For ANOVA, the denominator of the F-ratio is called the term. The error term provides a measure of the variance due to random (or unsystematic) differences. When the treatment effect is zero ($\displaystyle {H}_{{0}}$ is true), this term measures the same sources of variance as the numerator of the F-ratio, so the value of the F-ratio is expected to be nearly equal to .