If it is not too late, I suggest you to download -cmp- from SSC. I quote from the help "It offers far more flexibility in model construction. To take one arbitrary example, one could regress a continuous variable, on two endogenous variables, one binary and the other sometimes left-censored, instrumenting each with additional variables."
Nicola
At 02.33 28/06/2008 -0400, "yu chenyu" wrote:
>Dear Statalist,
>
>I have a question about specifying instruments in -ivprobit-, which
>estimates a binary probit model with endogenous continuous regressors.
>I want to regress Y on the endogenous variable of 'Lnincome' and
>exogenous variables X1 and X2. 'Lnincome' is dependent on instruments
>X2 and X3.
>
>I used the syntax - ivprobit Y X1 X2 (Lnincome=X2 X3) - . It seems
>that Stata 10 treats all variables as instruments for Lnincome, no
>matter whether I used 'twostep' or 'mle' option. So, X1, X2, X3 were
>all treated as regressors for Lnincome. Could you advise me how to
>remove X1 from the instruments? I want X1 to be related to Y, not to
>Lnincome. Are there some other ways to model this? Or, is it
>appropriate for me to estimate Lnincome first and then include the
>predicted Lnincome in the probit model?
*
* 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/
