The usefulness of the Stroop effect during sorting movements
-Aging differences in microslips and Stroop interference
rate-KazushiKitayama,KatsuyukiShibata*
KeiaiHospital * SchoolofSciencesCollegeofMedical,PharmaceuticalandhealthSciences,KanazawaUniversity Introduction Weareexposedtovarioustypesofinformationfrom ourperipheralenvironment,andweactwhileselecting thenecessaryinformation.However,itisdifficultto process information while simultaneously selecting numerouspiecesofinformation1-2）.Executivefunctionhas theabilitytoplan,inhibit,andcontrolattentionandactions3）. Inrecentyears,ithasbeenindicatedthatexecutive functiondisorderscontributetoelderlypeople’sfalls andthedeclineinactivitiesofdailyliving（ADL）4-5）．It issuggestedthatevaluatingelderlypeople’sexecutive functionishelpfulinevaluatingtheirbasicandapplied movements.Amongtheexecutivefunctiontests,the WisconsinCardSortingTest（WCST）isusedinclinical settings,butifitcanbeadministeredrelativelyeasily asStrooptasks,itmaybegenerallyusefulasaclinical evaluationtoolfortheelderly.Hasheretal.6）usedthe WCSTasarestraintfunctionthatcausedtheStroop effect.AcharacteristicoftheStroopeffectisthatthe output suppresses aresponse toeasy stimulus, and theoutputisataskthatrespondstodifficultstimulus. Additionally,theStroopeffecthasastrongcorrelationto Thisstudyaimstoverifythedifferencesintaskexecutionaccordingtooralandmanual classification,thecorrelationbetweenrequiredtimeandthenumberofmicroslips,andthe differenceintheStroopinterferencerateamongyoungpeopleandmiddle-agedpeople. Therewere16targetsinExperiment1（21to55yearsofage）;andinExperiment2,30 targetsintheyounggroup（ages20to39）and30inthemiddle-agedgroup（ages40to59）, allofwhomwerenormal,healthyadults.Weusedfourvarietiesof16classificationcards, andclassifiedeach.InExperiment1,wecalculatedtherequiredtimefortheoralandmanual tasksandtheStroopinterferencerate;andinExperiment2,wecalculatedtherequiredtime duringthemanualtasks,theStroopinterferencerate,andthenumberofmicroslips,and comparedtheoralandmanualtasks,aswellastheyounggroupandthemiddle-agedgroup. TheStroopinterferenceratewassignificantlyhigherduringtheoraltasksthanduringthe manualtasks（p<.05）.Comparingtheyounggrouptothemiddle-agedgroup,themiddle-agedgrouphadasignificantlyhigherStroopinterferencerateinbothoralandmanualtasks thantheyounggroup（p<.05）.Themiddle-agedgroup’snumberofmicroslipsincreased significantlycomparedtotheyounggroup（p<.05）.Asforthecorrelationbetweenthe requiredtimesandthenumberofmicroslips,ahighcorrelationwasshowninthe4varieties ofclassificationcards（p<.05）.Inthisstudy,theStroopinterferencerateonchoicereaction washigherinoraltasksthaninmanualtasks.Furthermore,oralreactiontimeswereslower thanmanualreactiontimesinbothsimpleandchoicereactiontasks.Itwasindicatedthatthe numberofmicroslipsincreaseswiththedelayintime,andacorrelationbetweenthenumber ofmicroslipsandtimewasshown. Abstract Strooptask,Stroopinterferencerate,Microslip,Manualtask,Aging Key WordstheMiniMentalStateExaminationdementiaevaluation7）, and the decreasein restraintfunctionis caused by cognitivefunctiondecline8-9）．WiththeStroopeffect, ratherthanthenamingofcolors,thereadingofthe letters/wordsisfast10）．Generally,aStrooptaskhasoral responses,andtherequiredtimeshowstherestraint reactionuntiltheresponse.Thistypeofrestraintreaction occursnotonlyinoralbutalsoinmanualresponses.In aclinicalsetting,makingasubjectiveevaluationwhile observingapatient’sbehavioriscommon.Itissuggested that quantitatively evaluating non-fluent movement isimportantininvestigatinghindrancesinbehavioral actions. Thisstudyfocusedontheusefulnessoftheevaluation oftheStroopeffectduringsortingmovements.Inour dailyactions,themicroslipphenomenonoftenoccurs,such ashesitatingbehavior,thetrajectoryofourmovements, andchangesintheshapeofourhands11）.Criteriaof themicroslipsis"Hesitations","TrajectoryChanges", "HandShapeChanges",and"TouchesandTake/Gives". IntheStrooptaskinthisstudy,thedegreeofexecutive functionisnotlimitedtorequiredtime.Inthisway,itis possibletodemonstratemovementpatternsasamicroslip phenomenon.
The goals of this research were to verify 1）the differencesbetweentheoralclassificationandmanual classification2）,thecorrelationbetweenthetimerequired for a task and the number of microslips, and 3） the differencesbetweentheyounggroupandthemiddle-aged groupinamanualStrooptask.
Methods
1. Experiment 1: The differences between oral classification and manual classification
1）Subjects Thesubjectswere16normal,healthyadults（aged35.5 ±10.1）.Wereceivedwrittenconsentfromthesubjects, explainingthepurposeofourresearchbothverballyand inwriting.Ourresearchwascarriedoutwiththeapproval oftheMedicalEthicsCommitteeofKanazawaUniversity （approvalnumber:353）. 2）ClassificationCards Theclassificationcardswerecomposedofthefollowing fourvarieties:16black-inkwordcards（red:4;blue:4; yellow:4;green:4）,16coloredcolorcards,32incongruent color-wordcards（Figure1）.Theclassificationcards werearrangedonawhiteboardinfour-cardcolumnsand four-cardrows.Thecardswere6×6squarecentimeter. 3）ExperimentSetting Thesubjects’measuredposturewasseated（Figure2）. Handmovementsduringtheexperimentwererecorded withavideocamera（Canon:iVIHR1）. 4）Implementation Themethodsinvolvedorallysorting（oralsort）and manuallysorting（manualsort）theclassificationcards. Theoralsortwas1）WordCards:WordReading,2）Color Cards:ColorNaming,3）Color-WordCards:ColorNaming, and4）Word-ColorCards:WordReading.Themanual sortwas1）WordCard:SortbyWord,2）ColorCards: SortbyColor,3）Color-WordCards:SortbyColor,and4） Word-ColorCards:SortbyWord.Theinstructionswere tocallthecolororwordofthecards,orputtheminthe classificationboxesbyhand. 5）AnalyzingMethod Thetimerequiredfortaskexecutionwasmeasured byastopwatchfromthevideorecordingofthesubjects’ vocalizations and movements. The measurement wasperformedbytwooccupationaltherapists,who determinedtheaveragerequiredtime.
6）StatisticalProcessing
TheStroopinterferencerate12）wascalculatedaccording to the calculation methods of Hakoda et al（Stroop interferencerate=（（3）requiredtimeofthecolor-word cards－（2）requiredtimeofcolorcards）/（2）required timeofthecolorcards ×100）.Thecomparisonof manualsortingandoralsortingwastestedbyatwo-way analysisofvariance.Additionally,toclarifythepresence orabsenceofinteractionbetweenthetwotasks,asa secondarytest,weperformedmultiplecomparisons, makingaBonferroniadjustmenttoapairedt-test.For thecomparisonoftheyounggroupandthemiddle-aged group,anunpairedt-testwasperformed.Thesignificance levelswerealllessthan5%.
2. Experiment 2: Differences among tasks and age group in manual sorting
1）Targets Therewere30targetsintheyounggroup（ages20 to39）and30targetsinthemiddle-agedgroup（ages40 to59）;allsubjectswerenormal,healthyadults,andall wereright-handed.Theaverageageintheyounggroup was30.2 ±4.6,andtheaverageageinthemiddle-aged groupwas49.3±6.0.Wereceivedwrittenconsentfrom
thesubjects,explainingthepurposeofourresearchboth verballyandinwriting.Ourresearchwascarriedout withtheapprovaloftheMedicalEthicsCommitteeof KanazawaUniversity（approvalnumber:353）.
2）ClassificationCards
The classification cards are the same as used in Experiment1. 3）ExperimentSettingandMethods Theexperimentsettingandmethodsarethesameasin Experiment1（Figure2）. 4）AnalysisMethods Theevaluationitemsmeasuredtherequiredtimeand thenumberofmicroslips.Themicroslipcalculationwas performedbytwooccupationaltherapistsandtheinter-rateragreementwasdeterminedbythekappacoefficient. 5）StatisticalProcessing Thecomparisonofmanualsortingandoralsortingwas testedbyatwo-wayanalysisofvariance.Toclarifythe presenceorabsenceofinteractionbetweenthetwotasks, asasecondarytest,weperformedmultiplecomparisons, makingaBonferroniadjustmenttoapairedt-test.For thecomparisonoftheyounggroupandthemiddle-aged group,anunpairedt-testwasperformed.Thesignificance levelswerealllessthan5%.Forthecorrelationcoefficient betweenthetimeandthenumberofmicroslips,Spearman’ srankcorrelationwascalculated.TheStroopinterference rate12）wascalculatedaccordingtothecalculationmethods ofHakodaetal.12）. FromthecalculatedStroopinterferencerate,apaired t-testwasperformedcomparingmanualsortingand oralsortingtasks,withaBonferroniadjustment.Again, forthecomparisonoftheyounggroupandthemiddle-aged group, an unpaired t-test was performed. The significancelevelswerealllessthan5%.Thestatistical analysissoftwareSPSS（Version20）wasusedduringthe statisticalprocessing.
Figure 1. Sorting Card
The classification cards are composed of the following four varieties: 16 black-ink word cards（red: 4; blue: 4; yellow: 4; green: 4）, 16 colored color cards, 32 incongruent color-word cards.
Figure 2. The setting of the experiment design
Results
1. Comparison of oral sorting and manual sorting
（Experiment1） Theaverageofrequiredtimefororalsortingwas7.7 ± 0.5secondsforwordcards,9.9 ± 0.6secondsforcolor cards,15.3 ± 1.4secondsforcolor-wordcards,and10.0 ± 0.6secondsforword-colorcards（Table1）.Inthe comparisonoftimeamongtasks,therequiredtimefor color-wordcardswassignificantlyslowerthantherest ofthecards（p<.05）.Theaverageofrequiredtimefor manualsortingwas21.0 ± 3.3secondsforwordcards, 19.2 ± 2.0secondsforcolorcards,25.7 ± 6.5seconds forcolor-wordcards,and19.7 ± 4.1secondsforword-colorcards（Table1）.However,aswiththeoralsorting, therewasasignificantdelay（p<.05）withcolor-word cardscomparedtotheothercardsinmanualsorting.The Stroopinterferenceratewassignificantlyhigherinthe oralsortingtasks（54.9%）comparedtothemanualsorting tasks（34.3%）（p<.05）（Figure3）.
2. Comparison of required time and the number of microslips in manual sorting （Experiment2）
Weconfirmedthattherewasasignificanttimedelay withcolor-wordcardscomparedtotheothercards（p< .05）. Therewere1.4microslipsforthewordcards,1.5for thecolorcards,6.5forthecolor-wordcards,and2.8for theword-colorcards.Thenumberofmicroslipsincreased significantlywiththecolor-wordcards,comparedtothe othercards（p<.05）（Figure4）. Ahighcorrelationcoefficientbetweentherequiredtime andthenumberofmicroslipswasconfirmedwith（r= .72）forwordcards,（r=.65）forcolor-wordcards,and（r =.49）forword-colorcards（p<.05）.Thecolorcards（r =.39）hadalowcorrelation（p<.05）（Figure5）.
3. Comparison of the young group to the middle-aged group in manual sorting（Experiment2）
Therewasnosignificantdifferencebetweenthegroups intherequiredtimeforwordcardsandcolorcards. However,therequiredtimeforcolor-wordcardsamong themiddle-agedgroup（28.7seconds）wassignificantly slowercomparedtotheyounggroup（23.1seconds）（p< .05）.Intheword-colorcardsasignificantdifferencewas confirmedbetweenthemiddle-agedgroup（21.4seconds） andtheyounggroup（19.2seconds）,withasignificant delayinthemiddle-agedgroup（p<.05）. Therewasnosignificantdifferencebetweenthegroups forthenumberofmicroslipsinthewordcardsandcolor cards.However,asignificantincreaseinthenumberof microslipsinthecolor-wordcardswasconfirmedforthe middle-agedgroup（7.6times）,comparedtotheyoung group（5.4times）（p<.05）.Asignificantdifferencewas alsoconfirmedintheword-colorcards,withthenumber ofmicroslipsincreasinginthemiddle-agedgroup（3.2 times）comparedtotheyounggroup（2.4times）（p<.05） （Figure6）. TheStroopinterferencerateinmanualsortingwas 27.9%fortheyounggroupand47.7%inthemiddle-aged group,withtheStroopinterferenceratesignificantly
Figure 3. Comparison of the oral task and the manual task （Stroop interference rate）
The Stroop interference rate was significantly higher in the oral sorting tasks（54.9%）compared to the manual sorting tasks （34.3%）.
Figure 4. Comparison between each the tasks （Number of microslips）
The average number of microslips increased significantly with the color-word cards, compared to the other cards.
Table 1. The average of required time for oral sorting and manual sorting oralsorting manualsorting wordcards 7.7 ± 0.5 21.0± 3.3 colorcards 9.9 ± 0.6 19.2 ± 2.0 color-wordcards 15.3 ± 1.4 25.7 ± 6.5 word-colorcards 10.0 ± 0.6 19.7 ± 4.1 （seconds） In the comparison of time among tasks, the required time for color-word cards was significantly slower than the rest of the cards （p < .05）. As with the oral sorting, there was a significant delay （p < .05） with color-word cards compared to the other cards in manual sorting.
higher（p<.05）inthemiddle-agedgroup,andwiththe middle-agedgroupexperiencingwordinterferencemuch morestronglythantheyounggroup（Figure7）. Consideration InExperiment1,thedifferencesinoralsortingand manualsortingwereexamined.InExperiment2,the differencesinthepresenceoftheStroopeffectinmanual sortingbetweenthetwoagegroupswereexamined.
1. Reaction differences in oral sorting and manual sorting
Toevaluatetheinhibitoryreactioncausedbythe
Figure 5. The correlation coefficient between the required time and the number of microslips
A high correlation coefficient between the required time and the number of microslips was confirmed with （r = .72） for word cards, （r = .65） for color-word cards, and （r = .49） for word-color cards （p < .05）. The color cards （r = .39） had a low correlation （p < .05）.
Figure 6. Comparison between the young group and the middle-aged group（Number of microslips）
The average number of microslips increasing in the middle-aged group compared to the young group.
Figure 7. Comparison of the young group and middle-aged group by the manual task（Stroop interference rate）
The Stroop interference rate in manual sorting was 27.9% for the young group and 47.7% in the middle-aged group, with the Stroop interference rate significantly higher in the middle-aged group, and with the middle-aged group experiencing word interference much more strongly than the young group.
Stroopeffect,therequiredtimeofthetaskswasdivided bytheStroopinterferencerate.TheStroopinterference ratewassignificantlyhigherinoralsortingthaninmanual sorting.Oyama13）statesthatoralreactionsareslowerthan manualreactions,suchaspushingbuttons.TheStroop interference-ratedatafromthisstudyconfirmsthatoral hasahigherinterferenceratevaluethanbuttonpushing, indicatingthatoralresponsesareslowerthanmovement responses.Thatis,wehaveinterpretedthatStroop interferencerateofmovementresponsesarelowerthan oral.
2. The Relationship between Required Time and the Number of Microslips
Thisstudyalsoconfirmedthedifferencebetween requiredtimeandmicroslipsinsortingtasks.Ahigh correlationbetweenrequiredtimeandmicroslipswas confirmed in some task. Suzuki et al. suggest that, regardlessofthedifficultyoftheconditions,thenumber ofmicroslipsincreaseswiththetimerequiredtocomplete thetask14）.Inotherwords,inadditiontothedifference inrequiredtime,theincreaseordecreaseinthenumber ofmicroslipswillbeoneindicatorofinhibitoryfunction evaluations.Suzuki14）pointsouttherelationshipbetween theenvironmentandtheagent/actorintheoccurrenceof microslips,andthatthefrequencyofmicroslipsincreases whenfactorsunnecessaryforaccomplishingthetaskare added.Weconsideredalowcorrelationthatrelationshipof thesubjectsandcolorwassmall.
3. The difference in movement reactions in the young group and middle-aged group
Thereisatendencyfortheinhibitionofunnecessary informationtowaneaspeopleage15）.Thedifficulty withappropriatelyselectingspecifiedinformationwas demonstratedbythedifferenceintimebetweenthe younggroupandthemiddle-agedgroup,withthetaskof distinguishingbetweenthecolor-wordcardsandword-colorcards. Suzukietal.pointoutthatinmovementselection,the strengthorweaknessofthetaskinfluencestheoccurrence ofmicroslips14）.Visualinformationwhensortingtheword cardsandcolorcardsmaybeunderstoodasasinglepiece ofinformationthatdoesnotaffectjudgment.However, formixedcolorandwordinformation,whenoneattempts tochooseacolor,thewordinformationinhibitsthecolor informationandthedelayinjudgmentduringsorting manifestsasahesitationinhandmovement,andincreases theoccurrenceofmicroslips.Thus,itispossiblethatthe declineofinhibitoryfunctionswithage8）hasaneffecton theoccurrenceofmicroslips. TheStroopinterferencerateishighearlyinlife,and withthedevelopmentoftheinhibitoryfunctionsfrom around7to10yearsofagetoaround17yearsofage,it tendstodecrease16-19）.Afterthat,theStroopinterference ratestabilizesduringadolescence16-17）.SasakiandHakoda20） completedStrooptaskswithelderlypeopleandchildren, andshowedthattheelderlysubjectshadahigherStroop interferencerate.Asforthedelayedreactiontimethat comeswithaging,thereismarkedlymoredelayinchoice reactionsthaninsimplereactions13）.Inotherwords theresultsofthisresearch,whichshowahigherStroop interferencerateinthemiddle-agedgroupthantheyoung group,showedthattheStroopinterferencerateincreases withage.Additionally,theStroopinterferenceratemay becomeanindicatorfortheevaluationofinhibitory functionswithaging. Thesignificanceofthisstudyisthepossibilityofthe Strooptaskwithage.Elderlyiseasytoacceptthesimple thingsthancomplexthings.Elderlyiseasytorefuse whenintroducingnewthings.Webelievethatasimple assessmentscaleisrequiredintheinhibitoryfunction evaluationfortheelderly.
Limitations and Future Research
Thisresearchislimitedinthattherewereonly30 youngpeopleand30middle-agedpeopleasthetargets, andsoitispossiblethattheresearchresultsarebiased andcannotbeconsideredasdatarepresentinganentire agegroup.Infutureresearch,itwillbenecessaryto collectmoredataonrequiredtimefororalsortingandfor manualsorting,andtoverifythecorrelationbetweenthe Stroopeffectinoraltasksandmanualtasks. Acknowledgements
I would like to express my sincere gratitude to my supervisor, Professor Shibata for providing this opportunityasaPh.D.studentinhislaboratory.Iamvery gratefultoKeiaihospitalstaffandDr.Doyoushitafortheir valuablecooperationinmyexperiments.Finally,Iwould liketoextendmyindebtednessformywife’sendless love,understanding,support,encouragementandsacrifice throughoutmystudy.
References 1） DamasioH,GrabowskiT,&TranelD,etal:Aneuralbasis forlexicalretrieval.Nature380:490-505,1996 2） TempiniM,PriceC,&JosephsO,etal:Theneural systemssustainingfaceandproper-nameprocessing.Brain 121:2103-2118,1998 3） Yogev-SeligmannG,HausdorffM,GiladiN:Therole ofexecutivefunctionandattentioningait.Movement Disorders23:329-342,2008 4） StroopJR:Studyofinterferenceinserialverbalreactions. JournalofExperimentalPsychology18:643-662,1935 5） Cahn-WeinerA,MalloyF,BoyleA,etal:Prediction offunctionalstatusfromneuropsychologicaltestsin community-dwellingelderlyindividuals.TheClinical Neuropsychologist,14:187-195,2000 6） HasherL,TonevST,LustigC,etal:Inhibitorycontrol, environmentalsupport,andself-initiatedprocessingin aging.EssaysinHonourofFergusCraik.EastSussex,UK, PsychologyPress:286-297,2001 7） UchiyamaH,QuoY,KameyamaKetal:Basicstudyof automaticdementiawithstroopeffect.BulletinofNagaoka UniversityofTechnology24:75-80,2002 8） HasherL,StoltzfusER,ZacksTR,etal:Ageandinhibition. JournalofExperimentalPsychology:Learning,Memory, andCognition,17:163-169,1991 9） HasherL,ZacksTR:Workingmemory,comprehension andaging.PsychologyofLearningandMotivation22:193-325,1988 10） MacLeodCM:Halfacenturyofresearchonthestroop effect.AnIntegrativeReview.PsychologicalBulletin109: 163-203,1991 11） ReedES,PalmerCF,SchoenherrD:Onthenatureand significanceofmicroslipsineverydayactivities.Journalof EcologicalPsychology4（1）:51-66,2009 12） HakodaY,SasakiM:Groupversionofthestroopand reverse-strooptest:Theeffectsofreactionmode,orderand practice.JapaneseJournalofEducationalPsychology38（4）: 389-394,1990 13） OyamaT:Historicalbackgroundandthepresentstatusof reactiontimestudies.Ergonomics21（2）:57-64,1985 14） Suzuki K., Mishima H, Sasaki M: Affordances and
variabilityofactions:Environment,actionsandmicroslips. JournalofJapanSocietyforFuzzyTheoryandSystems9 （6）:826-837,1997 15） GazzaleyA,CooneyJW,McEvoyK,etal:Top-down enhancementandsuppressionofthemagnitudeandspeed ofneuralactivity.JournalofCognitiveNeuroscience17（3）: 507-517,2005 16） RandG,WapnerS,McFarlamdJH,etal:Agedifferences inperformanceonthestroopcolor-wordtest.Journalof Personality31（4）:534-558,1963 17） SchadlerM,ThissenDM:Thedevelopmentofautomatic wordrecognitionandreadingskill.Memory&Cognition9 （2）:132-141,1981 18） CarrionLJ,OrzaGJ,SantamariaPJ:Developmentofthe inhibitorycomponentoftheexecutivefunctionsinchildren andadolescents.InternationalJournalofNeuroscience114 （10）:1291-1311,2004 19） PeruA,FaccioliC,TassinariG:Stroopeffectsfrom3to 10years:thecriticalroleofreadingacquisition.Archives ItaliennesdeBiologie144（1）:45-62,2006 20） SasakiM,HakodaY:Developmentandagingofstroop andreversestroopinterference.AnnualConventionofthe JapaneseAssociationofEducationalPsychology（27）:208-209,1985
