Can you make sense of this?
The blue line shows the ten-year trend in the marriage rate in Mississippi. The orange line represents how much milk we consume in the United States.
Uncanny, huh? Does drinking less milk result in fewer marriages, or do you suppose that fewer marriages lead to drinking less milk?
How about this one — German cars sold in the U.S. (green) versus the prevalence of suicides by car crash (blue)? Clearly, this graph shows that the number of such suicides is somehow correlated to the U.S. sales of Audis and BMWs.
Of course, the closeness of the trend lines is not supposed to make sense. It’s just a bit of geek humor. The charts are the work of Harvard law student Tyler Vigen, who wants to help destroy the idea that correlation = causation. He’s got a website dedicated to calling out such “spurious correlations.” Skeptics will find much to their liking there; for one thing, Vigen’s site is a great tool for reminding others to steer clear of the old cum hoc ergo propter hoc fallacy.
And if Vigen’s drollery doesn’t do it for you, maybe this xkcd joke will: