Testing difference between two Negative Binomial distributions

[cluhmann]

@crusher083

I reparametrized using Lognormal instead of Binomial, as I converted raw lethal cases to lethal cases per hospitalized patients.

This appears to have been the rationale for moving away from the negative binomial.

@dv_mepi

I would like to use BEST for a negative binomial model but would like to confirm how to apply it.

The original model is basically a BEST model (with negative biniomials instead of t distributions), though it omits all the decision theoretic procedures (defining ROPEs, comparing HDIs to ROPEs, power analyses, etc.) that Kruschke layers on top of the actual estimation/inference.