[R-sig-ME] response correlated among groups

[Rafael Wüest]

Dear list members

I would like to fit a model relating a gaussian response variable y to some continuous predictor variables (p1 and p2). 
The response is correlated among the classes of a factor (p3).

Is it possible to set up a mixed model that accounts for the correlations among classes of p3 in the response?

I would appreciate help on how to code such a model, if possible.

Many thanks in advance,
Rafael

-- 
Rafael Wüest
rafael.wueest at gmail.com
http://www.rowueest.net