NYPDzhuo爱

ResearchArticleViolenceandSexinTelevisionProgramsDoNotSellProductsinAdvertisementsBradJ.
BushmanInstituteforSocialResearch,DepartmentofPsychology,andDepartmentofCommunicationStudies,UniversityofMichiganABSTRACT—Adults(N5336)18to54yearsoldwatchedatelevisionprogramcontainingviolence,sex,bothviolenceandsex,ornoviolenceandsex.
Programswereshowninacomfortableroomcontainingpaddedchairsandtastysnacks.
Eachprogramcontainedthesame12ads.
Em-beddinganadinaprogramcontainingviolenceorsexreduced(a)viewers'likelihoodofrememberingthead-vertisedbrand,(b)theirinterestinbuyingthatbrand,and(c)theirlikelihoodofselectingacouponforthatbrand.
Theseeffectsoccurredformalesandfemalesofallages,regardlessofwhethertheylikedprogramscontainingvi-olenceandsex.
Theseresultsshowthatviolenceandsexintelevisionprogramsdonotsellproductsinadvertisements.
Thenumberonepriorityintelevisionisnottotransmitqualityprogrammingtoviewers,buttodeliverconsumerstoadvertisers.
—JeffSagansky,formerCBSprogrammingchief(quotedinKim,1994,p.
1434)Mostnetworkandcabletelevisioncompaniesare100%sup-portedbyadrevenue.
Advertisersarewillingtopayalotofmoneyforairtime.
Forexample,advertiserspaid$2.
3millionforasingle30-sadairingduringSuperBowlXXXVIII(Sutel,2004).
Thegoalofadvertisingistopresentgoodsorservicesinaneffectivemannersoindividualswillpurchasethem.
Televi-sionisanidealmediumtoadvertisegoodsorservicesbecausesomanypeopleownTVsets.
IntheUnitedStates,therearemoreTVsetsthantherearetoilets(AmericanPsychologicalAssociation,1993).
Televisionreachesmoreofanadvertiser'spotentialcustomersthandoesanyothermedium,andadultsspendsignificantlymoretimewithtelevisionthanwithanyothermedium(TelevisionBureauofAdvertising,2003).
Itiscommonlyassumedthattelevisedviolenceandsexsellproducts.
AlthoughTVprogramswithviolenceandsexdoat-tractyoungerviewers,overalltheyattractfewerviewersthanprogramswithoutviolenceandsex(Hamilton,1998;ParentsTelevisionCouncil,2003).
Thesmalleraudiencesizereducesthepotentialimpactofanadvertisement.
ThisimpactwouldbefurtherreducedifTVviewerscannotremembertheproductbeingadvertised.
Researchhasalreadyshownthattelevisedviolenceandseximpairmemoryforadvertisedproducts.
Ameta-analysisof16studiesinvolving2,474participantsfoundthatmemoryforadvertisedbrandswas27%lowerifadswereembeddedinaviolentprogramthanifthesameadswereem-beddedinanonviolentprogram(Bushman,2003;seealsoBushman&Phillips,2001).
Televisedsexalsomayimpairmemoryforadvertisedbrands(Bushman&Bonacci,2002).
Aplausiblecauseofthismemoryimpairmentisthatindi-vidualshavealimitedamountofattentiontodirecttowardtele-visionprograms(Lang,Newhagen,&Reeves,1996).
Researchsuggeststhatindividualspaymoreattentiontoviolentmediathantononviolentmedia(Furnham&Gunter,1987;Langetal.
,1996;Williamson,Kosmitzki,&Kibler,1995).
Individualsalsopaymoreattentiontosexualmediathantononsexualmedia(e.
g.
,Geer,Judice,&Jackson,1994;Geer&McGlone,1990).
Fromanevolutionaryperspective,thismakessense.
Ouran-cestorswhopaidattentiontoviolentandsexualcuesweremorelikelytosurviveandpassontheirgenesthanwerethosewhoignoredviolentandsexualcues.
Themoreattentionviewerspaytosexualorviolentprograms,thelessattentiontheyhaveavailableforthecommercialsembeddedinthoseprograms.
Eventhoughcommercialsoccurduringbreaksinaprogram,viewersmaystillbedevotingcognitiveresourcestothinkingabouttheprogramthatattractedtheirattentionsostrongly.
Thepresentstudyprovidesanimportantextensionoverpre-viousstudiesbecauseittestedwhethertelevisedviolenceandsexalsoinuencebuyingintentionsandconsumerbehaviors.
Thisresearchisimportantfortheoreticalandpracticalreasons.
AddresscorrespondencetoBradJ.
Bushman,InstituteforSocialResearch,426ThompsonSt.
,AnnArbor,MI48106;e-mail:bbushman@umich.
edu.
PSYCHOLOGICALSCIENCE702Volume16—Number9Copyrightr2005AmericanPsychologicalSocietyInthetheoryofreasonedaction,behavioralintentionsareviewedasanimportantmediatorofbehavior(e.
g.
,Ajzen&Fishbein,1977).
Thepresentstudyincludedameasureofbuyingintentions,whicharerelatedtoconsumerbehaviors(e.
g.
,Bagozzi,Baumgartner,&Yi,1991,1992;Shimp&Kavas,1984).
Specifically,thisstudytestedwhetherbuyingintentionsme-diatetheeffectofviolenceandsexinaTVprogramonthechoiceofacouponforanadvertisedproduct.
In2000,morethan77%ofAmericansusedcoupons,whichsavedthemmorethan$3.
6billion(CouponUsageStatistics,2000).
Ofcourse,thevalueofthegoodstheypurchasedwasmuchgreaterthanthefacevalueofthecouponstheyused.
Althoughcouponchoicesarenottechnicallyconsumerbehaviors(i.
e.
,theconsumerhastore-deemthecoupon),numerousstudieshaveshownthatcouponsincreaseproductsales(e.
g.
,Ailawadi,Lehmann,&Neslin,2001;Bawa&Shoemaker,1989;Lam,Vandenbosch,Hulland,&Pearce,2001;Schindler,1992).
Couponredeemersarealsoover7timesmorelikelytomakerepeatpurchasesthanarenonredeemers(Taylor,2001).
Insummary,thepresentstudytestedwhethertelevisedvio-lenceandsexreducememoryforanadvertisedproduct,in-tentionstobuythatproduct,andthelikelihoodofchoosingacouponforthatproduct.
Italsotestedwhethermemoryforbrandsisrelatedtobuyingintentionsandcouponchoices.
Televisedviolenceandsexwereexpectedtohaveanindirecteffectoncouponchoices,withbrandmemoryandbuyingin-tentionsservingasimportantmediators.
METHODParticipantsParticipantswere336adults18to54yearsold.
Theiragedis-tributionwasrepresentativeoftheagedistributionof18-to54-year-oldU.
S.
adultswholiveinhouseholdswithTV(i.
e.
,17%ages18–24,12%ages25–29,14%ages30–34,15%ages35–39,16%ages40–44,14%ages45–49,and13%ages50–54;NielsenMediaResearch,2000).
Theseagegroupsarethemostimportanttoadvertisers(Hamilton,1988).
ParticipantswererecruitedusingnewspaperadvertisementsincentralIowa(e.
g.
,DesMoines,Ames).
TheywerewarnedthatsomeoftheTVprogramscontainedviolenceandsex.
Partici-pantsreceived$20inexchangefortheirvoluntaryparticipation.
ProcedureParticipantsweretestedinsmallgroups,buteachworkedin-dependently.
TheyweretoldthattheresearcherswerestudyingattitudestowardTVprograms.
Thesessionswereconductedinacomfortablesetting.
Participantswereseatedinpaddedchairsandweregivensoftdrinksandsnacks(e.
g.
,potatochips,pret-zels,cookies).
Aftergivingtheirinformedconsent,participantswereran-domlyassignedtowatchaTVprogramthatcontainedviolence,sex,violenceandsex,ornoviolenceandsex.
Therewere84participants(42men,42women)ineachofthefourgroups.
Sothatthefourtypesofprogramswouldbeadequatelysampled,sixexemplarsofeachwereused(Wells&Windschitl,1999).
Adiewasrolledtodeterminewhatprogramtoshow.
Thesixviolentprogramswere''24,''''Cops,''''ProtectandServe,''''TerrorontheJob,''''TourofDuty,''and''X-Files.
''Alloftheseprogramshadaviolent(V)contentcode;nonehadasex(S)contentcode.
Thesixsexuallyexplicitprogramswere''AllyMcBeal,''''HowardStern,''''ManShow,''''SexintheCity,''''WildonE,''and''WillandGrace.
''Alloftheseprogramshadasex(S)contentcode;nonehadaviolent(V)contentcode.
Thesixviolentandsexualprogramswere''Angel,''''BuffytheVampireSlayer,''''CSIMi-ami,''''NYPDBlue,''''SouthPark,''and''WorldWrestlingEn-tertainmentRAW.
''Alloftheseprogramshadviolent(V)andsex(S)contentcodes.
Thesixneutralprogramswere''ADatingStory,''''America'sFunniestAnimals,''''HollywoodBeyondtheStars,''''It'saMiracle,''''MiraclePets,''and''TradingSpaces.
''Noneoftheseprogramshadviolent(V)orsex(S)contentcodes.
AlloftheprogramscontainingviolenceorsexwereratedTV-14(ParentsStronglyCautioned).
AlloftheneutralprogramswereratedTV-G(GeneralAudience).
Theadsoriginallyembeddedintheprogramswereeditedout,andthesame12(30-s)adswereembeddedineachprogram.
Becauseconsumersarequiteloyaltobrands(e.
g.
,Chaudhuri,2001),weusedadsforrelativelyunfamiliarbrands(i.
e.
,BodyFantasies,Dermoplast,FerraroRaffaelloChocolates,JoseOle,LibmanNittyGrittyRollerMop,Mederma,NatraTaste,NewSkin,NutraNails,Proheart6,SenokotNaturalVegetableLaxative,andSuddenChangeUndereyeLift).
Therewerethreecommercialbreaksatap-proximately12min,24min,and36minintoeachprogram.
Fouradswerepresentedineachbreak.
Tworandomordersofadswereused.
ImmediatelyafterviewingtheTVprogram,participantsratedhowexciting,boring,involving,humorous,violent,andsexuallyarousingtheythoughtitwas,usingascalerangingfrom1(notatall)to10(extremely).
Theviolenceandsexualarousalratingswereusedasmanipulationchecksfortheprogramcontentcodes.
TheotherratingswereusedascovariatestocontrolfordifferencesamongTVprogramsotherthandifferencesinviolentandsexualcontent.
AfterratingtheTVprogram,participantsreceivedsur-prisememorytests.
First,theylistedasmanybrandnamesastheycouldrecall.
Second,theyreceivedalistofproducts.
Foreachtypeofproduct(e.
g.
,sugarsubstitute),fourbrandswerelisted—theadvertisedbrand(e.
g.
,NatraTaste)andthreefoilbrands(e.
g.
,SugarTwin,SweetDeal,Sweet'n'Smart).
Aswiththeactualbrands,foilbrandsthatwouldberelativelyunfamiliartomostparticipantswereselected.
Foreachkindofproduct,participantscircledthebrandthatwasadvertised.
Volume16—Number9703BradJ.
BushmanNext,participantsreportedtheirbuyingintentionsusingthesamelist.
Foreachtypeofproduct,theycircledthebrandthey''wouldbemostlikelytobuy.
''Thenumberofcouponsselectedforadvertisedproductswasusedasameasureofconsumerbehavior.
Couponsforall40brandsonthelistwereavailable,andparticipantscouldchoose10.
Becauseinadequatefacevaluesoncouponscanreducetheireffectiveness(Chakraborty&Cole,1991;Cheong,2003),allofthecouponsofferedparticipants$1offthelistedprice—asub-stantialsavingsfortheproductsusedinthisstudy,andabout30centshigherthantheaveragecouponfacevalue(PromotionMarketingAssociation,2000).
Professionallyprinted(bogus)couponswereavailablefortheadvertisedandfoilbrands.
Next,participantsreportedtheirdemographics(i.
e.
,age,gender,ethnicbackground)andtheirtelevisionviewinghabits(i.
e.
,thenumberofhourseachweektheyspentwatchingTVandthepercentageoftimetheyspentwatchingviolentandsexualprograms).
Thelattermeasureswereusedtocontrolforhabitualexposuretotelevisedviolenceandsex.
Participantsalsore-portediftheyhadseentheassignedprogrambefore(thatepi-sodeoranyepisodes),iftheyhadseenanyoftheadsbefore,andiftheyhadheardofanyoftheadvertisedproductsbefore.
TheseresponseswereusedtocontrolforpreviousexposuretotheTVprogramsandads.
Finally,participantswerefullydebriefed.
Becausethecouponswerebogus,wegaveparticipantsthefacevalueofthecouponsinstead(i.
e.
,$10).
Thus,eachparticipantreceived$30($20forparticipatinginthestudyand$10fortheboguscoupons).
RESULTSPreliminaryAnalysesandManipulationChecksViolenceRatingsTypeofTVprogramhadasignificanteffectonviolenceratings,F(3,332)5173.
56,p.
05.
Thus,thedatawerecollapsedacrossexemplarsofprogramtypesforsubsequentanalyses.
DemographicDataThedemographicvariables(i.
e.
,sex,age,ethnicbackground)hadnomaineffectsonanyoftheoutcomemeasuresandwereTABLE1MeanViolenceandSexRatings,Memory,BuyingIntentions,andConsumerBehaviorsintheFourConditionsVariableConditionNeutralViolenceSexViolenceandsexViolencerating(1–10)1.
31b(0.
20)6.
12a(0.
20)1.
61b(0.
20)5.
82a(0.
20)Sexrating(1–10)1.
05b(0.
15)1.
43b(0.
15)3.
60a(0.
15)3.
42a(0.
15)NumberofbrandsrecalledAdjusted1.
21a(0.
12)0.
91ab(0.
12)0.
52c(0.
12)0.
72bc(0.
12)Unadjusted1.
18a(0.
12)0.
95ab(0.
12)0.
52c(0.
12)0.
70bc(0.
12)NumberofbrandsrecognizedAdjusted5.
76a(0.
27)4.
36b(0.
28)3.
54c(0.
28)3.
58c(0.
27)Unadjusted5.
70a(0.
28)4.
61b(0.
28)3.
44c(0.
28)3.
50c(0.
28)BuyingintentionsAdjusted5.
61a(0.
26)4.
15b(0.
26)4.
13b(0.
26)4.
16b(0.
26)Unadjusted5.
48a(0.
26)4.
23b(0.
26)4.
23b(0.
26)4.
13b(0.
26)CouponchoicesAdjusted5.
75a(0.
36)4.
31b(0.
37)4.
54b(0.
36)4.
10b(0.
36)Unadjusted5.
50a(0.
36)4.
41b(0.
36)4.
74ab(0.
36)4.
05b(0.
36)Note.
Standarderrorsareinparentheses.
AdjustedmeanswereadjustedforwhetherparticipantshadseentheTVprogram(theepisodeshownaswellasotherepisodes);whethertheyhadseentheadsbefore;whethertheywerefamiliarwiththeproductsadvertised;howexciting,boring,humorous,andinvolvingtheyratedtheTVprogram;howmanyhourstheyspentwatchingTVperweek;andthepercentageoftimetheyspentwatchingtelevisedsexandviolence.
Unadjustedmeanswerenotadjustedforanycovariates.
''Buyingintentions''referstothenumberofadvertisedproductsparticipantsintendedtobuy,and''couponchoices''referstothenumberofadvertisedproductswhosecouponswerechosen.
Withinarow,meanssharingthesamesubscriptarenotsignificantlydifferentatthe.
05significancelevel.
704Volume16—Number9ViolenceandSexDon'tSellnotinvolvedinanysignificantinteractions,ps>.
05.
Thus,thedatawerecollapsedacrossdemographicgroupsforsubsequentanalyses.
RandomOrdersofAdsTwodifferentrandomordersofadswereused.
Becauseran-domizationorderdidnotaffectanyoftheoutcomemeasures(ps>.
05),thedatawerecollapsedacrossthetwoordersofads.
FamiliarityWithAdsandBrandsAsexpected,mostparticipantshadnotseentheadsbeforeandwerenotfamiliarwiththeadvertisedbrands(meannumberofadsseenbefore51.
61,SE50.
13;meannumberoffamiliarbrands51.
93,SE50.
15).
PrimaryAnalysesAnalysisofcovariance(ANCOVA)wasusedtotesttheeffectsoftelevisedviolenceandsexontheoutcomemeasures.
CovariatesincludedwhetherparticipantshadseentheTVprogram(theepisodeshownaswellasotherepisodes);whethertheyhadseentheadsbefore;whethertheywerefamiliarwiththeadvertisedbrands;howexciting,boring,humorous,andinvolvingtheyratedtheTVprogram;howmanyhourstheyspentwatchingTVperweek;andthepercentageoftimetheyspentwatchingtele-visedviolenceandsex.
Theresults,however,weresimilarwhennocovariateswereincludedintheanalyses(seeTable1).
IfasignificanteffectwasfoundfortypeofTVprogram,aplannedcontrastwasperformedtotestwhetherparticipantswhosawaprogramwithoutviolenceandsexrespondeddifferentlyfromparticipantswhosawaprogramwithviolenceorsex.
Onthebasisofpreviousresearch,nodifferencesbetweenviolentandsexualprogrammingwerepredicted(Bushman&Bonacci,2002).
Thus,therespectivecontrastweightsfortheneutral,violent,sexual,andviolentsexualprogramswere3,11,11,and11.
Posthocttestswerealsoconductedtoexplorepossiblediffer-encesbetweenviolentandsexualprogramming(seeTable1).
CouponChoicesTypeofTVprogramsignificantlyinuencedwhetherpartici-pantsselectedcouponsforadvertisedbrands,F(3,320)54.
16,p<.
007(seeTable1).
Participantswhowatchedaprogramwithoutviolenceandsexselected33%morecouponsforadvertisedbrandsthandidparticipantswhowatchedaprogramwithviolenceorsex,F(1,320)511.
05,p<.
001,rpb5.
14.
1Posthoccomparisonsshowedsimilareffectsforviolentandsexualprogrammingoncouponchoices(seeTable1).
Onlyonecovariatewassignificant.
Themoretelevisionparticipantswatched,themorelikelytheyweretochoosecouponsforadvertisedbrands,F(1,320)54.
20,p<.
05,r5.
16.
BuyingIntentionsForeachtypeofproduct,participantsindicatedthebrandtheyweremostlikelytobuy.
TypeofTVprogramsignificantlyin-uencedbuyingintentions,F(3,320)58.
00,p<.
0001(seeTable1).
Participantswhowatchedaprogramwithoutviolenceandsexselected35%moreoftheadvertisedbrandsthandidparticipantswhowatchedaprogramwithviolenceorsex,F(1,320)523.
35,p<.
0001,rpb5.
23.
Posthoccomparisonsshowedsimilareffectsforviolentandsexualprogrammingonbuyingintentions(seeTable1).
Noneofthecovariatesweresignificant.
RecallTypeofTVprogramsignificantlyinuencedbrandrecall,F(3,320)56.
51,p<.
0003(seeTable1).
Brandrecallwas68%higherforpeoplewhosawaprogramwithoutviolenceandsexthanforpeoplewhosawaprogramwithviolenceorsex,F(1,320)513.
98,p<.
0002,rpb5.
18.
Therecallimpairmenttendedtobelargerforsexualprogrammingthanforviolentprogramming(seeTable1).
Onlyonecovariatewassignificant.
Brandrecallwaspositivelycorrelatedwiththenumberofad-vertisedproductsparticipantswerefamiliarwith,F(1,320)56.
94,p<.
009,r5.
27.
RecognitionTypeofTVprogramsignificantlyinuencedbrandrecognition,F(3,320)514.
25,p<.
0001(seeTable1).
Brandrecognitionwas51%higherforparticipantswhosawaprogramwithoutviolenceandsexthanforparticipantswhosawaprogramwithviolenceorsex,F(1,320)536.
62,p<.
0001,rpb5.
29.
Therecognitionimpairmenttendedtobelargerforsexualpro-grammingthanforviolentprogramming(seeTable1).
Brandrecognitionwasalsopositivelycorrelatedwiththenumberofproductsparticipantswerefamiliarwith,F(1,320)513.
71,p<.
0003,r5.
31.
TheRoleofMemoryandBuyingIntentionsinCouponChoicesStructuralequationmodelswerecomputedwithAMOSusingmaximumlikelihoodestimation(Arbuckle,1999).
Thehy-pothesizedmodelwasthatbrandmemoryandbuyingintentionsmediatetheeffectofTVviolenceandsexoncouponchoices.
Inspecifyingthemodel,adummyvariablewasusedtorepresentTVviolenceandsex(violentorsexualprogram51,neutralprogram50).
Brandmemorywastreatedasalatentvariable,measuredusingbrandrecallandrecognition.
Thevariance-covariancematrixusedfortheanalysesisgiveninTable2.
Thehypothesizedmediationmodeltthedataextremelywell,w2(5,N5336)54.
76,p<.
45,GFI(goodness-of-tindex)5.
995,1ThecomparativepercentagesreportedhereandlaterinResultswerecal-culatedbysummingthevaluesfortheparticipantswhowatchedviolentpro-gramming,sexualprogramming,andviolentsexualprogramming;dividingby3;andthendividingtheresultintothevaluefortheparticipantswhowatchedneutralprogramming:neutral/[(violence1sex1violenceandsex)/3].
Forex-ample,usingthedatainTable1,onecanseethatthenumberofadvertisedbrandswhosecouponswerechosenwas33%higherforparticipantswhosawaneutralprogramthanfortheothergroups:5.
75/[(4.
3114.
5414.
10)/3]51.
33,or33%higher.
Volume16—Number9705BradJ.
BushmanCFI(comparativetindex)51,RMSEA(rootmeansquareerrorofapproximation)5.
005.
AscanbeseeninFigure1,TVvio-lenceandseximpairedbrandmemory.
Brandmemoryandbuyingintentionswereimportantmediators.
Peopleintendedtobuybrandstheycouldremember,andtheychosecouponsforbrandstheyintendedtobuy.
AnalternativemodelincludeddirectpathsfromTVviolenceandsextobuyingintentionsandtocouponchoice.
Becausethemediationmodelisnestedwithinthisalternativemodel,achi-squaredifferencetestcanbeusedtocomparethetwomodels.
Thedifferencetestwasnonsignificant,w2(3,N5336)53.
68,p<.
30.
Thus,TVviolenceandsexdidnothaveadirecteffectonbuyingintentionsorcouponchoice.
TheeffectsofTVviolenceandsexoncouponchoicewereindirect,throughbrandmemoryandbuyingintentions.
Similaranalyseswereperformedtotesttheseparateeffectsofviolenceandsexonbrandmemory,buyingintentions,andcouponchoices.
TheseanalysesfoundsimilareffectsforTVviolenceandsex(seetheappendix).
DISCUSSIONItiscommonlyassumedthatviolenceandsexsellproducts,buttheymightactuallyhavetheoppositeeffect.
Thepresentstudyreplicatespreviousstudiesshowingthattelevisedviolenceandseximpairmemoryforadvertisedproducts.
Becausetelevisedviolenceandsexhadsimilareffectsonrecallandrecognition,thememoryfailureisprobablyduetoencodingratherthanre-trieval.
Televisedviolenceandsexalsodecreasedintentionstobuytheadvertisedbrandsandreducedthenumberofadver-tisedbrandswhosecouponsparticipantschose.
Memoryisanimportantoutcomemeasure,becauseitispositivelyrelatedtobuyingintentionsandcouponchoices.
Itistruethatviolentandsexyprogramsattractyoungviewers,andthisagegroupishighlydesirabletoadvertisers.
However,theeffectsinthepresentresearchheldformenandwomenofallages,regardlessofwhethertheylikedtowatchviolentorsexualprograms.
Theobtainedresultsareconsistentwiththetheoryofreasonedaction(e.
g.
,Ajzen&Fishbein,1977),whichpositsthatbe-havioralintentionsmediatetheeffectofapersuasivemessageonbehavior.
Inthepresentstudy,buyingintentionsmediatedtheeffectoftelevisedviolenceandsexoncouponchoices.
Thepresentstudyhasimportantpracticalimplicationsforsociety.
IfadvertisersrefusedtosponsorviolentandsexualTVprograms,suchprogramswouldbecomeextinct.
About60%ofTVprogramscontainviolenceorsex(Kunkeletal.
,1999;Na-tionalTelevisionViolenceStudy,1998).
TVviolenceandsexcanhaveanegativeimpactonsociety.
Researchfromhundredsofstudiesconductedoverseveraldecadeshasshownthattele-visedviolenceincreasessocietalviolence(e.
g.
,Anderson&Bushman,2002).
Sexuallyexplicitmediapromotesexualcal-lousness,cynicalattitudesaboutloveandmarriage,andper-ceptionsthatpromiscuityisthenorm(e.
g.
,Allen,Emmers,&Gebhardt,1995;Zillmann,2000).
Moreover,mediainwhichsexiscombinedwithviolencemayhaveparticularlyperniciouseffects(e.
g.
,Allen,D'Alessio,&Brezgel,1995;Linz,Donner-TABLE2Variance-CovarianceMatrixfortheStructuralEquationAnalysesMeasure123451.
TVviolenceorsex0.
18(0.
750).
18n.
29n.
23n.
14n2.
Brandrecall0.
0851.
174(0.
839).
58n.
33n.
18n3.
Brandrecognition0.
3491.
7377.
618(4.
313).
51n.
21n4.
Buyingintentions0.
2410.
8563.
3885.
833(4.
515).
37n5.
Consumerbehavior0.
2070.
6551.
9882.
95911.
229(4.
676)Note.
N5336.
Variancesandmeans(inparentheses)areonthediagonal;covariancesarebelowthediagonal.
Fordescriptivepurposes,correlationsareabovethediagonal.
np<.
05.
Fig.
1.
Effectoftelevisionviolenceandsexonviewers'consumerbehavior(i.
e.
,choosingcouponsforadvertisedbrands).
Theanalysisdepictedheretestedbrandmemoryandbuyingintentionsasmediatorsoftheeffect.
Asterisksindicatesignificantpathcoefcients,np<.
05.
Standardizedestimatesareshown.
706Volume16—Number9ViolenceandSexDon'tSellstein,&Penrod,1988).
Thus,sponsoringviolentandsexualprogramsmightbebadforsocietyandbadforbusiness.
Acknowledgments—ThisresearchwasfundedinpartbyaresearchgrantfromIowaStateUniversity.
IwouldliketothankRebeccaMarzenandLauraValenzianofortheirhelprecordingtheprograms,NicoleOswaltforherhelptestingparticipants,andMikeGillespieforhishelpwiththestructuralequationanalyses.
IalsowouldliketothankRowellHuesmannforhishelpfulfeedbackonadraftofthismanuscript.
REFERENCESAilawadi,K.
L.
,Lehmann,D.
R.
,&Neslin,S.
A.
(2001).
Marketresponsetoamajorpolicychangeinthemarketingmix:LearningfromProcter&Gamble'svaluepricingstrategy.
JournalofMarketing,65,44–61.
Ajzen,I.
,&Fishbein,M.
(1977).
Attitude-behaviorrelations:Atheo-reticalanalysisandreviewofempiricalresearch.
PsychologicalBulletin,84,888–914.
Allen,M.
,D'Alessio,D.
,&Brezgel,K.
A.
(1995).
Ameta-analysissummarizingtheeffectsofpornography:II.
Aggressionafterex-posure.
HumanCommunicationResearch,22,258–283.
Allen,M.
,Emmers,T.
,&Gebhardt,L.
(1995).
Exposuretopornographyandacceptanceoftherapemyths.
JournalofCommunication,45,5–26.
AmericanPsychologicalAssociation.
(1993).
Violenceandyouth:Psychology'sresponse.
Washington,DC:Author.
Anderson,C.
A.
,&Bushman,B.
J.
(2002).
Mediaviolenceandsocietalviolence.
Science,295,2377–2378.
Arbuckle,J.
L.
(1999).
AMOS(Version4.
01)[Computersoftware].
Chicago:SmallWatersCorp.
Bagozzi,R.
P.
,Baumgartner,H.
,&Yi,Y.
(1991).
Couponusageandthetheoryofreasonedaction.
AdvancesinConsumerResearch,8,24–27.
Bagozzi,R.
P.
,Baumgartner,H.
,&Yi,Y.
(1992).
Stateversusactionorientationandthetheoryofreasonedaction:Anapplicationtocouponusage.
JournalofConsumerResearch,18,505–518.
Bawa,K.
,&Shoemaker,R.
(1989).
Analyzingincrementalsalesfromadirectmailcouponpromotion.
JournalofMarketing,53,66–78.
Bushman,B.
J.
(2003,August).
Ifthetelevisionprogrambleeds,memoryfortheadvertisementrecedes.
PaperpresentedattheannualmeetingoftheAmericanPsychologicalAssociation,Toronto,Ontario,Canada.
Bushman,B.
J.
,&Bonacci,A.
M.
(2002).
Violenceandseximpairmemoryfortelevisionads.
JournalofAppliedPsychology,87,557–564.
Bushman,B.
J.
,&Phillips,C.
M.
(2001).
Ifthetelevisionprogrambleeds,memoryfortheadvertisementrecedes.
CurrentDirectionsinPsychologicalScience,10,44–47.
Chakraborty,G.
,&Cole,C.
(1991).
Theeffectsofcouponcharacter-isticsonbrandchoice:Alaboratorystudy.
PsychologyandMar-keting,8,145–159.
Chaudhuri,A.
(2001).
Thechainofeffectsfrombrandtrustandbrandaffecttobrandperformance:Theroleofbrandloyalty.
JournalofMarketing,65,81–93.
Cheong,K.
J.
(2003).
Observations:Arecents-offcouponseffectiveJournalofAdvertisingResearch,33,73–78.
Furnham,A.
,&Gunter,B.
(1987).
Effectsoftimeofdayandmediumofpresentationonimmediaterecallofviolentandnonviolentnews.
AppliedCognitivePsychology,1,255–262.
Geer,J.
H.
,Judice,S.
,&Jackson,S.
(1994).
Readingtimesforeroticmaterial:Thepausetoreect.
TheJournalofGeneralPsychology,121,345–352.
Geer,J.
H.
,&McGlone,M.
S.
(1990).
Sexdifferencesinmemoryforerotica.
CognitionandEmotion,4,71–78.
Hamilton,J.
T.
(1998).
Channelingviolence:Theeconomicmarketforviolenttelevisionprogramming.
Princeton,NJ:PrincetonUniver-sityPress.
Kim,S.
J.
(1994).
''Viewerdiscretionisadvised'':Astructuralapproachtotheissueoftelevisionviolence.
UniversityofPennsylvaniaLawReview,142,1383–1441.
Kunkel,D.
,Cope,K.
M.
,MaynardFarinola,W.
J.
,Biely,E.
,Rollin,E.
,&Donnerstein,E.
(1999).
SexonTV.
MenloPark,CA:KaiserFamilyFoundation.
Lam,S.
Y.
,Vandenbosch,M.
,Hulland,J.
,&Pearce,M.
(2001).
Eval-uatingpromotionsinshoppingenvironments:Decomposingsalesresponseintoattraction,conversion,andspendingeffects.
Mar-ketingScience,20,194–215.
Lang,A.
,Newhagen,J.
,&Reeves,B.
(1996).
Negativevideoasstructure:Emotion,attention,capacity,andmemory.
JournalofBroadcastingandElectronicMedia,40,460–477.
Linz,D.
G.
,Donnerstein,E.
,&Penrod,S.
(1988).
Effectsoflong-termexposuretoviolentandsexuallydegradingdepictionsofwomen.
JournalofPersonalityandSocialPsychology,55,758–768.
NationalTelevisionViolenceStudy.
(1998).
NationalTelevisionVio-lenceStudy(Vol.
3).
SantaBarbara:UniversityofCalifornia,SantaBarbara,CenterforCommunicationandSocialPolicy.
NielsenMediaResearch.
(2000).
Nielsentelevisionindex,nationalaudiencedemographics(Vol.
1).
NewYork:Author.
ParentsTelevisionCouncil.
(2003).
Sexlosesitsappeal:AstateoftheindustryreportonsexonTV.
RetrievedDecember15,2003,fromhttp://www.
parentstv.
org/ptc/publications/reports/stateindustrysex/main.
aspPromotionMarketingAssociation.
(2000).
Couponusagestatistics.
RetrievedJuly21,2004,fromhttp://couponing.
about.
com/cs/aboutcouponing/a/couponusage2000.
htmSchindler,R.
M.
(1992).
Acouponismorethanalowprice:Evidencefromashopping-simulationstudy.
PsychologyandMarketing,9,431–451.
Shimp,T.
A.
,&Kavas,A.
(1984).
Thetheoryofreasonedactionappliedtocouponusage.
JournalofConsumerResearch,11,795–809.
Sutel,S.
(2004).
NewcropofoddSuperBowladstodebut.
RetrievedJune2005fromAPOnline:http://newslink.
nandomedia.
com/SportServer/football/n/playoffs/v-fresno/story/1127283p-7844151c.
htmlTaylor,G.
A.
(2001).
Couponresponseinservices.
JournalofRetailing,77,139–151.
TelevisionBureauofAdvertising.
(2003).
MediacomparisonstudybytheTelevisionBureauofAdvertising.
(AvailablefromtheTelevisionBureauofAdvertising,3East54thStreet,NewYork,NY10022-3108)Wells,G.
L.
,&Windschitl,P.
D.
(1999).
Stimulussamplingandsocialpsychologicalexperimentation.
PersonalityandSocialPsycholo-gyBulletin,25,1115–1125.
Williamson,S.
S.
,Kosmitzki,C.
,&Kibler,J.
L.
(1995).
Theeffectsofviewingviolenceonattentioninwomen.
TheJournalofPsychol-ogy,129,717–721.
Volume16—Number9707BradJ.
BushmanZillmann,D.
(2000).
Inuenceofunrestrainedaccesstoeroticaonadolescents'andyoungadults'dispositionstowardsexuality.
JournalofAdolescentHealth,27S,41–44.
(RECEIVED2/2/04;REVISIONACCEPTED3/29/05;FINALMATERIALSRECEIVED4/6/05)APPENDIX:SEPARATEEFFECTSOFTVVIOLENCEANDSEXIntheanalysisoftheseparateeffectsofviolentprogramming,adummyvariablewasusedtorepresentTVviolence(violentprogram51,neutralprogram50).
Thehypothesizedmedia-tionmodeltthedataextremelywell,w2(5,N5168)55.
27,p<.
49,GFI5.
988,CFI5.
998,RMSEA5.
018.
Thedirect-effectsmodeldidnotimprovethet,w2(2,N5168)50.
76,p<.
69.
Similarly,intheanalysisoftheseparateeffectsofsexualprogramming,adummyvariablewasusedtorepresentTVsex(sexualprogram51,neutralprogram50).
Thehypothesizedmediationmodeltthedataextremelywell,w2(5,N5168)56.
67,p<.
25,GFI5.
985,CFI5.
991,RMSEA5.
045.
Thedirect-effectsmodeldidnotimprovethet,w2(2,N5168)51.
00,p<.
61.
Similarresultswereobtainedfortheeffectsofprogrammingthatwasbothviolentandsexual.
AdummyvariablewasusedtorepresentTVviolenceandsex(violentandsexualprogram51,neutralprogram50).
Thehypothesizedmediationmodeltthedataextremelywell,w2(5,N5168)53.
61,p<.
61,GFI5.
992,CFI51,RMSEA50.
Thedirect-effectsmodeldidnotimprovethet,w2(2,N5168)50.
76,p<.
68.
Twoadditionalmediationmodelscompareddifferencesbetweenviolentandsexualprogramming.
Forthesemodels,scoresforbrandrecallandbrandrecognitionwerestandardizedandsummedtocreateacompositevariableforbrandmemory.
Thedirect-effectsmodelshad0degreesoffreedom(i.
e.
,satu-ratedmodels)andarenotpresentedherebecausetheyshouldnotbeinterpreted.
Onemodelcomparedprogramswithviolenceandsextoprogramswithonlyviolence.
AdummyvariablewasusedtorepresentTVsex(violentandsexualprogram51,violentprogram50).
Thehypothesizedmediationmodeltthedatawell,w2(3,N5168)53.
39,p<.
34,GFI5.
99,CFI5.
99,RMSEA5.
028.
However,thestandardizedre-gressioncoefcientforthepathbetweentheindicatorvariableandthememoryvariablewasnonsignificant(coefcient5.
07,p5.
853).
Thus,sexdidnotaddanythingabovetheeffectofviolence.
Theothermodelcomparedprogramswithviolenceandsextoprogramswithonlysex.
AdummyvariablewasusedtorepresentTVviolence(violentandsexualprogram51,sexualprogram50).
Thehypothesizedmediationmodeltthedatawell,w2(3,N5168)51.
20,p<.
75,GFI5.
996,CFI51,RMSEA50.
However,thestandardizedregressioncoefcientforthepathbetweentheindicatorvariableandthememoryvariablewasnonsignificant(coefcient5.
17,p5.
814).
Thus,violencedidnotaddanythingabovetheeffectofsex.
708Volume16—Number9ViolenceandSexDon'tSell