Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


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

Re: st: Re: Stata logit interaction,


From   Maarten buis <maartenbuis@yahoo.co.uk>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Re: Stata logit interaction,
Date   Tue, 26 Apr 2011 16:41:36 +0100 (BST)

--- On Tue, 26/4/11, daniel klein wrote:
> Not the main topic of the question, but the claim ...
> 
> "[...] that by centering X1 and X2, we can allieviate
> multicolinearity problem. "
> 
> seems to be wrong 

That is true. It only reduces collinearity in a very special 
case of interaction variable, a variable interacted with
itself, i.e. a squared term. I guess that this is the source
of that misunderstanding. 

I did not want to go into this because as you said it is a 
minor point and because it is still very meaningful to center 
variables when you have interactions (or at least make sure 
that the value 0 represents a meaningful number within the 
range of your data). Once you do that you can more easily 
interpret the main effects: the effect of X1 when X2 is 
average. Imagine X2 is age; it is typically more interesting 
to look at the effect of X1 for an average aged individual 
than at the effect of X1 for a newly born baby... (things get 
even more interesting when X2 is year of birth).

-- Maarten

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

http://www.maartenbuis.nl
--------------------------

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index