James Stamey
Baylor University, USA
Title: Inference and sample size determination for clinical trial with rare events and potential misclassification
Biography
Biography: James Stamey
Abstract
We consider the case of comparing two or more proportions to a control in the case where interest is in an event such as an adverse event where the probability of occurrence is low. Also, we allow for potential misclassification in the assay. We investigate several different properties of a Bayesian inference procedure via simulation. The properties investigated include the following. We determine the bias and interval coverage for various non-informative priors to determine which prior is best. We also investigate the impact of misclassification on the bias and power. Finally, we overview a simulation based sample size determination procedure.
Speaker Presentations
Speaker PPTs Click Here