Message-ID: <20080402103622.M550@oikos.unam.mx>
Date: 2008-04-02T10:39:09Z
From: Leonel Arturo Lopez Toledo
Subject: how to know if random factors are significant?
In-Reply-To: <4949c7e60804020035g2934b01fx500ee31f8a1b1664@mail.gmail.com>
Dear all:
Thank you very much all for your comments. I?m trying everything you said and checked
previous messages from Marianne, John and Douglas Bates. I found that removing the
random factor reduced the p-values for fixed effects (as John mentioned) and this new
model has not better significant power indicated by the likelihood ration test.
So, even with a low Variance the random factor must be used in a mix model effects.
Thank you very much for your help. I'm understanding and learning so much with the "R-
help lists".
Leo
--
Este mensaje ha sido analizado por MailScanner
en busca de virus y otros contenidos peligrosos,
y se considera que est? limpio.
For all your IT requirements visit: http://www.transtec.co.uk