Okay, then the best way is to make sure everything is in one dataset and then use a Macro. For example: ods html file = <file location>/&filename; %macro response(response, filename); Proc reg data = dsn; by <group variable>; /* Group should be your variable that differentiates the responses */ where <group variable> = &response.; /* Response is the label that you have chosen to identify each group */ model Y = x1 x2 x3 x4.....xn /selection =stepwise slstay=0.05 slentry=0.05; model Y = x1 x2 x3 x4.....xn /selection =backwards slstay=0.05 slentry=0.05; run; %mend; %response(group1, Y1_stepwise) %response(group1, Y1_backwards) %response(group2, Y2_stepwise) %response(group2, Y2_backwards) I haven't got SAS on my computer, so couldn't test it. I hope you understand it though. The key thing is to create a Macro, so that you can expliciity label your output filenames easier.
... View more
