Biostatistics and Medical Informatics Department Seminar - Joseph Ibrahim, UNC Chapel Hill

Dr. Ibrahim will discuss how to use partial-borrowing scale transformed power prior (straPP) and the generalized scale transformed power prior (Gen-straPP) for accommodating settings in which the historical data and the current data involve different data types. straPP rescales the parameter vector from the historical data to alight with the new data. Gen-straPP adds robustness when the scaled parameters are not equal. He will discuss the advantages of straPP over other common priors.