Correlation Traps

One of the most seductive reasoning errors is treating correlation as causation. Ice cream sales and drowning deaths both rise in summer. Eating ice cream does not cause drowning — both are caused by hot weather.

Spurious correlations are everywhere. Countries that eat more chocolate win more Nobel Prizes. Bigger cities have both more churches and more crime. In each case, a hidden third variable explains both observations — national wealth in the first, population size in the second.