*
chapter_7_table_15

The following hypothetical salary data represents a nonorthogonal 
two-by-two factorial design. The first factor (sex) is crossed with 
college (degree or no degree). The primary question of interest is 
whether or not there is sex discrimination in terms of salary.

As before, this factorial ANOVA is analyzed using the GLM procedure. 
However, unlike the factorial ANOVAs that we have previously seen, this
one has "/ SS3" on the model statement line. This option says to use the
Type III sums of squares. Changing "SS3" to "SS1," "SS2," or "SS4" will 
produce the type I, II, and IV sums of squares, respectively (multiple 
sums of squares can also be specified by placing them after the forward 
slash on the model statement line). Recall that the different types of 
sums of squares (less Type IV) are formally compared in Tables 7.19 
through 7.22. 

Note that for the Type I sums of squares the order an effect enters 
the model statement line matters. Listing "A" first on the model 
statement line, for example, will lead to "Type I sums of squares 
where 'A' is entered first."
;

LIBNAME md 'd:\data files by type\sas data files\tables';
     Data c7t15;
       SET md.chapter_7_table_15;
PROC GLM DATA=c7t15;
CLASS educ sex;
MODEL salary = sex educ sex*educ / SS3;
RUN;