I wonder how people get these ideas. -intreg-
is emphatically not ordered probit.
It is a generalisation of -cnreg- and -tobit- in
which values of the response can be point
data, interval data, left-censored or right-censored.
Error terms are presumed normal (or Gaussian,
as some prefer to say). The calculation is
standard ML.
I pretty well copied that out of the manual, but
then I see that almost all of it is in the help.
It is really is difficult to do serious work
with Stata without access to the manuals.
Nick
[email protected]
Von Fintel Dieter
> Does anyone have any references to the estimation procedure
> which intreg uses. Some people have told me it is an ordered
> probit with pre-specified cutpoints, some have said it is a
> modified tobit? Is there any way of finding out more details
> - I do not have access to the base manual.
*
* 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/
