I know that, when performing power calculation based on logistic regression model, under the regular frequentist approach, power calculations become unstable if it is based on the model that causes perfect separation of the data.

But does the sample rule apply when I am doing the the bayesian power calculation with the logistic regression model? That is, if the logistic regression model that I am working with causes perfect separation of the data, would any bayesian power calculation based on this model be somehow unstable, perhaps by decreasing the standard deviation of the posterior of beta?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.