Lindley's paradox epitomizes the conflict between Bayesian and frequentist evidences in hypothesis testing. We scrutinize the point null approximation assumption using the interval null and the smooth formulations. Along with numerical calculation, under the smooth null formulation, we have found that Lindley's paradox prevails relatively independent of smoothness of the target function. However, it is less drastic as suggested earlier.
© 2008-2024 Fundación Dialnet · Todos los derechos reservados