Hi all,
I am trying to estimate an IV model with GLM in Stata for a fractional
response variable using the -qvf- command. I have a covariate which is
endogenous because of which I need to use an IV for endogeneity in GLM.
I came to know about the -qvf- command from this post:
http://www.stata.com/statalist/archive/2005-01/msg00464.html
However, I could not come to a conclusion as to whether -qvf- can be used
without any problem of interpretation of the coefficients when the IV is
for endogeneity. Prof. Wooldridge's work suggest the use of -glm-
family(binomial) link(probit) robust. Refer: (Unobserved Heterogeneity
and Estimation of Average Partial Effects, Section 4.2)
http://www.msu.edu/~ec/faculty/wooldridge/current%20research/ape2r8.pdf
Can any any of you tell me whether -qvf- with family(binomial) link(logit)
robust or -glm- with "family(binomial) link(probit) robust" or -qvf-
with "family(binomial) link(probit) robust" gives the best (or the least
problematic) estimation for a GLM with endogenous explanatory variables
while modelling fractional dependent variables.
Thanking you in advance,
Rijo John.
--
***************************************************
Rijo.M.John,Research Scholar
Indira Gandhi Institute of Development Research,
Film City Road, Goregaon East,
Mumbai, India-400065.
contact: (+91)9892412476
URL: http://rijojohn.bizhat.com
*
* 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/
