Empirical-likelihood-based hypothesis tests for the means of two zero-inflated populations

Abstract

In this paper, we propose a novel nonparametric method based on the adjusted empirical likelihood (AEL) and the transformed empirical likelihood (TEL), which are two modified versions of the empirical likelihood (EL). The proposed method is considered to conduct a hypothesis test for the equality of the means of two populations containing many zeroes. The asymptotic results have been established. Simulations are conducted under various scenarios to investigate the performance of the proposed method as well as in comparison with the EL, AEL, TEL and the normal approximation (NA) in terms of type I error rates and powers. All methods are applied to four datasets to illustrate the hypothesis testing procedure.