nlme: random effects

[Vincent.Duval@bc.boehringer-ingelheim.com]

Dear list,

This maybe a naïve question for some of you but I have been struggling for some time to find out a way to implement the following random effect.

Below is the basic model I have:

Base1.nlme=nlme(DV~Alpha*XX**Beta,

                                data="">   # groupedData depending of the variable ID

                                fixed=Alpha+Beta~1,

                                random=Alpha+Beta~1,

                                start=list(fixed=c(400,0.5)),

                                )

Subjects go four times into the same experiments (VISIT). Therefore I susspect some variability on parameters between occaasions.

Implementation of the varaibility between occasions improved the description of the data.

Base3.nlme=update(Base1.nlme,

                   random=Alpha+Beta~1|ID/VISIT)

but I can't manage to estimate a same variability on Alpha and Beta over the four periods and on top of it an inter-occasion variability for Beta

One of the ways I tried is bellow but didn't succeed more than the others,

Base4.nlme=update(Base3.nlme,

                        random=list(Alpha~1|ID, Beta~1|ID/VISIT))

Could someone give me a hint on how to start with this?

Thanks for your help,

Vincent

PS: I Use Splus7 under windows XP