eagereyesTV Episode 3: 3D Pie Charts For Science!

How do we read pie charts? This seems like a straightforward question to answer, but it turns out that most of what you’ve probably heard is wrong. We don’t actually know whether we use angle, area, or arc length. In a short paper at the VIS conference this week I’m presenting a study I ran to answer this question – a study using 3D pie charts!

This is the third eagereyesTV video. Watch it below or head over to YouTube. And please like and subscribe if you thought it was worthwhile, and let me know what you think!

Things I cover in the video:

• The 2016 pie chart studies with Drew Skau
• An explainer of what we found in those studies in an easier-to-digest format
• Write-up of the VIS 2019 short paper, the study I discuss in the video
• All papers and materials from that paper
• All the other pie charts articles here on my blog
  

Posted by Robert Kosara on October 20, 2019. Filed under pie-charts, eagereyestv.

---

Comments

Michael says…

Does this mean that normal pie charts (focus on "area") could be better encoded than donout charts (focus on "arc")?

Steve Horne says…

Thanks for the blog. Pie charts are not going to disappear anytime soon so the data visualization community should come up with some guidelines on how not to misuse them.

Annette Greiner says…

Thank you for making the effort to study this oddly captivating question. If I'm interpreting your latest study correctly, you are essentially claiming that if people used angle to determine proportion on a pie chart, they would assign the proportion of the projected angle without compensating for the percept of three-dimensionality, and similarly, if they used arc length, they would assign the proportion of the projected arc length without compensating. That strikes me as unlikely. How have you validated that model?