A robust rank test for location under asymmetry

Autor(es) e Instituição: 
Jimmy Corzo, Universidad Nacional de Colombia
Myrian Vergara, Universidad de la Salle, Colombia
Myrian Vergara

We propose a winsorized adaptative signed rank test for the location alternative for samples coming from asymmetric distributions. We give conditions under which the proposed test can have either greater or smaller acceptance breakdown, than the acceptance breakdown of a recently appeared adaptative rank test for location under asymmetry. Moreover, when symmetry of the sampled distribution can be justified the proposed test has greater acceptance breakdown than the winsorized signed rank test.