|Date: ||Wed, 14 Jul 2004 09:39:05 -0700|
|Reply-To: ||e dunbar <email@example.com>|
|Sender: ||"SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>|
|From: ||e dunbar <firstname.lastname@example.org>|
|Subject: ||Re: Comparing Independent Regression Coefficients|
|Content-Type: ||text/plain; charset="us-ascii"; format=flowed|
I just did this in a study.
I am interested in the thoughts of folks on this list about whether it is
better to use
standardized or unstandardized betas in computing the comparisons?
At 06:26 AM 7/14/2004, you wrote:
>From: Hector Maletta <email@example.com>
>Subject: Re: Comparing Independent Regression Coefficients
>I do not remember whether there is or not a way to compare the coefficients
>coming from different regressions. However, one different possibility is
>including in the regression a dummy variable to identify the two groups
>(coded 0 for cases in group 1 and 1 for cases in group 2) and then evaluate
>the significance of the beta coefficient for the dummy variable. This is not
>strictly equivalent but may be more illuminating.