Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: multiple imputation using Stata 11 ICE


From   maria villarroel <mvillarr@hotmail.com>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: multiple imputation using Stata 11 ICE
Date   Thu, 6 Oct 2011 14:16:34 -0400

Hi,

There appears to be an inconsistency between the number of imputations generated versus detected in the dataset. Could someone please suggest why it shows that I have M=0 imputations and clarify how I might correct the syntax? I am concerned that the additional observations generated via imputation are not being recognized. 

After running an ice command with 5 imputations in Stata 11, I opened the new imputed dataset and ran “mi query” to examine the results. The MI query results shows that there are M=0 imputation, the exact message is "data mi set wide, M=0 and date of the last update." However, I am expecting that M=5. The "mi describe" also shows that there are M=0 imputations. However, a tab of the system variable "__mj" correctly identifies imputations m= 0 to 5 and that the dataset has 6 observations for each case. 


Below is my syntax. The syntax runs without error messages. (Categorical vars: concur educ4 marst4 race1origlastct origlastgc. Continuous variable: q5 q6. All other variables are binary). 


mi set wide
mi svyset drgfinwt [pweight=drgfinwt], strata(sample) vce(linearized) singleunit(missing)

mi register imputed concur educ4 marst4 race1 aidsclose aidsnet alcvar avoidvar beenpaid4var drug23y drug2ever intgenmale lifeprt6p age3d1 maleage3d1 drug1ever multiple1y nevervar newprt1y orighadct orighadgc origheardct origheardgc origlastct origlastgc q5 varwithcsw 

mi register regular creg male s1 assignedmode interviewmethod numatt perpov sample w1- predphon urbcoded1-popchange4d4 centralloss10p- noncorefast

ice m.concur i.creg m.educ4 m.marst4 m.race1 q6 aidsclose aidsnet alcvar assignedmode beenpaid4var drug23y drug2ever interviewmethod intgenmale lifeprt6p age3d1 male maleage3d1 drug1ever medmsafast multiple1y urbcoded1 urbcoded2 urbcoded4 urbcoded5 nevervar newprt1y numatt orighadct orighadgc origheardct origheardgc m.origlastct m.origlastgc perpov popchange4d1 popchange6d4 q5 s1 varwithcsw smallmsafast smallmsaloss, ///

gen(m_) substitute(race1:i.race1 educ4:i.educ4 origlastct:ioriglastct origlastgc:i.origlastct concur:i.concur) saving(imputationthird, replace) m(5) seed(92911) cond(lifeprt6p multiple1y concur beenpaid4var multiple1y newprt1y varwithcsw: nevervar==0 \ origlastgc orighadgc: origheardgc==1 \ origlastct orighadct: origheardct==1) ///

eq(drug23y: perpov urbcoded4 _Icreg_4 lifeprt6p _Imarst4_3, /// 
drug1ever: alcvar popchange4d1 urbcoded2 _Icreg_4 numatt q5, /// 
educ4: _Iconcur_2 urbcoded5 s1 assignedmode, /// 
marst4: orighadct _Iconcur_2 beenpaid4var urbcoded1, /// 
age3d1: q5 urbcoded2 numatt, ///
**additional custom imputation lines omitted. 

Thanks. 

Maria  		 	   		  
*
*   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/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index