options pageno=1;
data reg;
input x y;
xy=x*y;
*Comment: The previous statement create a new variable, which is the
product of X*Y; 
datalines;
2 21
4 27
6 29
8 64
10 86
12 92
;;
proc print;
title 'Printout of Data';
run;
proc means maxdec=2 sum mean uss css;
*Comment:  USS will calcualte the sum of each value squared and CSS will
calculate the sum of squares for each variable;
title 'Descriptive statistics';
run;
proc reg;
model y=x/cli clm;
title 'Simple Linear Regression, Including Confidence Intervals';
run;