Geometric and Statistical Interpretations of Correlation

Reported in interviews at Two Sigma

Problem statement

The Pearson correlation coefficient $\rho(X, Y)$ is one of the most used -- and most misused -- quantities in statistics.

Provide at least three distinct interpretations of $\rho$: one geometric, one in terms of linear regression, and one as a standardized covariance. For each, give a brief derivation or justification. Then state the key limitation that all three interpretations share.

Show hint

What single algebraic object do covariance, the angle between centered vectors, and a standardized slope all have in common?

Relevant roles: Data Scientist, Applied Scientist and ML Engineer.

Related free problems