I love the above comic from XKCD. I thought of it in the middle of a great conversation recently on TalentNet Live that was, ostensibly anyway, about social media and best practices. Somehow we stumbled into the difference between correlation and causation. It’s one of my favorite things to talk about, which in and of itself is a little odd, to be sure. But for anyone who doesn’t know the difference, help is here.
Correlation is when two things tend to happen (or not happen) together. The example I used on the show was the correlation between ice cream consumption rates and drowning deaths. Statistically, these two data trends move together. When ice cream consumption rises, drowning deaths do as well. When ice cream sales start to drop, fewer people drown.
That’s not to say, of course, that people are drowning in ice cream. Rather, there is another factor in play, which is heat. When does ice cream tend to be consumed? When it is hot. And what do many people do when it is hot? Head to the water. Ice cream and drowning incidents have no relationship other than matching trends, both being related to a third factor. Correlation.
Causation, on the other hand, is when there are two data sets that move in harmony, but one is, in fact, causing the other. To follow the previous example, there is a causal relationship between the average daily temperature and ice cream consumption. The rising temperature inspired people to seek out a cool, tasty treat more readily than in the winter. Causation.
There also exists inverse correlations, in which two date sets tend to move in opposite directions. An increased number of years spent in school tends to be reflected in a lower unemployment rate. We all know that higher education isn’t a guarantee of a job, nor are people usually hired only because of their education, but there is certainly a relationship.
I hope now that I’ve given you a quick explanation, the comic at the top of the page is as funny to you as it is to me.