Peter Kampstra, writing in the Journal of Statistical Software, has demonstrated a very cool way of graphically demonstrating univariate data without resorting to a boxplot:
“Each bean consists of a density trace, which is mirrored to form a polygon shape. Next to that, a one-dimensional scatter plot shows all the individual measurements, like in a stripchart. The scatter plot is drawn using one small line for each observation in a batch… To enable easy comparison, a per-batch average and an overall average is drawn” (Kampstra 3).
What does this mean in actual practice? I’ll demonstrate using one of his examples:
The figures each show 3 plots (a bimodal, uniform and normal distribution). This makes it quite clear how the beanplot provides, what can be, very revealing information about various samples. The quartile information in the boxplots really gives us no clue as to how remarkably different each of the samples are.
I have to say I really like this tool and will definitely be incorporating it into my work.
Peter Kampstra, 2008. Beanplot: A Boxplot Alternative for Visual Comparison of Distributions. Journal of Statistical Software, 28.