Statalist The Stata Listserver


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

RE: st: RE: standardized parameter for dichotmious variable in a regr


From   "Maarten Buis" <M.Buis@fsw.vu.nl>
To   <statalist@hsphsun2.harvard.edu>
Subject   RE: st: RE: standardized parameter for dichotmious variable in a regr
Date   Tue, 22 Aug 2006 15:53:59 +0200

--- stata_user asked:
>  Is this means that we cannot take into account of the standardized
> coefficient of a dichotomious variable, I want to say that we can only
> compare between the standardized coefficients of continuous variables in
> order to see which one the most associated with the dependent variable

You can still add dichotomous variables to regression when you want to 
report standardized coefficients. So you can still take dichotomous 
covariates into account.

However, I think that the entire concept of looking at which independent 
variable is most associated with the dependent variable is often given 
way too much emphasis. Say you have two continuous independent variables: 
age and income. By looking at the standardized effects you claim that a 
standard deviation change in age is comparable to a standard deviation 
change in income. Admittedly standard deviations are more comparable than 
euros and years, but it is still not a very convincing comparison. 

In cases where we have a theoretical interest in which covariate has the 
strongest effect, the theoretical concepts underlying the covariates 
usually already have a comparable metric (because how would we form such 
a theory if they didn't?). In that case you should use that theoretical
information to create comparable variables (i.e. standardize you variable).

HTH,
Maarten

-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology 
Vrije Universiteit Amsterdam 
Boelelaan 1081 
1081 HV Amsterdam 
The Netherlands

visiting adress:
Buitenveldertselaan 3 (Metropolitan), room Z434 

+31 20 5986715

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


*
*   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