Robust Brown-Forsythe and Robust Modified Brown-Forsythe ANOVA Tests Under Heteroscedasticity for Contaminated Weibull Distribution

In this study, robust Brown-Forsythe and robust Modified Brown-Forsythe ANOVA tests are proposed to take into consideration heteroscedastic and non-normality data sets with outliers. The non-normal data is assumed to be a two parameters Weibull distribution. Robust proposed tests are obtained by usi...

Full description

Autores:
Karagöz, Derya
Saraçbasi, Tülay
Tipo de recurso:
Article of journal
Fecha de publicación:
2016
Institución:
Universidad Nacional de Colombia
Repositorio:
Universidad Nacional de Colombia
Idioma:
spa
OAI Identifier:
oai:repositorio.unal.edu.co:unal/66520
Acceso en línea:
https://repositorio.unal.edu.co/handle/unal/66520
http://bdigital.unal.edu.co/67548/
Palabra clave:
51 Matemáticas / Mathematics
31 Colecciones de estadística general / Statistics
Brown-Forsythe
Modified Brown-Forsythe
ANOVA
Weibull Distribution
ANOVA
Brown-Forsythe
Brown-Forsythe modificado
Distribución Weibull.
Rights
openAccess
License
Atribución-NoComercial 4.0 Internacional
Description
Summary:In this study, robust Brown-Forsythe and robust Modified Brown-Forsythe ANOVA tests are proposed to take into consideration heteroscedastic and non-normality data sets with outliers. The non-normal data is assumed to be a two parameters Weibull distribution. Robust proposed tests are obtained by using robust mean and variance estimators based on median/ MAD and median/Qn methods instead of maximum likelihood. The behaviors of the robust proposed and classical ANOVA tests are examined by simulation study. The results shows that the proposed robust tests have good performance especially in the presence of heteroscedasticity and contamination.