Date: Thu, 5 May 2005 16:34:55 -0500
Reply-To: "Nick ." <ni14@MAIL.COM>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: "Nick ." <ni14@MAIL.COM>
Subject: Logistic regression---Assumptions
Content-Type: text/plain; charset="iso-8859-1"
I would like to know which of the logistic assumptions are most important in terms of 'real life' modeling. I know that academic data sets usually satisfy the theoretic aspects of any statistical model, but in real life data sets don't behave all that nice. I wish to know which assumptions I can ignore using PROC LOGISTIC and which I must pay attention to. How does one deal with assumptions not being satisfied? You can't just go up to your boss and say, sorry, this data set doesn't meet the assumptions therefore I can't build a model. Your thoughts please.
Sign-up for Ads Free at Mail.com