Covariance/Correlation Matrix

CORE DEFINITION

The covariance matrix is a fundamental object in multivariate statistics, and its standardized version, the correlation matrix, is standard practice in psychometrics, finance, and principal component analysis. Its core proposition is that the linear association between every pair of variables can be arranged in one symmetric table, so the joint structure of the whole set becomes inspectable at once rather than pair by pair. The key qualification is that it captures linear relationships only, and it is unstable when the sample is small relative to the number of variables.