Chapter 10. Visualizing Proportions
We often want to show how some group, entity, or amount breaks down into individual pieces that each represent a proportion of the whole. Common examples include the proportions of men and women in a group of people, the percentages of people voting for different political parties in an election, or the market shares of companies. The archetypal such visualization is the pie chart, omnipresent in any business presentation and much maligned among data scientists. As we will see, visualizing proportions can be challenging, in particular when the whole is broken into many different pieces or when we want to see changes in proportions over time or across conditions. There is no single ideal visualization that always works. To illustrate this issue, I discuss a few different scenarios that each call for a different type of visualization.
Remember, you always need to pick the visualization that best fits your specific dataset and that highlights the key data features you want to show.
A Case for Pie Charts
From 1961 to 1983, the German parliament (called the Bundestag) was composed of members of three different parties, CDU/CSU, SPD, and FDP. During most of this time, CDU/CSU and SPD had approximately comparable numbers of seats, while FDP typically held only a small fraction of seats. For example, in the eighth Bundestag, from 1976–1980, CDU/CSU held 243 seats, SPD 214, and FDP 39, for a total of 496. Such parliamentary data is most commonly ...