comparing the vocabulary of different graded readi

ReadinginaForeignLanguageApril2008,Volume20,No.1ISSN1539-0578pp.43–69Comparingthevocabularyofdifferentgraded-readingschemesUdornWan-a-romMahasarakhamUniversityThailandAbstractThisstudycomparedgraded-readerwordlistswiththeGeneralServiceList(GSL;West,1953)andinvestigatedthewordsinthoselistsandthewordsactuallyusedingraded-readerbooks.Thewordlistsfromthe2majorgraded-readerseries,theGSL,andthewordsactuallyusedinthegradedreaderswereexaminedusingtheRangeprogram.Thecomparisonsshowedthatthelistsaredifferentfromeachotherlargelybecauseofthedifferentsizesofthelistsandbecauseofthewordstheycontainanddonotcontain.Inaddition,thewordsactuallyusedinthebooksdonotstickcloselytothewordsinthelistsonwhichtheyarebased,especiallyatLevel1.Conclusionsandimplicationsaredrawnforpracticeinextensivereadingprograms.Keywords:gradedlevels,graded-readingschemes,gradedreaders,wordlists,extensivereadingAgraded-readerschemeusuallyhaswordandstructureliststhataredividedintolevelstoguidewritersandeditorsindesigninggraded-readingbooks.ThefindingsofNationandWang’s(1999)researchshowthatmostgraded-readerschemessetupconditionsthatwillenhancevocabularylearning.Thelimitedvocabularyateachlevelwillberepeatedinbooksofthesamelevel.Wordsfromearlierlevelswillberepeatedveryoftenatsubsequentlevels,andthiswillprovidelearnerswithmoreopportunitiestoencounterthewords.Theserepetitionsarebelievedtobecrucialforestablishingwordknowledge.AccordingtoNationandWang,about10repetitionsaredesirable,butthemorethebetter.NationandWang(1999)alsofoundthat84.7%ofthewordsintheGeneralServiceList(GSL;West,1953)appearedintheOxfordBookworms’(OBW)lists,1showingthattheclassiclistofthe2,000GSLwordsisofpracticalusetowritersofgradedreaders.Ageneral-servicevocabularyisessentialforalllearners,nomatterthemodesinwhichandpurposesforwhichtheyareusingEnglishasaforeignorsecondlanguage.ThisclaimissupportedbythefindingthattheGSLprovidesaround82%averagecoverageofvariouskindsofwrittentexts(Hirsh&Nation,1992;Hwang&Nation,1989;Sutarsyah,Nation,&Kennedy,1994),2withhighercoverageformoreinformaltext.However,learnersneedvocabularysizesthatwillcoveratleast98%ofthetextstheyread(Hu&Nation,2000).AccordingtoNation(2006),forunsimplifiedtexts,thiswouldrequireavocabularysizeofapproximately7,000–9,000wordfamilies(i.e.,headwordstogetherwiththeirothercommonforms).Thenotionofvocabularysizehasbeentakenasaguidelinefordevisingaschemeforgradedreaders.Ideally,graded-readingschemes:Comparingthevocabularyofdifferentgraded-readingschemes44wouldtakelearnersstepbystepwith98%coverageateachstepuntiltheycanreadunsimplifiedtextwiththesamecoverage.Unfortunately,asNationandWang(1999)showed,mostschemesofgradedreadersarenotwelldesignedintermsofvocabularysize.FrequencycountsofEnglishsubstantiallyagreeonthehigh-frequencywords(Nation,2004).Becausethelevelsofgradedreadersmakeuseofthesehigh-frequencywords,thevariouswordlistsofgradedreadersarelikelytobecomposedofsubstantiallythesamewords.Designedasreadabletextsforsecondlanguagelearners,gradedreadersuseacontrolledvocabularyandstructuralfeaturesthatarearrangedinstagesorlevelsofincreasingdifficulty.Thesestagesorlevelsformgraded-readingschemes.Theprimarypurposeofthewordlistsintheseschemesistoprovideguidelinesforwritersandeditorsofgradedreaders.Publishersusuallysetthedifferentlevelsforgradedreadersaccordingtothenumberofheadwords,andwriterscanuseawiderangeofwordsinthelists,dependingonthestoryortopic.Presumably,vocabularyisselectedchieflyonthebasisoffrequency,butthewordlistmaybemodifiedforaparticulartitlebasedontherequirementsofthestory.Differentpublisherscannotbeguaranteedtomakelistswiththesamewordsandwiththesamenumberofheadwordsatthesamelevel.Becauseofthis,nosystematiccomparisonofthelevelsofthevariousschemeshasbeenmadebeyondreviewseveryfewyears,whichhavedealtwithcontent,features,andthenumberofheadwordsappearinginthecataloguestocomparedifferentschemes,asinHill’s(1997,2001)reviewsofgradedreaders.However,thesereviewsdidnotexaminethewordlistsindetailintermsofthewordsinthewordlistsandtheactualwordsusedinthebooks.Althoughmanyofthegraded-readerseriesontheworldmarketprobablydependonWest’s(1953)GSLasabasisforthechoiceofwordsusedinthebooks,forcommercialpurposes,thepublishershaveproducedwordlistsoftheirown,whicharelikelytobeconfidentialandunique.Variouswordlistshaveresulted,andthewordsincludedandthenumberoflevelsvarywiththegradingscheme.Littleisknownaboutthesimilaritiesofthewordlists.Onewaytocheckthisistocomparethewordlistsoftheseriestodeterminetheamountofoverlapbetweenthelists.Thepurposeofthisstudywastoexaminethewordlistsofgradedreadersindetail.Thisshouldanswerthequestionofwhetherthelistsfromthevariousseriesaresimilarenoughtouseasabasisforsettingupreadingschemesforanextensivereadingprogramorreadingacrossseries,whichpertainstolanguagelearningingeneralandvocabularyinparticular.Thestudycomparedsetsofwordlistsoftwomajorseries:thoseoftheOBWbytheOxfordUniversityPressandtheCambridgeEnglishReaders(CER)bytheCambridgeUniversityPress.Italsolookedattheamountofoverlapbetweenthewordsinthe