I have a problem with a linear lmer model. Indeed, I have several factors that interest me and I want to see if they have an impact on my data. I have 6 of them and for 4 of them, I have no worries. Whatever the order of my factors, it doesn’t change the results. However, for the other two, the result changes according to the place of the two factors, whereas I thought that the order of the factors didn’t influence the results. Results I obtain by performing a type 3 anova (Anova command). These are two related factors: to put it simply, there’s one species factor (n=4) and one colony factor (n=3), and I’ve got 2 species that are only represented in one park each and the other two in two parks. Can this explain the results? And if so, what would you advise me to do? Keep just one?
Thank you for your time.
Jules

I tried to change the order of factors and try different kind of anova but nothing worked.