Correlation

Tuesday, 2 July 2019

3:11 PM

Correlation The covariance of two r.v. is important as an indicator of the relationship between them. However, it heavily depends on units of X and Y (difficult interpretation, not scale-invariant). —+ The correlation coefficient p is often used instead. It is the covariance between the standardised versions of X and Y, or, explicitly, cov(X, Y) var(Y) Properties: p is dimensionless (no unit) o p always has a value between —I and I. o Positive (and negative) p means positive (and negative) linear relationship between X and Y o The closer IA is to I, the stronger is the linear relationship

 

Correlation examples 1.0 1.0 0.0 0.8 1.0 0.0 0.4 1.0 0.0 0.0 0.0 -1.0 -0.8 -1.0 0.0

 

 

Created with Microsoft OneNote 2016.