Skip to content

linear mixed effect model with ordered logit/probit link?

3 messages · byoung-inn bai, Arturo Coral Alamo, Spencer Graves

Original
byoung-inn bai
# 
Hello,

I'm working on the multiple categorical data (5-points scale) using linear 
mixed effect model and wondering if anyone knows about or works on the 
linear mixed effect model with ordered logit or probit link.
I found that the "lmer" function in R is very flexible and supports 
various models, but not ordered logit/probit models. I may conduct my 
analysis by turning my DVs into nested dichotonomies, but just wonder if 
there is anyway that I can do this without transforming my DVs. Any help 
or suggestion will be greatly appreciated.

Thanks,

Byoung-Inn



Byoung-Inn Bai
Ph.D. Candidate
University of Washington
Department of Political Science
Box 353530
e-mail: bbi68 at u.washington.edu
18 days later
Spencer Graves
# 
Have you received a reply to this?  I haven't seen one.  I just tried 
several things with RSiteSearch.  Searching for "ordered logit" produced 
70 hits.  One of the first 10 indicate, "The CRAN package MCMCpack has 
at least ordered probit." 
(http://finzi.psych.upenn.edu/R/Rhelp02a/archive/38663.html)

	  I haven't used ordered probits or logits or MCMC, but it would seem 
to me that if MCMCpack has ordered probits, it surely should also be 
able to handle mixed effects.

	  If you would like further assistance from this listserve, I suggest 
you read the posting guide (www.R-project.org/posting-guide.html). 
Doing so might increase the chances of obtaining a useful reply quickly, 
especially if your question includes an extremely simple example in a 
few lines of R code that someone can copy from your email into R and 
test a few ideas in a minute or so.

	  Best Wishes,
	  spencer graves
byoung-inn bai wrote: