Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


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

st: multiple imputation postestimation commands with complex survey data


From   "Cabrera-Nguyen, Elian" <pcabrera@wustl.edu>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   st: multiple imputation postestimation commands with complex survey data
Date   Mon, 28 Oct 2013 05:40:10 +0000

Dear Statalisters,

I am relatively new here, and I apologize in advance for any unintentional
breaches in etiquette. I promise to engage in appropriate acts of corporal
mortification to atone for any breaches of etiquette identified by fellow
Statalisters.

I am using Stata 12.1/SE on a Windows platform. I am analyzing complex
survey data.  Using -mi impute chained-, I have imputed M=100 data sets.

I am having difficulty finding out how to conduct two rather basic
analyses using -mi estimate: svy, subpop(indicator):-

Specifically, I have been trying to (a) perform a simple t-test to compare
means of a continuous variable between documented and undocumented
immigrants, and (b) conduct pairwise comparisons of mean scores of a
continuous variable by race (a polytymous variable). Prior to imputation,
I did so successfully by using the following code:

For a simple t-test:


svy, subpop(indicator): mean SCORE, over(IMSTAT)
test [SCORE]_subpop_1= [SCORE]_subpop_2

To conduct the ANOVA, I obtained mean scores for the continuous variable
by race and then performed pairwise comparisons as follows:

svy, subpop(indicator): regress SCORE i.RACE, baselevels
testparm i.RACE
pwcompare i.RACE, pveffects  mcompare(bonferroni)



However, after extensive research (including online reviews of the MI
literature specific to Stata; the Multiple Imputation section of the Stata
manual; and consulting wonderful texts including Carpenter and Kenward's
(2013) excellent book entitled _Multiple Imputation and its Practical
Applications_) , I have been unable to find a way to conduct an analog to
the above analyses using -mi estimate- postestimation commands. I am
probably missing something very basic, but I cannot figure out what it is.
Any guidance would be greatly appreciated, as would any well-deserved
mocking of my ignorance and stupidity.

Thanks you so much for your time and for this invaluable resource.

Sincerely,
 
Elián

Note: Email written using voice dictation software. Apologies for any
typos or grammatical errors


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index