Statalist


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

st: CMP bicensored tobit


From   Fabian Gouret <Fabian.Gouret@malix.univ-paris1.fr>
To   statalist@hsphsun2.harvard.edu
Subject   st: CMP bicensored tobit
Date   Thu, 30 Oct 2008 20:40:00 +0100

Dear all,

I want to estimate a system of three equations, more precisely three tobits,
where the error term are correlated.  Each tobit is censored at the left and at
the right (i.e. bicensored)

I think we can do that with the command CMP.

To see how the command function, I have tried to replicate only one of the
bicensored tobit equation of my system first with the command tobit, then with
CMP:

tobit Y X, ll ul

cmp (Y = X), ind("cond(Y, $cmp_cont, $cmp_left, $cmp_right)") qui

Unfortunately CMP considers only the censoring at the left.

Does anyone know how to consider a tobit censored at the left and at the right
with CMP?

I have also tried to use the interval censored macro $cmp_int generating a lower
bound Y1=Y with a dot for the lower value of Y, and generating a higher bound
Y2=Y with a dot for the higher value of Y. with the command intreg it gives the
same result as with the command tobit; but it does not function with

cmp(Y1 Y2= X), ind($cmp)

indicating "Fitting individual models as starting point for full model fit.
something that should be true of your data is not"

If someone understand my problem, it would be really great. Thanks a lot for
your kind attention

Fabian



----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.

-- 
Ce message a ete verifie par MailScanner
pour des virus ou des polluriels et rien de
suspect n'a ete trouve.

*
*   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/



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