next up previous contents
Next: 3.2.6.4 F Test Up: 3.2.6 Analysis of Variance Previous: 3.2.6.2 Fitted Model

3.2.6.3 Residual

The so called sum of squares of the residuals is computed by summing the squares of the observed data with the data of the fitted regression model.

\begin{displaymath}
SSE = \sum_{i=1}^n (y_i - \hat y_i)^2
\end{displaymath} (3.23)

The complete analysis of variance table is shown in Table 3.12. In the row `regression' the degree of freedom (df) and the sum of squares and the estimated mean square of the fitted model and in the next row, the `residual', the values for the difference of the calculated and observed data are listed. The values of F, R2, and RA2 from the first row are described later. The last row, `total', holds the sum of squares of the observed data.

Table 3.12: The analysis of variance table.
Source df Sum of Squares Mean Square
Regression p-1 SSR $\frac{SSR}{p-1}$
Residual n-p SSE $\frac{SSE}{n-p}$
Total n-1 SST  


next up previous contents
Next: 3.2.6.4 F Test Up: 3.2.6 Analysis of Variance Previous: 3.2.6.2 Fitted Model

R. Plasun