I have an experiment where I have placed the same treatments on two different shelves; the placement on the shelves is not an independent treatment, but rather a space constraint.
I do not want to treat the data from the treatments the same way because I expect there to be variability due to being on different shelves which may have different conditions.
How can I go about analyzing the data in a way that would account for this variable? Would it be a crossed effect? Or should I analyze the treatments separately going forward?
Thanks in advance for the any help or suggestions.
Go to Solution
Do you think that the difference between the shelves is a fixed effect (reproducible) or a random effect?
If the treatments were randomly assigned to the shelves, then the shelves act as a block. It is just a matter of determining if the block is a fixed or random effect.
If the placement of the treatments was not random, then they might be treated as whole plots.
I'm guess it is the former case.