Box Plot and Outlier Calculator

Compute five-number summaries, IQR fences, whisker endpoints, outliers, sorted data, and side-by-side box plots from raw data.

Results are calculated automatically as you enter data.

Enter values for each group. Empty cells are ignored, allowing unequal sample sizes.

Raw data

Quartile and fence formulas

▼ See explanations and tips below ▼

Related Calculators

What Is a Box Plot?

A box plot, also called a box-and-whisker plot, is a compact graph for summarizing a numeric dataset. Instead of showing every individual value, it shows the dataset’s center, middle spread, overall range, and possible outliers.

The main idea is simple: sort the values, divide them into quarters, and draw a box around the middle half of the data. The box begins at the first quartile, \(Q_1\), and ends at the third quartile, \(Q_3\). A line inside the box marks the median. Whiskers extend outward to show the smallest and largest non-outlier values, and any values beyond the outlier fences are plotted separately.

Box plots are especially useful when you want to compare several groups on the same scale. For example, a teacher might compare test scores across classes, a data analyst might compare delivery times across warehouses, or a researcher might compare measurements across treatment groups. Each box plot gives a quick visual summary of where the values tend to fall and how much they vary.

Why Box Plots Matter

A table of raw data can be precise, but it can also be hard to read. A box plot turns a list of numbers into a visual summary that highlights several important features at once:

  • Center: The median shows the middle of the sorted data.
  • Spread: The box width or height shows the spread of the middle \(50\%\) of the values.
  • Skew: A longer whisker on one side, or a median line that is far from the center of the box, may suggest that the data are stretched more in one direction.
  • Potential outliers: Points beyond the outlier fences are separated from the whiskers so they can be investigated.
  • Group differences: Multiple box plots make it easier to compare medians, variability, and unusual values across groups.

Box plots do not replace the raw data or a full statistical analysis. Their strength is that they show the structure of a numeric distribution quickly, especially during exploratory data analysis.

Key Terms to Know

  • Observation: One numeric value in the dataset.
  • Sorted data: The same values arranged from smallest to largest.
  • Median: The middle value of sorted data. If there are two middle values, the median is their average.
  • Quartiles: Values that divide sorted data into quarters. The first quartile \(Q_1\) is near the lower quarter, the second quartile is the median, and the third quartile \(Q_3\) is near the upper quarter.
  • Five-number summary: The minimum, \(Q_1\), median, \(Q_3\), and maximum.
  • Interquartile range: The spread of the middle half of the values, calculated as \(Q_3 - Q_1\).
  • Fence: A cutoff used to flag potential outliers.
  • Whisker: The line extending from the box to the smallest or largest non-outlier value.
  • Outlier: A value outside the lower or upper fence under the \(1.5 \times IQR\) rule.

How Box Plots and Outlier Fences Work

A box plot starts with sorted numeric data. For a group of \(n\) values, write the sorted observations as:

$$ x_{(1)} \le x_{(2)} \le \cdots \le x_{(n)} $$

The median is found first. If \(n\) is odd, the median is the middle sorted value. If \(n\) is even, the median is the average of the two middle sorted values:

$$ M = \frac{x_{(n/2)} + x_{(n/2 + 1)}}{2} $$

Next, the data are split into a lower half and an upper half. The first quartile \(Q_1\) is the median of the lower half, and the third quartile \(Q_3\) is the median of the upper half.

This calculator offers two quartile split methods:

  • Inclusive method: When the sample size is odd, the median is included in both the lower and upper halves before finding \(Q_1\) and \(Q_3\). When the sample size is even, the data are split evenly into the lower and upper halves.
  • Exclusive method: When the sample size is odd, the median is excluded from both halves before finding \(Q_1\) and \(Q_3\). When the sample size is even, the split is the same as the inclusive even-size split.

Once \(Q_1\) and \(Q_3\) are known, the interquartile range is:

$$ IQR = Q_3 - Q_1 $$

The lower and upper outlier fences are then calculated as:

$$ \text{Lower Fence} = Q_1 - 1.5 \times IQR $$
$$ \text{Upper Fence} = Q_3 + 1.5 \times IQR $$

A value is flagged as an outlier if it is less than the lower fence or greater than the upper fence:

$$ x < \text{Lower Fence} \quad \text{or} \quad x > \text{Upper Fence} $$

The whiskers do not usually stop at the fence values themselves. They stop at the smallest and largest actual observations that are still inside the fences. This is why a box plot can show a minimum or maximum value in the results table while the visible whisker ends at a different value.

Examples of Box Plots in Practice

Example 1: One Group With a Possible Outlier

Suppose the sorted data are:

$$ 4,\ 7,\ 8,\ 9,\ 10,\ 11,\ 13,\ 18,\ 31 $$

There are \(9\) values, so the median is the fifth value:

$$ M = 10 $$

Using the inclusive quartile method, the median is included in both halves:

$$ \text{Lower half} = 4,\ 7,\ 8,\ 9,\ 10 $$
$$ \text{Upper half} = 10,\ 11,\ 13,\ 18,\ 31 $$

So:

$$ Q_1 = 8, \qquad Q_3 = 13 $$

The interquartile range is:

$$ IQR = Q_3 - Q_1 = 13 - 8 = 5 $$

The fences are:

$$ \text{Lower Fence} = 8 - 1.5 \times 5 = 0.5 $$
$$ \text{Upper Fence} = 13 + 1.5 \times 5 = 20.5 $$

Any value greater than \(20.5\) is an outlier, so \(31\) is plotted separately. The upper whisker ends at \(18\), the largest actual value inside the fence.

Example 2: How the Quartile Method Can Change the Result

Use the same data:

$$ 4,\ 7,\ 8,\ 9,\ 10,\ 11,\ 13,\ 18,\ 31 $$

With the exclusive method, the median \(10\) is left out of both halves:

$$ \text{Lower half} = 4,\ 7,\ 8,\ 9 $$
$$ \text{Upper half} = 11,\ 13,\ 18,\ 31 $$

Now the quartiles are the averages of the two middle values in each half:

$$ Q_1 = \frac{7 + 8}{2} = 7.5 $$
$$ Q_3 = \frac{13 + 18}{2} = 15.5 $$

So:

$$ IQR = 15.5 - 7.5 = 8 $$

The fences become:

$$ \text{Lower Fence} = 7.5 - 1.5 \times 8 = -4.5 $$
$$ \text{Upper Fence} = 15.5 + 1.5 \times 8 = 27.5 $$

The value \(31\) is still an outlier, but the quartiles, IQR, and fences are different. This is why it is important to use the same quartile method when comparing results from different tools, classes, or reports.

Example 3: Comparing Two Groups

Imagine two groups of scores:

Group Values
Class A \(62, 71, 72, 80, 91, 99, 120\)
Class B \(55, 67, 70, 73, 88, 89, 94\)

Using the inclusive method, Class A has a median of \(80\), \(Q_1 = 71.5\), \(Q_3 = 95\), and \(IQR = 23.5\). Class B has a median of \(73\), \(Q_1 = 68.5\), \(Q_3 = 88.5\), and \(IQR = 20\).

The Class A box is shifted farther to the right and has a slightly larger middle spread. That suggests Class A has a higher center and somewhat more variation in the middle half of the scores. However, the interpretation only makes sense because both groups are on the same measurement scale.

How to Interpret the Result

The result is easiest to understand by reading the box plot from the inside outward.

The median line shows the center of the sorted data. Half of the observations are at or below the median, and half are at or above it, allowing for repeated values.

The box runs from \(Q_1\) to \(Q_3\). This middle section contains about half of the data. A wider box means the middle half of the observations is more spread out. A narrower box means those middle values are more tightly clustered.

The IQR is the numerical width of the box:

$$ IQR = Q_3 - Q_1 $$

A larger IQR means more variability in the middle half of the dataset. A smaller IQR means the middle half is more concentrated.

The whiskers extend to the smallest and largest actual values that are not outside the fences. When there are no outliers, the whiskers often reach the minimum and maximum values. When outliers are present, the whiskers stop before the minimum or maximum, and the outlying values are shown as separate points.

The outlier count tells you how many observations fall beyond the \(1.5 \times IQR\) fences. Outliers should be investigated, not automatically deleted. They may be data-entry errors, rare but valid observations, or important signals about the process being studied.

When comparing multiple groups, focus on whether the groups use compatible units and measurement contexts. A box plot can compare test scores with test scores or delivery times with delivery times, but it cannot make a meaningful visual comparison between unrelated units such as kilograms and minutes.

Common Mistakes and Misconceptions

Mistaking whiskers for the minimum and maximum. In a basic box plot without outlier markers, whiskers may reach the minimum and maximum. In a box plot with outlier fences, whiskers stop at the smallest and largest non-outlier values instead.

Confusing fences with whiskers. The lower and upper fences are calculated cutoff values. The whiskers usually end at actual data values inside those cutoffs, not at the fence values themselves.

Assuming all quartile methods are identical. Quartiles are not always computed the same way across calculators, spreadsheets, statistical software, and textbooks. Small odd-sized datasets are especially likely to show differences between inclusive and exclusive methods.

Mixing incompatible units. A multi-group box plot only makes sense when the groups are measured on the same scale or on directly comparable scales.

Treating outliers as automatic mistakes. A value outside the fence is unusual under the selected rule, but it is not necessarily wrong. It should be checked in context.

Using too little data. A box plot summary is weak when a group has only a few values. This calculator requires at least three finite numeric values in each included group.

Pasting formatted numbers as raw data. Separators and symbols matter. For example, a value written with a thousands separator such as 1,234 may be read as two separate values, \(1\) and \(234\), because commas are treated as delimiters.

When to Use Box Plots

Use a box plot when you want to summarize or compare numeric data, especially when the dataset is large enough that individual values are hard to inspect one by one.

Box plots are useful for:

  • comparing distributions across classes, groups, locations, products, or time periods;
  • checking whether one group has a higher or lower median than another;
  • comparing the spread of the middle half of several groups;
  • spotting possible outliers for follow-up;
  • summarizing skewed data where the median and IQR may be more informative than the mean and standard deviation;
  • creating a quick exploratory view before choosing a more detailed analysis.

A box plot is less useful when the exact shape of the distribution matters. For example, a box plot does not show whether the data have two peaks, gaps, or clusters inside the quartiles. In those cases, a histogram, dot plot, or raw-data table may be a better companion.

Limitations and Things to Keep in Mind

A box plot is a summary, so it hides some details. Two datasets can have similar medians and quartiles but very different internal patterns. Always inspect the raw data or use another graph when the detailed shape matters.

The \(1.5 \times IQR\) rule is a convention for flagging possible outliers. It is helpful for exploratory work, but it is not a universal proof that a value is wrong, impossible, or should be removed. In scientific, engineering, financial, health, or official reporting contexts, investigate outliers using the standards and subject-matter knowledge appropriate to the decision.

Quartile methods can differ. This calculator uses a split-half method with inclusive and exclusive options. Other tools may use interpolation-based percentile definitions or software-specific quantile methods. When results must match exactly, use the same quartile definition throughout the analysis.

The calculator works with raw numeric observations. It does not handle weighted data, grouped intervals, frequency-table notation, categorical labels, date-time values, unit conversions, or custom outlier multipliers. Blank cells are ignored, and only finite numeric values are included.

Displayed numbers may be rounded or formatted for readability. Very small values may display as zero, and very large or very small nonzero values may appear in scientific notation. For important work, keep the original data and check whether rounding affects your interpretation.

How to Use This Calculator

  1. Enter numeric observations in the raw-data table, or paste grouped data into the import area.
  2. Rename the group headings if the default labels do not describe your data.
  3. Add rows or groups when you need more observations or more datasets.
  4. Choose the inclusive or exclusive quartile method.
  5. Review the summary cards for the number of groups, total values, selected method, and outlier count.
  6. Read the box plot to compare medians, IQRs, whiskers, and outlier points.
  7. Check the results table for each group’s \(n\), minimum, \(Q_1\), median, \(Q_3\), maximum, IQR, and outlier values.
  8. Download the graph as a PNG if you need to save or share the visual.

For pasted data, use a CSV-style table when possible. Put group names in the first row, such as Class A,Class B, then place observations in rows below it. Blank cells are ignored. Avoid currency symbols, percent signs, unit suffixes, and thousands separators.

Frequently Asked Questions

What does the box in a box plot represent?

The box runs from \(Q_1\) to \(Q_3\), so it represents the middle half of the data. Its width or height is the interquartile range. A larger box means the middle half of the values is more spread out.

Are box plot outliers always errors?

No. An outlier is a value outside the calculated fences under the selected rule. It may be a mistake, but it may also be a valid rare value or a meaningful signal. Check the original data and the real-world context before removing it.

Why do my whiskers not reach the minimum and maximum?

When outliers are present, whiskers extend only to the smallest and largest actual values inside the fences. The minimum or maximum can appear separately as an outlier point if it falls outside the fences.

Why do inclusive and exclusive quartiles give different answers?

For odd sample sizes, the inclusive method includes the median in both halves before calculating \(Q_1\) and \(Q_3\), while the exclusive method leaves the median out. This can change the quartiles, IQR, fences, and outlier classification.

Can I compare two box plots with different units?

Not meaningfully. Box plots compare values along a shared numeric scale. If one group is measured in minutes and another in kilograms, the visual comparison does not describe a real difference in center or spread.

What happens when the IQR is zero?

An IQR of zero means \(Q_1\) and \(Q_3\) are equal. This often happens when many values are the same. The fences collapse to that same value, so values different from the shared quartile value may be flagged as outliers depending on the full dataset.

Sources and References

Books and Textbooks

  1. Barbara Illowsky and Susan Dean. Introductory Statistics 2e. OpenStax, 2023. Sections 2.3 Measures of the Location of the Data and 2.4 Box Plots.
  2. Wanhua Su. Introduction to Applied Statistics. MacEwan Open Books, 2024. Chapter 2, section 2.4 Five-Number Summary and Boxplot. DOI: 10.31542/b.gm.5.

Online and Research Sources

  1. National Institute of Standards and Technology. “1.3.3.7. Box Plot.” NIST/SEMATECH e-Handbook of Statistical Methods, accessed June 27, 2026. https://www.itl.nist.gov/div898/handbook/eda/section3/boxplot.htm
  2. Rob J. Hyndman and Yanan Fan. “Sample Quantiles in Statistical Packages.” The American Statistician, vol. 50, no. 4, 1996, pp. 361–365. Publication page accessed June 27, 2026. https://robjhyndman.com/publications/quantiles/
  3. Penn State Eberly College of Science. “3.2 - Identifying Outliers: IQR Method.” STAT 200, accessed June 27, 2026. https://online.stat.psu.edu/stat200/lesson/3/3.2