Alberto Cairo, who teaches Information Graphics and Visualization at the the University of Miami, has written about connected scatter plots on his blog (with a follow-up post here). He has been inspired, as I have, by Hans Rosling’s wonderful animated scatter plots. Cairo wondered how to accomplish something similar with a static chart, and he realized that one way is to use “connected” scatter plots. In these, each dot is associated with a specific year or month. He gives some nice examples, including some of his own work.
This reminded me that when I started looking at the data underlying the Reinhart-Rogoff debt controversy earlier this year, I wanted to find ways to visualize each country’s experience. The data consisted of annual Debt/GDP ratios and growth rates for various countries over half a century. Some analysts had portrayed all of these data points together in one grand scatter plot. The problem is that this format gives no sense of the changes experienced by each country over time. I wanted to see what how country’s story unfolded.
My first attempt to visualize this data was to create connected scatter plots for individual countries. Here is a chart for the U.K.:
We can see that it took twenty years for Britain to work its Debt/GDP ratio from about 250% after World War down to the now-infamous 90% threshold at the heart of the Reinhart-Rogoff controversy.
In the end, I choose to animate the data in something of the Rosling style, so I didn’t use the connected scatter plots. I am glad to have this excuse to share one of them.
Note: the chart above was based on my interpretation of the raw data used by Reinhart and Rogoff. The animations I made later were based on the data as interpreted by R&R themselves.