*
chapter_3_table_7_and_9

The data set analyzed to replicate Chapter 3 Tables 7 and 9 consist
of the number of drinks per week at a nine month follow-up for 
participants in five different therapy groups. The purpose of this 
data set is to assess the likelihood of the assumptions of the 
parametric statistical tests (in this case the assumptions of normality 
and homogeneity of variance). This data set is also of interest in 
order to explore potential transformations (e.g., the log 
transformation), such that the parametric assumptions may be more 
satisfied. 

Here we use PROC MEANS in order to replicate Tables 3.7 and 3.9. Note 
that the medianfor 'T1 CRA-D' (i.e., Group 1 in the Output) should be 
10.98, not 11.89 as is listed in Table 3.9. In addition, the kurtosis 
for 'T2 CRA-D' (i.e., Group 4 in the Output) in the book should be 
2.78, not 1.06 as is listed in Table 3.9.
;

LIBNAME md 'd:\data files by type\sas data files\tables';
     DATA c3t79;
       SET md.chapter_3_table_7_and_9_raw_data;
PROC MEANS DATA=c3t79 MEAN N MEDIAN MIN MAX STD SKEWNESS KURTOSIS;
CLASS group;
RUN;