Buckels Paul D Trapnell Delroy L Paulhus University of Manitoba Winnipeg Manitoba Canada University of Winnipeg Winnipeg Manitoba Canada University of British Columbia Vancouver British Columbia Canada article info Art ID: 10603
Download Pdf The PPT/PDF document "Trolls just want to have fun Erin E" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Trollsjustwanttohavefun ErinE.Buckels ,PaulD.Trapnell ,DelroyL.Paulhus UniversityofManitoba,Winnipeg,Manitoba,Canada UniversityofWinnipeg,Winnipeg,Manitoba,Canada UniversityofBritishColumbia,Vancouver,BritishColumbia,Canada articleinfoArticlehistory:Availableonlinexxxx DarkTetradDarkTriadAntisocialInternetbehavior Intwoonlinestudies(total 1.Introduction isthepracticeofbehavinginadeceptive,destructive,ordisruptivemannerinasocialsettingontheInternetwithnoapparentinstrumentalpurpose.Fromalay-perspective,InternettrollssharemanycharacteristicsoftheclassicJokervil- Despitepublicawarenessofthephenomenon,thereislittleempiricalresearchontrolling.Existingliteraturesarescatteredandmultidisciplinaryinnature(Hardaker,2010;Herring,Job-Sluder,&Scheckler,2002;McCosker,inpress;Shachaf&Hara,).Forinstance,ShachafandHara(2010)conductedinterviewsofWikipediatrolls,ndingthemesofboredom,attentionseeking,revenge,pleasure,andadesiretocausedamagetothecommunityamongtheirexpressedmotivationsfortrolling.Inotherresearch,Hardaker(2010)conductedacontentanalysisofUsenetpoststhat disruptiveaspectsmaydistinguishtrollingfromotherformsofon- lineantisociality,suchascyber-bullying,whereperpetratoridenti-tiesareusuallyclear(Lenhardt,2013)andtheintentismorestraightforward. Frequencyofactivityisanimportantcorrelateofantisocialusesoftechnology.Forinstance,cyber-bullyingisoftenperpetratedbyheavyInternetusers(Juvonen&Gross,2008),anddisagreeablepersonsusemobiletechnologiesmorethanothers notforsocial-izing,butforpersonalentertainment(Phillips&Butt,2006).Simi-larly,gamerswhoexpressnon-socialmotivationsforonlinegaming(e.g.,competition,personalachievement)demonstratelowerlevelsofagreeablenessandconscientiousnessthanothersGraham&Gosling,2013).Stillotherresearchhaslinkedlowagreeableness,lowconscientiousness,andhighextraversiontoheavyInternetuse( ThenoxiouspersonalityvariablesknownastheDarkTetradpersonality narcissism,Machiavellianism,psychopathy,andsadisticpersonality(Buckels,Jones,&Paulhus,2013;Furnham,Richards,&Paulhus,2013) areyettobeinvestigatedinthetroll-ingliterature.Theirrelevanceissuggestedbyresearchlinkingthesetraitstobullyinginbothadolescents(Fanti&Kimonis,)andadults(Baughman,Dearing,Giammarco,&Vernon, Correspondingauthor.Address:DepartmentofPsychology,UniversityofManitoba,190DysartRd.,Winnipeg,MB,R3T2N2,Canada.Tel.:+12049778445.E-mailaddress:(E.E.Buckels). PersonalityandIndividualDifferencesxxx(2014)xxx xxx ContentslistsavailableatScienceDirectPersonalityandIndividualDifferencesjournalhomepage:www.elsevier.c Pleasecitethisarticleinpressas:Buckels,E.E.,etal.Trollsjustwanttohavefun.PersonalityandIndividualDifferenceshttp://dx.doi.org/10.1016/j.paid.2014.01.016 2012;Jones&Paulhus,2010;Linton&Power,2013).Alsosugges-tiveisresearchshowingthatnarcissists(Ljepava,Orr,Locke,&Ross,2013)andthosewithantisocialpersonalitydisorder(Whaling,Rab,Carrier,&Cheever,2013)useFacebookmorefre-quentlythanothers,thusindicatingthatdarkpersonalitiesleavelargedigitalfootprints.OftheDarkTetrad,weexpectedeverydaysadism(Buckelsetal.,2013)toprovemostgermanetotrolling.Afterall,trollingcultureembracesaconceptvirtuallysynonymouswithsadisticpleasure:introll-speak,lulz. ToevaluatethepredictorsofInternettrolling,weadministeredquestionnairestostudentandcommunitysamplesintwoonlinestudies.InStudy1,wefocusedonpredictingenjoymentoftrolling,asopposedtootheronlinesocialactivities,suchasdebatingandchatting.WeexpectedthattheDarkTetradwouldbepositivelyassociatedwithatendencytoratetrollingasthemostfavoredactivityontroll-ablewebsites(denedhereaswebsitesthatper-mituserstointeractbypostingcomments).Wefurtherexpectedthatextraversionandagreeableness theBigFivedimensionsmostrelevanttotheDarkTetradssocialworld(Paulhus&Wil-liams,2002) wouldevidenceassociationswithtrolling,withapreferencefortrollingindicativeofextravertedbutdisagreeablepersonalitytraits. 2.Study1 2.1.Method2.1.1.Participantsandprocedure Werecruited418participants(42.4%female;age=29.2%,=11.0)fromAmazonsMechanicalTurkwebsite(www.mturk.com)tocompletesurveyquestionsonline.ThesamplewasrestrictedtorespondentsfromtheUnitedStates.Thekeyquestionsregardingtrollingandotheronlinebehaviorswereembeddedinalargerbatteryofpersonalityquestionnaires.Partic-ipantsreceivedmonetarycompensation($0.50)fortheirtime.2.1.2.Measures Twomeasuresofsadisticpersonalitywereadministered.FirstwastheShortSadisticImpulseScale(SSIS;OMeara,Davies,&Hammond,2011),containing10itemstoassessadispositionaltendencytoenjoyhurtingothers(e.g.,Hurtingpeopleisexciting=.88),ratedonve-pointscalesfrom1(stronglydisagree)to5stronglyagree).SecondwastheVarietiesofSadisticTendenciesscale(VAST;Paulhus&Jones,inpress),containingsixitemstoas-directsadism,(e.g.,Ienjoyhurtingpeople=.61)andsevenitemstoassessvicarioussadism(e.g.,InvideogamesIlikethereal-isticbloodspurts=.69),ratedonseven-pointscalesrangingfrom1(notatall)to7(verymuch).TheVASTdirectsadismsubscaleisconceptuallyequivalenttotheSSIS,andthescoreswerehighlycorrelatedinthissample(=.73,.001).Hence,westandardizedandsummedthemtocreateadirectsadismcompositescore. The27-itemShortDarkTriadscale(SD3;Jones&Paulhus,in)wasusedtoassessnarcissism(e.g.,Ihavebeencomparedtofamouspeople=.72),Machiavellianism(e.g.,Itsnotwisetotellyoursecrets=.80),andsubclinicalpsychopathy(e.g.,backneedstobequickandnasty=.79).The44-itemBigFiveInventory(John&Srivastava,1999)wasusedtoassessextraver-sion(=.87),agreeableness(=.79),conscientiousness(=.83),neuroticism(=.85),andopennesstoexperience(=.78) Finally,thesectiononInternetbehavioraskedparticipantstoestimatetheiroverallcommentingfrequency:Howmanyhoursperdaydoyouspendpostingcommentsonwebsitesnewssites,etc)?Asecondquestionprobedtheirpreferredactivitywhencommentingonline:Whatdoyouenjoydoingmostonthesecommentsites?withveresponseoptions:debatingis- suesthatareimportanttoyou,chattingwithotherusers, newfriends,trollingotherusers,andother(specify).Theorderoftherstfouransweroptionswasrandomized.Thosepartici-pantswhoindicatedthattheydidnotspendanytimepostingcom-mentswerelabeledasnon-commenters. 2.2.Results2.2.1.Onlinecommentingfrequency Acrossallparticipants,themeannumberofcommentinghoursperdaywas1.07,=1.77. Commentingtimewasassociatedwithlowerconscientiousnessscores,(418)=.001,andhigherscoresonallDarkTetradmeasuresexceptnarcissism:directsadism,(508)=.12,=.01,vicarioussadism,(508)=.21,.001,psychop-(512)=.12,=.005,andMachiavellianism,(512)=.16,.001;narcissism,(512)=.04,=.37. Commentingtimewasalsonegativelyassociatedwithage,(508)=.001,andmenreportedgreaternumbersofhourspostingcomments(=0.88,=0.78)thandidwomen(=0.49,=0.62),(505.91)=6.19,.001,=0.55.Incontrast,extraver-sion,agreeableness,neuroticism,andnarcissismwereallnon-sig-nicantpredictorsofcommentingfrequency,s-244;.500;.18.2.2.2.Favoredactivitywhencommenting Atotalof23.8%ofparticipantsexpressedapreferencefordebatingissues,21.3%preferredchatting,2.1%saidtheyespeciallyenjoymakingfriends,5.6%reportedenjoyingtrollingotherusers,and5.8%speciedanotheractivity.Theremaining41.3%ofpartic-ipantswerenon-commenters.Becauseoflowendorsementratesofthemakingfriendsoption,wecombined thatcategorywiththeothercategoryinthefollowinganalyses. AmultivariateanalysisontheDarkTetradrevealedasignicanteffectofactivitypreference:Wilks=0.97,(20,1646.00)=1.65,=.03.InspectionofthepatterndepictedinFig.1conrmedthat,asexpected,theDarkTetradscoreswerehighestamongthosewhoselectedtrollingasthemostenjoyableactivity.PlannedorthogonalcontrastsindicatedthattheeffectwassignicantforallDarkTetradmeasures:directsadism,(500)=3.03,=.003,=.27,vicarioussadism,(500)=2.91,=.004,=.26,psychopathy,(500)=3.09,=.002,=.28,narcissism,(500)=2.64,=.009,=.24,andMachiavellianism,(500)=2.78,=.006,=.25.Asec-ondmultivariateanalysisontheBigFivescoresindicatedthat,asexpected,participantswhochosetrollingastheirfavoriteactivitywerehigheronextraversion,(413)=2.02,=.04,=.20,andlow-eronagreeableness,(413)==.04,=.20,thanothers,butdidnotdifferonconscientiousness,neuroticism,oropenness,s.21. 3.Study2 AlimitationofStudy1isthatweaskedparticipantstoselecttheirfavoriteactivityfromalistofoptions.Thisnecessitatedacat-egoricalindexoftrollingthatlikelyunderestimatedtheeffects.HenceinStudy2,weassessedenjoymentofeachcommentingactivity(includingtrolling)onseparatecontinuousscales.ToruleoutthepossibilitythatoverallInternetuseexplainsrelationswithtrolling,wealsoincludedaquestionabouttotaltimespentonlineforuseasacontrolvariable.Finally,totriangulateontrollingwithmultiplemeasures,weconstructedasecondbriefindex:theAssessmentofInternetTrolling(GAIT)scale,whichassessedtrolling Becausethefrequencyscoreswerepositivelyskewed,weappliedasquareroottransformationtotherawscores,resultinginatransformedmeanof0.72hofcommentingperday,=0.75.Thetransformedscoreswereusedintheanalysesthatfollow.Thepatternofresultswasunchangedwhenallcategorieswereused. E.E.Buckelsetal./PersonalityandIndividualDifferencesxxx(2014)xxx xxx Pleasecitethisarticleinpressas:Buckels,E.E.,etal.Trollsjustwanttohavefun.PersonalityandIndividualDifferences(2014),http://dx.doi.org/10.1016/j.paid.2014.01.016 behavior,identication,andenjoyment.AsinStudy1,measuresoftheBigFivewereincludedforcomparison. Study2alsofeatureddatafromalargerandmorediversesam-ple,furnishinguswithenoughstatisticalpowertotesthypothesesabouttheuniquecontributionsoftheDarkTetrad.Forreasonsarticulatedearlier,weexpectedsadismtodominatepersonalityef-fectsontrolling.Thuswepredictedthattherelationsbetweensad-ismandtrollingwouldremainsignicantevenwhencontrollingforoverlapwithpsychopathy,narcissism,andMachiavellianism. 3.1.Method3.1.1.Participantsandprocedure Twolargesampleswerecollectedonline.Therstconsistedof188Canadianpsychologystudents(55%female;age=21.15,=3.63)whocompletedthequestionnairesforextracoursecred-itpoints.Thesecondconsistedof609UnitedStatesresidents(43%age=35.04,=12.98)recruitedonAmazonsMechan-icalTurkwebsite(MTurk;http://www.mturk.com).Thelatterpar-ticipantsreceived$0.50eachfortheirtime.Althoughallparticipantscompletedthedarkpersonalitymeasuresandtheshorttrollingscale,onlyasubsetofMTurkparticipants(=207;52%female;age=36.1,=13.4)completedthequestionnaireoncommentingbehavior,andonlythestudentsamplecompletedtheBigFiveInventory.3.1.2.Measures The44-itemBigFiveInventory(John&Srivastava,1999)wasusedtoassessextraversion(=.83),agreeableness(=.79),con-scientiousness(=.81),neuroticism(=.78),andopennesstoexperience(=.80). The27-itemShortDarkTriadscale(Jones&Paulhus,inpressassessednarcissism(=.75),Machiavellianism(=.80),andsub-clinicalpsychopathy(=.79).Toassesssadisticpersonality,weusedthe18-itemComprehensiveAssessmentofSadisticTenden-cies(CAST;Buckels&Paulhus,2013).TheCASTitemsetsucceedstheVASTusedinStudy1,boastingexpandedcontentcoverageandimprovedreliabilities.TheCASTassessesthreedistinctvarie-tiesofsadistictendencies:directphysicalsadism(e.g.,Ienjoyphysicallyhurtingpeople;veitems,=.80),directverbalsadism(e.g.,Ienjoymakingjokesattheexpenseofothers;sixitems,=.81),andvicarioussadism(e.g.,Ienjoyplayingthevillainingamesandtorturingothercharacters;sevenitems,=.81).Items wereratedon5-pointscalesrangingfrom1(stronglydisagree)to stronglyagree Wecomputedacompositefordirectsadismasthemeanofthe11physicalandverbalitems(=.86).Coefciental-phaforCASTtotalscoreswas.89. ThesectiononInternetuseaskedparticipants,HowmanyhoursperdaydoyouspendontheInternet?Participantsaddition-allyrespondedtoayes/noquestion,Doyoupostcommentsonwebsites(e.g.,YouTube,newssites,forums,etc.)?(evenoccasion-?).Thosewhoansweredyes(i.e.,thecommenters)wereaskedtoprovideadditionalinformationabouttheirpostingbehav-ior:Howmanyhoursperdaydoyouspendpostingcommentsonwebsites(e.g.,YouTube,newssites,forums,etc.)?WeregressedcommentinghoursonoverallInternethoursandsavedtheresidualscorestocreateasecondindexofcommentingfrequencywhencontrollingforoverallInternetuse.Finally,commentersratedtheirenjoymentofeachactivityusedinStudy1(debating,chatting,trolling,andmakingfriends)onscalesfrom1(notatallenjoyable)to7(veryenjoyable). Wealsoincludedfouritemsrelevanttotrollingthatwereinter-spersedintheothermeasures:Ihavesentpeopletoshockwebsitesforthelulz,Iliketotrollpeopleinforumsorthecommentssectionof,Ienjoygriengotherplayersinmultiplayergames,andThemorebeautifulandpureathingis,themoresatisfyingitistocor- ratedon5-pointscalesfrom1(stronglydisagree)to5(stronglyagree).Therstthreeitemsaddressedtrollingexperienceandenjoyingvariousformsoftrolling,whilethelastitemaddressedidenticationwithtrollingandInternetsubcultures.Meanresponsestothesefouritems(=.82)formedthecompositescorelabeled,GlobalAssessmentofInternetTrolling(GAIT). 3.2.Results Table1displayscorrelationsbetweentheDarkTetrad,com-mentingfrequency,andratedenjoymentofvariousactivities includingtrolling whilepostingcommentsonline.Asexpected,DarkTetradscoreswerepositivelycorrelatedwithcommentingfrequency.Partialcorrelations(alsodisplayedinTable1)indicatedthatcontrollingforInternetuse(=6.61,=3.14)didnotaffect Mean Personality Score (Standardized(Non-Commenter) Debating IssuesChattingTrolling OtherFavorite Activity when Posting Comments Online Machiavellianism Narcissism Direct Sadism Vicarious Sadism Fig.1.DarkTetradscoresasafunctionoffavoriteonlineactivityinStudy1.Errorbarsrepresentstandarderrors. ResponsestotheCASTarecommonlycollectedon7-pointscales;here5-pointratingscaleswereusedtostayconsistentwiththeSD3.Thisstatementisfromtheevolving(andoftentrolled)RulesoftheInternetE.E.Buckelsetal./PersonalityandIndividualDifferencesxxx(2014)xxx xxx Pleasecitethisarticleinpressas:Buckels,E.E.,etal.Trollsjustwanttohavefun.PersonalityandIndividualDifferenceshttp://dx.doi.org/10.1016/j.paid.2014.01.016 thoseassociations.Amongcommenters,commentingfrequency=0.82,=0.83)waspositivelycorrelatedwithoverallInternet(81)=.35,.001,butwaspositivelyassociatedwithtrollingenjoymentevenwhencontrollingforoverallInternetuse,(78)=.27,=.01. Alsoasexpected,sadism,psychopathy,andMachiavellianismscoreswerepositivelycorrelatedwithself-reportedenjoymentoftrolling,alls.37(seeTable1),evenwhencontrollingforoverallInternetuse,alls.39.Narcissism,incontrast,wasnotcorrelatedwithtrollingenjoyment,butwasinsteadpositivelycorrelatedwithenjoyingdebatingissuesimportanttothem.Vicarioussadismwassimilarlypositivelycorrelatedwithenjoymentofdebating,whileenjoymentofchattingwasnegativelycorrelatedwithpsychopathy. Turningnexttoscoresontheshorttrollingindex,GAITscoreswerestronglyassociatedwithcommenterstrollingenjoyment,(83)=.71,.001.Critically,GAITwaspositivelyassociatedwithscoresofallDarkTetradmeasures(seeTable1 andespeciallystronglywithsadism(s.55).Amongcommenters,GAITscoreswerepositivelycorrelatedwithcommentingfrequency,(80)=.43,.001,andremainedsowhencontrollingforoverallInternetuse,(78)=.47,.001.Acrossallparticipants,menhadstrongerGAITscores(=1.78,=0.92)thandidwomen(=1.34,=0.62),(530)=7.29,.001,=.63.ParticipantswithstrongerGAITscorestendedtobeyounger(.001),lowerinconscientiousness.001)andagreeableness(.001),andmar-ginallyhigherinopenness(=.13,=.08),ascomparedtotheirlowGAITcounterparts.Scoresonextraversionandneuroticismwereunrelated,s-244;.700;.50.3.2.1.UniquecontributionsoftheDarkTetradtotrolling ToexaminetheuniquecontributionsoftheDarkTetradontrollingenjoyment,weconductedamultipleregressionanalysiswithdatafromthesubsampleofcommenters.Totalsadism,Machiavellianism,narcissism,andpsychopathywereenteredaspredictorsoftrollingenjoyment.Thisanalysisindicatedthatsad-=0.53,(78)=4.21,=.002,andMachiavellianism,=0.23,(78)=2.23,=.03,wereuniquepredictorsoftrollingenjoyment.Incontrast,whencontrollingfortheotherDarkTetradscores,narcissismwasnegativelyassociatedwithtrollingenjoy-(78)==.001,andpsychopathywasunrelatedtotrollingenjoyment,=.89.ThepatternofassociationwasunaffectedbycontrollingforoverallInternetuse. WerananidenticalanalysisforscoresontheGAITscalewithdatafromthefullsample.Aswasthecasefortrollingenjoyment,sadismpredictedstrongerGAITscores,=0.61,(735)=15.41, .001,evenwhencontrollingforscoresontheDarkTriadmea- sures.Psychopathywasalsoaunique(thoughweaker)predictorofGAITscores,=0.10,(734)=2.43,.001.Machiavellianismandnarcissismwerenotsignicant,s-237;.600;.75.3.2.2.Mediationviatrollingenjoyment Becausetheassociationsbetweensadismandtrollingwerepar-ticularlystrong,weranamediationanalysistoexamineif,amongcommenters,ratedenjoymentoftrollingexplainedtherelation-shipbetweensadismandGAITscores. Inotherwords,wesoughttotestthehypothesisthatsadismleadstotrollingbehaviorsare.SignicancewastestedwithbothSobelstestandabootstrapped95%condenceintervalforthestandardizedindirecteffect(constructedwith10,000re-samplesandapercentiledistribution). Recallthat,amongcommenters,sadismscoreswerestronglyassociatedwithratedtrollingenjoyment.Aregressionanalysisindicatedthatenjoymentoftrollingwas,inturn,positivelyassoci-atedwithGAITscoreswhencontrollingforsadism,=0.48,(80)=6.31,.001.Themediatedeffectofsadismthroughenjoy-mentwassignicant,Sobels=4.10,.001;95%CI=[0.09,0.45].ItremainedsignicantevenwhencontrollingforscoresontheDarkTriad,Sobels=3.44,.001;95%CI=[0.10,0.51].Thedi-recteffectofsadismwassubstantiallyreduced,butremainedsig-nicantwhencontrollingfortrollingenjoyment,=0.40,(80)=5.07,.001,indicatingpartialmediation.Analternativemediationanalysisfoundnosupportfortheoppositecausaldirec-tion;thestandardizedindirecteffectofGAITonsadismviatrollingenjoymentwasnotsignicant,Sobels=0.74,=.46;95%CI=[0.09,0.25]. 4.Generaldiscussion ThepresentresearchwasthersttoexaminecomprehensivepersonalityprolesofInternettrolls.Acrosstwostudiesandtwomeasuresoftrolling,thepersonalityprojectionsoftrollsemergedinQuadrantIIoftheInterpersonalCircumplex(Wiggins,1995thatis,HighAgencyandLowCommunion(Jones&Paulhus,).Inotherwords,theydisplayedhighlevelsoftheDarkTetradtraitsandaBFIproleconsistentwiththosetraits.Itwassadism,however,thathadthemostrobustassociationswithtrollingofanyofthepersonalitymeasures,includingthoseoftheBigFive.Infact,theassociationsbetweensadismandGAITscoreswereso Table1AssociationsofDarkTetradvariableswithcommentingbehaviorandenjoymentinStudy2.PersonalityscaleCommentingfrequencyCommentingfrequency(controllingforoverallInternetuse)Ratedenjoyment:GAITscoresTrollingDebatingChattingMakingfriendsSadismTotal.43.21.07.12.68DirectSadism.35.17.65DirectPhysical.35.19.04.20.62DirectVerbal.32.15.56VicariousSadism.39.03.55Machiavellianism.33.14.34Narcissism.27.09.23.19.21.18Psychopathy.23.21.55.05;.01;.001.Alltestsaretwo-tailed.GAITscoreswereacompositeoffouritemscapturingtrollingbehavior,enjoyment,andidentity. Thepatternsofassociationswerecomparableacrossthestudentandcommunity ResultswereidenticalwhenomittingGAITitemsoverlappingwithenjoyment,leavingapuremeasureoftrollidentityandbehavior(2items;=.65).Theimpactofsadismontrollidentity/behaviorwasmediatedbyenjoyment:Sobels=2.90,=.004;95%CIforthestandardizedindirecteffect=[0.03,0.30].Thedirecteffectwasalsosignicant,95%CI=[0.28,0.61]withaBCacorrection.E.E.Buckelsetal./PersonalityandIndividualDifferencesxxx(2014)xxx xxx Pleasecitethisarticleinpressas:Buckels,E.E.,etal.Trollsjustwanttohavefun.PersonalityandIndividualDifferences(2014),http://dx.doi.org/10.1016/j.paid.2014.01.016 strongthatitmightbesaidthatonlinetrollsareprototypicaleverydaysadists(Buckelsetal.,2013).NotethattheDarkTetradassociationswerespecictotrolling.Enjoymentofotheronlineactivities,suchaschattinganddebating,wasunrelatedtosadism. SubsequentanalysesconrmedthattheDarkTetradassocia-tionswerelargelyduetooverlapwithsadism.Whentheiruniquecontributionswereassessedinamultipleregression,onlysadismpredictedtrollingonbothmeasures(trollingenjoymentandGAITscores).Incontrast,whencontrollingforsadismandtheotherDarkTetradmeasures,narcissismwasactuallynegativelyrelatedtotrollingenjoyment.GiventhatcontrollingforoverallInternetusedidnotaffecttheseresults,personalitydifferencesinbroaderten-denciesofInternetuseandfamiliaritycannotexplainthendings. InthenalanalysisofStudy2,wefoundclearevidencethatsa-diststendtotrollbecausetheyenjoyit.Whencontrollingforenjoyment,sadismsimpactontrollingwascutnearlyinhalf;andtheindirecteffectofsadismthroughenjoymentwassubstan-tial,signicant,andremainedsignicantwhencontrollingforoverlapwiththeDarkTriadscores.Thesendingsprovideapreli-minaryglimpseintothemechanismbywhichsadismfosterstroll-ingbehavior.Bothtrollsandsadistsfeelsadisticgleeatthedistressofothers.SadistsjustwanttohavefunandtheInternetistheir Study2alsofoundstrongpositiverelationsamongonlinecom-mentingfrequency,trollingenjoyment,andtrollingbehaviorandidentity,pointingtohighlevelsofconsistencyamongthemea-sures.Ourndingsaddtoaccumulatingevidencelinkingexcessivetechnologyusetoantisociality(Carr,2011;Juvonen&Gross,2008;Phillips&Butt,2006;Rosenetal.,2013).Thecausaldirectionoftheseassociationsisyetunclear.Doantisocialpersonsusetechnol-ogymorethanothersbecauseitfacilitatestheirnefariousgoals/motives?Thendingsofthisstudysuggestthatthisisindeedthecase,butmoreempiricalworkisneeded.Someauthorsgofur-thertoarguethatuseofInternettechnologyactuallyshiftsusersinanantisocialdirection(Carr,2011;Immordino-Yang,Christodou-lou,&Singh,2012;Suler,2004 TheInternetisananonymousenvironmentwhereitiseasytoseekoutandexploreonesniche,howeveridiosyncratic.Conse-quently,antisocialindividualshavegreateropportunitiestocon-nectwithsimilarothers,andtopursuetheirpersonalbrandofselfexpressionthantheydidbeforetheadventoftheInternet.Onlineidentityconstructionmaybeimportanttoexamineinre-searchontrolling,especiallyintermsofantisocialidentity(uszek&Hyland,2011;Walters,2007)anditsroleintrollingbehavior.Thetrollpersonaappearstobeamaliciouscaseofavir-tualavatar(Dunn&Guadagno,2012;McCreery,Krach,Schrader,&Boone,2012),reectingbothactualpersonality(Dunn&Guadag-no,2012;McCreeryetal.,2012)andonesidealself(Bessière,Seay,&Kiesler,2007).Ourresearchsuggeststhat,forthosewithsadisticpersonalities,thatidealselfmaybeavillainofchaosandmayhem theonlineTricksterwefear,envy,andlovetohate:thecyber- References Andreassen,C.S.,Grifths,M.D.,&Gjertsen,S.R.(2013).Therelationshipsbetweenbehavioraladdictionsandtheve-factormodelofpersonality.JournalofBehavioralAddictions,2,90 99. Baughman,H.M.,Dearing,S.,Giammarco,E.,&Vernon,P.A.(2012).Relationshipsbetweenbullyingbehavioursandthedarktriad:Astudywithadults.PersonalityandIndividualDifferences,52,571 575. Bessière,K.,Seay,A.F.,&Kiesler,S.(2007).Theidealelf:IdentityexplorationinWorldofWarcraft.CyberPsychologyandBehavior,10,530 535. Boduszek,D.,&Hyland,P.(2011).Thetheoreticalmodelofcriminalsocialidentity:Psycho-socialperspective.InternationalJournalofCriminologyandSociologicalTheory,4,604 615 Buckels,E.E.,Jones,D.N.,&Paulhus,D.L.(2013).Behavioralconrmationof everydaysadism.PsychologicalScience,24,2201 2209. Buckels,E.E.,&Paulhus,D.L.(2013).Comprehensiveassessmentofsadistictendencies(CAST).Unpublishedmeasure,UniversityofBritishColumbia. Carr,N.(2011).Theshallows:WhattheInternetisdoingtoourbrains.NewYork,NY:W.W.Norton&Company Dunn,R.A.,&Guadagno,R.E.(2012).Myavatarandme Genderandpersonalitypredictorsofavatar-selfdiscrepancy.ComputersinHumanBehavior,28,97 106. Fanti,K.A.,&Kimonis,E.R.(2013).Dimensionsofjuvenilepsychopathydistinguishbullies,bully-victims,andvictims.PsychologyofViolence,3,396 409. Furnham,A.,Richards,S.C.,&Paulhus,D.L.(2013).TheDarkTriadofpersonality:A10-yearreview.SocialandPersonalityCompass,7,199 216. Graham,L.T.,&Gosling,S.D.(2013).PersonalityprolesassociatedwithdifferentmotivationsforplayingWorldofWarcraft.Cyberpsychology,Behavior,andSocialNetworking,16,189 193. Hardaker,C.(2010).Trollinginasynchronouscomputer-mediatedcommunication:Fromuserdiscussionstoacademicdenitions.JournalofPolitenessResearch,6 Herring,S.,Job-Sluder,K.,&Scheckler,R.(2002).Searchingforsafetyonline:Managingtrollinginafeministforum.TheInformationSociety,18,371 384. Hyde,L.(1998).Trickstermakesthisworld.NewYork:FarrarStraus&Giroux Immordino-YangChristodoulou,J.A.,&Singh,V.(2012).Restisnotidleness:Implicationsofthebrainsdefaultmodeforhumandevelopmentandeducation.PerspectivesonPsychologicalScience,7,352 364. John,O.P.,&Srivastava,S.(1999).TheBig-Fivetraittaxonomy:History,measurement,andtheoreticalperspectives.InO.P.John&L.A.Pervin(Eds.),Handbookofpersonality:Theoryandresearch(pp.102 139).NewYork:Guilford Joiner,R.,Gavin,J.,&Dufeld,J.(2005).Gender,Internetidentication,andInternetanxiety:CorrelatesofInternetuse.CyberPsychology&Behavior,8,371 378. Jones,D.N.,&Paulhus,D.L.(2014).IntroducingtheShortDarkTriad(SD3):Abriefmeasureofdarkpersonalitytraits.Assessment,21,27 40. Jones,D.N.,&Paulhus,D.L.(2010).Differentprovocationstriggeraggressioninnarcissistsandpsychopaths.SocialPsychologicalandPersonalityScience,1 Jones,D.N.,&Paulhus,D.L.(2012).Differentiatingthedarktriadwithintheinterpersonalcircumplex.InL.M.Horowitz&S.N.Strack(Eds.),Handbookofinterpersonaltheoryandresearch(pp.249 267).Hoboken,NJ:Wiley&Sons Juvonen,J.,&Gross,E.F.(2008).Extendingtheschoolgrounds?Bullyingexperiencesincyberspace.JournalofSchoolHealth,78,496 505. Lenhardt,A.(2013).Teens,smartphones&texting.Washington,DC:PewResearchCentersInternet&AmericanLifeProject.Retrievedfrom Linton,D.K.,&Power,J.L.(2013).Thepersonalitytraitsofworkplacebulliesareoftensharedbytheirvictims:Isthereadarksidetovictims?PersonalityandIndividualDifferences,54,738 743. Ljepava,N.,Orr,R.R.,Locke,S.,&Ross,C.(2013).PersonalityandsocialcharacteristicsofFacebooknon-usersandfrequentusers.ComputersinHumanBehavior,29,1602 1607. McCosker,A.(2014).Trollingasprovocation:YouTubesagonisticpublics.Convergence:TheInternationalJournalofResearchintoNewMediaTechnologies(inpress).doi:1354856513501413. McCreery,M.P.,Krach,K.S.,Schrader,P.G.,&Boone,R.(2012).Deningthevirtualself:Personality,behavior,andthepsychologyofembodiment.ComputersinHumanBehavior,28,976 983. OMeara,A.,Davies,J.,&Hammond,S.(2011).Thepsychometricpropertiesandutilityoftheshortsadisticimpulsescale(SSIS).PsychologicalAssessment,23 Paulhus,D.L.,&Jones,D.N.(2014).Measuringdarkpersonalities.InG.J.Boyle,D.H.Saklofske&G.Matthews(Eds.),Measuresofpersonalityandsocialpsychological.SanDiego,CA:AcademicPress(inpress). Paulhus,D.L.,&Williams,K.M.(2002).Thedarktriadofpersonality:Narcissism,Machiavellianism,andpsychopathy.JournalofResearchinPersonality,36 Phillips,J.G.,&Butt,S.(2006).Personalityandself-reporteduseofmobilephonesforgames.CyberPsychology&Behavior,9,753 758. Rosen,L.D.,Whaling,K.,Rab,S.,Carrier,L.M.,&Cheever,N.A.(2013).IsFacebookcreatingiDisorders?Thelinkbetweenclinicalsymptomsofpsychiatricdisordersandtechnologyuse,attitudesandanxiety.ComputersinHumanBehavior,29,1243 1254. Shachaf,P.,&Hara,N.(2010).Beyondvandalism:Wikipediatrolls.JournalofInformationScience,36,357 370. Suler,J.(2004).Theonlinedisinhibitioneffect.CyberPsychology&Behavior,7 E.E.Buckelsetal./PersonalityandIndividualDifferencesxxx(2014)xxx xxx 5 Pleasecitethisarticleinpressas:Buckels,E.E.,etal.Trollsjustwanttohavefun.PersonalityandIndividualDifferenceshttp://dx.doi.org/10.1016/j.paid.2014.01.016 Walters,G.D.(2007).Thelatentstructureofthecriminallifestyle:Ataxometricanalysisofthelifestylecriminalityscreeningformandpsychologicalinventoryofcriminalthinkingstyles.CriminalJusticeandBehavior,34,1623 1637. Wiggins,J.S.(1995).InterpersonalAdjectiveScales:ProfessionalManual.Odessa,FL:PsychologicalAssessmentResources. Zweig,J.M.,Dank,M.,Yahner,J.,&Lachman,P.(2013).Therateofcyberdating abuseamongteensandhowitrelatestootherformsofteendatingviolence.JournalofYouthandAdolescence,42,1063 1077. E.E.Buckelsetal./PersonalityandIndividualDifferencesxxx(2014)xxx xxx Pleasecitethisarticleinpressas:Buckels,E.E.,etal.Trollsjustwanttohavefun.PersonalityandIndividualDifferences(2014),http://dx.doi.org/10.1016/j.paid.2014.01.016