Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | raoul reulen <r.c.reulen@gmail.com> |
| To | statalist@hsphsun2.harvard.edu |
| Subject | st: mi impute monotone inconsistencies |
| Date | Mon, 10 May 2010 01:12:41 +0100 |
Dear all I have a dataset with information on whether patients received a certain drug or not, and if so what dose of drug. Both the variables "drug (yes=1/no=0)" and "dose (continuous)" have missing values and the pattern in monotone. If drug is "no" then the dose is obviously "0", and if drug is "yes" then the dose cannot be "0". When I use an imputation model I do get a few inconsistencies, occasionally dose is "0" when drug is "yes" and occasionally dose is greater then "0" if the patient did not receive any drugs. .mi set mlong .mi register imputed drug dose .mi register regular case age year .mi misstable nested .mi impute monotone (logit) drug (pmm) dose = case age year , add(5) rseed(8833) What am I doing wrong? Thanks. Raoul * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/