Abstracting model prior and family for hierarchical ODE in U34

[hyunjimoon]

Goal: abstract U34 in this table to the highest level

1. How often does the measurement model (family) change? Can all (or some) changes be covered with prior change (e.g. adding hierarchy; family poisson to neg_binom is the same with gamma prior for rate)?

2. What scenarios does user decide to change prior distribution (not prior parameter)?

3. Could you give some feedback on the following prior type classification (details here)?

4. Confused about brms family quas

quasi(link = "identity", variance = "constant")
quasibinomial(link = "logit")
quasipoisson(link = "log")