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: RE: Descriptives using multiple-imputed complex survey design data


From   Stas Kolenikov <skolenik@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: RE: Descriptives using multiple-imputed complex survey design data
Date   Tue, 23 Feb 2010 13:50:21 -0600

This MI data set identifier may be string; -describe- and -destring- it if
needed.

I personally have certain reservations against using multiple imputation
with survey data. I have convinced myself that you are pulling oneself's leg
with stratification: the imputation models usually cut across strata, so
observations between strata cannot be assumed independent anymore. I am sure
there are other problems, too. So for a more appropriate calculation, you'd
probably want to drop stratification from your -svyset-. This essentially
boils down to

mim: <command> [pw=<sampling weight>], cluster( <PSU variable> )

if your original design was

svyset <PSU variable> [pw=<sampling weight>], strata( <strata variable>)

If you had finite population corrections and other stages of selection, then
I don't really know what to make of them.

On Tue, Feb 23, 2010 at 1:38 PM, Rosie Chen <jiarongchen2002@yahoo.com>wrote:

> Tony, thanks a lot for the timely help! Now I have renamed the dataset
> indeitifier variable according to your advice. Since I am using Stata 9.2
> and therefore don't have a pdf manual, may I ask for further clarification:
>
> After I run
> mim: svy: mean smoke
> Another error message appeared:
> observation identifier variable _mi is either missing or not numeric
>
> Is there an easy way to create such an observation identifier variable? If
> it is too complicated, I will just give up the idea of using Stata but use
> SAS for such an analysis. Thanks for your suggestion in advance!
>
> Rosie
>
>
>
> ----- Original Message ----
> From: "Lachenbruch, Peter" <Peter.Lachenbruch@oregonstate.edu>
> To: "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
> Sent: Tue, February 23, 2010 12:25:44 PM
> Subject: RE: st: RE: Descriptives using multiple-imputed complex survey
> design  data
>
> mim wants an identifier of the imputed data set.  If SAS produces such an
> indicator, simply rename it _mj or generate _mj=SASindicator.  Otherwise,
> you will need to create it/
>
> Tony
>
> Peter A. Lachenbruch
> Department of Public Health
> Oregon State University
> Corvallis, OR 97330
> Phone: 541-737-3832
> FAX: 541-737-4001
>
>
> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu [mailto:
> owner-statalist@hsphsun2.harvard.edu] On Behalf Of Rosie Chen
> Sent: Monday, February 22, 2010 10:06 PM
> To: statalist@hsphsun2.harvard.edu
> Subject: Re: st: RE: Descriptives using multiple-imputed complex survey
> design data
>
> Hello, does anyone know what this error message really means?: "imputation
> identifier variable _mj is either missing or not numeric".   I have 5
> datasets that have been imputed in SAS, and now would like to use them for
> descriptive analysis in Stata. Below is what I've done. What am I doing
> wrong? Thank you very much in advance!
>
>
> mimstack, m(5) sortorder(id)
> istub("data") nomj0
>
> compress
>
> inc1 was float now
> byte
> inc2 was float now
> byte
> inc3 was float now
> byte
> inc4 was float now
> byte
>
> svyset
> [pweight=aweight]
>  pweight: aweight
>          VCE: linearized
>     Strata 1: <one>
>         SU 1: <observations>
>        FPC 1: <zero>
>
> mim: svy: mean smoke
> error message: imputation
> identifier variable _mj is either missing or not numeric
>
>
> ----- Original Message ----
> From: Rosie Chen <jiarongchen2002@yahoo.com>
> To: statalist@hsphsun2.harvard.edu
> Sent: Tue, February 23, 2010 12:19:01 AM
> Subject: Re: st: RE: Descriptives using multiple-imputed complex survey
> design  data
>
> Thanks a lot for the helpful information, Tony and Steve! Will read the
> instruction and have a try soon.
>
> Rosie
>
>
>
>
> ----- Original Message ----
> From: "sjsamuels@gmail.com" <sjsamuels@gmail.com>
> To: statalist@hsphsun2.harvard.edu
> Sent: Mon, February 22, 2010 7:24:39 PM
> Subject: Re: st: RE: Descriptives using multiple-imputed complex survey
> design  data
>
> To Tony's response I would only add: read the help for Stata's survey
> commands ("help survey" will start you out. Since you have Stata 9.2,
> pdf's of the manual pages are not available, but the -help- system is
> also quite good.  The major difference between running survey commands
> in SAS and Stata is that in SAS, one specifies the survey design
> variables as statements in statements for each procedure; in Stata,
> one uses the -svyset- command prior to running other survey commands.
> I haven't run -mim-, but since it is a prefix command, your commands
> will have a double-prefix structure:
>
> mim , [mim options]:  svy:
>
>
>
> Steve
>
>
>
> On Mon, Feb 22, 2010 at 2:09 PM, Lachenbruch, Peter
> <Peter.Lachenbruch@oregonstate.edu> wrote:
> > a) read the volume on multiple imputation (pdf or manual)
> > b) consider using mim (see help file, or findit mim)
> > c) is 5 sufficient number of imputations?  I have found 20 usually to be
> a bit safer.
> >
> > Tony
> >
> > Peter A. Lachenbruch
> > Department of Public Health
> > Oregon State University
> > Corvallis, OR 97330
> > Phone: 541-737-3832
> > FAX: 541-737-4001
> >
> > -----Original Message-----
> > From: owner-statalist@hsphsun2.harvard.edu [mailto:
> owner-statalist@hsphsun2.harvard.edu] On Behalf Of Rosie Chen
> > Sent: Monday, February 22, 2010 1:28 PM
> > To: statalist@hsphsun2.harvard.edu
> > Subject: st: Descriptives using multiple-imputed complex survey design
> data
> >
> > Dear all,
> >
> >   I am not sure if my question is too simple or unclear, but I haven't
> received responses from the list. Could anyone give me some hint? I have
> done such an analysis before in SAS, but would like to learn how to do it in
> Stata. Thank you very much. Below is my question:
> >
> >    I wonder if someone can advise on the syntax I
> > should use for descriptive analysis (e.g., mean, cross-tab) for
> > multiple-imputed complex survey design data. Basically, 5 datasets have
> > been created through multiple imputation in IVeWare in SAS. How to
> > incorporate complex survey design features and the 5 imputed datasets
> > in descriptive analysis in Stata? By the way, my Stata version is 9.2.
> > Any sample syntax for mean and cross-tabs would be appreciated very
> > much. Thanks,
> >
> >    Rosie
> >
> >
> >
> > *
> > *   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/
> >
>
>
>
> --
> Steven Samuels
> sjsamuels@gmail.com
> 18 Cantine's Island
> Saugerties NY 12477
> USA
> 845-246-0774
>
> *
> *   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/
>
>
>
>
> *
> *   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/
>
>
>
> *
> *   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/
>



-- 
Stas Kolenikov, also found at http://stas.kolenikov.name
Small print: I use this email account for mailing lists only.
*
*   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