Evidence Primer #1 – Correlation, Causation, Coincidence

August 3, 2010

Since vaccines were first introduced, some people have been concerned about their safety. Most recent concerns in the UK have been around MMR and the HPV vaccines, with scare stories regularly featuring in the media. Often, these will feature parents who claim that their child became ill following a vaccine, blaming the vaccine for the illness.

Whilst in some very rare cases the vaccine may be responsible for the illness, in most cases it isn’t. The reason people claim there is a link is because of a fundamental misunderstanding of the concepts of Correlation, Causation and Coincidence.

Correlation or Causation?

When we have two sets of information (A and B) we sometimes see that there is a relationship between them, whereby as one thing increases, so does the other. In such cases, people might assume that this means A is causing B to rise. This is why many people believe that the illness that occurred after a vaccine was caused by the vaccine. However, this is a common misconception, as A may not cause B.

The best way to demonstrate that a correlation between A and B doesn’t have to imply that A causes B is with some examples:

Example 1 – A teacher tells her class that she will set a test for them on a particular topic the next day. In fact, she has set two tests on the topic, one in the morning and one in the afternoon. When she compares each child’s results, she notices that those who did well in the morning test also did well in the afternoon test. It is clear that there is a correlation between these results, but does that mean that one caused the other? It is difficult to see how performing well in one test would cause a child to perform well in the other. A much better explanation is that the correlation between the test results was caused by something else, such as the fact that children who did well in the first test did lots of revision the night before. These children are therefore more likely to also do well in the second test. In this case, a third factor (the amount of revision done) is responsible for the correlation.

Example 2 – Many people who go to bed with their shoes on wake up the following day with a headache. Does this mean that going to bed wearing shoes causes headaches? Again, a much better explanation exists for the correlation between morning headaches and sleeping with shoes on. A few drinks too many the night before is likely to be the culprit!

Coincidence

A correlation between factors doesn’t always require an explanation as in the examples above. In some cases there can be a correlation between things that happen at the same time but have no direct causal relationship or common cause. In these cases the correlation is purely coincidental. Our third example demonstrates this:

Example 3 – Since 1750 there has been a large decrease in the number of pirates sailing the seas. At the same time there has been an increase in global warming. Does this mean that global warming, as well as endangering polar bears, has endangered the pirate? Or does it mean that if more people became pirates we could reverse global warming?  Neither, as there is no direct relationship between the two. It is simply coincidental.

Causation, Correlation, Coincidence and Vaccines

Understanding these concepts may make you think differently about how to interpret the fact that some children get ill after being vaccinated. In most cases the link is purely coincidental, however in others there may be a common factor involved. For example, in the case of MMR and Autism the common factor may be age. Children with Autism are usually diagnosed with the condition between the ages of one and two, regardless of whether they have been vaccinated or not. However, the MMR vaccine is also administered around the same time, which leads some people to assume there is a direct link between the two when in fact there is no evidence that this is the case.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s