Date: Wed, 26 Jan 2005 16:19:06 -0500
Reply-To: Jeff Mather <Jmather@harthosp.org>
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: Jeff Mather <Jmather@harthosp.org>
Subject: using beta's to validate
Content-Type: text/plain; charset=US-ASCII
I have a patient population I would like to use to validate the
accuracy of a logistic regression model and I'm not sure how to do it.
I have the beta's each of the factors in the published model. I know
how to enter a single patients factors into the model, run through the
model equation to result in a predicted mortality, or odd's ratio for
that specific patient.
But I have >2000 patients in a database. I want to validate the
accuracy of the model on that population and compare expected mortality
with observed. I think this is probably a no-brainer, but I'm having a
block here. Hope someone can clear this for me.