Mixed models >> categorical data >> random effect functionalityA requested functionality under the Mixed models platform when working with Categorical data (binary) would be to be able to fit random effects.
More information from Categorical > Multi-level data > PROC GLIMMIX :
Being requested for multi-level data (Sequential monadic, diary studies, deprivation). Includes random effects for categorical and binary. Currently in SAS.
Statistics and AnalyticsRe: Mixed models >> categorical data >> random effect functionality - Status changed to: Yes, Stay Tuned!@Peter_Hersh we will have a new Generalized Linear Mixed Model personality available in JMP Pro 17.0, available in Fit Model, that can fit GLMMs for binomial and Poisson response distributions. If you are member of the JMP 17 Early Adopter Program you can try this and test it now.
Re: Mixed models >> categorical data >> random effect functionality@SamGardner I know GLMM has had an overhaul in JMP 17 is this now possible?
Re: Mixed models >> categorical data >> random effect functionalityGLMM's will be pretty hard to add. But, Generalized Estimating Equations are much simpler and achieve the same goal in that the inferences (tests and intervals) are robust to within subject correlation. The main difference is that inference with GEE's is at the population level rather than the subject level (no random effects), but from our meeting in October 2017 it sounds like population level inference is what you are after. We can see about adding GEE's as an option to GenReg.Re: Mixed models >> categorical data >> random effect functionality - Status changed to: Open for VotingRe: Mixed models>> categorical data>> random effect functionalityThis was reiterated in Categorical > Multi-level data > PROC GLIMMIX .
Re: Mixed models>> categorical data>> random effect functionality<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />Jeff-
Yes- that is correct. Thank you-
Scott
Re: Mixed models>> categorical data>> random effect functionalityreese.sj ,
Just to clarify, you'd like Fit Model to support Categorical responses when using the Mixed Model personality. Is that correct?
-Jeff
Re: Mixed models >> categorical data >> random effect functionality - Status changed to: Delivered
