Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Learning repetition rules and non-adjacent dependencies from human actions in nine-month-old infants
(USC Thesis Other)
Learning repetition rules and non-adjacent dependencies from human actions in nine-month-old infants
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Learning Repetition Rules and Non-adjacent Dependencies from Human
Actions in Nine-month-old Infants
by
Shiyang Lu
A Thesis Presented to the
FACULTY OF THE USC DORNSIFE COLLEGE OF LETTERS, ARTS, AND SCIENCES
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF ARTS
PSYCHOLOGY
August 2020
Copyright 2020 Shiyang Lu
ii
TABLE OF CONTENTS
L F
A......
I.1
E 1...5
M...6
R 9
D1 0
E 2.1 1
M.1 2
R..1 5
D1 5
G D...1 6
R.2 1
iii
LIST OF FIGURES
1 Images of the midpoint of human action clips………………………………….7
2 Examples of habituation materials in Experiment 1…………………………….7
3 Experiment 1 Results…………………………………………………………..10
4 Examples of habituation and test materials in Experiment 2………………….14
5 Experiment 2 Results…………………………………………………………..16
Abstract
Ourabilitytodetectandgeneralizepatternsisimportantfordifferentkindsoflearning,and
crucialforouradaptations. Indeed,infantsasyoungassevenmonthssucceededinlearning
simplerepetitionruleslikeABAandABBfromspeech. However,theevidenceforinfants’ability
tolearnsuchpatternsfromthevisualdomainsismixed. Whileinfantssucceededinlearning
thesepatternsfromsimultaneouslypresentedstimuli,theywerelessabletolearnfrom
sequentiallypresentedstimuli. Thisposedthequestionastowhyinfants’sequentialrulelearning
ischallengedinthevisualdomainbutnotinspeech. Recentstudiesonadults’non-adjacent
dependency(NAD)learning—whichcanbeconsideredamorecomplexrulethansimple
repetitions—suggestthattheirNADlearninginthevisualdomainisaidedbydynamicvisual
motion. Itispossible,then,thatinfants’rulelearninginthevisualdomainsisfacilitatedby
dynamicmotionaswell. Totestthispossibility,Itested9-month-oldinfantsontheirabilityto
learnrulesfromvisualhumanactions(adomainthatinvolvesdynamicvisualmotions).
Experiment1foundthattheseinfantssuccessfullylearnedtherulesfromvisualhumanactions,
supportingthepreviousclaim. Toseeifdynamicvisualmotionshaveabroaderimpactoninfants’
learningofdifferentpatterns,Ithentested9-month-oldinfants’abilitytolearnNADsfromvisual
humanactions. ResultsofExperiment2showedthatinfantslearnedNADsfromvisualhuman
actionswhentheywerefirstexposedtohumanactionsintheABApattern. Thisisthefirst
evidenceforinfants’NADlearninginavisualdomain,andtheageofNADlearningisyounger
thanwhattheliteraturehadsuggested. PossibleexplanationsforthesuccessinNADlearning
withhumanactionswerediscussed.
Keywords: Humanactions,Non-adjacentdependencies,Rulelearning
LearningRepetitionRulesandNon-adjacentDependenciesfromHumanActionsin
9-month-oldInfants
Ourworldisfilledwithspatialandtemporalregularitiesandpatterns,andourabilityto
detectandgeneralizefromsuchpatternsisessentialforvariouskindsoflearning. Thisabilityis
availabletoinfantsataveryyoungage: seven-month-oldinfantscanlearnsimplerepetitionrules
likeABA(e.g.,ga-ti-ga)andABB(e.g.,ga-ti-ti)fromspeech(Marcusetal.,1999). Inthevisual
domain,Saffranetal.(2007)foundthatinfantsofthesameagelearnedruleslikeABAfrom
picturesoffamiliarcategoriessuchascatsanddogs. Similarresultswerefoundwith
seven-month-oldinfants’learningofABAandABBrulesfrompicturesofhumanfaces(Bulf
etal.,2015). Thesestudiesseemtosuggestthatinfantsareabletodetectandgeneralizerulesin
thevisualdomainsaswellasintheauditorydomains.
Nevertheless,Johnsonetal.(2009)foundthatboth8-and11-month-oldinfantsfailedto
learntheABArulefromshapesbutsucceededwithruleslikeABBandAAB.Theythusarguedthat
therulelearningwithshapesisnotrobustinyounginfants,andnon-adjacentrepetitions(i.e.,
ABA)arepotentiallyhardertodetectthanadjacentones(i.e.,ABBandAAB).Howcanwe
reconcilethefindingsofthesethreestudiesinvolvingvisualstimuli? Onepossibilityisthat
familiarobjectcategories(e.g.,animalsandfaces)mightsupportruleencodingand
generalization. However,itisworthnotingthatbothSaffranetal.(2007)andBulfetal.(2015)
employedasequentialandsimultaneouspresentation,suchthattherewasalwaysatimeperiod
whenallthreeimageswereonthescreeninaleft-to-rightorder,eventhoughimageswithina
tripletcameuponthescreenatdifferenttimepoints. Onthecontrary,Johnsonetal.(2009)used
astrictlysequentialpresentation(i.e.,therewasneveratimewhenmorethanoneshapewas
presentonthescreen). Thesimultaneousaspectofthepresentationmightmakethetaskeasierin
Saffranetal.(2007)’sandBulfetal.(2015)’sstudiesbyalleviatingthememoryburden.
Therefore,thereiscurrentlynorobustevidenceforinfants’sequentialrulelearninginthevisual
domain. Why,though,areinfantschallengedbythesequentialrulelearning(andespecially
non-adjacentrepetitionrulessuchasABA)inthevisualdomainbutnotinspeech?
Researchonadults’non-adjacentdependency(NAD)learningprovidesapossible
explanation: dynamicvisualmotionsmightfacilitateourabilitytolearnpatternsfromthevisual
domains. NADsaretheco-occurrencesoftwoitemsnotimmediatelyadjacentinasequence. For
example,isand-ingcanbeconsideredanon-adjacentdependency,becausetheyalwaysco-occur
withatleastoneiteminbetweenthemin,say,thepresentprogressive: thecat ischasingthe
mouseorthecat isfuriouslychasingthemouse. AdultscanreadilylearnNADsfromanartificial
speechstream(Wangetal.,2019). However,theabilitytotracknon-adjacentdependenciesisnot
limitedtothelinguisticdomain. AdultscansuccessfullylearnNADsfromhumanaction
sequencesaswell(Endress&Wood,2011). Furthermore,LiandMintz(2015)foundthatadults
learnedNADsfromhumanactions,humanpostures,andobjectmovements,butfailedtodoso
forstaticimagesofobjectpostures. Itwasthushypothesizedthatdynamicvisualmotionsaided
NADlearninginthevisualdomain. Adults’learningofNADsfromotherauditorydomainsis
alsolessrobustthantheirNADlearninginspeechandinthedomainswithdynamicvisual
motions: adultscouldonlylearnNADsfromtoneswhentonesconstructingthedependencywere
perceptuallysimilartoeachother(Creeletal.,2004;Gebhartetal.,2009). SuccessfulNAD
learninginspeechanddynamicvisualmotionsseemstosuggestthatbotharegovernedbysimilar
principles. Inaway,wecanthinkofnon-adjacentrepetitionruleslikeABAasasimpleNAD
pattern,wherethefirstthethirditemsnotonlyco-occurbutalsoareidenticaltoeachother. Then,
itwillbereasonabletopositthat1)dynamicvisualmotionsalsofacilitaterulelearninginthe
visualdomain,and2)therulelearningindynamicvisualmotionsandspeecharegovernedby
similarprinciplesaswell.
Therefore,wedecidedtoturntohumanactions-adynamicvisualdomainthatissimilar
tospeechinthefollowingways: First,bothspeechandhumanactionsarepresentedintemporal
sequences,withnosalientpausesbetweenwordsoractions. Yet,theybothcontainothercues
signalingwordoractionboundaries(e.g.,Endress&Hauser,2010;Newtson&Engquist,1976).
Second,languageiscomprisedofhierarchicalstructuressuchthatsoundsgetorganizedinto
words,andwordsgetorganizedintowordphrases,andwordphrasesgetorganizedinto
sentences. Justlikelanguage,actionsequencescanbeperceivedashierarchicallyorganized smallersequenceswithasub-goaldirectingeachsmallersequence(Cooper&Shallice,2006).
Totestwhetherinfants’sequentialrulelearningabilityinthevisualdomaincanbeguided
bydynamicvisualmotions,Iusedaninfanthabituationparadigmtoexamine9-month-olds’
learningofABAandABBrulesfromvisualhumanactions(Experiment1). Recallthateven
11-month-oldinfantsfailedtofullylearntherepetitionrulesfromshapes(theyfailedtolearn
ABArulesin Johnsonetal.,2009). Iftheysucceed(whichtheydidinExperiment1),then
dynamicvisualmotionsindeedfacilitatesequentialrulelearninginthevisualdomains.
Furthermore,itwillbethenreasonabletopositthatdynamicvisualmotionscanalsofacilitatethe
learningofmorecomplexpatternsininfants. Sincetheadultliteraturesuggeststhatdynamic
visualmotionsfacilitateadults’NADlearning,ournextgoalistoexaminetheroleofdynamic
visualmotionsininfants’learningofNADs(Experiment2).
Earlyevidencesuggeststhat15-month-oldsbutnot12-month-oldscantrackNADsfrom
anartificialspeechstream(Gómez&Maye,2005). Thiscontrastswithinfants’earlyabilityto
trackadjacentdependencies(i.e.,transitionalprobabilities-theprobabilitythatoneitemoccurs
rightafteraspecificitem)inanartificialspeechstreamateightmonthsold(Saffranetal.,1996).
Thegapinthetimingbetweenthesuccessfullearningofadjacentdependenciesandthesuccessful
learningofNADspromptsresearcherstoinvestigateNADlearningatanearlierage. However,
theevidencepresentedislessclearasthereareconfoundsinthestudieswithyoungerinfants.
InanERPstudy,Friedericietal.(2011)foundthat4-month-oldinfantstrackedNADs
whenhearinganunfamiliarnaturallanguage. Theyfamiliarized4-month-oldmonolingual
GermaninfantstoItaliansentencesthatcontainedoneofthetwocorrectauxiliary-verbsuffix
pairings(puoX-areandstaX-ando). Theseauxiliary-verbsuffixpairingswereembeddedin
Italiansentences(e.g.,LasorellastacantandoorIlfratellostacantando). Infantswerethen
testedongrammaticalItaliansentencesandungrammaticalsentencesthatcontainedanincorrect
auxiliary-verbsuffixpairing(e.g.,puoX-andoorstaX-are). Amorepositive-goingwaveinthe
centro-parietalregionwasdetectedforgrammaticallyincorrectsentencescomparedto
grammaticallycorrectsentences. Friedericietal.(2011)thenconcludedthatafterabrief
exposuretoanunfamiliarnaturallanguage,4-month-oldswereabletoextractthenon-adjacent
dependenciesembeddedinthesequence. However,inthisdesign,aninfantmighthaveheard"La
sorellastacantando"inthelearningphaseand"Ilfratellostacantando"asagrammaticalstring
inthetestphase(A.D.Friederici,personalcommunication,April1,2020). Inthiscase,infants
mighthaverememberedthespecificauxiliary-verbstempairsinsteadofgeneralizingthe
non-adjacentdependenciesbetweenauxiliariesandverbsuffixes.
Inabehavioralstudy,MarchettoandBonatti(2015)familiarized12-month-oldinfants
withsequencesthatcontainedAXBpatternswhereAandBweresyllablesmutuallypredictiveof
eachother—i.e.,anon-adjacentdependencypair—andX wasoneofvarioussyllables. Inthis
study,theyincorporatedtwopairsofNADsinfamiliarization. Theninfantsheardboth
rule-words—wordsconformingtotheNADsheardduringfamiliarization,intheformofAA’Bor
AB’Bwhere A’andB’representitemsfromadifferentNADpair—andnon-words—words
violatingtheNADs,intheformofAB’X,XX’A,orXAA’.Thefindingdemonstratedthatwhen
therewerebrief200-msgapsinbetweenthesetri-syllabicwords,12-month-oldinfantslearned
theNADspresentedduringfamiliarization. UsingEEGmethods,Kabdebonetal.(2015)found
similarresultswith8-month-oldinfants. TheyemployedasimilardesignasMarchettoand
Bonatti(2015),butthetri-syllabicwordswereinsteadseparatedby25-mssubliminalpauses.
Resultsshowedthat8-month-oldinfantscouldreliablydistinguishrule-wordsfromnon-words,
indicatingsuccessfulNADlearning. However,bothofthesestudiesconfoundedNADcoherence
withpositionalcoherence. Inthetestmaterials: SequencesadheringtotheNADpatternsalsohad
itemsatthebeginningandendpositionsthatwereintheirtrainedpositions. Sequencesdeviating
fromtheNADpatternscontainedanedgepositionwithaniteminitsunattestedposition(i.e.,a
middleitemduringtrainingoccursinanedgepositionintesting,thusbreakingthepositional
coherence). Asedgepositionsarepotentiallymoresalient(Endress&Mehler,2009),this
positionalcoherencemightbenoticedbyinfantsaswell. Then,thepositionalcoherence,rather
thantheactualNADpatterns,mighthavebeenresponsibleforthecorrectresponses.
Todate,themostrobustevidenceforNADlearninginspeechiswith15-month-olds
(Gómez&Maye,2005). Eventhoughresearchershavebeentryingtopushtheacquisitionage
earlier,theevidencepresentedislessclearasthereareconfoundsinthestudieswithyounger
infants. Thelackofbehavioralevidenceininfantsyoungerthan15monthsofagealsomakesus
questionhowrobusttheactualNADlearningareinyoungerinfants. Thispromptsustolookat
NADlearninginadifferentdomain. Ifsuccessfullearningoccursinanotherdomain,theninfants
possessthecomputationalcapacityrequiredforNADlearning. Thequestionthenbecomeswhat
specificaspectsabouttheaforementionedexperimentsmadetheNADlearninginspeechless
robust. SinceresultsfromExperiment1andfindingsinadults’NADlearningliteratureseemto
suggestthatdynamicvisualmotionsareimportantforbothrulelearningandNADlearning,I
decidedtotest9-month-oldinfants’abilitytolearnNADsfromvisualhumanactionsin
Experiment2. ThiscanalsopotentiallyhelpanswerthepreviousquestionregardingtheNAD
learninginspeechwithinfants.
Experiment1
RecallthatSaffranetal.(2007)andBulfetal.(2015)reacheddifferentconclusionsfor
infants’repetitionrulelearningfromvisualstimulithanJohnsonetal.(2009). Whilefamiliar
categoriesmighthavefacilitatedinfants’visualrulelearning,simultaneouspresentationsmight
havemadethetaskeasierforinfantsaswell. Itseemedthatinfantswerenotabletolearnrules
robustlyfromsequentiallypresentedstimuli. Yet,morecomplexNADvisualpatternswere
learnablebyadultswithstimulidisplayingdynamicmotions(Endress&Wood,2011;Li&
Mintz,2015). Perhapsinfants’sequentialrulelearningwouldalsobefacilitatedbythesestimuli.
Totestthispossibility,Iexamined9-month-oldinfants’sequentialrulelearning(e.g.,ABAand
ABB)withvisualhumanactions-adomainwithdynamicvisualmotions. Ifinfantsareableto
learnsequentialrulesfromvisualhumanactions,then,combiningtheresultsoftheirfailureswith
otherstimuliinthevisualdomain,itsupportsthatdynamicvisualmotionscanfacilitateinfants’
sequentialrulelearning.
Method
Participants
InfantswererecruitedfromtheGreaterLosAngelesAreabyemailsandphonecalls. The
contactinformationweregeneratedfromadatabaseofparentswhohadexpressedinterestsin
havingtheirchildrenparticipateinresearchafterseeingouradvertisementsonFacebook. Parents
gaveinformedconsentbeforeinfantsstartedtheexperiment. Attheendoftheexperiment,we
gavetheparentat-shirtorasmalltoyfortheirchildasatokenofappreciation. Wetested18
full-terminfantsbetween8.5and9.5monthsofage(M =9.0months). Sixadditionalinfants
weretestedbutnotincludedintheanalysisduetofussiness(n=5)andprematurebirth(n=1).
Allexperimentsinthispaperwereapprovedbytheinstitution’sInstitutionalReviewBoard.
Apparatus
Infantssatontheirparents’lapsinadimroom,witha50-inchscreeninfrontofthem.
Parentsworeview-obstructingglassesandwereinstructedtonotinteractwiththeirinfantsduring
theexperiment. TheexperimentalmaterialswerepresentedusingthesoftwareHabit2.2.4
installedonanHPEliteOne800computerrunningWindows7(Oakesetal.,2019). Thestimuli
werepresentedonthescreeninfrontoftheinfant. Anexperimenter,blindtothestimulithat
infantswereviewing,live-codedwheninfantslookedatandlookedawayfromthescreenina
separatecontrolroom.
Stimuli
HabituationandtestmaterialswerehumanactiontripletssimilartothoseusedinLiand
Mintz(2015)andEndressandWood(2011). Eachactionclipwithinatripletlasted0.6second.
Eachactionclipstartedandendedwiththeanimatedhumanavatarinaneutraluprightposition
witharmsatthesidesandheadfacingforward. Thisensuredthatactionsequencesflownaturally
fromcliptoclip. SomeofthetripletsfollowedaABApattern(e.g.,headturn-legraise-head
turn),andothersfollowedaABBpattern(e.g.,headturn-legraise-legraise). Figure1contains
Figure1
ImagesofthemidpointofhumanactionclipsusedinExperiment1and2
Figure2
ExamplesofhabituationmaterialsfromtheABAandABBconditionsinExperiment1
framesexcerptedfromthehumanactionclipsusedinbothExperiment1and2. Theyarethe midpointsofeachactionclipsthatdepictthemaximumextentofmovements.
Thehabituationmaterialswerecreatedfromeightuniquehumanactionclips,half
assignedtoclassAandtheotherhalfassignedtoclassB.Then,theAandBactionclipswere
combinedexhaustivelytocreate16uniqueABAand16uniqueABBhumanactiontriplets. Infants
eithersawABAhumanactiontriplets(ABAhabituationcondition)orABBhumanactiontriplets
(ABBhabituationcondition)duringhabituation. Eachhabituationtrialconsistedofadifferent
pseudo-randomsequenceofthe16uniquehumanactiontriplets. Withinatrial,tripletswere
separatedbya0.75-secondblankscreentoaidsegmentation. Therefore,eachhabituationtrial
videocouldlastforatmost40seconds. Figure2showsthesamplematerialsforeachofthe
habituationconditions.
Thetestmaterialswerecomprisedoffourcompletelynovelhumanactionclips,two
assignedtoclassAandtwotoclassB.FouruniqueABAactiontripletsandfouruniqueABB
actiontripletsweregeneratedfromthesefournovelactionclips. Thetestphaseconsistedofeight
testtrials,witheachtrialcontainingrepetitionsofasingleactiontriplet,witha0.75-secondblank
screenseparatingtherepetitions. Foreachinfant,halfofthetesttrialsfollowedthepatternseen
duringhabituation(i.e.,consistent),andhalffollowedaninconsistentpattern(i.e.,ABBpatternfor
infantshabituatedtotheABApattern,andABApatternforinfantshabituatedtotheABBpattern).
Procedures
Weusedavisualhabituationproceduresimilartothoseusedinotherexperimentsthat
testedforinfants’visualrulelearning(Johnsonetal.,2009;Saffranetal.,2007). Theexperiment
consistedofahabituationphaseandatestphase. Theexperimentstartedwithanattention-getting
videoonthescreen. Assoonastheinfantattendedtothescreen,thehabituationphasestarted. A
habituationtrialbeganonceaninfantorientedtothescreenandendedwhentheinfantlooked
awayfromthescreenformorethantwoconsecutiveseconds. Thevideoloopedifthetrialwas
notterminatedbeforethevideoreachedtheend. Whenahabituationtrialended,an
attention-gettingvideoappearedonthescreentorecapturetheinfant’sattentionbeforethenext
habituationtrialstarted. Thehabituationphaseendedwhentheaverageoftheinfant’slooking
timestothelastthreehabituationtrialswaslessthan50%oftheaveragelookingtimetothefirst
threetrials,orwhentheinfantreachedthemaximumof25habituationtrials.
Movingintothetestphase,infantssawbothtrialswithconsistentpatternsandthosewith
inconsistentpatterns. Thetrialtypesinterleavedinthetestphase,andthetypeofthefirsttrialwas
counterbalancedacrossinfants. Atesttrialstartedonceaninfantattendedtothescreenandended
whentheinfantlookedawayfromthescreenfortwoconsecutiveseconds. Ifinfantsweretolearn
thenon-adjacentrule(i.e.,ABA)embeddedinhabituation,thenwewouldexpectthemtolook
longerattheinconsistenttesttrialsascomparedtotheconsistenttesttrials.
Results
Wefirstexcludedtesttrialswithlookingtimeslessthan1.8seconds,becausethiswasthe
timeneededforseeingatleastoneiterationoftheactiontriplet(2). Foreachinfant,wethen
excludedtrialswithlookingtimesthatwere1.5timestheinterquartilerangehigherthanthe
upperquartileorthatmuchlowerthanthelowerquartileforthatinfant(8). Afterthesetwosteps,
wehadatotalof134testtrialsfromthe18infants,andeachinfanthadatleastsixvalidtrials.
Foreachinfant,wethencalculatedtheiraveragelookingtimestotheconsistenttrials(M =9.79s,
SD=4.67s)andtheiraveragelookingtimestotheinconsistenttrials(M =12.41s,SD=6.50s).
Theselookingtimeswerenotnormallydistributedacrosssubjects(W =.73,p<.001for
consistenttrials;W =.88,p<.001forinconsistenttrials;). However,wewereinterestedinthe
differencesacrossconsistentandinconsistenttesttrialswithineachindividual. Thedifferencesof
thelookingtimesforconsistentandinconsistenttrialswerenormallydistributed(M
diff
=-2.63s,
SD
diff
=5.54s,W =.95,p=.390).
NosignificantdifferencewasfoundbetweenresultsfrominfantshabituatedtotheABA
conditionandfromthosehabituatedtotheABBcondition(t(16)=.02,p=.983). Therefore,we
combinedthedatafromtwohabituationconditionsforthefollowinganalysis. Wethenconducted
1 aone-tailedt-testtocomparethedifferencesinlookingtimestogrammaticalandungrammatical
trialsacrossthetwohabituationconditions. Aone-tailedt-testwasjustifiedbecausewe
hypothesizedthedirectionofthedifferencebeforehand,andthedifferencesoflookingtimesfor
consistentandinconsistenttrialswerenormallydistributed. Infantslookedsignificantlylongerto
theinconsistenttrialsthantheconsistenttrials(t(17)=-2.01,p=.030,Cohen’sd =.47,see
Figure3). Thirteenoutofthe18infantslookedlongertoinconsistentversusconsistenttrials
(c 2
(1,N =18)=2.72,p=.049,withYates’correction).
Figure3
By-subjectmeandifferencescorebetweenconsistentandinconsistenttesttrialsinExperiment1
Discussion
Inthisexperiment,wetested9-month-oldinfants’abilitytolearnsequentialrepetition
rules(i.e.,ABAandABB)fromvisualhumanactions. Resultsshowedthatthese9-month-old
infantslearnedtheabstractruleinthehabituationphaseandgeneralizedittothenewtesthuman
actions. ThisfindingpresentsacontrasttothefindingsofJohnsonetal.,2009,where8-and
11-month-oldsshowedevidenceoflearningwhenhabituatedtoABBsequencesandtestedon
ABAvs. ABBsequences,butnoevidenceoflearningwhenhabituatedtoABAsequences. It
1 supportstheargumentthatdynamicvisualmotionsmightfacilitateinfants’sequentialrule learninginthevisualdomain.
Currently,sequentialABArulelearninghasbeenfoundinyounginfantswithtwo
domains: speechandvisualhumanactions. Withintheauditorydomains,infantscouldlearn
sequentialrepetitionrulesfromspeechbutnotfrommusicaltones(Marcusetal.,2007). Within
thevisualdomains,infantscouldlearnfromhumanactionsbutnotfromshapes(Saffranetal.,
2007). Thesuccessfulsequentialrulelearninginthedomainsofspeechandvisualhumanactions
suggeststhatbothmightbegovernedbysimilarprinciples. Itisworthnotingthatbothofthem
consistofstimulithatareperceivableoutcomesofhumanactions: infantsmightmapthose
stimuliontotheirownbodies,makingtheencodingandmemorizationofthestimulimorerobust,
thusenablingthemtolearnrulesfromthesetwodomains. Thispromptedustofurtherexplore
infants’abilitytolearnrulesfromvisualhumanactions. Towardsthisend,wewantedtotest
sequentialrulelearningwithvisualhumanactionstoamoredifficultcase,withactualitem
dependenciesinsteadofrepetitions.
Experiment2
InExperiment1,wefoundthat9-month-oldinfantslearnedtherepetitionruleswith
humanactions,supportingtheclaimthatinfants’sequentialrulelearningcouldbefacilitatedby
dynamicvisualmotions. Thisparallelsthefindingsfromadults’NADlearningliteraturethat
dynamicvisualmotionssupportadults’NADlearning. Ifweconsidersequentialrulelearningas
aneasierversionofNADlearningwherepeoplearetrackingrepetitionsinsteadofactualitem
dependencies,thenitwouldbereasonabletopositthatinfants’NADlearningisfacilitatedby
dynamicvisualmotionsaswell. Totestthisclaim,Iexamined9-month-olds’learningof
non-adjacentdependencies(i.e.,intheformofaXbwhereaandbrefertoaspecificitemandX
referstooneitemfromaclassofitems)withvisualhumanactionsinExperiment2. Recallthat
themostrobustevidenceforinfants’NADlearningcurrentlyiswith15-month-olds(Gómez&
Maye,2005). AsafirststepatexaminingvisualNADlearning,weaddedaprimingphaseinthe
1 beginningofthenormalhabituationproceduretogivetheseinfantssomeextrahelp. Thepriming
phaseconsistedofhumanactiontripletsfollowingtheABApattern(similartothoseseenin
Experiment1). ByexposinginfantstotheABApatternfirst,wehopedthatthiswouldhighlight
thestartandendpositionsofthetripletsforinfants,thushelpingthembetternoticethe
non-adjacentdependenciesembeddedinhabituation.
Method
Participants
InfantswererecruitedfromthesamedatabaseusedinExperiment1. Parentswere
contactedbyemailsandphonecalls. Theygaveinformedconsentbeforetheirchildstartedthe
experiment. Attheendoftheexperiment,wegavetheparentat-shirtorasmalltoyfortheirchild
asatokenofappreciation. Wetested18full-terminfantsbetween8.5and9.9monthsofage(M =
9.1months). Sevenadditionalinfantsweretestedbutnotincludedintheanalysisduetofussiness
(n=6)andexhibitingmaximallookingtimestomorethanhalfofthetesttrials(n=1).
Apparatus
TheexperimentsetupandtheapparatususedwerethesameasinExperiment1.
Stimuli
TheexperimentalmaterialswerehumanactionclipssimilartothoseusedinExperiment1
(seeFigure1). Eachactionclipwithinatripletlasted0.6second. Eachactionclipstartedand
endedwiththeanimatedhumanavatarinaneutraluprightpositionwitharmsatthesidesand
headfacingforward.
TheprimingphaseconsistedofABAhumanactiontripletssimilartothoseusedin
Experiment1. TheprimingmaterialswerecreatedfromtwoclassAhumanactionclipsandthree
classBhumanactionclips. TheywereexhaustivelycombinedtocreatesixuniqueABAhuman
actiontriplets. Eachprimingtrialconsistedofauniquelyrandomorderofthesixtriplets. Triplets
withinaprimingtrialwereseparatedbya0.75-secondblankscreen.
1 Thehabituationphasewascomprisedoftennovelhumanactionclips. Fourofthehuman
actionclipswereusedtocreatetwoNADs: a_bandc_d (here,eachletterrepresentsaunique
humanactionclip). Oftheremainingsixclips,three(X
1 3
)occurexhaustivelyinthemiddle
positionofa_b,andtheotherthree(X
4 6
)occurexhaustivelyinthemiddlepositionofc_d
(creatingatotalofsixuniquetriplets). Researchhassuggestedthatthreeinterveningitemshave
enoughvariabilityforpeopletolearntheNADs(Wangetal.,2019). Theassignmentfromaction
clipstothelettersherewererandomforeachinfant. Eachhabituationtrialhadtwoblocksofthe
sixrandomlyorderedtripletsthatwereseparatedby0.75-secondblankscreens.
Thetestphasecontainedeightdifferenttesttriplets: fourgrammaticaltripletsandfour
ungrammaticaltriplets. GrammaticaltripletswereoftheformaBb,wherea_bisoneoftheNADs
inhabituationandBrepresentsoneofthemiddleitems(i.e.,classB)inpriming. Ungrammatical
tripletswereoftheformaBd,combiningthefirstactionclipofoneNAD(a_),amiddleitemin
priming,andthelastactionclipofanotherNAD(_d). Thus,alltesttripletsthemselveswere
novelandhadzeroadjacenttransitionalprobabilities,butalltheactionclipswithinthetriplets
werefamiliartoinfants. However,thegrammaticaltripletsfollowedtheNADpatternsandthe
ungrammaticaltripletsviolatedthepatterns. AsinExperiment1,agiventesttrialconsistedof
repetitionsofoneactiontriplet.
Figure4showsasamplehabituationtrial,andanexampleforeachofthetestitemtypes.
Procedures
TherewasaprimingphaseaddedbeforethehabituationphaseofExperiment2. Thus,
Experiment2containedaprimingphase,ahabituationphase,andatestphase.
Theexperimentstartedwithanattention-gettingvideoonthescreen. Assoonastheinfant
attendedtothescreen,theprimingphasestarted. Aprimingtrialstartedonceaninfantfixatedon
thescreen,andendedwhentheinfantlookedawayfortwoconsecutivesecondsorwhenthe
infantsawall6tripletsinatrial(atotalof15.3seconds). Aprimingphaseendedwhentheinfant
reachedacumulativelookingtimeofatleast30seconds,orwhentheinfantfinishedeight
1 Figure4
Examplesofhabituationmaterialsandgrammaticalvs. ungrammaticaltestitemsinExperiment2
primingtrialsbeforehittingthefirstcriterion. SinceExperiment1showedthat9-month-old
infantscouldlearntheABApatternfromhumanactions,exposuretothesesequencescould
potentiallyhighlightthenon-adjacentrelationandmakeitmoresalientforinfantswhenthey
processthenon-adjacentdependencytripletsinhabituation.
Thehabituationphasefollowedtheprimingphase. Ahabituationtrialstartedoncethe
infantlookedatthescreenandendedwhentheinfantlookedawayfromthescreenformorethan
twoconsecutiveseconds,orwhentheinfantsawall12tripletsinthatparticulartrial(atotalof
30.6seconds). Whenahabituationtrialended,anattention-gettingvideooccurredtore-capture
theinfant’sattention. Thehabituationphaseendedwhentheaverageoftheinfant’slookingtimes
tothelastthreehabituationtrialswaslessthan50%oftheaveragelookingtimetothefirstthree
trials.
1 Thetestphasestartedonceaninfantmetthehabituationcriterion. Infantssawboth
grammaticalandungrammaticaltesttrialsinthetestphase,withtrialtypesinterleavedwithinthe
testphase. Thetypeofthefirsttrialwascounterbalancedacrossinfants. Atesttrialstartedonce
theinfantorientedtothescreen,andendedwhenaninfantlookedawayfromthescreenfortwo
consecutiveseconds,orwhenaninfantsawsixrepetitionsofthetriplet. Ifinfantsareableto
learntheNADsembeddedinthehabituation,thenweexpectthemtolooklongeratthe
ungrammaticaltripletsthanthegrammaticaltriplets.
Results
AsinExperiment1,wefirstexcludedtrialswithlookingtimesthatwereshorterthan1.8
seconds(4)orwereoutlierswithineachinfant(9). Afterthesetwosteps,wehadatotalof131
testtrialsforthe18infants,andeachinfanthadatleastfivevalidtrials. Foreachinfant,we
calculatedtheiraveragelookingtimestothegrammaticaltrials(M =6.36s,SD=2.10s)and
theiraveragelookingtimestotheungrammaticaltrials(M =7.70s,SD=3.06s). Thedifferences
ofthelookingtimesforgrammaticalandungrammaticaltrialswerenormallydistributed(M
diff
=
-1.34s,SD
diff
=2.42s,W =.96,p=.520). Aone-tailedt-testshowedthatinfantslooked
significantlylongertotheungrammaticaltrialsthanthegrammaticaltrials(t(17)=-2.35,p=
.016,Cohen’sd =.55,seeFigure5). Thirteenoutof18infantslookedlongertoungrammatical
versusgrammaticaltesttrials(c 2
(1,N =18)=2.72,p=.049,withYates’correction).
Discussion
Inthisexperiment,weexamined9-month-oldinfants’abilitytolearnNADsfromvisual
humanactionswhentheywerebrieflyexposedtoasimplenon-adjacentrepetitionrule(i.e.,ABA)
first. Resultsshowedthat9-month-oldinfantscouldlearnNADsfromvisualhumanactionsif
theyhadbriefexposuretoABApatternsfirst. Thisisthefirstevidencethatweknowofforvisual
non-adjacentdependencylearningininfants. Severalreasonsmightexplainthesuccessful
learninghere: 1)thebriefexposuretoABApatternsmighthavehighlightedthestartandend
positionsofthetripletsforinfants,thusmakinglaterdependencieseasiertodetect;2)as
1 Figure5
By-subjectmeandifferencescorebetweengrammaticalandungrammaticaltesttrialsin Experiment2
suggestedbyinfants’successfullearningofABArulesfrombothspeechandhumanactions,
infantsmightbeabletolearnnon-adjacentpatternsmorereadilyfromspeechanddomainsthat
involvedynamicvisualmotions. Thesetwopossibleexplanationsarenotmutuallyexclusive.
Furtherexperimentsareneededtobetterunderstandthesetwopossibleexplanations.
GeneralDiscussion
Toinvestigatetheroleofdynamicvisualmotionsininfants’ruleandNADlearning,I
conductedtwoexperimentswith9-month-oldinfants. Resultsshowedthat9-month-oldinfants
wereabletolearnbothrepetitionrules(e.g.,ABAandABB)andnon-adjacentdependenciesfrom
visualhumanactions. ThesuccessfullearninginbothExperiment1and2suggeststhatinfants’
sequentialrulelearningandNADlearningarefacilitatedbydynamicvisualmotions.
Howdothesefindingsinformtheliteratureoninfantvisualpatternlearning? Rabagliati
etal.(2019)recentlyperformedametaanalysisoftheexperimentsexamininginfants’rule
learningandproposedthatinfantswereonlyabletolearnrulesfromthosestimulithatwere
1 consideredmeaningfulorcommunicative. However,theydidnotseparatesequentialand
simultaneouspresentations. RecallthatSaffranetal.(2007)andJohnsonetal.(2009)found
differentresultsforinfants’rulelearningfromvisualstimuli,andtheyemployeddifferent
presentationmethodsintheirexperiments. ThefindingsmighthavedifferedifRabagliatietal.
(2019)weretoseparatelylookatexperimentswithsequentiallypresentedstimuliand
experimentswithsimultaneouslypresentedstimuli. Furthermore,inafollow-upstudy,Rabagliati
etal.(2019)foundthat7-month-oldinfantslearnedABAandABBrulesfrommanualgestures
(akintosignlanguagegestures)onlyiftheywereprimedtoconsiderthesegesturesas
communicative. Inapre-exposurephase,halfoftheinfantsviewedashortvideoinwhichone
actorusedASLgesturestocommunicatewiththeotheractorandtheinfant,andtheotheractor
respondedinspeech. Theotherhalfviewedashortvideoinwhichtwoactorssimultaneously
producedthesamegesturesequencebutwerenotfacingtowardseachotherortheinfant. Infants
whosawthesecondvideofailedtolearntherepetitionrulesfromsequentiallypresentedgestures.
Theauthorsarguedthatinfants’failureinthesecondscenariowasbecausetheydidn’tconsider
gesturesascommunicativeinthefirstplace. Thisisinterestingbecauseourstudiesfoundthat
9-month-oldinfantslearnedABArulesfromhumanactions,andtherandomlycombinedactions
presentedcouldhardlybeconsideredascommunicative. Furthermore,gesturescanbeconsidered
asaspecifictypeofhumanactions,andtheyinvolvedynamicvisualmotions. Therearetwo
possibleexplanationsforthedifferenceinourfindings: 1)Infantswhoparticipatedinourstudy
wereolderthanthoseinRabagliatietal.(2019). Nine-month-oldinfantsmighthavebetter
computationalcapacitiesthan7-month-oldinfants. However,thisalonewouldnotbeenoughto
explainwhy7-month-oldslearnedthecommunicativegestures. While7-month-oldssucceeded
withrulelearningwithspeech,itispossiblethattheywerejustconsideringthegesturespresented
asapartofspeechinsteadofprocessingthemasmotions. Ifthedifferenceismainlyduetoage,
thenthesuccessfullearningofcommunicativegesturesin7-month-oldsmightindeedsuggestthat
rulelearningismoreenhancedwithspeech. 2)Itcouldalsobethecasethatinfantscouldlearn
rulesfromspeechandhumanactionsequallywell,butthedifferencecamefromthemovements
weused. Inourstudy,bigmusclemovementswereusedinsteadofthefinegesturemovements.
Bigmusclemovementsarepotentiallymoresalienttodetectandsegmentthanfinegesture
movements,resultinginbetterrulelearning. Furthermore,aswehavenotedinthediscussionof
Experiment1,speechandhumanactionsarecurrentlytheonlytwodomainsinwhichsuccessful
sequentialrulelearninghasoccurredforinfants,andbothdomainscontainstimulithatare
perceivableoutcomesofhumanactions. Itcouldbethatinfantsweremappingthestimulionto
theirownbodies,thusenablingthemtobetterrepresentandencodethespeechsoundsand
actions,whichmayresultinbetterlearning. Thismighthelpexplainwhyourstudyfounda
successfullearningwhileRabagliatietal.(2019)’sdidnot. Asbigmusclemovementsaremore
salienttodetectandsegment,infantsmightbeabletobettermapthesemovementsontotheirown
bodiesthantomapfinegesturemovements. Inthiscase,communicativesignalsmightalso
facilitatelearning,butmightnotbenecessary.
Infants’successfulrulelearninginExperiment1providesfurtherevidenceforaclaim
firstmadeintheadults’NADlearningliterature: dynamicvisualmotionsfacilitateruleandNAD
learninginthevisualdomain. Thispromptedmetofurtherinvestigatethisclaimandseewhether
dynamicvisualmotionsplayaroleininfants’NADlearning. Furthermore,sincethecurrent
findingsoninfants’NADlearningareunclearastowhethersuccessfulNADlearningactually
occurred(duetopotentialconfoundsintheexperiments),therewasanotherreasonforustolook
atNADlearningfromadifferentdomain. Ifsuccessfullearningoccursinanotherdomain(inthis
case,visualhumanactions),theninfantspossesstherequiredcomputationalcapacityforNAD
learningatayoungeragethanwehadexpected. InExperiment2,9-month-oldinfantslearned
NADsfromhumanactionswhentheywerefirstexposedtosimplenon-adjacentrules(inthis
case,ABApatterns). Thisshowsthat9-month-oldinfantshavethecomputationalcapacitiesto
learnNADs. However,itisnotclearhowmuchtheinitialprimingphasecontributedtothis
success. Infantsmighthavebeenmoreawareofthebeginningandendpositionsduetotheinitial
priming,thusresultinginthesuccessfullearningoflaterNADs. Futurestudiesneedtoexamine
whether9-month-oldinfantscanlearnNADsfromhumanactionswithouttheaidofpriming.
Eitheroutcomewillhaveusefulimplicationsforinfants’sequentialrulelearningandNAD learning.
If9-month-oldinfantscanstilllearnNADsfromhumanactions,thenat9monthsofage
infantsalreadypossessthecomputationalcapacitiesrequiredforNADlearning. Recentevidence
fromourlabsuggeststhat9-month-oldsarenotabletolearnNADsfromartificialspeechwithout
thesupportofABApatterns(Chiang,2018). Thequestionthenbecomeswhatspecificallyis
makingtheNADlearninginspeechlessrobustinyounginfantsthanthatofhumanactions. One
possibilityisthatinfants’NADlearninginspeechbuildsontheirNADlearninginhumanactions.
Recallourpreviousconjecturethatinfantsmapspeechandhumanactionsontotheirownbodies.
Iftheenhancedsequentialrulelearninginspeechandhumanactionsisindeedduetoinfants’
abilitytomapstimuliontotheirownbodies,thenwewouldexpecttoseeanearlyabilitytolearn
NADsfromvisualhumanactionsastheseactionscanbeeasiertomapontoone’sownbodythan
speechsounds—theoutcomesofvisualhumanactionsaremoreperceivableandthusbetterfor
ruleencoding. AsthelearningofNADsispotentiallymorechallengingthanthatofrepetition
rules,infantsmightbeabletolearnNADsfromhumanactionsatanyoungeragethanfrom
speech.
If9-month-oldinfantsfailtolearnNADsfromhumanactionswithouttheaidofpriming,
thenitwillbeinterestingtoconsiderwhatmakestheprimingphaseeffectiveforsubsequentNAD
learningandtowhatextenttheprimingpatternneedstobesimilartothesubsequentNAD
patterns. Isitnecessaryforinfantsneedtopayattentiontothebeginningandendofthetriplet?
OrisitthattheycanreadilylearnNADsaslongastheyknowthatthereissomesortofpatterns
embeddedinthesequences? Futureexperimentscananswerthesequestionsandhelpusgain
moreinsightsintotheprimingeffectdisplayedhere.
Insummary,Ipresentedtwoexperimentsinvestigating9-month-oldinfants’learningof
sequentialrepetitionrules(e.g.,ABAandABB)andnon-adjacentdependenciesfromvisual
humanactions. InExperiment1,9-month-oldinfantswereabletolearnsequentialrepetition
rulesfromvisualhumanactions. InExperiment2,9-month-oldinfantssucceededatlearning
2 non-adjacentdependenciesfromvisualhumanactionswhentheywerefirstexposedtovisual humanactionsintheABApattern. Findingsofthesetwostudiessupporttheargumentthat dynamicvisualmotionsfacilitateinfants’learningofabstractrulesandpatterns.
2 References
Bulf,H.,Brenna,V.,Valenza,E.,Johnson,S.P.,&Turati,C.(2015).Manyfaces,onerule:The
roleofperceptualexpertiseininfants’sequentialrulelearning.FrontiersinPsychology,6,
1595.
Chiang,C.Non-adjacentdependencytrackingin9-month-oldinfants:Usingafasterspeechrate
tofacilitatelearning.MasterThesis.UniversityofSouthernCalifornia,2018.
Cooper,R.P.,&Shallice,T.(2006).Hierarchicalschemasandgoalsinthecontrolofsequential
behavior.
Creel,S.C.,Newport,E.L.,&Aslin,R.N.(2004).Distantmelodies:Statisticallearningof
nonadjacentdependenciesintonesequences.JournalofExperimentalPsychology:
Learning,Memory,andCognition,30(5),1119.
Endress,A.D.,&Hauser,M.D.(2010).Wordsegmentationwithuniversalprosodiccues.
CognitivePsychology,61(2),177–199.
Endress,A.D.,&Mehler,J.(2009).Primitivecomputationsinspeechprocessing.TheQuarterly
JournalofExperimentalPsychology,62(11),2187–2209.
Endress,A.D.,&Wood,J.N.(2011).Frommovementstoactions:Twomechanismsforlearning
actionsequences.CognitivePsychology,63(3),141–171.
Friederici,A.D.,Mueller,J.L.,&Oberecker,R.(2011).Precursorstonaturalgrammarlearning:
Preliminaryevidencefrom4-month-oldinfants.PLoSOne,6(3).
Gebhart,A.L.,Newport,E.L.,&Aslin,R.N.(2009).Statisticallearningofadjacentand
nonadjacentdependenciesamongnonlinguisticsounds.PsychonomicBulletin&Review,
16(3),486–490.
Gómez,R.,&Maye,J.(2005).Thedevelopmentaltrajectoryofnonadjacentdependency
learning.Infancy,7(2),183–206.
Johnson,S.P.,Fernandes,K.J.,Frank,M.C.,Kirkham,N.,Marcus,G.,Rabagliati,H.,&
Slemmer,J.A.(2009).Abstractrulelearningforvisualsequencesin8-and
11-month-olds.Infancy,14(1),2–18.
2 Kabdebon,C.,Pena,M.,Buiatti,M.,&Dehaene-Lambertz,G.(2015).Electrophysiological
evidenceofstatisticallearningoflong-distancedependenciesin8-month-oldpretermand
full-terminfants.BrainandLanguage,148,25–36.
Li,J.,&Mintz,T.H.(2015).Constraintsonlearningnon-adjacentdependencies(nads)ofvisual
stimuli.,InCogsci.
Marchetto,E.,&Bonatti,L.L.(2015).Findingwordsandwordstructureinartificialspeech:The
developmentofinfants’sensitivitytomorphosyntacticregularities.JournalofChild
Language,42(4),873–902.
Marcus,G.F.,Fernandes,K.J.,&Johnson,S.P.(2007).Infantrulelearningfacilitatedbyspeech.
PsychologicalScience,18(5),387–391.
Marcus,G.F.,Vijayan,S.,Rao,S.B.,&Vishton,P.M.(1999).Rulelearningbyseven-month-old
infants.Science,283(5398),77–80.
Newtson,D.,&Engquist,G.(1976).Theperceptualorganizationofongoingbehavior.Journalof
ExperimentalSocialPsychology,12(5),436–450.
Oakes,L.M.,Sperka,D.,DeBolt,M.C.,&Cantrell,L.M.(2019).Habit2:Astand-alone
softwaresolutionforpresentingstimuliandrecordinginfantlookingtimesinorderto
studyinfantdevelopment.BehaviorResearchMethods,51(5),1943–1952.
Rabagliati,H.,Ferguson,B.,&Lew-Williams,C.(2019).Theprofileofabstractrulelearningin
infancy:Meta-analyticandexperimentalevidence.DevelopmentalScience,22(1),e12704.
Saffran,J.R.,Aslin,R.N.,&Newport,E.L.(1996).Statisticallearningby8-month-oldinfants.
Science,274(5294),1926–1928.
Saffran,J.R.,Pollak,S.D.,Seibel,R.L.,&Shkolnik,A.(2007).Dogisadogisadog:Infant
rulelearningisnotspecifictolanguage.Cognition,105(3),669–680.
Wang,F.H.,Zevin,J.,&Mintz,T.H.(2019).Successfullylearningnon-adjacentdependenciesin
acontinuousartificiallanguagestream.CognitivePsychology,113,101223.
Abstract (if available)
Abstract
Our ability to detect and generalize patterns is important for different kinds of learning, and crucial for our adaptations. Indeed, infants as young as seven months succeeded in learning simple repetition rules like ABA and ABB from speech. However, the evidence for infants' ability to learn such patterns from the visual domains is mixed. While infants succeeded in learning these patterns from simultaneously presented stimuli, they were less able to learn from sequentially presented stimuli. This posed the question as to why infants' sequential rule learning is challenged in the visual domain but not in speech. Recent studies on adults' non-adjacent dependency (NAD) learning—which can be considered a more complex rule than simple repetitions—suggest that their NAD learning in the visual domain is aided by dynamic visual motion. It is possible, then, that infants' rule learning in the visual domains is facilitated by dynamic motion as well. To test this possibility, I tested 9-month-old infants on their ability to learn rules from visual human actions (a domain that involves dynamic visual motions). Experiment 1 found that these infants successfully learned the rules from visual human actions, supporting the previous claim. To see if dynamic visual motions have a broader impact on infants' learning of different patterns, I then tested 9-month-old infants' ability to learn NADs from visual human actions. Results of Experiment 2 showed that infants learned NADs from visual human actions when they were first exposed to human actions in the ABA pattern. This is the first evidence for infants' NAD learning in a visual domain, and the age of NAD learning is younger than what the literature had suggested. Possible explanations for the success in NAD learning with human actions were discussed.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Mechanisms underlying acquisition of non-adjacent dependencies
PDF
Can infants discriminate between declaratives and interrogatives?
PDF
Statistical learning is a two-step process
PDF
Learning about word learning in ambiguous situations
PDF
Motor cortical representations of sensorimotor information during skill learning
PDF
Building blocks of action and goal comprehension in a newborn visual system
PDF
When and how infants discriminate between declaratives and interrogatives
PDF
The psychology and psychophysics of voice recognition
PDF
Cognitive-linguistic factors and brain morphology predict individual differences in form-sound association learning: two samples from English-speaking and Chinese-speaking university students
PDF
Sensitive, specific, and generative face recognition in a newborn visual system
PDF
When things are left unsaid: existential and anaphoric implicit objects in discourse
PDF
The role of individual variability in tests of functional hearing
PDF
Active state learning from surprises in stochastic and partially-observable environments
PDF
Human–building integration: machine learning–based and occupant eye pupil size–driven lighting control as an applicable visual comfort tool in the office environment
PDF
Implicit theory ideology about human attributes and aggression
PDF
Rationality and the primacy of the occurrent
Asset Metadata
Creator
Lu, Shiyang
(author)
Core Title
Learning repetition rules and non-adjacent dependencies from human actions in nine-month-old infants
School
College of Letters, Arts and Sciences
Degree
Master of Arts
Degree Program
Psychology
Publication Date
07/29/2020
Defense Date
04/17/2020
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
human actions,non-adjacent dependencies,OAI-PMH Harvest,rule learning
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Mintz, Toben H. (
committee chair
), Manis, Franklin R. (
committee member
), Zevin, Jason D. (
committee member
)
Creator Email
helen.lv08@gmail.com,lushiyan@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-350355
Unique identifier
UC11663435
Identifier
etd-LuShiyang-8818.pdf (filename),usctheses-c89-350355 (legacy record id)
Legacy Identifier
etd-LuShiyang-8818.pdf
Dmrecord
350355
Document Type
Thesis
Rights
Lu, Shiyang
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
human actions
non-adjacent dependencies
rule learning