Covariance

In probability theory and statistics, covariance measures how two random variables change together. Covariance is positive when higher values of one variable generally coincide with higher values of the other, and lower values tend to coincide as well—that is, when the variables move in the same direction. Covariance is negative when higher values of one variable tend to coincide with lower values of the other, indicating movement in opposite directions. The magnitude of covariance is harder to interpret because it is not standardized and depends on the variables’ units of measurement. The correlation coefficient is a normalized measure derived from covariance; its magnitude indicates the strength of the linear relationship between the variables.