普通心理學(xué)9：智力與心理測驗

Chapter9

IntelligenceandPsychologicalTesting智力和心理測驗心理學(xué)(家)的特點?EmpiricismTheoreticaldiversitySocio_historicalcontextMultifactorialcausation

CulturalheritageHeredityandenvironmentSubjectivityandexperience本章的主要內(nèi)容智力測驗測驗的一般概念智力測驗的進(jìn)化智力測驗的基本問題極端智力智力遲緩智力超常智力發(fā)展遺傳的作用環(huán)境的作用文化的作用智力測量與研究的新方向問題智力測驗有什么用?如何判斷一個智力測驗?智力遲緩是否先天決定?天才是“白癡”？IQ能決定什么？。。。。TypesofTestMentalability(心理能力)Personality（人格）KeyConceptsinPsychologicalTestingMentalabilitytestsIntelligencetests(智力測驗)一般心理能力:信息加工的速度/深度與準(zhǔn)確性;Aptitudetests（才能測驗）某一類心理潛能（specific）:verbalreasoning,numericalreasoning,abstractreasoning,perceptualspeedandaccuracy,mechanical,spelling,languageusage…Achievementtests（成就測驗）已學(xué)習(xí)過的知識的掌握程度KeyConceptsinPsychologicalTestingPersonalitytestsPersonalitytraits(人格特質(zhì))Personalityscale(人格量表)insteadoftestKeyConceptsinPsychologicalTestingPsychologicaltestisastandardized（標(biāo)準(zhǔn)化）

measureofasampleofaperson’sbehavior.Uniformadministration；Uniformscoring；Norm(常模）_M,SDStandardizationgroup（標(biāo)準(zhǔn)化群體）KeyConceptsinPsychologicalTestingReliability（信度）Correlationcoefficient（相關(guān)系數(shù)）Test-Retestreliability(再測信度)PracticeeffectAlternativeorparallelformsreliability(復(fù)本信度)（A、B卷）Split-halfreliability(分半信度)/內(nèi)在一致性信度（internalconsistencyreliability）Above0.70KeyConceptsinPsychologicalTestingFigure9.3Correlationandreliability.AsexplainedinChapter2,apositivecorrelationmeansthattwovariablescovaryinthesamedirection;anegativecorrelationmeansthattwovariablescovaryintheoppositedirection.Thecloserthecorrelationcoefficientgetstoeither–1.00or+1.00,thestrongertherelationship.Ataminimum,reliabilityestimatesforpsychologicaltestsmustbemoderatelyhighpositivecorrelations.Mostreliabilitycoefficientsfallbetween70and.95.Figure9.2Test-retestreliability（重測信度）.

Subjects’scoresonthefirstadministrationofanassertivenesstestarerepresentedontheleft,andtheirscoresonasecondadministrationofthesametestafewweekslaterareshownontheright.Ifsubjectsobtainsimilarscoresonbothadministrations,asintheleftgraph,thetestmeasuresassertivenessconsistentlyandhashightest-retestreliability.Iftheygetverydifferentscoresonthesecondadministration,asintherightgraph,thetesthaslowreliability.Validity（效度）TheabilityofatesttomeasurewhatitisdesignedtomeasureContentvalidity（內(nèi)容效度）Achievementtest/educationaltestCriterion-relatedvalidity（效標(biāo)關(guān)聯(lián)效度）預(yù)測能力（aptitudetest）Constructvalidity（結(jié)構(gòu)效度）Abstractpersonalqualities,suchascreativity,intelligence,extraversionHypotheticalconstruct(假設(shè)的結(jié)構(gòu))KeyConceptsinPsychologicalTestingFigure9.4Criterion-relatedvalidity（效標(biāo)關(guān)聯(lián)效度）.Toevaluatethecriterion-relatedvalidityofapilotaptitudetest,apsychologistwouldcorrelatesubjects’testscoreswithacriterionmeasureoftheiraptitude,suchasratingsoftheirperformanceinapilottrainingprogram.Thevalidityofthetestissupportedifasubstantialcorrelationisfoundbetweenthetwomeasures.Iflittleornorelationshipexistsbetweenthetwosetsofscores,thedatadonotprovidesupportforthevalidityofthetest.Figure9.5Constructvalidity（結(jié)構(gòu)效度）.

Psychologistsevaluateascale’sconstructvaliditybystudyinghowscoresonthescalecorrelatewithavarietyofvariables.Forexample,someoftheevidenceontheconstructvalidityoftheExpressionScalefromthePsychologicalScreeningInventoryissummarizedhere.Thisscaleissupposedtomeasurethepersonalitytraitofextraversion.Asyoucanseeontheleftsideofthisnetworkofcorrelations,thescalecorrelatesnegativelywithmeasuresofsocialintroversion,socialdiscomfort,andneuroticism,justasonewouldexpectifthescaleisreallytappingextraversion.Ontheright,youcanseethatthescaleiscorrelatedpositivelywithmeasuresofsociabilityandself-acceptanceandanotherindexofextraversion,asonewouldanticipate.Atthebottom,youcanseethatthescaledoesnotcorrelatewithseveraltraitsthatshouldbeunrelatedtoextraversion.Thus,thenetworkofcorrelationsdepictedheresupportstheideathattheExpressionScalemeasuresextraversion.信度VS效度如果有人用你的成人身高來評價智力,你感覺如何,這一測量是可信的,但它有效嗎?SirFrancisGalton(1869)高爾頓FamilytreestudiesSensorytestReactiontime,sensitivitytohigh-pitchedsounds,colorperception,etc..Correlatedwithschoolorprofessionalperformance;Correlation（相關(guān)）,percentiletestscore（比例分）..

智力測量的歷史高爾頓有關(guān)智力測量的主要思想智力可定量個體差異呈鐘型曲線可由客觀測驗測得兩套測驗成績之間的相關(guān)程度可通過統(tǒng)計分析確定智力由遺傳決定；AlfredBinet(比納)andTheodoreSimon(1905)西蒙測驗內(nèi)容:判斷和推理等抽象思維(非機(jī)械記憶力);MentalageHowtocollectitems?60%passinthegroup.比奈-西蒙測驗的特點測驗分?jǐn)?shù)是當(dāng)前操作的評估,并非對天生智力的評估;測驗是為了更好地分班和教育,并非為了污蔑他們;強(qiáng)調(diào)訓(xùn)練可以影響智力發(fā)展;依據(jù)經(jīng)驗編制測驗--看是否有效?--并非依據(jù)某一智力理論;LewisTerman(特曼，1916,37,60,73,86)Stanford-BinetIntelligenceScale測驗內(nèi)容：推理能力；IntelligenceQuotient(WilliamStern,1914)WIQ)IQ=MA/CAx100測驗修訂的理由用于不同的年齡及人群更新不適應(yīng)社會發(fā)展的條目更新常模DavidWechsler(韋克斯勒，1939,49,55,67,91)WechslerAdultIntelligenceScaleDeviationIQ(離差智商)

IQ=100+15(X-M)/SDM，SD為常模特點，X為原始分。Figure9.7Thenormaldistribution.

Manycharacteristicsaredistributedinapatternrepresentedbythisbell-shapedcurve.Thehorizontalaxisshowshowfaraboveorbelowthemeanascoreis(measuredinplusorminusstandarddeviations-標(biāo)準(zhǔn)差).Theverticalaxisisusedtographthenumberofcasesobtainingeachscore.Inanormaldistribution,thecasesaredistributedinafixedpattern.Forinstance,68.26%ofthecasesfallbetween+1and–1standarddeviation.ModernIQscoresindicatewhereaperson’smeasuredintelligencefallsinthenormaldistribution.OnmostIQtests,themeanissetatanIQof100andthestandarddeviationat15.Thus,anIQof130meansthatapersonscored2standarddeviationsabovethemean.AnydeviationIQscorecanbeconvertedintoapercentilescore,whichindicatesthepercentageofcasesobtainingalowerscore.ThementalclassificationsatthebottomofthefigurearedescriptivelabelsthatroughlycorrespondtorangesofIQscores.Dointelligencetestsmeasurepotentialorknowledge?IntellectualpotentialNotaffectedbyyourknowledgebackgroundApplyrelativelycommonknowledgeAblendofpotentialandknowledge關(guān)于智力測驗的主要問題Predictschoolperformance.50s-.60swith學(xué)業(yè)成績可解釋成績總變異量的.25-.36;.60s-.80swith學(xué)歷IQ說明1/4的社會地位變異,1/6的經(jīng)濟(jì)收入變異;父母社會經(jīng)濟(jì)地位對社會地位變異的解釋量為1/3,對經(jīng)濟(jì)收入的解釋量為1/5.預(yù)測力不如父母社會經(jīng)濟(jì)地位?IQ測驗可以預(yù)測什么？MeasurementAbstractreasoningandverbalfluency(!!)Socialcompetence,practicalproblemsolving,creativity,mechanicalingenuity,artistictalent(??)WorkbyRobertSternberg(1981)R.J.Sternberg(IBMProfessorofPsychology,DepartmentofPsychology,YaleUniversity)原因分析Figure9.8Laypersons’conceptionsofintelligence.

RobertSternbergandhiscolleagues(1981)askedparticipantstolistexamplesofbehaviorscharacteristicofintelligence.Theexamplestendedtosortintothreegroupsthatrepresentthethreetypesofintelligencerecognizedbytheaverageperson:verbalintelligence,practicalintelligence,andsocialintelligence.Jobcomplexitymayplayarolebetweenmentalabilityandjobperformance（Schmidt&Hunter,2004)highJobcomplexityWesterncultureswithEuropeanrootsRapidinformationprocessingDecisiverespondingThenotionthatabilitycanbequantifiedNonWesternculturesPeoplewillbegoodatdoingthingsthatareimportanttothem.(theingredientsofintelligenceareculturalspecific)JapanChina,IndiaTheme3:Psychologyevolvesinasocio-historicalcontext.AreIQtestswidelyusedinothercultures?MentalRetardation(心理遲鈍)DiagnosisbasedonIQandadaptivetesting（AmericanAssociationonMentalRetardation-AAMR)—2-3%IQ：2ormoreSDbelowmeanAdaptiveskilldeficits（適應(yīng)技巧缺陷）Originationbeforeage184levels:mild(輕度),moderate（中度）,severe（重度）,profound（深度）Mildmostcommonbyfar（pp.355）--85%2.1ExtremesofIntelligence:MentalRetardation心理遲緩的診斷:在適應(yīng)性技能等方面表現(xiàn)出缺陷Figure9.9Theprevalenceandseverityofmentalretardation.

Theoverallprevalenceofmentalretardationisroughly1to3%ofthegeneralpopulation.Thevastmajority(85%)oftheretardedpopulationismildlyretarded.Onlyabout15%oftheretardedpopulationfallsintothesubcategoriesofmoderate,severe,orprofoundretardation.Causes:BiologicalDownsyndrome(唐氏綜合癥)：21th.染色體三體型；Phenylketonuria(苯丙酮尿癥/肝毒性癡呆癥)：aminoacidphenylalanine（苯基丙氨酸）---（某種代謝酶的缺陷）--tyrosine(酪氨酸）--（合成各種兒茶酚氨）；Hydrocephaly(水腦)Environmental（Figure9.10）IQ<50IQ60-69:ExtremesofIntelligence:MentalRetardation1:unskilledlaborer;2:semiskilledlaborer;3:skilledmanuallaborer;4:artisantrade;5:professionalnon-manual;Identificationissues–idealsvs.practiceGiftednessvs.highintelligenceIQ2SDabovemeanstandardCreativity,leadership,specialtalent?Stereotypes–weak,sociallyinept,emotionallytroubledLewisTerman(1925)–largelycontradictedstereotypesEllenWinner(1997)–moderately(IQ130-150)foundlygifted(IQ>180)Giftednessandhighachievement–beyondIQ2.2ExtremesofIntelligence:Giftedness(天才)ExceptionalintelligenceExceptionalcommitment(MOTIVATION)ExceptionalcreativityeminenceRenzulli(1986)creativityGenerationofideasthatareOriginalNovelUsefulCreativeachievementsresultfromlongtimehardwork;Dependondivergentthinking;

CorrelatesofcreativityHighlycreativepeople:moreautonomous,introverted,opentonewexperiences;self-confident,andselfaccepting;IQof120Connectionbetweenexceptionalcreativityandmentalillness;Figure9.27Estimatedprevalenceofpsychologicaldisordersamongpeoplewhoachievedcreativeeminence.

Ludwig(1995)studiedbiographiesof1004peoplewhohadclearlyachievedeminenceinoneof18fieldsandtriedtodeterminewhethereachpersonsufferedfromanyspecificmentaldisordersinhisorherlifetime.Thedatasummarizedhereshowtheprevalenceratesfordepressionandforamentaldisorderofanykindforfourfieldswherecreativityisoftenthekeytoachievingeminence.Asyoucansee,theestimatedprevalenceofmentalillnesswasextremelyelevatedamongeminentwriters,artists,andcomposers(butnotnaturalscientists)incomparisontothegeneralpopulation,withdepressionaccountingformuchofthiselevation.Hereditystudies(遺傳研究)Familyandtwinstudies(家族和雙子研究)Heritabilityestimates:aheritabilityratioisanestimateoftheproportionoftraitvariabilityinapopulationthatisdeterminedbyvariationsingeneticinheritance.(遺傳估計)Intelligence:HeredityorEnvironment?LaunchVideoFigure9.12StudiesofIQsimilarity.

ThegraphshowsthemeancorrelationsofIQscoresforpeopleofvarioustypesofrelationships,asobtainedinstudiesofIQsimilarity.Highercorrelationsindicategreatersimilarity.TheresultsshowthatgreatergeneticsimilarityisassociatedwithgreatersimilarityinIQ,suggestingthatintelligenceispartlyinherited(compare,forexample,thecorrelationsforidenticalandfraternaltwins).However,theresultsalsoshowthatlivingtogetherisassociatedwithgreaterIQsimilarity,suggestingthatintelligenceispartlygovernedbyenvironment(compare,forexample,thescoresofsiblingsrearedtogetherandrearedapart).(DatafromMcGueetal.,1993)Figure9.13Theconceptofheritability.

Aheritabilityratioisanestimateoftheportionofvariationinatraitdeterminedbyheredity—withtheremainderpresumablydeterminedbyenvironment—asthesepiechartsillustrate.Typicalheritabilityestimatesforintelligencerangebetweenahighof70%andalowof50%,althoughsomeestimateshavefallenoutsidethisrange.Bearinmindthatheritabilityratiosareestimatesandhavecertainlimitationsthatarediscussedinthetext.Whatistheexactlymeaningofheritabilityestimates?智力變異由遺傳變異決定的比例；例如，性別的可遺傳性估計為99%身高的可遺傳估計為90%；快樂的可遺傳估計為40-50%；心理特質(zhì)的可遺傳性估計：20-60%EnvironmentstudiesAdoptionstudies(收養(yǎng)研究)foster&biologicalparents;SiblingsUnrelatedchildren(comparedgroup)Intelligence:HeredityorEnvironment?LaunchVideo遺傳、環(huán)境和IQ的關(guān)系此圖表明遺傳和環(huán)境對IQ得分的貢獻(xiàn)。父子間的IQ相似（遺傳的影響），但是父親一兒子的IQ都與社會地位相關(guān)（環(huán)境的影響）環(huán)境對IQ的影響被撫養(yǎng)的兒童的IQ得分黑人和混血兒由中等階層的白人家庭撫養(yǎng)，他們的IQ值高于正常。因為孩子們來自于IQ值低于平均值的組中，這一數(shù)據(jù)顯示了環(huán)境對IQ的影響EnvironmentCumulativedeprivationhypothesis(累積的剝奪假設(shè)）Home-environmentandschoolingeffectsOrphanages,poverty…TheFlynneffect(JamesFlynn):GenerationalincreasesinmeasuredIQ;InteractionTheconceptofthereactionrange（遺傳決定的可塑范圍）Intelligence:HeredityorEnvironment?LaunchVideoFigure9.15Reactionrange.

Theconceptofreactionrangepositsthathereditysetslimitsonone’sintellectualpotential(representedbythehorizontalbars),whilethequalityofone’senvironmentinfluenceswhereonescoreswithinthisrange(representedbythedotsonthebars).Peopleraisedinenrichedenvironmentsshouldscorenearthetopoftheirreactionrange,whereaspeopleraisedinpoor-qualityenvironmentsshouldscorenearthebottomoftheirrange.GeneticlimitsonIQcanbeinferredonlyindirectly,sotheoristsaren’tsurewhetherreactionrangesarenarrow(likeTed’s)orwide(likeChris’s).Theconceptofreactionrangecanexplainhowtwopeoplewithsimilargeneticpotentialcanbequitedifferentinintelligence(compareTomandJack)andhowtwopeoplerearedinenvironmentsofsimilarqualitycanscorequitedifferently(compareAliceandJack).HeritabilityasanExplanationAurthurJensen(1969)Heritalityestimates80%HerrnsteinandMurray(1994)–TheBellCurveRacialdifferencesinaverageintelligence;J.PhillipeRushton&ArthurJensen(2005);LindaGottfredson(2005)GeneticsfactorsaccountforabouthalfofthegapbetweenracesinaverageIQ;CulturalDifferencesinIQFigure9.16Geneticsandbetween-groupdifferencesonatrait.

Kamin’sanalogy(seetext)showshowbetween-groupdifferencesonatrait(theheightofcornplants)couldbeduetoenvironment,evenifthetraitislargelyinherited.Thesamereasoningpresumablyappliestoethnicgroupdifferencesinthetraitofhumanintelligence.EnvironmentasanExplanation:LeonKamin’scornfieldanalogyCulturalbiasonIQtest:Featuredstudy

(Steele&Aronson,1995)X:stereotypevulnerabilityTotestproblem-solvingstrategies;Totestgeneralverbalability;Y:GREverbal.4.4文化與IQ測驗的有效性刻板印象威脅ExploringBiologicalindexesandcorrelatesofIntelligenceAttempttofocusonculturefreemeasuresofintelligenceReactiontime(RT)asmeasureofmentalspeed

(ArthurJensen,HansEysenck)Modestcorrelation(0.2-0.3)withIQtest;Inspectiontime:timetomakeaperceptualdiscriminationaccordingtoacriterion(r=0.3--0.4)withIQNewDirectionsintheStudyofIntelligenceExposuredurationProbabilityofcorrectresponse.8514msTask:whetherthelongerlineisontheleftorright?BiologicalindexesofintelligenceBrainsizebyheadsize,r=.15withIQBrainsize(byMRI),r=.35withIQIQandlongevity1932L-IQ組（85以下）；H-IQ組（115以上）1997，L-IQ組存活率數(shù)<H-存活率聰明人更長籌?原因:遺傳?生活條件?健康習(xí)慣?CognitiveprocessesinintelligentbehaviorPrimarilyfocused:Howmanyamountofintelligenceofpeoplehave?---measureapproachUpdatedfocused:Howdopeopleusetheirintelligence?—cognitiveprocessingapproach!NewDirectionsintheStudyofIntelligenceSternberg’striarchictheory