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

RE: st: Factor variables and mim

From   "Lachenbruch, Peter" <>
To   <>
Subject   RE: st: Factor variables and mim
Date   Mon, 10 Aug 2009 11:42:52 -0700

This sounds like a nice option.  Does this imply that I can run a
stepwise regression or LARS by using cmdok?


Peter A. Lachenbruch
Department of Public Health
Oregon State University
Corvallis, OR 97330
Phone: 541-737-3832
FAX: 541-737-4001

-----Original Message-----
[] On Behalf Of Yulia
Marchenko, StataCorp LP
Sent: Sunday, August 09, 2009 12:14 PM
Subject: Re: st: Factor variables and mim

Fred Wolfe <> pointed out that although -mi
estimate- supports factor variables, it does not support -xtmixed-:

> Yes, -mim- does not accept factor variables - too bad.  and -mi
> doesn't work with xtmixed or mfp
> mim works with xtmixed.
> From my perspective, if you want to use multiply imputed data sets and
> -xtmixed-, it appears that -ice- and -mim- are what must be used, but
> without factor variables - something that could be done in Stata 10.1.

Fred is right that -mi estimate- does not yet officially support
However, -mi estimate- provides a way of obtaining results from
virtually any
estimation command via option -cmdok-.  For example, you can type:

. mi estimate, cmdok: xtmixed y x1 i.x2 || group:

to obtain MI results from -xtmixed- with factor variables.

*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index