Question

Correlation: Correlation Does Not Mean Causation

One of the major misconceptions about correlation is that a relationship between two variables means causation; that is, one variable causes changes in the other variable. There is a particular tendency to make this causal error, when the two variables seem to be related to each other.

What is one instance where you have seen correlation misinterpreted as causation? Please describe.

Answer 1

Correlation is a quantitative measure that gives the degree of linear association between two variables. It doesn,t in any form imply that one is causing the other. That is causation.

For example, there is a correlation between ice cream sales and sunglasses sale.

As temperature increases, both ice cream & sunglasses sales increase and as temperaure decreases, both ice cream & sunglasses sale decreases, hence a clear correlation between ice cream and sunglasses sale, but it no way implies that increased sunglasses sale are causing the ice cream sales to increase and vice vera,that would be an absurd assumption. The temperature is the cause.

Hence, correlation and causation are separate topics, there might be a causal relationship between two things that are correlated, but that is not necessary.