[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: gologit2

From   Richard Williams <>
Subject   Re: st: gologit2
Date   Tue, 15 Apr 2008 15:28:35 -0500

At 10:58 AM 4/15/2008, Christian Bustamante wrote:
Thanks for your responses, but I need to understand this right, so I
have more questions....

If the problem is in only a small subset of variables, it could be
reasonably use gologit? (I suppose, that you can find these variables
with the -brant, detail- command.
Ideally, you have some great theory about when and where and why the proportional odds assumption is violated. In the absence of a great theory, you can use mindlessly empirical means, such as the brant test, or gologit2's own -autofit- option. I often describe autofit as the lesser of three evils: while you don't like to use mindless empiricism to make your decisions, you also don't like to have models whose assumptions are clearly violated (ologit) or models that may have far more parameters than is necessary (mlogit).

When I did the Brant test, some chi-squares values of variables
appears as negatives (and p-value equal to 1), how can in be possible?
I've never seen this before.  You might post your commands and output.

Which representative author has written about this estimation technique?
I don't know how representative I am, but like Maarten suggested before there is a lot of material on my gologit2 support page, including stuff I have written and links to other sources. Again the link is

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

* For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index