s-news
[Top] [All Lists]

GEE and intra-class correlation

To: List 0-S-Plus <s-news@lists.biostat.wustl.edu>
Subject: GEE and intra-class correlation
From: Emil Coman <emilcoman@yahoo.com>
Date: Mon, 22 May 2006 10:10:39 -0700 (PDT)
Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; h=Message-ID:Received:Date:From:Subject:To:MIME-Version:Content-Type:Content-Transfer-Encoding; b=3RGiojfhDJDfirEgefF6ZcF92jGjBqaSB267EkF1xDSh1ELt1wFyRHI+ZlKZP0xoTYay82ryzCK33qTE3839npyqtpjpMqu2iJWnZEyqH5IHtJTJPkAJrM87h+o/oCrnznhZdAOdTbylE6mezidSpvtd1/eiCVc3TvDcwqiArZs= ;
now that I got some responses to a basic question, here's another one...
I found in an article a reference to the possibility of
getting the ICC in S-Plus: "the intra-class correlation coefficient may be
obtained from the off-diagonal elements of the working correlation matrix
called "gee$worcor" (figure 3, p. 1274, in
Paul Burton, Lyle Gurrin, Peter Sly (1998).
Extending the simple linear regression model to account for correlated
responses: An introduction to generalized estimating equations and multi-level
mixed modelling,
Statistics in Medicine, Volume 17, Issue 11 , Pages 1261 - 1291
).

Does anyone know how to get the value of ICC? I've tried
gee$worcor
and
summary(gee$worcor)
 after running gee, but this does not do it, I only get
NULL
and respectively
Length Class Mode
0 NULL

... thanks, emil

I am using a command like
gee(formula = Y~ X+ Z+ W, cluster = C, variance = "glm.scale", data = "">


Yahoo! Messenger with Voice. PC-to-Phone calls for ridiculously low rates.
<Prev in Thread] Current Thread [Next in Thread>
  • GEE and intra-class correlation, Emil Coman <=