José Enrique Chacón Durán , Jesús Montanero Fernández
, Agustín García Nogales
, P. Pérez
An interesting problem in classical statistical inference is that of testing a point (or precise) null hypothesis. From a Bayesian point of view, a major handicap arises when testing a precise hypothesis, as its final probability is zero for continuous posterior distributions. Two Bayesian methods to deal with this problem not implying modifications of the prior distribution are suggested.
© 2008-2024 Fundación Dialnet · Todos los derechos reservados