Quickly Adding Depth to Flat Figures

Last week I attended a talk by Christopher Love, a faculty member at MIT. I noticed a simple visual effect he used several times in his presentation to give depth and dimension to what would otherwise be a flat figure.

On one slide, he illustrated a crowd of people as a way to talk about the effects of a treatment. The figure could have looked something like this, if he'd left it flat.

Now, if had been me, I'd probably have just left it like this. But not Dr. Love. He went the extra mile and added depth by changing the size and color. He made figures in the foreground bigger and darker, while figures in the background were smaller and lighter.

It takes 30 seconds to add these extra details, and the result is more pleasing and interesting. Now you have this same trick in your toolbox.

Graphical Perception

Not all quantitative visualizations are created equal. Graphical perception refers to the "visual decoding of information encoded on graphs". Human beings are good at decoding some types of "visual encodings", but not others. For example, we are much better at understanding the relative differences between data points on a linear scale, but not so good at gauging angles or areas. Presenting your information in "easily decoded" formats will help your audience digest the material more accurately and more easily.

In 1984, researchers William S. Cleveland and Robert McGill published a paper quantifying the ability of people to accurately decode different types of visual information. The results of this paper are summarized in this blog post at FlowingData.com.