Good morning, thank you for the quick response. I already knew this paper.The situation I have is somewhat different, being in this case more similar to anova with several ways, and can be hypothetically described as follows:3 factors (treatments): site, spacing and dosage (all categorical) Only one response variable: metric In the presented situation, the response variable strongly departs from the assumptions for the realization of an anova of 3 way. There is software that performs what I need, just two-way no more factors, called Past. https://folk.uio.no/ohammer/past/ grato.
... View more