Hierarchy of Statistical Goals

February, 2003

Hierarchy of Statistical Goals: 17

Bayesian Inference

The criticism of Bayesian inference is that we usually do not know the prior probability of a beneficial treatment

As we have seen, the predictive values are very sensitive to the choice of prior distribution

Possible remedies:

·Use data from previous experiments

·Use subjective opinion or consensus of experts

·Do a sensitivity analysis over many different choices for the prior distribution

·Use frequentist approaches