A Note for Likelihood Ratio Methods for Testing the Homogeneity of a Three-Sample Problem with a Mixture Structure

  • Pengcheng Ren
  • , Guanfu Liu*
  • , Xiaolong Pu
  • , Xingyu Yan
  • *Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Recently, our paper entitled “Generalized fiducial methods for testing the homogeneity of a three-sample problem with a mixture structure” is published in Journal of Applied Statistics 50 (2023), pp. 1094–1114. In simulation studies of this paper, the likelihood ratio method is regarded as a comparison method with the generalized fiducial methods. However, the construction of the likelihood ratio method and its asymptotic theories were not provided. It is worth noting that under the null model, the proportion parameter disappears, and it is unidentifiable. Hence, the classic theory of the likelihood ratio method is not applicable to the testing problem we consider. In consideration of the advantages owned by the likelihood ratio method, it is necessary to separately to establish the likelihood ratio method and study its asymptotic theory. Therefore, we write this note to highlight this method.

Original languageEnglish
Pages (from-to)114-133
Number of pages20
JournalSankhya A
Volume87
Issue number1
DOIs
StatePublished - Feb 2025

Keywords

  • Homogeneity test, likelihood ratio test, mixture, three-sample problem

Fingerprint

Dive into the research topics of 'A Note for Likelihood Ratio Methods for Testing the Homogeneity of a Three-Sample Problem with a Mixture Structure'. Together they form a unique fingerprint.

Cite this