As a bit of a data geek, I found this hilarious. This graph from Businessweek shows examples where correlation (two trends that look similar) is not causation (trend A did not cause trend B). I’ve included the funniest one above, but there’s more.
I often wonder how the way we represent data shapes the way we view the world, and the decisions we make. Rene Descartes came up with the Cartesian Plane, two perpendicular number lines. This technically connected geometry and algebra, but the more obvious effect is that you can plot two sets of data against one and other. As income goes up, so does house size, as temperature goes down, so does amount of clothing worn. I wonder though, does this emphasis on two limit our understanding of the world?
Very few aspects of reality are determined by one factor, especially social aspects. You see numerous studies linking this or that to cancer, but cancer is actually caused by a complex interplay of many factors, genetic, behavioral, and seemingly random. I’ve seen studies where they say that red wine can fight heart disease, but what if you drink red wine and smoke and have a history of heart disease and do steroids? How do you fit that on a Cartesian Plane?
There is of course multivariate analysis, which basically does what it says on the box, analyzes multiple variables. You rarely if ever see that in popular media or conversation though. It’s not a great news story if you say that multiple factors are involved in cancer and we don’t really understand it. If licking rabbits can show a decline, however, that’s a story.
I myself only usually deal with two variables at a time (in infographics), because that’s what I’m comfortable with, and what I think people are comfortable understanding. Quite often, however, I wonder what I’m missing.