I want to fit a model with both fixed and random effects but has a binary DV. JMP gives me options for Nominal Logistic / GLM but announces an error when I try to fit the model. Any suggestions are welcome? I figure that JMP must be able to do something at least equivalent to this if it can do do choice models (which this is, but Im having problems getting that function to work with my design).
I have used JMP 7.0 and know that one cannot use Random effects with logistic regression in that program.
1. Is this still the case for JMP 10.0?? Just want to know since I would consider upgrading (purchasing) if it does this, but definitely not if it does not.
2. Also, will JMP give model selection critieria such as AIC values?
Thank you in advance!
i also didn't like the fact that the logistic regression platform can't take a random effect.
as for deer&dog's question, the logistic regression platform give AIC and BIC by default but in the linear regression you will need to ask for it as you can see in the attached output. i produced it from the "big class" data file.
Thanks Ron! I'll check my output (just starting an analysis) to see if Jmp 7 produces AIC and BIC. I am amazed that JMP 10 doesn't handle random effects in the log regression. That has been lacking for years now. Bummer.
Can JMP Pro 13 fit a model with both fixed and random effects but has a binary dependent variable?