Figure 1 of Wilson (2019) When is a harmonic mean p-value not a Bayes factor?, in response to Held (2019) On the Bayesian interpretation of the harmonic mean p-value.
Empirical type I errors for the eight competing methods for each study at nominal level of 0.0001 based on non-normal error terms from a chi-squared distribution with two degrees of freedom.
This package provides the data and calculations for the example used in the article "Bayesian Inference Demystified for Software Engineering Researchers: a Step-by-Step Introduction." It consists of t
