Publicação

Interval Estimators for a Binomial Proportion: Comparison of Twenty Methods

Ver documento

Detalhes bibliográficos
Resumo:In applied statistics it is often necessary to obtain an interval estimate for an unknown proportion (p) based on binomial sampling. This topic is considered in almost every introductory course. However, the usual approximation is known to be poor when the true p is close to zero or to one. To identify alternative procedures with better properties twenty non-iterative methods for computing a (central) two-sided interval estimate for p were selected and compared in terms of coverage probability and expected length. From this study a clear classification of those methods has emerged. An important conclusion is that the interval based on asymptotic normality, but after the arcsine transformation and a continuity correction, and the Add 4 method of Agresti and Coull (1998) yield very reliable results, the choice between the two depending on the desired degree of conservativeness.
Autores principais:Pires , Ana M.
Outros Autores:Amado , Conceição
Assunto:confidence interval binomial distribution proportion test normal approximation arcsine transformation continuity correction bootstrap HPD credibility intervals
Ano:2008
País:portugal
Tipo de documento:artigo
Tipo de acesso:unknown
Instituição associada:Instituto Nacional de Estatística
Idioma:inglês
Origem:REVSTAT-Statistical Journal
_version_ 1869054297011912704
author Pires , Ana M.
author2 Amado , Conceição
author2_role author
author_facet Pires , Ana M.
Amado , Conceição
author_role author
country_str portugal
creators_json_txt [{\"Person.name\":\"Pires , Ana M.\"},{\"Person.name\":\"Amado , Conceição\"}]
datacite.creators.creator.creatorName.fl_str_mv Pires , Ana M.
Amado , Conceição
datacite.rights.fl_str_mv http://purl.org/coar/access_right/c_abf2
datacite.subjects.subject.fl_str_mv confidence interval
binomial distribution
proportion test
normal approximation
arcsine transformation
continuity correction
bootstrap
HPD credibility intervals
datacite.titles.title.fl_str_mv Interval Estimators for a Binomial Proportion: Comparison of Twenty Methods
dc.creator.none.fl_str_mv Pires , Ana M.
Amado , Conceição
dc.format.none.fl_str_mv application/pdf
dc.identifier.none.fl_str_mv https://doi.org/10.57805/revstat.v6i2.63
dc.language.none.fl_str_mv eng
dc.publisher.none.fl_str_mv Statistics Portugal
dc.rights.none.fl_str_mv http://purl.org/coar/access_right/c_abf2
dc.source.none.fl_str_mv REVSTAT-Statistical Journal; Vol. 6 No. 2 (2008): REVSTAT-Statistical Journal; 165–197
REVSTAT; Vol. 6 N.º 2 (2008): REVSTAT-Statistical Journal; 165–197
2183-0371
1645-6726
dc.subject.none.fl_str_mv confidence interval
binomial distribution
proportion test
normal approximation
arcsine transformation
continuity correction
bootstrap
HPD credibility intervals
dc.title.fl_str_mv Interval Estimators for a Binomial Proportion: Comparison of Twenty Methods
dc.type.none.fl_str_mv http://purl.org/coar/resource_type/c_6501
description In applied statistics it is often necessary to obtain an interval estimate for an unknown proportion (p) based on binomial sampling. This topic is considered in almost every introductory course. However, the usual approximation is known to be poor when the true p is close to zero or to one. To identify alternative procedures with better properties twenty non-iterative methods for computing a (central) two-sided interval estimate for p were selected and compared in terms of coverage probability and expected length. From this study a clear classification of those methods has emerged. An important conclusion is that the interval based on asymptotic normality, but after the arcsine transformation and a continuity correction, and the Add 4 method of Agresti and Coull (1998) yield very reliable results, the choice between the two depending on the desired degree of conservativeness.
dirty 0
eu_rights_str_mv unknown
format article
id revstat_9624165cc440fd2be2ca2fd6fa87f2ec
identifier.doi.fl_str_mv https://doi.org/10.57805/revstat.v6i2.63
inst_facet_str urn:organizationAcronym:revstat-statistical journal{{{_:::_}}}Instituto Nacional de Estatística
instacron_str REVSTAT-Statistical Journal
institution Instituto Nacional de Estatística
instname_str Instituto Nacional de Estatística
language eng
network_acronym_str revstat
network_name_str REVSTAT-Statistical Journal
oai_identifier_str oai:revstat:article/63
organization_str_mv urn:organizationAcronym:revstat-statistical journal
person_str_mv Pires , Ana M.
Amado , Conceição
publishDate 2008
publisher.none.fl_str_mv Statistics Portugal
repo_facet_str urn:repositoryAcronym:revstat{{{_:::_}}}REVSTAT-Statistical Journal
reponame_str REVSTAT-Statistical Journal
repository_id_str urn:repositoryAcronym:revstat
service_str_mv urn:repositoryAcronym:revstat
spelling en-USInterval Estimators for a Binomial Proportion: Comparison of Twenty MethodsPires , Ana M.Amado , Conceiçãoconfidence intervalbinomial distributionproportion testnormal approximationarcsine transformationcontinuity correctionbootstrapHPD credibility intervalsCopyright (c) 2008 REVSTAT-Statistical Journalhttp://purl.org/coar/access_right/c_abf2https://creativecommons.org/licenses/by/4.0https://doi.org/10.57805/revstat.v6i2.63DOIoai:revstat:article/63OAIhttps://revstat.ine.pt/index.php/REVSTAT/article/view/63URLhttps://doi.org/10.57805/revstat.v6i2.63DOIhttps://revstat.ine.pt/index.php/REVSTAT/article/view/63/67URLHasVersion2008-06-24T00:00:00Zen-USIn applied statistics it is often necessary to obtain an interval estimate for an unknown proportion (p) based on binomial sampling. This topic is considered in almost every introductory course. However, the usual approximation is known to be poor when the true p is close to zero or to one. To identify alternative procedures with better properties twenty non-iterative methods for computing a (central) two-sided interval estimate for p were selected and compared in terms of coverage probability and expected length. From this study a clear classification of those methods has emerged. An important conclusion is that the interval based on asymptotic normality, but after the arcsine transformation and a continuity correction, and the Add 4 method of Agresti and Coull (1998) yield very reliable results, the choice between the two depending on the desired degree of conservativeness.Statistics Portugalapplication/pdfen-USREVSTAT-Statistical Journal; Vol. 6 No. 2 (2008): REVSTAT-Statistical Journal; 165–197pt-PTREVSTAT; Vol. 6 N.º 2 (2008): REVSTAT-Statistical Journal; 165–1972183-03711645-6726engjournal articlehttp://purl.org/coar/resource_type/c_6501literatureVoRhttp://purl.org/coar/version/c_970fb48d4fbd8a85
spellingShingle Interval Estimators for a Binomial Proportion: Comparison of Twenty Methods
Pires , Ana M.
confidence interval
binomial distribution
proportion test
normal approximation
arcsine transformation
continuity correction
bootstrap
HPD credibility intervals
status SINGLETON
status_str VoR
subject.fl_str_mv confidence interval
binomial distribution
proportion test
normal approximation
arcsine transformation
continuity correction
bootstrap
HPD credibility intervals
title Interval Estimators for a Binomial Proportion: Comparison of Twenty Methods
title_full Interval Estimators for a Binomial Proportion: Comparison of Twenty Methods
title_fullStr Interval Estimators for a Binomial Proportion: Comparison of Twenty Methods
title_full_unstemmed Interval Estimators for a Binomial Proportion: Comparison of Twenty Methods
title_short Interval Estimators for a Binomial Proportion: Comparison of Twenty Methods
title_sort Interval Estimators for a Binomial Proportion: Comparison of Twenty Methods
topic confidence interval
binomial distribution
proportion test
normal approximation
arcsine transformation
continuity correction
bootstrap
HPD credibility intervals
topic_facet confidence interval
binomial distribution
proportion test
normal approximation
arcsine transformation
continuity correction
bootstrap
HPD credibility intervals
url https://doi.org/10.57805/revstat.v6i2.63
visible 1