How to Make a Histogram (and How It Differs from a Bar Graph)
Written by Chart Maker Hub · Published · 3 min read
A histogram answers one question: how are my values spread out? It groups numbers into equal-width ranges, called bins, and draws a bar for how many values land in each range. This guide shows how to make one with the free histogram maker and how to read it.
Histogram vs bar graph: what’s the difference?
They look alike, but they answer different questions.
| Histogram | Bar graph | |
|---|---|---|
| Data you enter | Raw numbers, one per observation (40 test scores) | One value per category (sales per region) |
| X-axis | A continuous number scale split into bins | Separate categories |
| Bars | Touch, because the bins are adjacent ranges | Have gaps, because categories are distinct |
| Bar order | Fixed by the number scale | Can be sorted, for example largest first |
| Question | How are the values distributed? | Which category is bigger? |
If you already have one total per category, you want a bar graph. If you have a list of individual measurements, you want a histogram.
1. Collect one value per observation
Enter or paste raw values in a single column: one exam score, delivery time or height per row. Don’t pre-count them; the tool does the counting. You can paste thousands of values from a spreadsheet column.
2. Choose the bins
The bin width changes the picture more than any other setting.
- Automatic (the default) uses the Freedman–Diaconis rule, which sets the bin width from the interquartile range and the number of values, and falls back to Sturges’ rule when that isn’t possible. It is a sensible starting point.
- Number of bins lets you ask for roughly k bins; the width is rounded to a tidy number.
- Bin width lets you set meaningful ranges, such as 10-point score bands or 5-minute time bands, and choose where the first bin starts.
Try two or three widths before deciding. Too few bins hide the shape; too many turn it into noise.
3. Read the shape
- Symmetric: values cluster in the middle and tail off evenly on both sides.
- Right-skewed: a long tail toward high values, common for incomes, prices and waiting times. The mean is pulled above the median.
- Left-skewed: a long tail toward low values, such as scores on an easy test.
- Two peaks (bimodal): often a sign that two different groups are mixed together, such as weekday and weekend orders.
- Gaps and isolated bars: possible outliers or data-entry errors worth checking.
Turn on Mark the mean to see where the average falls relative to the peak.
4. Counts or percentages?
Show counts when the sample size matters to your reader. Switch Bar height shows to Percent of observations when you compare histograms of groups with different sizes, so the shapes are on the same scale.
5. Common histogram mistakes
- Unequal bin widths. Bar height then no longer reflects density, and readers draw the wrong conclusion. Keep bins equal.
- Gaps between bars. Gaps suggest separate categories. Histogram bars should touch.
- Too few observations. With fewer than about 20 values, the shape depends heavily on the bin choice. A box plot or a simple dot list may be more honest.
- Comparing many groups with overlapping histograms. Use side-by-side box plots instead.