Speaking Stata: Creating and Varying Box Plots

Abstract

Box plots have been a standard statistical graph since John W. Tukey and his colleagues and students publicized them energetically in the 1970s. In Stata, graph box and graph hbox are commands available to draw box plots, but sometimes neither is sufficiently flexible for drawing some variations on standard box plot designs. This column explains how to use egen to calculate the statistical ingredients needed for box plots and twoway to re-create the plots themselves. That then allows variations such as adding means, connecting medians, or showing all data points beyond certain quantiles.