0

I am have a data set with multiple visits with 2 treatment arms and a Vehicle group. Also i have a variable say "SSA" having two values 1 and 0. here 1 stands for responder and 0 for non-responder subjects. while performing Proc Freq for chi-square statistics i am getting the following error. Here is the code i used

 PROC FREQ DATA=P&V1;
       TABLE TREATMENT*SSA/CHISQ ;
       WHERE TREATMENT IN (1 &TR1); *(&tri for treatment 2 and treatment 3);
 RUN;

NOTE: No statistics are computed for TREATMENT * SSA since SSA has less than 2 no missing levels. WARNING: No OUTPUT data set is produced because no statistics can be computed for this table, which has a row or column variable with less than 2 no missing levels. this error is for my last visit where i am having a single value 0 in all treatment groups and Vehicle.

ved_null_
  • 71
  • 2
  • 10
  • please suggest is there any way to get P-values for those visits? – ved_null_ Nov 25 '13 at 05:11
  • Pretty sure you cannot do a Chi sq test using a field with no variation. – scott Nov 25 '13 at 14:17
  • @scott i also wonder why my statistician asking for it when i am saying there is no variation. Is there any PROC Freq method to get the P-values – ved_null_ Nov 26 '13 at 03:20
  • As far as I know, there's no method to get p-values from PROC FREQ in this situation because p-values don't make sense here. – Alex A. Dec 12 '14 at 17:55

0 Answers0