Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

RE: st: Log transformation of Dependent and Independent Variables


From   "Nick Cox" <n.j.cox@durham.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   RE: st: Log transformation of Dependent and Independent Variables
Date   Wed, 7 Sep 2005 12:07:51 +0100

I agree, on the understanding that when you say "dispersed" 
you mean "skewed". Taking logs on both sides makes sense 
if the relationship is multiplicative rather 
than additive, but in no sense is it compulsory. 

In addition to this, taking logs of dummy variables 
is problematic as log 0 is indeterminate. With 
any coding of dummy variables as two positive 
values, say a, b > 0, then I guess you would find 
that mapping a dummy to log a, log b boils down 
to the same thing, just messier. 

Nick 
n.j.cox@durham.ac.uk 

Chris Rohlfs
 
> yes, this is a common thing to do.  wage regressions are a 
> good example. 
> economists usually take the log of the dependent variable 
> (wages).  and 
> then some regressors (such as education) typically enter the equation 
> linearly -- including some dummy variables like race or sex dummies. 
> while other regressors (such as parents' income) typically 
> enter in log form.

ALICE DOBSON

> > The dependent variable in my data set is highly dispersed 
> and I intend to 
> > take a log transformation. However, there are a few dummy 
> independent 
> > variables. Can I simply take a log transformation of the DV 
> and a few 
> > dispersed IVs and let the dummy variables and the less 
> dispersed variables 
> > remain as they are?

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   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