COHEN KULIK AN KULIAUTHORPETE A COHEN Assistan Director Offic o Instru - PDF document

COHEN KULIK AN KULIAUTHORPETE A COHEN Assistan Director Offic o Instructiona Service anEducationa Research an Adjunc Assistan Professo o Psychology 3 Webste Hall Dartmout College Hanover N 03755 Specializations: Instructiona evaluation researc o colleg teaching researc synthesisJAME A KULIK Researc Scientis an Associat Director Cente foResearc o Learnin an Teaching 10 E Madison Universit oMichigan An Arbor M 48109 Specializations: Researc synthesiseducationa evaluationCHEN-LI KULIK Assistan Researc Scientist Cente fo Researc oLearnin an Teachin 10 E Madison Universit o Michigan AnArbor M 48109 Specializations: Researc synthesis individualize instruction psychologica measurement24 at FLORIDA ATLANTIC UNIV on January 14, 2011http://aerj.aera.netDownloaded from EDUCATIONA OUTCOME O TUTORINresearc design the th mos selecte researc result (i.e. thos frojourna articles woul provid th bes basi fo estimate o effec sizeThi meta-analysi confirm som thing tha hav lon bee suspecteabou tutoring I shows a man commentator hav suggested tha tutorinbenefit bot tutor an tutee o bot th cognitiv an affectiv levels Iaddition i specifie th averag strengt o tutorin effects an i identifieth setting an condition wher effect ar strongest Finally th metaanalysi raise som ne question abou tutoring I challenge otheinvestigator an reviewer t identif th ke variable underlyin variatioi tutorin outcomesREFERENCECOHEN J Statistical power analysis for the behavioral sciences (Rev ed.) Ne YorkAcademi Press 1977DEVIN-SHEEHAN L. FELDMAN R S. & ALLEN V L Researc o childre tutorinchildren A critica review Review of Educational Research, 1976 46, 355-385ELLSON D G Tutoring I N Gag (Ed.) The psychology of teaching methods. Chicago Universit o Chicag Press 1976FITZ-GIBBON C T An analysis of the literature of cross-age tutoring. WashingtonD.C. Nationa Institut o Education 1977 (ERI Documen ReproductioServic No E 14 807)GLASS G V Primary secondary an meta-analysi o research Educational Re­searcher, 1976 5, 3-8GLASS G V & SMITH M L Meta-analysi o researc o clas siz an achievementEducational Evaluation and Policy Analysis, 1979 7 2-16HARTLEY S S Meta-analysi o th effect o individuall pace instructioi mathematic (Doctora dissertation Universit o Colorado 1977) Disser­tation Abstracts International, 1977 38(1-A), 4003 (Universit Microfilm No77-29,926)KULIK J A. KULIK C.-L C & COHEN P A A meta-analysi o outcom studieo Keller' Personalize Syste o Instruction American Psychologist, 1979 34, 307-318 (aKULIK J A. KULIK C-L C & COHEN P A Researc o audiotutoria instructionA meta-analysi o comparativ studies Research in Higher Education, 1979 11, 321-341 (bMCGAW B. & GLASS G V Choic o th metri fo effec siz i meta-analysisAmerican Educational Research Journal, 1980 17, 325-337MOHAN M Peer tutoring as a technique for teaching the unmotivated. Fredonia N.Y.Stat Universit o Ne York Teache Educatio Researc Center 1972 (ERIDocumen Reproductio Servic No E 06 154)ROSENSHINE B. & FURST N The effects of tutoring upon pupil achievement: A research review. Washington D.C. Offic o Education 1969 (ERI DocumenReproductio Servic No E 06 462)SMITH M L Publicatio bia an meta-analysis Evaluation in Education: An International Review Series, 1980 4(1), 22-2424 at FLORIDA ATLANTIC UNIV on January 14, 2011http://aerj.aera.netDownloaded from COHEN KULIK AN KULIthe wer teaching bu the als gaine a bette understandin o thesareasTutorin program apparentl hav muc smalle effect o th selfconcept o children Neithe tutor no tutee change i self-estee a a resul o tutorin programs Th literatur contain anecdota report odramati change i self-concep brough abou b tutorin programs buquantitativ studie d no suppor thes reports Dramati change i selfestee appea t b atypicalGenerally ou result wer consisten wit finding o othe reviewers A numbe o reviewers fo example hav reporte tha structure tutorinprogram produc especiall stron effect (Ellson 1976 Rosenshin & Furst 1969) Althoug w foun tha bot structure an unstructureprogram produce measurabl effects th effect fro th structure program wer indee stronger Lik othe reviewer (e.g. Fitz-Gibbon 1977)w als foun tha th degre o effectivenes o tutorin program dependeo whethe standardize o locall develope test wer use i th evaluations Effect wer stronge whe measure wit locall develope testsResult fro ou analysi agree especiall closel wit result reporteb Hartle (1977) Sh reported fo example a E o . fo tutorin studiei mathematics Ou result wer identica fo studie o mathematic teachin (i.e. ou E als equale .6) bu w foun somewha smalle tutorineffect i th are o readin instruction Hartle als reporte tha somadditiona factor wer relate t th siz o progra effects includin thtyp o repor (dissertatio versu publi schoo report an th typ oexaminatio (loca versu standardized) Ou meta-analysi als showe thathes tw factor wer relate t th siz o effectsReviewer an meta-analyst therefor agree t a larg exten o thfact abou tutorin programs Bu achievin consensu o th interpretatioo thes facts ma tur ou t b a mor difficul matter Relationship thao th surfac loo straightforwar ofte tur ou t b comple o closeexamination Thi ma b th cas wit som o th relationship betweestud feature an tutorin outcome note b th reviewersA goo exampl i th relationshi betwee th sourc o a stud an thsiz o th reporte effects W foun i ou meta-analysi tha journaarticle reporte stronge finding tha dissertations an thi resul ha beereporte frequentl i othe meta-analyse (Smith 1980) Selectiv publicatio o researc result certainl ca accoun fo th differenc i findingfro dissertation an journa articles Bu whic results—thos fro journaarticle o thos fro dissertations—shoul w accep a th mor accurateI th selectio proces tha eventuall lead t publicatio o researc ibase o th strengt o findings the th leas selecte result (i.e. thosfro dissertations woul provid th bes basi fo estimatin th siz otutorin effects If th selectio proces i base instea o th qualit o24 at FLORIDA ATLANTIC UNIV on January 14, 2011http://aerj.aera.netDownloaded from EDUCATIONA OUTCOME O TUTORINTABL IMean and Standard Error of Effect Size for Tutor Achievement when Studies are Classified in Various Ways Categor o StudNumbe oStudieEffec SizCategor o StudNumbe oStudieMeaStandar ErroImplementatioSubstitut2.4.1SupplemenTuto TraininN11.2.3.1.1Ye2.3.1Cross-ag TutorinN1.2.0Ye2.3.1Structure TutorinN1.3.0Ye2.3.1Duratio o Treatmen0- Week3 .5.25-1 Week2.3.119-3 Week9 .1.0Clas Leve o Tutee1-2.3.14-7 .1.17-2 .6.2Clas Leve o Tutor1-1.2.04-1.4.27-1.2.1Subjec MatterMat1.6.2ReadinAVerag Abilit o TutoVer LoLo211.2.4.2.0.1.0Middl8 .2.1Sourc o StudDissertatio2.2.0Publishe7 .2.1Stud Yea1966-1971.3.11971-1972.3.11976-1985 .2.1* Significan differenc amon effec size fo categorie o thi variable,/ 0.0524 at FLORIDA ATLANTIC UNIV on January 14, 2011http://aerj.aera.netDownloaded from COHEN KULIK AN KULIprograms Th averag E i th nin studie wa .09 th standar errowa .042 Clearly thi effec wa ver smal an wa no larg enoug t bconsidere statisticall reliableEffects on Student Tutors O th 6 studie w located 3 examine achievemen effect o tutors1 investigate change i self-concep o tutors an 5 examine change ituto attitude towar th subjec matte bein taughtAchievement. I 3 o th 3 studie investigatin effect i thi areastudent wh serve a tutor performe bette tha di contro student oexamination i th subjec bein taught I th remainin 5 studies exam score wer bette fo student no servin a tutors O th 3comparisons 1 reporte statisticall significan results an i eac cas th favore student servin a tutors Th averag E i th 3 wa .33 th standar erro wa .09 Onl 1 o 1 stud featureinclude i thi analysi wa significantl relate t effec size—th kin o matte taugh i th tutorin progra (Tabl II)Attitude toward subject matter. I fou o th fiv studie investigatineffect i thi area attitude wer positiv amon thos servin atutors i th othe study th student wh di no serv a tutor hel thmor positiv attitudes Onl on stud showe a statisticall significandifferenc i subjec matte attitude o tutor an conventiona studentsan thi stud favore tutors Th averag E fo attitud towar th subjecwa .42 th standar erro wa .46 Cohe (1977 referre t effect o thimagnitud a moderat i sizeSelf concept. A tota o 1 studie reporte o effect o tutorin programo self-concept o student wh serve a tutors I 1 studies self-concepwa highe fo tutor tha fo thos wh di no serv a tutors i thremainin 4 studies self-concep wa highe fo thos wh di no serv atutors Fou studie showe statisticall significan differences an i eaccas th differenc favore student wh serve a tutors Th averag effec tuto self-concept however wa small Th mea E wa .18 th standarerro wa .12DISCUSSIOTh messag fro th educationa literatur o tutorin program seemclea enough Thes program hav definit an positiv effect o thacademi performanc an attitude o thos wh receiv tutoring Tutorestudent outperforme thei peer o examinations an the expresse morpositiv attitude towar th subject i whic the wer tutored Tutorinprogram als ha positiv effect o childre wh serve a tutors Thestutor no onl develope mor positiv attitude towar th subject tha24 at FLORIDA ATLANTIC UNIV on January 14, 2011http://aerj.aera.netDownloaded from Siz Yea differenc amon effec size variable,/ at FLORIDA ATLANTIC UNIV on January 14, 2011http://aerj.aera.netDownloaded from COHEN KULIK AN KULITABL I Mean and Standard Error of Effect Size for Tutee Achievement when Studies are Classified in Various Ways Effec SizCategor o StudNumbe oStudieMeaStandarErroImplementatioSubstitut2.4.1Supplemen2.3.1Tuto TraininN1.3.0Ye3.4.0Cross-ag TutorinN2.2.0Ye2.4.1Structure TutoringN2.2.0Ye2.5.1Rando Assignmen t GroupN2.3.0Ye3.4.1Contro fo Instructio EffecDifferen Instructor1.3.1Sam Instructo4.4.0Contro fo Autho Bias**Instructo Develope Tes1.8.2Commercia Standardize Tes4.2.0Duratio o Treatment**0- Week6 .9.25-1 Week3.4.019-3 Week1.1.0Clas Leve o Tutee1-3.4.04-1.2.17-6 .3.1Clas Leve o Tutor1-1.3.14-1.6.27-1.2.010-11.3.1Subjec Matter*Mat1.6.1Readin3.2.0Othe4 .3.1Averag Abilit o TuteLo4.4.0Middl 1.3.1Leve o Achievemen Measure**Low-orde1.7.1Mixe3.2.024 at FLORIDA ATLANTIC UNIV on January 14, 2011http://aerj.aera.netDownloaded from EDUCATIONA OUTCOME O TUTORINnatio performanc wa bette i th conventiona class an i 1 studyther wa n differenc betwee tutee an conventiona students A tota o2 o th comparison reporte a statisticall significan differenc i resultfro th tw teachin approaches Result o 1 o thes studie favoretutees an result o 1 stud favore conventiona instruction Clearly a distinc majorit o studie favore tuteesTh averag E i th 5 studie wa .40 th standar erro o E wa.069 A averag E o .4 mean tha i a typica clas tutorin raise thperformanc o tutore student b approximatel two-fifth o a standarddeviation-unit Pu i anothe way th averag chil i th tutore grouscore a th 66t percentil o th student i th untutore o controgroup Cohe (1977 ha referre t effect o thi magnitud a modes isizeAlthoug th effec o tutorin wa modes i th typica study th siz oeffec varie fro stud t stud (Figur 1) Th larges E (2.3 cam fro stud fro Moha (1972) Fou othe studie als reporte larg effectFIGUR 1 Distributio showin th effect o tutorin o tute achievemen i5 studies3-2-1 1 2 EFFECT O TUTE ACHIEVEMEN (ES24 at FLORIDA ATLANTIC UNIV on January 14, 2011http://aerj.aera.netDownloaded from COHEN KULIK AN KULICharacteristics of Studies Th 6 studie use i thi analysi wer o man differen types Tdescrib th mai feature o th studies w define 1 variables Fou oth variable describe th type o tutorin program use i th studieswhethe th tutorin wa structure o nonstructured whethe th tutorin cross-ag o not whethe tutorin wa a supplemen t o a substitutfo classroo instruction an whethe o no tutor receive training Th thre variable describe aspect o th experimenta desig o thstudies rando versu nonrando assignmen o student t comparisogroups contro fo teache effect b usin th sam teacher fo bot an contro groups an contro o autho bia o test throug o standardize examinations Si othe variable describe feature o cours setting includin duratio o th program clas leve o tutorsclas leve o tutees subjec matter averag abilit leve o tutees an leve skill teste o examinations Finally tw variable describe publicatio o th studies th manne o publicatio o th stud an th yeao publicationStudy Outcomes Th 6 studie describe effect o tutorin program o bot tutor antutees Thes effect wer i thre majo areas studen achievemen ameasure o examinations favorabilit o studen attitude towar thsubjec matter an favorabilit o studen self-concept T quantif theffect o tutorin program i eac o thes areas w use th Effec Siz(ES) define a th differenc betwee th mean o tw group divide bth standar deviatio o th contro grou (Glass 1976) Fo studie thareporte mean an standar deviation fo bot experimenta an controgroups w calculate E fro th measurement provided Fo les fullreporte studies w calculate E fro statistic suc a / an F, usinprocedure describe b McGa an Glas (1980)RESULTI thi sectio w first describ th effect o tutorin program o thchildre wh receive tutoring W the tur t th effect o tutorinprogram o childre wh serve a tutor i thes programsEffects on Tutees A tota o 5 o th 6 studie reporte result o academi achievemeno tutore students 9 studie reporte o self-concept an 8 reporte oattitud towar subjec matterAchievement. I 4 o th 5 achievemen studies th examinatio performanc o student wh wer tutore wa bette tha th examinatioperformanc o student i a conventiona class i 6 othe studies exami24 at FLORIDA ATLANTIC UNIV on January 14, 2011http://aerj.aera.netDownloaded from 1 T A at FLORIDA ATLANTIC UNIV on January 14, 2011http://aerj.aera.netDownloaded from COHEN KULIK AN KULIo subjectiv impression an informa observation an thu ar o limitescientifi value othe report describ soun experimenta studie o tutoringI suc studies investigator usuall compar th performanc o equivalengroup o student assigne t classroom wit an withou tutorin programs Comparison ofte focu o learnin gain i th tw type oclassrooms an sometime als cove affectiv growt o tutore anuntutore studentsSevera majo review o suc studie hav appeare i th educationaliteratur i recen year (Devin-Sheehan Feldman & Allen 1976 Ellson1976 Fitz-Gibbon 1977 Rosenshin & Furst 1969) Eac o th reviewconclude tha tutorin program ca contribut t th academi growt oth childre wh receiv th tutorin an probabl t th growt o thchildre wh provid th tutorin a well However tw o th review(Ellson 1976 Rosenshin & Furst 1969 reporte tha thes contributionha bee clearl demonstrate onl fo well-structure an cognitiveloriente programs Sinc eac o th fou review use relativel informanarrativ an bo scor technique fo summarizin findings non provide statement abou th siz o th learnin gain t b expecte fro o abou th condition unde whic positiv effect ar likel toccur Fo mor precis conclusions mor forma revie method arneededI 197 Hartle introduce th us o mor powerfu revie method intthi area Th methodolog tha sh employed calle "meta-analysis, wafirs describe b Glas (1976 i hi presidentia addres t th AmericaEducationa Researc Association Meta-analysi i simpl th statisticaanalysi o a larg collectio o result fro individua studie fo th purposo integratin th findings Applyin thi metho t finding o mathematic i elementar an secondar schools Hartle showe no onl tha effect o tutorin wer positive bu tha the wer stronge tha thos suc othe individualizin teachin method a computer-base instruction programme instruction an instructio wit individua learnin packages Hartle als showe tha th effect o tutorin wer significantlstronge i som situation tha i othersAlthoug Hartley' stud advance knowledg o tutorin considerablyhe wor wa stil somewha limite i scope First Hartley' analysi onl mathematic teaching Hartle di no determin whethe program ha th sam remarkabl effect i al subject o whethesuc effect wer restricte t mathematic teaching Second Hartle studieonl achievemen effects Sh di no determin whethe tutorin hapositiv o negativ effect o othe instructiona outcomes suc a attitudtowar school attitud towar schoo subjects o o self-conceptHartley' analysi als suffere fro som possibl methodologica weaknesses First he stud aggregate effect o thos bein tutore an othos providin tutoring Othe reviewer hav preferre t loo a thes a23 at FLORIDA ATLANTIC UNIV on January 14, 2011http://aerj.aera.netDownloaded from American Educational Research Journal Summer 1982, Vol. 19, No. 2, Pp. 237-248 Educationa Outcome o TutoringA Meta-analysi o FindingPETE A COHENDartmouth College JAME A KULI an CHEN-LI C KULIUniversity of Michigan A meta-analysis of findings from 65 independent evaluations of school tutoring programs showed that these programs have positive effects on the academic performance and attitudes of those who receive tutoring. Tutored students outperformed control students on examinations, and they also developed positive attitudes toward the subject matter covered in the tutorial programs. The meta-analysis also showed that tutoring programs have positive effects on children who serve as tutors. Like the children they helped, the tutors gained a better understanding of and developed more positive attitudes toward the subject matter cov­ered in the tutorial program. Participation in tutoring programs had or no effect, however, on the self-esteem of tutors and tutees. Th tutorin program offere i man elementar an secondar schooltoda diffe i a importan wa fro yesterday' tutoria programs I mosmoder programs childre ar tutore b peer o paraprofessional rathetha b regula schoo teacher o professiona tutors Th us o pee an tutor ha dramaticall affecte th availabilit o tutorinprograms N longe a luxur availabl onl t a aristocrati elite tutorin toda ar ope t boy an girl i ordinar classroom throughouth countryHundred o teacher an researcher alread hav writte report o theffect o suc program o children Althoug som o th report ar baseTh materia i thi repor i base upo wor supporte b th Nationa Scienc Foundatiounde Gran No SE 79-20742 An opinions findings an conclusion o recommendationexpresse i thi repor ar thos o th author an d no necessaril reflec th view o thNationa Scienc Foundation23 at FLORIDA ATLANTIC UNIV on January 14, 2011http://aerj.aera.netDownloaded from http://aerj.aera.netJournalAmerican Educational Research http://aer.sagepub.com/content/19/2/237The online version of this article can be found at: DOI: 10.3102/00028312019002237 1982 19: 237Am Educ Res JPeter A Cohen, James A. Kulik and Chen-Lin C. KulikEducational Outcomes of Tutoring: A Meta-analysis of Findings    Published on behalf of   American Educational Research Associationand http://www.sagepublications.com can be found at:American Educational Research JournalAdditional services and information for      http://aerj.aera.net/alertsEmail Alerts:   http://aerj.aera.net/subscriptionsSubscriptions:   http://www.aera.net/reprintsReprints:   http://www.aera.net/permissionsPermissions:   http://aer.sagepub.com/content/19/2/237.refs.htmlCitations:   at FLORIDA ATLANTIC UNIV on January 14, 2011http://aerj.aera.netDownloaded from