For my greenhouse experiment with potted trees, I am planing a fractional factorial design. The experiment has six factors (treatments) each with two factor-levels, and I want to use a model/package, which creates me the right experimental design (factor combination; blocking). Currently, I am using the package ‘planor’, but any other package would be also fine to solve my problem. Here some information about the experiment:
Factors 1 to 5: are nutrients (factor levels: present/absent) and leaf wettness (wet/dry).
Factor 6: temperature (ambient/hot)
Since I have to simulate the temperature/hot factor in a climate chamber, including the combinations with factors 1-5, I have to consider the temperature-factor, as a blocking-factor. My problem: treatment-, and blocking-factor are combined. I can use two climate chambers resulting in 4 blocks: block 1 = hot, block 2 = hot, block 3 = ambient, block 4 = ambient.
So far as I know, it should be possible to specify this by using the hierarchy-, and base-arguments of planor (see below). However, I was not able to combine block and temp. Maybe I am using also the wrong model to describe my experiment properly? For the statistcal analysis I prefer to use a liner mixed-effect model
Thank you in advance, Uli
Here, the R code:
trial.fac1<- planor.factors( factors=list( Block=c(1,2,3,4), N=c("low","high"), Mg=c("low","high"), Mn=c("low","high"), S=c("low","high"), temp=c("ambient","hot"), wet=c("dry","wet"), hierarchy=(~temp/Block) ))
key1 <- planor.designkey(factors = trial.fac1, model = mod1, nunits = 32, base = ~Block+temp )
Des <- planor.design(key1, randomize = ~temp/Block)