84  Pie Charts — Type 2

Same Observations, Different Variables

The secont type of pie chart is the same observations, different variables variant. This one is a bit more tricky to work with, but there is an elegant solution. In ?fig-Bayerpiecharts we have three pie charts which I’ve adapted from a review article. This review article reported efforts from bayer Pharmaceuticals headquarters in Berlin to reproduce a number of studies in various biological research fields. In the original review, Bayer reported that only about a quarter of the 67 studies surveyed could be reproduced in their hands. They rightly highlight reproducibility as an important topic, so it’s a bit dismaying that they didn’t provide the actual data for their figure. Here, I’ve simulated some values that fit the struture. Each pie chart depicts the same observations but different variables.

Same observation, different variables pie charts.

Aside from the aforementioned difficulties in reading pie charts, this collection is disappointing because a very important and interesting question of the study is obscured. We want to know the story of each study assessed here. That is what value does each observation take in each of the three pie charts. It’s just not possible to convey this in three discrete pie charts. The solution would be a Sankey diagram, but that is outside the scope of this course.