Sum Of Squares

Sum of the squared deviations from the mean

Adding the deviations would total zero - not very useful. But squaring the deviations, then adding them gives us very useful information.

SAS calls it "Corrected SS". What SAS calls the "Uncorrected SS" won't be used in Stats 1.

Can think of it on a histogram as constructing squares. When the distance from the mean is great, the square is very large. We are really calculating the area of the squares.

Other measures of variability or dispersion.

page revision: 5, last edited: 20 Sep 2008 00:10