A1347
Title: Novel constraints in empirical likelihood for ranked set sampling
Authors: Soohyun Ahn - Ajou University (Korea, South) [presenting]
Abstract: An innovative empirical likelihood (EL) method tailored for ranked set sampling (RSS) is introduced, which effectively harnesses the inherent ranking information within the sampling process. The approach imposes a novel constraint, ensuring that the sum of the within-stratum probabilities for each rank stratum equals $1/H$, where H represents the number of rank strata. This constraint streamlines the application of EL across both balanced and unbalanced RSS, eliminating the need for subjective weight selection in unbalanced cases. The efficacy of the method is demonstrated through its application to one-sample mean testing, supported by a comprehensive numerical study and analysis of two real-world datasets, body fat data and dental size data.
