1 paper · 1 filter
Luke Hagar, Paul Gustafson
To design many Bayesian studies, the sample size is chosen to attain sufficient power to reject a null hypothesis or a desired probability that an interval estimate is sufficiently…