[R] Nested ANOVA in R

Majid Iravani majid.iravani at wsl.ch
Thu May 3 11:17:42 CEST 2007

Dear R users,

I installed an experiment as following setup:

- Four plant species
- Seed addition as main factor with two levels: control and seed addition.
- Four replicates (sites)
- 8 plots in each site arranged as two rows (each rows 4 plots) one meter 
far from each other
- Four sub-plots in each plot (therefore 32 sub-plots in each site)
- In the four plots in each site (first row of plots) seeds of each of four 
plant species were sowed onto one sub-plot by random (therefore 4 sub-plots 
sown by a plant species in four different plots in each site).
In the other four plots (second row of plots) one sub-plot selected by 
random as control for each of four plant species (therefore 4 sub-plots in 
four different plots in each site as control for a plant species).

Now, my first question is about the name of this design. I thought this is 
a 3-factor split-plot arrangement.

My second question is about data analysis for this experiment in R. I would 
like to test the effect of seed addition and sites on the seedling 
emergence of each plant species separately using a three- way ANOVA (??). 
Because of the hierarchical experimental design, the experiment must be 
analyzed by nested ANOVA but I do not know how I must arrange dataset and 
then how to write ANOVA model in R.

I would greatly appreciate if somebody gives me some comments on these cases

Thank you very much in advance,

  Majid Iravani
  PhD Student
  Swiss Federal Research Institute WSL
  Research Group of Vegetation Ecology
  Zürcherstrasse 111  CH-8903 Birmensdorf  Switzerland
  Phone: +41-1-739-2693
  Fax: +41-1-739-2215
  Email: Majid.iravani at wsl.ch

More information about the R-help mailing list