I get the impression that a lot of people around the web confound absolutely everything when talking about correlations on the topic of social sciences.