How to Present Error Bars and Uncertainty in Lab Report Figures
By Dr. Zubair Khalid, DVM, MS, PhD ·
Quick Answer
- Choose error bars that match your claim: SD describes spread in the sample, SEM describes precision of the mean estimate, and CI communicates a range with a stated confidence level.
- Label every error bar in the figure caption with the statistic name, the number of replicates, and the definition of the error bar so readers can interpret the graph without guessing.
- State the limitation in your methods section that error bars do not show statistical significance unless the figure caption explicitly says the bars represent a confidence interval for a difference.
At a Glance
| Error Bar Type | What It Shows | Best Used When | Common Label in Figure Caption |
|---|---|---|---|
| Standard Deviation (SD) | Variability of individual data points around the mean | Describing the spread of your raw measurements in the sample | "Error bars represent SD" |
| Standard Error of the Mean (SEM) | Precision of the sample mean as an estimate of the population mean | Comparing means across groups when sample size is equal | "Error bars represent SEM" |
| Confidence Interval (CI) | Range that likely contains the true population mean with a stated confidence level | Making inference about the population or comparing groups | "Error bars represent 95% CI" |
Understanding What Error Bars Communicate
Error bars are visual summaries of uncertainty that accompany point estimates such as means or medians. In a lab report figure, the point estimate is usually the mean of several replicate measurements, and the error bar extends above and below that point to show how much the measurements vary or how uncertain the mean is. The choice of error bar changes the interpretation of the figure, so the reader needs to know which statistic you plotted.
The three most common error bars in life-science lab reports are the standard deviation, the standard error of the mean, and the confidence interval. Each answers a different question about your data. The standard deviation describes how spread out the individual measurements are around the mean. The standard error of the mean describes how far the sample mean is likely to be from the true population mean. The confidence interval gives a range of plausible values for the true mean at a chosen confidence level, typically 95 percent.
A common source of confusion is that SEM bars are always shorter than SD bars for the same data because the standard error divides the standard deviation by the square root of the sample size. This visual difference can make SEM bars look more impressive, but it does not mean the data are less variable. The SEM is a statement about the mean, not about the individual observations.
When you write a lab report, the figure caption must state which error bar you used. A caption that says "error bars represent standard deviation" tells the reader that the bars show the spread of the data. A caption that says "error bars represent 95 percent confidence intervals" tells the reader that the bars show the plausible range for the true mean. Without this label, the reader cannot know whether overlapping bars mean anything about statistical significance.
Selecting the Right Error Bar for Your Claim
Use SD When Describing Data Spread
If your goal is to show how much individual replicate measurements vary, plot the standard deviation. This is appropriate when you want the reader to see the heterogeneity in your sample. For example, if you measured the body weight of ten animals per treatment group, SD bars show the range of body weights you actually observed. This is useful when the variability of the subjects is itself part of the scientific story.
SD bars are also appropriate when you plan to discuss whether the variability differs between groups. If one treatment group has a much larger SD than another, that observation may be biologically meaningful. The figure should make that difference visible.
Use SEM When Comparing Means
If your primary claim is about the mean value of each group and you want to compare means across groups, the standard error of the mean is a common choice. SEM bars give a visual sense of how precisely each mean is estimated. When sample sizes are equal across groups, SEM bars can be compared visually because the same sample size applies to every bar.
The limitation of SEM bars is that they do not tell the reader anything about the spread of the individual data points. Two groups can have identical SEM bars but very different SDs if the sample sizes differ. For this reason, some journals and instructors prefer that you report SD when the figure is meant to describe the data and SEM only when the figure is explicitly about the precision of the mean.
Use CI When Making Inference
Confidence intervals are the most informative error bar for inference because they combine the mean, the variability, and the sample size into a single range. A 95 percent confidence interval means that if you repeated the experiment many times and calculated the interval each time, approximately 95 percent of those intervals would contain the true population mean.
CI bars are useful when you want to communicate the uncertainty around a treatment effect or when you want to compare groups. When two 95 percent confidence intervals do not overlap, the difference between the means is often statistically significant, but the converse is not necessarily true. Overlapping CI bars do not prove that the groups are the same. This distinction matters in a lab report discussion because readers may draw incorrect conclusions from overlapping bars.
Labeling Error Bars in Figures and Captions
The figure caption is the contract between you and the reader. It must state the type of error bar, the number of replicates, and the definition of the error bar. A complete caption might read: "Mean enzyme activity for each treatment group. Error bars represent the standard deviation of three replicate measurements per group." This caption tells the reader exactly what the bars mean and how many measurements went into each mean.
The axis labels should also be complete. The y-axis should state the variable name and the unit of measurement. The x-axis should state the group names or the independent variable levels. If the figure contains multiple panels, each panel needs its own label or a clear key.
The figure itself should be readable without the reader having to search the methods section. A reader should be able to look at the figure, read the caption, and understand what was measured, how many replicates were used, and what the error bars represent. This is a core principle of transparent research reporting, and it applies to student lab reports as well as to published papers.
Practical Steps for Adding Error Bars to Your Figure
Step 1: Calculate the Summary Statistics
Before you can draw error bars, you need to calculate the mean and the measure of variability for each group. Use a spreadsheet or statistical software to compute the mean, the standard deviation, the standard error, and the confidence interval for each group. Record the sample size for each group because the SEM and CI both depend on the number of replicates.
Step 2: Choose the Error Bar Type Based on Your Claim
Decide what you want the figure to communicate. If the figure is descriptive and you want to show the spread of the data, use SD. If the figure is comparative and you want to show the precision of each mean, use SEM. If the figure is inferential and you want to show a range of plausible values for the true mean, use CI. Write down your choice before you make the figure so you do not change it later to make the results look different.
Step 3: Plot the Points and Add the Error Bars
Plot the mean for each group as a point or bar. Add the error bars using the values you calculated. Most spreadsheet and graphing programs have an error bar feature that lets you specify custom values. Enter the exact values for the upper and lower bounds of each error bar instead of letting the software guess.
Step 4: Write the Caption That Matches the Figure
Write the caption immediately after you make the figure. State the error bar type, the sample size, and the definition. Check that the caption matches the values you plotted. If you plotted SD bars, the caption must say SD. If you plotted 95 percent CI bars, the caption must say 95 percent CI.
Step 5: Check the Figure in the Context of Your Report
Place the figure in your report and read the surrounding text. Verify that the text describes the same error bar that the caption names. If the text says "the variability was high" but the figure shows SEM bars, the text and figure are inconsistent because SEM bars do not show the variability of the individual data points.
Common Failure Patterns in Error Bar Presentation
Using SEM Bars to Imply Low Variability
A frequent mistake is plotting SEM bars and then writing in the text that the data were tightly clustered. SEM bars are always shorter than SD bars for the same data, so they can create a false impression of low variability. If your claim is about the spread of the data, you must use SD bars. If your claim is about the precision of the mean, SEM bars are appropriate, but the text should say "the mean was estimated precisely" instead of "the data were not variable."
Overlapping Error Bars and Significance Claims
Another common failure is concluding that two groups are not different because their error bars overlap. This conclusion is only valid under specific conditions, and it depends on the type of error bar and the sample size. Overlapping SD bars tell you almost nothing about statistical significance. Overlapping 95 percent CI bars suggest that the difference may not be significant, but they do not prove it. If you want to make a claim about significance, you need to run a statistical test and report the test result in the text.
Inconsistent Error Bar Types Across Figures
Some lab reports use SD bars in one figure and SEM bars in another figure without explaining why the choice changed. This inconsistency confuses the reader and makes the report look careless. Choose one error bar type for the whole report unless there is a specific reason to change it, and state that reason in the text.
Missing Sample Size Information
An error bar without a stated sample size is uninterpretable. The SEM and CI both depend on the number of replicates, so the reader needs to know the sample size to understand the bars. Always state the sample size in the caption or in the methods section.
Records and Measurements for Error Bar Calculations
The quality of your error bars depends on the quality of your raw data records. Keep a laboratory notebook or electronic record that includes the individual replicate measurements for each group, beyond the calculated means. This record allows you to recalculate the error bars if you need to change the error bar type or if a reviewer asks for additional statistics.
Record the following for each group in your experiment:
- The individual replicate values
- The number of replicates
- The mean of the replicates
- The standard deviation of the replicates
- The standard error of the mean
- The confidence interval and the confidence level used
This record also supports the data management and sharing expectations that apply to research funded by agencies such as the National Institutes of Health. The NIH Data Management and Sharing Policy describes expectations for planning and managing scientific data, and keeping clear records of your raw measurements is the foundation of that practice.
Writing the Methods Section for Error Bars
The methods section of your lab report should state how you calculated the error bars and which statistic you plotted. A clear methods statement might read: "Data are presented as mean plus or minus standard deviation. Statistical analysis was performed using a two-sample t test with a significance threshold of p less than 0.05." This statement tells the reader exactly what the error bars mean and how the statistical analysis was conducted.
If you used confidence intervals, state the confidence level. If you used SEM, state the sample size. If you used SD, state that the bars represent the spread of the individual measurements. The methods section and the figure captions should be consistent with each other.
The EQUATOR Network provides access to reporting guidelines that help researchers write transparent and complete methods sections. While these guidelines are designed primarily for published research, the same principles apply to student lab reports. A clear methods section that states the error bar type and the statistical approach is a sign of careful scientific writing.
Describing Error Bars in the Results and Discussion
In the results section, describe what the figure shows without overinterpreting the error bars. You might write: "The mean enzyme activity was higher in the treatment group than in the control group, and the SD bars show substantial overlap between the groups." This sentence describes the figure accurately without claiming significance.
In the discussion section, interpret the error bars in the context of your research question. If the CI bars are wide, you might write that the estimate of the treatment effect is imprecise and that a larger sample size would be needed to draw firm conclusions. If the SD bars are large, you might discuss the biological variability in the system.
Avoid writing that overlapping error bars prove that two groups are identical. The absence of evidence for a difference is not evidence of absence. This distinction is important in the discussion section, where readers expect you to interpret your findings honestly.
The Role of Error Bars in Transparent Reporting
Transparent reporting of error bars is part of the broader practice of honest scientific communication. The Committee on Publication Ethics Core Practices describe expectations for data integrity and responsible research conduct. While these practices are aimed at journals and authors, the underlying principles apply to student lab reports as well. Reporting your error bars accurately and labeling them clearly is a form of data integrity.
The National Library of Medicine Bookshelf provides access to authoritative biomedical texts that discuss research methods and statistical concepts. These resources can help you understand the mathematical definitions of SD, SEM, and CI and the conditions under which each statistic is appropriate.
Choosing Between Parametric and Nonparametric Approaches
The error bars described in this article assume that your data are approximately normally distributed and that the mean is a meaningful summary of the data. If your data are heavily skewed or contain outliers, the mean may not be the best summary, and SD, SEM, and CI based on the normal distribution may be misleading.
In such cases, you might report the median and the interquartile range instead of the mean and SD. The interquartile range describes the middle 50 percent of the data and is robust to outliers. If you use the median and interquartile range, the figure caption must say so, and the methods section must explain why you chose a nonparametric approach.
This decision should be made before you analyze the data, not after you see the results. Choose the summary statistic that matches the distribution of your data and the question you are asking.
Sample Size and Its Effect on Error Bars
The sample size has a direct effect on SEM and CI bars. The SEM is calculated by dividing the SD by the square root of the sample size, so larger samples produce smaller SEM bars. The CI also narrows as the sample size increases because the standard error decreases and the critical value from the t distribution changes.
This relationship means that you cannot compare SEM or CI bars across groups with different sample sizes without accounting for the sample size. If one group has five replicates and another group has twenty replicates, the SEM bars will look very different even if the underlying variability is the same. The figure caption should state the sample size for each group so the reader can interpret the bars correctly.
SD bars do not have this problem because the SD is a property of the data and does not depend on the sample size in the same way. However, the SD estimate becomes more reliable as the sample size increases, so small samples produce noisy SD estimates.
Common Questions About Error Bars in Lab Reports
Should I Use SD or SEM in My Lab Report?
Use SD when you want to show the spread of the individual data points. Use SEM when you want to show the precision of the mean estimate. State your choice in the figure caption and be consistent across all figures in the report.
Do Overlapping Error Bars Mean the Difference Is Not Significant?
No. Overlapping error bars do not prove that two groups are the same. The relationship between error bar overlap and statistical significance depends on the type of error bar, the sample size, and the statistical test. Run the appropriate statistical test and report the result in the text.
How Do I Calculate the Standard Error of the Mean?
The standard error of the mean is the standard deviation divided by the square root of the sample size. Most spreadsheet programs have a function that calculates the standard deviation, and you can divide that value by the square root of the sample size to get the SEM.
What Does a 95 Percent Confidence Interval Mean?
A 95 percent confidence interval is a range of values that would contain the true population mean in approximately 95 percent of repeated experiments. It combines the mean, the variability, and the sample size into a single range.
Can I Use Error Bars for Nonparametric Data?
If your data are not normally distributed, consider reporting the median and interquartile range instead of the mean and SD. The figure caption must state which summary statistics you used.
How Many Replicates Do I Need for Reliable Error Bars?
More replicates generally produce more reliable estimates of the mean and the variability. The exact number depends on the variability of your system and the size of the effect you are trying to detect. Plan your sample size before you collect data and state your rationale in the methods section.
Professional Escalation Criteria
If you are preparing a lab report for a course and your instructor has specific requirements for error bars, follow those requirements even if they differ from general practice. If you are preparing a manuscript for publication, consult the journal's instructions to authors and the relevant reporting guideline from the EQUATOR Network. If you are analyzing data for a regulated study, such as a preclinical safety study, consult the applicable regulations and your institutional review board before choosing error bars.
If you are uncertain about which error bar to use or how to interpret your results, ask your instructor, supervisor, or a statistical consultant before you submit the report. A statistical error in a figure can undermine the credibility of the entire report, and it is better to ask for help than to submit a figure that misleads the reader.
Frequently Asked Questions
What is the difference between SD and SEM error bars?
SD error bars show the spread of the individual data points around the mean. SEM error bars show the precision of the mean as an estimate of the population mean. SEM is always smaller than SD for the same data because it divides SD by the square root of the sample size.
How do I know which error bar to use in my figure?
Choose the error bar that matches your claim. If you are describing the variability of the data, use SD. If you are comparing the precision of means across groups, use SEM. If you are making an inference about the population mean, use a confidence interval.
What should I write in the figure caption about error bars?
State the type of error bar, the number of replicates, and the definition. For example, "Error bars represent the standard deviation of three replicate measurements per group." This caption tells the reader exactly what the bars mean.
Do error bars show statistical significance?
Error bars alone do not show statistical significance. The relationship between error bar overlap and significance depends on the type of error bar and the sample size. You need to run a statistical test and report the test result to make a claim about significance.
Can I use confidence intervals instead of SD or SEM?
Yes. Confidence intervals are often more informative than SD or SEM because they combine the mean, the variability, and the sample size into a single range. State the confidence level, such as 95 percent, in the caption.
What if my data are not normally distributed?
If your data are skewed or contain outliers, consider reporting the median and interquartile range instead of the mean and SD. State this choice in the caption and explain it in the methods section.
How many replicates do I need for meaningful error bars?
The number of replicates depends on the variability of your system and the size of the effect you are studying. More replicates give more reliable estimates. Plan your sample size before collecting data and state your rationale in the methods section.
Should I use the same error bar type in every figure?
Yes, unless you have a specific reason to change the type. Consistent error bar types across figures make the report easier to read and interpret. If you change the type, explain why in the text.
Using the Evidence
| Source | Best use in this topic | Important limitation |
|---|---|---|
| Research Methods Resources | official guidance | Check the linked page for current local requirements |
| EQUATOR Network | official guidance | Check the linked page for current local requirements |
| Core Practices | official guidance | Check the linked page for current local requirements |
Related Articles
- Factorial Designs: When and How to Use Them
- Peer Review Metrics: What They Are and How to Use Them
- Data Presentation in Research: Tables, Figures, and Text
- Control Groups in Experiments: How to Choose and Use Them
- How to Present Data in Research Papers: Methods, Tables, and Figures
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.