turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

- JMP User Community
- :
- Discussions
- :
- Fitting generalized linear mixed models

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Nov 19, 2011 3:06 PM
(1736 views)

Hi everybody,

I'd like to fit a generalized linear mixed model onto my data. It seems that the Fit Model platform does not offer this option- is there another way of doing this in JMP?

Thanks for your help

Philip

9 REPLIES

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Nov 19, 2011 3:58 PM
(874 views)

I am using JMP 7, but it does offer generalized models under the 'fit model' platform.

Fit model, then under "Personality" (at upper right), you find different options including "Generalized linear model". It then gives you the option to specify the the distribution (binomial, poisson or exponential). One can also go directly to logistic regression (binomial or ordinal) as a "personality" within the 'fit model'. Is this what you are looking for, or are you trying to find something else such as a mixed model (GLMM)?

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Nov 19, 2011 4:02 PM
(874 views)

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Nov 19, 2011 4:15 PM
(874 views)

Ok. I looked into this question *a lot* in the past, although my memory of what I found is a bit old. Here's a try.

You can do ANOVA with random and nested effects in JMP. Once you put the variable into the 'x' box, you highlight it and click on 'attributes' and "random" is an option. But that assumes your y variable has a somewhat reasonable normal distribution.

You can't do other GLMM (poison, binomial) with random effects. It looks like you can when you set up the model within 'fit model'. But, when you hit 'run model', it tells you that the generalized linear model does not take random effects. At least that is the case with JMP 7. But I looked into whether I could do this with a more recent version (JMP 9) and believe the answer is still no.

I turned to Stata and SAS for that analysis (survival with families as a random effect). I found Stata easier to use, but it would be helpful to know SAS or R these days. That was a few years ago, but it seems the same versions of JMP are still available.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Nov 19, 2011 4:16 PM
(874 views)

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Nov 19, 2011 4:19 PM
(874 views)

Thanks a lot , it's a pity JMP can't do that (yet). I definitely can't fit GLMMs with my JMP 8.

All the best

P

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Nov 19, 2011 4:42 PM
(874 views)

well if someone using JMP 9 knows otherwise and it can fit a GLMM, let us know

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

May 23, 2013 9:16 PM
(874 views)

Hi,

Any updates on this threat for JMP 10? Looking to run a GLMM.

Thanks in advance.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

May 23, 2013 10:20 PM
(874 views)

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Aug 24, 2016 3:48 PM
(874 views)