topic Re: multiple imputation question - no variability across imputations? in SAS Procedures

multiple imputation question - no variability across imputations?

deleted_user — Tue, 13 Jan 2009 17:05:39 GMT

I have been running multiple imputations on a dataset I have. The problem I'm getting is that there is no variability across imputations - the estimate is the same for each imputation. Can someone tell me what the problem is?

Re: multiple imputation question - no variability across imputations?

lamenramen — Tue, 29 Mar 2011 07:19:03 GMT

I too have had this issue, but I'm not sure what to do about it. I can't generate a covariance matrix either when this happens, which makes mianalyze not work.

Re: multiple imputation question - no variability across imputations?

Sekhar — Thu, 31 Mar 2011 19:38:02 GMT

Hi,

If you want to impute the missing values with mean or median or 0 or with any stats, you have to use Proc STDIZE procedure.The below is the example

proc stdize data=abc method=median reponly;
run;

For missing values if we want to impute with zeroes then the example would be

proc stdize data=abc missing=0 reponly;
run;

Kindly let me know if this useful or not.