- A box and whisker plot summarizes a numeric distribution through its median, quartiles, whiskers, and any separately plotted potential outliers.
- Read the median first, then compare the box widths, whisker lengths, and flagged points across groups that share the same scale.
- The box contains the middle 50% of values, so its length shows how tightly or widely that central half is spread.
- Before comparing calculations, check how the chart defines quartiles and whiskers. Do not assume its whiskers represent the raw minimum and maximum.
A long list of numbers can hide the pattern you need. A box and whisker plot shows where values cluster, how far they spread, and which observations sit far from the rest.
You can use this chart to compare order values across regions, response times across support teams, or test scores across classes. Once you know what each line means, a box plot becomes a quick way to spot differences worth a closer look.
What is a box and whisker plot?
A box and whisker plot, also called a box plot or boxplot, summarizes a numeric distribution with five reference values. These are the minimum, first quartile, median, third quartile, and maximum. If the chart flags potential outliers, check its legend to identify what the whisker endpoints represent.
The CDC box plot guidance describes the chart as a graphic representation of the five-number summary that shows numeric data distribution. This compact format helps you see center and spread without reading every observation.
Why does that matter? Suppose you have delivery times across four warehouses, where averages may hide uneven performance. Side-by-side box plots reveal differences in typical times, variation, and unusually long delays.
A useful chart should also prompt your next question. If one warehouse has a much wider box, you might ask whether carrier mix, staffing, distance, or order type explains that variation.
Why box plots are useful for comparing distributions
Box plots work especially well when you align several groups on one numeric scale. Your eye can compare four features quickly:
Median: Which group has the higher or lower typical value?
Middle spread: Which group has the wider interquartile range?
Whiskers: How do their lengths and endpoints compare across groups?
Potential outliers: Which group contains observations that deserve investigation?
NIST calls box plots an excellent tool for conveying location and variation information, especially between groups, in its NIST box plot guidance. That is why the chart is useful for business comparisons such as order value by sales region or ticket resolution time by support team.
Imagine that two support teams share the same median resolution time. If Team A has a narrow box while Team B has a wide one, Team B shows more variation. Next, examine ticket type, severity, staffing, or escalation patterns rather than stop at the shared median.
How to read the five-number summary in a box plot
Start by finding the line inside the box. That line marks the median, which splits the ordered values into two equal halves. Next, look at the box edges, then the whiskers, and finally any points drawn beyond them.
Statistics Canada says a box plot visualizes the five-number summary and helps show center, spread, skewness, and potentially unusual observations in its Statistics Canada explanation.
Before you interpret a chart, check its legend or documentation. Confirm what the whisker endpoints represent and how the chart calculates its quartiles.
Minimum, Q1, median, Q3, and maximum
To understand the five-number summary, sort the data from smallest to largest and identify these points:
Minimum: The smallest value in the full data set.
Q1: The lower quartile, with about 25% of values at or below it.
Median: The middle value by rank, also called Q2.
Q3: The upper quartile, with about 75% of values at or below it.
Maximum: The largest value in the full data set.
The box stretches from Q1 to Q3, so it contains the middle 50% of observations. A longer box means that central half has greater variation. Meanwhile, the median line shows where the center falls inside that range.
If the median sits close to one edge of the box, treat its placement as a prompt to inspect the underlying data. Ask which observations or group characteristics could explain what you see.
Before treating a whisker endpoint as the raw minimum or maximum, confirm the chart's stated method. When you work with interactive data visualizations, review the settings so you know what each endpoint represents.
IQR, whiskers, and potential outliers
The interquartile range, or IQR, measures the spread of the middle half of your data:
`IQR = Q3 - Q1`
A common convention uses the IQR to calculate two inner fences. NIST defines IQR as Q3 minus Q1 and lists the formulas in its NIST outlier fences:
`Lower fence = Q1 - 1.5 × IQR`
`Upper fence = Q3 + 1.5 × IQR`
For the 1.5 × IQR method used here, values below the lower fence or above the upper fence are potential outliers. Draw each whisker to the most extreme observed value inside its fence. A fence is a calculated boundary, while a whisker ends at an observed value.
Treat an outlier flag as a question, not a deletion instruction. A large transaction could be a data-entry error, a valid order, or a sign of fraud.
Subject knowledge determines whether it needs correction, separate analysis, or no action. Some organizations use business rules instead of the 1.5 × IQR convention.
How to make a box and whisker plot step by step
You can build a box plot by hand with a short data set. For larger sets, spreadsheet or analytics software handles the arithmetic, but knowing the steps helps you check the output.
Use this process:
Sort the values from smallest to largest.
Find the median of the full set.
Split the data into lower and upper halves, excluding the median when the set has an odd number of values.
Find the median of the lower half to get Q1.
Find the median of the upper half to get Q3.
Subtract Q1 from Q3 to calculate the IQR.
Calculate the lower and upper fences with the 1.5 × IQR rule.
Draw the box from Q1 to Q3 and add a line at the median.
Extend whiskers to the most extreme observed values inside the fences.
Plot observations beyond the fences as separate points.
If you analyze spreadsheet data, record the quartile method used in your hand calculation. Check the tool's documentation before comparing its output with your result.
A simple box plot example
Consider this sorted data set:
`3, 5, 7, 10, 12, 14, 18`
There are seven values, so the fourth value is the median: 10. Excluding that median, the lower half is `3, 5, 7`, whose middle value gives Q1 = 5. The upper half is `12, 14, 18`, whose middle value gives Q3 = 14.
The five-number summary is:
Minimum = 3
Q1 = 5
Median = 10
Q3 = 14
Maximum = 18
Now calculate the interquartile range:
`IQR = 14 - 5 = 9`
Next, calculate the fences:
`Lower fence = 5 - (1.5 × 9) = -8.5`
`Upper fence = 14 + (1.5 × 9) = 27.5`
Every observed value falls between -8.5 and 27.5, so there are no potential outliers. The whiskers can extend to 3 and 18 because both values fall within the fences.
This hand calculation uses the median-of-halves convention and excludes the overall median from each half. State that method when you need someone else to reproduce the result.
How to interpret box and whisker plots across groups
When several box plots share a scale, use the same reading order every time. A consistent sequence keeps one dramatic whisker or outlier from distracting you from the full comparison.
Compare medians. Which group has the higher or lower center?
Compare IQRs. Which box is longer, and therefore more variable in its middle 50%?
Inspect whiskers. Are the tails similar, or does one side extend farther?
Review potential outliers. Which observations merit a follow-up question?
Suppose you compare monthly order values across three sales regions. The West has the highest median, the East has the narrowest IQR, and the Central region has several high potential outliers. You can describe those visible differences directly, but the chart alone does not prove statistical significance or explain the cause.
Whisker and median placement can guide your next question. If one whisker extends farther or the median sits off-center, inspect the underlying observations before drawing a conclusion.
Ask a follow-up question after the chart reveals a pattern
A distribution chart starts the analysis. Once you notice a wider IQR, an off-center median, or a potential outlier, ask what segment, driver, or recent change could explain it.
For example, if one support team has a wider spread in resolution time, break the data down by issue type, customer tier, or escalation status. If a region has unusually large orders, check whether a few account types or products account for the difference.
ThoughtSpot product material says SpotIQ uses automated insight detection algorithms to analyze large data volumes. That positioning supports a useful habit: move from seeing a pattern to testing an explanation.
After a chart reveals a visible difference, you can ask follow-up data questions about the segment or driver behind it. This keeps the focus on investigating the pattern rather than stopping at the summary.
When to use a box plot, and when to choose another chart
Choose a box plot when your main goal is to compare numeric distributions across meaningful groups. It is particularly useful when you need a compact view of center, middle spread, tail length, and potential outliers.
Good uses include comparing transaction values by region, wait times by clinic, test scores by class, or delivery delays by carrier. Box plots also fit reports where space is limited because each group takes little room.
Choose another chart when your question changes:
Histogram: Use it to see distribution shape, gaps, and multiple peaks within one group.
Density plot: Use it to compare smoothed distribution shapes, especially with larger samples.
Scatter plot: Use it to examine the relationship between two numeric variables and retain individual points.
Line chart: Use it when change over time is the main question.
A box plot is a compact summary, so it may not show every individual observation. Pair it with the raw data when your question depends on specific values.
When precision matters, label the method used for the chart and compare groups calculated with the same stated rules. In governed analytics work, reusable analytics logic can help keep repeated definitions consistent across analyses.
Conclusion
To read a box and whisker plot, start with the median and check the box for the middle 50%. Then inspect whiskers and potential outliers before comparing groups on one scale.
The chart becomes more useful when you ask what caused the pattern. To match more visuals to your questions, explore data visualization and start sharper follow-up analysis.
