Tag Archives: Statistics

Correlation CAN Imply Causation!

When analysing results in science we often look for something called ‘correlation’. This is a term that we give for a relationship or connection between two things. Finding and defining these relationships can be very useful to understand the way our world works.

However we can often fall into a trap, assuming that if there is a correlation, then that MUST mean one thing causes the other. Sometimes this is the case, for example there is a correlation between time spent driving a car and likelihood of a motor accident. Of course the more time you spend in a car, the higher the chance you may end up in a motor accident. This isn’t always the case though. Rather amusingly, there is a strong correlation between number of pirates and global average temperature (as number of pirates decrease, temperature rises). Now we can instantly tell this seems nonsense and is probably just a coincidence but other examples could be less clear.

To explain this further, the fantastic Youtuber’s at MinutePhysics produced this video which we thought we’d share!