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: oaxaca with multiply imputed data?


From   "Muehleck, Kai" <Muehleck@dzhw.eu>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   st: oaxaca with multiply imputed data?
Date   Fri, 4 Oct 2013 14:32:54 +0000

Dear all,

I'm using a Blinder-Oaxaca decomposition to analyse the gender-wage gap by use of Ben Jann's Oaxaca ado. The analysis is based on multiply imputed data. mi estimate does not support oaxaca but it is possible to run the decomposition model together with mi estimate by using the cmdok (command ok) option. 

Technically this works. However, I don't know if this also correct from a statistical point of view. The results look ok. They do (of course) differ from the results without imputed data but not in an extreme and unexpected manner (coefficients differ to some extent and SE are smaller using the imputed data). I have also checked whether Stata does apply the Rubin rule to the coefficient estimates with oaxaca and that looks correct as well. Still, I am not sure whether this is really ok and whether I can trust the t-statistics etc. as well.

Any thoughts on this highly appreciated!

Best regards,
Kai

Deutsches Zentrum für Hochschul- und Wissenschaftsforschung GmbH
Goseriede 9 | 30159 Hannover | www.dzhw.eu

Dr. Kai Mühleck
Abteilung Hochschulforschung
Arbeitsbereich Steuerung, Finanzierung, Evaluation
Telefon +49 (0)511 1220-456 | Fax +49 (0)511 1220-250 
E-Mail muehleck@dzhw.eu 

Registergericht: Amtsgericht Hannover | HRB 210251
Geschäftsführer: Dr. Bernhard Hartung
Vorsitzender des Aufsichtsrats: Ministerialdirigent Peter Greisler





*
*   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