Essay Prompts: Correlation is not Causation

Here’s an interesting fact: Humans are stupid. I don’t mean to imply we don’t have intelligence (although some who claim to have a high IQ, well, let’s just say they get elected to public office on other qualities). Rather, we put our trust in things we shouldn’t (and not just politicians). We are horrible at judging risk. We often see things that just are not there. We often believe the most ridiculous nonsense about cooking, such as fresh ground salt is better.  Worst of all, we often confuse correlation with causation.

  • Correlation: a mutual relationship or connection between two or more things
  • Causation: the relationship between cause and effect; causality.

Here are two examples of this confusion I’ve seen in my news feeds and on FB:

One fellow wrote about a friend of his that got a good report from the doctor in a followup visit, stating: “He’s definitely living proof that God answers prayers.” No, he isn’t. It is wonderful that the friend is doing better, but there is no causality here. You can pray to God or a Saint and get better, but that is not proof that God or the Saint was *why* you got better, no matter what the Pope says. There is no proven or provable, testable, repeatable method of showing that one action causes the other.

Another fellow wrote about high tax states, citing a article that he believed said that “high state taxes cause people to leave those states, making it very difficult to actually increase tax revenue, no matter how high their tax rates get.” Again, there’s no causality here. Yes people leave high tax states, but they leave low tax states as well, for many many reasons. Taxes may be a reason, but typically it isn’t the precipitating reason. And in the absence of clear evidence that whereever taxes are higher, people leave, this is just correlation. There are many high tax states where people don’t leave (witness property values in California) and high tax countries where people don’t leave. Further, “people” is far too nebulous a category, for all people are not the same. Are you talking those with wealth? Those on fixed incomes? Retirees? People in particular industries? The issue is just too nebulous to attribute to causality.

Always remember storks and babies. I had a statistics professor explain it this way: You may think there is a causality between storks and babies, because whereever there are more babies, there are more storks. But that’s a spurious correlation: there are more babies where there are more storks because there are more babies in big cities, and big cities often have zoos, which have more storks. No causality there.

Keep this in mind as you read your papers.

Share

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.