* 
Chapter 4, Table 1

The hypothetical example shown in Table 4.1 consists of systolic blood pressure scores for 
20 participants in four treatment groups for hypertension. The participants were randomly 
assigned to the four treatment groups of interest: drug therapy, biofeedback, diet, and 
combination. Note that for this example, there are not equal sample sizes in each group; 

LIBNAME md 'd:\data files by type\sas data files\tables';
DATA c4table1; 
SET md.chapter_4_table_1;

* Our first hypothesis of interest is whether the drug therapy and feedback groups differ in 
their effectiveness in treating hypertension. To test this hypothesis, we perform a simple 
pairwise comparison for the drug therapy and feedback groups. An illustrative example of the 
calculations for this analysis is shown in Table 4.2, while the SPSS output and contrast 
coefficients for this analysis are shown in Table 4.4. The following SAS syntax is used to 
obtain the pairwise comparison of means between the drug therapy and the biofeedback groups 
in Chapter 4, Table 1;

PROC GLM; 
CLASS cond; 
MODEL bloodpr = cond;
CONTRAST 'Drug Therapy vs. Biofeedback' cond 1 -1 0 0;
RUN;

* Our second hypothesis of interest is whether the average of the drug therapy, biofeedback, 
and diet groups differs from the combination group in effectiveness in treating hypertension. 
Using the data from Table 4.1, we test this hypothesis by performing a complex comparison of 
means by comparing the average of the means of the drug therapy, biofeedback, and diet groups 
to the mean of the combination group. An illustrative example of the calculations for this 
analysis is shown in Table 4.3, while the SPSS output and contrast coefficients for this 
analysis are shown in Table 4.4. The following SAS syntax is used to obtain the complex 
comparison of means between the average of the drug therapy, biofeedback, and diet groups and 
the combination group in Chapter 4, Table 1;

PROC GLM; 
CLASS cond; 
MODEL bloodpr = cond;
CONTRAST 'Drug Ther, Biofeed, and Diet vs. Combination' cond 1 1 1 -3;
RUN;