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 at the end of May, and its replacement, statalist.org is already up and running.


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

Re: st: interpretation of exponentiated standardized coefficients in OLS


From   Maarten buis <maartenbuis@yahoo.co.uk>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: interpretation of exponentiated standardized coefficients in OLS
Date   Wed, 14 Apr 2010 05:35:42 -0700 (PDT)

--- On Wed, 14/4/10, Sergi VIDAL TORRE wrote:
> Suppose I have a standardized coefficient in an OLS
> regression. I now (for some obscure reason) wouldlike to
> eponentiate it in order to interpret the value as a
> percentual change in the dependent variable (as an
> odds-ratio in logisitc regression). Is is possible to
> interpret the values aresulting from such a transformation
> as an analogous change in percentages?

As a rule, exponentiating the coefficients does _not_ lead
to the interpretation of factor change in dependent variable
for a unit change in the independent variable. 

There are exception:
1) you used a log() inside link function, this is for instance 
what happens when you do a logit, where the link function is
the log of the odds.
2) you have log transformed the dependent variable prior to 
estimation your model. HOWEVER, you are now modeling the 
conditional geometric mean rather than the conditional
arithmatic mean, this is usually _not_ what you want. 

Under normal circumstances you would want to use -glm-
with the -log()- link, see for instance this post from 
earlier today:
<http://www.stata.com/statalist/archive/2010-04/msg00726.html>

If you are really sure you want to model the conditional
geometric mean, you can look at:
Roger Newson (2003) "Stata tip 1: The eform() option of regress"
The Stata Journal, 3(4):445.
<http://www.stata-journal.com/article.html?article=st0054>

Hope this helps,
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