Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: comparing biprobit marginal effects CIs


From   "Dimitriy V. Masterov" <dvmaster@gmail.com>
To   Statalist <statalist@hsphsun2.harvard.edu>
Subject   Re: st: comparing biprobit marginal effects CIs
Date   Tue, 16 Apr 2013 13:09:19 -0700

On Tue, Apr 16, 2013 at 12:45 PM, Garrido, Melissa
<melissa.garrido@mssm.edu> wrote:
> Hi Dimitriy,
> I'm not sure how to answer your confidence interval question off the top of my head, but if you use the factor notation with option #1 (i.variable name) on the variable "vote", your AME equals .095 exactly.

Thanks for solving this piece of my puzzle. The factor notation seems
to do the trick:

biprobit (private = years loginc i.vote) (vote = year loginc logptax ), robust;
margins, dydx(vote) predict(pmarg1) force;

I've also made some progress on the other issues. I believe the
problem has to do with the standard error for rho, which is very high
and percolates to the treatment effect. I think that methods (2) and
(3) do not account for the uncertainty in rho. I guess the bootstrap
for the final results is the way to go here.

DVM
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index