[R] aov and missing values
Prof Brian Ripley
ripley at stats.ox.ac.uk
Fri Nov 3 15:31:36 CET 2000
> Date: Fri, 03 Nov 2000 16:10:54 +0200
> From: "Pedro J. Aphalo" <pedro.aphalo at joensuu.fi>
>
> I am learning R, and although I have looked in the documentation, I may
> be asking something obvious. Sorry, if that is the case.
(No, the problem is with your expectations from the theory.)
> In a split-plot design if there is a missing subunit summary gives me a
> table with two rows for the same factor, one in the error within section
> and one in the section using error between units. With no data missing
> the table is "normal". How does one interpret the table when data is
> missing?, or is it that aov cannot cope with missing values in this
> case?
aov() does cope, but the conventional analysis is indeed as you
describe. Factors do have effects in more than one stratum. For example,
consider the oats example in Venables & Ripley (1999). If I omit
observation 70, I get
> summary(oats.aov)
Error: B
Df Sum of Sq Mean Sq F Value Pr(F)
Nf 1 841.89 841.890 0.2242563 0.6605042
Residuals 4 15016.57 3754.142
Error: V %in% B
Df Sum of Sq Mean Sq F Value Pr(F)
Nf 1 1100.458 1100.458 1.747539 0.2187982
V 2 1156.821 578.410 0.918521 0.4335036
Residuals 9 5667.471 629.719
Error: Within
Df Sum of Sq Mean Sq F Value Pr(F)
Nf 3 19921.29 6640.431 37.23252 0.0000000
Nf:V 6 408.96 68.160 0.38217 0.8864622
Residuals 44 7847.41 178.350
so Nf appears in all three strata, not just the last one.
The `recovery of intra-block information is needed'.
If you have an unbalanced layout (e.g. with a missing value) use
lme to fit a model. V&R do this example in lme too.
--
Brian D. Ripley, ripley at stats.ox.ac.uk
Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/
University of Oxford, Tel: +44 1865 272861 (self)
1 South Parks Road, +44 1865 272860 (secr)
Oxford OX1 3TG, UK Fax: +44 1865 272595
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
More information about the R-help
mailing list