Excel Notched Box Plot: What It Shows and How to Make One

Excel for Microsoft 365 can make a standard Box & Whisker chart, but it does not have a built-in setting to add notches. To show a notched box plot, calculate the notch limits from each group’s median, quartiles, and sample size, then present those limits alongside the Excel chart or build a custom plot; the notch is an approximate interval around the median, not a different quartile calculation.

What a notched box plot shows

A box plot summarizes a numeric sample with its median, lower and upper quartiles, and whiskers. A notched box plot adds a narrowing around the median. The notch represents an approximate confidence interval for the population median, commonly calculated as the sample median plus or minus 1.57 times the interquartile range divided by the square root of the sample size. The interquartile range, or IQR, is Q3 minus Q1.

The visual purpose is comparison. If notches for two groups do not overlap, that is often treated as evidence that their medians differ. It is a useful visual convention, but not a substitute for a formal statistical test: the interval is approximate, depends on the data and sample size, and its interpretation can be unreliable for small or unusual samples. Overlapping notches do not prove that medians are equal.

A notch is also not the same thing as the box itself. The box still runs from Q1 to Q3, and the horizontal line through it still marks the median. The narrowed section simply draws attention to uncertainty around that median. Some notch limits can fall outside the box, especially with small samples or particular distributions; that is a limitation of the approximation, not automatically a spreadsheet error.

Make a standard box plot in Excel

For a standard chart, arrange each group in its own column, with a heading at the top and observations underneath. For example, put one product’s measurements in A2:A21 and a second product’s measurements in B2:B21, with headers in row 1. Each column becomes a box in the chart. Keep the underlying values as numbers and avoid mixing totals or summary statistics into the sample data.

Select the data, including the group headings.
Choose Insert > Insert Statistic Chart > Box and Whisker.
Select the chart, open Format Data Series, and review Series Options. Depending on your needs, adjust the quartile calculation and choose whether to show outlier points, inner points, mean markers, or a mean line.
Add a chart title and axis titles, and check that the category labels identify the groups clearly.

Excel’s whiskers extend to the most extreme observations that remain within 1.5 times the IQR from the quartiles; values beyond that range are displayed as outliers when outlier points are shown. The whiskers therefore do not necessarily reach the sample minimum and maximum. If you need the raw extremes, show the points or calculate those values separately.

The desktop ribbon path above is for Excel for Microsoft 365 on Windows. On a Mac, the chart type is also available from Insert, though the formatting pane and labels can be arranged differently. Excel for the web may display charts created in desktop Excel, but chart creation and formatting options can differ; use the Windows or Mac desktop app if the Box & Whisker chart command is unavailable.

Calculate notch limits with worksheet formulas

Excel does not calculate or draw a notch automatically, but worksheet formulas can produce the two limits. Consider this example sample in A2:A11: 10, 12, 13, 14, 15, 16, 18, 20, 22, and 30. Use the inclusive quartile convention consistently if you compare these calculations with a chart set to use inclusive quartiles.

With the sample values in A2:A11, enter this in B2 to calculate Q1. It returns 13.25.

=QUARTILE.INC(A2:A11,1)

With the sample values in A2:A11, enter this in B3 to calculate the median. It returns 15.5.

=MEDIAN(A2:A11)

With the sample values in A2:A11, enter this in B4 to calculate Q3. It returns 19.5.

=QUARTILE.INC(A2:A11,3)

With the sample values in A2:A11, enter this in B5 to count the observations. It returns 10.

=COUNT(A2:A11)

With Q1 in B2, the median in B3, Q3 in B4, and the sample size in B5, enter this in B6 for the lower notch limit. It returns about 12.397.

=B3-1.57*(B4-B2)/SQRT(B5)

With the same cell layout, enter this in B7 for the upper notch limit. It returns about 18.603.

=B3+1.57*(B4-B2)/SQRT(B5)

For this example, the IQR is 19.5 − 13.25, or 6.25. The notch half-width is approximately 1.57 × 6.25 ÷ √10, or 3.103. Adding and subtracting that from 15.5 gives limits of about 18.603 and 12.397. These results are approximate because the constant and formula describe a statistical convention; do not round intermediate values if you want the displayed limits to be as precise as possible.

Ways to show the result beside the chart

The simplest reliable option is to keep the Excel box plot and add the calculated notch limits in a small table beside it. Include the group name, median, lower limit, and upper limit. This preserves Excel’s automatic box plot and makes the interval values explicit, without suggesting that the chart itself contains notches. If you compare several groups, calculate the limits separately for each column and keep the quartile method consistent.

You can also annotate the chart with text or lines, but take care with scale and alignment. A manually positioned line may not stay aligned with a data value if the axis range, chart size, or data changes. If you add annotations, label them clearly as approximate median intervals and recheck their positions after changing the chart. A table is often easier to maintain and less likely to mislead.

For a genuinely notched silhouette, a custom XY scatter chart is possible, but it takes more preparation. Calculate each group’s quartiles, median, notch limits, and whisker endpoints; assign each group a numeric horizontal center; then provide ordered X-Y coordinates for the box outline, including inward points at the notch limits and a center point at the median. Plot those coordinates as straight-line scatter series, and add separate series for whiskers, caps, and outliers. This is a hand-built chart, not an Excel Box & Whisker chart, so it will not update its statistical structure automatically when data changes. Test the axis scaling and labels carefully.

Choosing quartiles and reading outliers

Quartile conventions matter because different methods can produce different Q1 and Q3 values, particularly in small samples. Excel’s Box & Whisker chart offers inclusive and exclusive median calculations in its series options. The formulas above use QUARTILE.INC, so select the corresponding inclusive convention when comparing the chart’s box with the worksheet calculations. If you use another statistical package or published results, check its quartile definition before treating small differences as mistakes.

Outliers in a box plot are observations outside the 1.5-IQR fences, not necessarily bad data. They may be valid rare values, recording errors, or results from a different process. Investigate their source before excluding them. Also remember that the notch formula uses sample size: missing or blank cells are not counted by COUNT, while a numeric zero is counted. Use the number of valid observations for the group rather than a total row count that includes headings or empty cells.

Common problems and practical checks

No Box & Whisker command: Confirm you are using a desktop Excel version that supports the chart type and have selected numeric observations arranged by group. In older releases, the chart type may not be available.
Groups appear in the wrong direction: Put each group in a separate column, or use Switch Row/Column on the Chart Design tab if Excel interprets the rows as categories.
Quartiles do not match a manual calculation: Check whether the chart uses inclusive or exclusive quartiles, and use the same convention in the formulas.
Whiskers seem to omit minimum or maximum values: They generally end at the most extreme values inside the 1.5-IQR fences; outside values are plotted separately as outliers.
Notch limits look unusually wide: Review the sample size, quartiles, and data values. The approximation can be unstable for small samples, and limits can extend beyond the box.

Before interpreting a comparison, check group sizes and distributions. A large difference in sample size, strong skew, or many tied values can make a simple visual comparison harder to interpret. When the conclusion matters, use an appropriate statistical method for comparing medians rather than relying only on whether the notches overlap.

Excel charts and other options compared

Excel’s native chart is the best choice when you want a quick, refreshable summary with quartiles, whiskers, and optional outlier or mean markers. It is easier to update than a hand-built notched graphic, but it does not provide notches. Adding calculated limits in a companion table offers a practical compromise: the chart remains native, and the uncertainty information is visible without pretending Excel drew a notch.

A custom XY plot gives you more control over a notched outline but requires coordinate tables and maintenance. Statistical software can be preferable when you need notches for many groups, reproducible analysis, or more control over statistical assumptions. For instance, R’s ggplot2 boxplot geometry has a notch option. Regardless of the tool, report the sample size and explain what interval the notch represents so readers do not confuse it with the quartile box or whisker range.

Frequently Asked Questions

Does Excel have a notched box plot chart?
Not as a built-in option in Excel for Microsoft 365. Excel can create a standard Box & Whisker chart, and you can calculate notch limits separately or construct a custom XY chart.

What do the notches mean on a box plot?
They usually show an approximate confidence interval around the median, commonly using median ± 1.57 × IQR ÷ √n. The notches are intended to help compare medians, not to show the range of the data.

Do non-overlapping notches prove that medians differ?
No. Non-overlap is a visual indication often used to suggest a difference, but the interval is approximate. Use a suitable statistical test when you need a formal conclusion; overlap alone does not establish equality.

Why does Excel’s box plot not reach the minimum and maximum?
The whiskers generally reach the most extreme values within 1.5 times the IQR from the quartiles. More extreme observations are shown as outlier points when that option is selected.

Can I use the notch formula for every sample size?
You can calculate it, but the approximation can be misleading for small samples or unusual distributions. Report the sample size, inspect the data, and avoid treating the notch as a guaranteed confidence interval in every situation.

Try it: the Date Difference Calculator: days, months, years and weekdays between two dates, with the DATEDIF formulas.

More in How-to basics

Sources: Function behaviour and menu paths: Microsoft Support documentation for Excel. Method & sources

Spotted an error in this guide? Let us know.