Date:         Wed, 8 Dec 2004 14:22:50 -0500
Reply-To:     "Li,Qian [Ontario]" <Qian.Li@EC.GC.CA>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         "Li,Qian [Ontario]" <Qian.Li@EC.GC.CA>
Subject:      Validating a regression model
Content-Type: text/plain; charset="us-ascii"

Dear SAS-L,

I am using multiple stepwise regression to develop a prediction model.
The procedure I am using is Proc Reg. Dataset is separated into
Development_data and Validation_data.

proc reg data=development_data outest=beta;
        model pollutant = &predtors/vif
                                            selection=stepwise
                                    slentry=0.05
                                    slstay=0.05
                                            rsquare;
run;


My question here is, instead of using the "outest" file to test the
validation dataset, is there any "better" way to validate the model
using a new dataset? Or is there any similar statement such as SCORE
statement in Logistic Procedure?

Thanks,
Qian