Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: Heckman+Heteroskedasticity

From   "Rahman,L (pgr)" <>
To   <>
Subject   st: Heckman+Heteroskedasticity
Date   Sun, 24 Nov 2002 12:46:13 -0000

Dear Stata users,

Apologies if you have already received this email - I was not sure
whether it got sent as I didn't receive a copy:

I have a few questions regarding the heckman estimator:
1) Is there a way to accomodate heteroskedastic errors into the heckman
ml estimator?
2) My second problem is similar to others I have seen in the archives
but I haven't seen a solution -  I am getting very different results
when I use the ml over the two-step estimator. I thought the 2-step is
consistent but that the ml is more efficient. However, I am getting
completely different coefficients and s.e. from both methods. Unlike
previous queries in the archive my log likelihood seems to be maximising
ok (no concave functions at the last iteration - although there are some
at the beginning which I thought was fine). 
Any suggestions on how to proceed on either query is much appreciated.
Thank you,

Lupin Rahman
London School of Economics
Houghton Street, London
Tel: +44 207 955 6677

*   For searches and help try:

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