Dear all,   I am trying to do multiple imputation, by treatment group.    i am running the following code:   proc MI data= raw  seed=46485 nimpute=5  out=  temp ;    class trt01p ;    var trt01p base month1 month3 ;    monotone reg (month3 = base month1/details) discrim( trt01p /classeffects = include ) ; run;   the log always came up with a warning message:   WARNING: The imputed variable TRT01P in the MONOTONE statement is the leading variable in the VAR list. Missing values for this variable will not be imputed.   What is the issue here? should I use 'by treatment' other than 'class treatment', and also do not include treatment in the var list and monmontone statement?   Thanks very much   SL     ... View more

Dear all,   Recently I was running sas proc mixed to validate my counterpart's results. Below is the statement I used (-note, 'Base' in the model statement):    proc mixed data=indata method=reml;    class TRTPN(ref='1') SITEGR1 AVISITN SUBJID; model CHG = BASE  TRTPN SITEGR1 AVISITN TRTPN*AVISITN  / DDFM=KENWARDROGER; repeated AVISITN / type=un sub=SUBJID; lsmeans trtpn*avisitn/pdiff cl ; run;   My counterpart runs like this:   proc mixed data=indata method=reml;    class TRTPN(ref='1') SITEGR1 AVISITN SUBJID; model CHG =  TRTPN SITEGR1 AVISITN TRTPN*AVISITN  BASE  / DDFM=KENWARDROGER; repeated AVISITN / type=un sub=SUBJID; lsmeans trtpn*avisitn/pdiff cl ; run;     Surprisingly, the standard error estimates from the lsmeans statement are quite different. Take one example: mean (SE) for model-adjusted means for one visit of one treatment group: 9.78 (4.896)           9.78 (2.150)   The adjusted means are the same but the standard errors change a lot; and the standard error values depend on the location/order of the covariates.   I do not understand why. Am i doing something wrong or the variance-covariance changes with the order of the covariates?   thanks all.               ... View more