How to Read a Box Plot and Understand Its Data

A box plot, also called a box-and-whisker plot, compresses a set of numerical data into a few landmarks. It helps you see a typical value, how tightly values cluster, and whether unusually distant observations appear. It does not show every data point or reveal the cause of a pattern.

Before you start, identify what is being measured and check the axis labels and units. If several plots appear together, note which category belongs to each one. Box plots can follow different conventions, especially for whiskers and outliers, so use a caption or legend when available.

What You’ll Need

A box plot to read, including its axis labels, units, and any legend or caption.
If you want to verify or build the plot: the underlying numerical data and the plotting method or software used.

Step-by-Step

Read the scale. Find the axis with the numerical values. Check its units and tick marks, and see whether it starts at zero or uses a narrowed range. Estimate positions between ticks carefully; do not assume equal-looking distances represent equal values if the scale is not linear. The scale tells you how to translate the plot’s parts into actual numbers.
Identify the group and variable. Read the title, axis labels, and legend to learn what each plot represents. For example, a plot might show the time taken by one group or compare times across several groups. Confirm whether larger values mean more, longer, or a higher measurement; the plot’s shape alone does not establish what is being measured.
Locate the median. Find the line inside the box. It marks the median: half the observations are at or below it and half are at or above it, allowing for tied values. Estimate its value from the axis. The median describes the center by position, not necessarily the arithmetic average, and it may not be exactly centered in the box.
Read the box edges. The lower edge usually marks the first quartile (Q1), and the upper edge usually marks the third quartile (Q3). About one-quarter of observations lie at or below Q1, and about three-quarters lie at or below Q3. The box contains the middle half of the observations. Read both edges from the scale rather than judging their values by appearance alone.
Compare the box’s height or length. Subtract Q1 from Q3 to get the interquartile range (IQR), the spread of the middle 50% of values. A larger IQR means that middle portion is more spread out; a smaller one means it is more tightly grouped. Compare this spread only for plots measured in compatible units and shown on the same scale.
Check the whiskers. Whiskers extend from the box toward lower and higher values, but their exact meaning depends on the plotting convention. They may reach the minimum and maximum, or the most extreme values within a rule such as 1.5 IQR from a quartile. Check the key or caption before calling whisker endpoints the overall minimum and maximum.
Inspect separate points. Dots beyond the whiskers are often flagged as potential outliers under the plot’s chosen rule. Read their approximate values and note which side they fall on. A flagged point is not automatically an error or an unusual event; it is a value worth checking against the data and its context.
Describe the overall pattern. State the approximate median, the middle-half range, and whether one side is more spread out or has distant points. A longer upper whisker or a median off-center in the box can suggest asymmetry, but the plot alone does not prove a cause. For a comparison, check each group on the same scale and mention overlap as well as differences.

Common Mistakes

Assuming whiskers always show the minimum and maximum. Some plots stop at the most extreme non-outlier values. Confirm the convention before describing the full range.
Calling the median the average. The median is the middle position in ordered data. The mean is calculated by adding values and dividing by their count; a box plot generally does not display it unless marked separately.
Reading the box as a range containing all values. The box covers only the middle 50%. Values can lie below or above it, including any points drawn beyond the whiskers.
Assuming a longer box means more observations. Box length indicates spread, not sample size. A plot may show sample size separately, but the box itself does not tell you how many observations were collected.
Treating apparent overlap as proof of no difference. Overlapping boxes do not, by themselves, establish whether groups differ meaningfully or statistically. Use suitable analysis and context for that question.
Comparing plots with different scales. Different axis ranges can make similar spreads look unlike one another. Check tick marks, units, and axis limits before comparing.

Tips

Say “about” when estimating values from a graphic; exact quartiles may not align with labeled ticks.
When comparing groups, read medians first, then compare IQRs, whiskers, and outliers. This keeps center and spread distinct.
Remember that quartile calculation methods can differ slightly, especially in small data sets. If exact values matter, consult the source data or plotting notes.
Use the box plot as a summary, not a replacement for the data. A histogram, dot plot, or table can reveal clusters, gaps, and repeated values hidden by the summary.
Describe what the graphic shows without claiming why it happened. Other information is needed to explain causes or judge whether a pattern matters.

Troubleshooting

The plot has no visible median line: It may overlap a box edge when quartiles or values coincide, or the line may be hard to see. Check the legend and, if possible, the source data or an accessible version of the chart.

The whiskers seem inconsistent: Look for a note defining their rule and how outliers are marked. If no explanation is provided, report what is visible without asserting that the whiskers represent the minimum and maximum.

You cannot estimate a value confidently: Find the nearest labeled ticks and estimate only if the axis spacing allows it. If the scale is truncated, logarithmic, or unlabeled, consult the chart notes before interpreting distances.

Two plots seem hard to compare: Verify they use the same units and scale, and check whether they summarize comparable data. If sample sizes or distributions matter, seek those details or inspect the underlying observations.

Frequently Asked Questions

What does the line inside a box plot show?

It usually marks the median, the midpoint of the ordered observations. A separate symbol or line may represent the mean if the chart says so.

What does the box represent?

The box runs from the first quartile to the third quartile and contains the middle 50% of values. Its length shows the interquartile range.

What do dots beyond the whiskers mean?

They are commonly flagged as potential outliers according to the chart’s rule. They are still observations and should not be discarded without checking the data and context.

Can a box plot show the exact data values?

Usually not. It summarizes the distribution, so many different data sets can produce similar box plots. Separate points may show some individual values, but not necessarily all of them.

How can I tell if data are skewed?

An off-center median and unequal whisker lengths can suggest asymmetry. Treat this as a clue rather than a precise measure; a histogram or the raw data can show the distribution more fully.

Further reading: NIST/SEMATECH e-Handbook of Statistical Methods. Details can change, so check the official page for anything that matters to you. Spotted a mistake? Tell us. How we write an answer.