Statalist The Stata Listserver

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

st: RE: iv probit

From   "Schaffer, Mark E" <>
To   "Michael Barnes" <>, <>
Subject   st: RE: iv probit
Date   Fri, 18 May 2007 08:07:08 +0100


> -----Original Message-----
> From: Michael Barnes [] 
> Sent: 18 May 2007 01:35
> To:
> Subject: iv probit
> Hi:
> I'm trying to do IV probit on Stata, but am having very 
> little success.  My model is:
> y=bo + b1x1 +b2x2 + e, z;  where y is binary (0,1), x1 is a 
> set of binary
> (0,1) rhs variables (both exogenous and endogenous), x2 is a 
> set of continuous variables (both exogenous and endogenous), 
> and z is a set of instruments for the endogenous variables in 
> x2 and x1.  Depending on the specification of the model, it 
> may be overidentified.
> I am having problems doing a two-stage (iv) probit approach.  
> Do you know of a canned module available for stata that will 
> give me correct standard errors and efficient estimates with 
> the set up I am using?

Which version of Stata are you using?  Depending on the answer, it's
either built-in (which suggests you don't have Stata 9) or there's a
user-written IV probit available in the usual way.

findit iv probit

from within Stata should direct you to it.


>  If not, what is the best way to 
> approach this problem?  Should I generate the first stage 
> predicted values seperately using OLS and probit, and then 
> use the generated predicted values in my second stage probit 
> regression?  If so how do I construct my var/cov. matrix?
> Thanks for your help,
> Michael
> _________________________________________________________________
> More photos, more messages, more storage-get 2GB with Windows 
> Live Hotmail. 
> TAGHM_migration_HM_mini_2G_0507

*   For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index