WWW.PUBLICATIONETHICS.ORG
RETRACTION GUIDELINES
Whichpublicationsshouldberetracted?
Ifonlyasmallpartofanarticlereportsaweddata,andespeciallyifthisistheresultofgenuineerror,thenthe
problemisbestrectiedbyacorrectionorerratum.(Thetermerratumusuallyreferstoaproductionerror,causedby
thejournal.Thetermcorrigendum(orcorrection)usuallyreferstoanauthorerror.)Partialretractionsarenothelpful
becausetheymakeitdifcultforreaderstodeterminethestatusofthearticleandwhichpartsmaybereliedupon.
Similarly,ifonlyasmallsectionofanarticle(e.g.afewsentencesinthediscussion)isplagiarised,editorsshould
considerwhetherreaders(andtheplagiarisedauthor)wouldbebestservedbyacorrection(whichcouldnotethe
factthattextwasusedwithoutappropriateacknowledgement)ratherthanretractingtheentirearticlewhichmay
containsound,originaldatainotherparts.
Retractionshouldusuallybereservedforpublicationsthataresoseriouslyawed(forwhateverreason)thattheir
ndingsorconclusionsshouldnotbereliedupon.
Ifredundantpublicationhasoccurred(i.e.authorshavepublishedthesamedataorarticleinmorethanonejournal
withoutappropriatejustication,permissionorcrossreferencing)thejournalthatrstpublishedthearticlemay
issueanoticeofredundantpublicationbutshouldnotretractthearticleunlessthendingsareunreliable.Any
journalsthatsubsequentlypublisharedundantarticleshouldretractitandstatethereasonfortheretraction.
Ifanarticleissubmittedtomorethanonejournalsimultaneously,andisacceptedandpublishedinbothjournals
(eitherelectronicallyorinprint)atthesametime,precedencemaybedeterminedbythedateonwhichalicenceto
publishoracopyrighttransferagreementwassignedbytheauthors.
Incasesofpartialoverlap(i.e.whenauthorspresentsomenewndingsinanarticlethatalsocontainsasubstantial
amountofpreviouslypublishedinformation)editorsneedtoconsiderwhetherreadersarebestservediftheentire
articleisretractedorwhetheritwouldbebesttoissueanoticeofredundantpublicationclarifyingwhichaspects
hadbeenpublishedpreviouslyandprovidingappropriatecross-referencestotheearlierwork.Thiswilldependon
theamountofoverlap.Editorsshouldbearinmindthatthemainpurposeofretractionsistocorrecttheliterature
andensureitsintegrityratherthantopunishauthorswhomisbehave.
Onlypublisheditemscanberetracted.Guidelinesondealingwithredundantpublicationsidentiedinsubmitted
manuscriptscanbefoundintherelevantCOPEowchart[http://publicationethics.org/les/u2/01A_Redundant_
Submitted.pdf].Postinganalversiononawebsiteconstitutespublicationevenifanarticlehasnotappeared(or
willnotappear)inprint.Ifanarticleisretractedbeforeitappearsintheprintversionofajournal,theelectronic
versionshouldberetainedonthejournal’swebsitewithaclearnoticeofretractionanditshouldbeincludedon
bibliographicdatabases(e.g.withadigitalobjectidentier[doi]orotherpermanentcitationthatwilllocateit)
evenifitdoesnotappearintheprintedjournalandthereforedoesnotreceiveapageallocation.Thisisbecause
electronicversionsmayalreadyhavebeenaccessedandcitedbyresearcherswhoneedtobealertedtothefactthat
the article has been retracted.
ARCHIVED