How to Read a Box Plot

A box plot shows five numbers: minimum, first quartile, median, third quartile and maximum. The box holds the middle half of the data, the line inside it is the median, and the whiskers reach the farthest points inside the fences. This guide reads one worked plot step by step, and shows how Tukey and Excel methods draw different boxes.

By the end of this guide you will be able to read any box plot in seconds: where the middle half of the data sits, where its centre is, how far the whiskers reach, and whether any points are outliers. You will also know why two box plots of the same data can look different β€” and which numbers to check before you compare them.

The five numbers

A box plot is built from five numbers, and the whole plot is drawn from them. They are the minimum, the first quartile (Q1), the median, the third quartile (Q3) and the maximum. Each one marks a point on the number line, and the plot turns those five points into a box, a line and two whiskers.

NumberWhat it isWhere it sits on the plot
MinimumThe smallest valueThe end of the lower whisker, unless it is an outlier
First quartile (Q1)The 25th percentileThe bottom edge of the box
MedianThe middle valueThe line inside the box
Third quartile (Q3)The 75th percentileThe top edge of the box
MaximumThe largest valueThe end of the upper whisker, unless it is an outlier

Read the five numbers in order and you have read the plot. Everything else in this guide is about what each part of the plot does with them.

The box holds the middle 50%

The box runs from Q1 at its bottom edge to Q3 at its top edge. Its height is the interquartile range, IQR = Q3 βˆ’ Q1. The middle half of the data sits inside the box: a quarter of the values are below it and a quarter are above it.

A wide box means the middle half is spread out; a narrow box means it is tightly bunched. The box does not hold all the data β€” only the middle 50%. The whiskers and any dots hold the rest.

For the 20 values in this guide the box runs from 17.5 to 27.5, so the IQR is 10.0. The worked plot later in this guide reads the whole plot from those numbers.

The line in the box is the median β€” not the mean

The line drawn inside the box is the median: the middle value when the data are sorted, with half the values below it and half above it. It is not the mean, and the plot does not show the mean at all.

That matters because the mean and the median are different numbers whenever the data are skewed. The mean is pulled towards the long tail β€” a few very large values drag it up β€” while the median stays where the middle of the data is. A box plot shows the median precisely because it is not pulled around by outliers.

Where the line sits inside the box tells you about the shape of the middle half. If it sits in the middle of the box, the middle half is roughly symmetric. If it sits near the bottom edge, the middle half is bunched low with a long tail above it. If it sits near the top edge, the reverse is true.

For the 20 values in this guide the median is 22.5, and it sits in the middle of the box (17.5 to 27.5). The data are deliberately symmetric β€” the values pair up to sum to 45 β€” so the mean is 22.5 as well. In most real data the two are not equal; read the mean vs median vs mode guide to see how they differ.

Whiskers reach the farthest point inside the fences

The whiskers are the lines that run out from the box, but they do not reach the minimum and the maximum. They stop at the farthest data point that is still inside the fences.

The fences are drawn 1.5 Γ— IQR beyond the box: the lower fence is Q1 βˆ’ 1.5 Γ— IQR and the upper fence is Q3 + 1.5 Γ— IQR. Any value inside the fences is ordinary; any value outside them is an outlier, and the whisker stops at the last ordinary value.

For the 20 values in this guide the fences are 17.5 βˆ’ 15 = 2.5 and 27.5 + 15 = 42.5. Every value lies inside them, so the whiskers reach the true minimum (12) and maximum (33). When a value does fall outside, the whisker stops short β€” that is the subject of the next section.

The length of a whisker tells you about the tail. A long whisker means the values stretch a long way from the box; a short whisker means they stop close to it.

Outliers are plotted as dots

A value outside the fences is plotted as a separate dot beyond the end of the whisker. The dot is still a real data value β€” the plot flags it for a closer look, it does not delete it.

Add a value of 50 to the 20 values in this guide and it lands outside the upper fence of 42.5. It appears as a dot at 50, while the upper whisker stays at 33, the farthest value inside the fence. The dot is the plot telling you that 50 is a long way from the rest of the data.

Do not delete a dot just because it is a dot. Check first whether it is a typo, a measurement error or a genuine extreme β€” an unusually large order, a record temperature, a very high income. If it is genuine, it belongs in the data and in the story. The how to find outliers using IQR guide works through the 1.5 Γ— IQR rule step by step.

Some plots also mark points beyond 3 Γ— IQR with a larger symbol β€” the β€œfar out” values. The 1.5 Γ— IQR fence is the usual rule; 3 Γ— IQR is the stricter one for values that are extreme even among outliers.

Worked plot: reading the shared dataset

Here is the plot this guide has been building towards. The 20 values are: 12, 15, 14, 18, 16, 20, 22, 19, 17, 21, 23, 25, 24, 26, 28, 27, 29, 30, 31, 33. Sorted, they run from 12 to 33.

NumberValue
Minimum12
First quartile (Q1)17.5
Median22.5
Third quartile (Q3)27.5
Maximum33

The IQR is 27.5 βˆ’ 17.5 = 10.0. The fences are 17.5 βˆ’ 1.5 Γ— 10.0 = 2.5 and 27.5 + 1.5 Γ— 10.0 = 42.5. No value falls outside them, so there are no dots and the whiskers reach 12 and 33.

Read it in order: the box (17.5 to 27.5) holds the middle 10 values; the median line at 22.5 sits in the middle of the box; the two whiskers are the same length β€” 12 to 17.5 is 5.5 and 27.5 to 33 is 5.5. That symmetry is no accident: the values pair up to sum to 45, so the plot is balanced. Draw this plot yourself with the box plot maker.

The quartile-method trap: Tukey vs Excel

Here is the trap. Those five numbers depend on how the quartiles are calculated, and there is more than one way to do it. Tukey’s method (the exclusive hinges) and Excel’s QUARTILE.INC give different boxes from the same data.

MethodQ1MedianQ3IQRFences
Tukey17.522.527.510.02.5 to 42.5
Excel (QUARTILE.INC)17.7522.527.259.53.5 to 41.5

The median is the same under both methods (22.5), so a matching median does not mean matching quartiles. The box edges differ by 0.25 at each end, the IQR differs by 0.5, and the fences move by 1.0. On a small dataset the difference is larger β€” and it can move a value across a fence, turning an ordinary point into a dot.

Before you compare two box plots, check which method each one uses. The box plot maker draws both side by side, and the quartile calculator gives you both sets of numbers.

Comparing two groups

Box plots earn their keep when you put two side by side. Split the 20 values into two groups of 10: the first ten (12, 15, 14, 18, 16, 20, 22, 19, 17, 21) and the last ten (23, 25, 24, 26, 28, 27, 29, 30, 31, 33).

GroupQ1MedianQ3IQR
First 101517.5205
Last 102527.5305

The boxes do not overlap: the first group’s Q3 (20) sits below the second group’s Q1 (25). The medians are 10 apart (17.5 against 27.5) and the whole second group sits higher than the first. When you compare groups, look at three things: where the boxes sit (the centres), how wide they are (the spread of the middle half) and how long the whiskers are (the tails). The box plot maker draws multi-group plots, and the one-way ANOVA calculator tests whether the group means differ.

Frequently Asked Questions

Is the line in the box the mean?

No. The line inside the box is the median β€” the middle value when the data are sorted. The mean is not drawn on a box plot at all. The two are different numbers whenever the data are skewed: the mean is pulled towards the long tail, the median is not. If you need the mean, work it out with the statistics calculator β€” a box plot will not give it to you.

What do the whiskers show?

The whiskers run from the box to the farthest data point that is still inside the fences, which are Q1 βˆ’ 1.5 Γ— IQR and Q3 + 1.5 Γ— IQR. They do not reach the minimum and maximum unless every value is inside the fences. Points outside the fences are plotted as separate dots beyond the whiskers.

Why do two box plots of the same data look different?

Because the quartiles can be calculated in more than one way. Tukey’s method (exclusive hinges) and Excel’s QUARTILE.INC give different Q1 and Q3 for the same data β€” for the 20 values in this guide, Tukey gives 17.5 and 27.5 while Excel gives 17.75 and 27.25. The median is the same under both, so a matching median does not mean matching quartiles. Check which method a plot uses before you compare two plots.

Check Tools

Try the numbers from this guide with the site’s free calculators:

Box Plot Maker
Draw the plot from this guide β€” Tukey or Excel quartiles, 1.5Γ—IQR fences and multi-group comparison.
Try the box plot maker β†’

Quartile Calculator
Find quartiles and the IQR β€” the numbers behind the box edges.
Try the quartile calculator β†’

Stem-and-Leaf Plot Maker
See the same five numbers as a stem-and-leaf plot β€” every value kept, sorted into stems.
Try the stem-and-leaf plot maker β†’

Explore All Statistical Tools
See every calculator on the site in one place.
Browse all our statistical tools β†’

Check Articles

Go deeper with these guides:

How to Find Outliers Using IQR
The 1.5Γ—IQR rule step by step, with the fences worked out on real data.
Read the guide β†’

Mean vs Median vs Mode
Mean, median and mode are three different answers to ‘what is typical?’ β€” and the trimmed mean sits between them.
Read the guide β†’

What Is Skewness and Kurtosis
How a long tail pulls the mean away from the median β€” and what the box plot shows instead.
Read the guide β†’