L8–Puttingitalltogether(measuringpersonality&intelligenceinscience)ReliabilityvsValidity

• deVaus(2002):Avalidmeasurealwayshitsthetargetassessedconstruct);areliablemeasurealwayshitsthesameplaceonthetarget

• Reliabilityisanecessarybutnotsufficientconditionforvalidity• Attheconceptuallevel,avalidmeasureisalwaysreliable• Inpractice,thedistinctionbetweenreliabilityandvalidityisratherfuzzy:

o Cronbach’salphascanbeanindexofinternalconsistencyreliabilityorinternalconsistencyvalidity(relatedtoconstructvalidity)

o Byassessingparallel-formsreliabilitywearealsoassessingconcurrentvalidity(criterion-relatedvalidity)andtoadegreeconvergencevalidity

o Assessmentofinter-raterreliabilityiscloselyrelatedtocontentvalidityo Finally,test-retestreliabilitycanbeusedasanindexofexternalvalidity

Maximisingpsychometricvalues

• Increasesamplesizeo SEMisinverselyproportionatetothesquarerootofsamplesize

• Allowforsufficient(item/participant)meaningfulvariabilityo Q(respondent-centred)vsR(stimulus-centred)analysis

§ Non-discriminatingitems/responses–Discriminability(invariableresponses,itemdifficultyorclarity,ceiling/flooreffect)

§ Minimisationofserialeffects§ Participant-wiseinvariable,random,acquiescentresponses

ú Homogenousandinverselykeyeditemstendtoincreasereliability

ú E.g.“Itendtobesociable”vs“Itendtobesolitary”• Conceptually&empiricallyvaliddimensionality

o Shouldmakepsychologicalsense• Developa“sane”researchdesign/methodology

o Dataanalysiscannot(always)accountfor(fix)designerrors§ E.g.constantmeasurementscannot(really)bestatisticallyanalysed

• Aconstantprocess:o Constructs(re)defined

(Theoretical/Operational(re)assessment)o Measurementscreated&refined

(empiricalstandardization)o Latentstructure(dimensions)expanded/clarified

BeyondCTT

• GeneralisabilityTheory(Cronbachetal.,1972)o Focusonhowwellandunderwhatconditionscantest(observed)scoresbe

generalisedo Attemptstounderstandthevariabilitycomponentsofthescoresandpredict

thescores’valuesundervariouscircumstanceso Itaddssystematicerrorintheobservedscoresandattemptstomapitand

eliminate(control)it• ItemResponseTheory(IRT)

o Mathematicallymapsthecharacteristicsofmeasurementitems(e.g.difficulty,discrimination),andmodelsthemagainstparticipants’ability(latentattribute)onastudyconstruct

o Canbeusedto“accurately”predictresponse-patternsinagivenmeasurement,andthusevaluateandrefinethepsychometricpropertiesofthatinstrument

Measurementstandardisation&normativescores

• Theprocessofpsychometricevaluationcan(should)leadtomeasurement(andbyextensionconstruct)standardisation

o Itisassumedthatstandardisedmeasurementshaveidentifiedthevaluesofthetruescores(±SEM)inagivenpopulationorsetting(seealsoConfidenceIntervals)

o Standardornormativescoresareuniversallyunderstoodunitsintests(e.g.percentiles)thatallowfortherelativeevaluation(profiling)ofaperson’sperformance/attributedonagivenconstruct

ProfilingandPsychography(summingupthewholeperson)

• Criminalpersonalityprofiling“Thecrimesceneis,inessence,apersonalityprojection”(Turco,1990)

o Eliminatingsuspectso Usedwithunusualcrimeso Adaptiveinterrogationtechniqueso Identifyunknownoffenders

• Psychography,psychobiography,&psychohistoryo Identifyandexplainissuesandthemesthroughoutaperson’slifefroma

psychologicalperspective• Freud(1910)laidthefoundationsthroughhis“proscriptiveguidelines”thatallowfor

theassessmentofvalidityandreliabilityelementsinpersonology• ErikEriksongreatlycontributedtopsychographythroughhistriplebookkeeping

approach(Body-Ego-Family/Culture)

v MainReading:- Ch.6,p.79-83

Somepsychographical/profilingarticles(notneededforthecourse):

- Capps,D.(2004).ApsychobiographyofJesus.InE.H.,Harold&W.G.,Rollins(Eds),Psychology&theBible:Anewwaytoreadthescriptures(Vol.4,FromJesustoChrist,pp.59-70).Westport,CT:PraegerPublishers.

- Cribb,C.&Gregory,A.H.(1999).Stereotypesandpersonalitiesofmusicians.JournalofPsychology,133,104-114.

- Grivet-Shillito,M.L.(1999).CarlGustavbeforehebecameJung.JournalofAnalyticalPsychology,44,87-100.

- Kaufman,J.(2001).TheSylviaPlatheffect:Mentalillnessineminentcreativewriters.JournalofCreativeBehavior,35,37-50.

- McDermott,J.(2001).EmilyDickinsonrevisited:Astudyofperiodicityinherwork.AmericanJournalofPsychiatry,158,686-690.

- Preti,A.,DeBiasi,F.,&Miotto,P.(2001).Musicalcreativityandsuicide.PsychologicalReports,89,719-727.

- Runyan,W.M.(1981).WhydidVanGoghcutoffhisear?Theproblemofalternativeexplanationsinpsychobiography.JournalofPersonality&SocialPsychology,40,1070-1077.

- Therivel,W.A.(1999).WhyMozartandnotSalieri.CreativityResearchJournal,12,67-76.

- Young,M.S.,&Pinsky,D.(2006).Narcissismandcelebrity.JournalofResearchinPersonality,40,463-471.

L15–AlbertBandura:SocialCognitiveTheoryIOutline

1. Introduction• Bandura’s(1925-)theoryisoneofthemostwidelyaccepted• Banduraintegratedandclarifiedconceptsfrompreviouspsychologicalforces• Personalityismouldedbyaninteractionofbehaviour,personalfactors

(especiallycognition),andtheenvironment.2. AssumptionsofSocialCognitiveTheory

• Directandvicariouslearning• Peoplenotonlythink,butarecapableofthinkingaboutthinking

(metacognition)• ReciprocalDeterminism

o Banduraput‘cognition’andthe‘person’backintopersonality3. Observationallearning

• Thecoreofobservationallearningismodelling• Factorsthatdeterminewhetherwewilllearnfromamodel

4. Bandura’sfour-stepmodelofobservationallearning• Attend• Remember• Reproduce• Motivationbecauseofreinforcement

5. RelatedResearch:• Aggressionissociallylearnt(Bandura,Ross,&Ross,1963;Paik&Comstock,

1994)• Higherorderformssuchasmoraljudgementsaresociallylearnt(Banduraand

McDonald,1963;Bandura,1991;Cowan.Langer,Heavenrich&Nathanson,1969)

Bandura(1925-)&theadvantageofhindsight

• Integrateandclarifyconceptsfrompreviouspsychologicalforces• Freud’sdescriptionoftheidentificationprocessandtheimportanceof

parentalauthoritywasmadetestable• Banduraarguedthattheeffectsofcontingentrewardsandpunishments

arecognitivelymediated,andisthusmorereadilyacceptableasatheoryofpersonality.

DifferencesbetweenradicalbehaviourismandtheSocialCognitiveapproach

AssumptionsofBandura’sSocialCognitiveTheory

1. Vicariouslearning• Wecananddolearnthingsthroughdirectexperience,butmuchofour

behaviourisshapedthroughtheobservationofothers.Bandura(1986,p.19)statedthat“virtuallyalllearningphenomena,resultingfromdirectexperience,canoccurvicariouslybyobservingother’sbehaviouranditsconsequencesforthem”.

2. Theimportanceofcognition• “Peopleformbeliefsaboutwhattheycando,theyanticipatethelikely

consequencesofprospectiveactions,theysetgoalsforthemselves,andtheyotherwiseplancoursesofactionthatarelikelytoproducedesiredoutcomes”(Bandura,1991,p.248).

3. Reciprocaldeterminism• Herebehaviour(and/orpersonality)ismouldedbythereciprocalinteractionof

personalfactors(especiallycognition),environmentaleventsandourbehaviour.Thepersonisseenasawholeentityforwhompreviouslearninghistory,expectationsofmastery(efficacy),andcurrentinterpretationsoftheworldallinteracttoregulatebehaviour.

Bandura’s(1994)conceptofreciprocaldeterminism

Bandura’s(1994)conceptofreciprocaldeterminism:Appliedexample

Factorsthatdeterminewhetherwewilllearnfromamodel

1. Thecharacteristicsofthemodel:wearemostlikelytomodelhighstatusindividuals,competentindividuals,andpowerfulpeople

2. Thecharacteristicsoftheobserver:peoplewholackstatus,skillorpoweraremostlikelytomodel,i.e.,childrenandnovices

3. Consequencesofbehaviour:thegreaterthevaluethattheobserverplacesonthebehaviour,themorelikelythatthebehaviourwillbelearned

Bandura’sfour-stepmodelofobservationallearning

1. Attendtothemodel2. Rememberwhatisseenandheard3. Reproducethememoryduringimitation4. MotivationbecauseofReinforcementofaccurateperformanceoftheobserved

behaviour5. Todemonstratethatobservationallearninghasoccurred,theimitatedactionmust

consistofanewlyorganisedpatternofbehaviouralresponsesnotpreviouslyintheobserver’srepertoire.

Reciprocaldeterminismistriatic.Personalfunctionarisesfrommutualinteractionbetweenthreeinterlockingfactors:1.Person(P):Thecognitiveandemotion/affectivesystem2.Behaviour(B):Theindividual’sbehaviour3.Environment(E):PhysicalandpersonalenvironmentEachofthesefactorscausallyinfluencestheother,withdifferentinfluencesoccurringindifferentcontext

Thestrengthofthesefactorsneednotbeallequal/allmakeanequalcontributionTheawarenessoftheB-Erelationshipalsoinfluencesthecomplexityofyourpersonality

Mediationmodel

RelatedResearch:Aggressionissociallylearnt

• Bandura,RossandRoss(1961;1963)testedthisideaamongst4-5yearoldswhowatchedamaleandfemaleadult(model)playwithapopularinflatedBobdoll.Therewerefourconditions:live;videotape;cartoon;control.

Numberofaggressiveacts(Bandura,Ross,&Ross,1963)

• PaikandComstock(1994)conductedameta-analysisonover200

studiescompletedduring1957and1990.Theyfoundthat10viewersout100wouldbeaffectedbytheviolencetheyseeonTVorfilm.

Relatedresearch:Higherorderforms(moraljudgement)aresociallylearnt

• BanduraandMcDonald(1969)foundamongst5-to-11-yearchildrenathighandlowlevelsofmoraljudgmentexposedtoadultmodelsshowedasubstantialchangeintheirmoraljudgementlevel.Thischangeinmoraljudgementpersisted(2weekslater)whentheadultmodelwasnolongerpresent.

• Oneofthemostcogentargumentsagainstsociallearningastheprimevariableinthelearningofmoraljudgementsisthefactthatlowerlevelsofmoraljudgementspredominateatearlierages.

v Furtherreadings:

- Bandura,A.(1986).Socialfoundationsofthoughtandaction.EnglewoodCliffs,NJ:Prentice-Hall.

- Bandura,A.(1999).Socialcognitivetheoryofpersonality.InD.Cervone&Y.Shoda(Eds.),Thecoherenceofpersonality:Social-cognitivebasesofconsistency,variability,andorganization(pp.185-241).NewYork:GuilfordPress.

- Caprara,G.V.,&Cervone,D.(2000).Personality:Determinants,dynamicsandpotentials.USA:CambridgeUniversityPress.

L35–AgeDifferencesinIntelligenceOutline

• TheScottishMentalSurveys• StabilityvsChange• Longitudinalvscross-sectionalmethods

o Evidencefromcross-sectionalstudieso EvidencefromlongitudinalStudies

• DomainspecificknowledgeandexpertiseLearningOutcome

• Befamiliarwithsomeofthekeylargedatabasesusedinlifespanresearch• Abilitytoevaluatethestrengthsandweaknessesofcross-sectionalvslongitudinal

research• Knowhowgroupfactors/aspectsofintelligenceprogressacrossthelifespan

StudyingIntelligenceOverTime:TheScottishMentalSurveys(SMS)

• SMS1932:o ALL87,498childrenborn1921takeMorayhousetesto +1000followedupwithSB;followedintoadulthood

• SMS1947o ALL70,805childrenbornin1936testedo +36-daysample(social/demographicinfo)o +6-daysample(SB2,followedfor16years)

• Laterfollow-ups:o AberdeenBirthCohort1921(from1997)o LothianBirthCohort1921(from1999)o AberdeenBirthCohort1936o LothianBirthCohort1936

MorayHouseTest#12Exampleitems:12.Finistofishaswingisto:

a. Featherb. Airc. Birdd. Saile. Herring

36.“Tragu”ischeaperthan“vashol,”and“vashol”isdearerthan“spongop.”Whichisthedearest?42.Underlinethe“different”wordin:eye,pen,nose,chin,ear

• Originalpurpose:selectionintohighschool(1925)o ~45minslongo 71itemso maxscore=76

DoesIntelligenceChangeOverTime?“Stability”vs“Change”

• Stability=Therank-orderofpeoplestaysthesame

o EMPIRICALTEST:Correlationsb/wintelligenceattime1versustime2o THEORETICALMEANING:Intelligenceisstableovertime

• Change=Theabsolutelevelofatraitdifferso EMPIRICALTEST:Changestointelligencemeansattime2comparedtotime1o THEORETICALMEANING:Intelligencechanges/develops/declinesovertime

• NOTE:Testscoresmayremainconsistent(i.e.,rank-orderisthesame)evenaslevelsmaychange(i.e.,scoresforthewholegroupincreaseordecreaseovertime)

Example1:Stablewithchange

• 5peopletakeanIQtest• Theytakeit20yearslater• IQchangesforeveryone

o Mean(T1)=104o Mean(T2)=110

• IQisperfectlystableo r=1.00

Example2:Unstable,nochange

• 5peopletakeanIQtest• Theytakeit4yearslater• MeanIQdoesnotCHANGE

o Mean(T1)=104o Mean(T2)=104

• IQisNOTstableo r=0.06

GodfreyThomson DeveloperoftheMorayHouseTests

IntelligenceTestScoresareStable

Butwhataboutveryyoungchildren?

v Mackintosh(2012).IQandHumanIntelligence(2ndEd).NewYork,OxfordUniversity

Press.

LifespanResearch:Methods

Cross-SectionalResearch:TheWechsler-BellevueScales(1939)

• DeclineonbothverbalandperformanceIQ• HUGEdifferencesforperformanceIQ(>25IQpointslostbyage47!!)

v Mackintosh(2012).IQandHumanIntelligence(2ndEd).NewYork,OxfordUniversity

Press.

Cross-SectionalResearch:WAIS-R(1981)

• DifferencesonWAIS-RPerformanceo 10IQpointslostbyage45o 30IQpointslostbyage72

• DifferencesonWAIS-RVerbalo Nolossbyage45o ~7-pointlossbyage72

v Mackintosh(2012).IQandHumanIntelligence(2ndEd).NewYork,OxfordUniversity

Press.Cross-SectionalResearch:AgeDifferencesGf/Gctheoryabilities

v Horn,J.,Cattell,R.B.(1967).Agedifferencesinfluidandcrystallizedintelligence.ActaPsychologica,26,107-109.

v Park,D.C.,Lautenschlager,G.,Hedden,T.,Davidson,N.S.,Smith,A.D.,&Smith,P.K.(2002).Modelsofvisuospatialandverbalmemoryacrosstheadultlifespan.PsychologyandAging,17,299-320.

Whydoweseesuchlargedecreases??

• IQdropsfrom100to73fromage16-47[Wechsler-Bellevue]

o Anaverage16-year-olddeclinessomuchthattheyarementallyretardedby47?

• Whathasgonewrong?• ΔAGEconfoundedwithΔYEARBORN• Differenceineducation:

o <20%finishHS1970;~60%finishHS2010

• Flynneffect:3IQpointsperdecadeo 45yoscore74becausetheyare45

VSo Thegroupof45-year-oldswould

alwayshavescored74(evenwhentheywere16)?

COHORTEFFECTS

Cross-SectionalvsLongitudinalResearchCross-Sectional

• CohortDifferenceso Educationo Technology/complexityo Particularevents(e.g.,1930sdepression;WW1

conscriptionetc.)o Flynneffecto Test/testingfamiliarityo Ageofretiremento %ofwomeninworkforceo #ofchildrenperfamilyo Maternalage

CompareLongitudinalvsCross-Sectional

• Cross-SectionalResearchOver-EstimatesCognitiveDecline• BUTdifferentsizesfordifferentcohorts

SOLONGITUDINALMUSTBEABETTERRESEARCHDESIGN?

• SEATTLESTUDY:N=50020-80yoin1956• Followedevery7years;+newcohortevery7years

v Schaie,K.W.(2005).Developmentalinfluencesonadultintelligence:TheSeattle

LongitudinalStudy.Oxford,OxfordUniversityPress.Whydon’twejustlookatlongitudinal?Longitudinal

• Updatedtests/measures• Retesteffects• SelectiveAttrition• Takesalifetimetodo…

• Retest–likelyinflatesscoresathigherages(alsoboredom/motivation)

o under-estimatescognitivedecline• Selectiveattrition–intelligencerelatestohealth/death,incarceration,geographic

mobility,etc.o under-estimatescognitivedecline

Shouldwebeworriedabouttheagingworkforce?Worldpower-brokersaregenerallyover60

• Roleofdomainspecificknowledge• Domainspecificknowledgeincreasesacrossthelifetime• Roleofexpertise• Deliberatepracticeof10,000hoursormore• Strategiesforsuccessfulperformanceshifttodomain

specificknowledgeandexpertise

v Ackerman,P.L.(1996).Atheoryofadultintellectualdevelopment:Process,Personality,Interests,andKnowledge.Intelligence,22,227-257.

v FurtherResources:- Deary,I.J.(2012).Intelligence.AnnualReviewofPsychology,63,453–482.

[Agingsectionfromp.470]- Mackintosh(2012).IQandHumanIntelligence(2ndEd).NewYork,Oxford

UniversityPress.[Ch8;TheStabilityofIQandtheRiseandFallofIntelligence]

SampleQuestions:Q1.

• Namethetworesearchdesignsthatcanbeusedtostudyagedifferencesinintelligence.

• Brieflydescribeeachofthesedesigns.• Brieflystatetheresearchfindingsonagedifferencesinintelligenceforeach

ofthesedesigns.• Criticallyevaluatethestrengthsandweaknessesandeachofthesedesigns.

Q2.

• Definestabilityandchangeastheyapplytoindividualdifferencesinintelligence.

• Whatevidenceistherethatintelligencechanges?• Whatevidenceistherethatintelligenceisstableovertime?

Q3.

• Brieflydescribecross-sectionalandlongitudinalmethodsforexaminingagedifferences.

• Whataretheweaknessesofcross-sectionalresearch?• Whataretheweaknessesoflongitudinalresearch?• Howdothemajorgroupfactorsofintelligencechangeacrossthelifespan?