writing question and need a sample draft to help me learn.
Winners and Losers in laws and criminality?
Bi-weekly papers (EVERY 2 Weeks) should be no less than two typed pages, and no more than three typed pages in length. Here we are looking for concise, cogent points, not epics; YOU SHOULD HAVE SUPPORT FOR YOUR ASSERTIONS BASED ON READINGS! Do NOT turn in garbage papers to me. (I find that to be insulting, and no one likes feeling insulted.)
Follow APA Guidelines.
Basically a two page paper with your thoughts and reaction to the reading with a little summary
Requirements: See instructions
JournalofCommunicationISSN0021-9916ORIGINALARTICLEEmotions,Partisanship,andMisperceptions:HowAngerandAnxietyModeratetheEffectofPartisanBiasonSusceptibilitytoPoliticalMisinformationBrianE.WeeksDepartmentofCommunicationStudies,UniversityofMichigan,AnnArbor,MI48109,USACitizensarefrequentlymisinformedaboutpoliticalissuesandcandidatesbutthecircum-stancesunderwhichinaccuratebeliefsemergearenotfullyunderstood.Thisexperimentalstudydemonstratesthattheindependentexperienceoftwoemotions,angerandanxiety,inpartdetermineswhethercitizensconsidermisinformationinapartisanoropen-mindedfashion.Angerencouragespartisan,motivatedevaluationofuncorrectedmisinformationthatresultsinbeliefsconsistentwiththesupportedpoliticalparty,whileanxietyattimespromotesinitialbeliefsbasedlessonpartisanshipandmoreontheinformationenviron-ment.However,exposuretocorrectionsimprovesbeliefaccuracy,regardlessofemotionorpartisanship.Theresultsindicatethattheuniqueexperienceofangerandanxietycanaffecttheaccuracyofpoliticalbeliefsbystrengtheningorattenuatingtheinfluenceofpartisanship.Keywords:Misinformation,Misperceptions,Emotion,Fact-checking,Corrections,MotivatedReasoning,Anger,Anxiety.doi:10.1111/jcom.12164Politicsarenotdevoidofemotion.Rather,emotionsareacorecomponentofcitizen-shipthatshapehowpeopleseetheirpoliticalworld(Marcus,2002).Emotionshaveapowerfulinfluenceonpoliticalattitudesandbehaviors(Brader,2005;Marcus,MacK-uen,&Neuman,2011;Nabi,2003),butmighttheyalsoaffectwhatcitizensperceivetobepoliticalreality?Politicalmisperceptions,whicharecharacterizedaspersonalbeliefsaboutpoli-ticsthatareinaccuratebasedonthebestavailableevidence(Kuklinski,Quirk,Jerit,Schwieder,&Rich,2000),areprevalentinAmericanpoliticsandheldaboutarangeofpoliticalissuesandcandidates(Jerit&Barabas,2012;Weeks&Garrett,2014).TheprevailingunderstandingforwhymisperceptionsemergeisbasedonthetheoryofCorrespondingauthor:BrianE.Weeks;e-mail:beweeks@gmail.comJournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation699
EmotionsandMisperceptionsB.E.Weekspartisanmotivatedreasoningandsuggeststhatpoliticalinformationconsistentwithanindividual’sideologyorpartisanshipismorelikelytobebelieved,regardlessofitsveracity(e.g.,Garrett&Weeks,2013;Nyhan&Reifler,2010).Certainly,partisan-shipmattersforpoliticalbeliefsandmisperceptionsoftenfallalongpartylines,asDemocratsandRepublicansvaryintheiraccuracyregardingthefactsaboutpolitics(Gaines,Kuklinski,Quirk,Peyton,&Verkuilen,2007).However,thispartisan-basedexplanationforpoliticalmisperceptionsmaybeincomplete,asresearchhasnotfullyidentifiedtheconditionsunderwhichpoliticallybiasedevaluationsofmisinforma-tionoccur.Thisstudycontributestoourunderstandingofpoliticalmisperceptionsbyargu-ingthatcitizens’emotionalexperiencescanenhanceorlessenpartisanship’sinfluenceoninaccuratebeliefs.Thefocushereisontheeffectsoftwonegativeemotions:anxi-etyandanger.Althoughtheliteratureisbeginningtorecognizetheinfluenceoftheseemotionsonmotivatedreasoning,thedifferentialeffectsofangerandanxietyinthisprocessarenotfullyunderstood,norhavetheybeenexploredinthecontextofmis-perceptions.Priorresearchhasfocusedprimarilyontheinfluenceofanxiety,showingthattheexperienceofthisemotioncanattenuatetheeffectofexistingattitudesonpoliticalevaluations(MacKuen,Wolak,Keele&Marcus,2010;Marcusetal.,2011;Redlawsk,Civettini,&Emmerson,2010).Lessattentionhasbeenpaidtotheeffectsofangeronmotivatedreasoning,thoughMacKuenetal.(2010)foundthatangryindi-vidualswerelesslikelytoseekoutattitude-challenginginformationandlesswillingtocompromisepolitically.Whilepriorworksuggestsanxietyandangermayuniquelyaffectpartisanprocessing,ithasneitherisolatedthecausalinfluenceofeachemo-tion,norhasitfullyassessedwhethertheseemotions’influenceextendstoexplicitlypartisanmotivationsortotheaccuracyofpoliticalbeliefs.Toaddressthesetheoreticalgaps,thisexperimentalstudymanipulatesemotionalstates,partisancues,andtheinformationenvironment(presenceofacorrectivemes-sageornot)tobetterunderstandhowpoliticalmisperceptionsareformed.Theresultsdemonstratethatparticipantsattimesrespondtouncorrectedmisinformationindif-ferentwayswhenangryoranxious,resultingindivergentbeliefsaboutaseriesofinaccuratepoliticalclaims.Whenfalseclaimsgouncorrected,angerexacerbatestheinfluenceofpartisanshipandmakesparticipantsmoresusceptibletoparty-consistentmisinformation—claimstheyarepredisposedtobelievebecauseoftheirpoliticalaffiliation.Conversely,anxietylessenstheinfluenceofpartisanshipwhenconsideringuncorrectedmisinformationandincreasesthelikelihoodthataclaimdisseminatedfromtheopposedpoliticalpartyisbelieved.Thisstudyprovidesamorecompletetheoreticalexplanationforhowpoliticalmisperceptionsareinitiallyestablished,aswellasacausalaccountofhowangerandanxietyuniquelyaffectpartisanprocessingofpoliticalnewsandinformation.PartisanshipandpoliticalmisperceptionsTheformationandpersistenceofpoliticalmisperceptionsareoftenattributedtoapoliticallymotivatedneedforconsistency(e.g.,Meirick,2013;Nyhan&Reifler,2010;700JournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation
B.E.WeeksEmotionsandMisperceptionsWeeks&Garrett,2014;though,cf.Garrett,Nisbet,&Lynch,2013).Thetheoryofmotivatedreasoningsuggestsindividualsattimesevaluateinformationinabiasedmannerinordertoremainconsistentwithpriorattitudesorbeliefs(Kunda,1990).Inthecaseofpolitics,thismeanscitizensareoftenmotivatedtoconsidernewinforma-tionthroughapartisanorideologicallens(Taber&Lodge,2006).Partisanmotivatedreasoningispresumedtobedrivenbyaffective-basedgoalsthatleadcitizenstoeval-uateattitude-consistentinformationasstrong,convincing,andvalid,whileincon-sistentinformationisconsideredweakandunconvincing(Taber&Lodge,2006).Inthecontextofmisperceptions,thetheorysuggeststhatpoliticalclaimsconsistentwithone’spartisanshiporideologyaremorelikelytobeaccepted,regardlessoftheirverac-ity.Incontrast,claimsaremorelikelytoberejectediftheychallengeexistingpartisanattitudes.Empiricalevidencesupportsthiscontention,asmisinformationregardingpoliticalcandidatesorissuesistypicallymorelikelytobebelievedwhenitiscon-sistentwithpartyidentificationorideology(Jerit&Barabas,2012;Nyhan&Reifler,2010;Weeks&Garrett,2014).However,citizensdonotalwaysinterpretpoliticalinformationinawaythatisbiasedtowardtheirpartisanleaningsbutinsteadattimesconsiderinforma-tioninamoreeven-handedmanner(Bullock,2009;Druckman,2012;MacKuenetal.,2010).Thiswillingnesstosetasidepartisanshipisevidentwithmisinfor-mationtoo,ascitizensoccasionallybelievefalse,attitude-discrepantclaimsandrejectattitude-consistentones.Forexample,apollconductedin2012foundnearlyone-thirdofDemocratsincorrectlybelievedordidnotknowifIraqpossessedweaponsofmassdestructionpriortotheU.S.invasionin2003,amisperceptionthatseemstocontradictDemocrats’oppositiontothewaratthattime(Valentino,2012).Surveysfieldedduringthe2012presidentialelectionshowbetweenapproximately10and20%ofDemocratsbelievedorwereunsurewhetherBarackObamawasbornintheUnitedStates,whilebetween30and50%ofRepublicansdismissedthisclaimasfalse(Cassino,2013;Gallup,2011).1Soalthoughcitizensaremorelikelytoholdattitude-consistentmisperceptions,partisanmotivatedreasoningappearstohavelimits.Theinfluenceofpartyidentificationmightbeconditionalbutscholarshavenotfullyidentifiedwhenpartisanshipaffectsfalsebeliefs.Thefollowingsectionsarguethatanxietyandangermoderatepartisanship’sinfluencewhencitizensconsidertheveracityofmisinformation.AnxietyandangerDespitesharinganegativevalence,theemotionsanxietyandangerexhibitseveraldistinctqualitiesthatshouldleadtodifferentresponsestomisinformation.Emo-tionsare“internal,mentalstatesrepresentingevaluative,valencedreactionstoevents,agents,orobjectsthatvaryinintensity…[t]heyaregenerallyshort-lived,intense,anddirectedatsomeexternalstimuli”(Nabi,1999,p.295).Emotionsplayakeyroleinhowpeopleinteractwithstimulitheyencounterintheirenvironmentanddiffer-entemotionsaredistinguishedbyuniquegoalsandmotivations,cognitiveappraisals,andactiontendencies(Frijda,1986;Lazarus,1991;Roseman,Wiest,&Swartz,1994).JournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation701
EmotionsandMisperceptionsB.E.WeeksAnxietyisanaversiveandmotivationalstatethatoccursinresponsetothreateningstimuliinone’senvironment(Eysenck,Derakshan,Santos,&Calvo,2007;Marcus,Neuman,&MacKuen,2000).Anxietyisassociatedwithuncertainty,alackofper-sonalcontrol,andhighlevelsofphysiologicalarousal,allofwhichleadpeopletodevelopstrategiestoreducesuchfeelings(Eysencketal.,2007).Angerisanegativelyvalencedemotionthatistriggeredwhenanindividual’sgoalsareblocked,whenonefeelsslighted,orwhenaperceivedinjusticeorviolationofstandardshasoccurred(Carver&Harmon-Jones,2009).Angerisassociatedwithperceptionsofcertaintyandindividualcontrol(Lerner&Keltner,2001)andresultsinbehaviorsthatseekrestitution,oftenbycausingharmtoorpunishingothers(Carver&Harmon-Jones,2009;Frijda,1986).Althoughangerandanxietyoftenco-occurandarenotmutuallyexclusive,evidencedemonstratestheseemotionscanresultindivergentpoliticaloutcomes(Huddy,Feldman,&Cassese,2007;Lerner&Keltner,2001).Anxietyincreasespoliticalinformationseeking,learning,anddeliberation,whereasangerdepresseseachandpromotesclose-mindedness(MacKuenetal.,2010;Redlawsketal.,2010;Valentino,Hutchings,Banks,&Davis,2008).Angerandanxietyarealsoassociatedwithdifferencesinsupportforvariouspoliticalpolicies.Anxiety(orfear)canincreasesupportforconciliatoryandprecautionarypolicieswhereasanger(orknownthreat)boostspunitivepolicysupport(Goodall,Slater,&Myers,2013;Nabi,2003).Theseuniqueoutcomesofangerandanxietymakeitclearthatconceptualizationsofemotionsneedtomovebeyondpositiveornegativevalence(Nabi,2010).EmotionandpartisanbiasPoliticallymotivatedreasoningisthoughttobe“drivenbyautomaticaffectiveprocessesthatestablishthedirectionandstrengthofbiases”(Taber&Lodge,2006,p.756)buttheobserveddifferencesbetweenanxietyandangersuggestnotallaffectiveresponses,eventhoseofthesamevalence,workinthesameway.Evidencesuggeststheexperienceofanxietycandiminishtheeffectsofmotivatedreasoningaspeopleformpoliticalevaluations.AccordingtothetheoryofAffectiveIntelligence,novelanduncertainpoliticalsituationsmaytriggerfeelingsofanxiety(Marcusetal.,2000).Thisanxietymotivatespeopletotakeactionagainstthatimpendingthreatbymorecare-fullyconsideringeachofthevariousalternativesattheirdisposal,eveninformationthatchallengesexistingattitudes.Becausetheypaycloserattentiontotheinformationathandandaremorelikelytoweighopposingviewpoints,thesubsequentevaluationsmadebyanxiousindividualsarelesslikelytobebasedonpriorattitudes,partisan-ship,orideologyandaremoreheavilyinfluencedbycontemporaryinformation(Brader,2005;MacKuenetal.,2010;Marcusetal.,2000).Forexample,voterswhowereanxiousabouttheirpreferredparty’scandidateweremorelikelytoreconsidertheirvotechoiceandpaycloserattentiontothecandidates’policypositionsandpersonalitycharacteristics,whichattimesresultedinanintentiontovoteagainsttheirparty’scandidate(Marcusetal.,2011).Redlawsketal.(2010)directlytestedthehypothesisthatanxietyisassociatedwithareductioninattitudinalbiasbyexposing702JournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation
B.E.WeeksEmotionsandMisperceptionsparticipantstolargeamountsofpoliticallyincongruentinformation.Theyfindevi-denceofan“affectivetippingpoint”inwhichencounteringahighvolumeofnegativeinformationaboutapreferredpoliticalcandidatewasassociatedwithincreasesinanxiety,whichwassubsequentlyrelatedtomorenegativeevaluationsoftheinitiallylikedcandidate.Thisevidencesuggeststhatanxietymayattenuatepartisanbias.Conversely,angermayenhancethemotivatedreasoningprocess.Angerarisesinresponsetoaversivestimuliintheenvironmentandinordertodealwiththisanger,peoplemaybecomedefensive,dismissattitude-challenginginformation,seekinfor-mationthatsupportstheirposition,orrelymoreheavilyonpriordispositionsintheirevaluations(MacKuenetal.,2010;Marcusetal.,2011).Angryindividualsarealsolesslikelytoconsiderandlearnfromnewinformation(Valentinoetal.,2008).Takentogether,thispriorworksuggeststhatanxietyandangershoulddiffer-entiallyinteractwithpartisanshiptoaffectbeliefs,butmanyimportantquestionsremain.ThecurrentstudybuildsonworkbyRedlawsketal.(2010)andMacKuenetal.(2010)inseveralimportantways.First,evidenceofacausaleffectofangerandanxietyonmotivatedreasoninghasbeenelusive.2Neitherstudydirectlymanipulatedtheseemotions,whichprohibitscleartestsaboutthedifferentialeffectsofanxietyandangerinthemotivatedreasoningprocess(andRedlawsketal.onlyexaminedanxiety).Second,emotionsinthesestudiesweregeneratedbyexposingparticipantstovariousamountsofinformationthateitherchallengedorwasconsistentwiththeirpreferences.Whatisstilluncleartheoreticallyiswhethergeneralanxietyorangeraboutapoliticalissue,independentofthetypeorvolumeofinformationexposure,canaffectpartisanprocessing.Third,thetwostudiesdidnottestwhethertheeffectsofanxietyandangerextendtoexplicitlypartisanmotivations.TheRedlawsketal.studywaslimitedtoaprimaryelectioncontextanddidnottestfortheeffectsofpartisanship.Similarly,MacKuenetal.examinedtheeffectsofemotionsonpoliticalbiasinthecontextofapoliticalpolicyanddidnotexaminetheroleofpartisanaffili-ation.Theomissionofpartisanshipinthesestudiesisimportantinlightofcriticismsthatemotionsareoftenrationalizationsofpartisanpreferences(Ladd&Lenz,2008).Yet,researchhasfrequentlyshownthattheeffectsofemotionsonpoliticalbehaviorareexogenoustoorinteractivewithpartisanship(seeBrader,2011)andthecurrentstudyisdesignedtofurthertestthispossibility.Finally,whileexistingstudiesfocusonhowemotionsmaymoderatetheeffectofpoliticalbiasoninformationseekingandevaluationsofpoliticalissuesandcandidates,thepresentstudyexamineswhethertheseemotionsaffectassessmentsofpoliticalfacts,whichmayhelpexplainsomeuncertaintiesregardingwhenpartisanbiasinfluencesbeliefaccuracy.Anxiety,anger,andpoliticalmisperceptionsPeopleshouldrespondtopoliticalmisinformationquitedifferentlywhenexperi-encinganxietyoranger.Anxietyshouldincreasethelikelihoodthatindividualscarefullyconsiderawiderrangeofviewpoints,eventhosetheymaydisagreewith,whichshoulddiminishtheinfluenceofpartisanshipandincreasetheimpactofcontemporaryinformationinevaluation(MacKuenetal.,2010;Marcusetal.,2011;JournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation703
EmotionsandMisperceptionsB.E.WeeksRedlawsketal.,2010).Inthecaseofmisinformation,thissuggeststhatanxiousindi-vidualsexposedtocounterattitudinalcorrectionsshouldmorecarefullyweighfalseclaimsversusfacts,whichshouldincreasethecorrections’effectiveness.Specifically,whenacorrectionisprovided,anxiousindividualsexposedtomisinformationfromthein-partywillholdmoreaccuratebeliefsthanthoseinaneutralemotionalstate(H1a).Althoughtheinfluenceofanxietyinitiallyappearstobenormativelybeneficialfordemocracybecauseitmayleadtomorerationalupdating,thereisalsothepossibilityforadverseeffects(Marcus,2002).Forinstance,peopleprimedtoexperiencehighlev-elsofanxietywhowerethenexposedtoananti-free-speechmessageweresignificantlylesstolerantpoliticallycomparedtothosewhoreceivedalowanxietyprime(Marcus,Sullivan,Theiss-Morse,&Stevens,2005).Inthecaseofpoliticalmisinformation,anx-ietycouldmakeindividualsmoresusceptibletouncorrectedfalseclaimsthatreflectnegativelyontheirownparty.Becausetheyarelesslikelytobedrivenbypartisanmotivations,anxiouscitizensshouldpauseandmorecarefullyconsiderthenatureofthemisinformationwhenitconflictswiththeirpartyaffiliation(Marcusetal.,2011).Asaresult,anxiousindividualswhoreceiveuncorrectedmisinformationfromtheout-partywillholdlessaccuratebeliefsthanthoseinaneutralemotionalstate(H1b).Anger,however,shouldenhancetheinfluenceofpartisanshiponfalsebeliefs.Angermakespeopledefensiveandincreasesthelikelihoodthattheyignoreinfor-mationthatchallengestheirattitudesandpaycloserattentiontoattitude-consistentinformation(Marcusetal.,2011).Asaresult,angerleadstoevaluationsthatarebasedmoreheavilyonexistingdispositions,suchaspartisanship.MacKuenetal.(2010)foundthatangryindividualssoughtoutmoreinformationthatconfirmedtheirpriorattitudes,becameresoluteinthosepriorattitudes,andwerelesswillingtocompro-misepolitically.Extendingthesefindingstomisperceptions,angershouldresultinlessopennesstonew,attitude-challenginginformationcontainedinafact-checkingmessageandanunwillingnesstoheartheotherside.Thus,angershouldenhancetheinfluenceofpartisanshiponbeliefs,aspeoplewillbemorelikelytofallinlinewiththeirpoliticalpartywhentheyareangry.Asaresult,angrycitizensshouldbemoresusceptibletoinaccurateclaimsendorsedbyamemberoftheirpreferredpoliticalpartyandlesslikelytoacceptcorrectionsthatindicatetheirparty’sclaimsarewrong.Inparticular,whenacorrectionisprovided,angryindividualsexposedtomisinforma-tionfromthein-partywillholdlessaccuratebeliefsthanthoseinaneutralemotionalcondition(H2a).Also,angryindividualsexposedtouncorrectedmisinformationfromthein-partywillholdlessaccuratebeliefsthanthoseinaneutralemotionalstate(H2b).MethodAnonline,between-participantsexperimenttestedthesehypotheses.English-speakingadultsoverage18wholiveintheUnitedStatesandhaveaccesstotheInternetparticipatedinthestudy(N=768).ThesamplingfirmQualtricsrecruitedthesampleusinganopt-inpanel,andthedatawerecollectedbetween25Marchand2April2014.Thesamplewasdiverseintermsofage(M=47.99,SD=14.05),704JournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation
B.E.WeeksEmotionsandMisperceptionsgender(50%male),race(83.3%White,8.5%Black,and7%Hispanic),education(22.0%highschoolorless,25.3%4-yearcollegedegree,and11.6%postgraduate),andpoliticalpartyaffiliation(43.6%Democratorleaning,29.6%Republicanorleaning,and23.7%Independent).ProcedureQualtricsfirstsentpanelmembersanemailaskingthemtoparticipateinthestudy.Afterprovidingconsent,participantswererandomlyassignedtotakeatleast2min-utestowritesomethingabouteitherimmigrationreformorthedeathpenalty(issuemanipulation)thatmakesthemeitherangryoranxious(emotionmanipulation).Par-ticipantsinacontrolconditionwroteaboutsomethingthatmakesthemrelaxed.Sim-ilartaskshavebeenusedinpriorresearchandallowfortheindependentinductionofeachemotionrelatedtothetargetissue(Lerner&Keltner,2001;Valentino,Brader,Groenendyk,Gregorowicz,&Hutchings,2011).Althoughthiswritingtaskgener-atesemotionssomewhatartificially,intherealworldpeoplefrequentlyusestrongemotionallanguagetoexpresstheiropinionsonline(Papacharissi,2012),andnewsorganizations’websitesarefilledwithuncivilandemotionalcomments(Coe,Kenski,&Rains,2014).Thewritingtaskisthusareasonablerepresentationofhowpeo-pleexpressthemselvespoliticallyoutsideoftheexperiment.Immediatelyfollowingthewritingtask,participantsprovidedtheircurrentemotionalstatebyreportingtheextenttowhichtheywereexperiencingaseriesofemotions.Matchingtheissuetheywroteaboutintheemotionalinduction,participantsnextreadanewsarticleattributedtotheAssociatedPressthatdiscussedhowpublicmis-perceptionswerecirculatingaroundthepoliticaldebateoneitherimmigrationorthedeathpenalty(seeAppendixS1,SupportingInformation,forsamplestimuli).ThearticlesweremanipulatedtocontaineitherA)aseriesoffourinaccurateanduncor-rectedclaims(misinformation-only)abouttheissue,orB)thesameclaimsaswellascorrectiveinformationprovidingexplicitevidenceshowingwhythestatementsarefalse(correction).Thetwoissueswereusedasaformofstimulussampling,whichimprovesconstructandexternalvalidity(Wells&Windschitl,1999).ThepoliticalaffiliationofthesourceoftheoriginalmisinformationwasalsomanipulatedsuchthattheinaccurateinformationwasattributedtoeitherCongressionalRepublicansorDemocrats.Thecorrectiveinformationinthecorrectionconditionwasattributedtoafictionalindependentfact-checkingorganization.Thenewsarticleswerefictitiousbutwerebasedonrealisticpresentationsofmisinformationinthemedia.Outsideofthelaboratory,thepubliccanbeexposedtomisinformationpresentedwithoutexplicitcorrectionsvianewsorpoliticalwebsites,blogs,ande-mail(Garrett,2011;Weeks&Southwell,2010),whilenewsorganizationsandfact-checkingorganizationssuchasFact-check.orgoftenrebutpoliticalinaccuraciesbypresentingboththeorig-inalclaimandanevidence-basedcorrection(Gottfried,Hardy,Winneg,&Jamieson,2013).Theclaimswereexplicitlyfalseandbasedonrealmisperceptionssurround-ingeachissue.Thefactualstatementsusedascorrectionswerecreatedusingevi-dencefromreportsbyexpertsatgovernmentalorganizationsandnonpartisaninterestJournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation705
EmotionsandMisperceptionsB.E.Weeksgroups.Immediatelyafterreadingtheirassignednewsarticle,participantsreportedtheirbeliefsintheclaimspresentedinthearticle.Thestudyconcludedbyassessingpsychologicalindividualdifferencesanddemographics.MeasuresEmotionDummycodedvariableswerecreatedtorepresenttheemotionalinductioncondi-tions.Boththeangerandanxietyconditionswerecodedhighandtheneutralcontrolconditionwasusedasthereferencegroup.Allmodelsintheanalysisusetheinductionconditiondummyvariablesaspredictors.SourceofmisinformationThesourceoftheoriginalmisinformationwasmanipulatedsuchthatclaimswereattributedtoeitherCongressionalRepublicansorDemocrats.Adichotomousvari-ablewascreatedtorepresentwhetherthesourceofthemisinformationwasfromthepoliticalpartyparticipantssupportedoropposed.If,forexample,aself-reportedRepublican(orleaning)sawanarticleinwhichthefalseclaimswereattributedtoCongressionalRepublicans,thatcombinationwouldbecodedas“in-party.”IfaDemocratsawthatsamearticlethatcombinationwouldbecodedas“out-party.”Thisresultedinadichotomousvariableinwhich287participantsreadanarticleattributingthemisinformationtotheout-party,while275participantssawclaimsstemmingfrommembersoftheirin-party(codedhigh).ThiscodingexcludestrueIndependentswhodidnotleantowardoneparty,participantsassociatedwithathirdparty,orthosewhodidnotrespondtothepoliticalaffiliationquestion.CorrectionThepresenceofacorrection(codedhigh)wasmanipulatedsuchthatinaccurateclaimswerepresentedeitherwithorwithoutevidenceindicatingtheclaimswerenottrue.PoliticalknowledgePoliticalknowledgeisoftenassociatedwithpoliticalmisperceptionsandwasincludedasacontrolinthemodels(Nyhan&Reifler,2010).Politicalknowledgewasmeasuredusingquestionsaboutthecurrentmake-upofCongress,thepartyofthePresidentwhoappointedthecurrentChiefJusticeoftheSupremeCourt,thecurrentunemploy-mentrate,andthecurrentU.S.SecretaryofState.Theitemsweresummed,creatingarangeofpossiblescoresfrom0to4(α=.65,M=2.00,SD=1.37).BeliefinfalseclaimsBeliefaccuracywasthedependentvariableandwasmeasuredbyaskingpartic-ipantstoreporttheextenttowhichtheybelievedeachofthefourclaimstheyreadtobetrueorfalse.Participantswereasked:“Thearticleyoujustreadmadeseveralclaimsaboutimmigration(thedeathpenalty).Whatdoyouthinkabouttheaccuracyofthesestatements?Foreachstatementlistedbelow,pleasetellusifyou706JournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation
B.E.WeeksEmotionsandMisperceptionsthinkitisdefinitelytrue,probablytrue,probablyfalse,definitelyfalseoryouareunsure?”Participantswhoreadtheimmigrationarticlewereaskedtoreporttheirbeliefinthefollowingfalseclaims:(a)Illegalimmigrantsareabletoreceivegovernmentwelfarebenefitslikefoodstampsandhousingbenefits;(b)ThemajorityofimmigrantsdonotlearntospeakEnglish;(c)Immigrantsaremorelikelytobecomecriminalsthannativeborncitizens;and(d)Proposedlegislationwillallowanycurrentillegalimmigranttobecomeacitizen.Participantswhoreadthedeathpenaltyarticlereportedtheirbeliefinfoursepa-rateclaims:(a)ThenumberofinmatesexecutedintheUnitedStatesisgrowing;(b)MinoritiesaremorelikelytobeexecutedthanWhites;(c)Itislegaltoexecutejuve-nilesintheUnitedStates;and(d)Publicsupportforthedeathpenaltyhasincreasedinrecentyears.Priortoanalyses,responseswererecodedsothatthe“unsure”responsewaslocatedatthemidpointoftherevised5-pointscale.Highervaluesonthescalerepresentgreateraccuracyinassessingthesefalseclaims.3Onaverage,participants’perceptionsofthefourimmigrationclaimswerenotveryaccurate,asthemeanwasatthemidpointofthe5-pointscale(fouritemsaveragedscale,α=.75,M=2.99,SD=1.06).Beliefsaboutthedeathpenaltyclaimsweremoreaccurate,asthemeanfellbetweenthe“unsure”and“probablyfalse”responseoptions(fouritemsaveragedscale,α=.74,M=3.19,SD=1.07).At-testconfirmedthatbeliefsaboutthedeathpenaltyweresignificantlymoreaccuratethanthoseaboutimmigration,t(766)=2.52,p<.05.Theissueswerecombinedforanalyses(M=3.09,SD=1.07)buttheissuevariableservedasacontrolthroughout.Althoughthesescaleditemsareusedtotestthehypotheses,convertingthebeliefitemsintodichotomousmeasuresofaccuracyhelpsillustratethedistributionofcorrectresponses.Allfourclaimsaboutbothissueswerefalse,soparticipantswhoresponded“probablyfalse”or“definitelyfalse”werecreditedwithprovidingthecorrectoraccurateresponse.Acrosstheentiresample,participantsprovidedthecorrectresponsetojustunderhalfofthefourclaims(range0to4;M=1.89,SD=1.39).Table1displaysthepercentageofparticipantswithcorrectbeliefsbycondition,aswellasthemeannumberofcorrectresponsesineachcondition.Acrosstheentiresample,oneinfiveparticipants(20.8%)didnotprovideacorrectresponsetoanyofthefourclaims,while17.1%wereaccurateinassessingallclaims.4ResultsEmotioninductioncheckAninductioncheckdemonstratesthewritingtaskgeneratedtheintendedemotionalresponse.Emotionalresponseswereassessedonascalefrom1(veryslightlyornotatall)to5(extremely)andrandomlypresenteditemswereusedtomeasureanger(angry,outraged,anddisgusted;α=.90,M=1.83,SD=1.08),anxiety(anxious,afraid,andnervous;α=.84,M=1.83,SD=0.93),andenthusiasm(enthusiastic,JournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation707
EmotionsandMisperceptionsB.E.WeeksTable1PercentageofParticipantsWithCorrectBeliefsbyConditionNumberofCorrectBeliefs(%)Condition01234MeanNumberCorrect(SD)EmotionAnger19.125.218.818.818.21.94(1.40)Anxiety22.520.725.316.515.11.91(1.34)Neutral21.419.617.323.817.92.06(1.42)PartyofsourceIn-party21.823.318.918.517.51.87(1.41)Out-party16.720.623.020.619.22.04(1.36)InformationenvironmentMisinformationonly29.426.024.514.55.61.50(1.22)Correction11.118.116.724.230.02.45(1.38)IssueImmigration23.823.320.217.914.81.75(1.40)Deathpenalty17.821.221.520.219.42.16(1.34)Entiresample20.822.320.819.017.11.89(1.39)Note:Beliefswerecollapsedintoadichotomousmeasureandconsideredaccurateifrespondentreportedthatthefalseclaimwas“probably”or“definitely”false.hopeful,andproud;α=.84,M=2.69,SD=1.14).Angerandanxietywerepositivelycorrelated,(r=.56,p<.001),whileangerandenthusiasmwerenegativelycorrelated(r=−.06,p<.10[two-tailed]).Anxietyandenthusiasmwerepositivelycorrelated(r=.12,p=.001).Participantsaskedtowritesomethingthatmadethemangryreportedmoreanger(M=2.10,SD=1.17)thanparticipantsintheanxiety(M=1.78,SD=1.06)orneutralconditions(M=1.42,SD=0.75),F(2,764)=23.61,p<.001.Thedifferenceinangerbetweentheangerandanxietyconditionswasalsosignificant,t(597)=3.50,p<.001.Participantswhowroteaboutwhatmadethemanxiousreportedmoreanxiety(M=1.95,SD=1.00)thanparticipantsintheanger(M=1.78,SD=0.88)orneutralconditions(M=1.71,SD=0.89),F(2,764)=4.04,p<.05.Adirectcomparisonofanxietybetweentheanxiousandangryconditionswassignificant,t(597)=2.15,p<.05.Finally,thoseintheanxietyconditionfeltmoreanxietythananger,t(284)=3.23,p<.001,whileparticipantsintheangerconditionexperiencedgreaterangerthananxiety,t(313)=5.35,p<.001.StatisticalanalysesAlthoughmaineffectsoftheexperimentalconditionswerenothypothesized,theresultofanordinaryleastsquaresregressionpredictingbeliefaccuracywithouttheinteractiontermsisreported(Table2,Model1).Thehypothesesweretestedusingtwoseparate“moderatedmoderation”regressionmodels,alsoknownasthree-way708JournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation
B.E.WeeksEmotionsandMisperceptionsTable2EffectsofAnxietyandAnger,Partisanship,andCorrectionsonBeliefModel1Model2Model3Anxiety−.14(.11)−.42(.19)*−.15(.11)Anger−.09(.11)−.09(.11).26(.18)Correction(codedhigh).78(.08)***.70(.15)***.98(.15)***In-party(codedhigh)−.13(.08)−.35(.15)*.15(.15)Anxiety×In-party—.63(.24)**—Anxiety×Correction—.27(.24)—Correction×In-party—.23(.21)−.36(.22)#Anxiety×Correction×In-party—−.71(.35)*—Anger×In-party——−.64(.23)**Anger×Correction——−.45(.24)#Anger×Correction×In-party——.79(.34)*Issue(deathpenaltycodedhigh).27(.08)**.27(.08)**.27(.08)**Politicalknowledge.15(.03)***.15(.03)***.15(.03)***Constant2.18(.18)***2.25(.19)***2.02(.20)***Observations561561561F(df)20.46(6,554)13.07(10,550)***13.16(10,550)***R2.18.192.193Notes:Unstandardizedregressioncoefficientsreported.Standarderrorislistedinparentheses.p-valuesaretwo-tailed.#p<.10.*p<.05.**p<.01.***p<.001.interactions(Hayes,2013).Thefirstmodeltestedthehypothesesaboutanxiety,whilethesecondexaminedpredictionsrelatedtoanger.Thehypothesesmakedifferentpre-dictionsabouttheinteractiveeffectsofemotionsandpartisanshiponbeliefaccuracydependingontheinformationenvironmentandthethree-wayinteractionmodelsallowforsuchtests.Table2(Models2and3)illustratesthatthethree-wayinterac-tionbetweentheemotion,partisansource,andcorrectionvariablesissignificantinboththemodelforanxiety,b=−.71(.35),p=.04(two-tailed)andanger,b=.79(.34),p=.02(two-tailed).Thesesignificantcoefficientsindicatethattheextenttowhichthetwoemotionsandpartisanshipinteracttoaffectbeliefsiscontingentonthepres-enceofacorrection.Whatthecoefficientsdonotindicate,however,istheprecisenatureoftheseinteractions.Inordertoexaminetheseconditionaleffectsandtestthespecifichypothesesposedinthisstudy,thesesignificantinteractionswerefur-therprobedusingPROCESS(Hayes,2013,p.307).Probingtheinteractionsallowsfordirectcomparisonsofthehypothesizedgroupdifferences.PROCESSprovidesinfer-entialteststhatexplainthenatureoftheemotion×partisansourceinteractionsinboththemisinformation-onlyandcorrectionconditions.InreportingtheresultsofthehypothesestestsbelowIfocusonthecoefficientsfortheseinferentialtests(whichinthiscasesignifyadjustedmeandifferencesbetweengroups)aswellastheestimatedmarginalmeansforthegroupsbeingcompared.JournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation709
EmotionsandMisperceptionsB.E.WeeksTestsofmaineffectsThecoefficientsreportedinTable2(Model1)indicatetherewerenomaineffectsoftheangeroranxietyconditionsonparticipants’beliefs(ps>.20).Thecoefficientforthevariablerepresentingpartisanshipwasnegative,b=−.13(.08),suggestingthatparticipantswhoreceivedmisinformationfromthein-partyheldlessaccuratebeliefsthanthosewhoreceivedthesameinformationfromtheout-party.However,thiseffectwasnotsignificant(p=.11)andcannotbetakenasstrongevidenceregardingthedirectinfluenceofpartisanship.Thelackofmaineffectssuggeststhatemotionsandthepartisansourceoftheinformationalonedidnotsignificantlyaffectbeliefaccuracy.5Thereweremaineffectsoftheissue(beliefsaboutthedeathpenaltyweremoreaccurate)andcorrectionmanipulations,asparticipantswhoreceivedcorrec-tiveinformationheldsignificantlymoreaccuratebeliefsthanthosewhoreceivedonlythemisinformation.Thecoefficientforthecorrectionvariablesuggeststhatallthingsequal,receivingacorrectionwasestimatedtoimprovebeliefaccuracyby.78units(t=9.29,p<.001).Inaddition,thecontrolvariablepoliticalknowledgewaspositivelyrelatedtobeliefaccuracy,b=.15,p<.001.HypothesestestsH1apredictedthatwhenacorrectionisavailable,anxiousindividualsexposedtomisinformationfromthein-partyholdmoreaccuratebeliefsthanthoseintheneutralcondition.Inotherwords,correctionsthatchallengeone’spartisanloyaltiesaremoreeffectivewhenanxietyispresent.Thishypothesiswastestedbyexaminingthecoefficientfortheanxiety×partisansourceinteractionwithinthecorrectioncondition.Thecoefficientwasnotsignificant,b=−.08(.25),t=−.33,p=.75,indi-catingthatanxietyandthepartisansourcedonotinteracttoaffectbeliefsinthecorrectioncondition.Exploringthisnonsignificantinteractionfurtherillustratesthatfalseclaimsattributedtothein-partywerenomorelikelytoberejectedintheanxietycondition(M=3.29,SE=.15)thanintheneutralcondition(M=3.52,SE=.12)whenacorrectionwasavailable,b=−.23(.20),t=−1.18,p=.24(seein-party,righthalfofFigure1).Thus,H1aisnotconfirmed.H1bpositedthatwhenmisinformationgoesuncorrected,anxiousindividualswhoreceivefalseclaimsfromtheout-partywillholdlessaccuratebeliefsthanthoseintheneutralcondition.Inthemisinformation-onlycondition,theinteractionbetweentheanxietyandpartisansourcevariablesispositiveandsignificant,b=.63(.24),t=2.61,p=.009(two-tailed).Probingthisinteractionindicatesthatbeliefsofanxiousindividualswhowereexposedtouncorrectedmisinformationattributedtotheout-party(M=2.52,SE=.15)weresignificantlylessaccuratethanthoseintheneutralcondition(M=2.94,SE=.11),b=−.42(.19),t=−2.25,p=.02(two-tailed)(seeout-party,lefthalfofFigure1).H1bisthereforesupportedasanxiety(comparedtoneutralemotionalstate)reducedbeliefaccuracywhenevaluatinguncorrectedmisinformationstemmingfromtheout-party.H2apredictedthatwhenacorrectionisavailable,beliefsofangryindividualsexposedtomisinformationfromthein-partywillbelessaccuratethanthosein710JournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation
B.E.WeeksEmotionsandMisperceptions12345Belief AccuracyIn-PartyMisinformation OnlyOut-PartyIn-PartyCorrectionOut-PartyNeutralAnxietyFigure1Predictedbeliefaccuracybyexperimentalconditionandanxietyinduction.Notes:Barsrepresentestimatedmarginalmeansbycondition,aftercontrollingforissue,theangercondition,andpoliticalknowledge.Beliefismeasuredona1–5scalewithhigherscorescorrespondingtomoreaccuracy.Errorbarsdenotestandarderror.aneutralemotionalstate.Thishypothesisistestedbyfocusingontheinteractionbetweentheangerdummyandpartisansourcevariableswithinthecorrectioncondition.Thenonsignificantcoefficientforthisinteractiondemonstratesthattheinfluenceofangerandpartisanshiparenotdependentononeanotherinthecor-rectioncondition,b=.15(.25),t=.60,p=.55.Additionalassessmentsofthedatafurtherindicatethatwhenacorrectionispresented,falseclaimsfromthein-partyarenomorelikelytobebelievedwhenparticipantsareangry(M=3.41,SE=.15)thanwhentheyareinaneutralemotionalstate(M=3.46,SE=.12),b=−.04(.19),t=−.23,p=.82(seein-party,righthalfofFigure2).Asaresult,H2aisnotsupported.H2bpredictedthatwhenmisinformationisnotcorrected,angryindividualswhoreceivefalseclaimsfromthein-partywillholdlessaccuratebeliefsthanindividualsinaneutralemotionalstate.Inthemisinformationonlycondition,theinteractionbetweentheangerandpartisansourcevariablesisnegativeandsignificant,b=−.64(.23),t=−2.73,p=.01(two-tailed),suggestingthatangerenhancestheeffectofpartisanshiponbelief.Probingtheinteractionrevealedthatbeliefaccuracyforangryparticipantsexposedtouncorrectedmisinformationfromthein-party(M=2.45,SE=.13)wassignificantlylowerthanthoseintheneutralcondition(M=2.83,SE=.11),b=−.38(.18),t=−2.14,p=.03(two-tailed)(seein-party,lefthalfofFigure2).ThefindingsthereforesupportH2b;angerreducedbeliefaccuracywhenevaluatinguncorrected,party-consistentmisinformation.DiscussionEmotionsplayanimportantroleinhowpeoplerespondtouncorrectedpoliticalmisinformation.ThisstudyprovidescausalevidencethattheindependentexperienceJournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation711
EmotionsandMisperceptionsB.E.Weeks12345Belief AccuracyOut-PartyIn-PartyMisinformation OnlyIn-PartyOut-PartyCorrectionNeutralAngerFigure2Predictedbeliefaccuracybyexperimentalconditionandangerinduction.Notes:Barsrepresentestimatedmarginalmeansbycondition,aftercontrollingforissue,theanxietycondition,andpoliticalknowledge.Beliefismeasuredona1–5scalewithhigherscorescorrespondingtomoreaccuracy.Errorbarsdenotestandarderror.ofangerandanxietyattimeshasdifferentconsequencesforpoliticalmisperceptionsbyheighteningordampeningtheinfluenceofpartisanship.Wheninitiallyfacedwithinaccurateclaimsaboutpolitics,angrypeoplearemorelikelytoprocesstheinformationinapartisanmanner,whichresultsinbeliefsthatreinforcetheirpartyaffiliation.Incontrast,anxietyreducestherelianceonpartisanshipatthisinitialstageandleadstobeliefsthatareconsistentwithinformationcontainedinthemessage.Inthisstudypoliticalmisperceptionsappeartostemnotfrompartisanshipalone,butratherfromaninteractionbetweenemotions,partyidentification,andtheinformationenvironment.Relyingextensivelyonthetheoryofpartisanmotivatedreasoning,previousresearchhasshownthatpartisanshipisrelatedtoinaccuratepoliticalbeliefs(Nyhan&Reifler,2010;Weeks&Garrett,2014).However,todatethisworkhasnotfullyillustratedwhenpartyaffiliationimpactsmisperceptions.Thecurrentstudyprovidesoneexplanation.Theresultsheresuggestthatthepartisanmotivatedreasoningpro-cessmaybefacilitatedbyangerratherthananxietyorgeneralnegativeaffect,asangerresultedininitialbeliefsthatwereconsistentwithpartyidentification.Thisraisesthepossibilitythatthe“automaticaffectiveprocesses”thatarethoughttoleadtobiasedprocessingofinformationmightbelimitedtothediscreteemotionangerratherthanothernegativeemotionsorgeneralnegativeaffect(Taber&Lodge,2006,p.756).Itmaybethatpoliticalattitudesalonearenotenoughtodrivepartisanprocessingofmisinformation,butratherattitudesthataretiedtoangerorresentment.Weknowtherearelimitstopartisanmotivatedreasoning(Druckman,2012;MacKuenetal.,2010;Redlawsketal.,2010),andpeopledonotnecessarilybelievederogatoryclaimsabouttheothersidesimplybecausetheyareRepublicanorDemocrat.Rather,thecombinationofangerandpartisanshipmightbewhatleavesthemmisinformed.This712JournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation
B.E.WeeksEmotionsandMisperceptionsprocesscouldhelpexplainwhymanyRepublicansbelievefalseinformationaboutBarackObama.Forexample,64%ofRepublicanssaiditwas“probablytrue”thatObamawashidinginformationabouthisbirthplacein2013(Cassino,2013),whichroughlycoincideswiththepercentageofRepublicans(58%)whoexpressedangeratObamaduringthattimeperiod(Steinhauser,2013).Theinteractionofangerandpartisanshiptofacilitatebeliefinuncorrectedmisinformationisespeciallytroublinggiventhatangeralsodepressesinformationseekingandincreasesselectiveexposure(MacKuenetal.,2010;Valentinoetal.,2008).Angerthereforehasthepotentialtocreatemediadietsinwhichpeopleareprimarilyexposedtolike-mindedmessages,whichmayfurtherenhanceangerdirectedatpoliticalopponents(Slater,2007).Itwillbeimportantforfutureresearchtoexaminethepotentialreinforcinganddynamicinfluenceofanger.Theeffectsofanxietyprovideatheoreticalexplanationforwhypeopleoccasion-allybelievemisinformationthatisinconsistentwiththeirpartisanaffiliation(e.g.,DemocratswhobelieveObamawasnotbornintheUnitedStates).Anxiouspar-ticipants(vs.neutral)whoonlyreceivedmisinformationsawasignificantdecreaseinbeliefaccuracywhenexposedtoinaccurateinformationstemmingfromtheout-party.Partisanmotivatedreasoningwouldsuggesttheseindividualsshouldrejectthismisinformationbecauseitchallengestheirpartisanship.Yetthatisnotwhathap-pened.InsightsfromAffectiveIntelligencetheoryhelpusunderstandwhythisseem-inglycounterintuitiveresultshouldactuallybeexpected.Recallanxietyarisesoutofnovelpoliticalcircumstancesanddepressestheroleofpriorattitudesandincreasestheinfluenceofcontemporaryinformationinevaluations(MacKuenetal.,2010;Red-lawsketal.,2010).Inthepresentresearch,considerDemocratsaskedtowritesome-thingthatmakesthemanxiousaboutimmigrationreform.Thisexperiencedanxietymayhavesignaledthatsomethingiswrongwiththeirpositionortheirparty’sstance,whichshouldincreasethewillingnesstoconsidertheinformationathand(Marcusetal.,2011).WhentheseDemocratsaresubsequentlypresentedwithmisinformationoriginatingfromtheRepublicanPartytheydonotautomaticallyprocessthisinfor-mationinapartisanway.Instead,theirbeliefsareinfluencedmorebythecontentofthemessage.Theyaremorelikelytopaycloserattentiontowhattheout-partyhastosayand,whencorrectiveinformationisnotavailable,haveahigherprobabilityofbeingmisinformed.Althoughanxietyhastypicallybeendiscussedasanemotionthatfacilitatesdemocraticthinkingandopen-mindedness,thissuggestsaparadoxofanx-iety.Thatis,anxietypromotescriticalthinkingandlearningaboutpolitics,butitmaybackfireiftheinformationconsideredisinaccurateormisleading(Marcus,2002).Thisstudydoesprovidereasonstobehopefulaboutthenatureofpoliticalmis-perceptions.Correctionstomisinformationwereeffective,eveninthefaceofemo-tionalexperiencesandpartisanmotivations.Whethercorrectionsareeffectivehasbeenaprimaryquestiondrivingmisperceptionresearchandtheresultsofpriorworkaremixed.Howaretheseconflictingfindingstobereconciled?First,correctionstoattitude-consistentbeliefscanbeeffectiveaslongastheydon’trequireachangeinattitude(Ecker,Lewandowsky,Fenton,&Martin,2014).TheclaimsinthisstudyareJournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation713
EmotionsandMisperceptionsB.E.Weekslikelynotacriticalcomponentofpeople’spartisanidentity,soadmittingtheyarefalseafterseeingstrong,correctiveevidencedidnotrequirepeopletochangetheirfeelingsabouttheirpoliticalparty.Inthiscase,thecorrectionsmayupdatebeliefsbutattitudestowardthepoliticalpartyremainunchanged.Second,thecorrectionsmighthavebeensuccessfulhereduetoa“tippingpoint.”Evenstrongpartisanmotivatedreasonersarewillingtogiveupontheirpositionwhenexposedtoanabundanceofinformationtellingthemtheyarewrong(Redlawsketal.,2010).Giventhatthecorrectionspro-videdstrongargumentsforwhyeachoftheclaimswasfalse,itispossiblethatevenangryparticipantsforwhomthecorrectionswereinconsistentwiththeirpartisanshipcouldnolongerengageinmotivatedprocessing.Thismayalsohelpexplainwhysomepriorexperimentalresearchhasfailedtofindcorrectiveeffectsforcounterattitudinalmisperceptions.Inmanyofthosestudies,thecorrectioncomesintheformofabriefpassageembeddedinanewsarticle(e.g.,Nyhan&Reifler,2010)orasubtleidentifi-cationofthemisinformation(Garrett&Weeks,2013),whichmaynotofferenoughcontradictoryevidencetoreachatippingpoint.Whilebothofthesepossibilitiesareplausible,itwillbecriticalforfutureresearchtofurtheroutlinetheconditionsandprocessesrelatedtosuccessfulcorrectivemessages.Thestudy’sexperimentaldesignprovidestheopportunitytoexaminecausalfac-torsthatinfluencemisperceptionsbutitalsocreatessomelimitations.Notably,theresultsreportedherearebasedonasingleexperiment,makingreplicationessential.Thecontrolleddesignusingalarge,politicallyanddemographicallydiversesample,aswellastheuseofstimulisampling,suggestthefindingsarenotlimitedtoasinglepoliticalissueorparticularsubgroupbutfutureresearchshouldexaminetheseeffectsindifferentcontexts.Astrengthofthisstudyisthatitprovidesacausaltestsoftheuniqueeffectsofanxietyandangeronpartisanbias.However,thebenefitsofthisdesigncomewithsomedrawbacks.Theemotionalinductionsdidnotdirectparticipantstowriteaboutaparticularpartisantargetoftheiremotionrelatedtotheissue.Itispossiblethatparticipantsintheanxietyconditionnotedsomethingabouttheout-party’sstanceontheissuethatmadethemanxious.Thismayhavelimitedthestudy’sabilitytofindeffectsinthecorrectioncondition,giventhatanxietyaboutone’sownpartyorpositiononanissue,nottheopposition’s,attenuatesbias(Marcusetal.,2011).Thislimitationisatradeoffthatwasnecessarytotestthepossibilitythattheseemotionsuniquelyaffectpartisanmotivatedreasoning.Somescholarsarguethatemotionsareoftentheresultofpreexistingpoliticalattitudes(Ladd&Lenz,2008),soastringenttestofthetheoryrequiredthatangerandanxietybeinducedwithoutanyexplicitpartisancues.Inaddition,therewasconcernthatdirectingattentiontopartisanconsiderationsabouttheseissueswouldhaveinflatedangerintheanxietycondition,thuspreventingacleantestofthedistincteffectsofeachemotion.Soalthoughthewritingtaskcomeswithdisadvantages,itwasnecessarytoisolatetheeffectsofangerandanxietyfromeachother,aswellasfrompartisanship.Theemotioninductionraisesanothernotablelimitation.Participantsintheneu-tralconditionwerenotaskedtowriteaboutimmigrationorthedeathpenalty,which714JournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation
B.E.WeeksEmotionsandMisperceptionscreatesthepossibilitythatmeresalienceoftheissue,nottheexperienceofemotions,affectedtheresults.Itisalsopossiblethattheemotionalinductionscreatednega-tivethoughtsabouttheissue,whichdirectlyaffectedbeliefaccuracy.However,theevidencesuggeststhisthreatislimitedastherewasnomaineffectofemotiononbelief,indicatingthattheinductiontaskalonedidnotaffectperceptions.Inaddition,thedifferentialinteractiveeffectsofanxietyandangerareconsistentwiththeoryanddemonstratethattheuniqueexperienceoftheseemotions,notsimplywritingabouttheissue,influencedwhatpeoplebelieve.Anotherlimitationinvolvesthefalseclaimsused.Themisinformationherewasnotwidelyknownanddidnotreceivemediaattentionpriortofieldingthestudy.Itispossibletheeffectsobservedheremightbedifferentwithlong-standingmisper-ceptionsthathavebeeninthepublicsphereforyears.Studiesattemptingtocorrectwell-knownmisperceptionshaveshownthetasktobemoredifficult(e.g.,Nyhan&Reifler,2010)andtheformatofthecorrectionsusedheremayworkdifferentlywithprominentmisperceptionsintherealworld.Despitetheselimitations,thisstudyoffersauniquetheoreticalcontributiontothestudyofpoliticalmisperceptions.Usingahighlycontrolledexperimentaldesign,thisresearchdirectlyaddressesthequestionofwhypeopleholdpoliticalmisperceptionsbyprovidingcausalevidencethattwonegativeemotions—anxietyandanger—uniquelyworkinconnectionwithotherfactorstoinfluencepeople’sresponsestoinaccurateinformation.Itisclearfromthisworkthatpoliticalbeliefsareformednotbyasingleconsiderationbutratherbyasetofinteractinginfluencesincludingemotion,partisanship,andtheinformationenvironment.Examininghowthesecomponentsworktogether—notinisolation—willbetterserveusasweseektounderstandtheproblemofpoliticalmisperceptions.AcknowledgmentsTheauthorthanksKellyGarrettforhisguidanceonthisprojectandalsoDaveEwoldsen,ArielHasell,LanceHolbert,ShannonMcGregor,andEmilyMoyer-Guséfortheircommentsandsuggestionsatvariousstages.ThisresearchwassupportedbytheNationalScienceFoundationunderGrantIIS-1149599,awardedtoKellyGarrett.Anyopinions,findings,andconclusionsorrecommendationsexpressedinthismaterialarethoseoftheauthoranddonotnecessarilyreflecttheviewsoftheNationalScienceFoundation.Notes1Redlawsk,Tolbert,&McNeely(2014)findfearexacerbatedeffectsofsymbolicracismonevaluationsofObamaandasimilarmechanismmayexistwithDemocrats’misperceptionsaboutObama.2Brader(2005)demonstratedexperimentallythatanxietycanaltercandidatepreferences.However,hisstudywasnotdirectlyinterestedinemotions’effectsonpartisanmotivatedreasoning,nordidheexaminetheeffectsofanger.JournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation715
EmotionsandMisperceptionsB.E.Weeks3Someevidenceindicatesthatsurveyresponsestofactualquestionsdonotalwaysrepresenttruebeliefs,butratherareexpressionsofpartisanloyalty(e.g.Bullock,Gerber,Hill,&Huber,2013;Prior,Sood,&Khanna,2013).Althoughthissuggeststhatpartisandifferencesinfactualperceptionsmaybeoverstated,researchindicatesthatself-reportedmisperceptions,whethertruebeliefsornot,canindependentlyaffectdemocraticoutcomessuchassupportforpoliticalpolicies(Kuklinskietal.,2000)andvotechoice(Weeks&Garrett,2014).4Thestudydoesnottakeintoaccountparticipants’priorknowledgeontheseissues,whichcouldaffectbeliefaccuracy.Forexample,beingknowledgeableabouttheseissuesshouldimproveaccuracy,anditispossiblethatparticipantswhoreportedaccuratebeliefsdidsobecausetheyknewtherightanswerfromthestartandnotbecausethecorrectionswereeffective.However,giventhattheexperimentaldesignincorporatedrandomassignment,anybiasgeneratedfrompriorknowledgeshouldbeevenlydistributedthroughoutthesample.5ItispossiblethatrespondentswouldnotbelievescenariosinwhichDemocratsspreadmisinformationaboutimmigrationorRepublicanspromotefalseclaimsaboutthedeathpenalty,giveneachparty’sgeneralpositiononthoseissues.Twoanalysesofvariance(controllingfortheemotionandcorrectionconditionsandpoliticalknowledge)demonstrateamaineffectofparticipants’politicalaffiliation,asRepublicansheldlessaccuratebeliefsthanDemocratsaboutimmigration(F(1,270)=6.62,p=.01),whileDemocratswerelessaccuratethanRepublicansaboutthedeathpenalty(F(1,275)=4.52,p=.03).Howevertherewasnomaineffectofthepoliticalpartyofthesourceofthemisinformationonbeliefforeitherimmigration(F(1,270)=1.44,p=.23)orthedeathpenalty(F(1,275)=1.33,p=.25).Moreimportantly,therewasnointeractionbetweentheparticipants’partyaffiliationandthepartisansourceofthemisinformationforeitherissue(immigration:F(1,270)=2.38,p=.12;deathpenalty:F(1,275)=2.28,p=.13).ThisindicatesthatneitherRepublicans’norDemocrats’beliefsabouttheseissuesdependedonthepoliticalpartyattributedasthesourceofthemisinformation.ReferencesBrader,T.(2005).Strikingaresponsivechord:Howpoliticaladsmotivateandpersuadevotersbyappealingtoemotions.AmericanJournalofPoliticalScience,49,388–405.doi:10.1111/j.0092-5853.2005.00130.x.Brader,T.(2011).Thepoliticalrelevanceofemotions:“Reassessing”revisited.PoliticalPsychology,32,337–346.doi:10.1111/j.1467-9221.2010.00803.x.Bullock,J.G.(2009).PartisanbiasandtheBayesianidealinthestudyofpublicopinion.JournalofPolitics,71,1109–1124.doi:10.1017/S0022381609090914.Bullock,J.G.,Gerber,A.S.,Hill,S.J.,&Huber,G.A.(2013).Partisanbiasinfactualbeliefsaboutpolitics.WorkingPaper19080,NationalBureauofEconomicResearch.Retrievedfromhttp://www.nber.org/papers/w19080Carver,C.S.,&Harmon-Jones,E.(2009).Angerisanapproach-relatedaffect:Evidenceandimplications.PsychologicalBulletin,135,183–204.doi:10.1037/a0013965.Cassino,D.(2013,January17).Conspiracytheoriesprosper:25%ofAmericansare“truthers.”FairleighDickinsonUniversity’sPublicMindPoll.Retrievedfromhttp://www.scribd.com/doc/120815791/Fairleigh-Dickinson-poll-on-conspiracy-theories716JournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation
B.E.WeeksEmotionsandMisperceptionsCoe,K.,Kenski,K.,&Rains,S.A.(2014).Onlineanduncivil?Patternsanddeterminantsofincivilityinnewspaperwebsitecomments.JournalofCommunication,64,658–679.doi:10.1111/jcom.12104.Druckman,J.N.(2012).Thepoliticsofmotivation.CriticalReview:AJournalofPoliticsandSociety,24,199–216.doi:10.1080/08913811.2012.711022.Ecker,U.K.H.,Lewandowsky,S.,Fenton,O.,&Martin,K.(2014).Dopeoplekeepbelievingbecausetheywantto?Pre-existingattitudesandthecontinuedinfluenceofmisinformation.Memory&Cognition,42,292–304.doi:10.3758/s13421-013-0358-x.Eysenck,M.W.,Derakshan,N.,Santos,R.,&Calvo,M.G.(2007).Anxietyandcognitiveperformance:Attentioncontroltheory.Emotion,7,336–353.doi:10.1037/1528-3542.7.2.336.Frijda,N.H.(1986).Theemotions.NewYork,NY:CambridgeUniversityPress.Gaines,B.J.,Kuklinski,J.H.,Quirk,P.J.,Peyton,B.,&Verkuilen,J.(2007).Samefacts,differentinterpretations:PartisanmotivationandopiniononIraq.JournalofPolitics,69,957–974.doi:10.1111/j.1468-2508.2007.00601.x.Gallup.(2011,May13).Obama’sbirthcertificateconvincessome,butnotall,skeptics.Retrievedfromhttp://www.gallup.com/poll/147530/obama-birth-certificate-convinces-not-skeptics.aspxGarrett,R.K.(2011).Troublingconsequencesofonlinepoliticalrumoring.HumanCommunicationResearch,37,255–274.doi:10.1111/j.1468-2958.2010.01401.x.Garrett,R.K.,Nisbet,E.C.,&Lynch,E.K.(2013).Underminingthecorrectiveeffectsofmedia-basedpoliticalfactchecking?Theroleofcontextualcuesandnaïvetheory.JournalofCommunication,63,617–637.doi:10.1111/jcom.12038.Garrett,R.K.,&Weeks,B.E.(2013).Thepromiseandperilofreal-timecorrectionstopoliticalmisperceptions.PaperpresentedattheProceedingsoftheACM2013ConferenceonComputerSupportedCooperativeWork(CSCW)(pp.1047–1058),SanAntonio,TX.NewYork,NY:ACM.Goodall,C.E.,Slater,M.D.,&Myers,T.A.(2013).Fearandangerresponsestolocalnewscoverageofalcohol-relatedcrimes,accidents,andinjuries:Explainingnewseffectsonpolicysupportusingarepresentativesampleofmessagesandpeople.JournalofCommunication,63,373–392.doi:10.1111/jcom.12020.Gottfried,J.A.,Hardy,B.W.,Winneg,K.M.,&Jamieson,K.H.(2013).Didfactcheckingmatterinthe2012presidentialcampaign?AmericanBehavioralScientist,57,1558–1567.doi:10.1177/0002764213489012.Hayes,A.F.(2013).Introductiontomediation,moderation,andconditionalprocessanalysis.NewYork,NY:TheGuilfordPress.Huddy,L.,Feldman,S.,&Cassese,E.(2007).Onthedistinctpoliticaleffectsofanxietyandanger.InW.R.Neuman,G.E.Marcus,A.N.Crigler,&M.Mackuen(Eds.),Theaffecteffect:Dynamicsofemotioninpoliticalthinkingandbehavior(pp.202–230).Chicago,IL:UniversityofChicagoPress.Jerit,J.,&Barabas,J.(2012).Partisanperceptualbiasandtheinformationenvironment.TheJournalofPolitics,74,672–684.doi:10.1017/S0022381612000187.Kuklinski,J.H.,Quirk,P.J.,Jerit,J.,Schwieder,D.,&Rich,R.F.(2000).Misinformationandthecurrencyofdemocraticcitizenship.TheJournalofPolitics,62,790–816.doi:10.1111/0022-3816.00033.Kunda,Z.(1990).Thecaseformotivatedreasoning.PsychologicalBulletin,108,480–498.doi:10.1037/0033-2909.108.3.480.JournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation717
EmotionsandMisperceptionsB.E.WeeksLadd,J.M.,&Lenz,G.S.(2008).Reassessingtheroleofanxietyandvotechoice.PoliticalPsychology,29,275–296.Lazarus,R.S.(1991).Emotionandadaptation.NewYork,NY:OxfordUniversityPress.Lerner,J.S.,&Keltner,D.(2001).Fear,anger,andrisk.JournalofPersonalityandSocialPsychology,81,146–159.doi:10.1037/0022-3514.81.1.146.MacKuen,M.,Wolak,J.,Keele,L.,&Marcus,G.E.(2010).Civicengagements:Resolutepartisanshiporreflectivedeliberation.AmericanJournalofPoliticalScience,54,440–458.doi:10.1111/j.1540-5907.2010.00440.x.Marcus,G.E.(2002).Thesentimentalcitizen:Emotionindemocraticpolitics.UniversityPark:ThePennsylvaniaStateUniversityPress.Marcus,G.E.,MacKuen,M.,&Neuman,W.R.(2011).Parsimonyandcomplexity:Developingandtestingtheoriesofaffectiveintelligence.PoliticalPsychology,32,323–336.doi:10.1111/j.1467-9221.2010.00806.x.Marcus,G.E.,Neuman,W.R.,&MacKuen,M.(2000).Affectiveintelligenceandpoliticaljudgment.Chicago,IL:UniversityofChicagoPress.Marcus,G.E.,Sullivan,J.T.,Theiss-Morse,E.,&Stevens,D.(2005).Theemotionalfoundationofpoliticalcognition:Theimpactofextrinsicanxietyontheformationofpoliticaltolerancejudgments.PoliticalPsychology,26,949–963.doi:10.1111/j.1467-9221.2005.00452.x.Meirick,P.C.(2013).Motivatedmisperception?Party,education,partisannews,andbeliefin“deathpanels”.Journalism&MassCommunicationQuarterly,90,39–57.doi:10.1177/1077699012468696.Nabi,R.L.(1999).Acognitive-functionalmodelfortheeffectsofdiscretenegativeemotionsoninformationprocessing,attitudechange,andrecall.CommunicationTheory,9,292–320.doi:10.1111/j.1468-2885.1999.tb00172.x.Nabi,R.L.(2003).Exploringtheframingeffectsofemotion:Dodiscreteemotionsdifferentiallyinfluenceinformationaccessibility,informationseeking,andpolicypreference?CommunicationResearch,30,224–247.doi:10.1177/0093650202250881.Nabi,R.L.(2010).Thecaseforemphasizingdiscreteemotionsincommunicationresearch.CommunicationMonographs,77,153–159.doi:10.1080/03637751003790444.Nyhan,B.,&Reifler,J.(2010).Whencorrectionsfail:Thepersistenceofpoliticalmisperceptions.PoliticalBehavior,32,303–330.doi:10.1007/s11109-010-9112-2.Papacharissi,Z.(2012).Withoutyou,I’mnothing:PerformancesoftheselfonTwitter.InternationalJournalofCommunication,6,1989–2006.Prior,M.,Sood,G.,&Khanna,K.(2013).Youcannotbeserious:Dopartisansbelievewhattheysay?WorkingPaper.Retrievedfromhttp://government.arts.cornell.edu/assets/psac/sp13/Prior_PSAC_Apr15.pdfRedlawsk,D.P.,Civettini,A.J.,&Emmerson,K.M.(2010).Theaffectivetippingpoint:Domotivatedreasonersever“getit”?PoliticalPsychology,31,563–593.doi:10.1111/j.1467-9221.2010.00772.x.Redlawsk,D.P.,Tolbert,C.J.,&McNeely,N.A.(2014).Symbolicracismandemotionalresponsestothe2012presidentialcandidates.PoliticalResearchQuarterly,67,680–694.doi:10.1177/1065912914531091.Roseman,I.J.,Wiest,C.,&Swartz,T.S.(1994).Phenomenology,behaviors,andgoalsdifferentiatediscreteemotions.JournalofPersonalityandSocialPsychology,67,206–221.doi:10.1037/0022-3514.67.2.206.718JournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation
B.E.WeeksEmotionsandMisperceptionsSlater,M.D.(2007).Reinforcingspirals:Themutualinfluenceofmediaselectivityandmediaeffectsandtheirimpactonindividualbehaviorandsocialidentity.CommunicationTheory,17,281–303.doi:10.1111/j.1468-2885.2007.00296.x.Steinhauser,P.(2013,October7).CNNshutdownpoll:Plentyofblametogoaround.Retrievedfromhttp://politicalticker.blogs.cnn.com/2013/10/07/cnn-shutdown-poll-plenty-of-blame-to-go-around/Taber,C.S.,&Lodge,M.(2006).Motivatedskepticismintheevaluationofpoliticalbeliefs.AmericanJournalofPoliticalScience,50,755–769.doi:10.1111/j.1540-5907.2006.00214.x.Valentino,B.(2012).SurveyonforeignpolicyandAmericanoverseascommitments.YouGov.Retrievedfromhttp://www.dartmouth.edu/∼benv/data.html.Valentino,N.A.,Brader,T.,Groenendyk,E.W.,Gregorowicz,K.,&Hutchings,V.L.(2011).Electionnight’salrightforfighting:Theroleofemotionsinpoliticalparticipation.JournalofPolitics,73,156–170.doi:10.1017/S0022381610000939.Valentino,N.A.,Hutchings,V.L.,Banks,A.J.,&Davis,A.K.(2008).Isaworriedcitizenagoodcitizen?Emotions,politicalinformationseeking,andlearningviatheinternet.PoliticalPsychology,29,247–273.doi:10.1017/S0022381610000939.Weeks,B.E.,&Garrett,R.K.(2014).Electoralconsequencesofpoliticalrumors:Motivatedreasoning,candidaterumors,andvotechoiceduringthe2008U.S.presidentialelection.InternationalJournalofPublicOpinionResearch,26,401–422.doi:10.1093/ijpor/edu005.Weeks,B.,&Southwell,B.(2010).Thesymbiosisofnewscoverageandaggregateon-linesearchbehavior:Obama,rumors,andpresidentialpolitics.MassCommunicationandSociety,13,341–360.doi:10.1080/15205430903470532.Wells,G.L.,&Windschitl,P.D.(1999).Stimulussamplingandsocialpsychologicalexperimentation.PersonalityandSocialPsychologyBulletin,25,1115–1125.doi:10.1177/01461672992512005.SupportingInformationAdditionalsupportinginformationmaybefoundintheonlineversionofthisarticle:AppendixS1.(a)Misperceptionsaboutimmigrationmakereformharder.(b)Mispercep-tionsaboutthedeathpenaltyclouddebate.JournalofCommunication65(2015)699–719©2015InternationalCommunicationAssociation719