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 at the end of May, and its replacement, statalist.org is already up and running.


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

RE: st: Rounding multiple imputation variables


From   Erum Ikramullah <eikramullah@childtrends.org>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   RE: st: Rounding multiple imputation variables
Date   Mon, 2 Aug 2010 16:59:12 -0400

I am imputing my data with the following code:

mi impute mvn twobio par_ed p_rlgn othp_age othp_empenr othp_edcont othp_re2 = mpf r_edcont rempenr_p race4c agebaby1
r1236100 if finalsamp_m=3D3D=3D3D1, add(100);

The imputation does work as I am able to see all my missing filled in:
Multivariate imputation                 Imputations =3D3D      100
Multivariate normal regression                added =3D3D      100
Imputed: m=3D3D1 through m=3D3D100                  updated =3D3D        0

Prior: uniform                           Iterations =3D3D    10000
                                            burn-in =3D3D      100
                                            between =3D3D      100

               |              Observations per m
               |----------------------------------------------
      Variable |   complete   incomplete   imputed |     total
---------------+-----------------------------------+----------
        twobio |       3509            7         7 |      3516
        par_ed |       3210          306       306 |      3516
        p_rlgn |       3109          407       407 |      3516
      othp_age |       3273          243       243 |      3516
   othp_empenr |       3251          265       265 |      3516
   othp_edcont |       3186          330       330 |      3516
      othp_re2 |       3370          146       146 |      3516
--------------------------------------------------------------
(complete + incomplete =3D3D total; imputed is the minimum across m  of the number of filled in observations.)

The problem is that I want to run two of my predictor variables as categorical using xi.  However, Stata does not allow for xi on multiply imputed data.  This is why I then want to round the values.  So, basically, instead of a value of 1.3492 for race, I would want a value of 1.  Anyone know how to do this rounding of multiply imputed variables??

Thanks,
Erum


-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Richard Goldstein
Sent: Monday, August 02, 2010 2:56 PM
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: Rounding multiple imputation variables

you don't tell us how you are imputing the data but it appears you are
not using the correct -mi impute-; you do not say how many categories
there are but you should look in the manual for -mi impute logit- and
-mi impute mlogit-

Rich

On 8/2/10 2:43 PM, Erum Ikramullah wrote:
> Hello,
> 
> I am working on a project in which I am imputing missing data using Sata 11  mi impute command.  I have 2 categorical variables with missing data that I would like to impute, however when I run mi: estimate the model does not run on the multiply imputed categorical data, understandably.  I have been advised to round the imputed values for the categorical values and then use those new variables in my mi: estimate model.  Does anyone have code that will allow me to do this rounding?
> 
> Thanks very much!!
> 
> Erum Ikramullah
> Senior Research Analyst
> 4301 Connecticut Avenue, NW Suite 350
> Washington, DC 20008
> www.childtrends.org
*
*   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/

*
*   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