理解的認(rèn)知?dú)v程研究-嘉義大學(xué)課件_第1頁(yè)
理解的認(rèn)知?dú)v程研究-嘉義大學(xué)課件_第2頁(yè)
理解的認(rèn)知?dú)v程研究-嘉義大學(xué)課件_第3頁(yè)
理解的認(rèn)知?dú)v程研究-嘉義大學(xué)課件_第4頁(yè)
理解的認(rèn)知?dú)v程研究-嘉義大學(xué)課件_第5頁(yè)
已閱讀5頁(yè),還剩165頁(yè)未讀 繼續(xù)免費(fèi)閱讀

下載本文檔

版權(quán)說(shuō)明:本文檔由用戶提供并上傳,收益歸屬內(nèi)容提供方,若內(nèi)容存在侵權(quán),請(qǐng)進(jìn)行舉報(bào)或認(rèn)領(lǐng)

文檔簡(jiǎn)介

CognitiveInvestigationsofComprehensionProcesses曾玉村YuhtsuenTzengLearning&CognitionLab,NationalChungChengUniversityPresentedatGSME,NCYU,May,04,2011CognitiveInvestigationsofCo1TounderstandishumannatureComprehension/understandingrestsonmemoryComprehensionpromotereasoning,problemsolving,creation…TounderstandistoactivateoldknowledgeTounderstandistoafirststepoflearningEducationshouldaimatunderstandingTounderstandishumannatureC2Whatiscomprehension/understanding?LanguageNon-languageWhatiscomprehension/understa3Whatislanguagecomprehension?ComplexcognitiveprocessesTheconstructionofacoherentmentalrepresentationofthetext(e.g.,Kintsch,1988;Graesser&Clark,1985;Trabasso&vandenBroek,1985)ThereaderidentifiesmeaningfulrelationsbetweenpartsofthetextbetweentextandbackgroundknowledgeWhatislanguagecomprehension4Whatiscomprehension?

TextinputprovidesrawelementsforcomprehensionWhatiscomprehension?

Texti5Whatiscomprehension?

ReadersconnecttextelementstobuildanetworkrepresentationWhatiscomprehension?

Reader6WhatdotheLSmodelssay?

Howdoreadersachievecomprehension?Asreadersproceedthroughthetext,theyattempttomaintaincoherenceforeachnewtextsegment(sentence)(Re)activateinformationfrompriortextorbackgroundknowledgeE.g.,referentialandcausalcoherenceNaming/lexicaldecision/speededrecognitionaswellasreadingtimestudies(e.g.,O’Brien&Myers,1989;McKoon&Ratcliff,1990)InferencesarecrucialtogobeyondtextinputsWhatdotheLSmodelssay?

Ho7FormalizationoftheLSmodelAttentional/WorkingmemorylimitationsComprehensioninvolvesfluctuationofactivationfromTextinputCarryoverReinstatement(fromrepresentationofpriortext)ActivationofbackgroundknowledgeCohortactivation(seebelow)landscapeofactivationsFormalizationoftheLSmodelA8理解的認(rèn)知?dú)v程研究-嘉義大學(xué)課件9TheLSmodel:ConstructingamemoryrepresentationConnectionsareidentifiedbetweentextitemsthatareactivatedsimultaneouslyTheepisodicmemoryrepresentationisupdatedasaresultofeachnewactivationvector,viaanasymptotic(delta)ruleTheLSmodel:Constructingam10LSmodel:DynamicinteractionbetweenactivationandrepresentationCo-activatedelementsformacohortintheepisodicrepresentationCohortactivationprovidesadditionalactivationduringreadingprocessConsistentwithcurrentmodels(e.g.,Construction-Integrationmodel,Kintsch1988;Resonancemodel,O’Brien&Myers)LSmodel:Dynamicinteraction11理解的認(rèn)知?dú)v程研究-嘉義大學(xué)課件12MathematicalFormalizationsofLSmodelThreephasesofLSmodelDetermineinputComputecohortUpdateepisodicmemoryMathematicalFormalizationsof13DetermineInputDetermineInput14ComputeCohortActivation-Equation1ComputeCohortActivation-Equ15TransformConnectionStrengths-Equation2TransformConnectionStrengths16SigmoidTransformationSigmoidTransformation17ComputeExpectancy-

Equation3ComputeExpectancy-

Equation318UpdateConnectionStrengths-Equation4UpdateConnectionStrengths-E19PartialMemoryRepresentationatReadingCycle3PartialMemoryRepresentation20PartialMemoryRepresentationatReadingCycle13PartialMemoryRepresentation21TheLSprogram:InputTheLSprogram:Input22TheLSprogram:InputTheLSprogram:Input23TheLSprogram:

OutputofactivationpatternsTheLSprogram:

Outputofact24TheLSprogram:

OutputofconnectionpatternsTheLSprogram:

Outputofcon25理解的認(rèn)知?dú)v程研究-嘉義大學(xué)課件26理解的認(rèn)知?dú)v程研究-嘉義大學(xué)課件27Offlinememoryrepresentationoftextcomprehension(Tzeng&Chen,2006)

TextbaseSituationmodelOfflinememoryrepresentation28Offlinememoryrepresentationoftextcomprehension(曾玉村&陳秋芬,修改中)Offlinememoryrepresentation29Howdoescausalstructureoftextaffectonlinereadingprocesses?Aminimalistview:readersmakeminimumamountinferences(McKoon&Ratcliff,1992)AbottomupapproachofinferencemakingsMany“inferences”arepassivememoryactivationprocessesReaderswillmakeinferencesonlyifNoreadilyavailableinformationinWMCoherencebreaksintextsHowdoescausalstructureoft30Howdoescausalstructureoftextaffectonlinereadingprocesses?Aconstructionistview:readerswouldsearchformeaning(Graesser,Singer,&Trabasso,1994)Readerswouldmakemany(butnotall)typesofinferencestosatisfycoherenceassumptionandexplanationassumptionCoherenceassumption:needlocal&globalcoherencetobuildsituationmodelsExplanationassumption:toexplainwhycertaineventoccurHowdoescausalstructureoft31AcontrastingpredictionAminimalistviewpredictsnoinferenceswillbemadeiftextshavelocalcoherenceAconstructionistviewpredictsthatinferenceswillbemadeeveniftextshavelocalcoherencebutdonothaveglobalcoherenceAcontrastingpredictionAmini32LabelHierarchicalVersionSequentialVersion1.Setting2.Initiating13.Goal14.Action15.Outcome16.Reaction17.Initiating28.Goal29.Action210.Action311.Outcome212.Action413.Action514.Outcome3從前有個(gè)男孩叫吉米有一天他看到他的朋友騎著一輛新腳踏車吉米也想要買一輛新腳踏車他去告訴媽媽他媽媽拒絕幫他買腳踏車.吉米很傷心他媽媽告訴他應(yīng)該要自己存一些錢吉米想要賺一些錢.他就在雜貨店找到工作他替雜貨店外送貨品他賺了不少錢.他去一家百貨公司他走到二樓他買了一部新腳踏車.從前有個(gè)男孩叫吉米有一天他看到他的朋友騎著一輛新腳踏車吉米也想要買一輛新腳踏車他去告訴媽媽他媽媽答應(yīng)幫他買腳踏車.吉米很高興他媽媽告訴他應(yīng)該要自己存一些錢吉米想要賺一些錢.他就在雜貨店找到工作他替雜貨店外送貨品他賺了不少錢.他去一家百貨公司他走到二樓他買了一個(gè)新的籃球.LabelHierarchicalVersionSeque33Doreadersmakeonlineglobalinferences?Procedures22collegestudents32storieswereconstructedoutof16scenarios,halfsequential&halfhierarchicalParticipantsengagedinself-pacedreadingwhilerecordingtheireyemovementsusingEyelinkIIDoreadersmakeonlineglobal34Eyetrackingsystem

Eyetrackingsystem35EyetrackingsystemEyetrackingsystem36ResultsofeyemovementdataFocusingontwotypesofsentencesMajorgoals&consequences(i.e.,3th、5th、8th、11th&14thsentences)Spillovereffectsofthem(i.e.,4th、6th、9th&12thsentences)ResultsofeyemovementdataFo37ResultsofeyemovementdataFirstGaze:

thesumoftimeareader’sfirstlookofasentencetothefirsttimes/heexitsthatsentence

nodifferencesonbothversionsexceptforoutcome3longerdurationforsequentialthanhierarchicalstories(889vs.777msrespectively,t=2.56,p=.011)

allotherps>.05

ResultsofeyemovementdataFi38ResultsofeyemovementdataSecondGaze:thesumoftimereaders’fixationsontheirsecondpassofasentencelongerforsequentialversionthanhierarchicalversiononoutcome1(1159vs.818ms,t=2.61,p=.009)Resultsofeyemovementdata39ResultsofeyemovementdataTotalduration:thesumoffirst-gazeandsecond-gazelongerforsequentialversionthanforhierarchicalversiononoutcome1(2258vs.1789ms,t=2.65,p=.009)longerforAction1forsequentialversion(aspilloverregionofGoal1)(2350vs.1968ms,t=2.06,p=.04)ResultsofeyemovementdataTo40ResultsofeyemovementdataPatternoffirstGazeCognitiveloadingforbothversionisthesamesincethewordingsandsyntaxarealmostidenticalResultsofeyemovementdataPa41ResultsofeyemovementdataPatternofsecondgazeLongerfixationtimeforsequentialversionsonOutcome1Readersattempttogobacktointegratesentencestobuildacoherentrepresentationsinceitstartanewepisodeafterthissentence(i.e.,lackofglobalcoherence)ResultsofeyemovementdataPa42ResultsofeyemovementdataPatternoftotalgazeOutcome1marksthefirstdifferencebetweentwoversionsoftextthereforereaderspaydifferentamountoftimethiseffectspillbacktoAction1ThereisacoherencebreakbetweenthefirstandthesecondepisodeLackofglobalcoherenceinsequentialstoriescausesreadersspendmoretimeonthefirstepisodeResultsofeyemovementdataPa43ConclusionsofeyemovementdataCausalstructureaffectonlinereadingtimesFixationtimeincreaseiftextslackofcross-episode(global)coherenceHighcausalcoherenttextsresultinfewerlookingbackbehaviorsSupportaconstructionistviewConclusionsofeyemovementda44KeyIssuesofcomprehensionresearchSuccessfuldecodingFormtext-basedrepresentationsextractpropositionsfromtexts(Kintsch&vanDijk,1978)Formsituationmodel:whatthetextisaboutGobeyondtextAddingbackgroundkknowledge(vanDijk&Kintsch,1983;Kintsch,1988)KeyIssuesofcomprehensionre45InferencesarecrucialforconstructingsituationmodelThreeturtlesfloatingonalog.Severalfishswimmingunderthelog.Inferencesarecrucialforcon46WhatareinferencesInformationnotexplicitintext&readersretrievefrommemory(Kintsch,1998;Graesser,Singer,&Trabasso,1994;Till,Mross&Kintsch,1988)「他把鎖好的門打開」「鑰匙」「氣憤的老婆把杯子往地板上摔」「碎掉」WhatareinferencesInformation47閱讀理解的推論能力Weaver與Kintsch(1991)估計(jì)讀者每讀一則文章之?dāng)⑹?,平均必須進(jìn)行上打的推論才能真正完全理解文意理解需要多少推論以及何種推論有待釐清閱讀理解的推論能力Weaver與Kintsch(19948閱讀理解的推論能力最小量論(minimalist)(McKoon&Ratcliff,1992)若文本具有局部連貫性,讀者則沒有進(jìn)行整體推論連貫之必要,除非相關(guān)訊息可免費(fèi)取得建構(gòu)論(constructivist)(Graesser,Singer,&Trabasso,1993)讀者會(huì)進(jìn)行必要之推論以獲取文章整體意義,滿足讀者尋求意義的需求閱讀理解的推論能力最小量論(minimalist)49閱讀理解的推論能力發(fā)展傳統(tǒng)閱讀發(fā)展停留於整體能力指標(biāo)描述傳統(tǒng)閱讀能力發(fā)展研究往往缺乏理論推論能力發(fā)展研究較具特定性推論能力發(fā)展研究有潛力提早診斷閱讀理解困難閱讀理解的推論能力發(fā)展傳統(tǒng)閱讀發(fā)展停留於整體能力指標(biāo)描述50兒童讀者推論能力的發(fā)展Song和Fisher(2005,2007)發(fā)現(xiàn),幼兒階段已開始發(fā)展代名詞使用與理解的能力。針對(duì)兒童讀者所做的研究,發(fā)現(xiàn)代名詞推論(pronounresolution)的能力可以有效預(yù)測(cè)兒童的閱讀理解表現(xiàn)。(Hickmannetal.,1995;Kail&Hickmann,1992;Oakhill&Yuill,1986)。兒童讀者推論能力的發(fā)展Song和Fisher(2005,51中文兒童推論能力發(fā)展曾玉村、陳明蕾與陳沛嵐(Tzeng,Chen,&Chen,2007)尋找中文讀者特有的理解特性小六學(xué)生閱讀完整指稱詞的文章與極少指稱詞(正常文章盡量刪除指稱詞)的文章的理解並無(wú)差異小四學(xué)生對(duì)於極少指稱詞文章的理解低於完整指稱詞的文章黃秋華、曾玉村與陸偉明(Huang,Tzeng,&Lu,2009)則進(jìn)一步尋求年幼讀者指稱詞推論能力的發(fā)展的線上指標(biāo)指稱詞推論能力介於四到六年級(jí)似乎有明顯變化六年級(jí)對(duì)於零代詞的處理能力似乎已經(jīng)發(fā)展成熟中文兒童推論能力發(fā)展曾玉村、陳明蕾與陳沛嵐(Tzeng,C52研究目的了解兒童因果推論與指稱詞推論在理解歷程中之發(fā)展與可能的交互影響同時(shí)收集線上與離線資料研究目的了解兒童因果推論與指稱詞推論在理解歷程中之發(fā)展與可能53研究假設(shè)與預(yù)測(cè)邏輯短文前後句有無(wú)代名詞或不同強(qiáng)度之因果關(guān)係時(shí),會(huì)影響讀者閱讀理解的心理表徵,進(jìn)而影響回答問(wèn)題之正確率與閱讀時(shí)間預(yù)測(cè)六年級(jí)優(yōu)於四年級(jí)年級(jí)與短文連貫性有交互作用高因果優(yōu)於低因果短文有無(wú)代名詞短文僅對(duì)四年級(jí)有影響研究假設(shè)與預(yù)測(cè)邏輯54實(shí)驗(yàn)方法受試者:臺(tái)南市國(guó)小學(xué)生兩班排除特殊兒童以及上學(xué)期國(guó)語(yǔ)科成績(jī)低於兩個(gè)標(biāo)準(zhǔn)差者一班四年級(jí)學(xué)生23人一班六年級(jí)學(xué)生30人實(shí)驗(yàn)方法受試者:55實(shí)驗(yàn)材料材料的形式為前後兩句的極短文由大學(xué)生事先評(píng)定後再挑選與修改短文共54篇:實(shí)驗(yàn)短文36篇,filler18篇操弄因果連貫性高因果連貫:米奇和高菲穿著外套跳繩,他流很多汗低因果連貫:米奇和高菲剪頭髮動(dòng)不停,他喝到髒水操弄指稱詞關(guān)連性有指稱詞連貫:米奇和高菲穿著外套跳繩,他流很多汗無(wú)指稱詞連貫:米奇和高菲穿著外套跳繩,流很多汗短文主角均為兒童熟悉的八個(gè)卡通人物進(jìn)行變換實(shí)驗(yàn)材料材料的形式為前後兩句的極短文56實(shí)驗(yàn)材料實(shí)例高因果連貫+有指稱詞連貫米奇和高菲穿著外套跳繩,他流很多汗。高因果連貫+無(wú)指稱詞連貫米奇和高菲穿著外套跳繩,流很多汗。低因果連貫+有指稱詞連貫米奇和高菲剪頭髮動(dòng)不停,他喝到髒水。低因果連貫+無(wú)指稱詞連貫米奇和高菲剪頭髮動(dòng)不停,喝到髒水實(shí)驗(yàn)材料實(shí)例高因果連貫+有指稱詞連貫57實(shí)驗(yàn)程序指導(dǎo)語(yǔ)練習(xí)階段,熟悉程序與按鍵的操作正式實(shí)驗(yàn)所需時(shí)間約為40分鐘指導(dǎo)語(yǔ)練習(xí)題正式題(分成兩階段)實(shí)驗(yàn)程序指導(dǎo)語(yǔ)指導(dǎo)語(yǔ)練習(xí)題正式題(分成兩階段)58實(shí)驗(yàn)程序?qū)嶒?yàn)材料用ExperimentBuilder呈現(xiàn)每一短文分成三個(gè)畫面:米奇和高菲穿著外套跳繩,他流很多汗。米奇穿外套作運(yùn)動(dòng)嗎?實(shí)驗(yàn)程序?qū)嶒?yàn)材料用ExperimentBuilder呈現(xiàn)米59實(shí)驗(yàn)程序受試者依自己的速度按空白鍵閱讀短文受試者在第三個(gè)畫面必須回答問(wèn)題正確按「○」的按鍵錯(cuò)誤按「ㄨ」的按鍵紀(jì)錄受試者反應(yīng)正確率及每句閱讀時(shí)間實(shí)驗(yàn)程序受試者依自己的速度按空白鍵閱讀短文60實(shí)驗(yàn)設(shè)計(jì)2(年級(jí))x2(因果連貫)x2(代名詞連貫)年級(jí)為受試者間設(shè)計(jì)因果連貫與代名詞連貫為受試者內(nèi)設(shè)計(jì)實(shí)驗(yàn)設(shè)計(jì)2(年級(jí))x2(因果連貫)x2(代名詞連貫)61結(jié)果:答對(duì)率

高因果有代名詞高因果無(wú)代名詞低因果有代名詞低因果無(wú)代名詞四年級(jí)0.88(.12)0.90(.08)0.82(.11)0.76(.19)六年級(jí)0.94(.08)0.93(.09)0.86(.12)0.94(.07)結(jié)果:答對(duì)率

高因果高因果低因果低因果四年級(jí)0.88(.162理解的認(rèn)知?dú)v程研究-嘉義大學(xué)課件63結(jié)果:答對(duì)率三因子交互作用不顯著六年級(jí)答對(duì)率高於四年級(jí)(F=22.601,p=.000<.05)六年級(jí)的部分因果關(guān)係與代名詞有交互作用(F=6.857,p=.01<.05)在高因果關(guān)係條件下,代名詞的效果沒有顯著差異(F=.113,p=.738>.05)在低因果關(guān)係的條件下,無(wú)代名詞的答對(duì)率優(yōu)於有代名詞(F=9.983,p=.003<.05)在有代名詞的條件下,高因果優(yōu)於低因果(F=7.934,p=.007<.05)在無(wú)代名詞的條件下,因果關(guān)係沒有顯著差異(F=.476,p=.493>.05)結(jié)果:答對(duì)率三因子交互作用不顯著64結(jié)果:答對(duì)率四年級(jí)的部分因果關(guān)係與代名詞交互作用沒有顯著因果關(guān)係的主要效果達(dá)顯著(F=13.579,p=.000<.05)高因果優(yōu)於低因果代名詞的主要效果沒有顯著(F=.277,p=.600>.05)結(jié)果:答對(duì)率四年級(jí)的部分65結(jié)果:答對(duì)率年級(jí)較大正確率較高四年級(jí)完全受因果連貫強(qiáng)度影響六年級(jí)也受因果連貫性影響閱讀高因果連貫短文,不受代名詞有無(wú)之影響閱讀低因果連貫短文,沒有代名詞之短文較佳結(jié)果:答對(duì)率年級(jí)較大正確率較高66結(jié)果:第二句的閱讀時(shí)間

(單位:ms)

高因果有代名詞高因果無(wú)代名詞低因果有代名詞低因果有代名詞四年級(jí)1250.54(382.57)1508.43(568.58)1739.73(740.69)2002.46(923.61)六年級(jí)1016.93(282.56)1205.58(377.34)1464.08(552.78)1705.45(430.44)結(jié)果:第二句的閱讀時(shí)間高因果高因果低因果低因果四年級(jí)12567理解的認(rèn)知?dú)v程研究-嘉義大學(xué)課件68結(jié)果:第二句的閱讀時(shí)間三因子交互作用不顯著六年級(jí)閱讀時(shí)間短於四年級(jí)(F=10.905,p=.001<.05)六年級(jí)的部分因果關(guān)係與代名詞沒有交互作用因果關(guān)係的主要效果達(dá)顯著(F=37.731,p=.000<.05)高因果連貫閱讀時(shí)間較短代名詞的主要效果達(dá)顯著(F=7.780,p=.006<.05)有代名詞閱讀時(shí)間較短結(jié)果:第二句的閱讀時(shí)間三因子交互作用不顯著69結(jié)果:第二句的閱讀時(shí)間四年級(jí)的部分因果關(guān)係與代名詞交互作用沒有顯著因果關(guān)係的主要效果達(dá)顯著(F=11.882,p=.001<.05)高因果連貫閱讀時(shí)間較短代名詞的主要效果沒有顯著(F=3.331,p=.071>.05)結(jié)果:第二句的閱讀時(shí)間四年級(jí)的部分70結(jié)果:第二句的閱讀時(shí)間年級(jí)大閱讀時(shí)間較短具有較高因果連貫性文章,對(duì)四六年級(jí)均較易閱讀具有明確代名詞連貫性文章,有助於六年級(jí)閱讀(四年級(jí)尚未定)結(jié)果:第二句的閱讀時(shí)間年級(jí)大閱讀時(shí)間較短71進(jìn)行中未來(lái)三年更多發(fā)展性資料涵蓋二到六年級(jí)學(xué)生陸續(xù)收集中進(jìn)一步探討推論能力的大腦機(jī)制進(jìn)行中未來(lái)三年更多發(fā)展性資料涵蓋二到六年級(jí)學(xué)生陸續(xù)收集中72Vocabularyinstruction&comprehensionWordgenerationprogram(Snow&Lawrence,2009)Directteachcross-subjectvocabulariesto7tgradersacrossdifferentcontextsVocabularyinstruction&compr73ScienceunderstandingSciencelearningisaffectedbylanguageGoodsciencelearningrequirecausalunderstandingDeepKnowingwhyAcollaboratedprojectonearlysciencelearning(Stein,2010)BasedontheoriesofcoherenceScienceunderstandingSciencel74Knowingwhatv.s.knowingwhystatechangesofwaterKnowingwhatisonlyafirststepKnowingwhyistrueunderstandingHowtomakeyoungchildrenknowwhy?MolecularconceptsarecrucialCanweteachmolecularconceptsto4thgraders?KnowingsciencealsoinvolvequantificationsUnitsofmeasurementbecomeimportantKnowingwhatv.s.knowingwhy75ComparisonofU.S.andTaiwan4thGradeMeasurementScoresatPretest–Fall

AllGroupsComparisonofU.S.andTaiwan76

ComparisonofU.S.andTaiwan4thGradeMeasurementScores–Spring

ComparisonofU.S.andTaiwan77

ComparisonofFallandSpring4thGradeTaiwanMeasurementScoreswithTypicalInstruction

ComparisonofFallandSpring78

ComparisonofFallandSpring4thGradeU.S.MeasurementScoreswithTypicalInstruction

ComparisonofFallandSpring79

U.S.andTaiwanComparison:

U.S.InstructioninMeasurementversusTaiwanNaturalTeaching

U.S.andTaiwanComparison:

80

ComparisonofU.S.ExperimentalandControlConditionsfromPretoPostTest

ComparisonofU.S.Experiment81

ComparisonofExperimentalandControlConditions–Fall4thGradeTaiwanMeasurementScores

ComparisonofExperimentalan82EarlyLearninginthePhysicalSciences

Whatistobetaught?Howcanweshowthatcoreconcepts(thosethatarechosenforinstruction)arelinkedtoextremelyimportantissuesandproblems.Globalwarmingandclimatechange?Nuclearreactors:thecontrolofradioactivesubstances?Earthsciences:natureofearthquakes,volcaniceruption?Theminingofcoal,uranium,oil?Sustainableandcleanenergysources?EarlyLearninginthePhysical83感謝

國(guó)科會(huì)99-2420-H-194-003-

宗卓章基金會(huì)

中正大學(xué)前瞻計(jì)畫

感謝

國(guó)科會(huì)99-2420-H-194-003-

宗卓84中正大學(xué)認(rèn)知與學(xué)習(xí)實(shí)驗(yàn)室成員

連啟舜博士

陳明蕾博士

吳裕聖博士

蔡瓊賢

黃秋華

蔡幸錦

郭淑雅

國(guó)立嘉大附小師生

嘉義市博愛國(guó)小

嘉義市興嘉國(guó)小

嘉義縣三興國(guó)小

臺(tái)南市公園國(guó)小

臺(tái)南市裕文國(guó)小

中正大學(xué)認(rèn)知與學(xué)習(xí)實(shí)驗(yàn)室成員

連啟舜博士

陳明蕾博士

吳裕聖85CognitiveInvestigationsofComprehensionProcesses曾玉村YuhtsuenTzengLearning&CognitionLab,NationalChungChengUniversityPresentedatGSME,NCYU,May,04,2011CognitiveInvestigationsofCo86TounderstandishumannatureComprehension/understandingrestsonmemoryComprehensionpromotereasoning,problemsolving,creation…TounderstandistoactivateoldknowledgeTounderstandistoafirststepoflearningEducationshouldaimatunderstandingTounderstandishumannatureC87Whatiscomprehension/understanding?LanguageNon-languageWhatiscomprehension/understa88Whatislanguagecomprehension?ComplexcognitiveprocessesTheconstructionofacoherentmentalrepresentationofthetext(e.g.,Kintsch,1988;Graesser&Clark,1985;Trabasso&vandenBroek,1985)ThereaderidentifiesmeaningfulrelationsbetweenpartsofthetextbetweentextandbackgroundknowledgeWhatislanguagecomprehension89Whatiscomprehension?

TextinputprovidesrawelementsforcomprehensionWhatiscomprehension?

Texti90Whatiscomprehension?

ReadersconnecttextelementstobuildanetworkrepresentationWhatiscomprehension?

Reader91WhatdotheLSmodelssay?

Howdoreadersachievecomprehension?Asreadersproceedthroughthetext,theyattempttomaintaincoherenceforeachnewtextsegment(sentence)(Re)activateinformationfrompriortextorbackgroundknowledgeE.g.,referentialandcausalcoherenceNaming/lexicaldecision/speededrecognitionaswellasreadingtimestudies(e.g.,O’Brien&Myers,1989;McKoon&Ratcliff,1990)InferencesarecrucialtogobeyondtextinputsWhatdotheLSmodelssay?

Ho92FormalizationoftheLSmodelAttentional/WorkingmemorylimitationsComprehensioninvolvesfluctuationofactivationfromTextinputCarryoverReinstatement(fromrepresentationofpriortext)ActivationofbackgroundknowledgeCohortactivation(seebelow)landscapeofactivationsFormalizationoftheLSmodelA93理解的認(rèn)知?dú)v程研究-嘉義大學(xué)課件94TheLSmodel:ConstructingamemoryrepresentationConnectionsareidentifiedbetweentextitemsthatareactivatedsimultaneouslyTheepisodicmemoryrepresentationisupdatedasaresultofeachnewactivationvector,viaanasymptotic(delta)ruleTheLSmodel:Constructingam95LSmodel:DynamicinteractionbetweenactivationandrepresentationCo-activatedelementsformacohortintheepisodicrepresentationCohortactivationprovidesadditionalactivationduringreadingprocessConsistentwithcurrentmodels(e.g.,Construction-Integrationmodel,Kintsch1988;Resonancemodel,O’Brien&Myers)LSmodel:Dynamicinteraction96理解的認(rèn)知?dú)v程研究-嘉義大學(xué)課件97MathematicalFormalizationsofLSmodelThreephasesofLSmodelDetermineinputComputecohortUpdateepisodicmemoryMathematicalFormalizationsof98DetermineInputDetermineInput99ComputeCohortActivation-Equation1ComputeCohortActivation-Equ100TransformConnectionStrengths-Equation2TransformConnectionStrengths101SigmoidTransformationSigmoidTransformation102ComputeExpectancy-

Equation3ComputeExpectancy-

Equation3103UpdateConnectionStrengths-Equation4UpdateConnectionStrengths-E104PartialMemoryRepresentationatReadingCycle3PartialMemoryRepresentation105PartialMemoryRepresentationatReadingCycle13PartialMemoryRepresentation106TheLSprogram:InputTheLSprogram:Input107TheLSprogram:InputTheLSprogram:Input108TheLSprogram:

OutputofactivationpatternsTheLSprogram:

Outputofact109TheLSprogram:

OutputofconnectionpatternsTheLSprogram:

Outputofcon110理解的認(rèn)知?dú)v程研究-嘉義大學(xué)課件111理解的認(rèn)知?dú)v程研究-嘉義大學(xué)課件112Offlinememoryrepresentationoftextcomprehension(Tzeng&Chen,2006)

TextbaseSituationmodelOfflinememoryrepresentation113Offlinememoryrepresentationoftextcomprehension(曾玉村&陳秋芬,修改中)Offlinememoryrepresentation114Howdoescausalstructureoftextaffectonlinereadingprocesses?Aminimalistview:readersmakeminimumamountinferences(McKoon&Ratcliff,1992)AbottomupapproachofinferencemakingsMany“inferences”arepassivememoryactivationprocessesReaderswillmakeinferencesonlyifNoreadilyavailableinformationinWMCoherencebreaksintextsHowdoescausalstructureoft115Howdoescausalstructureoftextaffectonlinereadingprocesses?Aconstructionistview:readerswouldsearchformeaning(Graesser,Singer,&Trabasso,1994)Readerswouldmakemany(butnotall)typesofinferencestosatisfycoherenceassumptionandexplanationassumptionCoherenceassumption:needlocal&globalcoherencetobuildsituationmodelsExplanationassumption:toexplainwhycertaineventoccurHowdoescausalstructureoft116AcontrastingpredictionAminimalistviewpredictsnoinferenceswillbemadeiftextshavelocalcoherenceAconstructionistviewpredictsthatinferenceswillbemadeeveniftextshavelocalcoherencebutdonothaveglobalcoherenceAcontrastingpredictionAmini117LabelHierarchicalVersionSequentialVersion1.Setting2.Initiating13.Goal14.Action15.Outcome16.Reaction17.Initiating28.Goal29.Action210.Action311.Outcome212.Action413.Action514.Outcome3從前有個(gè)男孩叫吉米有一天他看到他的朋友騎著一輛新腳踏車吉米也想要買一輛新腳踏車他去告訴媽媽他媽媽拒絕幫他買腳踏車.吉米很傷心他媽媽告訴他應(yīng)該要自己存一些錢吉米想要賺一些錢.他就在雜貨店找到工作他替雜貨店外送貨品他賺了不少錢.他去一家百貨公司他走到二樓他買了一部新腳踏車.從前有個(gè)男孩叫吉米有一天他看到他的朋友騎著一輛新腳踏車吉米也想要買一輛新腳踏車他去告訴媽媽他媽媽答應(yīng)幫他買腳踏車.吉米很高興他媽媽告訴他應(yīng)該要自己存一些錢吉米想要賺一些錢.他就在雜貨店找到工作他替雜貨店外送貨品他賺了不少錢.他去一家百貨公司他走到二樓他買了一個(gè)新的籃球.LabelHierarchicalVersionSeque118Doreadersmakeonlineglobalinferences?Procedures22collegestudents32storieswereconstructedoutof16scenarios,halfsequential&halfhierarchicalParticipantsengagedinself-pacedreadingwhilerecordingtheireyemovementsusingEyelinkIIDoreadersmakeonlineglobal119Eyetrackingsystem

Eyetrackingsystem120EyetrackingsystemEyetrackingsystem121ResultsofeyemovementdataFocusingontwotypesofsentencesMajorgoals&consequences(i.e.,3th、5th、8th、11th&14thsentences)Spillovereffectsofthem(i.e.,4th、6th、9th&12thsentences)ResultsofeyemovementdataFo122ResultsofeyemovementdataFirstGaze:

thesumoftimeareader’sfirstlookofasentencetothefirsttimes/heexitsthatsentence

nodifferencesonbothversionsexceptforoutcome3longerdurationforsequentialthanhierarchicalstories(889vs.777msrespectively,t=2.56,p=.011)

allotherps>.05

ResultsofeyemovementdataFi123ResultsofeyemovementdataSecondGaze:thesumoftimereaders’fixationsontheirsecondpassofasentencelongerforsequentialversionthanhierarchicalversiononoutcome1(1159vs.818ms,t=2.61,p=.009)Resultsofeyemovementdata124ResultsofeyemovementdataTotalduration:thesumoffirst-gazeandsecond-gazelongerforsequentialversionthanforhierarchicalversiononoutcome1(2258vs.1789ms,t=2.65,p=.009)longerforAction1forsequentialversion(aspilloverregionofGoal1)(2350vs.1968ms,t=2.06,p=.04)ResultsofeyemovementdataTo125ResultsofeyemovementdataPatternoffirstGazeCognitiveloadingforbothversionisthesamesincethewordingsandsyntaxarealmostidenticalResultsofeyemovementdataPa126ResultsofeyemovementdataPatternofsecondgazeLongerfixationtimeforsequentialversionsonOutcome1Readersattempttogobacktointegratesentencestobuildacoherentrepresentationsinceitstartanewepisodeafterthissentence(i.e.,lackofglobalcoherence)ResultsofeyemovementdataPa127ResultsofeyemovementdataPatternoftotalgazeOutcome1marksthefirstdifferencebetweentwoversionsoftextthereforereaderspaydifferentamountoftimethiseffectspillbacktoAction1ThereisacoherencebreakbetweenthefirstandthesecondepisodeLackofglobalcoherenceinsequentialstoriescausesreadersspendmoretimeonthefirstepisodeResultsofeyemovementdataPa128ConclusionsofeyemovementdataCausalstructureaffectonlinereadingtimesFixationtimeincreaseiftextslackofcross-episode(global)coherenceHighcausalcoherenttextsresultinfewerlookingbackbehaviorsSupportaconstructionistviewConclusionsofeyemovementda129KeyIssuesofcomprehensionresearchSuccessfuldecodingFormtext-basedrepresentationsextractpropositionsfromtexts(Kintsch&vanDijk,1978)Formsituationmodel:whatthetextisaboutGobeyondtextAddingbackgroundkknowledge(vanDijk&Kintsch,1983;Kintsch,1988)KeyIssuesofcomprehensionre130InferencesarecrucialforconstructingsituationmodelThreeturtlesfloatingonalog.Severalfishswimmingunderthelog.Inferencesarecrucialforcon131WhatareinferencesInformationnotexplicitintext&readersretrievefrommemory(Kintsch,1998;Graesser,Singer,&Trabasso,1994;Till,Mross&Kintsch,1988)「他把鎖好的門打開」「鑰匙」「氣憤的老婆把杯子往地板上摔」「碎掉」WhatareinferencesInformation132閱讀理解的推論能力Weaver與Kintsch(1991)估計(jì)讀者每讀一則文章之?dāng)⑹?,平均必須進(jìn)行上打的推論才能真正完全理解文意理解需要多少推論以及何種推論有待釐清閱讀理解的推論能力Weaver與Kintsch(199133閱讀理解的推論能力最小量論(minimalist)(McKoon&Ratcliff,1992)若文本具有局部連貫性,讀者則沒有進(jìn)行整體推論連貫之必要,除非相關(guān)訊息可免費(fèi)取得建構(gòu)論(constructivist)(Graesser,Singer,&Trabasso,1993)讀者會(huì)進(jìn)行必要之推論以獲取文章整體意義,滿足讀者尋求意義的需求閱讀理解的推論能力最小量論(minimalist)134閱讀理解的推論能力發(fā)展傳統(tǒng)閱讀發(fā)展停留於整體能力指標(biāo)描述傳統(tǒng)閱讀能力發(fā)展研究往往缺乏理論推論能力發(fā)展研究較具特定性推論能力發(fā)展研究有潛力提早診斷閱讀理解困難閱讀理解的推論能力發(fā)展傳統(tǒng)閱讀發(fā)展停留於整體能力指標(biāo)描述135兒童讀者推論能力的發(fā)展Song和Fisher(2005,2007)發(fā)現(xiàn),幼兒階段已開始發(fā)展代名詞使用與理解的能力。針對(duì)兒童讀者所做的研究,發(fā)現(xiàn)代名詞推論(pronounresolution)的能力可以有效預(yù)測(cè)兒童的閱讀理解表現(xiàn)。(Hickmannetal.,1995;Kail&Hickmann,1992;Oakhill&Yuill,1986)。兒童讀者推論能力的發(fā)展Song和Fisher(2005,136中文兒童推論能力發(fā)展曾玉村、陳明蕾與陳沛嵐(Tzeng,Chen,&Chen,2007)尋找中文讀者特有的理解特性小六學(xué)生閱讀完整指稱詞的文章與極少指稱詞(正常文章盡量刪除指稱詞)的文章的理解並無(wú)差異小四學(xué)生對(duì)於極少指稱詞文章的理解低於完整指稱詞的文章黃秋華、曾玉村與陸偉明(Huang,Tzeng,&Lu,2009)則進(jìn)一步尋求年幼讀者指稱詞推論能力的發(fā)展的線上指標(biāo)指稱詞推論能力介於四到六年級(jí)似乎有明顯變化六年級(jí)對(duì)於零代詞的處理能力似乎已經(jīng)發(fā)展成熟中文兒童推論能力發(fā)展曾玉村、陳明蕾與陳沛嵐(Tzeng,C137研究目的了解兒童因果推論與指稱詞推論在理解歷程中之發(fā)展與可能的交互影響同時(shí)收集線上與離線資料研究目的了解兒童因果推論與指稱詞推論在理解歷程中之發(fā)展與可能138研究假設(shè)與預(yù)測(cè)邏輯短文前後句有無(wú)代名詞或不同強(qiáng)度之因果關(guān)係時(shí),會(huì)影響讀者閱讀理解的心理表徵,進(jìn)而影響回答問(wèn)題之正確率與閱讀時(shí)間預(yù)測(cè)六年級(jí)優(yōu)於四年級(jí)年級(jí)與短文連貫性有交互作用高因果優(yōu)於低因果短文有無(wú)代名詞短文僅對(duì)四年級(jí)有影響研究假設(shè)與預(yù)測(cè)邏輯139實(shí)驗(yàn)方法受試者:臺(tái)南市國(guó)小學(xué)生兩班排除特殊兒童以及上學(xué)期國(guó)語(yǔ)科成績(jī)低於兩個(gè)標(biāo)準(zhǔn)差者一班四年級(jí)學(xué)生23人一班六年級(jí)學(xué)生30人實(shí)驗(yàn)方法受試者:140實(shí)驗(yàn)材料材料的形式為前後兩句的極短文由大學(xué)生事先評(píng)定後再挑選與修改短文共54篇:實(shí)驗(yàn)短文36篇,filler18篇操弄因果連貫性高因果連貫:米奇和高菲穿著外套跳繩,他流很多汗低因果連貫:米奇和高菲剪頭髮動(dòng)不停,他喝到髒水操弄指稱詞關(guān)連性有指稱詞連貫:米奇和高菲穿著外套跳繩,他流很多汗無(wú)指稱詞連貫:米奇和高菲穿著外套跳繩,流很多汗短文主角均為兒童熟悉的八個(gè)卡通人物進(jìn)行變換實(shí)驗(yàn)材料材料的形式為前後兩句的極短文141實(shí)驗(yàn)材料實(shí)例高因果連貫+有指稱詞連貫米奇和高菲穿著外套跳繩,他流很多汗。高因果連貫+無(wú)指稱詞連貫米奇和高菲穿著外套跳繩,流很多汗。低因果連貫+有指稱詞連貫米奇和高菲剪頭髮動(dòng)不停,他喝到髒水。低因果連貫+無(wú)指稱詞連貫米奇和高菲剪頭髮動(dòng)不停,喝到髒水實(shí)驗(yàn)材料實(shí)例高因果連貫+有指稱詞連貫142實(shí)驗(yàn)程序指導(dǎo)語(yǔ)練習(xí)階段,熟悉程序與按鍵的操作正式實(shí)驗(yàn)所需時(shí)間約為40分鐘指導(dǎo)語(yǔ)練習(xí)題正式題(分成兩階段)實(shí)驗(yàn)程序指導(dǎo)語(yǔ)指導(dǎo)語(yǔ)練習(xí)題正式題(分成兩階段)143實(shí)驗(yàn)程序?qū)嶒?yàn)材料用ExperimentBuilder呈現(xiàn)每一短文分成三個(gè)畫面:米奇和高菲穿著外套跳繩,他流很多汗。米奇穿外套作運(yùn)動(dòng)嗎?實(shí)驗(yàn)程序?qū)嶒?yàn)材料用ExperimentBuilder呈現(xiàn)米144實(shí)驗(yàn)程序受試者依自己的速度按空白鍵閱讀短文受試者在第三個(gè)畫面必須回答問(wèn)題正確按「○」的按鍵錯(cuò)誤按「ㄨ」的按鍵紀(jì)錄受試者反應(yīng)正確率及每句閱讀時(shí)間實(shí)驗(yàn)程序受試者依自己的速度按空白鍵閱讀短文145實(shí)驗(yàn)設(shè)計(jì)2(年級(jí))x2(因果連貫)x2(代名詞連貫)年級(jí)為受試者間設(shè)計(jì)因果連貫與代名詞連貫為受試者內(nèi)設(shè)計(jì)實(shí)驗(yàn)設(shè)計(jì)2(年級(jí))x2(因果連貫)x2(代名詞連貫)146結(jié)果:答對(duì)率

高因果有代名詞高因果無(wú)代名詞低因果有代名詞低因果無(wú)代名詞四年級(jí)0.88(.12)0.90(.08)0.82(.11)0.76(.19)六年級(jí)0.94(.08)0.93(.09)0.86(.12)0.94(.07)結(jié)果:答對(duì)率

高因果高因果低因果低因果四年級(jí)0.88(.1147理解的認(rèn)知?dú)v程研究-嘉義大學(xué)課件148結(jié)果:答對(duì)率三因子交互作用不顯著六年級(jí)答對(duì)率高於四年級(jí)(F=22.601,p=.000<.05)六年級(jí)的部分因果關(guān)係與代名詞有交互作用(F=6.857,p=.01<.05)在高因果關(guān)係條件下,代名詞的效果沒有顯著差異(F=.113,p=.738>.05)在低因果關(guān)係的條件下,無(wú)代名詞的答對(duì)率優(yōu)於有代名詞(F=9.983,p=.003<.05)在有代名詞的條件下,高因果優(yōu)於低因果(F=7.934,p=.007<.05)在無(wú)代名詞的條件下,因果關(guān)係沒有顯著差異(F=.476,p=.493>.05)結(jié)果:答對(duì)率三因子交互作用不顯著149結(jié)果:答對(duì)率四年級(jí)的部分因果關(guān)係與代名詞交互作用沒有顯著因果關(guān)係的主要效果達(dá)顯著(F=13.579,p=.000<.05)高因果優(yōu)於低因果代名詞的主要效果沒有顯著(F=.277,p=.600>.05)結(jié)果:答對(duì)率四年級(jí)的部分150結(jié)果:答對(duì)率年級(jí)較大正確率較高四年級(jí)完全受因果連貫強(qiáng)度影響六年級(jí)也受因果連貫性影響閱讀高因果連貫短文,不受代名詞有無(wú)之影響閱讀低因果連貫短文,沒有代名詞之短文較佳結(jié)果:答對(duì)率年級(jí)較大正確率較高151結(jié)果:第二句的閱讀時(shí)間

(單位:ms)

高因果有代名詞高因果無(wú)代名詞低因果有代

溫馨提示

  • 1. 本站所有資源如無(wú)特殊說(shuō)明,都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請(qǐng)下載最新的WinRAR軟件解壓。
  • 2. 本站的文檔不包含任何第三方提供的附件圖紙等,如果需要附件,請(qǐng)聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶所有。
  • 3. 本站RAR壓縮包中若帶圖紙,網(wǎng)頁(yè)內(nèi)容里面會(huì)有圖紙預(yù)覽,若沒有圖紙預(yù)覽就沒有圖紙。
  • 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
  • 5. 人人文庫(kù)網(wǎng)僅提供信息存儲(chǔ)空間,僅對(duì)用戶上傳內(nèi)容的表現(xiàn)方式做保護(hù)處理,對(duì)用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯,并不能對(duì)任何下載內(nèi)容負(fù)責(zé)。
  • 6. 下載文件中如有侵權(quán)或不適當(dāng)內(nèi)容,請(qǐng)與我們聯(lián)系,我們立即糾正。
  • 7. 本站不保證下載資源的準(zhǔn)確性、安全性和完整性, 同時(shí)也不承擔(dān)用戶因使用這些下載資源對(duì)自己和他人造成任何形式的傷害或損失。

評(píng)論

0/150

提交評(píng)論