groups

[deleted_user]

My surveys requires over 5,0000 interviewees rank two conditions
separately from 1 to 10. I wish to assign interviewees into 10*10 =
100 groups such as group (1,1), the group that interviewees rank both
conditions 1 or group (1,2), the group that interviewees rank the first
condition 1 and the second condition 2 etc.
Do you know any way to help me out?
what happens if we rank three conditions?
Based on the group, i wish to do statistics such as mean, median, standard deviation, t-test for the diffrent between groups.
Thank you so much for your help,

[Patrick]

I believe for the first 3 measures a procedure like proc means or even proc tabulate would do the job.
Just use your two variables with the conditions as classification vars.

I'm having limited stats knowledge only, but isn't a t-test for comparison of 2 distributions only. As you're having balanced data I believe it's Proc ANOVA you're looking for.


HTH
Patrick