Help using Base SAS procedures

Possible? Multiple imputation of nonparametric data and performing Wilcoxon or Friedman

Reply
N/A
Posts: 1

Possible? Multiple imputation of nonparametric data and performing Wilcoxon or Friedman

Hi,

does anybody know if this is POSSIBLE? Or am I trying the impossible?Smiley Wink

Performing a multiple imputation (propensity score method) on nonparametric data with missing values. After that performing a nonparametric procedure (Wilcoxon, Friedman). And then pooling it (mianalyze).

This is how far i got:

proc mi data=WORK.MASTER_TEST_SAS seed=21355417 nimpute=5 mu0=50 10
180 out=outmi;
mcmc chain=multiple displayinit initial=em(itprint);
var MZP GAF;
run;

proc npar1way wilcoxon correct=no data=outmi;
class MZP;
var GAF;
exact wilcoxon;
by _Imputation_;
OUTPUT out=outnpar1way;
run;

So far everything woked fine. But know I dont know how to perform the
MIANALYZE step correctly:

proc mianalyze data=outnpar1way;
MODELEFFECT P2_WIL?????????;
STDERR P2_WIL??????????;
run;

Please, if anybody has a good idea I would be more than happy...pSmiley Wink

Thanks,

ed

Ask a Question
Discussion stats
  • 0 replies
  • 506 views
  • 0 likes
  • 1 in conversation