search

open-AIMS / bayesnec

Bayesian No-Effect-Concentration estimation in R
https://open-AIMS.github.io/bayesnec
Other
11 stars 6 forks source link

Consider adding decision rules to fit beta_binomial and negativebinomial to overdispersed binomial and poisson data automatically #149

Open beckyfisher opened 1 year ago

beckyfisher commented 1 year ago

At the moment bayesnec requires the user to examine the dispersion estimate and refit the model manually calling the appropriate family to handle the over dispersed case. Given the decision criteria are relatively clear in this case (dispersion significantly different from 1), it would make sense to automate re-fitting using the appropriate family.