Date:         Tue, 28 Oct 2008 06:13:06 -0700
Reply-To:     shiling99@YAHOO.COM
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Shiling Zhang <shiling99@YAHOO.COM>
Organization: http://groups.google.com
Subject:      Re: To statisticians: Units,
              standardization and regression modelling
Comments: To: sas-l@uga.edu
Content-Type: text/plain; charset=ISO-8859-1

First, centering variables do not change the values of coefs except
constant
Second, standardized variables may change coefs. All measures will be
at standardized level. It has the advantage to compare the sensitivity
among independent variables (sort of) in the sense of one standard
unit.
Third, when a scale changes, then the corresponding coef will be
factored by the amount of scale. Selecting the proper scale in your
research area will make it easy to understand for everyone else in you
area.

All three are not very important in a regression analysis. Your
research area may prefer one to the other.

HTH

On Oct 27, 4:13 am, "cat.." <cat....@gmail.com> wrote:
> Hi there,
>
> Let's say I want to fit a logistic model Logit P(diabetis) = Age +
> Glucose.
>
> I usually do not ask for standardizing the covariates before entering
> them into the model, only centering them, just to make the
> interpretation of exp(beta) easier (= OR when the covariate is
> incremented by 1 <unit used in the model>). Centering the covariate
> impacts the value of the intercept, reducing them will impact the
> value of their coefficient in the model.
>
> But now, I have a doubt: What about the units ? What senses does it
> make to add years and mg/dl ?
>
> Any thought about that ?
>
> What is your practice in modeling ? Do you always only work with
> standardized variables ?
>
> Thanks to any contributor.
>
> Catherine.