Skip to content
Prev 6607 / 20628 Next

pMCMC and HPD in MCMCglmm

Hi,
when I set the fixed effect priors for the first example with repeated 
measures I  follow the indication of both the coursenote and past post (https:
//stat.ethz.ch/pipermail/r-sig-mixed-models/2010q3/004415.html):

V=diag(n)*((varUNITS+varCLASS)+pi^2/3)

where 
n= number of fixed effects
varUNITS= residual variance (var of e)
varCLASS = variance of random effect (var of Zu)

You suggest  to assume 1 for the residual variance (varUNITS) + ~2 for the 
random variable variance (varCLASS). 

How do you decide to set ~2 for varCLASS? is there an indication about a 
possible range of this value?

Thank a lot

Massimo

 -----------------------
 Massimo Fenati
 DVM
 Padova - Italy
an
of
of
exact
it