Covariance
lastmod 2023-10-26created 2014-10-2310 edits
In probability theory and statistics, covariance is a value that indicates the degree to which two random variables are related. (It is distinct from variance, which indicates the dispersion of a single variable.) If, when one of the two variables tends to increase, the other also tends to increase, the covariance will be positive. Conversely, if, when one of the two variables tends to increase, the other tends to decrease, the covariance will be negative. Covariance thus shows whether a relationship tends upward or downward, but since its value changes with the scale of the units in which the two variables are measured, it is unsuitable for gauging the strength of the relationship through correlation analysis. Correlation analysis uses the population correlation coefficient ρ as the measure of the strength of a relationship.
Formula
Covariance_formula.png