Short answer
A box plot summarizes a distribution: the line inside the box is the median, the box spans the middle 50 percent of values from the first to the third quartile, the whiskers reach the most extreme values within 1.5 times the box height, and dots beyond them are flagged as outliers.
Key takeaways
- The box holds the middle half of the data; its length is the interquartile range (IQR).
- The line in the box is the median, not the mean.
- In the common Tukey convention, whiskers stop at the last value within 1.5 × IQR of the box.
- Box plots are best for comparing several groups side by side.
- They hide the shape inside the box, so check a histogram when the shape matters.
A box plot squeezes a whole distribution into a few marks. Once you know what each mark means, it's one of the fastest ways to compare groups: delivery times by warehouse, test scores by school, response times by server. The statistician John Tukey popularized the form in the 1970s, and most chart tools still draw it his way.
Here's how to read one, part by part, and when it's the wrong chart.
What does each part of a box plot show?
The box is the middle half of the data, the line in it is the median, and the whiskers and dots show how far the rest spreads.
Imagine 21 delivery times in minutes, sorted from fastest to slowest. A box plot of them has five parts:
| Part | What it marks | In plain words |
|---|---|---|
| Line inside the box | Median (50th percentile) | Half the deliveries were faster than this |
| Bottom of the box | First quartile, Q1 (25th percentile) | A quarter were faster than this |
| Top of the box | Third quartile, Q3 (75th percentile) | Three quarters were faster than this |
| Whiskers | Most extreme values within 1.5 × IQR of the box | The normal range of the rest |
| Dots beyond the whiskers | Values further out | Unusual deliveries worth checking |
The height of the box, Q3 minus Q1, is called the interquartile range or IQR. It's a measure of spread that ignores the extremes, which makes it more stable than the full range.
How are the whiskers calculated?
In the usual convention, each whisker runs to the furthest data point that is still within 1.5 times the IQR from the box.
Say Q1 is 24 minutes and Q3 is 36. The IQR is 12, and 1.5 × 12 is 18. So the upper fence sits at 36 + 18 = 54 minutes, and the lower fence at 24 − 18 = 6. The top whisker ends at the slowest delivery that's still 54 minutes or less. A delivery that took 70 minutes is drawn as a dot.
Not every tool does it this way. Some draw whiskers to the minimum and maximum, and some to fixed percentiles such as the 5th and 95th. Check the documentation or add a note to the chart, because readers will assume the 1.5 × IQR rule.
How do you read the shape from a box plot?
Look at where the median sits inside the box and how long each whisker is.
A median in the middle of the box with whiskers of similar length suggests a fairly symmetric distribution. A median pushed toward the bottom of the box with a long upper whisker suggests right skew: most values are low, with a tail of high ones. Delivery times, incomes and website load times usually look like this.
What a box plot can't tell you is what happens inside the box. Two peaks, a gap or a cluster all disappear into the same rectangle. That's its main weakness.
When is a box plot the right choice?
When you need to compare the distribution of one measure across several groups.
Box plots shine side by side. Eight warehouses, eight boxes on one axis, and you can see at a glance which one is slowest on a typical day (median), which is least predictable (tallest box) and which has the most extreme delays (dots). Doing the same with eight histograms takes a full page.
They're less useful when:
- You have one group. A histogram shows its shape in more detail.
- You have very few values. With eight data points, quartiles are shaky; plot the points themselves.
- Your audience hasn't seen box plots before. General readers often don't know what the box means, so a dot plot or a simple range chart with a labeled median is easier.
How can you make box plots easier to read?
Add the raw points, sort the groups, and label the median.
A few changes help almost every box plot:
- Overlay the data points as faint dots, slightly jittered sideways, if there are fewer than a few hundred per group. Readers can then see clusters the box would hide.
- Sort groups by median so the order carries meaning, the same advice that applies to bar charts.
- Label the median of each box, or at least the one your story is about.
- Show the count for each group. A box built from 12 values deserves less trust than one built from 1,200.
- Say how the whiskers were drawn in a short note, especially if your tool doesn't use 1.5 × IQR.
If you're deciding between a box plot, a histogram and a scatter plot, our guide to choosing a chart walks through the questions that point to each.
Do box plots work for engagement or analytics data?
Yes, and they handle skewed metrics like post impressions better than averages do.
Social and web analytics are heavily skewed: most posts get modest numbers and a few get huge ones. An average gets dragged up by those few, while a box plot shows the typical post (median), the usual range (box) and the outliers (dots) separately. Comparing post formats as box plots, one box per format, is a quick way to see whether one format really does better or just had one lucky hit.
Questions
What does the line in the middle of a box plot mean?
It's the median: half the values are below it and half above. It is not the average. If the median sits near one end of the box, the middle half of the data is skewed toward that side.
What do the dots outside a box plot mean?
They are values more than 1.5 times the interquartile range beyond the box, under the most common convention. They're flagged as possible outliers, which means worth a look, not automatically wrong.
Is a box plot the same as a box-and-whisker plot?
Yes. Box plot and box-and-whisker plot are two names for the same chart.
When should I use a histogram instead of a box plot?
Use a histogram when you have one group and the shape matters, for example whether there are two peaks. Use box plots when you're comparing several groups and need a compact summary of each.