Spurious Correlations

I read today that heavy use of social media correlates well with the divorce rate.  I’m thinking is it social media that creates marital difficulty or is it people who are in unhappy marriages spend more time on social media?  Which is the cause and which is the effect?

Statisticians say that correlation does not imply causation.  Tyler Vigen, a Harvard Law student, posts daily graphs at Spurious Correlations that show some interesting relationships between random data.

For example the more lawyers there are in Virginia, the lower the number of divorces becomes. And the divorce rate in Maine closely matches the graph for the consumption of Margarine there.

You can also use the generator at the site to make your own spurious correlations.