The mean sum of squares
associated with error variation (MSE) is the ratio of the sum of squares
associated with error variation to the degrees of freedom. In regression analysis the MSE is the
variance term. The MSR is used to
specify the average amount of variation explained about the regression line by
the model parameters.
MSE
= SSE / dfSSE
MSR
= SSR / dfSSR
where, df is degrees of freedom.