Linear regression implies causality?

Hey Luis,

Thank you for your question!

Indeed neither correlation, nor regression necessarily imply causation. When faced with a problem, very careful analysis needs to be performed to determine the true cause-and-effect relationship.

This lesson aims at telling the difference between correlation and regression, which could be a subtle one.

Kind regards,
365 Hristina

my confusion is because in the lessons and in the 365 booklet it says that linear regression is used for causal relationships, why? causal and causality is different?

I think Luis,  one could make a regression with two totally different and unrelated variables Y and a X and still get a Beta coefficient. So regression does not mean causality, the causality remains in the theory or context behind the model.

luis, causality or direction of effect must first be theoretically derived before it can be assumed in a regression model. Thus, one cannot "search" for causality with the regression, the regression can only be used if a causal relationship is assumed.