st: Size Inconsistency Using CMP

[bahareh sehatzadeh <sehatzadeh@gmail.com>]

Greetings,
I am using cmp model (conditional mixed process estimator) as a two
stage model with the first stage as a linear regression and the second
stage as an ordered probit one. Stata reports two individual models
for each stage and one mixed-process model. The number of observations
for the first stage individual model is 722 and for the second stage
individual model 822. I understand the difference here and have no
problem with it. However in the mixed-process regression the number of
observations is reported as 827. I want to know if Stata is using
observations with missing data in the regression process or it is just
a mistake in reporting the number. To deal with the problem I tried to
delete observations with missing data and made a new database with
only 722 observations and then ran the model. The results I got were
completely different and the coefficients were unexpected. I will be
grateful if someone can advise me on this issue. Thank you for your
consideration.
Sincerely,
Bahareh Sehatzadeh.
*
*   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/