Box plots squeeze a whole distribution into a small shape, so many groups fit side by side.
They work best for comparison, not for studying one distribution in detail.
Every part of the plot is drawn from the quartiles and the median.
The bottom of the box is the first quartile, the top is the third quartile, and the line inside is the median.
The box height is the IQR, so a tall box means the middle half of the values is widely spread.
The whiskers usually stop at the last value inside the IQR fences, 1.5 IQRs beyond the box.
Values beyond the whiskers are drawn as single points, which is the IQR outlier rule made visible.
A median line near one end of the box, or one whisker much longer than the other, shows skew.
Box plots of a dozen groups side by side show differences in typical level and spread at once.
Because only quartiles are drawn, 2 very different shapes can produce the same box.
A distribution with 2 peaks can look like an ordinary box, hiding its most interesting feature.
Box plots also hide how many values each group has, so a box from 5 values looks as solid as one from 5,000.
Overlaying the individual points, or printing the count beside each box, covers both weaknesses.
In practice box plots suit reports that compare many teams, products or cities.
The usual misreading is taking the whisker ends as the minimum and maximum, when points beyond them are the extremes.