Correlation Coefficient: Definition
You'll learn how to measure the strength and direction of linear relationships in a standardized way.
What Is the Correlation Coefficient?
The Pearson correlation coefficient, usually denoted ρ (rho) for populations or r for samples, is simply standardized covariance. While covariance tells us whether two random variables move together, its value depends on the units of measurement. The correlation coefficient removes this problem by dividing covariance by the product of the two standard deviations:
where σ_X and σ_Y are the standard deviations of X and Y respectively.
Why Standardize?
Think of covariance like measuring distance in different units—sometimes meters, sometimes miles. The correlation coefficient converts everything to a universal scale: it always falls between -1 and +1, regardless of the original units.
- ρ = +1: Perfect positive linear relationship
- ρ = -1: Perfect negative linear relationship
- ρ = 0: No linear relationship
- Values in between indicate the strength of the linear association
Dimensionless Nature
Because we divide covariance (which has units like "dollars × hours") by a product of standard deviations (also "dollars × hours"), the units cancel out completely. This makes correlation dimensionless—a pure number with no units attached. You can compare correlations across completely different contexts: stock prices and temperatures, heights and weights, etc.
This standardization also means correlation is invariant under linear transformations: if you change X to aX + b (scale and shift), the correlation remains unchanged.
Key Takeaway: The Pearson correlation coefficient is covariance divided by the product of standard deviations, yielding a dimensionless measure of linear association that always ranges from -1 to +1.