I have estimated the following mixed model:
mod<-lmer(fights ~ sex*age + neighbour + (1|my random factors)
sex has 2 categories: a ans b, and neighbor has three categories: b, c, d, where b is the same in both factors.
My model was working, but returns the warning message that one column was dropped. I assume that it is one of the two bs. The summary gives me a result for a compared to b, for c compared to d, and for the interaction sex with age. That's all I need.
But I'm not sure if the model has calculated what I think that it was doing. If I would choose another way to get these results, I would need two separate models (one with age * sex and a second with neighbour) - that's not what I want. Can someone tell me if lmer can handle such "tricky" constellations and indeed dropped the column I'm not interested in (b in factor neighbour)?