A while ago, I wrote about connected scatter plots. This graphic form is quite unusual in the news, but it may come in handy to display covariation. Today, The New York Times has published this —the print version of the graphic is quite large, by the way; I'm glad that the Times' graphics desk is still willing to challenge its more traditional readers a bit:
As I've discussed before, it's reasonable to argue that two stacked line charts would be equally —or more— effective and intuitive than the connected scatter plot. If you are a visualization researcher in need of a topic to study, it'd be really interesting to conduct an experiment comparing how easily readers interpret both graphic forms. I'm in favor of the connected scatter plot in certain cases, but I'd like to have some evidence either to confirm or nullify my hunch.