R Programming By Example

Graphing disaggregated data with boxplots

Creating bar graphs is useful when presenting results to people who are not familiar with statistics, but the fact that bar graphs aggregate information (just as we did in the bar graphs for top performers) means that, in reality, we lose information due to the reduction. If you're working with people who understand what quartiles are, then boxplots may be a useful visualization. They are an easy way to see individual distributions for different levels of a variable.

Each box represents the first quartile at the bottom, the third quartile at the top, and the median on the line in the middle. The lines that extend vertically reach up to any observation within 1.5 * IQR, where the interquartile range ...