Model fit is singular - LMM
I've been advised to use a LMM because my data is binary and I'm completely lost. My dependent variable is Recall (binary). Each participant (n=30) was shown the same words and then split into groups (a-e) to counterbalance my colours. I have text, background and timepoint as my fixed effect variables and have group and participant in my random effects grouping factors. I was told my analysis wouldn't run with interaction effects so I've removed them but I keep getting this warning now and I'm not sure how to fix it. Any help at all would be appreciated!!
Comments
I will forward this to our expert but what I would do is start with the very simplest model, and keep adding the most important factors until the model stops running. I suspect the data are too sparse to support a relatively complex model (as is often the problem when random effects are added).
EJ
Oh I see Frantisek has already addressed this.