Correlation

Lore & Background

The concept of correlation is central to statistics, describing the degree to which two variables are linearly related. The most common measure is the Pearson product-moment correlation coefficient, developed by Karl Pearson from an idea by Francis Galton. This coefficient is calculated as the ratio of covariance to the product of standard deviations, yielding a value between -1 and +1. A value of +1 indicates a perfect positive linear relationship, -1 a perfect negative linear relationship, and 0 no linear relationship. However, correlation is not the same as dependence: uncorrelated variables may still be dependent. The presence of correlation alone is insufficient to infer causation, a principle often summarized as 'correlation does not imply causation.' Other coefficients, such as Spearman's rank correlation coefficient, are more robust and detect less structured relationships. The concept has been generalized to mutual information and distance covariance.

Reader's Guide

Correlation is a foundational concept in statistics, used to quantify the strength and direction of a linear relationship between two variables. Its primary significance lies in its ability to indicate predictive relationships, as seen in practical applications like an electrical utility adjusting power production based on weather-correlated demand. However, its limitations are critical: correlation does not imply causation, and uncorrelated variables may still be dependent. The Pearson correlation coefficient, the most common variant, is sensitive only to linear relationships, while other coefficients like Spearman's rank correlation address nonlinear associations. The coefficient ranges from -1 to +1, with values near zero indicating weaker linear dependence. Despite its widespread use, correlation must be interpreted cautiously, as it does not establish cause-and-effect. Its legacy includes informing fields from economics to medicine, where it helps identify associations but requires further analysis to confirm causality. The development of alternative measures, such as mutual information, reflects ongoing efforts to capture more complex dependencies.

Did You Know?

• The Pearson correlation coefficient is defined only if both standard deviations are finite and positive.
• The absolute value of the Pearson correlation coefficient is not bigger than 1, as a corollary of the Cauchy–Schwarz inequality.
• Correlation is not the same as dependence: uncorrelated variables may still be dependent.
• The concept of correlation has been generalized to mutual information and distance covariance.