In my dataset from an actual clinical trial, I have 64 ID numbers (numbered 1 to 64), where 32 goes to treatment group A and 32 goes to treatment B and a score (outcome)
So I want to add a hypothetical variable where I want to generate a variable (let's call it Gender/Sex, so 2 levels) that is sex is evenly and randomly distributed within each group (16 males and 16 females per treatment group). How do I do this randomly?
If you use survey select you can name your strata:
proc surveyselect data='trial' method=srs n= 'sample size'
seed= 'number' out='trial sample;
strata "enter strata here, gender sex with no column, try both and see what gives you the desired output"
run;
I put everything you'll need to type in quotes. Hope this helps.
Well, that statisticians out there probably have the best answers for this. My way would be:
data want;
call streaminit(123);
set have;
male_cnt=0;
female_cnt=0;
if rand("uniform") <0.5 then do;
if male_cnt <16 then do;
sex="male";
male_cnt=male_cnt+1;
end;
else do;
sex="female";
female_cnt=female_cnt+1;
end;
else do;
/* same as above but reveresed */
end;
run;
Probably a bit simplistic and not so random, but if it doesn't matter too much...
If you use survey select you can name your strata:
proc surveyselect data='trial' method=srs n= 'sample size'
seed= 'number' out='trial sample;
strata "enter strata here, gender sex with no column, try both and see what gives you the desired output"
run;
I put everything you'll need to type in quotes. Hope this helps.
thanks that worked for me. I used that to randomly select the men.. and then if the sex variable is missing, i pop recoded that to women.
It's finally time to hack! Remember to visit the SAS Hacker's Hub regularly for news and updates.
Learn the difference between classical and Bayesian statistical approaches and see a few PROC examples to perform Bayesian analysis in this video.
Find more tutorials on the SAS Users YouTube channel.
Ready to level-up your skills? Choose your own adventure.