> It is pleasing to see error bars on a graph in a mainstream publication. Those are "dynamite plots", a favorite of scientists, but the statisticians I know p

> It is pleasing to see error bars on a graph in a mainstream publication. Those are "dynamite plots", a favorite of scientists, but the statisticians I know prefer to use boxplots or dotplots instead.