Summary
Contemporarycognitivepsychologyisabroadsubjectfield.Itsresearchinvolvesallaspectsofhumancognition,anditsconceptsandtheoriesalsopenetratealmostpsychologyAllbranchesofthefield.Interestingly,theconnectionistresearchthatemergedduringthesecondcognitiverevolutiondidnotmakeinformationprocessingpsychology,whichischaracterizedbycomputeranalogy,withdrawfromtheforefrontofpsychology.Itevenstillexistsasthemainbodyofcognitivepsychology..Therefore,tograspthefullpictureofcontemporarycognitivepsychology,wemustfirstgraspthebasicviewpoints,researchcontentandresearchmethodsofinformationprocessingcognitivepsychology.
Itisanactivityprocessofproducingsecondaryinformationwithhighvaluecontentandconvenientforuserstouseonthebasisoforiginalinformation.Thisprocesswilladdvaluetotheinformation.Onlyonthebasisofproperprocessingofinformationcannewandeffectiveinformationorknowledgebeproducedtoguidedecision-making.Informationprocessingpsychologyistostudythetransmission,conversion,storageandfunctionmechanismofinformationintheinformationprocessingsysteminordertouncovertheactivitiesinthe"blackbox"ofthemind.
Basicviewpoints
Asweallknow,computersprovidethemostconvenientmetaphorforinformationprocessingpsychology—comparethehumanbraintoacomputer.Severaloperationsactonsymbols,andprocessinginformationisstillstoredinsymbolform.Thestructureandprocessofprocessingcanbevisuallyexpressedasflowchartsoras"boxes-and-arrowsmodels".Thehumanbrainisaninformationprocessingsystem.Itcaninput,encode,store,extract,copy,andtransmitphysicalsymbolsthatcharacterizeinformation.Thecompletionofthisprocessisaseriesofprocesses,anddifferentprocessingtasksandprocessingstagesaredifferent.Cognitivestructureiscompleted.Theserelativelyindependentcognitivestructuresareconnectedbackandforth,andhavehierarchicaldifferences.Theyare"machines"inthehumanbrainsimilartoartificialintelligencemachines.
Informationprocessingsystem,alsoknownas"SymbolOperationSystem",ismainlycomposedoffourparts,namelysensor,effector,processor,memorydevice,amongwhichsensoristhedevicethatreceivesinformation,Thatis,thesensorysystem;theprocessoristhecontrolpartoftheentireinformationprocessingsystem,whichdeterminesthegoal,planandexecutionoftheinformationprocessing,includingthecontrolprocessorandworkingmemoryinthefigure;thememorydevice,mainlyreferstothepermanentMemoryisanimportantpartoftheinformationprocessingsystem,whichstoresalargenumberofsymbolstructurescomposedofvarioussymbolsconnectedinacertainrelationship,thatis,information;theeffectoristhepartoftheinformationprocessingsystemthatrespondstotheinformation,whichistheentiresystemThefinalstructurecontrolstheoutputofinformation.Thecorehereistheprocessor,whichalsoincludesthreefunctionalparts:
(1)Asetofbasicinformationprocesses,suchasmakinganddeletingsymbols,makingnewsymbolstructuresandcopyingandchangingexistingonesStructure,markingexternalstimuliwithsymbolsorsymbolstructure,distinguishingandcomparingsymbolstructure,anddeterminingresponseaccordingtosymbolstructure;(2)Short-termmemory,maintainingbasicinformationprocessinputandoutputSymbolicstructure;
(3)Thenarratorintegratesthebasicinformationprocessandshort-termmemorytodeterminethesequenceofthebasicinformationprocess.Thedescriptionandregulationofthesequenceofthebasicinformationprocessconstitutetheprogram,whichisthebehaviormechanismoftheinformationprocessingsystem."Informationprocessingpsychologyisthestudyofthetransmission,conversion,storage,andmechanismofinformationinaninformationprocessingsysteminordertouncovertheactivitiesinthe"blackbox"ofthemind.
Mainresearch
Informationprocessingpsychologyhasawiderangeofresearchcontents,mostofwhichfocusonfouraspects:perceptualprocessingandpatternrecognition,attention,memory,andproblemsolving.
PerceptualProcessing
PerceptionItistheprocessofdeterminingthemeaningofastimulus,includingtheorientation,selection,organizationandinterpretationofthestimulus.Informationprocessingpsychologyfocusesonthemeticulousstudyoftheperceptualprocess,therebyrevealingitsinternalinformationprocessingmethods.Theseprocessingmethodsmainlyinclude:
(1)Data-drivenprocessingandconcept-drivenprocessing(Lindary&Norman,1977).Theformerreferstoprocessingstartingfromstimulus,alsocalledbottom-upprocessing;thelatterreferstothesubject’sperceptionofperceptionTheprocessingthatbeginswiththegeneralknowledgeoftheobjectisalsocalledtop-downprocessing.
(2)Seriesprocessingandparallelprocessing.Theformerreferstoaprocessingmethodinwhichinformationprocessingiscarriedoutstepbystepinadeterminedorder,andthelatterItreferstotheprocessingmethodinwhichmultiplestimulusinformationcanbeprocessedsimultaneouslyindifferentinformationprocessingunits.
(3)Overallprocessingandlocalprocessing,theformerreferstotheoverallcharacteristicsofthestimulus,thelatterItreferstothelocalcharacteristicsofperceptualstimuli,andinmanycases,perceptualprocessingshowsobviousoverallsuperiorityeffects,suchaswordsuperiorityeffects,objectsuperiorityeffects,etc.
Patternrecognition
Thefocusofthefieldofperceptionresearchispatternrecognition.Patternrecognitionreferstotheidentificationandclassificationofexternalstimuli.Anotherreasonwhycognitivepsychologistsareinterestedinpatternrecognitionisthattheywantcomputerstosimulatethisabilityofpeople.,Tomakeitmoreintelligent.Therearethreerepresentativetheoriesininformationprocessingpsychologythatcanexplaintheprinciplesofinformationrecognition,understandinganddecipheringtoacertainextent.
Templatematchingmodel
Thetemplatematchingmodel(templatematchingmodel)isproposedbasedontherecognitionmodeofthemachine.Itscentralideaisthatthehumanmemorysystemstoresvariousstimulitemplates.WheninputThestimulusinformationmatchesacertainstoredtemplate,andthestimulusinformationcanbedecipheredandrecognized.However,itisobviouslynotenoughtoexplainthehighlyflexiblecognitiveprocessofhumanbeingsonlywiththismechanicalandstrictlycorrespondingrigidtemplatematching.Itisalsouneconomical.
Prototypematchingmodel
Theprototypematchingmodelimprovesthetemplatematchingmodel.ItisbelievedthatpeoplemaystoredifferentExternalstimulusstrictlycorrespondstothetemplate,butageneralizedrepresentationofatypeofstimulus,thatis,prototype,sotheprototypeisacomprehensiveandabstractproduct.Aftertheexternalstimulusinformationisinput,theinformationprocessingsystemrecognizesitaccordingtothedegreeofmatchingbetweentheinputinformationandtheprototypeInformation,thestimulusinformationisgenerallyrecognizedashavingthemostGoodmatchingprototype,andgiveitacertainmeaning,sothatitcanbeunderstood.Obviously,theprototypematchingmodelencompassesthetemplatematchingmodel,andatthesametimeovercomestheshortcomingsofthetemplatematchingmodelofinflexibilityanduneconomical,andisalsosupportedbymorelifeexperienceandpsychologicalexperiments.However,thepatternrecognitiontheorythatcanbetterreflectthemeaningofsymbolprocessingisstillafeatureanalysismodel.
Featureanalysismodel
Thefeatureanalysismodelbelievesthatafterthesubjectreceivestheinputinformation,itwillfirstperformfeatureanalysisonit,andthencomparetheresultsoftheanalysiswiththelengthThefeaturetablesofvariousthingsintimememoryarecompared,andrecognitionisobtainedoncethebestmatchisobtained.Thefeatureanalysismodelismoreflexibleandeconomicalinexplaininghumanpatternrecognition,anditalsohasahigherabstraction.Buttherearestillsomequestions,suchashowalargenumberoffeaturetablesarestored,andwillitincreasetheburdenofmemory.
Thestudyofattention
Thepsychologicalmechanismofattentionisoneoftheearliestexperimentalsubjectsinmoderncognitivepsychology.Mainlythebinaurallisteningtechnology.SinceBroadbentfirstproposedtheearlyselectionmodel,cognitivepsychologistshavesuccessivelyputforwarddifferentviewpointssuchasthemid-termselectionmodel,thelateselectionmodelandtheresourcerestrictiontheory.
Filtermodel
Inthebinaurallisteningexperiment,Terryfoundthattheinformationofthesubject’sfollowingearswaspaidattentionto,soitcouldbefurtherprocessed.Non-followerinformationisnotnoticedandcannotbeprocessed.BritishpsychologistBrudbentputforwardtheattentionfiltertheorybasedonfurtherexperiments(Broadbent,1958).Thetheorybelievesthattheprocessingcapacityofthehumancerebralcortexisverylimited.Inordertohaveabottleneckfilterintheinformationinputchannel,theinformationpassingthroughthefilterisfurtherprocessedtobeidentifiedandstored,andtherestoftheinformationisblockedattheadvancedlevel.Outsidethehub.Thiskindoffilterworksaccordingtothe"allornothing"principle,onechannelpassestheinformationandtheotherchannelsareclosedatthesametime.Therefore,thistheoryisalsocalledthesingle-channeltheory,whichbelievesthatthefilterislocatedbeforeperception,soinformationselectionoccursintheearlystageofinformationprocessing,soitiscalledtheearlyselectionmodel.
Attenuatormodel
Thereisevidencethatsomeparticularlystrongstimuliorinformationthatisparticularlymeaningfultotheinformationreceiver,evenifitisnotenteredfromthefollowerTheprocessingsystemcanalsobeidentified,soAmTreisman(AmTreisman)proposedanattenuatormodel.Theattenuatormodelbelievesthatthefilteringdeviceintheinformationchannelworksaccordingtotheattenuationmode,buttheactivationthresholdofparticularlymeaningfulitems(suchasyourownname)islow,soitcanbeactivatedandidentified.Thisfilteringdeviceiscalledthehubfilter.Itseemsthatthechoiceofattentionnotonlydependsonthecharacteristicsofsensoryinformation,butalsodependsontheroleofthecentralfilter,soitisalsocalledthemid-termselectionmodel.
Thelateselectionmodel
Thelateselectionmodelbelievesthatselectiveattentionoccursinthelatestageofinformationprocessing,andthefilteringdeviceislocatedbetweenperceptualprocessingandworkingmemory.Themodelassumesthatallinputinformationreacheslong-termmemoryandactivatesrelateditemsinit,andthencompetesfortheprocessingofworkingmemory.Informationwithhighperceptualintensityorsignificantmeaningisfurtherprocessedandthenreacted.Thismodelcanexplainthephenomenonofattentionallocationwell,becausealltheinputinformationhasbeenprocessed;itcanalsoexplainthatparticularlymeaningfulinformationiseasytoattractpeople’sattention,becausetheactivationthresholdoftheseitemsstoredinlong-termmemoryisveryhigh.low.However,thismodelisnoteconomical,anditassumesthatallinputinformationisprocessedbythehub.
Completelydifferentfromtheabove-mentionedfiltertheoryisthetheoryofmentalintelligencelimitationproposedbyD.Kahneman.Thistheoryregardsattentionasthesourceofmentalintelligence,andthetotalamountofhumanmentalintelligenceislimited.Ifataskdoesnotexhaustallthesourcesofintelligence,thenitcanbedirectedtoanothertask.Whenapersonisfacedwithmorethanonetask,thepersonmustallocatethesourceofmentalintelligence,andtheamountofmentalintelligenceoccupiedbyeachtaskwillberelativelyreduced,andtheactivityefficiencywillbecorrespondinglyreduced.Thistheorycanwellexplainthephenomenonofattentiondistributionandrelatedexperimentalresults,butitcannotpredicthowmanypeople’smentalsourcesofintelligenceareandhowtheyaredistributed.Soitisstillnotcompletelysatisfactory.
Accordingtothepointofviewofinformationprocessing,memoryistheprocessofinformationinput,encoding,storageandextraction.Itcanmorefullyreflecttheworkflowofinformationprocessingsystems,soitisalsothecoreofinformationprocessingpsychologyresearch.Oneofthecontents.
Researchonmemory
Accordingtotheviewpointofinformationprocessing,memoryistheprocessofinput,coding,storageandextractionofinformation,whichcanmorefullyreflecttheworkflowoftheinformationprocessingsystem.Soitisalsooneofthecorecontentsofinformationprocessingpsychologyresearch.Theresearchoninformationprocessingofmemoryfocusesonthreeaspects:thestructure,representationandcapacityofmemory,mainlythestructureandinformationrepresentationofmemory.
Researchonmemorystructure
Researchonmemorystructurestartsfromtwokindsofmemorytheories.Asearlyas1890,AmericanpsychologistJamesproposedtheconceptsofprimarymemoryandsecondarymemory.Theformerreferstoshort-termmemoryandthelatterreferstolong-termmemory.AtkinsonandShifrey(1968)proposedathree-levelinformationprocessingmodelformemory.Themodelbelievesthatthememorystructureisfixed,whilethecontrolprocessisvariable.Memoryconsistsofthreestoragesystems:sensorymemory,short-termmemoryandlong-termmemory.Intheprocessofinformationprocessing,theexternalinformationfirstentersthesensorymemorythroughthesensoryorgans,wheretheinformationiskeptforaveryshorttime,onlyabout1second,andthenthenoticedinformationisrecognizedandenteredintotheshort-termmemory.Short-termmemoryisabufferforinformationprocessing.Theinformationinitisatthecenterofconsciousactivity.However,thecapacityhereislimited,only7±2informationblockscanberetained,andtheinformationretentiontimeisonlyabout1minute,unlessKeeprepeatingtheinformation.Retellingcanmakepartoftheinformationintheshort-termmemoryenterthelong-termmemory.Thecapacityofthelong-termmemoryisverylarge,andtheretentiontimeoftheinformationcanalsobeverylong.Itisourinformationdatabase,andthelargeamountofknowledgeandexperiencewehaveaccumulatedarestoredhere..Theinformationinthelong-termmemorycanbereturnedtothestateofconsciousnessundertheactionoftheactivationsignalfortheapplicationofthecognitivesystem.
Therearealsodifferentopinionsonthemulti-storagetheoryofmemory,andtherepresentativeoneisthetheoryofCrick’sprocessinglevel.Theprocessingleveltheorybelievesthatthemulti-storagestructuredoesnotexist,andthelengthoftheinformationretentiontimeisnotduetothedifferenceinthesystem,butsimplybecauseithasbeenprocessedatdifferentlevels.Informationprocessingleavestracesofmemory,somemoryisaby-productofinformationprocessing.
Researchontherepresentationofmemoryinformation
Therepresentationofmemoryinformationismainlytheinformationrepresentationoflong-termmemory,andtheinformationoflong-termmemoryisalsocalledknowledge,Thatis,personalknowledge.Peopledivideknowledgeintotwocategories:declarativeknowledgeandproceduralknowledge.Therepresentationmethodsofdeclarativeknowledgearebothsituationalandsemantic.Amongthem,therepresentationtheoryofsemanticmemoryinformationmainlyincludestwotypes:networkmodelandfeatureanalysismodel.Theformerbelievesthatthehumanbrain’smemoryofsemanticsisstoredhierarchicallyintheformofanetwork.Allconceptsaredividedintoseverallevelsaccordingtothelogicalsubordinaterelationship,andtheconceptsofeachlevelareconnectedinturntoformahierarchicalnetwork.Thefeatureoftheconceptisattachedtoeachnodeofthenetwork;thelatterconsidersthattherepresentationoftheconceptdependsonthefeatureset.Anyconceptincludesadefiningfeaturesetandadescriptivefeatureset.Themorethefeaturesofthetwoconceptscross,theoverlapoftheconceptsThemore,theclosertherelationship.
Investigationonproblem-solvendum
Inthemid-1950s,twopublicationsmarkedthebeginningofresearchonproblem-solvingbyinformationprocessingtheory.OneisBruner,Goodnow,andAustin(1956)"ThinkingResearch",whichmainlystudiesthecognitiveprocessofclassifyingstimulusinformation;theotherisaresearchpaperpublishedbyNewell,ShawandSimon,whichtheydescribeinthispaperAnengineeringresearchsituation.Thegoalofthisengineeringresearchistodevelopprogramsforadigitalcomputersothatitcansolveproblems,andinfact,thisistherealbeginningofartificialintelligence.Thefirststepoftheworkistocollectasmuchinformationaboutpeopleaspossibletosolveproblems,andbasedontheanalysisofthedata,tocompileaproblem-solvingprogramsimilartothatusedbypeople;thesecondstepistocollectdetailedinformationaboutpeopleandcomputerswhentheysolvethesameproblem.Onthepremiseofanalysisandcomparison,thecomputerprogramismodifiedtoprovideacomputeroperationmodeclosertohumanbehavior.Oncethesimulationinaparticulartaskissuccessful,theresearchercantrytousethesameinformationprocessingseriesandproceduresinawiderrangeoftasks.Theproblem-solvingtheorygraduallyformedinthesestudieswaslaterfullyexpressedintheirlongwork"HumanProblemSolving"(Newell&Simon,1972,pp.787-868).Thismonographcanbecalledaclassicworkonproblem-solvingresearchinmoderncognitivepsychology.NewellandSimonbelievethatproblemsolvingisaprocessofcontinuousheuristicsearchintheproblemspace.
Anothermajorresearchonproblemsolvingisresearchbyexpertsandnovices.Therearetwomaindifferencesbetweenexpertsandnovicesinproblemsolving:problemrepresentationanddifferenttypesofproblemsolvingmethods.Expertsarebasedonknowledgeinaspecificfield,andtheyhavericherproblemrepresentationsthannovices.Theserepresentationscanbeseenasschemascontainingsub-schemes.Forexperts,theinformationcontainedintheschemaisofteninformationorganizedaccordingtobasiclaws,ratherthanthesurfaceinformationfromtheproblem.Expertsaremorelikelytousetheinformationgivenbytheproblemtousethepositivepushstrategytoadvancetotheproblemtargetstate.Novicesmayproposehypothesesbasedonpossiblesolutionsandthentestthehypotheses.Thisstrategyhaslessperformance.Throughthepracticeofapplyingstrategies,expertswillautomatevariousrecalledandeasy-to-executeoperationsduringtheforwardpushprocess(VanLehn,1989).Throughsuchautomationandschematization,expertswillshifttheburdenofworkingmemorytonoresources.Thelimitedlong-termmemorycanfurtherimprovetheperformanceandaccuracyofproblemsolving.Butnovicesmustusetheirworkingmemorytosearchforprobleminformationandfindmultiplealternativestrategies.Hehasnoworkingmemoryspacetomonitortheirproblem-solvingprogressandaccuracy.
Atpresent,theresearchonproblemsolvingininformationprocessingpsychologyisstillveryimmature.Thisfieldalsoneedstoaccumulatemoredata,ormoreeffectiveresearchmethodsareneeded.
ResearchMethod
TheabilityofinformationprocessingpsychologytoemergeintheAmericanpsychologyworlddominatedbybehaviorismiscloselyrelatedtoitsbreakthroughinmethodology.Specifically,themostrepresentativemethodsofinformationprocessingpsychologyinclude:factor-basedexperimentalmethods,eyemovementresearchmethods,oralreportanalysismethods,andcomputersimulationmethods.
Factor-basedexperimentalmethodis
Theexperimentalmethodsofinformationprocessingpsychology,especiallytheexperimentswithreactiontimeandhomeworkscoresasindicators,arepaidmoreattention.Comparedwithotherbranchesofpsychology,thereactiontimemethodisitsmosteffectiveandtypicalexperimentalmethod,whichmainlyincludesthesubtractionreactiontimemethodandtheadditivefactormethod.
Thesubtractivereactiontimemethodreferstowhentwoinformationprocessingserieshavetherelationshipofinclusionandinclusion,thatis,inadditiontoalltheprocessesoftheotherinformationprocessingseries,thereisalsoauniqueInformationprocessingstageorprocess,thetimedifferencebetweenthesetwoprocessingseriesisthetimerequiredforthisuniqueinformationprocessingstageorprocess.Justasthediscriminativereactionincludesalltheprocessingstagesofthesimplereaction,andatthesametimeitalsohasamentalprocessingstageofsignaldiscriminationthatisnotavailableinthesimplereaction,thenthetimerequiredforthediscriminativementalprocessingcanbeobtainedbysubtractingthereactiontime.Obviously,themorecomplexthementalprocessing,thelongertheprocessingtimerequired.
Butthismethodalsohasitsweaknesses.Theuseofthisexperimentrequirestheexperimentertohaveapreciseunderstandingoftheseriesofpsychologicalprocessesbetweenthestimulusandtheresponsecausedbytheexperimentaltask,andrequirestwosubtractivetasks.Itisgenerallydifficulttomatchsharedpsychologicalprocessesstrictly.Theseweaknessesgreatlylimitthewidespreaduseofthesubtractivereactiontimemethod.
Inthemiddleofthe20thcentury,Sternbergputforwardtherulesofaddition,calledadditivefactorsmethods.Thismethodisthedevelopmentandextensionofthesubtractivereactiontimemethod.Thelogicoftheadditivefactorreactiontimeexperimentis:iftheeffectsoftwofactorsaremutuallyrestricted,thatis,theeffectofonefactorcanchangetheeffectofanotherfactor,thenthetwofactorsonlyactonthesameinformationprocessingstage;Theeffectsofeachfactorareindependent,thatis,theycanbeaddedtogether,thenthesetwofactorseachactondifferentinformationprocessingstages.Inthisway,throughunivariateandmultivariateexperiments,thevariousstagesoftheinformationprocessingprocessaredeterminedfromthetimechangeofthecompletionofthejob.Therefore,theimportantthingisnottodistinguishtheprocessingtimeofeachstage,buttoidentifyeachprocessingstageandprocessingsequence,andconfirmtheexistenceofdifferentprocessingstages.Theadditivefactormethodassumesthatwhentwofactorsaffecttwodifferentprocessingstages,theywillhaveindependenteffectsonthetotalreactiontime.Thatis,regardlessofthelevelofonefactor,theotherfactorwillhaveaconstanteffectonthereactiontime.Thisistheso-calledadditiveeffectofthetwofactors.Thebasicmethodofadditivefactormethodistoexplorefactorswithadditiveeffecttodistinguishdifferentprocessingstages.
EyeMovementResearchMethods
Eye-movementiseyemovement.Ithasthreebasicmethods:fixation,saccadesandpursuitMovement)(YanGuoli,2004,pp.5-13).Thesethreeeyemovementmethodsareoftenintertwined.Thepurposeistoselectinformationandimagethestimulitobenoticedinthefoveaareatoformaclearimage.Eyemovementcanreflecttheselectionmodeofvisualinformation,whichisofgreatsignificanceforrevealingtheinternalmechanismofinformationprocessing.Theeyemovementexperimentreliesonaneyetracker,whichisaninstrumentthatrecordseyemovementinformation.
Accordingtoresearchreportspublishedinrecentyears,thecommonlyuseddataorparametersforpsychologicalresearchusingeyetrackersmainlyinclude:
(1) Eyetrackingdiagram;
(2) Eyemovementtime;
(3) Eyemovementdirection anddistance;
(4) Pupilsizeandblinking.
Theresearchofeyemovementpsychologyisanascendantfield,anditstechnicalmethods,researchideasandsubjectareasinvolvedarestillintheprocessofrapiddevelopment.
Oralreportanalysismethod
Speakingreportanalysismethod(protocolanalysis),alsoknownasthink-loud,isamethodwhereparticipantsreportthemselvesaloudTheideawhenperformingacertainoperationtoexplorethemethodofinternalcognitiveprocess.Spokenreportsaremostlycarriedoutduringtheoperation,andcanalsobenarratedthroughrecallaftertheoperation.Intheoralreportexperiment,thesubjectsarerequiredtoreportthedetailedcontentoftheirthinkingduringtheoperationaloudandtruthfully,sothattheinternalthinkingprocessisexternallyverbal,buttheyarenotrequiredtoexplainthesituationorthinkingprocess.Whatthesubjectsreportedshouldmainlybetheinformationretainedinshort-termmemorythatwillsoondisappear.Usingthismethod,Naxen,Simonandothershavemadecertainachievementsincognitiveresearch.Theoralreportanalysismethodhasbeenacceptedandadoptedbymanyinformationprocessingpsychologists.
Computersimulationmethod
Computersimulation(computersimulation)isthemostrepresentativeresearchmethodofinformationprocessingpsychology,whichistounderstandpeoplethroughcomputersimulationofmentalprocessTheprocessofmentalactivityitselfisalogicalanalysisoftheprocessofhumaninternalinformationprocessing.Computersimulationisusuallycombinedwiththeoreticalanalysis,startingfromthreeaspects:programreduction,processanalysis,andprogramsimulation.