Wednesday, October 23, 2013

Noticing misleading graphs

I recently attended Edward Tufte's one-day course "Presenting Data and Information." I recommend it for anyone with a strong interest in producing or consuming complex, information-rich graphics, articles or other presentations. (If the $380 cost is an obstacle, one could also consider just purchasing some or all of the four books that come with the course; Tufte's excellent presentation drew heavily on the books.) For me, I believe the course will be helpful in various activities, including my magazine job, this blog and the book I am working on. In any case, the course has definitely sharpened my eye for noticing when info is being presented in a misleading, uninteresting or otherwise bad way. Here via Quartz is an example of a not-very-informative chart, produced by no less an entity than Apple. Quartz's analysis is here.

Me: For resting on one's laurels, and obscuring one's lack of recent progress, a focus on a cumulative number will work for a while. (Of course, if things are really bad, that will become noticeable as the curve approaches flatness.)

By the way, one thing I didn't particularly expect was Tufte's enthusiasm for the work of Richard Feynman, especially his famous Feynman diagrams. But in looking into that further, I find that I worked on a Michael Shermer column at Scientific American years ago (either fact checking it or copy editing it--can't remember) that involved Tufte, Feynman and the latter's diagram-illustrated van. As a further aside, I'm pleased to see that this TV movie is coming out: "The Challenger Disaster" (originally "Feynman and the Challenger").

No comments: