Dear,
Does somebody has an example code of proc mcmc for a hierarchical random effects model of at least 3 levels? I want to fit such a bayesian model to data with proc MCMC but can't find really how to. My model looks like this:
Y_{ijk} | \mu_{jk} ~ N(\mu_{jk}, \sigma^2_w)
where \mu_{jk} | m_k ~ N(m_k, \sigma^2_{b2})
and m_k ~ N(\beta, \sigma^2_{b1})
One could, for instance, take m_k as a school effect, \mu_{jk} as a nested class effect (within the school k) and i=1,...,n_{jk} as the number of observations (students) in class j of school k. The eventual parameter of interest is mainly \beta. All help will be greatly appreciated!
Thanks in advance,
Bert
