Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: conditional test after logistic regression

From   Maarten buis <>
Subject   Re: st: conditional test after logistic regression
Date   Wed, 14 Apr 2010 16:03:08 +0000 (GMT)

--- On Wed, 14/4/10, Ricardo Ovaldia wrote:
> The article that I am reading states the null in terms of
> beta - Ho: (Bx|By)=0 
> Would your answer change if y=x^2?
> So that the model would be:
> logistic case x x2
> where x2=x^2
> and the null is now,  Ho: (Bx2|Bx)=0 ?

Ok, I think I see what you mean. I would then state the H0 as
(Bx2|Bx=0)=0 or probably more interesting is (Bx|Bx2=0)=0. 
The latter hypothesis would mean whether there is an effect, 
given that the effect is non-quadratic. You could test that 
as follows:

logistic case x 
test x

Hope this helps,

Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen


*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index