What a histogram shows that a list cannot
A histogram groups values into equal-width bins and draws a bar per bin, height equal to the count. The bars touch, because the axis is a number line with no gaps. What you get is the shape of the data at a glance: where the values pile up, how they spread, whether they lean left or right, whether they split into two humps, and whether a few values sit far away from the rest. Those four questions (centre, spread, shape, outliers) are the first examination any dataset deserves, and the histogram answers three of them by itself.
Twenty commute times in minutes: 24, 28, 31, 22, 35, 27, 29, 26, 33, 25, 30, 27, 24, 38, 29, 26, 31, 28, 23, 27. Mean 28.15, median 27.5, standard deviation 4.04, range 22 to 38. With 8 bins the histogram shows a main pile from 24 to 33 and a thinning right tail toward 38. Mean and median sit close (28.15 vs 27.5), matching the picture: a roughly symmetric heap with a slight right lean.
The bin count is an editorial decision
Too few bins and every dataset looks like a smooth mound: bimodality, gaps and spikes get averaged away. Too many and the chart fragments into noise, one or two counts per bar, and you start seeing patterns that are not there. Start near the square root of the count (√20 ≈ 4 or 5 here is on the low side; 8 to 10 reads better), then try two more settings and keep the one where the shape holds still. A shape that survives several bin counts is a property of the data. A shape that appears at exactly one setting is a property of the setting.
Confusing a histogram with a bar chart. Bars here stand for ranges of a continuous variable and must touch; gaps between histogram bars would claim that some values are impossible. Bar charts compare separate categories and keep their gaps. Different data, different grammar.
Reading shapes
Symmetric: mean and median agree, and normal-based methods are on the table. Right-skewed: a long tail of large values (incomes, house prices); the mean gets pulled toward the tail, so quote the median alongside. Bimodal: two humps usually means two populations were mixed (heights of a mixed-sex sample, times from two different machines). Split the data by the suspect variable before summarizing.
Frequently asked questions
How many bins should I use?
Start near the square root of your count, then try a couple of settings above and below. Keep the count where the main shape stops changing. The tool lets you change it live for exactly this reason.
What is the difference between a histogram and a bar chart?
A histogram bins a continuous variable into ranges, and its bars touch. A bar chart compares distinct categories, and its bars stand apart. Histograms show distribution shape; bar charts compare amounts.
Why does my histogram change when I change the bins?
Binning is a summary; every summary discards something. If the story flips between settings, the dataset is small or the pattern is fragile, and both facts are worth knowing before you publish the chart.
Can a histogram show two variables?
Not directly. Use a scatter plot for pairs, or draw two histograms side by side with the same bins and axis to compare groups fairly.