The cutpoints in Stata's -oprobit- give the correct probabilities for
the categories for the case of all covariates equal to zero (including a
constant term). Note there is no constant in the -oprobit- results. (To
see this, try -oprobit- with no covariates). Sure, you can make the
first cutpoint zero, but then you have to have a non-zero constant in
the model. It's the same result, either way. By the way it's not the
length of the intervals that matters - it is PHI(c_k) - PHI(c_k-1) that
gives the probbailities, where PHIis the standard normal CDF.
Al Feiveson
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Rachel
Sent: Monday, April 30, 2007 10:25 AM
To: statalist
Subject: st: Number of cutpoints in -oprobit-?
My question is about the number of cutpoints identified in an ordered
probit model with J potential outcomes. According to the manual,
Stata's -ologit- and -oprobit- commands identify J-1 cutpoints.
To take the example [R] ologit: suppose that a car can have 5 categories
of repair records, from Poor to Excellent. In this example, -ologit-
estimates 4 cutpoints, where /cut1=-2.77. The manual states that "For
domestic cars, the probability of a poor record is the probability that
u-j<-2.77.
My question is why oprobit identifies /cut1 at all, rather than
normalizing it to zero as certain other statistical packages (such as
Limdep) do. My understanding is that the ordered probit/logit is
identified by the relative lengths of the intervals--the cutpoints
themselves have no meaning. And since the length of the interval between
infinity and the first cutpoint is always infinite, why is this first
cutpoint identified at all?
Thanks in advance,
Rachel
*
* 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/
*
* 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/
