Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: losing observations while correcting for selection bias using selmlog


From   "Ramani Gunatilaka" <[email protected]>
To   [email protected]
Subject   st: losing observations while correcting for selection bias using selmlog
Date   Sat, 19 Apr 2008 13:57:01 +0600

Dear Statalist,
I am using selmlog to estimate wage functions. The full sample is of
labour market participants.
For the multinomial logit part, I have five categories - 1.
unemployed, 2. contributing family workers, own account workers,
employers, 3. public employees 4. private formal employees 5. informal
employees.
I have wage data only for categories 3, 4 and 5. I want to estimate
wage functions for each of these. As suggested by Rafa de Hoyes in an
old Statapost (12 April 2006), I defined separate wage variables for
each of these wage categories.
w3 is the wage data for public employees. There are 3723 of them.

Now when I run the following model for the public sector, the eventual
wage equation ends up with only about 400 observations.

selmlog w3 $basefactors1, select(selmcat=$basefactors1 $selfactors) showmlogit

Can somebody please tell me why this is so?
BTW, I intend bootstrapping to fix the standard errors but will embark
on that time-consuming exercise once I sort out this problem.
Thank you in advance,
Ramani
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index