Box Plot Maker
Compare groups by median, spread and outliers. One column per group, then export a PNG.
About this tool
A box plot compresses a whole group of numbers into five: the minimum, the first quartile, the median, the third quartile and the maximum. That is enough to compare several groups at a glance — which one sits higher, which one is more consistent, which one has a straggler pulling its average around. Response times across servers, scores across classes, yields across suppliers all fit this shape.
The reason to reach for it instead of a bar chart of averages is that averages lie by omission. Two teams can average the same number while one is tightly clustered and the other swings between extremes, and a bar chart draws them identically. The box makes that difference impossible to miss, and the outlier dots point straight at the values worth investigating.
Quartiles are computed with linear interpolation, the same method Excel's PERCENTILE and NumPy use, so the numbers here match what your spreadsheet reports. Paste columns straight from a sheet — tabs, commas and aligned spaces all parse, and the header row becomes the group labels under each box.
Frequently asked questions
How do I read a box plot?
The box spans the middle half of the data: its bottom is the first quartile, its top the third, and the line inside is the median — not the mean. The whiskers reach out to the furthest values still considered ordinary, and anything beyond them is drawn as a separate dot. So the box tells you where the bulk sits, the whisker length tells you how far the tails run, and the dots tell you which values stand apart.
How are outliers decided?
By the standard 1.5×IQR rule. The interquartile range is the height of the box, and any value more than 1.5 times that distance beyond either quartile is plotted as an individual point. When no value crosses that line the whiskers stop at the actual minimum and maximum, so a plot with no dots means every value was within the usual range — not that outliers were hidden.
What layout should the data be in?
One column per group, with the group name in the header row. The columns do not need to be the same length; shorter ones simply have fewer values, and blank cells are skipped rather than counted as zero. That matters when comparing an experiment with 40 samples against a control with 25.
When is a box plot better than a histogram?
Use a histogram when you have one set of numbers and want to see its exact shape, including whether it has two peaks. Use a box plot when you are comparing several groups: five boxes side by side make differences in level and spread obvious in a way five stacked histograms never do. The trade-off is that a box plot smooths the shape away — two very different distributions can produce nearly identical boxes.
Does my data leave the browser?
No. Quartiles, whiskers and outliers are all computed on this page, and nothing is sent anywhere. Salary bands, lab measurements and patient readings can be compared here without handing them to a server.