Confound it! Correlation is (usually) not causation! But why not?

By Stephen Downes - Stephen's Web. Confound it! Correlation is (usually) not causation! But why not?
gwern branwen, LessWrong, Jul 27, 2014

When somebody proposes a simple mechanism to improve (say) learning outcomes, they're most always wrong. But why? It's because they have ascribed a simple cause-effect relation onto a complex phenomenon. More...