How to get standard deviations on multiply imputed AND weighted data

[ejay0503]

Hello, all

Perhaps this is an easy question (but I can't figure out somehow!). I am using national data that is weighted so I multiple imputed the data using the weight. I am having trouble getting "pulled" standard deviations of the variables...I would greatly appreciate your help! Please see my code below. 

(proc mi.....truncated) 

proc surveymeans NOBS MEAN STDERR STDdata=perm.enjoy19m2 ;
strata RAESTRAT; cluster RAEHSAMP; weight R11WTRESP;
var age ;

domain _imputation_ ;
ods output domain = out ; run;

proc mianalyze data = out ; modeleffects mean;
stderr stderr ; run ;

Best regards, 

[SAS_Rob · Accepted solution]

In order to get a combined estimate you would need a measure of variation for the standard deviation.  You might be able to calculate one using the Delta method or something similar but this is not something that SAS reports.

[ejay0503]

It's very good to know that SAS doesn't report SDs on imputed data. Thank you very much for your help, Rob!