st: MI estimate error

["Michael Beck" <michael.beck@edu.unibe.ch>]

Hi everybody

I have a problem with estimations using stata11 and the mi estimate command.
I am trying to run certain multinominal logits with different dependent
variable. The data is consisting of 5 already imputed datasets, on which I
recoded some variables and then used the "mi import:"  command to create a
usable data file. I also  created an "original data" file containing missing
values. My MI data is now consisting of m=0 (Containing missing values) and
m=1 - m=5. 

With regards to content I am checking the probabilities for the transitions
to lower secondary schools for different nationalities as well as the
probabilities for being in a certain category of proficiency levels.

So far so good, when I am running the following command 

mi estimate, post: mlogit uebertritt staat2 staat3 staat4 staat5,
baseoutcome(1) rr vce(cluster XYZ)

(uebertritt being the transition to different school forms, staat2-5 being
nationalities und XYZ being a variable identifying school classes).

Everything is working fine. But when I exchange the dependent variable
trying to compute the following

mi estimate, post: mlogit manote staat2 staat3 staat4 staat5, baseoutcome(2)
rr vce(cluster XYZ)

(manote being the grade in math)

Stata is telling me:

mi estimate: omitted terms vary
    The set of omitted variables or categories is not consistent between m=1
and m=5; this is not
    allowed.  To identify varying sets, you can use mi xeq to run the
command on individual
    imputations or you can reissue the command with mi estimate, noisily


I have no idea what the problem is and how to deal with it.  So if anyone
knows this error and has an idea, how I could fix it, I would be very happy,
if someone could help me with this one.

Kind regards
Michael  


 


*
*   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/