CORRELATION VS CAUSATION

Vagdevi Kommineni
2 min readOct 15, 2018

--

We commonly come across these two terms in almost all walks of our lives related to statistics. There is also a misconception about their implication. Let’s look at what they are:

Correlation is a statistical measure, indicating the size and direction of a relationship between two variables.

Causation indicates that the occurrence of one event is the result of the occurrence of another event; implying the causal relationship between the two variables.

It is worth noting that correlation does not imply causation.

For example, sales of ice-creams and sales of sun-screen lotions would be in correlation. They both are relatively higher during summers than in any other seasons. Though they exhibit high correlation, it is really common sense to think that neither of them is the cause of the occurrence of the others. The cause of the both is the summer being very hot. Consider more examples below:

An intuition of Correlation and Causation
Example

Correlation is generally denoted by ‘r’ and its value lies in [-1,1].

If the correlation coefficient of the two variables is 1, they are positively correlated, meaning that both of their values increase or decrease simultaneously.

If the correlation coefficient of the two variables is -1, they are negatively correlated, meaning that one of their values increases while that of the other’s decreases simultaneously.

If the coefficient of the two variables is 0, then there is no relation between the two variables.

--

--