[19135] in s-news-athena
Re: [S] Pr(F) in anova()
daemon@ATHENA.MIT.EDU (Bill Venables)
Tue Sep 7 22:41:14 1999
Message-Id: <199909080230.MAA04740@snowy.nsw.cmis.CSIRO.AU>
To: Doug Moog <dbm3@po.cwru.edu>
Cc: s-news@wubios.wustl.edu
In-Reply-To: Your message of "Tue, 07 Sep 1999 13:41:44 -0400."
<4.1.19990907133810.00941090@pop.cwru.edu>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Wed, 08 Sep 1999 12:28:43 +1000
From: Bill Venables <venables@acland.qld.cmis.csiro.au>
> I'm running Splus 2000, Windows 95, and I'm confused about what I'm reading
> in the output of anova() applied to a linear model, lm(). I will run a
> model with one or more terms and get a certain set of F and Pr(F) values.
> Then when I add a new term to the model, the F values of the previous terms
> increase, and their Pr(F) values decrease. But since the terms are added
> sequentially, shouldn't the F values remain the same when you add a new
> term? How could the earlier terms be affected by the later terms? Or is it
> not sequential after all? The anova() output says "Terms added sequentially."
Almost anything can happen as you add new terms, even in a sequence of
sequential anovas, because the denominator of the F-statistics is always taken
from the largest fitted model and that can go up or down depending on how the
reduction in the numerator SSq compares with the reduction in the denominator
degrees of freedom.
Linear models, for all its (genuine) simplicity can be a tricky and subtle
business and in my (highly opinionated) view trying to find you way round a
linear model using sums of squares and anova tables is to run a real risk of
missing the whole point. Just a note of caution...
--
-----------------------------------------------------------------
Bill Venables, Statistician, CMIS Environmetrics Project.
Physical address: Postal address:
CSIRO Marine Laboratories, PO Box 120,
233 Middle St, Cleveland, Queensland Cleveland, Qld, 4163
AUSTRALIA AUSTRALIA
Telephone: +61 7 3826 7251 Email: Bill.Venables@cmis.csiro.au
Fax: +61 7 3826 7304
-----------------------------------------------------------------------
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
