Statalist


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

Re: st: How to tell with transformation is better?


From   Maarten buis <maartenbuis@yahoo.co.uk>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: How to tell with transformation is better?
Date   Thu, 28 Aug 2008 14:50:48 +0100 (BST)

--- Richard Goldstein <richgold@ix.netcom.com> wrote:
> I have no idea what you mean by "very low" when referring to the
> value of r-squared

I always tell my students that their choice of romantic partner is
pre-determined by their school, and the social background of their
parents. My students immediately rebel against that. Next, I show a
regression explaining education of partner by own education and
education of the father, which has a r-squared of about .4 and tell
them that this model does not explain about 60% of the variation, and
ask them wether they are happy with this model. The answer is usually
no, and then I explain to them they are being inconsistent... 

In other words a very low r-squared is very likely to be the state of
the (social) world, and not a problem with your model.

-- Maarten

BTW if I had to classify an R-squared of .4, than I would classify it
as very high. The correlation between characteristics of partners is
usually exceptionally strong, most other correlations in the social
sciences are much much weaker. In many other cases I would not be
surprised when seeing an R-squared of only 1%.


-----------------------------------------
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

http://home.fsw.vu.nl/m.buis/
-----------------------------------------

Send instant messages to your online friends http://uk.messenger.yahoo.com 
*
*   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   |   What's new   |   Site index