The case for a fully robust hierarchical Bayesian statistical analysis of clinical trials

View/ Open
Date
2009-06Author
Pericchi Guerra, Luis Raúl
Pérez, María Eglée
Metadata
Show full item recordAbstract
Two clinical trials (past and present), conjugate priors are compared
with (Robust) Cauchy- Student-t and what we call Berger's prior.
The behavior of the latter Robust Bayesian methods is
qualitatively di erent from Conjugate Bayesian Methods and
arguably much more reasonable and acceptable to the practitioner
and regulatory agencies.