Correlation and causation

You may have heard the phrase ‘correlation does not imply causation’ or ‘correlation is not causation’, but what does this mean?

Earlier, you saw how two variables can be correlated, but common sense suggests that ice cream sales do not cause homicides. While that example was exaggerated, in some cases this might not be so obvious. It might not surprise you that people in the northern hemisphere generally spend more when it’s cold than when it’s hot, but does this mean that cold weather causes increased spending?

Humans are prone to fabricating patterns when two variables appear to be so closely associated – assuming one is dependent on the other – and implying a cause and effect relationship (i.e. A causes B). In reality, there are many other possibilities, including:

• the opposite is true: B causes A

• A and B are correlated but caused by C (a ‘confounder’)

• another variable is involved – A causes B but only when D occurs

• A and B are correlated but it’s just a coincidence.

In our example of cold weather spending, the increase in spending may not be caused by the lower temperatures, but rather the cold weather happens to coincide with festive sales (Black Friday, Cyber Monday and new year sales).