ResearchArticlePathologicalVideo-GameUseAmongYouthAges8to18ANationalSt - PDF document

ResearchArticlePathologicalVideo-GameUseAmongYouthAges8to18ANationalStudyDouglasGentileIowaStateUniversityandNationalInstituteonMediaandtheFamily,Minneapolis,MinnesotaABSTRACT—Researchershavestudiedwhethersomeyouthare‘‘addicted’’tovideogames,butpreviousstudieshavebeenbasedonregionalconveniencesamples.Usingana-tionalsample,thisstudygatheredinformationaboutvid-eo-gaminghabitsandparentalinvolvementingaming,todeterminethepercentageofyouthwhomeetclinical-stylecriteriaforpathologicalgaming.AHarrispollsurveyedarandomlyselectedsampleof1,178Americanyouthages8to18.About8%ofvideo-gameplayersinthissampleex-hibitedpathologicalpatternsofplay.Severalindicatorsdocumentedconvergentanddivergentvalidityofthere-sults:Pathologicalgamersspenttwiceasmuchtimeplayingasnonpathologicalgamersandreceivedpoorergradesinschool;pathologicalgamingalsoshowedco-morbiditywithattentionproblems.Pathologicalstatussignicantlypredictedpoorerschoolperformanceevenaftercontrollingforsex,age,andweeklyamountofvideo-gameplay.Theseresultsconrmthatpathologicalgamingcanbemeasuredreliably,thattheconstructdemonstratesvalidity,andthatitisnotsimplyisomorphicwithahighamountofplay.Manyparents(andtheoccasionalprofessional)haveremarkedthattheywereworriedabouttheirchildrenbeing‘‘addicted’’tovideogames.Isthissimplyhyperbole,meaningonly‘‘mychildplaysalot,andIdon’tunderstandwhy?’’Atrueaddiction,evenabehavioraladdiction,hastomeanmuchmorethanthatsomeonedoessomethingalot.Accordingtopsychiatricandpsychologicalexperts,ithastodamagemultiplelevelsoffunctioning,suchasfamily,social,school,occupational,andpsychologicalfunctioning.Cliniciansandsocialworkersinmultiplecountrieshavereceivedrequestsfortreatmentofvid-eo-gameaddiction.Inresponse,researchershaveconductedseveralstudiesanddeterminedthattheideacannotbedis-missed.WhencriteriasimilartothoseusedintheDiagnosticandStatisticalManualofMentalDisorders(DSM;e.g.,criteriasimilartothoseusedtodenepathologicalgambling)wereappliedtovideogamers,asubstantialnumberappearedtoex-hibitdamagedfunctioningonmultiplelevels.Itwasimpossible,however,totellhowrepresentativeofgamersthesestudieswere.ThepurposeofthepresentstudywastosurveyanationalpanelofAmericanchildrenages8to18todeterminetheirvideo-gaminghabits,theirparents’involvementingaming,andthepercentageofyouththatappeartomeetDSM-stylecriteriaforpathologicalgaming.Thepast15yearshaveseenarevolutionintheuseofdigitaltechnologies,withprocessingpower,accesstodigitaltechnol-ogies,andchildren’suseofsuchtechnologiesincreasingdra-matically.Theamountoftimechildrenandadolescentsspendwithvideogameshasbeenincreasingsteadily(Anderson, AddresscorrespondencetoDouglasGentile,DepartmentofPsy-chology,IowaStateUniversity,W112LagomarcinoHall,Ames,IA50011,e-mail:dgentile@iastate.edu. PSYCHOLOGICALSCIENCEVolume20—Number5Copyright2009AssociationforPsychologicalScience nitionssimilartotheDSMcriteriaforpathologicalgambling.Thisapproachappearstobeavalidstartingpointbecausepathologicalvideo-gameuseandpathologicalgamblingarebothassumedtobebehavioraladdictions(TejeiroSalguero&Ber-n,2002).Bothgamblingandvideogamesareformsofgames.Assuch,theyareinitiallyplayedasaformofenter-tainment,becausetheyarestimulatingandproducepositive(andsometimesnegative)emotions.Peoplegambleorplayvideogamesformanyreasons,includingtorelax,toexperiencecompetenceandautonomy,andtoescapefromdailyconcerns(Grifths,2003;Ryan,Rigby,&Przybylski,2006).Gamblingorgamingmayproducea‘‘ow’’state,inwhichtheplayerisfo-cused,mayloseasenseofplaceortime,hasasenseofcontrol,andndstheactivityintrinsicallyrewarding(Csikszentmihalyi,1990).Theactivityisnotpathologicalatrst.Butgamblingbecomespathologicalforsomepeoplewhenitbeginstoproduceseriousnegativelifeconsequences.Thelistofpathological-gamblingcriteriainthefourtheditionoftheDSM(DSM–IV)demonstratesthatanysinglesymptomisnotpathological.Havingzerotofourofthesymptomsiscon-sideredtobewithinthenormalrange,andaperson’sgamblingisconsideredpathologicalonlyafterithasresultedinproblemsinseveralareasofhisorherlife.Usingthisclinicalapproachtodeningpathologicalvideogamingappearsappropriateforinitialinvestigations,asitprovidesasomewhatcleandistinctionbetweenbeinghighlyengagedinabehavioranddoingitinsuchawayastoincurdamagetoseveralareasofone’slife.Althoughthereislittleresearchonwherethedividinglineis,beinghighlyengagedinabehaviorappearstobeboththeoreticallyandempiricallydistinctfrombeingaddicted(Charlton,2002).Althoughcasestudiesofpathologicalvideo-gameuseweredocumentedasearlyas1983,scienticstudiesrstbegantobereportedinthemid-1990s(Fisher,1994;Grifths&Dancaster,1995;Grifths&Hunt,1998).Mostofthepublishedstudiesofsupposedcomputer,Internet,andvideo-gameaddictionhavefocusedoneitherthereliabilityofvariousdenitionsofpatho-logicaluseortheconstructvalidityofpathologicaluse.Forexample,TejeiroSalgueroandBersaben(2002)createdanine-itemquestionnaireassessingvideo-gameuse,basingtheirquestionnaireonDSMcriteriaforpathologicalgamblingandsubstanceabuse.Theyreportedreasonablereliabilityandfactorstructureforthisquestionnaire,aswellassomeevidenceofitsconstructvalidity(i.e.,scoresindicatingpathologicalplaycor-relatedwithamountofvideo-gameplaying,self-perceptionsofplayingtoomuch,andameasureofpsychologicaldependenceondifferenttypesofdrugs).Thepresentstudyassessedvideo-gameusewithan11-itemscalebasedontheDSM–IVcriteriaforpathologicalgambling.ThestudyfollowedDSMdiagnosticcriteriaforotherdisordersinconsideringgamingtobepathologicalifthegamerexhibitedatleasthalf(6)ofthesymptoms.AlthoughthesymptomsweresimilartoDSM–IVcriteriaforpathologicalgambling,theyalsosharecorecharacteristicswithotherdenitionsofaddictions,suchasBrown’scorefacetsofaddiction(Brown,1991):salience(theactivitydominatestheperson’slife,eithercognitivelyorbehaviorally),euphoriaorrelief(theactivityprovidesa‘‘high’’orthereliefofunpleasantfeelings),(overtime,agreateramountofactivityisneededtoachievethesame‘‘high’’),withdrawalsymptoms(thepersonexperiencesunpleasantphysicaleffectsornegativeemotionswhenunabletoengageintheactivity),(theactivityleadstoconictwithotherpeople,work,obligations,ortheself),andrelapseandrein-statement(thepersoncontinuestheactivitydespiteattemptstoabstainfromit).Thereisstillnoagreementastowhetherpathologicalgamingisadiscreteproblem,andthepurposeofthisstudywasnottoresolvethatdebate,butrathertoprovidesomenewrelevantdataandtoexploreapproachestodeningpathologicalgaming.Thisistherststudytousealarge-scale,nationallyrepresentativesampleofyouthtostudythereliabilityofmeasuresofpatho-logicalvideogaming,thevalidityofthisconstruct,andtheprevalenceofpathologicalvideogaming.Inaddition,thissampleallowsustoprovidenationaldataabouttrendsinvideo-gameuse,parentalmonitoringofgaming,andchildren’splayingofMature-ratedgames.ParticipantsAnationalsampleof1,178U.S.residents,ages8through18,wassurveyedbyHarrisPolls.ThissamplewasastratiedrandomsampleofHarrisInteractive’son-linepanelandwasrecruitedthroughpassword-protectede-mailinvitationstoparticipateina20-minomnibussurvey.Thesamplesizeyieldedresultsaccurateto3%witha95%condenceinterval.Thesampleincluded588malesand590females,andapproximately100participantsofeachagefrom8through18(minimum98,for8-year-olds;maximum119,for16-year-olds).Allregionsofthecountrywererepresentedinthisstudy,asthesampleincluded253participantsintheEast,369intheSouth,289intheMidwest,and267intheWest.Theethnic-racialmakeupwas66%White,17%BlackorAfricanAmeri-can,3%AsianorPacicIslander,1%NativeAmerican,7%mixed,and2%other(4%declinedtoanswer).SixteenpercentreportedbeingofHispanicorigin.InstrumentsandparticipantsweretreatedinaccordancewiththecodeandstandardsoftheCouncilofAmericanSurveyResearchOrganizationsandthecodeoftheNationalCouncilofPublicPolls.ProcedureInterviewswereconductedbyHarrisInteractive,usingaself-administeredon-linequestionnaireviaHarris’sproprietary,Web-assistedinterviewingsoftware.Thesoftwarepermitson-linedataentrybytherespondents.Interviewsaveraged20minVolume20—Number5 DouglasGentile inlengthandwereconductedbetweenJanuary17andJanuary23,2007.MeasuresThesurveyincludedseveralscales,includingthepreviouslymentioned11-itempathological-gamingscalebasedontheDSM–IVcriteriaforpathologicalgambling.Becausethereisnoclearstandardforhowtomeasurepathologicalgamingorhowtoscoresymptomchecklistsofpathologicalgaming,participantswereallowedtorespond‘‘yes,’’‘‘no,’’or‘‘sometimes’’toeachsymptom.Inaddition,thesurveyincludedseveralitemsassessingchildren’svideo-gamehabits(adaptedfromtheGeneralMediaHabitsQuestionnaireandtheAdultInvolvementinMediaScale;Andersonetal.,2007;Gentile,Lynch,Linder,&Walsh,2004).Theseitemsmeasuredweeklyamountofvideo-gameplay,knowledgeofgameratings,householdrulesformediause,schoolperformance,attentiondifculties,involvementinphysicalghts,andphysicalhealth.DataWeightingDatawereweightedtoreectthedistributionofkeydemo-graphicvariables(age,gender,race-ethnicity,highestlevelofeducation,parents’education,urbanicity,andregion)inthegeneralpopulationof8-to18-year-oldsintheUnitedStates.Theweightswerecalculatedbasedonthe2006CurrentPopu-lationSurvey(U.S.CensusBureau,2006).Allreportedresultsincorporatethisweighting.RESULTSVideo-GameUseTheresultsofthesurveyindicatethatmost(88%)Americanyouthbetweenages8and18playvideogamesatleastocca-sionally.Theaverage(median)reportedfrequencyofplayingwasthreeorfourtimesaweek(Table1),withboysplayingmorefrequentlythangirls,0.98.Theaverageamountofplayingtimewas13.2hrperweek(13.1).However,therewasasizabledifferencebetweenboys’averageplayingtime(16.4hr/week,14.1)andgirls’averageplayingtime(9.2hr/week,(1034).001,0.57.Giventrendsfortelevisionusage,onemightexpectvideo-gameusagetoincreaseacrosselementaryschool,peakataboutmiddleschool,anddropoffacrosshighschool(Hustonetal.,1992).Thepatternofvideo-gameusageacrossagesissimilar,butnotidenticalto,thetypicalpatternwithtelevision.Thefrequencyofvideo-gameplayappearedtoberelativelysteadyfromages8to13,andtodecreasethereafter(seeFig.1).Thelinearandquadraticchangeswerebothstatisticallysignicant,(1,1166).001,and(1,1166)respectively.Theamountofvideo-gameplayappearedtoin-creaseatmiddle-schoolage(12–14),butdidnotdropconsis-tentlyforallolderages(seeFig.2).However,thedifferencesinamountofplaybetweenageswerenotstatisticallysignicant(althoughwhenageswerecombinedintothreegroups,theyweremarginallysignicant).Therefore,althoughadolescentsplayvideogameslessfrequentlyastheygrowolder,theyappeartoincreasetheirplayingtimepersession. TABLE1Respondents’FrequencyofPlayingVideoGamesandMeansofObtainingMature-RatedVideoGamesSexAgerangeOverallBoysGirls8–1112–1415–18Frequencyofvideo-gameplayAtleastonceaday2333122626175or6timesaweek132061614113or4timesaweek161714171912Onceortwiceaweek161518161616Acoupleoftimesamonth96111078Aboutonceamonth436454Lessthanonceamonth73125413Never123216919Obtainedamature-ratedvideogame...Asagift263516163134Withownmoney,andparentsknewaboutit2233972337Withparents’money,andparentsknewaboutit1318671617Withownmoney,andparentsdidnotknowaboutit471229Withparents’money,andparentsdidnotknowaboutit120102Percentagewhoownamature-ratedgame395420224156Allnumbersinthetablearepercentages.Volume20—Number5 Video-GameUse Overall,onlyabouthalfthehomesrepresentedinthesurveyhadrulesaboutvideogames.Forty-fourpercentofrespondentssaidtherewererulesaboutwhentheywereallowedtoplayvideogames,46%reportedhavingrulesabouthowlongtheywereallowedtoplay,and56%saidtheyhadrulesaboutthekindsofgamestheywereallowedtoplay.Becauseseveralstudieshavedemonstratedbothshort-andlong-termnegativeeffectsofplayingviolentvideogames,thesurveyaskedtheyouthhowtheyhadgottenM-rated(‘‘Mature’’)videogames,iftheyhad(seeTable1).AlargepercentageoftheyouthownedM-ratedgames:22%of8-to11-year-olds,41%of12-to14-year-olds,and56%of15-to18-year-olds(39%of15-and16-year-olds).BoysweremorethantwiceaslikelyasgirlstohaveobtainedM-ratedgames,whetherasagiftorthroughapurchaseusingtheirownortheirparents’money;7%ofboysadmittedthattheyhadpur-chasedsuchagamewiththeirownmoneywithouttheirparents’knowledge.PrevalenceofSymptomsofPathologicalVideo-GameUseAsTable2shows,mostofthesymptomsofpathologicalvideo-gameuseweredemonstratedbyonlyasmallpercentageofyouthgamers.Somesymptomsweremoretypicalthanothers,however.Skippinghouseholdchorestoplayvideogameswasthedisruptionmostoftenreported(33%oftheyouthresponded‘‘yes,’’andanadditional21%responded‘‘sometimes’’).Atleastonefthofrespondentssaidthattheyhadplayedtoescapefromproblems(25%responded‘‘yes’’),thattheyhadskippedtheirhomeworktoplay(23%),thatvideogameshadhighcognitivesalienceforthem(21%),andthattheyhaddonepoorly Weekly Amount of Video-Game Play (hr/week)891011121314151617188–1112–1415–18 Fig.2.Averageweeklyamountofvideo-gameplaybyage. 8–1112–1415–1889101112131415161718 Fig.1.Frequencyofvideo-gameplaybyage.Theresponsescalewasverballyanchoredasfollows:0lessthanonceamonthaboutonceamonthacoupleoftimesamonthonceortwicea3or4timesaweek5or6timesaweek,and7atleastonceadayVolume20—Number5 DouglasGentile onschoolworkoratestbecauseofplaying(20%).If‘‘some-times’’iscountedasa‘‘yes,’’then7ofthe11symptomswereendorsedbyatleast20%ofyouth.Theotherpotentiallyprob-lematicsymptomswereendorsedbyfarfeweryouthgamers,withtheleastlikelysymptombeingstealingvideogamesorstealingmoneytobuygames(2%‘‘yes’’andanadditional2%‘‘sometimes’’).Boysweremorelikelythangirlstoreporteachofthesymptoms,withtheexceptionthatmoregirlsthanboysreportedtryingtoreducetheirvideo-gameplay(Table2).PrevalenceofPathologicalVideo-GameUseAsnoted,anindividualwasconsideredtobeapathologicalgamerifheorsheexhibitedatleast6ofthe11criteriaonthesymptomchecklist.However,itwasunclearwhethera‘‘some-times’’responseshouldbeconsideredequivalenttoa‘‘yes’’response,equivalenttoa‘‘no’’response,orsomethinginbe-tween.Therefore,threeversionsofchecklistscoreswerecal-culated.VersionAconsidered‘‘sometimes’’tobeequivalenttoa‘‘yes.’’Thisapproachyieldedreasonablereliability,andthehighestprevalenceofpathologicalgaming(.77;19.8%ofgamers).VersionBwasthemostconservativeapproach,con-sidering‘‘sometimes’’tobeequivalenttoa‘‘no.’’Thisapproachalsoyieldedreasonablereliability,andthelowestprevalenceofpathologicalgaming(.74;7.9%ofgamers).VersionCconsidered‘‘sometimes’’tobeequivalenttohalfofa‘‘yes’’(yes1,sometimes.5,no0).ThisapproachyieldedthehighestreliabilityandaprevalenceveryclosetothatofVersionB(.78;8.5%ofgamers).VersionCwasadoptedasthe‘‘best’’ap-proach,asitwasconservativeinassessingprevalencewhilestillallowingparticipantswho‘‘sometimes’’experiencedsymptomstobeconsidered.Itshouldbenoted,however,thattheconstructvalidityanalysesreportedhereyieldedalmostidenticalresultswhentheothertwoversionswereused.Boysexhibitedagreaternumberofsymptoms(2.2)thangirls(1.7),.99.Notethattheaveragenumberofsymptomsreportedwasnothighforeithergroup.However,thepercentageofrespondentsexhibitingatleastsixofthesymptomswasverydifferentforboysandgirls;11.9%ofboysand2.9%ofgirlswereclassiedaspathologicalbythiscriterion,1,178)Therelationsbetweenthenumberofreportedsymptomsandtheprimarycontinuousdependentvariablesweretestedforlinearitybyvisualexaminationofscatterplotsandbyre-gressionanalyses.Ingeneral,thequadratictermsweresmall,butsignicantbecauseofthelargesamplesize.Therelationstendedtoappearlinearfromzerouptoaboutsixsymptoms,withtheslopechangingataboutthatpoint.Thisndingpro- TABLE2PercentageoftheSampleReportingEachSymptomofPathologicalVideo-GameUseTotalsample‘‘Yes’’responses‘‘Yes’’‘‘Sometimes’’BoysGirlsOvertime,haveyoubeenspendingmuchmoretimethinkingaboutplayingvideogames,learningaboutvideo-gameplaying,orplanningthenextopportunitytoplay?21192911Doyouneedtospendmoreandmoretimeand/ormoneyonvideogamesinordertofeelthesameamountofexcitement?89123Haveyoutriedtoplayvideogameslessoftenorforshorterperiodsoftime,butareunsuccessful?22223Doyoubecomerestlessorirritablewhenattemptingtocutdownorstopplayingvideo2621Haveyouplayedvideogamesasawayofescapingfromproblemsorbadfeelings?25202919Haveyoueverliedtofamilyorfriendsabouthowmuchyouplayvideogames?14101710Haveyoueverstolenavideogamefromastoreorafriend,orhaveyoueverstolenmoneyinordertobuyavideogame?2231Doyousometimesskiphouseholdchoresinordertospendmoretimeplayingvideo33214024Doyousometimesskipdoinghomeworkinordertospendmoretimeplayingvideo23192915Haveyoueverdonepoorlyonaschoolassignmentortestbecauseyouspenttoomuchtimeplayingvideogames?20122611Haveyoueverneededfriendsorfamilytogiveyouextramoneybecauseyouspenttoomuchmoneyonvideo-gameequipment,software,orgame/Internetfees?96134Foreachsymptom,chi-squaretestswereusedtocompareprevalenceamongboysandgirls.Overall,thenumberofsymptomsreportedwassigniÞcantlydifferent(.001)betweenboys(2.8)andgirls(1.3).Also,theprevalenceofpathologicalgaming(i.e.,displayingatleast6ofthe11symptoms)wassigniÞcantlydifferent(.001)betweenboys(12%)andgirls(3%).Volume20—Number5 Video-GameUse videdadditionaljusticationfortheuseofsixsymptomsasaConstructValidityofPathologicalVideo-GameUseAlthoughquestionnairespacewaslimited,severalitemswereincludedtomeasuretheconstructvalidityofpathologicalgaming.Theoretically,pathologicalgamers,comparedwithothergamers,shouldspendmoretimeplaying,playmorefre-quently,knowmoreaboutgames,havemorehealth-relatedproblemsthatcouldbeassociatedwithgameuse,andgetworsegradesinschool;theyshouldalsobemorelikelytofeel‘‘ad-dicted’’togames,tohavefriendstheythinkare‘‘addicted’’togames,andtohavevideo-gamesystemsinthebedroom.Dem-onstratingsuchassociationswouldindicateconvergentvalidity,andfailingtodemonstratethemwouldbeevidenceofalackofconstructvalidity.Similarly,someitemsonthesurveyprovidedtheopportunitytodemonstratedivergentvalidity.Astheorycurrentlystands,thereisnoreasontoassumethatpathologicalgamingwouldberelatedtothetypeofschoolachildattends,havingaTVinthebedroom(giventhegenerallyhighpreva-lence),usingtheInternettohelpwithhomework,orachild’sageorrace.Someofthesevariablesmaybefoundtoberelevantasmorestudiesareconducted,butthereiscurrentlynotheoreticalreasontoassumetheywouldbe.Tables3and4showhowpathologicalgamersandnon-pathologicalgamerscomparedonanumberofdimensions.Pathologicalgamershadbeenplayingformoreyears,playedmorefrequentlyandformoretime,knewmoreofthevideo-gameratingsymbols,receivedworsegradesinschool,weremorelikelytoreporthavingtroublepayingattentioninschool,weremorethantwiceaslikelytohavebeendiagnosedwithanattention-decitdisorder,hadmorehealthproblemsthatwerelikelytohavebeenexacerbatedbylonghoursofplayingvideogames(e.g.,handpainandwristpain),andweremorelikelytoreporthavingfelt‘‘addicted’’togamesandhavingfriendstheythoughtwere‘‘addicted’’togames.Pathologicalgamerswerealsosignicantlymorelikelytohavebeeninvolvedinphysicalghtsinthepastyear.Also,aspredicted,patho-logicalgamersweremorelikelytohaveavideo-gamesystemintheirbedrooms.Thetestsofdivergentvalidityshowedthat,aspredicted,pathologicaluseofvideogameswasnotsystematicallyrelatedtoage,frequencyofusingtheInternettodohomework,havingaTVinthebedroom,ortypeofschoolattended.Equalpercent-ages(within1%)ofpathologicalandnonpathologicalgamersattendedprivateorparochialschools,attendedpublicschools,orwerehome-schooled.Approximatelyequalpercentagesofpathologicalandnonpathologicalgamerscomefromeachrace,6.0,n.s. TABLE3ComparisonsofPathologicalandNonpathologicalGamers:ContinuousVariablesVariable95%CIMSDMSDMeannumberofyearsplayingvideogames5.53.26.63.20.34(0.12,0.56).98Meanfrequencyofplayingvideogames(0leastonceaday)4.02.36.31.11.28(1.16,1.40).99Meanweeklyamountofvideo-gameplay(hours)11.812.624.616.00.88(0.62,1.15).99Meannumberofvideo-gameratingsymbolsknown3.42.34.22.10.36(0.16,0.56).98Meangradesusuallyreceived6.11.54.80.76(0.49).99Frequencyoftroublepayingattentiontoclassesatschool)2.50.93.00.55(0.33).99Overallhealth(1notatallhealthyextremelyhealthy)3.10.73.00.60.15(0.34,0.04).95Frequencyofhandorngerpain(1everyday)2.71.23.01.30.24(0.01,0.47).94Frequencyofwristpain(1almosteveryday)2.71.23.11.40.31(0.06,0.56).98Frequencyofneckpain(1almosteveryday)2.91.43.21.40.21(0.00,0.43).91Frequencyofblurredvision(1almosteveryday)2.71.32.91.30.15(0.06,0.56).82Frequencyofheadaches(1almosteveryday)4.11.54.51.50.27(0.05,0.48).81Meanage13.13.013.32.70.07(0.13,0.27).66FrequencyofusingtheInternettodohomework(1almostalways)3.01.03.11.10.10(0.14,0.33).67Respondents’self-ratingsofhowmuchtheyareaffectedbyviolenceinthegamestheyplay,comparedwithotherstudentsofthesameage(1alotlessalotmore)1.71.01.91.10.19(0.05,0.43).93SigniÞcanceofthedifferencebetweenpathologicalandnonpathologicalgamerswasdeterminedbytest.ConÞdenceintervals(CIs)forthevalueswerecalculatedfollowingthemethodofBonett(2008),whichhasbeendemonstratedtoworkproperlywithunequalsamplesizesandunequalvariances.Volume20—Number5 DouglasGentile Thesetestsprovidesomeevidenceofconvergentanddivergentconstructvalidityforpathologicalvideo-gameuse,whendenedasdamagingfunctioninginseveralareas(e.g.,school,social,family,andpsychologicalfunctioning).Theamountoftimespentplayingisnotacriterionforpathologicalvideo-gameuse,justashowmuchonedrinksisnotacriterionforalcoholdependence.However,becausetheamountoftimespentplayingvideogamesisaconsistentnegativepredictorofschoolperformance(An-dersonetal.,2007;Gentileetal.,2004),andbecausepatholog-icalgamersspendtwiceasmuchtimeplayinggameseachweekasnonpathologicalgamersdo(Table3),itwouldbestrongerevidenceforpathologicalgamingiftheclassicationwereasignicantpredictorofschoolperformanceaftercontrollingforamountoftimespentplayingvideogames.Thispossibilitywastestedwithananalysisofcovariance.Evenaftercontrollingforsex,age,andweeklyamountofvideo-gameplaying,pathologicalstatuswasasignicantpredictorofschoolperformance,(1,1003)27.7,.001,.027,rep.99.Inanationalsampleofyouthages8to18,8.5%ofvideo-gameplayersexhibitedpathologicalpatternsofplayasdenedbyexhibitingatleast6outof11symptomsofdamagetofamily,social,school,orpsychologicalfunctioning.Althoughthisper-centagemayatrstappeartobehigh,itisverysimilartotheprevalencedemonstratedinmanyotherstudiesofpathologicalvideo-gameuseinthisagegroup,includingstudiesinothernations(e.g.,theprevalencerateof9.9%amongSpanishado-lescents,reportedbyTejeiroSalguero&Bersaben,2002).Thisdenitionshowedreasonablereliabilityandconstructva-lidity.Itsconvergentvaliditywasdemonstratedbythendingsthatpathologicalgamers(comparedwithnonpathologicalgamers)spenttwiceasmuchtimeplayinggames(24hr/week),weremorelikelytohavevideo-gamesystemsintheirbedrooms,reportedhavingmoretroublepayingattentionatschool,re-ceivedpoorergradesinschool,hadmorehealthproblems,andweremorelikelytofeel‘‘addicted.’’DivergentvaliditywasdemonstratedbythendingsthatpathologicalgamerswerenotmorelikelytohavetelevisionsintheirbedroomsortousetheInternetforhomework;moreover,pathologicalstatuswasnotrelatedsignicantlytoage,race,ortypeofschoolattended.Pathologicaluse(oraddiction)mustmeanmorethan‘‘doitalot.’’Aspredicted,pathologicalstatuswasasignicantpre-dictorofpoorerschoolperformanceevenaftercontrollingforsex,age,andweeklyamountofvideo-gameplay.Thisistherststudyofpathologicalcomputer,Internet,orvideo-gameusetoconductsuchastrongtest,andtheresultsdemonstratethatpathologicaluseisnotsimplyisomorphicwithexcessiveplayorhighengagementwithvideogames.Onecouldhypothesizethatthethird-personeffect(thebeliefthatoneislessaffectedbymediathanotherpeopleare)mightbestrongeramongpathologicalgamersthanamongnonpathologi-calgamersbecausetheyshouldhaveastrongerpositivefeelingaboutvideogames.Thethird-personeffectwasdemonstratedinthisstudy(76%ofgamersbelievedthemselvestobelessaffectedbyvideo-gameviolencethanotherstudentstheirage,whereasonly4%believedthemselvestobemoreaffected),butpathologicalstatuswasnotsignicantlyrelatedtoit.Thendingthatpathologicalgamersweretwiceaslikelyasnonpathologicalgamerstohavebeendiagnosedwithanatten-tionproblem,suchasattention-decitdisorderorattention-decit/hyperactivitydisorder,helpstodemonstratethecurrentlimitsofknowledgeaboutpathologicalgaming.Onecouldin-terpretthisndingasapredictabletypeofcomorbidity,giventhatmanyaddictionsarecomorbidwithotherproblemsandthatInternet‘‘addiction’’haspreviouslybeenfoundtobecorrelatedwithattentionproblems(Yooetal.,2004).Still,itisnotclearwhether(a)pathologicalplayisentirelydistinctfromotherpathologies,(b)pathologicalplayisacontributingfactortothedevelopmentofotherproblems,or(c)theotherproblemscon-tributetopathologicalplay. TABLE4ComparisonsofPathologicalandNonpathologicalGamers:DichotomousVariablesVariablegamers(%)gamers(%)ratio95%CIHasbeendiagnosedwithanattentionproblem(e.g.,attention-decitdisorderorattention-decit/hyperactivitydisorder)11.025.32.77(1.66,4.65).99Hasfelt‘‘addicted’’tovideogames21.165.47.00(4.34,11.28).99Hasfriendswhoare‘‘addicted’’tovideogames56.877.22.67(1.56,4.67).99Hasbeeninaphysicalghtinthepastyear12.324.12.28(1.36,3.82).99Hasavideo-gamesysteminthebedroom40.864.42.69(1.71,4.22).99HasaTVinthebedroom62.469.01.36(0.86,2.18).81SigniÞcanceofthedifferencebetweenpathologicalandnonpathologicalgamerswasdeterminedbychi-squaretest. Forastrongertestofconstructvalidity,pathologicalstatuswasrecalculateddroppingthetwoschool-relateditems(skippinghomework,doingpoorlyonanassignmentortest;veofnineitemswererequiredforpathologicalstatus),andtheanalysisofcovariancewasconductedagain.Aftercontrollingforsex,age,andweeklyamountofvideo-gameplaying,pathologicalstatus(denedwithoutschool-relatedsymptoms)wasagainasignicantpredictorofschoolperfor-(1,1019)Volume20—Number5 Video-GameUse Theprimarylimitationofthisstudyisitscorrelationalnature.Itdoesnotprovideevidenceforthepossiblecausalrelationsamongthevariablesstudied.Itiscertainlypossiblethatpathologicalgamingcausespoorschoolperformance,andsoforth,butitisequallylikelythatchildrenwhohavetroubleatschoolseektoplaygamestoexperiencefeelingsofmastery,orthatattentionproblemscausebothpoorschoolperformanceandanattractiontogames.Furthermore,thedataonpathologicalgamingshouldbeconsideredtobeexploratory,asnostandarddenitionofpathologicalgamingexists,andfurtherresearchmaydeterminethattherearebetterwaystodeneit.Arelatedlimitationisthenatureoftheresponsescale;forexample,itisunclearwhetherallyouthwouldinterpretthedifferencebe-tween‘‘yes’’and‘‘sometimes’’thesameway.However,thefactthattherelationsbetweenpathologicalstatusandtheothervariablesweresimilarregardlessofhowwedened‘‘some-times’’suggeststhatthislimitationmaynotbeamajorproblem.Nonetheless,inaclinicalsetting,thisapproachwouldserveonlyasascreenforadetailedclinicalinterviewinwhichtheactualseverityoftheproblemscouldbeassessed.Furthermore,ifthesequestionsweretobeusedinaclinicalsetting,theyshouldbetimeboundedthesamewaythatsubstance-useitemsare(e.g.,‘‘Inthepastyear,howmuchhaveyou...?’’).ThisstudyisalsolimitedbecauseitincludedonlyyouthwhohadaccesstotheInternet.However,asof2005,87%ofyouthages12through17wereusersoftheInternet(Hitlin&Rainie,2005).Althoughthisgroupasawholemaybemorelikelythanotherstoplayvideogames(becausetheyhavecomputeraccess),comparisonsweremadeonlyamonggameplayers.Thisstudy’sprimarystrengthisthatitisnationallyrepre-sentativewithin3%.Nonetheless,ityieldsfarmorequestionsthananswers.Wedonotknowwhoismostatriskfordevelopingpathologicalpatternsofplay,whatthetimecourseofdevelopingpathologicalpatternsis,howlongtheproblemspersist,whatpercentageofpathologicalgamersneedhelp,whattypesofhelpmightbemosteffective,orevenwhetherpathologicalvideo-gameuseisadistinctproblemorpartofabroaderspectrumofdisorders.Thepresentstudywasdesignedtodemonstratewhetherpathologicalgamingisanissuethatmeritsfurtherat-tention.Withalmost1outof10youthgamersdemonstratingreal-worldproblemsbecauseoftheirgaming,wecanconcludethatitis.Acknowledgments—ThesupportofSuzanneMartin,HarrisInteractive,CraigAnderson,andDouglasBonettisgratefullyacknowledged.Anderson,C.A.,Gentile,D.A.,&Buckley,K.(2007).Violentvideogameeffectsonchildrenandadolescents:Theory,research,andpublicpolicy.NewYork:OxfordUniversityPress.Bonett,D.G.(2008).Condenceintervalsforstandardizedlinearcontrastsofmeans.PsychologicalMethods,99–109.Brown,R.I.F.(1991).Gaming,gamblingandotheraddictiveplay.InJ.H.Kerr&M.J.Apter(Eds.),Adultplace:Areversaltheoryap-(pp.101–118).Amsterdam:Swets&Zeitlinger.Charlton,J.P.(2002).Afactor-analyticinvestigationofcomputer‘addiction’andengagement.BritishJournalofPsychologyChiu,S.-I.,Lee,J.-Z.,&Huang,D.-H.(2004).VideogameaddictioninchildrenandteenagersinTaiwan.CyberPsychology&Behavior,571–581.Csikszentmihalyi,M.(1990).Flow:Thepsychologyofoptimalexperi-.NewYork:HarperandRow.Fisher,S.(1994).IdentifyingvideogameaddictioninchildrenandAddictiveBehaviors,545–553.Gentile,D.A.,Lynch,P.J.,Linder,J.R.,&Walsh,D.A.(2004).Theeffectsofviolentvideogamehabitsonadolescenthostility,ag-gressivebehaviors,andschoolperformance.JournalofAdoles-,5–22.Grifths,M.(2000).Doesinternetandcomputer‘‘addiction’’exist?Somecasestudyevidence.CyberPsychology&Behavior,211–Grifths,M.(2003).Internetgambling:Issues,concerns,andrecom-CyberPsychology&Behavior,557–568.Grifths,M.D.,&Dancaster,I.(1995).TheeffectofTypeAperson-alityonphysiologicalarousalwhileplayingcomputergames.AddictiveBehaviors,543–548.Grifths,M.D.,&Hunt,N.(1998).DependenceoncomputergamesbyPsychologicalReports,475–480.Hitlin,P.,&Rainie,L.(2005).Teens,technology,andschool.RetrievedJune13,2007,fromhttp://www.pewtrusts.com/pdf/Huston,A.C.,Donnerstein,E.,Fairchild,H.,Feshbach,N.D.,Katz,P.A.,Murray,J.P.,etal.(1992).Bigworld,smallscreen:TheroleoftelevisioninAmericansociety.Lincoln:UniversityofNebraskaJohansson,A.,&Gotestam,K.G.(2004).Internetaddiction:Charac-teristicsofaquestionnaireandprevalenceinNorwegianyouth.ScandinavianJournalofPsychology,223–229.Kipnis,D.(1997).Ghosts,taxonomies,andsocialpsychology.canPsychologist,205–211.Nichols,L.A.,&Nicki,R.(2004).DevelopmentofapsychometricallysoundInternetaddictionscale:Apreliminarystep.PsychologyofAddictiveBehaviors,381–384.Ryan,R.M.,Rigby,C.S.,&Przybylski,A.(2006).Themotivationalpullofvideogames:Aself-determinationtheoryapproach.tivationandEmotion,347–363.Shaffer,H.J.,Hall,M.N.,&VanderBilt,J.(2000).‘‘Computerad-diction’’:Acriticalconsideration.AmericanJournalofOrtho-,162–168.Shaffer,H.J.,&Kidman,R.(2003).Shiftingperspectivesongamblingandaddiction.JournalofGamblingStudies,1–6.Shaffer,H.J.,LaPlante,D.A.,LaBrie,R.A.,Kidman,R.C.,Donato,A.N.,&Stanton,M.V.(2004).Towardasyndromemodelofad-diction:Multipleexpressions,commonetiology.HarvardReviewofPsychiatry,367–374.TejeiroSalguero,R.A.,&Bersaben,R.M.(2002).Measuringproblemvideogameplayinginadolescents.,1601–U.S.CensusBureau.(2006).Currentpopulationsurvey.Availablefromhttp://www.census.gov/cps/Volume20—Number5 DouglasGentile Yee,N.(2001).TheNorrathianscrolls:AstudyofEverQuest(Version.RetrievedMarch6,2009,fromhttp://www.nickyee.com/eqt/Yee,N.(2002).AriadneÑunderstandingMMORPGaddictionRetrievedMarch3,2009,fromhttp://www.nickyee.com/hub/Yoo,H.J.,Cho,S.C.,Ha,J.,Yune,S.K.,Kim,S.J.,Hwang,J.,etal.(2004).AttentiondecithyperactivitysymptomsandInternetPsychiatryandClinicalNeurosciences,487–494.Young,K.S.(1997).Internetaddiction:Theemergenceofanewclinicaldisorder.CyberPsychology&Behavior,237–244.4/24/08;REVISIONACCEPTEDVolume20—Number5 Video-GameUse DouglasGentile,‘‘PathologicalVideo-GameUseAmongYouthAges8to18:ANationalStudy,’’PsychologicalScience,Volume20,Number5,pp.594–602.Inthisarticle,Imadeanerrorinclaimingthatthestudywasnationallyrepresentativewithin3%.Becauseofthetypeofsamplethatwasused,itisimpossibletoputacondenceintervalaroundthe8.5%prevalencerateofpathologicalgam-ingreported.Thesamplewasfromapanelrecruitedthroughmultiplemeans,andthedatawereweightedbyage,gender,race-ethnicity,highestlevelofeducation,parents’education,urbanicity,andregioninordertomakethedatageneralizabletothetotalpopulation.Iincorrectlyassumedthatacon-denceintervalcouldbecalculatedsimilarlytohowitiscalcu-latedforarandomprobabilitysample(suchasthesampleobtainedinarandom-digittelephonesurvey).However,itisnotpossibletocalculatetheexactcondenceintervalforourdatabecausethetheoreticalsamplingframeisnotdetermined,althoughsomestudiessuggestthatthecondenceintervalwouldbesimilarto(andperhapsbetterthan)thecondenceintervalsobtainedwithrandomprobabilitysamples(Taylor,2007).Nevertheless,themainconclusionofthestudystillstands:Byplausibleclinicalstandards,anontrivialpercentageofyouthexhibitsignsofaddictivebehaviorwithregardtovideogames.Thatis,thestudyprovidesevidencefortheconvergentanddivergentvalidityofpathologicalvideo-gameuse,anddemonstratesthat8.5%oftheyouthwhoplaygameswouldbeclassiedaspathologicalgamersaccordingtoadenitionsimilartothatforpathologicalgambling.Furthermore,theprevalenceofpathologicalgamingfoundinthisstudyissimilartothatfoundinotherstudiesusingsimilardenitionsofpathologicalgaming.Eveniffuturestudiesusingdifferentap-proachestodenitionorsamplingndthatthepercentageislower,thiswouldnotdetractfromthecentralpointthatpatho-logicalvideo-gameuseisaproblemthatislargeenoughtowarrantconcernandfurtherstudy.Taylor,H.(2007).Thecaseforpublishing(some)onlinepolls.PollingReport,1–6. PSYCHOLOGICALSCIENCEVolume]]]Copyright2009AssociationforPsychologicalScience