Evaluating the power and the size of multivariate Levene’s test using Monte Carlo simulation and bootstrap - DOI: 10.4025/actascitechnol.v28i1.1293

Detalhes bibliográficos
Autor(a) principal: Cirillo, Marcelo Angelo
Data de Publicação: 2008
Outros Autores: Ferreira, Daniel Furtado, Sáfadi, Thelma
Tipo de documento: Artigo
Idioma: por
Título da fonte: Acta scientiarum. Technology (Online)
Texto Completo: http://www.periodicos.uem.br/ojs/index.php/ActaSciTechnol/article/view/1293
Resumo: This work aimed to evaluate the I-type error rate and the power of Levene’s multivariate test using different location parameters to compare k covariance matrices. Monte Carlo’s simulation and bootstrap were used for this purpose. Configurations were considered with different degrees of correlations, heterogeneity of covariance matrices, sample sizes and number of variants. Results showed that Levene’s multivariate test centered on the mean is more powerfull in small sample size and with low degree of heterogeneity among covariance matrices; the bootstrap approach controlled the I-type error when median was used as scale parameter.
id UEM-6_00b72c4ae90c5276903d95990cd9a9c9
oai_identifier_str oai:periodicos.uem.br/ojs:article/1293
network_acronym_str UEM-6
network_name_str Acta scientiarum. Technology (Online)
repository_id_str
spelling Evaluating the power and the size of multivariate Levene’s test using Monte Carlo simulation and bootstrap - DOI: 10.4025/actascitechnol.v28i1.1293Estudo do poder e tamanho do teste de Levene multivariado via simulação Monte Carlo e bootstrap - DOI: 10.4025/actascitechnol.v28i1.1293simulação bootstraptaxa de erro tipo Ipoder do teste de Levene1.02.02.00-5 EstatísticaThis work aimed to evaluate the I-type error rate and the power of Levene’s multivariate test using different location parameters to compare k covariance matrices. Monte Carlo’s simulation and bootstrap were used for this purpose. Configurations were considered with different degrees of correlations, heterogeneity of covariance matrices, sample sizes and number of variants. Results showed that Levene’s multivariate test centered on the mean is more powerfull in small sample size and with low degree of heterogeneity among covariance matrices; the bootstrap approach controlled the I-type error when median was used as scale parameter.Este trabalho tem por objetivo avaliar a taxa de erro tipo I e o poder do teste de Levene multivariado definido em função de diferentes parâmetros de posição para comparar k-matrizes de covariâncias. Para isso foram utilizadas simulação Monte Carlo e bootstrap. São contempladas situações com diferentes graus de correlação, heterogeneidade entre as matrizes de variâncias e covariâncias, tamanho amostral e número de variáveis. Concluiu-se que o teste de Levene centrado na média foi mais poderoso para amostras pequenas e com baixa heterogeneidade das matrizes de covariâncias; a abordagem bootstrap controlou a taxa de erro tipo I quando a mediana foi utilizada como parâmetro de posição.Universidade Estadual De Maringá2008-03-20info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionapplication/pdfhttp://www.periodicos.uem.br/ojs/index.php/ActaSciTechnol/article/view/129310.4025/actascitechnol.v28i1.1293Acta Scientiarum. Technology; Vol 28 No 1 (2006); 105-112Acta Scientiarum. Technology; v. 28 n. 1 (2006); 105-1121806-25631807-8664reponame:Acta scientiarum. Technology (Online)instname:Universidade Estadual de Maringá (UEM)instacron:UEMporhttp://www.periodicos.uem.br/ojs/index.php/ActaSciTechnol/article/view/1293/724Cirillo, Marcelo AngeloFerreira, Daniel FurtadoSáfadi, Thelmainfo:eu-repo/semantics/openAccess2024-05-17T13:02:35Zoai:periodicos.uem.br/ojs:article/1293Revistahttps://www.periodicos.uem.br/ojs/index.php/ActaSciTechnol/indexPUBhttps://www.periodicos.uem.br/ojs/index.php/ActaSciTechnol/oai||actatech@uem.br1807-86641806-2563opendoar:2024-05-17T13:02:35Acta scientiarum. Technology (Online) - Universidade Estadual de Maringá (UEM)false
dc.title.none.fl_str_mv Evaluating the power and the size of multivariate Levene’s test using Monte Carlo simulation and bootstrap - DOI: 10.4025/actascitechnol.v28i1.1293
Estudo do poder e tamanho do teste de Levene multivariado via simulação Monte Carlo e bootstrap - DOI: 10.4025/actascitechnol.v28i1.1293
title Evaluating the power and the size of multivariate Levene’s test using Monte Carlo simulation and bootstrap - DOI: 10.4025/actascitechnol.v28i1.1293
spellingShingle Evaluating the power and the size of multivariate Levene’s test using Monte Carlo simulation and bootstrap - DOI: 10.4025/actascitechnol.v28i1.1293
Cirillo, Marcelo Angelo
simulação bootstrap
taxa de erro tipo I
poder do teste de Levene
1.02.02.00-5 Estatística
title_short Evaluating the power and the size of multivariate Levene’s test using Monte Carlo simulation and bootstrap - DOI: 10.4025/actascitechnol.v28i1.1293
title_full Evaluating the power and the size of multivariate Levene’s test using Monte Carlo simulation and bootstrap - DOI: 10.4025/actascitechnol.v28i1.1293
title_fullStr Evaluating the power and the size of multivariate Levene’s test using Monte Carlo simulation and bootstrap - DOI: 10.4025/actascitechnol.v28i1.1293
title_full_unstemmed Evaluating the power and the size of multivariate Levene’s test using Monte Carlo simulation and bootstrap - DOI: 10.4025/actascitechnol.v28i1.1293
title_sort Evaluating the power and the size of multivariate Levene’s test using Monte Carlo simulation and bootstrap - DOI: 10.4025/actascitechnol.v28i1.1293
author Cirillo, Marcelo Angelo
author_facet Cirillo, Marcelo Angelo
Ferreira, Daniel Furtado
Sáfadi, Thelma
author_role author
author2 Ferreira, Daniel Furtado
Sáfadi, Thelma
author2_role author
author
dc.contributor.author.fl_str_mv Cirillo, Marcelo Angelo
Ferreira, Daniel Furtado
Sáfadi, Thelma
dc.subject.por.fl_str_mv simulação bootstrap
taxa de erro tipo I
poder do teste de Levene
1.02.02.00-5 Estatística
topic simulação bootstrap
taxa de erro tipo I
poder do teste de Levene
1.02.02.00-5 Estatística
description This work aimed to evaluate the I-type error rate and the power of Levene’s multivariate test using different location parameters to compare k covariance matrices. Monte Carlo’s simulation and bootstrap were used for this purpose. Configurations were considered with different degrees of correlations, heterogeneity of covariance matrices, sample sizes and number of variants. Results showed that Levene’s multivariate test centered on the mean is more powerfull in small sample size and with low degree of heterogeneity among covariance matrices; the bootstrap approach controlled the I-type error when median was used as scale parameter.
publishDate 2008
dc.date.none.fl_str_mv 2008-03-20
dc.type.driver.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
format article
status_str publishedVersion
dc.identifier.uri.fl_str_mv http://www.periodicos.uem.br/ojs/index.php/ActaSciTechnol/article/view/1293
10.4025/actascitechnol.v28i1.1293
url http://www.periodicos.uem.br/ojs/index.php/ActaSciTechnol/article/view/1293
identifier_str_mv 10.4025/actascitechnol.v28i1.1293
dc.language.iso.fl_str_mv por
language por
dc.relation.none.fl_str_mv http://www.periodicos.uem.br/ojs/index.php/ActaSciTechnol/article/view/1293/724
dc.rights.driver.fl_str_mv info:eu-repo/semantics/openAccess
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
dc.publisher.none.fl_str_mv Universidade Estadual De Maringá
publisher.none.fl_str_mv Universidade Estadual De Maringá
dc.source.none.fl_str_mv Acta Scientiarum. Technology; Vol 28 No 1 (2006); 105-112
Acta Scientiarum. Technology; v. 28 n. 1 (2006); 105-112
1806-2563
1807-8664
reponame:Acta scientiarum. Technology (Online)
instname:Universidade Estadual de Maringá (UEM)
instacron:UEM
instname_str Universidade Estadual de Maringá (UEM)
instacron_str UEM
institution UEM
reponame_str Acta scientiarum. Technology (Online)
collection Acta scientiarum. Technology (Online)
repository.name.fl_str_mv Acta scientiarum. Technology (Online) - Universidade Estadual de Maringá (UEM)
repository.mail.fl_str_mv ||actatech@uem.br
_version_ 1799315331571777536