廣告調(diào)查（第二版）英文 指導手冊 Davis-Instructors-Manual-ch08 Observation Research Human and Automated

AdvertisingResearch:Instructor’sManual

Copyright?2012PearsonEducation,Inc.publishingasPrenticeHall

PAGE

PAGE

AdvertisingResearch:Instructor’sManual

Copyright?2012PearsonEducation,Inc.publishingasPrenticeHall

8. ObservationResearch:HumanandAutomated

ChapterGoals

Thechapterwillhelpstudentsincreasetheirunderstandingof:

? howhumanandautomatedobservationisused.

? thestrengths,weaknesses,andlimitationsofeachformofobservation.

? therangeofinsightsprovidedbyeachformofobservationandhowtheseinsightscontributetoadvertisingdecision-making.

NotestotheInstructor

TheChapterLectureprovidesaguidetokeytopicsandcontent.ThePowerPointslidesarenamed:davis_adresearch_ch8.ppt.

Twoonlinesupplementalreadingsillustratetheuseandpowerofobservationresearch.

The“VideoConsumerMappingStudy”isthelargestobservationalstudyofmediauseandconsumptioneverconducted.Theresearch,conductedonbehalfoftheNielsen-fundedCouncilforResearchExcellencebyBallStateUniversity’sCenterforMediaDesignandSequentPartners,observedandrecordedthemediausageof476adultsinsixgeographicallydispersedDesignatedMarketAreas(DMAs).Mediausagewasrecordedbyextensivelytrainedobserversevery10secondsforanindividual’sfullwakingdayperiod.Observationratherthansurveyswereusedtocollectthedatabecausepriorresearchhaddemonstratedthatmediausageiscomplex,multifacetedandnotaccuratelyreportedinsurveysorotherself-reportingapproaches.

Thesecondreportisnotaresearchstudyperse,butitdoesuseobservationtoevaluatethesuccessofasocialmediacampaign.Thecampaign,designedbyAmnestyUK,wasdesignedtoraiseawarenessofissuesrelatedtoviolenceagainstwomen.Thesuccessofthecampaignwasexclusivelymeasuredthroughthecollectionofobservationaldatathatreportedthenumberofpeoplewho:

? sentsupportingemails,

? changedtheirFacebookorTwitteravatars,

? Tweetedthetargetmessageand/orbecameaTwitterfollower,

? tookapledge,

? visitedatargetwebsite.

Theextenttowhichthecampaignaccomplisheditsgoals,asreflectedinthepriorobservationaldata,isprovidedinthereport.Asyoudiscussthecampaignanditsresults,besuretonoticehowtheobservationalmeasuresusedtoevaluatethecampaignaredirectlyrelatedtothecampaign’sapproachandgoals.

ChapterLecture

Lecturediscussestwotypesofobservationresearchcommonlyusedtoinformadvertisingdecision-making:

? Humanobservation,whichusesaresearchertoobserveotherpeople’sbehaviors.

? Automatedobservation,whichusescomputersormechanicaltrackingdevicestoobservebehaviors,forexample,websurfingcookies.

Slide8-2

Slide8-3

I.HumanObservation

Humanobservationresearchismostappropriateinfourtypesofsituations:

? Whereobservationsofbehavioraremoreinsightfulthandescriptionsofbehavior

? Whererespondentsmaybeunabletoverbalizetheirattitudes

? Whensurveymeasuresofattitudesmaynotaccuratelypredictactualbehaviors

? Whenbehaviorsthemselvesarethebestsourceofinsight

A.QualitativeorQuantitative?

Observationresearchcanbeeitherquantitativeorqualitativedependinguponsamplesize,howsamplewasselectedandthetypeofdatacollected.

Considerasituationwhereyouwanttoknowifshowingyoungchildrenadvertisingencouragingphysicalactivityactuallymotivatesthemtobemoreactive.Youselectagroupofchildrenandthenrandomlyassignthemtooneoftwoconditions:

Slide8-4

Slide8-5

Slide8-6

Slide8-7

Slide8-8

Slide8-9

ChildreninthetreatmentgroupwatchanepisodeofTheSimpsonsrecordedfromatelevisionbroadcastembeddedwiththree:30advertisementspromotingphysicalactivity.Childreninthecontrolgroupwatchthesameepisodeembeddedwithchild-appropriateadvertisements,butwithoutphysicalactivityadvertisements.

Afterviewing,childrenaregiventimetoplayoutdoors.Observerswatchthechildren.Datacouldeitherbequalitativeorquantitative.

? Qualitativewouldconsistofresearchertakingnotesaboutchildren’splayaswellasrecordingthoughtsaboutwhetherornotanyspecificinstanceofplaycouldbeconsidered“physicalactivity.”

? Quantitativewouldconsistofresearcherpreparing(inadvance)alistofactivitiesconsidered“physicalactivity”and,whenalistedactivityoccurs,recordtheamountoftimespentonthatactivity.

B.AspectsofHumanObservationResearch

Humanobservationresearchischaracterizedintermsoffourdimensions:(1)typeofsituationinwhichobservationtakesplace(naturalorartificial),(2)observerpresence(openversusdisguised),(3)levelofobserverparticipation(activeorpassive)and(4)formofdatarecording(unstructuredversusstructured).

1.Situation:NaturalVersusArtificial

Naturalobservationofindividuals,situations,objects,oreventstakesplacenaturallyasbehaviorsunfoldattheirownpaceintheirownenvironment.Examplesinclude:

? countingthenumberandgenderofindividualswhovisitafastfoodrestaurant

? observingsalesclerksastheyservetheircustomers

? observingthelabel-readingbehaviors

? recordingthetimeshoppersspendreadingvariouspoint-of-purchasedisplays

Naturalobservationisappropriatewhenthetargetbehaviorsarerepetitive,

frequent,and/oroccurwithinareasonablyshorttimeframe.

Slide8-

10

Slide8-

11

Slide8-

12

Slide8-

13

Slide8-

14

Slide8-

Artificialobservationrecordstargetbehaviorsoreventsincontextofafabricatedsituation.Theresearchertakesanactiveroleinpromptingthetargetbehavior,forexample:

? Mysteryshopperresearch

? Commercialtestfacilities

Advantages:(1)permitsresearchertospeedupthedatagatheringprocessbyinitiatingdesiredsituationratherthanwaitingforittooccurnaturallyand(2)permitsresearchertocontrolextraneousvariablesthatmightimpactwhatisbeingobserved.

2.ObserverObtrusiveness:OpenVersusDisguised

Thisreferstotheextenttowhichthepresenceofanobserverisknownbyindividualsunderobservation.

Openobservationoccurswhenthepresenceoftheobserverisexplicitlyknown.Disguisedobservationhidespresenceoftheobserver.

Researchhasdemonstratedthattheknownpresenceofanobserverhasgreatpotentialforalteringthebehaviorsofthepersonbeingobserved.Asaresult,disguisedobservationtypicallyprovidesbetter,morerealisticdata.

3.ObserverParticipation:ActiveVersusPassive

Activeobservertypicallytakespartintheactivitiesbeingobserved.A“mysteryshopper”thatinitiatesconversationswithsalespersonnelisanactive

participantintheresearchprocess.

Passiveobservertypicallywatcheswithoutinterferingorinteractingwiththepeopleorobjectsbeingobserved.A“mysteryshopper”whowatcheshowsalespersonnelinteractwithrealshoppersisapassiveobserver.

Approacheshavecomplimentarystrengthsandweaknesses.

? Anactiveobserverisclosertothesourceofthedataandhasopportunitytodirectlyinteractwiththosebeingobserved.Activeobserversalsoallowtheobservertodirectactivitiestofocusonkeyissues.Mustbecarefulnottointroducebias.

? Passiveobservationrequirestheresearchertotakeeventsastheycome.

15

Slide8-

16

Slide8-

17

Slide8-

18

Slide8-

19

Itcanbetime-consuming,butprovidesthecleanestdatabecauseeventsandbehaviorsunfoldwithoutanyartificialinfluence.

4.DataRecording:StructuredVersusUnstructured

Structureddatarecordingknowsdesiredtypesofinformationandbehaviorsinadvance.Desireddataisrecordedonachecklistorobservationformwhileallotherinformationandbehaviorsareignored.

Unstructureddatarecordingrecordsobservedbehaviorsinverbalform,typicallyasanarrativeorfieldnotes.Therearenorestrictionsonthetypesofbehaviorsobserved,althoughitisexpectedthatpre-identifiedtargetbehaviorswillbespecificallynoted.

Approacheshavecomplementarystrengthsandweaknesses.

? Structuredobservationsrequiremoretimeinvestmentbefore

observation(inordertocreatetheobservationform),butdatacollection

andanalysisaremuchmoreefficient.

? Unstructuredobservationsprovidegreateropportunityfordiscoveriesinthefieldbutcanbetimeconsumingasdatamustbecodedpriortoanalysis.

C.ExamplesofObservationResearch

1.DisconnectBetweenAttitudesandBehaviors

Thisobservationisanaturalsituationincludingdisguisedobserver,passiveobserver,andstructureddatacollection.

Background:GreenGlobe21isaninternationalcertificationprogramdesignedtopromotesustainabletravelandtourism.TheGreenGlobe21(GG21)ecolabelcanbeplacedonproducts,places,ormaterials.Surveyresearchfoundtourists’attitudestowardstheGG21ecolabelwereverypositiveandthattheyappearedtohaveahighawarenessofsustainabilityissues.

ResearchQuestion:Dothesesupportiveattitudestranslateintoactualbehaviors?

Methodology:Atouristcenterwasselectedastheplaceofobservation.Theobservationportionofthestudywascarriedoutintwophases.(1)Baselinedatawascollected.Here,visitorwalkingpatternsaroundthecenterwereobservedandcoded.Thenumberandtimespentlookingatmaterials(bothGG21labeledandnonlabeled)wasobservedandrecorded.(2)Alarge,

attention-gettingdisplayandslideshowexplainingGreenGlobe21wasplacedinthefrontofthecenter.ThegoalofthedisplaywastopassivelypromptvisitorstorecalltheirexpressedpositiveattitudestowardecolabelingandpayattentiontoGG21labeledmaterials.Identicaltothebaselinephase,walkingpatternsandmaterialinteractionswereobservedandrecorded.

Results:Therewasasignificantdisconnectbetweenattitudesandbehavior.Eventhoughvisitors’walkingpatternsshowedthatGG21labeledmaterialshadtheopportunitytobeviewed,barelyanyvisitorstookthetimetostopandexaminethesematerials.Thiswastrueofallvisitors,butespeciallythosewhoshowedhighlevelsofenvironmentalconcernandawarenessofGG21inpost-visitinterviews.

Slide8-

20

Case2:Self-ReportedDataRegardingAttitudesandBehaviorsMayNotbe

Accurate

Thisexampleusesobservationthatisnatural,disguised,passive,andstructured.

ResearchQuestion:Doindividuals’self-reportedbehaviorsaccuratelyrepresenttheiractualbehaviorswithregardtoalcoholpurchase?

Methodology:Conductfocusgroupinterviewstouncoverattitudesand

reportedbehaviorsregardingalcoholpurchase.Conductobservationsofalcohol

purchasebehaviors.

Results:Verbalreportssignificantlyoverstatedbrandimportanceandinvolvementinpurchasedecision,specifically:

? Consumersclaimedtheyregularlyreadproductlabelingandtookovertenminutesin-storeforaroutineshop…this

contrastedwiththeobservationalfindingsthatshowedthatfewer

than10%ofconsumersreadlabelsand,onaverage,shopperstooklessthanthreeminutestoenter,makeadecision,payandleave.

? Ininterviewsandfocusgroups,consumersconsistentlymisrepresentedbrandloyalty.Recalledbehaviorgaveanaverageusageoffourbrandswhileobservationalresearchfindingsshowedthenumbertobesomewhereclosertosixbrands.

? Fewrespondentswereobservedtonoticeposters,giftswithpurchaseorbonusstockoffers.Intheinterviewsrespondentsclaimedtheseofferswerenoticedandimpactedonmanypurchasedecisions.

? Abouthalftherespondentsthoughttheyhadboughtaproductonspecial.Thefigureobtainedthroughobservation,wasclosertotenpercent.

Slide8-

21

Slide8-

22

Slide8-

23

Case3:BehavioralRatherthanAttitudinalChangeistheFocus

Thisexampleusesobservationthatisartificial,open,passive,andstructured.

ResearchQuestion:Haveactualbehaviors,ratherthanjustattitudes,changedinresponsetomessageexposure?Specifically,researcherswantedtoknowiftheycouldpromotehealthychickenhandlingbehaviorsthroughacommunicationprogram.

Methodology:Pre-postexperimentaldesignwherethebehaviorsofthoseexposedtothecommunicationprogramwerecomparedtoindividualsinacontrol(unexposed)group.Targetedbehaviorsweretheuseofseparateoradequatelywashedanddriedchoppingboardsandknivesbetweenthepreparationofrawchickenandotherfoods,orthepreparationofallotherfoodspriortopreparingchicken.

Results:Short-termfoodhandlingbehaviorsimprovedintheexposedversusnon-exposedgroup.

II.AutomatedObservation

Humanbehaviorsareobserved,butobservationaldataiscollectedbymachineratherthanpeople.Machinesmonitorandtrackconsumers’behaviors,typicallycollectingquantitativedatainastructured,disguisedmanner.

Therearetwomainformsofautomatedobservation(1)directmonitoringofconsumerbehaviors.Occurswhencookiesareusedtotrackanindividual’sonlinebehaviorsorwhenloyaltycardsareusedatthesupermarket(2)monitoringtheproductsofconsumerbehaviors.Occurswhencomputersareusedtomeasureonline“buzz”ortomonitorchangesinbrandperceptionsinblogsorotherformsofconsumergeneratedmedia.

A.ObservingOnlineBehaviors

Websitesdonottypicallydeliveradvertisingfromtheirownserversbutareinsteadfedadvertisementsfromanadvertisingnetwork’sserver.Itisthenetwork’sserverthatdecideswhichadwillbeseenineachindividualviewingsituation.Cookieshelptheadvertisingnetworkdeterminewhichadsshouldbeservedtoeachindividual.

DoubleClick’sDARTisoneofthelargestadservingnetworks.DART’sgoalistouseobservationorpastbehaviorstoincreaseadvertisingrelevanceand,byimplication,increaseadvertisingresponse.

“OurclientsstoretheiradsonDoubleClick’sadservers.WhenyouvisitaWebpageonwhichaclientisusingDoubleClicktechnologytodeliverads,codingthatthewebsitepublisherplacedintheWebpagetellsyourcomputer’sbrowsertosendarequestforanadtotheDoubleClickadserver.WhentheDoubleClickadserverreceivesarequest,itwillselectanadbasedonthecriteriathattheclienthaschosentogetherwithanyinformationloggedagainsttheuniquecookieid.”

FetchBackusesadifferentapproach.Itusesobservationtomatchadvertisingexposuretopriorwebsitevisits,displaysadvertisingforawebsitetopastvisitorstothatsite.

CoremetricsLIVEProfiletrackscustomersandprospectsastheyinteractwithbusinessesonline,acrossmultipleadnetworks,orviaemail,video,affiliatesitesandsocialmedia.Itthenintegratesthisdataandofflineinformation,

providingasinglecomprehensiveviewofeachvisitor’sbehaviorovertimeand

acrosschannels.

Slide8-

24

Slide8-

25

GhosteryisaFirefoxadd-onthatmakeswebobservationexplicitandtransparent.

B.ObservingOfflineBehaviors

Storeloyaltycardsareonewaythatobservationtakesplaceoffline.Loyaltycardsarethecreditcardorkeychain-sizedcardswithabarcodeormagneticstripeofferedbymostlargeretailchains,particularlysupermarkets,pharmacies,andclothingstores.

Grocerystoreloyaltycardsworkasfollows:

? Whenscannedatcashregister,thecardunlocksspecialdiscountsofferedto"loyal"members.

? Inreturnforsavings,cardholdersagreetoallowthegrocerystoretotracktheirpurchaseseachtimetheyshop.

? Grocerystoresusethisinformationtodecidewhichproductstocarry,whatpricestocharge,andinsomecases,totargetconsumerswithspecificcouponsandpromotionsonbehalfof

grocerymanufacturers.

Storeusesaggregatedatainternally(andsometimesexternally)aspartofitsmarketingresearch.Onanindividuallevel,theobservationaldataprovidedbythesecardsallowsmarketerstodeterminesuccessofpromotionaleffortsortocustomizeofferstospecifictypesofbehaviorsorpurchasepatterns.

Slide8-

26

Slide8-

27

C.ObservingConsumerGeneratedMedia

Consumergeneratedmedia(CGM)“describestheevolvingconsumer-createdspaceontheInternet.CGMreferstoabroadrangeofonlineword-of-mouthvehiclesincluding,butnotlimitedto,consumer-to-consumeremail,postingsonpublicInternetdiscussionboards/forums,Usenetgroupsandlistservs,

consumerratingsWebsitesorfourms,blogs,moblogs(siteswhereuserspost

digitalimages/photos/movies)vlogs(videoblogs),socialnetworkingwebsites,andindividualwebsites.”

Thevastnumberofsourcestobemonitored(nowestimatedtoexceed100million)necessitatesautomatedobservation.AllinvolvedcompaniesprovideacoresetofinformationbasedontheircontinuousobservationoftensorevenhundredsofmillionsofCGMsources.InformationtypicallyprovidedbyCGMobservationincludes:

? Coverage:Numberoftimesabrandorissueismentioned.

? Depth:Howdeeplydoesthepostingdiscussabrand?IsitjustapassingmentionordoestheCGMgointothesubjectin-depthwithnumerouscommentsandlinks?

? Content:Wasthepostingdesignedtosolveaproblem,comparedifferentbrands,passalonginformation,advocateforacause,criticizetheorganizationorsimplyallowtheauthortorant?

? Sentiment:WhatwasthetoneandaffectoftheCGM.Wasitgenerallypositive,negativeorneutral?(Slide8-27providesanexampleofsentimentmonitoring.)

Thepriorlistdescribestwotypesofobservations:purelydescriptive(content,depth)andanalytical(contentandsentiment).

AdvertisersuseobservationsofCGMtoinformplanninganddecision-makinginfourprimaryways.

1.CampaignMonitoring

AdvertiserscanrelatetrendsandshiftsinCGMtotheircommunicationprogram,identifyingaspectsofaprogramthatseemtobeworkingwellandthosethatareproblematicandneedadditionalattention.

Slide8-27showsthekeyeventsforthefirsttenmonthsofanewsoftdrink’scommunicationcampaignchartedagainsttheamountanddirectionofconsumerbuzz.Trendsindicatethefollowingwithregardtoimpactofthecommunicationprogram:

Slide8-

28

Slide8-

29

? SincethiswasanewsoftdrinktherewaslittleCGMpriortotheproductintroductionandthestartoftheadvertisingcampaign.AlloftheCGMmonitoredduringthisperiodwaspre-announcementspeculation.

? ThetelevisioncampaignlaunchedattheendofMarch.ThecampaignwasextremelywellreceivedandgeneratedincreasingamountsofCGM.AnexaminationofsentimentindicatedthattheadditionalCGMsentimentwasoverwhelminglypositive.Asthecampaigncontinuedtorun,somesignsof“wear-out”occurredastheamountofpositivesentiment(whilestillhigh)begantoslowlydecline.

? AviralvideocampaignwaslaunchedinJulytoannounceasummerpromotion.Itappearsthatthiscampaignwasadisaster.NegativeCGMmentionsbegintospikeveryshortlyafterthelaunchoftheviralcampaignwhilepositiveCGMdropsdramatically.

? Theinitialviralcampaignisreplacedbyasecondcampaignseveralweekslater.Thissecondviralcampaignworkswelltoreversethepriortrend,generatinghighlevelsofpositivesentimentwhilesignificantlyreducingtheamountofnegativesenti