[19157] in s-news-athena
[S] lme Confidence Intervals
daemon@ATHENA.MIT.EDU (Steven.Delean@jcu.edu.au)
Wed Sep 8 20:56:22 1999
Message-Id: <3.0.6.32.19990909105332.00c777b0@pop.jcu.edu.au>
Date: Thu, 09 Sep 1999 10:53:32 +1000
To: s-news@wubios.wustl.edu
From: "Steven.Delean@jcu.edu.au" <Steven.Delean@jcu.edu.au>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Dear S users,
I am using lme, with a beta spline basis for the fixed effects, to model
mammal growth, and would appreciate any comment on the estimation of
simultaneous confidence intervals for the mean profile. I am aware that
there are some difficulties in estimating CI's incorporating both sources
of variance, but would be grateful for any information, including the
current state of the literature. I have investigated using Frank Harrell's
rm.boot function (in Hmisc) to obtain bootstrapped CI's which account for
within subject correlation structure. Has anyone used, or see any problems
with that approach? I will summarise to the list.
Many thanks,
Steve
[S+ 3.3, Win 95, nlme 3.1]
<><><><><><><><><><><><><><><><><><><><><>
Steven Delean
PhD Student / Statistical Consultant
School of Tropical Environment Studies and Geography
James Cook University
Townsville
QLD. 4811
Australia
Telephone : 07 - 4781 5560
Fax : 07 - 4781 4020
steven.delean@jcu.edu.au
-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news
