Correlation Equals Causation…?

Why can’t we figure it out?! 

If there is one thing that is drilled into every science, psychology and math major’s head, it is that a correlation does not equal a causation. This makes me wonder why this seemingly blatant academic maxim is so largely ignored in “the real world”. It is one thing for individuals to fall victim to cognitive biases, but it seems to me that society as a whole has fallen victim to the correlation-causation problem; and the authorities aren’t doing anything to help this invisible crisis, in fact, they’re making it worse!

Illusory correlation is a cognitive bias that occurs when two unrelated events are inaccurately believed to be causally related. I worry that many of the headlines containing results from scientific studies show lack of awareness about, as well as a propagation of, illusory correlations. For example, from Montreal Gazzette articles published this week, “Heat waves can boost chance of early births: study”, “Failure to shed baby weight can lead to heart disease, diabetes”, and “Kids can get sick in doctor’s waiting rooms”. I, and I’m guessing most of my peers, see such statements and immediately begin asking questions like “who was this measured in?”, “who funded this study?” and of course the dreaded, “could a third variable account for this relationship?”. The fundamental definition of a scientific experiment is that it involves following systematic procedures in an attempt to disprove a hypothesis, not just merely describing things as they are. Many of the headlines in today’s media are written as if they are conclusions set in stone — kids get sick because they go to waiting rooms, failing to shed baby weight causes diabetes, heat waves induce labor. The reality is that, most often, reviewing the study even briefly would make one realize that the actual finding was that two variables are linked. Scientific papers are very conservative when it comes to claiming causal links. Still, the explanations provided by the media are presented as if they have been confirmed.

The case of the MMR (mumps, measles, rubella) vaccine is the famous example of how assuming that correlation means causation can be very harmful. In 1998, a fraudulent research paper in the medical journal The Lancet claimed that autism was caused by the MMR vaccine. The media naively reported these findings, with Jenny McCarthy, a Playboy Playmate, actress, and model acting as the primary spokesperson for the cause. Because of this negative publicity, vaccination rates dropped drastically. The true link between autism and vaccines was the result of a third variable, namely, age. Most individuals with autism are diagnosed around the age of 3, and infants typically receive the MMR vaccine when they are slightly older than a year, therefore, onset of autism and application of an MMR vaccine are related. Every child diagnosed with autism had had the vaccine in the not-so-distant past. The overlooked point was that children who did not have autism had also received the same vaccine.

This reasoning is an example of a logical fallacy; diagnosis of autism does relate to having had received an MMR vaccine, but this does not mean that the vaccine causes autism. Claiming that to be true is, in my opinion, only slightly less absurd than claiming that moving to a poor country will increase penis size. But freedom of expression is a right to which every North American is entitled, even when they are wrong. And for this reason, things like this and this keep happening.

Yes, that’s right measles is back. This just proves that ignorance is not always bliss. In some cases, ignorance will cause the return of a previously eliminated (at least in North America) virus. Now that, is scary stuff.

Click here to interact with the map.