Many students asked me what exactly is r-value. They are always confused with the idea of correlation and causation. If you’re still confused, let me first establish that r-value implies correlation between the two variables and you definitely hear that correlation does not imply causation in lectures.
Correlation does not imply causation!
Firstly, causality is ideological. If data is correlated, we need to still determine if one causes the other. Take for example, ice cream sales and murder. Rates of violent crime and murder have been known to jump when ice cream sales do too. So they are correlated. But buying ice cream doesn’t turn you into a killer!
Other examples include Education & Wage, Obesity & Debt Bubble, etc.
And to solve this problem, there is the need for hypothesis testing.
[…] For regression, I hope students know when to use to calculate sample values, and the intuition behind it. Also understand that r-value indicates correlation, and not causality, know that its independent of scaling and translation. You can read a bit here. […]