Residual analysis

Consider the plot of the (X , Y) values given in the table in the previous page. Also, imagine the overlap of the plot of the values with the regression line of y and x. The values of a and b are computed on the previous page on a condition that


is minimum. That is, the residual sum of squares will be minimum if we use the values of a and b in the previous page to form the linear regression equation.


Definition:

Residual variance is a measure of the variation of the y values about the regression line. Residual variance is given by the following equation :


The dot separating the y and x subscripts indicates that the regression line is that of y on x, because we are predicting y values from the corresponding x values.
Definition:

The square root of the residual variance is called the standard error of estimate.