Re: st: Test of ordered probit vs ordinary probits

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

At 03:26 PM 10/31/2007, Richard Williams wrote:

> Assumptions of ologit & oprobit can also be tested via my -gologit2- command.
To elaborate on that - see example 1 in the gologit2 help.  Slightly 
modifying that example,

use http://www.indiana.edu/~jslsoc/stata/spex_data/ordwarm2.dta, clear
gologit2 warm yr89 male white age ed prst, store(unconstrained) link(p)
gologit2 warm yr89 male white age ed prst, pl lrf store(constrained) link(p)
lrtest constrained unconstrained, stats

This yields an LR chi square test of 56.14 with 12 d.f. The 
assumptions of the ordered probit model are violated.  If you are 
familiar with SPSS Plum, this is the same # that its test of parallel 
lines gives you.

Or, you could do

omodel probit  warm yr89 male white age ed prst

which gives you an approximate chi-square of 55.81 with 12 d.f.

Note that this just tells you that somewhere, someplace, somehow, the 
parallel lines assumption is violated.  The brant test (after ologit) 
and/or gologit2's autofit option can help you to identify where the 
assumption is violated.  This can be much more parsimonious than, 
say, immediately jumping to mlogit every time the assumptions of the 
ordered logit model are found to be violated.




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

*
*   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/