Hi I do not understand why, in the next example, type III SS(A) is greater than type II SS(A). Type III SSA is SC(A/B,A*B,X) while Type II SSA is SC(A/B,X) so, should not be always greater type II SS than type III SS?

data one;input a b y x; cards; 1 1 3.1 2 1 1 5 4.1 2 1 6 6 2 1 8 8.6 1 2 2 3 1 2 1 2 2 2 3 4.9 2 2 4.2 5 ; proc glm data=one;class a b;model y=a b a*b x / ss1 ss2 ss3 ss4; run; Source DF Type II SS Mean Square F Value Pr > F

a 1 0.00114134 0.00114134 0.01 0.9472 b 1 4.48520264 4.48520264 20.34 0.0204 a*b 1 0.17922373 0.17922373 0.81 0.4337 x 1 4.36355090 4.36355090 19.79 0.0211

Source DF Type III SS Mean Square F Value Pr > F

a 1 0.04119809 0.04119809 0.19 0.6947 b 1 3.61338699 3.61338699 16.39 0.0271 a*b 1 0.17922373 0.17922373 0.81 0.4337 x 1 4.36355090 4.36355090 19.79 0.0211

Thanks for your help

Javier