Chapter 3 Transfer of learning paradoxes for learners

EducationalResearch31(1999)577}589

Chapter3

Transferoflearning:paradoxesforlearners

P.R.J.Simons

DepartmentofEducationalSciences,UniversityofNijmegen,P.O.Box9104,6500HDNijmegen,Netherlands

Abstract

Therearethreekindsoftransfer:frompriorknowledgetolearning,fromlearningtonewlearning,andfromlearningtoapplication.Thecentralthesisofthischapteristhatallthreeshouldstartfromthedilemmasorparadoxeslearnersmayhavewhentryingtoreachtransfer.Sixofthesearedescribed.Theyare\"ndingrelevantpriorknowledge;theparadoxoftacitknowledge;usingrelevantpriorknowledgewhilelearning;recognizingrelevantsituationsandconditions;theparadoxofneartransferandfartransfer;andtheparadoxical`whataoftransfer,includinglearningtolearn.Inordertooptimizetransferoneneedstohelplearnerssolvetheproblemstheyencounterandto\"ndwaysoutofthevariousdilemmasandparadoxes.Forthethreekindsoftransferexemplarystudiesaredescribedthatillustratethekindsofsolutionsdevelopedandtheire!ectiveness.󰀁1999ElsevierScienceLtd.Allrightsreserved.

1.Introduction

Transferoflearningoccurswheneverpreviouslylearnedknowledgeandskillsa!ectthewayinwhichnewknowledgeandskillsarelearnedandperformed(Cormier&Hagman,1987).Therearethreewaysinwhichtransfercanoccur:frompriorknowledgeandskillstonewlearning,fromnewknowledgeandskillstonewlearningsituations(learningnowpreparingforlaterlearning),andfromnewknowledgeandskillstoapplicationsinworkanddailylife(learningforpractice)(Simons,1990).Anexampleofthe\"rstcategoryoftransferisa(Dutch)studentwholearnstoreadanEnglishtextandusesknowledgeandskillsfromhisorhernativelanguage(Dutch).Anexampleofthesecondcategoryoftransferisastudentwhoislearningtouseatext-processingprogramlikeWordPerfectbutknowsthatheorshehastolearnMicrosoftWordlater.HowcanthisstudentlearntouseWordPerfectinsuchawaythatheorshewillbeabletolearnWordaseasilyaspossiblelater?Anexampleofthe\"nalcategoryisastudentwholearnstospeaktheFrenchlanguageintheclassroomandhastodoitinreal-lifecontextwhenreceivingaFrenchguest.

0883-0355/99/$-seefrontmatter󰀁1999ElsevierScienceLtd.Allrightsreserved.PII:S0883-0355(99)00025-7

578P.R.J.Simons/Int.J.Educ.Res.31(1999)577}589

Inlinewiththethreetypesoftransferarethreeresearchquestions.The\"rstis`howcanwefacilitateorincreasetheuseofpriorknowledgeininstructionalsituations?aThesecondis`howcanwedesigninstructionalsituationsinsuchawaythatchancesofbetterlaterlearningaremaximized?aThethirdis`howcanwebuildinstructionalororganizationalenvironmentsthathelpfacilitatetransferfromonesituation(learn-ing)toanothersituation(application)?aIntheliteratureontransferthethirdresearchquestionisemphasizedmost.Intermsofthisquestion,someresearchersfocusonthemorespeci\"cquestionofhowinstructionalenvironmentscanbedesignedinsuchawaythatlateruseoftheknowledgeandskillsinworkingordailylifeconditionsareoptimized.Othersstudythequestionhowworkingconditionscanbedesignedinsuchawaythatpeopleusewhattheylearnedinatrainingoratschoolintheirwork.Theothertwomajorquestions(\"rstandsecond,above)areeitherignoredorstudiedunderotherheadings(suchaspriorknowledge,tacitknowledge,andpreparationforlifelonglearning).

Theoretically,however,thebasicmechanismsandprinciplesoftransferareeitheridenticalorverysimilarforthethreekindsoftransfer(Cormier&Hagman,1987;Simons,1990).Inallthreecases,thefundamentalproblemishowknowledgeandskillslearnedinoneormoresituationsareusedorappliedinoneormoreothersituations.The\"rstsituationcanbeaspeci\"clearningsituation(questions2and3)oranentirerangeofdi!erentlearningsituations(question1).Thesecondsituationcaneitherbealearningsituationorarangeoflearningsituations(questions1and2)oranentirerangeof`applicationanduseasituations(question3).Table1illustratesthesepossibilities.

Furthermore,thesethreeprimaryresearchquestionsaremostlytreatedfromtheperspectiveofthedesignofinstructionalororganizationalenvironments.Fromtheconstructivistviewpoint,however,thelearners'perspectivesmustbeconsidered\"rst,beforeproperenvironmentscanbedesigned.Inthischapter,ananalysisispresentedoflearnerperspectivesontransferproblems.Indoingthis,thetransferdilemmasorparadoxesthatlearnersconfrontwhenlearninghavebeensought.Inwhatfollows,thethreetypesoftransferareillustratedwithaseriesofstudies.Someoftheparadoxesarespeci\"ctooneofthetransferquestions,othersaregeneric.

Table1

Threeresearchquestionsinrelationto\"rstandsecondsituations

SecondsituationsLearningsituation(s)

RangeofapplicationorusesituationsQuestion3:Fromlearningtoapplication(s)*

Firstsituations

Onespeci\"clearningsituationRangeoflearningsituations

Question2:Fromlearningtonewlearning

Question1:Fromknowledgeandskillstonewlearning

P.R.J.Simons/Int.J.Educ.Res.31(1999)577}589579

2.Paradoxes

Whatareparadoxes?PerhapsthemostfamousbooksaboutparadoxesisCharlesHandy's(1994)Theageofparadox.Hewrote:`Everycoin,Inowrealize,hasatleasttwosides,buttherearepathwaysthroughtheparadoxesifwecanunderstandwhatishappeningandarepreparedtoactdi!erentlya(p.17).QuinnandCameron(1988)de\"nedparadoxesas`contradictorymutuallyexclusiveelementsthatarepresentandoperateequallyatthesametimea(p.5).Thisisaratherstrictde\"nitionofparadoxes.AlessrigidoneisfromFletcherandOlwyler(1997):aparadoxisa`seeminglycontradictorystatementthatmaynonethelessbetruea(p.7).Itisimportanttorealizethatparadoxesareonlyseeminglycontradictory.Thereisacontradictionfromacertainperspectiveonly,notfromotherperspectives.Therefore,paradoxicalthink-ing\"tswellintoconstructivism:itfocusesthinkingontheperspectivesofactorsintheirenvironmentsandhowtheserestricttheirthinking.Inchangingourperspectiveormentalmodel,paradoxicalcontradictionsmaybeovercome.Wemay\"ndwaystoreconciletheseeminglyirreconcilabletensions.

Findinganddescribingparadoxesmayaccomplish\"veaims(Simons,1998):ELookingfordimensionsanddegreesinsteadofdichotomies.Insteadofadichotomousdistinctionbetweenshort-andlong-termlearning,forexample,thereislearningforthenextday,thenextweek,thenextyear,fortheexam,forlife,etc.EFocusingonhowthetwosidesofacoincanbecombinedinsteadofontheircontrast.Forinstance,\"ndingwaystocombinestructureandfreedom.

EFindingwaystousetheonesideofthecointostrengthentheother.Canalearner,forinstance,reachbettershort-termresultsbyfocusingonlong-termresults?ESca!olding,i.e.,replacingtheonesidegraduallybytheother.Canwe,forexample,helplearnerstolearnmoreandmoreindependentlythroughagradualincreaseofself-control?

EFindingdeterminingconditions.Forinstance,underwhatconditionstousesup-portandwhentowithholdit?

2.1.Paradoxofxndingpriorknowledge

Manypriorstudieshavedemonstratedtheimportanceofinformalandformalpriorknowledge(seeDochy,1992).Thispriorknowledgecanbedescribedasalltheknowledgelearnershaveavailablewhenenteringalearningenvironment,thatispotentiallyrelevantforlearningnewknowledge.Thispriorknowledgeisoftentacitorinert.Moreover,implicitandexplicitknowledgemaybewrong(ascomparedtoscienti\"corculturalstandards)oronlypartiallycorrect.Aswasshowninmanystudies,wrongorpartiallycorrectpriorknowledgetendstoresistchange.Evenafteryearsofformaleducation,somemisconceptions(re)appearedintheanswersstudentsgavewhensolvingproblems(Eylon&Linn,1988).

Ontheotherhand,studentsalsohavemuchpriorknowledgethatiscorrectandrelevant.Whenlearning,studentsusemuchofthispriorknowledgewithout

580P.R.J.Simons/Int.J.Educ.Res.31(1999)577}589

awareness.Itentersautomaticallyinthelearningprocess.Forinstance,inlearningfromatextoneautomaticallyusesallkindsofdomain-speci\"cknowledgeandreadingskills.Otherkindsofknowledgedonotenterautomaticallyinthelearningprocess.Thelearnermustdothispurposefullyorshouldbehelpedbyaninstructionalsystemtousethisknowledge.Theamountofpriorknowledgeastudenthasisapowerfulpredictorofnewlearning(Dochy,1992).

Fewstudentsusemuchoftheirpriorknowledgeintentionally,spontaneously,andactively(seeforinstanceDeJong&Simons,1990;Dochy,1992).Priorknowledgemayposeseveralcontradictoryproblemsforstudentsthatmayexplainwhytheytendnottousemuchoftheirpriorknowledgeactively.Thereissomuchpriorknowledgethatlearnersmaynotknowwhichpriorknowledgeisrelevantinacertainlearningenvironmentandwhichisnot.Partoftheproblemforalearner,then,isselectingtherelevantpriorknowledgefromanalmostendlessamountofexplicitknowledgethatisavailable.Howcanalearnermakethedecisionastowhatpriorknowledgetoincludeandwhichtoignore?Furthermore,heorshemaynotknowwhichpartofthepriorknowledgeiscorrectandwhichisincorrectorpartiallycorrectaccordingtoscienti\"candorculturalstandards.

2.2.Paradoxoftacitknowledge

Automatizedimplicitortacitknowledgewillsometimescomeintoactionwithoutawarenessorevenin#uenceonthepartofthelearner.Onceasituationisrecognizedorde\"nedasoneinwhichacertainsetofroutinesorautomatizedproceduresisrelevant,thetransferprocess`runsonitsownaandperhapscannotbestopped.Butthisposesaproblemifitisincorrectorpartiallycorrectaccordingtoscienti\"corculturalstandards.Howcanalearnerbecomeawareofhisorhermisconceptionsiftheyareso`tacita?Somestudentsmaynotbeawareofthedi!erentkindsofpriorknowledge(formalandinformal,correctandincorrect)andtheiradvantagesanddisadvantages.

2.3.Paradoxofusingrelevantpriorknowledge

Oncestudentsareawareofrelevantpriorknowledge,theymaydecidenottouseit;theymay,forexample,beafraidofinterferenceofnewandpriorknowledge.Althoughitseemslogicaltomakeuseofallthepriorknowledgeyouhavetofacilitateyourlearning,therearealsoseveralgoodreasonsnottodoso.Learnersmaynotbeawareoftheimportanceoftheactiveuseofpriorknowledge.Usingpriorknowledgemayrequireagreatdealofwork,itmaycreateconfusion,itmaydistractyoufromthemainpoints,anditmaymakeyourlearningtooidiosyncratic.Thus,fromtheperspectiveofthelearner,theproblemiswhentousepriorknowledgeactivelyandwhentoprotectoneselffromitsin#uences.Inthisactiveuseofpriorknowledgetherearetwootherdilemmas:whentocompareoldandnewideasactivelyandwhentotestandevaluatethee!ectivenessofnewknowledgeandskillsinnewsituations.

P.R.J.Simons/Int.J.Educ.Res.31(1999)577}589581

2.4.Paradoxofrecognizingrelevantsituationsandconditions

Thisistheproblemofrecognizingrelevantsituationsandconditions.Whereasthepreviousparadoxfocusesontheactiveuseofpriorknowledge(ornot),thepresentonereferstotherecognitionofsituationsandconditionswheretransfercouldoccur.Inthesecaseslearnerssimplydonotseethattwoormoresituationsorconditionsaresimilar.Whenisasituationsimilartoanotherone?Indeed,therearesomanydimensionsonwhichsituationsdi!er(time,place,content,culture,mood,etc.).Bereiter(1995)describesthemainproblemoftransferasatransferbetweensituations.Therearesomanyenvironmentalconditionsthatmaydeterminetheuseofacertainstrategyorofcertainknowledge.Howcanlearners\"ndtheirwaysintheseenviron-mentalconditions?

Themainproblemforlearnersisthattheyoftentimesdonotandcannotknowwhenandwheretheyshouldandwillusewhattheylearned.Howcanoneprepareforsituationsonecannotknow?Especiallyinthecaseoffartransfer,wheretherearemanydi!erentapplicationpossibilities,itisdi$cultforlearnersto\"ndwaystopreparethemselvesforlateruseoftheirknowledge.Theonlytwothingsalearnercando(seealsoBereiter,1995)istostriveforrealanddeepunderstanding(optimizingtheaccessibilityoftheknowledge)andtocollectknowledgeaboutthesituationalcondi-tions.Inthisrespectthereisperhapsadispositionfortransferorasetofabilitiesrelatedtotransfer:`transfer-abilitya(Simons,1990).2.5.Paradoxofnearandfartransfer

Anotherparadoxrelatestothedistinctionbetweennearandfartransfer.Inneartransferthereisacloseconnectionbetweenthelearningsituation(orthepriorknowledge)andtheapplication(orthenewlearningsituation).Infartransferthedistancebetweenpriorknowledgeorlearningandapplication(orthesecondlearningsituation)ismuchgreater(seeMayer&Greeno,1972).Thisisnodichotomy;rather,itisadimensionofdistance.Thisdistancecansometimesbemeasuredormanipulated(seeBassok&Holyoak,1989).Basically,however,thedistanceisasubjectivemeasurethatvariesamongindividuals(Simons&Verscha!el,1992).Animportanthypothesisisthatonehastododi!erentthingsfornearandfartransfer;astrategyforneartransfermaybeinappropriateorine!ectiveforfartransferandtheotherwayaround.Forneartransferoneneedsthe`lowroadtotransfera(Salomon&Perkins,1989):toautomatizeandpracticeinasmallrangeofsituations(contextualization)(seeSimons,1990).Forfartransfer,however,the`highroada(Salomon&Perkins,1989)isbetter:decontextualizationandpracticeinavarietyofdi!erentsituationsareimportantstrategies.Forlearners,thebasicparadoxiswhethertogoforneartransferandtocon\"netherangeofsituations,focusingonpracticeandautomatization,ortogoforfartransfer,searchingfordecontextualizationandvariety.2.6.Theparadoxical`whataoftransfer

Whatshouldalearnertakewithhimorherfromonelearningsituationtoanother?SalomonandPerkins(1989)wrote:`Ingeneral,the&what'(oftransfer)mightbe

582P.R.J.Simons/Int.J.Educ.Res.31(1999)577}589

asubroutinedevelopedinthelearningcontextbutalsousefulinthetransfercontext,anoverarchingprincipleabstractedinthelearningcontextbutapplicableinthetransfercontext,apieceoffactualknowledgeusefulinbothbutinquitedi!erentways,alearningstrategythatbecomesusedinnewdomains,acognitivestyle,orevenacomplexstrategyofapproachingnewproblemsa(p.116).Collins,BrownandNewman(1989)distinguishfourtypesoftransferableelements:(a)domain-speci\"cknowledge(concepts,rules,algorithms);(b)heuristicproblem-solvingstrategies;(c)strategiesforself-regulation;and(d)learningstrategies.

Forlearnerstheparadoxamountstochoosingamongthevariouselementsthatcouldbetransferred.Whatshouldandcouldthelearnertakewithhimorhertoothersituations?Isitpossibletocombineseveralelements?Onespeci\"cdilemmaishowtotransferlearningskillstonewlearningsituations.Ineverylearningsituationtherearealsoopportunitiestolearnaboutthelearningprocessesthemselves.Whenpeoplealwayslearninthesameway,theywillprobablydevelopahabittolearnjustinthatway.Thiswayoflearningbecomesthenaturaloneandpeopleevenadapttheirconceptionsoflearninginthatdirection(Simons,1997).Thisbecomes,however,problematicwhennewwaysoflearningareneeded,forinstancewhenschoolsmakenewdemands,whenastudentchangesfromoneschooltoanother,oratthetransitionfromschooltowork.Thedilemmaforalearnerinthisrespectishowtolearnabouttheotherkindsoflearningsituations:akindoffartransferoflearningskillsandstrategies.

3.Threesetsofstudies

Inthissectionstudiesthatillustratetheinterrelationshipamongthethreeresearchquestionsarereviewed.Thestudiesallfocusedonhelpinglearnerstosolvesomeoralloftheparadoxesdescribedabove.

3.1.Dealingwithpriorknowledgeoflearners

Howcaninstructionalsystemshelpstudentsusetheircorrectorpartiallycorrectpriorknowledgewithoutcreatinginterferencewithnewknowledgeandatthesametimetodealwithwrongorpartiallyincorrectpriorknowledgethattendstoresistchange?Althoughmanystudiesdemonstratingtheproblemsofstudentsandinstruc-tionalsystemswithpriorknowledgearenowavailable,fewhaveshownhowtheymightbesolved.StrikeandPosner(1985)proposedatheoreticalframeworkaboutcognitiveaccommodation.Studentsshouldfeeldissatisfactionwiththeirexistingconceptions.Theyshouldunderstandnewscienti\"cnotionslyingbehindthenewconceptions.Thesenewconceptionsshouldenablestudentstosolveproblems.Theyshouldalsohaveopportunitiestoactuallytestandevaluatetheirnewconceptions.Prawat(1989)stressedtheimportanceofmetacognitiveawareness.`Studentsshouldbeawareoftheirpre-conceptionsbeforetheywillbeabletorestructurethem.Studentsmust\"rstrecognizethatthenewinformationisrelatedtowhattheyalreadyknow;theythenhavetolinkthisinformationtotwotypesofpriorknowledge*that

P.R.J.Simons/Int.J.Educ.Res.31(1999)577}589583

whichisconsistentwiththescienti\"cnotionsandthatwhichisincompatiblewiththosenotions.Itisthelatterconnectionthatleadstotherealizationthattheirownideasarenotcompleteorsatisfyingexplanationsandthatthescienti\"cviewisamoreconvincingandpowerfulalternativea(pp.12}13).AsSiegler(1983)showed,however,awarenessisnotenough:cognitiveaccommodationalsoimpliesacceptance.Studentsshould\"ndthenewnotionsplausible.NussbaumandNovick(1982)proposedtheso-called`exposingeventsathatevokestudents'pre-conceptions.HewsonandHew-son(1984)emphasizedcomparisonandcontrastingactivitiesthatstudentsshoulduse.InspiredbythesetheoriesAli(1990)andBiemans(Biemans&Simons,1995,1996)createdaninstructionalstrategy(theCONTACTstrategy)consistingof\"vesteps.Studentsshouldbe(1)helpedinsearchingtheirownrelevantpre-conceptions(throughanexposingevent);(2)stimulatedtocomparetheirpre-conceptionswiththenewinformation;(3)askedtoformulatethenewidea;(4)askedtoapplythenewconceptioninaconcreteproblemand(5)askedtoevaluatetheadequacyofthenewconceptioninrelationtostep4.Ofcourse,thisstrategyhelpstosolveonlyafewoftheproblemsofstudentsmentionedabove.Itisastrategythatisespeciallyusefulindomainswheretherearemanymisconceptionstobecorrectedthroughinstruction.Furthermore,itdoesnothelpstudentstosolvethemetacognitiveproblemsmen-tioned,likebecomingawareofthefunctionsofthedi!erentkindsofpriorknowledge,thecriticalroleofmisconceptions,andsoon.Finally,itisastrategythatcanonlybeappliedwhenthereissomeknowledgeaboutfrequentlyoccurringmisconceptions.InaseriesofstudiesBiemans(1997)showedthattheCONTACTstrategysigni\"-cantlyimprovedthelearningofstudentsof11}13yearsofage.Ninelessonsaboutphysicalgeography,includingconceptssuchasequator,earthrotation,condensation,rain,atmosphericpressure,wind,andweatherwerepresentedbycomputers.Eachlessonstartedwithaconcreteproblemorquestiontobesolvedbythestudentinorderto\"ndhisorherpre-conceptions(step1).Anexampleofsuchaquestionis`WhenColumbussetsailin1492,thewinddidnotblowhimstraighttotheequator.InsteadhewasblowntoAmericawithacurvetotheright.Canyouexplainthis?aSubjectshadtochoosefromsixalternativesandtogiveargumentationfortheirchoices.Thedistracteranswersrepresentedmisconceptionsfoundinpriorstudiestobetypicalforstudentsofthisagelevel.

Next,thenewinformationwaspresented.Studentswereaskedtocomparetheiroldideawiththeinformationpresented(step2).Thentheoriginalquestion(fromstep1)wasposedanothertimeandstudentshadtoselectthecorrectanswerandtogiveanargumentationforitagain(step3).Followingthat,anewbutsimilarconcreteproblemwaspresentedwhichstudentshadtosolvewiththenewlyacquiredknow-ledge(step4).Finally,studentshadtoevaluatethenewknowledgeandcomparetheoldandthenewknowledgeanothertime(step5).Ifstudentsgavethewronganswerinstep3,theywereaskedtore-readtheinformation.Foreachofthe\"vestepstherewasextrainformationavailableaboutitsrelevance(Whyisitimportant?)andhowtoperformit.Duringsomeofthelessonsthe`whyaand`howainformationwaspresentedtoallstudents.Inotherlessonsthestudentscouldconsultthisinformationwhenevertheywanted.The\"vestepswerealsovisuallyillustratedbywayofgraphicmetaphors.

584P.R.J.Simons/Int.J.Educ.Res.31(1999)577}589

Thedependentvariablesconsistedofmultiple-choicetests.Theimmediateachieve-menttestcontainednineitems(threereproduction,threeinsightandthreeproblem-solvingitems).Twoweeksafterthelastlessonwascompletedaretentiontestconsistingof18reproduction,18insightand18problemsolvingitemswasadminis-tered.Attheendofeachlessonsubjectsalsosolvedaconcreteproblemthatwasusedinthebeginningphasetoelicittheirpre-conceptions.Theanswerstotheseproblemswerecalledthe`\"nalideaa.Thefullstrategyofallstepswascontrastedwitha`before}afteraconditionwherestudentsonlyexecutedthe\"rstandthelaststepofthestrategyandreceivednofeedbackorextrahelpandwithacontrolcondition.Inonestudy,theCONTACTstrategyresultedinsigni\"cantlyhigherscoresonallthetestsandsubtests(i.e.,reproduction,insight,application)thanthebefore}aftercondition.Inaddition,thelatterstudentsperformedsigni\"cantlybetterthanstudentsinthecontrolcondition.TheresultsofanothersuggestedthatitwaspossibletoteachstudentstousetheCONTACTstrategyontheirown.Afterawhile,studentsweretaughthowtousetheCONTACTstepsindependently.Thehelpo!eredbythecomputerwassca!oldedandfadedoutsystematicallyandgradually.Thisstudyshowedthatalearningtolearnapproachresultedinbetterphysical}geographicalconceptions(\"nalidearesponses)andinhigherlearningperformanceonthemul-tiple-choiceteststhantheregularCONTACTstrategy.OnemayconcludefromthisstudythatitispossibletoteachstudentstolearnbythemselvesaccordingtotheprinciplesoftheCONTACTstrategy.3.2.Learningnowforlaterlearning

Ideally,learningenvironmentsdonotonlyhelpstudentsgetinvolvedinproductivelearningprocessesandachievehighlearningresults,butalsopreparethemforlaterlearning(see,forexample,Candy&Crebert,1990).Howcaninstructionalsystemshelpstudentstoprepareforfuturelearningsituations?Teurlings(1993)conductedastudythatdemonstrateshowtodothis.AdultstudentslearnedinaseriesoflessonshowtoworkwithWordperfect5.1(WP).Inthecontrolconditionthelessonswereorganizedliketheynormallyare.IntheexperimentalconditionstudentslearnednotonlytotypewritewithWP,butalsohowtolearntherestofWP(to\"ndtheirwayinthecomputer)ontheirown.Thiswasdoneusingthefollowingprocedure.

First,therewasavideodemonstratingthatlearningWPisnotamatterofphysicalskillacquisition(manystudentsthinkthatitislikelearningtotypewrite)butamatteroflearningtothinklikethecomputer.Second,bydrawingananalogywithacomplexbuildingstudentslearnhowto\"ndtheirwayinthecomputer:wheretolookatthescreen;howto\"ndoutwhereoneisintheprogram;whatthebasicstructureoftheprogramis.Third,studentslearnedhowtoworkwiththevarioushelpfunctions(aminimalmanual,on-linesupport).Fourth,thestudentswerestimulatedtoexperimentandtotryoutnewthings:nothingseriouscanhappenaslongasyoudonotmakeoneortwomistakes(likeformatc:).Fifth,theactualskilltrainingwasbroughtbacktothebareessentialstomaketimefortheinstructioninlearningskills.Thismeant,forinstance,thatstudentsonlylearnedtounderlinetextandthattheywereencouragedtolearnsimilarprocedureslikemakingtextboldontheirown.Theresultssuggested

P.R.J.Simons/Int.J.Educ.Res.31(1999)577}589585

thatthestudentsintheexperimentalconditionnotonlylearnedmorethanthoseinthecontrolcondition,but,moreimportantly,weremuchfasterandmuchmoree!ectiveinlearninganewpartofWPontheirown,bothimmediatelyafterthetrainingandsixweekslater.3.3.Learningfortransfertowork

Forthislastsectiontheunderlyingquestionis:howcaninstructionalandwork-relatedenvironmentalconditionsbedesignedinsuchawaythatlearnersarehelpedtosolvetheparadoxestheyencounter?Severalstudieswerecarriedoutfocusingonthistransferfromtrainingtoworksituations.Subjectswereadultsworkinginorganiza-tions.Inthedesign,therearethreewaystoproceed:focusingonfartransfer;focusingonneartransferandfocusingonorganizationalmeasures(withrespecttobothfarandneartransfer).Inordertopromotefartransferlearningenvironmentscanbedesignedtooptimizethechancefortransfertooccurfromthetrainingsituationtoseveraldi!erent`reallifeasituations.

Inessencethisisaquestionofimprovingtheaccessibilityofmemoryrepresenta-tions.Simons(1990)hasdescribedtheconditionsfoundintheliteraturethatincreasetheaccessibilityofinformationandskillsinmemory.Examplesinclude:

1.Increasingtheconnectednessofthememoryrepresentationbymakingtherela-tionsbetweenconceptsmoreexplicitorbycon\"ningtheinstructiontokeyconceptsthatareverycentralintherelationalnetwork;

2.Increasingthesubjectiverelevancethatpeopleattachtotheknowledgeandskillstobelearnedbyexplanationoftheirutilityorbygivingtheassignmenttostudentsto\"ndoutwhattherelevanceisontheirown;

3.Improvingthebindingoflearningwithvaryingcontextsofwork,forinstancethroughonthejobtrainingorbyusingthepracticalexperiencesthatworkerscanbringwiththem;

4.Increasingthemetacognitiveskillsofworkersbyteachingthemdirectlyorbycreatinglearningenvironmentsthatcalluponthesekindsofskills;

5.Broadeningthegeneralityofknowledgeandskillsbygivingopportunitiesforre#ectionorbyo!eringavarietyofpractice;and

6.Organizingana!ectiveclimatedirectedattransfer(a`transferculturea,seePea,1987).Intheliteratureabouttransferoftraininginindustrialorworksettings,however,theattentionforneartransferprevailsovertheattentionforfartransfer.BroadandNewstrom(1992),Claas,PouwerandThijssen(1986)andGaines-RobinsonandRobinson(1989)reviewedtheempiricalstudiesdoneandconcludedthat,bywayofexample,thefollowingprinciplesshouldguidethedesignofinstructionaimedpre-dominantlyatneartransfer:

1.Explicitlearninggoalsshouldbeformulated,aimingeitherforneartransferorforfartransfer;

586P.R.J.Simons/Int.J.Educ.Res.31(1999)577}589

2.Thesegoalsshouldfollowfromaneeds-assessmentandconnecttoprofessionalpractice;

3.Thegoalsshouldbeformulatedasconcretelyaspossible;

4.Thegoalsshouldbetunedtotheenteringlevelofthelearnersandanadaptationtodi!erencesinenteringlevelshouldbepossible;

5.Atthestartofacourseinformationshouldbegivenaboutthegoalsandthecontentsinordertocorrectincorrectexpectations;

6.Thechoiceoflearningcontentsshouldbesuchthatthereisaconnectionwithconcreteorsimulatedworksituations;

7.Thecriteriathatwillbeusedtojudgelearningperformanceshouldbeclear;and8.Learningperformanceshouldbeevaluatedandtestedregularly.Then,feedbackcanbegivenabouttheprogressandtheexistinggapsinknowledgecanbedetermined.Apartfromthesetwokindsoftransfer-promotinginstructionaldesigns,therearealsotransfer-promotingorganizationaldesigns(see,forinstance,Baldwin&Ford,1988;Claasetal.,1986).Thesearedesignsattheleveloftheorganizationthatpromotestransfertotheworksituation.Thismeans,ontheonehand,thatsomeofthe(instructional)designelementsaretranslatedintoactivitiesofmanagers.Ontheotherhand,therearespecialmeasuresoftheorganizationthatcanbetakenbeforeandafterlearningtakesplaceandthathelptheprocessoftransfertotheworkplace.Inessence,theseactivities\"tintothe`trainingforimpactaapproachdevelopedbyGaines-RobinsonandRobinson(1989).Themainstrategiesofthisapproachcanbesum-marizedasfollows:

1.Atforehand,importantdecision-makersshouldcommitthemselves.

2.Theyshouldagreethatthegoalsofthetrainingareimportantfortheorganizationandtheyshouldconvinceotherrelevantpeoplethatthisisthecase

3.Severalpeopleintheorganizationshouldcontrolthatthegoalsmatchrealandimportantbusinessneeds.

4.Anadequateselectionofparticipantsisimportant(havingtherequiredpriorknowledgeandbeingreadytolearn).

5.Agoodsupport-systemintheworkenvironmentisthoughttobeimportant(forinstanceexemptingworkersfromtheirjobwhenfollowingacourseandhavingsomeonetakeovertheirwork).

6.Aftertraining,itse!ectsshouldbemadeexplicit:Boththedirectlearningresultsandthein#uenceontheworkbehaviorandontheorganizationasawhole.Claasetal.(1986)reviewedtherelevantliteratureandfoundempiricalsupportfortene!ectiveorganizationalelements.Examplesinclude:

1.Gapsbetweenthelearninggoalsandorganizationalgoalsshouldbediagnosedandshouldleadtosupplementingintegrationmeasures;

P.R.J.Simons/Int.J.Educ.Res.31(1999)577}589587

2.Thereshouldbean`integrationcounselorawhocreatesandmonitorsthecondi-tionsforoptimalintegration;

3.Thetimespanbetweenthelearningandtheuseinworkenvironmentshouldbeasshortaspossible;and

4.Thelearnershouldgetampleopportunitiestousethelearnedknowledgeandskillsintheworkenvironment.Inseveralstudiestheuseinpracticeofthevariousinstructionalandorganizationalnearandfartransfer-promotingelementswasinvestigated.The\"ndings(forexample,Geelen&Hendriks,1998;Kerstens,1990;TerHeinen&Termote,1990)showedthatmostofthepossibletransfer-promotingmeasureselementswerenotevidentindailypractice.Elementsofneartransfertendtodominate.TerHeinenandTermote(1990))showedthattheextenttowhichtransferelementswereinplacewassigni\"cantlycorrelatedwiththeadequacyofapplicationoftheoryinwork-practice.

Lether(1997)studiedthetransfere!ectsofacourseforinsuranceagentswhohadtobecomemoreemployable.Telephoneoperatorswhopreviouslysoldhomeownersandautomobileinsurancehadtolearnhowtoselltravelinsurance.Inatrainingprogramtheylearnedallaboutthecontentsoftravelinsurance.Furthermore,theylearnedstrategiesforcross-selling(howtoseduceclientstobuyotherproductsthantheonestheyaskfor)andhowtobeclient-centered(howtoactinaccordancewiththeneedsoftheclient).Mostoftheinstructionaldesignelementsintendedtopromotetransferasdescribedabovewereincludedinthetrainingprogram.Theorganizationaldesignelements,however,werenotincludedatall.

Thereweretwovariantsofthetraining.Inthe\"rst,therewasaregularclassroominstruction.Inthesecond,the`contacthoursawiththeteacherwereminimal;studentslearnedindependentlyusingaspeciallypreparedself-instructionpackage.Onanimmediatepost-testtheself-instructiongroupoutperformedtheclassroomgroupsigni\"cantly.Asimulatedclient(aso-called`mysterycallera)phonedallparticipantstodeterminewhetherthetelephoneoperatorswouldtrytodocross-sellinginunobservedpractice.Iftheywouldnotdothisspontaneously,thesimulatedclientwouldgivehintslike`Iwillsoontakeaholidaya.Iftheoperatorsstillwouldnottrytosell,theclientwoulddirectlyaskforthetravelinsurance.Theresultsofthismysterycallingwerequiteclear:almostnoneofthe43operatorso!eredtravelinsurancesspontaneously.Whenpromptedasmallnumberofpeopleo!eredaninsurancethemselves,butmostofthemconnectedtheclienttothetravelinsurancedepartment.Therewasasmall(butsigni\"cant)di!erenceinthesetransferresultsinfavoroftheself-instructiongroup.

Instructionaimedatpromotingtransferapparentlydidnotresultinahighdegreeoftransfereitherundertheself-instructionortheclassroomcondition.Theself-instructionstudents,however,hadslightlyhighertransfer`scoresathandidthoseintheclassroomgroup.Perhapssomeoftheorganisationaldesignelementsshouldhavebeenincludedinordertoreallyhelptheinsurancesellerstoovercomethetransferparadoxes.Othercausesofthelowtransfermightbethattheworkersfeltinsecure,thattheydidnotlikethephenomenonofcross-selling,orthattheyhadtoworkundergreattimepressure.

588P.R.J.Simons/Int.J.Educ.Res.31(1999)577}589

4.Conclusions

Therearethreeseparatebutinterrelatedtransferquestionsthathavemuchincommon.Theyalldealwiththeuseofknowledgeandskillslearnedinonesituationinothersituations,betheylearningorworksituations.Inpromotingtransfer,itisusefultostartfromthelearner'sperspective.Whataretheparadoxesthatlearnersencounterwhentheyaremovingtowardtransfer?Inpromotingtransferitisimportanttotaketheseparadoxesintoaccountandtotrytohelplearnersto\"ndsolutionsorwaystodealwiththecontradictionsinherentinthem.Thestudiesmentionedhereshowedhowthiscanbedoneandthatimportante!ectscanbereached.Transfercanbepromotedinvarioussettingsifandwhenwehavethelearner'sperspectiveasthefocus.Youngstudentscanbehelpedtousetheirpriorknowledgemoreactivelyandtoovercomesomeoftheirpre-conceptions.Moreover,theycanlearnhowtodothisontheirown.Adultscanlearnhowtolearntotransfertheirknowledgeandskillswhenalearningtolearnapproachisembeddedintotheirregulartraining.References

Ali,K.(1990).Instructiestrategieenvoorhetactiverenvanpreconcepties.[Instructionalstrategiesfortheactivationofpriorknowledge.]Dissertation,TilburgUniversity.

Bassok,M.,&Holyoak,K.J.(1989).Interdomaintransferbetweenisomorphictopicsinalgebraandphysics.JournalofExperimentalPsychology:Learning,MemoryandCognition,15,153}166.

Baldwin,T.T.,&Ford,J.K.(1988).Transferoftraining:Areviewanddirectionsforfutureresearch.PersonnelPsychology,41,63}105.

Biemans,H.(1997).Fosteringactivationofpriorknowledgeandconceptualchange.Dissertation,Univer-sityofNijmegen.

Biemans,H.J.A.,&Simons,P.R.J.(1995).Howtousepreconceptions?Thecontactstrategydismantled.EuropeanJournalofPsychologyofEducation,10,243}259.

Biemans,H.J.A.,&Simons,P.R.J.(1996).Acomputer-assistedinstructionalstrategyforpromotingconceptualchange.InstructionalScience,24,157}176.

Broad,M.L.,&Newstrom,J.W.(1992).Transferoftraining.Actionpackagedstrategiestoensurehighpayowfromtraininginvestments.Reading:Addison-Wesley.

Candy,P.C.,&Crebert,R.G.(1990).Teachingnowforlearninglater:Thetransferoflearningskillsfromtheacademytotheworkplace.Paperpresentedatthe8thAustralianlearningandlanguageconference.Brisbane,July.

Claas,L.Pouwer,M.,&Thijssen,J.G.L.(1986).Transferinhetkadervanbedrijfsopleidingen.[Transferinthecontextofindustrialtraining.]InM.G.M.Geerdinketal.(Red.)Wetensvaardigheden.DenHaag:ROI/CIVOB.

Collins,A.,Brown,J.S.,&Newman,S.E.(1989).Cognitiveapprenticeship:Teachingthecraftofreading,writingandmathematics.InL.B.Resnick,Knowing,learningandinstruction.EssaysinhonourofRobertGlaser(pp.453}494).Hillsdale:Erlbaum.

Cormier,S.M.,&Hagman,J.D.(1987).Introduction.InS.M.Cormier&J.D.HagmanTransferoflearning;contemporaryresearchandapplications(pp.1}8).SanDiego:AcademicPress.

DeJong,F.P.C.M.,&Simons,P.R.J.(1990).Cognitiveandmetacognitiveprocessesofself-regulatedlearning.InJ.M.Pieters,P.R.J.Simons,&L.deLeeuw,Researchoncomputer-basedinstruction(pp.81}100).Lisse:SwetsandZeitlinger.

Dochy,F.J.R.C.(1992).Assessmentofpriorknowledgeasadeterminantforfuturelearning.DoctoralDissertation.Heerlen:OpenUniversity.

Eylon,B.,&Linn,M.C.(1988).Learningandinstruction:Anexaminationoffourresearchperspectivesinscienceeducation.ReviewofEducationalResearch,58,251}301.

P.R.J.Simons/Int.J.Educ.Res.31(1999)577}589589

FletcherJ.,&Olwyler,K.(1997).Paradoxicalthinking.SanFrancisco:Berrett-Koehler.

Gaines-Robinson,D.,&Robinson,J.C.(1989).Trainingforimpact.SanFrancisco:Jossey-Bass.

Geelen,M.,&Hendriks,E.(1998).ProjectmanagementAchmea.Thesis.Nijmegen:UniversityofNijmegen.Handy,C.(1994).Theageofparadox.Cambridge:HarvardBusinessSchool.

Hewson,P.W.,&Hewson,M.G.A.(1984).Theroleofconceptualcon#ictinconceptualchangeandthedesignofscienceinstruction.InstructionalScience,13,1}13.

Kerstens,J.(1990).Transfer-bevorderendeactiviteitenindeopleidingtotpsychiatrischverpleegkundige.[Transfer-promotingactivitiesinthetrainingofpsychiatricnurses.]Thesis.Nijmegen:UniversityofNijmegen.

Lether,E.B.(1997).Transfer-e!ectvaneenbedrijfsopleiding.[Transfer-e!ectofacorporatetraining.]Thesis.Nijmegen:UniversityofNijmegen.

Mayer,R.,&Greeno,J.G.(1972).Structuraldi!erencesbetweenlearningoutcomesproducedbydi!erentinstructionalmethods.JournalofEducationalPsychology,63,165}173.

Nussbaum,J.,&Novick,S.(1982).Alternativeframeworks,conceptualcon#ictandaccommodation:Towardsaprincipledteachingstrategy.InstructionalScience,11,183}200.

Pea,R.D.(1987).Socializingtheknowledgetransferproblem.InternationalJournalofEducationalResearch,11,639}663.

Prawat,R.S.(1989).Promotingaccesstoknowledge,strategyanddispositioninstudents:Aresearchsynthesis.ReviewofEducationalResearch,59,1}41.

Quinn,R.,&Cameron,K.S.(1988).Paradoxandtransformation.Cambridge:Ballinger.

Salomon,G.,&Perkins,D.N.(1989).Rockyroadstotransfer:Rethinkingmechanismsofaneglectedphenomenon.EducationalPsychologist,24,113}142.

Siegler,R.S.(1983).Fivegeneralizationsaboutcognitivedevelopment.AmericanPsychologist,38,263}277.Simons,P.R.J.(1990).Transfervermogen.[Transfer-ability]Inaugurallecture.Nijmegen:QuickPrint.Simons,P.R.J.(1997).Ontwikkelingvanleercompetenties.[Developmentoflearningcompetences.]OpleidingenOntwikkeling,10,17}20.

Simons,P.R.J.(1998).Paradoxesoflearningandteaching.InvitedaddresspresentedattheBritishPsychologicalSociety,Brighton,March27.

Simons,P.R.J.,&Verscha!el,L.(1992).(Red.).Transfer:onderzoekenonderwijs.[Transfer:researchandeducation.]Tijdschriftvooronderwijsresearch,17,3}16.

TerHeinen,B.,&Termote,A.(1990).Hetgeleerdeinpraktijkgebracht:eenonderzoeknaardeinvloedvantransfer-bevorderendemaatregelenindepraktijk-opleidingtotradio-diagnostischlaboratoryworker[Practisingwhatyoulearned:researchonthein#uenceoftransfer-promotingmeasuresintrainingradio-diagnosticlaboratoryworkers].Thesis.Nijmegen:UniversityofNijmegen.

Strike,K.A.,andPosner,G.J.(1985).Aconceptualchangeviewoflearningandunderstanding.InL.H.T.WestandA.L.Pines,Cognitivestructureandconceptualchange(pp.211}231).Orlando:AcademicPress.

Teurlings,C.C.J.(1993).Lerentekstverwerken:eennieuwperspectief.[Learningtouseawordprocessor:anewperspective.]Tilburg:TilburgUniversity.

P.Robert-JanSimonsstudiedpsychologyattheUniversityofAmsterdam(Nether-lands)andreceivedhisPh.D.fromtheUniversityofTilburg.Hisdoctoralthesiswasontheroleofconcreteanalogiesinlearning.Since1990hehasservedasaprofessorofeducationalpsychologyattheUniversityofNijmegen(Netherlands)andcurrentlyisthescienti\"cdirectoroftheNijmegenInstituteforResearchonPedagogyandEducation.Hisresearchliesintheareaofthepsychologyoflearning.