Side-By-Side Boxplot vs Comparative Dotplot

A side-by-side boxplot puts one boxplot per group along the same axis, so a difference in center reads as a shift and a difference in spread reads as a wider or narrower box. For example, if class A has median 78 with $Q_1 = 70$ (the first quartile) and $Q_3 = 84$ (the third quartile), while class B has median 85 with $Q_1 = 81$ and $Q_3 = 89$, then B sits 7 points higher with a smaller interquartile range (IQR) of $89 - 81 = 8$ against A's $84 - 70 = 14$. This site computes quartiles with the median-excluded (TI-84) rule. Because each box is built from only five numbers, a side-by-side boxplot can hide a gap, a cluster, or a second peak that a dotplot would show.

Because every dot is still one observation, a comparative dotplot compares groups without summarizing anything away, so gaps, clusters, and repeated values stay visible. For example, group A with values 4, 5, 5, 6 and group B with values 3, 5, 5, 7 share the same mean of $20 / 4 = 5$, but B has a range of $7 - 3 = 4$ against A's $6 - 4 = 2$. A side-by-side boxplot would also show that difference in spread, but it would hide the fact that both groups pile up at 5. Comparative dotplots work best for small data sets, since with hundreds of points the stacks grow too tall to read.