A modified single-stage sampling procedure for heteroscedasticity analysis of means with unbalanced design

Wei Ming Wang, Chun Che Wen, Rajdeep Das, Miin Jye Wen

Research output: Contribution to journalArticlepeer-review

Abstract

The analysis of means (ANOM) is a method that can compare the mean of each treatment with the overall mean. As one-way ANOVA, the conventional ANOM is not robust under unequal variances. We applied two inference procedures, single-stage and modified single-stage sampling, to solve heteroscedastic analysis of means (HANOM) under unbalanced design. Using proposed weighted average across groups, the influence of the unknown mean and variance are both eliminated. To validate the stability of HANOM under several scenarios of unequal variances and sample sizes, simulation studies of empirical type I error rate are conducted to test the quality of procedures. A real application is provided for illustrating these two procedures clearly. In addition, we build an user-friendly interface to show the results of the HANOM by using Shiny package in R software.

Original languageEnglish
Pages (from-to)2277-2290
Number of pages14
JournalCommunications in Statistics: Simulation and Computation
Volume53
Issue number5
DOIs
Publication statusPublished - 2024

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Modelling and Simulation

Fingerprint

Dive into the research topics of 'A modified single-stage sampling procedure for heteroscedasticity analysis of means with unbalanced design'. Together they form a unique fingerprint.

Cite this