« Fewer posts this week | Main | The simplest chart: one data series »



Do you have the same objection to bubbles instead of dots on a 2-D scatter plot where bubble size is used to represent a third dimension? I wish I could include an example from my research where , because the data are dense, they work far better than actually trying to show the third dimension.


John, it's well known that real quantities are very hard to interpret when shown as areas (see processtrends.com for an example of Cleveland's hierarchy of interpretable graphics). I like to use bubbles as a "bonus" information feature on a scatter graph, but not to make an important point. I always want the important data to be in the first two dimensions of the scatter graph, not the area of the bubbles.

Bubbles are best when they show small integers: it's easy to see the difference between a bubble of area=2.0, and one of area=1.0 or 3.0, especially if the graph comes with a key.

Jon Peltier

Bubbles are also effective when they are not used to display a continuous variable, but a variable with discrete values. What comes to mind is the dots representing cities on a map, where a tiny dot is for <10,000 people, the next size is 10-25k, then 25-50k, 50-100, 100-250, etc.

But for gereral use, unless you drastically skew the scale, bubbles are only pretty markers.

The comments to this entry are closed.


Link to Principal Analytics Prep

See our curriculum, instructors. Apply.
Kaiser Fung. Business analytics and data visualization expert. Author and Speaker.
Visit my website. Follow my Twitter. See my articles at Daily Beast, 538, HBR.

See my Youtube and Flickr.

Book Blog

Link to junkcharts

Graphics design by Amanda Lee

The Read

Keep in Touch

follow me on Twitter