On Friday Vanderbilt’s Curb Center for Art, Enterprise, & Public Policy hosted a “boot camp,” featuring half a dozen workshops and presentations on creative practice. I attended a talk by Eric Rodenbeck, founder and creative director of Stamen Design, on the topic of data visualization. As is my practice these days, I took some sketchnotes on the talk, and I thought I would share them here on the blog:
Rodenbeck showed several fascinating examples of data visualization, including a bubble chart showing transactions on the NASDAQ stock exchange and this video by Scott Manley showing all the asteroids discovered between 1980 and 2011:
Rodenbeck also made a few points that I thought were particularly interesting.
- In many areas of life, we are collecting massive amounts of data on an ongoing basis. As the amount of data on a topic increases, visualizations of that data become more interesting–and the probability that you’ll see something cool in those visualizations approaches 100%.
- When trying to design useful visualizations of a data set, it’s best to start small. Take a slice of the data, try out a few visualizations, and look for stories. When a story emerges from a particular visualization, train that visualization on the larger data to see if that story continues.
- If you want others to explore your data set, show them the data using an appropriate visualization and let them “play” with the data (thought sliders and other controls). They’ll start to ask questions about the data that they wouldn’t be able to ask without this “play” time.
- The idea that there’s one visualization to rule them all is a fiction. Different visualizations tell different stories about data.
I was glad to see so many of my colleagues (at least 40 of them) in the room learning about data visualization. I think it’s important that we teach our students how to read and make sense of these kinds of data visualizations–if not create such visualizations themselves.