Re: st: Constraints in mprobit

[Richard Williams <Richard.A.Williams.5@ND.edu>]

At 08:16 PM 5/25/2006, Emily Clough wrote:
> I can get yours to work, but not mine.  Here's an example of what I've
> been getting (the "Equation[1] not found" error message is the one I get
> consistently):
>
>
> . constraint 1 [1]ref_Chances=0
Try referring to the equation # rather than the value of the dv, e.g.

constraint 1 [#1]ref_Chances=0

There are different ways of referring to equations (value of y, value 
label, equation #), and if one way doesn't work, sometimes another 
way will.  Stata seems to be a little inconsistent across commands as 
to how constraints are implemented, or whether they are even allowed at all.

Incidentally, note that in Rich Gate's example, he used the value 
labels of Y when referring to the equations.  If he had instead done 
the seemingly equivalent

. constraint 1 [2]age = [3]age

. mprobit insure age male nonwhite site2 site3, constraints(1)
(note: constraint number 1 caused error r(303))

you see that he would get an error. (But do the same thing with 
mlogit and it runs fine).

As a sidelight, you'll see that there are several little differences 
between mlogit and mprobit.  e.g. mlogit reports a LR chi-square, 
mprobit reports a Wald chi-square.  mlogit uses the value labels of y 
for the equations, mprobit (sometimes) uses labels like _outcome_1, 
_outcome_2.  They even differ in how you are allowed to abbreviate 
constraints!  I don't think Stata was aiming for perfect consistency 
between the two commands.  mlogit seems to have always had some sort 
of privileged status among Stata commands, e.g. it can do things with 
equation labels that other commands can't do - and my impression is 
that Stata kind of regrets that it did that!  Perhaps StataCorp feels 
that it is stuck with the way it did mlogit but it doesn't have to do 
the same things with new commands.


-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
FAX:    (574)288-4373
HOME:   (574)289-5227
EMAIL:  Richard.A.Williams.5@ND.Edu
WWW (personal):    http://www.nd.edu/~rwilliam
WWW (department):    http://www.nd.edu/~soc 
*
*   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/