出版社:Dep. of Statistical Sciences "Paolo Fortunati", Università di Bologna
摘要:The Skew-Normal (SN) class of densities was introduced by A. Azzalini in 1985. His aim was to determine a statistical model, including normal densities, for handling a precise kind of departure from normality the skewness. In this paper we show the "strange" behaviour of the SN class from the point of view of likelihood based inference. Finally we compare the usual frequentist approach with a Bayesian analysis proving that the former corresponds to a "strange" prior distribution of the parameters of the SN densities.