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 on April 23, and its replacement, statalist.org is already up and running.


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

st: Micombine Logit and Logit for Multiple Imputations


From   Kristen Budd <buddkm@gmail.com>
To   statalist <statalist@hsphsun2.harvard.edu>
Subject   st: Micombine Logit and Logit for Multiple Imputations
Date   Wed, 29 Sep 2010 16:02:24 -0400

Hi everyone.  I am having difficulty with my imputed data set and
standard errors.  I conducted multiple imputations (10) using ICE due
to the level of missing data.  I get vastly different standard errors
when using the command -- for varlist x1 x2 x3: logit y X -- compared
to when using the command -- for varlist x1 x2 x3: micombine logit y
X.  When using "micombine logit" the standard errors are much bigger.
The standard errors should be approximate to each other, but they are
not.

I am not sure why I am encountering this problem.  I am working with
STATA 10.1.  Any helpful input or ideas would be greatly appreciated.
An example with women is below.  Using "logit" the s.e. is 0.104 --
increasing to 0.358 when using the "micombine logit" command.

---------------------------------------------------
 Variable |        logit            mi
-------------+------------------------------------
  women |  .46984281    .47091965
             |  .10455369    .35799366
             |     0.0000       0.1884
   _cons |  2.1079187    2.1025736
             |  .06864507    .23416615
             |     0.0000       0.0000
---------------------------------------------------
                             legend: b/se/p

Best,

Kris

*
*   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