The old miest (version 1.15) works fine for "svymean" and "svylogit".
Leonelo
>>> Kenneth Scheve <kenneth.scheve(a)yale.edu> 06/17/03 11:34AM >>>
The new version of miest 2.0 is the one we recommend everyone to use. It
has been updated for Stata 7.0 and written for the most used procedures in
stata (about 25 different commands).
miest 1.15 was written with Stata 5.0 in mind and was written to work with
virtually every command used in stata at that time. It now is somewhat
unreliable in that it will bomb or give nonsense answers for some commands.
Which is of course why we updated the proc. Generally, however, for many
commands the 1.15 version works just fine. Whether it works for "svy"
commands probably depends on which svy command you use. My recommendation
would be to use the 1.15 version for the model you are estimating but check
by hand a few parameter estimates (the formulas that miest calculates for
combining parameter estimates are very simple and available in our article,
Schafer's book, etc.). If it works, then you should be fine. Again though
the version of the program that we are supporting and recommending for use
is the most recent one on Gary's web cite--2.0.
Ken
At 11:17 AM 6/17/2003 -0400, Leonelo Bautista wrote:
>Ken:
>I have miest versions 1.15 and 2.0 (downloaded from Gary King's page
>today). Version 1.15 runs with the svy commands. However, version 2.0 does
>not. Does version 2.1 run with svy? If so, where can I get it? If not,
>would results from version 1.15 be correct?
>Thanks for your help.
>
>
>
>Leonelo E. Bautista, MD, MPH, DrPH
>Assistant Professor
>Department Preventive Medicine & Biometrics
>Uniformed Services University of the Health Sciences
>4301 Jones Bridge Road
>Bethesda, MD 20814
>Room A1039
>Phone: (301)2953712
>Fax:(301)2951854
>
> >>> Kenneth Scheve <kenneth.scheve(a)yale.edu> 06/17/03 10:02AM >>>
>
>I don't have Stata 8.0 yet but in Stata 7.0, miest saves the matrices in
>memory. So if you want to access _mib or _miVCE, use stata's matrix
>commands. For example, after estimation type.
>
>matrix list _mib
>matrix list _miVCE
>
>Finally, just a reminder that you should be using the latest version (2.1)
>of the miest and misum procedures.
>
>Ken
>
>At 09:46 AM 6/17/2003 -0400, Leonelo Bautista wrote:
> >Hi,
> >I'm running "miest" in Stata 8. After the estimation "miest" closes all
> >active datasets. Where does miest "save the multiple imputation parameter
> >estimates in the vector _mib and the corresponding variance-covariance
> >matrix in _miVCE"?
> >Thanks in advance,
> >
> >
> >
> >Leonelo E. Bautista, MD, MPH, DrPH
> >Assistant Professor
> >Department Preventive Medicine & Biometrics
> >Uniformed Services University of the Health Sciences
> >4301 Jones Bridge Road
> >Bethesda, MD 20814
> >Room A1039
> >Phone: (301)2953712
> >Fax:(301)2951854
The new version of miest 2.0 is the one we recommend everyone to use. It
has been updated for Stata 7.0 and written for the most used procedures in
stata (about 25 different commands).
miest 1.15 was written with Stata 5.0 in mind and was written to work with
virtually every command used in stata at that time. It now is somewhat
unreliable in that it will bomb or give nonsense answers for some commands.
Which is of course why we updated the proc. Generally, however, for many
commands the 1.15 version works just fine. Whether it works for "svy"
commands probably depends on which svy command you use. My recommendation
would be to use the 1.15 version for the model you are estimating but check
by hand a few parameter estimates (the formulas that miest calculates for
combining parameter estimates are very simple and available in our article,
Schafer's book, etc.). If it works, then you should be fine. Again though
the version of the program that we are supporting and recommending for use
is the most recent one on Gary's web cite--2.0.
Ken
At 11:17 AM 6/17/2003 -0400, Leonelo Bautista wrote:
>Ken:
>I have miest versions 1.15 and 2.0 (downloaded from Gary King's page
>today). Version 1.15 runs with the svy commands. However, version 2.0 does
>not. Does version 2.1 run with svy? If so, where can I get it? If not,
>would results from version 1.15 be correct?
>Thanks for your help.
>
>
>
>Leonelo E. Bautista, MD, MPH, DrPH
>Assistant Professor
>Department Preventive Medicine & Biometrics
>Uniformed Services University of the Health Sciences
>4301 Jones Bridge Road
>Bethesda, MD 20814
>Room A1039
>Phone: (301)2953712
>Fax:(301)2951854
>
> >>> Kenneth Scheve <kenneth.scheve(a)yale.edu> 06/17/03 10:02AM >>>
>
>I don't have Stata 8.0 yet but in Stata 7.0, miest saves the matrices in
>memory. So if you want to access _mib or _miVCE, use stata's matrix
>commands. For example, after estimation type.
>
>matrix list _mib
>matrix list _miVCE
>
>Finally, just a reminder that you should be using the latest version (2.1)
>of the miest and misum procedures.
>
>Ken
>
>At 09:46 AM 6/17/2003 -0400, Leonelo Bautista wrote:
> >Hi,
> >I'm running "miest" in Stata 8. After the estimation "miest" closes all
> >active datasets. Where does miest "save the multiple imputation parameter
> >estimates in the vector _mib and the corresponding variance-covariance
> >matrix in _miVCE"?
> >Thanks in advance,
> >
> >
> >
> >Leonelo E. Bautista, MD, MPH, DrPH
> >Assistant Professor
> >Department Preventive Medicine & Biometrics
> >Uniformed Services University of the Health Sciences
> >4301 Jones Bridge Road
> >Bethesda, MD 20814
> >Room A1039
> >Phone: (301)2953712
> >Fax:(301)2951854
Ken:
I have miest versions 1.15 and 2.0 (downloaded from Gary King's page today). Version 1.15 runs with the svy commands. However, version 2.0 does not. Does version 2.1 run with svy? If so, where can I get it? If not, would results from version 1.15 be correct?
Thanks for your help.
Leonelo E. Bautista, MD, MPH, DrPH
Assistant Professor
Department Preventive Medicine & Biometrics
Uniformed Services University of the Health Sciences
4301 Jones Bridge Road
Bethesda, MD 20814
Room A1039
Phone: (301)2953712
Fax:(301)2951854
>>> Kenneth Scheve <kenneth.scheve(a)yale.edu> 06/17/03 10:02AM >>>
I don't have Stata 8.0 yet but in Stata 7.0, miest saves the matrices in
memory. So if you want to access _mib or _miVCE, use stata's matrix
commands. For example, after estimation type.
matrix list _mib
matrix list _miVCE
Finally, just a reminder that you should be using the latest version (2.1)
of the miest and misum procedures.
Ken
At 09:46 AM 6/17/2003 -0400, Leonelo Bautista wrote:
>Hi,
>I'm running "miest" in Stata 8. After the estimation "miest" closes all
>active datasets. Where does miest "save the multiple imputation parameter
>estimates in the vector _mib and the corresponding variance-covariance
>matrix in _miVCE"?
>Thanks in advance,
>
>
>
>Leonelo E. Bautista, MD, MPH, DrPH
>Assistant Professor
>Department Preventive Medicine & Biometrics
>Uniformed Services University of the Health Sciences
>4301 Jones Bridge Road
>Bethesda, MD 20814
>Room A1039
>Phone: (301)2953712
>Fax:(301)2951854
Hi,
I'm running "miest" in Stata 8. After the estimation "miest" closes all active datasets. Where does miest "save the multiple imputation parameter estimates in the vector _mib and the corresponding variance-covariance matrix in _miVCE"?
Thanks in advance,
Leonelo E. Bautista, MD, MPH, DrPH
Assistant Professor
Department Preventive Medicine & Biometrics
Uniformed Services University of the Health Sciences
4301 Jones Bridge Road
Bethesda, MD 20814
Room A1039
Phone: (301)2953712
Fax:(301)2951854
Hello,
I'm using Amelia to impute data and Clarify to run regressions on the
five imputed datasets produced by Amelia. Clarify does not produce
the stats. for model fit (I'm particularly interested in R2, Adj. R2,
and F). Up to this point, I've been using Stata's regular regression
command (outside of Clarify) to generate the same regression on all
five imputed datasets to get these values, which I then average over
the five sets of results to get the overall figures for the MI
results. But if there are lots of regressions, this is pretty time
consuming.
Has anyone out there written any code to automate this procedure in
Stata? (I don't have a lot of programming experience.) If so, would
you be willing to share it?
Thanks,
Strom
--
Strom C. Thacker
Susan Louise Dyer Peace Fellow
Hoover Institution
Stanford University
Stanford, CA 94305-6010
650-725-3432
650-723-1687 (fax)
sthacker(a)bu.edu
http://www.bu.edu/sthacker
