of Statistical Reporting

Report the Exact Test and Its Parameters

Every statistical test you perform should be identified by name, and you should report the specific parameters that allow readers to understand what you tested. For a t-test, this means reporting the t-value, the degrees of freedom, and the p-value. For an analysis of variance, you report the F-value, the degrees of freedom for both the numerator and denominator, and the p-value. For a chi-square test, you report the chi-square value, the degrees of freedom, and the p-value.

The format for these reports follows a standard convention. A t-test result might appear as t(28) = 3.42, p = 0.002, where 28 is the degrees of freedom. An ANOVA result might appear as F(2, 45) = 5.67, p = 0.006, where 2 and 45 are the degrees of freedom for the between-groups and within-groups variance, respectively. A chi-square result might appear as χ²(1, N = 80) = 6.21, p = 0.013.

These notations are not arbitrary. The degrees of freedom tell readers how many independent pieces of information went into the estimate, which affects how much confidence they can place in the result. The test statistic itself allows readers to check your work if they have access to your raw data. The p-value tells them the probability of observing a result at least as extreme as yours, assuming the null hypothesis is true.

Distinguish Statistical Significance from Practical Importance

A statistically significant result means that the observed effect is unlikely to have occurred by chance, given your sample size and variability. It does not mean that the effect is large, important, or practically meaningful. A very small effect can be statistically significant if your sample size is large enough, and a large effect can fail to reach significance if your sample is small or your data are highly variable.

Your lab report should address both dimensions. Report the p-value to establish statistical significance, but also describe the magnitude of the effect in the context of your experiment. If you measured the growth of plants under two light conditions, report also that the difference was significant but also how many centimeters the plants grew on average under each condition. This contextual information helps readers judge whether the effect matters for the question you are investigating.

Match Your Reporting to Your Experimental Design

The statistical tests you use must match the design of your experiment. A paired t-test is appropriate when you measure the same subjects before and after a treatment. An independent t-test is appropriate when you compare two separate groups. A one-way ANOVA is appropriate when you compare three or more groups on a single factor. A repeated-measures ANOVA is appropriate when you measure the same subjects under multiple conditions.

Your lab report should explain why you chose each test. This explanation demonstrates that you understand the logic of statistical inference, beyond the mechanics of running a test in software. For example, if you used a paired t-test, explain that each subject served as its own control, which reduced the influence of individual differences on your results.

Practical Workflow for Reporting Statistics

Step 1: Record Your Raw Data and Methods Before Analysis

Before you run any statistical test, document your data collection procedures and your raw measurements. This record should include the number of subjects or samples, the units of measurement, and any data that you excluded and why. The NIH Data Management and Sharing Policy describes expectations for preserving and sharing research data, and the same principles apply to your lab work. A complete data record allows you to verify your statistical results and provides the evidence you need if a reviewer questions your analysis.

Step 2: Choose Your Statistical Test Based on Your Design

Select your test before you look at your data. This decision should be based on your experimental design and your research question, not on which test gives you a significant result. If you are comparing two groups, use a t-test. If you are comparing three or more groups, use an ANOVA. If you are examining the relationship between two continuous variables, use correlation or regression.

Document your choice and your reasoning. This documentation is part of transparent reporting, which the EQUATOR Network promotes as essential for research credibility. When you write your lab report, include a sentence in your methods section that states which test you used and why it was appropriate for your design.

Step 3: Run the Test and Record the Output

When you run your statistical test, record the full output from your software. This output should include the test statistic, the degrees of freedom, and the p-value. Do not rely on your memory or on a screenshot that you might lose. Copy the relevant values into your lab notebook or a dedicated analysis file.

If your software reports a p-value as something like 0.000, report it as p < 0.001 instead of p = 0.000. A p-value cannot actually be zero, and reporting it as such suggests that you do not understand what the value represents.

Step 4: Write the Results in Standard Format

Write your results section using the standard format for each test. Include the test statistic, the degrees of freedom, and the p-value in a single sentence or parenthetical note. For example, "The mean height of plants grown under red light was significantly greater than the mean height of plants grown under blue light, t(28) = 3.42, p = 0.002."

If you ran multiple tests, report each one separately. Do not combine results from different tests into a single sentence unless they are directly related to the same comparison.

Step 5: Create Figures and Tables That Match Your Text

Your figures and tables should present the same information as your text, not contradict it. If your text reports a significant difference between groups, your figure should show that difference visually. If your text reports no significant difference, your figure should not exaggerate the apparent difference through a misleading scale or truncated axis.

Label every figure and table with a descriptive caption that identifies the variables, the units, and the type of error bars shown. This caption is part of your statistical reporting, and it should be specific enough that a reader can understand the figure without referring to the text.

Options and Tradeoffs in Statistical Presentation

Standard Deviation, Standard Error, and Confidence Intervals

The choice of error bars depends on what you want to communicate. Standard deviation describes the spread of your individual data points around the mean. Standard error describes the precision of your estimate of the mean. Confidence intervals describe the range of values within which the true population mean is likely to fall, given your sample.

Use standard deviation when you want to show the variability of your measurements. Use standard error when you want to show how precisely you have estimated the mean. Use confidence intervals when you want to show the uncertainty around your estimate in a way that supports inferential claims.

Your figure caption must state which type of error bar you used. A caption that simply says "error bars" without specifying the type is incomplete and can mislead readers. If you used standard error bars, readers might interpret them as standard deviation and draw incorrect conclusions about the variability of your data.

Parametric and Nonparametric Tests

Parametric tests, such as the t-test and ANOVA, assume that your data follow a normal distribution and that your groups have similar variances. Nonparametric tests, such as the Mann-Whitney U test and the Kruskal-Wallis test, do not make these assumptions. They are appropriate when your data are skewed, ordinal, or otherwise not suitable for parametric analysis.

Your lab report should state which type of test you used and why. If you used a nonparametric test, explain that your data did not meet the assumptions of the parametric equivalent. This explanation shows that you checked the assumptions and made an informed choice, instead of defaulting to the first test your software offered.

One-Tailed and Two-Tailed Tests

A two-tailed test checks for a difference in either direction. A one-tailed test checks for a difference in a specific direction. One-tailed tests have more statistical power to detect an effect in the predicted direction, but they cannot detect an effect in the opposite direction.

Use a two-tailed test unless you have a strong theoretical reason to predict the direction of the effect before you collect your data. If you use a one-tailed test, state this choice explicitly in your methods section and explain your justification. Reporting a one-tailed test without justification can be seen as an attempt to make a marginal result appear more significant than it is.

Records and Measurements

What to Record in Your Lab Notebook

Your lab notebook should contain the raw data from your experiment, the units of measurement, the date and time of each measurement, and any observations about the conditions under which you collected the data. This record is the foundation of your statistical analysis, and it should be complete enough that another researcher could reproduce your analysis from your notes alone.

Record any data that you excluded from your analysis and the reason for the exclusion. Excluding outliers is sometimes justified, but you must document the decision and the criteria you used. The Committee on Publication Ethics treats data fabrication and falsification as serious misconduct, and incomplete or selective reporting of data can raise similar concerns.

How to Document Your Statistical Software

Record the name and version of the statistical software you used, along with the specific commands or procedures you ran. This documentation allows readers to understand exactly how you computed your results and to reproduce your analysis if they have access to your data.

If you used a spreadsheet program for your analysis, note that this is generally less rigorous than using dedicated statistical software. Spreadsheet programs can produce incorrect results for some statistical procedures, and they do not provide the same level of documentation and reproducibility as statistical packages.

How to Present Descriptive Statistics

Descriptive statistics summarize your data before you run any inferential tests. Report the mean and standard deviation for each group, or the median and interquartile range if your data are not normally distributed. Include the sample size for each group, since this information is essential for interpreting the inferential statistics that follow.

Present descriptive statistics in a table or in the text of your results section. If you use a table, include a caption that identifies the variables, the groups, and the statistics reported. If you use the text, write a sentence such as "The mean height of plants grown under red light was 24.3 cm (SD = 3.1), and the mean height of plants grown under blue light was 18.7 cm (SD = 2.8)."

Common Failure Patterns in Statistical Reporting

Reporting p-Values Without Test Statistics

A p-value without the test statistic and degrees of freedom is incomplete. The reader cannot verify the result or understand the strength of the evidence. This pattern is common in student lab reports, where the writer reports that a result was significant but does not provide the supporting details.

The fix is to report the full test result in a standard format. For a t-test, write t(degrees of freedom) = value, p = value. For an ANOVA, write F(numerator df, denominator df) = value, p = value. This format is concise and provides all the information a reader needs to evaluate your result.

Using the Wrong Number of Decimal Places

Reporting too many decimal places suggests false precision, while reporting too few can obscure meaningful differences. A p-value of 0.123456789 implies that your measurement was precise to nine decimal places, which is almost never the case. A p-value of 0.1, on the other hand, does not tell the reader whether the result was significant at the 0.05 level.

The standard convention is to report p-values to two or three decimal places, with values below 0.001 reported as p < 0.001. Report test statistics and descriptive statistics to a number of decimal places that matches the precision of your measurements. If you measured plant height to the nearest millimeter, report means to one decimal place in centimeters, not to three or four.

Confusing Error Bar Types

Error bars are a common source of confusion in lab reports. A figure with error bars that are not labeled forces the reader to guess what they represent. If the error bars represent standard error but the reader assumes they represent standard deviation, the reader will underestimate the variability of the data.

The fix is to label your error bars in the figure caption and to choose the type of error bar that matches your claim. If you are making an inferential claim about the difference between groups, use confidence intervals or standard error. If you are describing the variability of your measurements, use standard deviation.

Reporting Non-Significant Results as Trends

A non-significant result means that your data do not provide sufficient evidence to reject the null hypothesis. It does not mean that there is a trend, a tendency, or a near-significant effect. Phrases such as "approached significance" or "showed a trend toward significance" are misleading and should be avoided.

If your result is not significant, report the p-value and state that the difference was not statistically significant. You may discuss possible reasons for the non-significant result in your discussion section, but you should not describe it as a trend or a near-miss.

Omitting the Sample Size

The sample size is essential for interpreting any statistical result. A p-value from a study with 5 subjects is not comparable to a p-value from a study with 500 subjects. Your lab report should state the sample size for each group in your methods section and repeat it in your results section when you report your descriptive statistics.

Using Figures That Contradict the Text

A figure that exaggerates a small difference or obscures a large difference can contradict the statistical results reported in your text. Truncated axes, misleading scales, and inappropriate error bars can all create this problem. Your figures should present your data honestly and should be consistent with the statistical results you report.

Quality and Welfare Controls

Checking Your Work Before Submission

Before you submit your lab report, check every statistical result against your raw data and your software output. Verify that the test statistic, degrees of freedom, and p-value match what you recorded when you ran the analysis. Verify that your figures show the same results as your text. Verify that your error bars are labeled correctly and that they match the type of error bar you describe in your caption.

This checking process is part of responsible research practice. The Committee on Publication Ethics expects authors to take responsibility for the accuracy of their work, and the same expectation applies to student lab reports. A careful check before submission can catch errors that would otherwise undermine your conclusions.

Getting Feedback From Peers or Instructors

Ask a classmate or your instructor to review your results section before you submit it. A fresh reader can often spot errors that you have overlooked, such as a missing p-value or an unclear figure caption. This feedback is especially valuable for learning the conventions of statistical reporting, since these conventions are often implicit and are best learned through practice and correction.

Using Reporting Guidelines as a Checklist

The EQUATOR Network provides reporting guidelines for many types of research studies. While these guidelines are designed for published research, they can serve as a checklist for your lab report. Look for a guideline that matches your study design and use it to check that you have reported all the necessary information.

For a simple lab experiment, you may not need a formal guideline, but you should still check that your report includes the essential elements: the research question, the methods, the statistical tests, the results, and the interpretation. A complete report allows a reader to understand what you did and why your conclusions follow from your data.

Limitations and Professional Escalation

When to Seek Help From a Statistician

If you are unsure which statistical test to use, or if your data do not meet the assumptions of the test you planned to use, seek help from a statistician or your instructor. This is especially important for complex experimental designs, such as those with multiple factors, repeated measures, or nested data structures. Guessing at the wrong test can produce misleading results that are difficult to correct after the fact.

When to Consult Your Instructor About Reporting Standards

Different courses and different journals have different standards for statistical reporting. Some require specific formats for p-values, while others prefer confidence intervals. Some require that you report effect sizes, while others do not. If you are unsure about the standards for your assignment, ask your instructor before you submit your report.

When to Escalate Concerns About Data Integrity

If you suspect that your data are fabricated, falsified, or otherwise compromised, you should report your concerns to your instructor or research supervisor. The Committee on Publication Ethics treats data fabrication and falsification as serious misconduct, and addressing these concerns promptly is part of responsible research practice. This applies to your own work and to the work of others that you are asked to review or verify.

Safety and Regulatory Context

Institutional Policies on Data Management

Many institutions have policies on data management, data sharing, and data retention. These policies may specify how long you must keep your raw data, how you should store it, and who has access to it. The NIH Data Management and Sharing Policy describes expectations for NIH-funded research, and similar principles may apply to your institution even if your work is not federally funded.

Responsible Authorship and Acknowledgment

If you worked with a lab partner or a research group, you should acknowledge their contributions appropriately in your report. The Committee on Publication Ethics identifies authorship and acknowledgment as core ethical responsibilities, and the same principles apply to student work. Discuss authorship and acknowledgment with your collaborators before you submit your report, and make sure that everyone who contributed is appropriately credited.

Researcher Identity and Record Keeping

Maintaining an accurate record of your research activities is part of professional practice. The ORCID for Researchers describes how researchers can maintain a persistent identifier that links their work across publications and projects. While you may not need an ORCID iD as a student, developing the habit of accurate record keeping will serve you well in your future career.

Frequently Asked Questions

Why do I need to report the degrees of freedom in my lab report?

Degrees of freedom tell the reader how many independent pieces of information went into your estimate, which affects how much confidence they can place in your result. Without degrees of freedom, the reader cannot fully evaluate your test statistic or your p-value.

What is the difference between standard deviation and standard error?

Standard deviation describes the spread of your individual data points around the mean. Standard error describes the precision of your estimate of the mean. Use standard deviation for descriptive purposes and standard error or confidence intervals for inferential claims.

How many decimal places should I use for p-values?

Report p-values to two or three decimal places, and report values below 0.001 as p < 0.001. Do not report p = 0.000, since a p-value cannot actually be zero.

Can I report a non-significant result as a trend?

No. A non-significant result means that your data do not provide sufficient evidence to reject the null hypothesis. Describing it as a trend or a near-miss is misleading and should be avoided.

What should I do if my data do not meet the assumptions of a parametric test?

Use a nonparametric test instead, such as the Mann-Whitney U test or the Kruskal-Wallis test. State in your methods section that your data did not meet the assumptions of the parametric test and explain why you chose the nonparametric alternative.

How do I choose between a one-tailed and a two-tailed test?

Use a two-tailed test unless you have a strong theoretical reason to predict the direction of the effect before you collect your data. If you use a one-tailed test, state this choice explicitly in your methods section and explain your justification.

What should I include in a figure caption for a bar chart with error bars?

Your caption should identify the variables, the units, the type of error bar, and the sample size. For example, "Mean plant height under red and blue light. Error bars represent standard deviation, n = 15 per group."

Why is it important to report the sample size in my results section?

The sample size is essential for interpreting any statistical result. A p-value from a study with 5 subjects is not comparable to a p-value from a study with 500 subjects. Reporting the sample size allows the reader to judge the strength of your evidence.

Using the Evidence

Related Articles

• Types Of Statistical Errors
• Statistical Synonyms: A Guide to Terminology in Statistics
• Effect Size in Statistics: Interpreting the Magnitude of Findings
• Gene Ontology Enrichment Analysis: Avoiding Common Interpretation Errors
• Which R-Value Represents the Strongest Correlation? A Guide to Interpreting Correlation Coefficients

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.