Re: st: IV estimation in a negative binomial framework

[Austin Nichols <austinnichols@gmail.com>]

Arka Roy Chaudhuri <gabuisi@gmail.com>:
What do you mean by "in my case qvf seems to perform better than
ivpois" exactly?  Are you running a simulation with a known effect?
-qvf- is for measurement error only, but my simulations prefer -ivpois-
for every type of endogeneity, including measurement error.
Note that -gmm- supplants -ivpois- if you are running modern Stata.

On Fri, Aug 12, 2011 at 4:31 PM, Arka Roy Chaudhuri <gabuisi@gmail.com> wrote:

> I had a question about -ivpois- vs -qvf-.Is there any reason for
> -ivpois- to be theoretically better than -qvf-? I looked at the
> example in the ivpois help file which shows that ivpois performs
> better than qvf but in my case qvf seems to perform better than
> ivpois. So I was wondering if there are any theoretical reasons for
> why ivpois is better than qvf and the essential differences between
> the two.
*
*   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/