[R] split-plot design?

[ripley@stats.ox.ac.uk]

On Fri, 17 May 2002, Jerome Goudet wrote:

>
> Dear R-gurus,
>
> We are planning an experiment to test if plants produced by selfing are less
> fit than those produced by outcrossing.  We have plants from three different
> alpine valleys, picked randomly among all the possible valleys. In each valley,
> we have a number of individuals, also picked at random. seeds from this
> individuals were brought back to the green house and sawned. when they
> flowered, five of the flowers were selfed and five outcrossed. the number of
> seeds produced by each flower was then recorded.
>
> Would you all agree that valleys and individuals are random factors? If so,
> would you also agree that the data should be analysed as a split-plot design:
>
> seeds.nb~valley+ind%in%valley+treat+treat:valley+treat:ind%in%valley

No, no.  I don't quite understand the sampling, but this is definitely not
a split-plot design (and there is only one treatment).  And split-plot
experiments are usually analysed conditional on the plots, just as
randomized block designs are.

If I understand aright, you have three valleys, which are blocks, and ten
flowers from each divided into two treatments.  Or perhaps you have
three valleys, several plants from each, and ten flowers grown from each
plant.  The plants would still be the blocks.

Given your responses are counts, I think your main worry will be a
suitable modle for count data: negative binomial, perhaps?


> Is there a way in R to specify which factor are random? Is there a way in R to
> specify which MS should be used a a denominator?

Error(), which is used for split-plot designs.

>
> Thanks a lot for your enlightenments and  best wishes.
>
>
> Jérôme Goudet
> Inst. Ecology
> UNI Lausanne
> Switzerland
>
> -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
> 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
> _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
>

-- 
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._