Hi All,
I was reading a stack overflow, which seemed to indicate that that if the difference of the likelihoods of two GLMER models is close, that the boundary (singular) fit warning may be a 'false positive'. I was wondering if folks could verify if this true and if anyone knows of a test that can be run to see if the warning is a false positive or if there is a way to evaluate the extent to which the issues leading to warning are impacting the quality of the model? Thanks in advance for any advice.
Best,
Lucas