[R-sig-ME] non-singularity with glmer() in a logit mixed model

[Fernando Pedro Bruna Quintas]

Hi, and thank you for your idea, David Duffy. I am sorry if I was not clear. My use of the expression "more restrictive definition" was not in the sense of three categories of the same variable. I tried to say that I have two dependent binomial variables and one  them has less ones. The reason I mentioned that is because they are about a similar phenomenon, that is why I want to compmare the results and the reviewer ask us to use the same explanatory variables even if they are not significant. Indeed, variable y.A is "R&D" and y.B is "process innovation". There is all possible combinations of values for the two variables in different firms: firms with 0-0 in both variables, 1-0, 0-1, or 1-1. Therefore, I need to estimate two models and compare the results. I am having troubles with the model for R&D (in a developing country!).

Best regards

Fernando

________________________________
De: David Duffy <David.Duffy using qimrberghofer.edu.au>
Enviado: lunes, 15 de febrero de 2021 6:14
Para: Fernando Pedro Bruna Quintas <f.bruna using udc.es>; r-sig-mixed-models using r-project.org <r-sig-mixed-models using r-project.org>
Asunto: Re: non-singularity with glmer() in a logit mixed model

> The focus of the paper is to compare the results of two dependent variables, Y.A and Y.B for firms nested in 24 regions.
> I am estimating mixed logit models [...] the definition of Y.A is much more restrictive though more interesting than the definition of Y.B.
> Therefore, for Y.A there are far fewer ones for firms in the 24 regions.

I don't know if this is helpful, but you might try a multinomial (or even ordinal, if that is appropriate) outcome (no, y-loose, y-strict) GLMM. There are
a few suitable R packages (eg MCMCglmm).

	[[alternative HTML version deleted]]