I've been playing around with mixed model code over the last few days, and it seems that no matter what I do, some of my lsmeans (for 2 different interactions) always come up as non-estimable. Any thoughts on why this might be happening? Thanks in advance!
↧