Date:         Wed, 5 Jul 2006 12:20:31 -0400
Reply-To:     Peter Flom <Flom@NDRI.ORG>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Peter Flom <Flom@NDRI.ORG>
Subject:      Re: A LOGISTIC REGRESSION question
Comments: To: Ganesh bala <ganeshkumar77@YAHOO.COM>
In-Reply-To:  <200607051441.k65Akcwq000521@mailgw.cc.uga.edu>
Content-Type: text/plain; charset=US-ASCII

I don't see a reason to cut your sample size

Peter

>>> Ganesh bala <ganeshkumar77@YAHOO.COM> 7/5/2006 10:41 am >>>
Hi All,

I plan to build a logistic model to predict response. My dataset
contains
500K observations :

response =.33%
non_response =99.67%


Since I have a disproportional sample of responders vs. non_responders,
I
am thinking that I should be taking a 1:4 split between them
respectively.
Is this correct? Also, should I follow any other procedures (offset??)
while doing so?
Any ideas would be much appreciated.

Thanks,
Nash