p really only varies among species (years, reps). I think Zipkin had observation level covariates associated with strata (j) and substrata (k). I don't. All p are the same here. So this piece of the model could be simplified. This would have 2 benefits:
faster model runs
might make it easier to fill all nodes of p, even when k isn't same for all j
preally only varies among species (years, reps). I think Zipkin had observation level covariates associated with strata (j) and substrata (k). I don't. All p are the same here. So this piece of the model could be simplified. This would have 2 benefits:p, even whenkisn't same for allj