Locally most powerful invariant tests for departures from exponentiality with incomplete data (Q1085540)

This paper aims to introduce a somehow simpler derivation of the locally most powerful scale and/or location invariant test for \(H_ 0: \theta =\theta_ 0\), where \(\theta\) is a shape parameter ranging over an open set \(\Theta\subseteq {\mathbb{R}}\) and \(\theta_ 0\) corresponds to a simplifying assumption. As an application, locally most powerful scale invariant tests for exponentiality versus some typical failure distributions are obtained, based on a type II censored sample.