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

Re: st: gologit2

From   Maarten buis <>
Subject   Re: st: gologit2
Date   Mon, 14 Apr 2008 08:41:47 +0100 (BST)

--- Christian Bustamante <> wrote:
> I'm doing a research using the ologit command. After the estimation,
> I ran the Brant test and the Approximate likelihood-ratio test to
> verify the parallel regression assumption. The result was that isn't,
> because the chi-squared value of the test was too high (I'm right, am
> I?).

If you are unsure, than go through the logic of testing: formulate the
null-hypothesis. The help-file of -brant- (part of the user written 
-Spost- package) states: "-brant- performs a Brant test of the parallel
regression assumption (also called the proportional odds assumption)
after ologit." This means that the null-hypothesis is that the parallel
regression assumption holds. The p-value gives you the probability of
finding a sample which deviates from this assumption as much or more if
the null-hypothesis is true. If it is smaller than 5% (=.05) then we
say that that is so unlikely that we probably haven't drawn a "weird"
sample but that the null hypothesis is wrong. 
> So, I found a web page which says that it happened, you should use
> the Generalized Ordered Logistic Model (gologit2 command). My
> questions are:
> 1. Is it the right solution?
> 2. Can I verify the parallel regression assumption? Or it doesn't
> matter at this point?
> 3. Which postestimation commands will be useful?

These questions are answered in detail here:

Hope this helps,

Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands

visiting address:
Buitenveldertselaan 3 (Metropolitan), room Z434

+31 20 5986715

Yahoo! For Good helps you make a difference
*   For searches and help try:

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