Cole & Greer (2013) - Audience Response to Brand Journalism: The Effect of Frame, Source, and Involvement

http://jmq.sagepub.com/Communication QuarterlyJournalism & Masshttp://jmq.sagepub.com/content/90/4/673The online version of this article can be found at:DOI: 10.1177/107769901350316015 October 2013 2013 90: 673 originally published online Journalism & Mass Communication QuarterlyJames T. Cole II and Jennifer D. GreerInvolvementAudience Response to Brand Journalism: The Effect of Frame, Source, andPublished by:http://www.sagepublications.comOn behalf of:Association for Education in Journalism & Mass Communicationat: can be found Journalism & Mass Communication Quarterly Additional services and information for http://jmq.sagepub.com/cgi/alerts Email Alerts: http://jmq.sagepub.com/subscriptions Subscriptions: http://www.sagepub.com/journalsReprints.nav Reprints: http://www.sagepub.com/journalsPermissions.nav Permissions: by guest on November 20, 2014 jmq.sagepub.com Downloaded from by guest on November 20, 2014 jmq.sagepub.com Downloaded from What is This?- Oct 15, 2013 OnlineFirst Version of Record - Nov 18, 2013 Version of Record>> by guest on November 20, 2014 jmq.sagepub.com Downloaded from by guest on November 20, 2014 jmq.sagepub.com Downloaded from Journalism & Mass Communication Quarterly90(4) 673 690 2013 AEJMCReprints and permissions:sagepub.com/journalsPermissions.nav DOI: 10.1177/1077699013503160jmcq.sagepub.comPublics, Policies, and Social ControlAudience Response toBrand Journalism: TheEffect of Frame, Source,and InvolvementJames T. Cole II1 and Jennifer D. Greer2AbstractThisstudyexaminedreactionstobrandjournalisminlightofframe,source,and product involvement. Participants in an experimental study viewed a custom magazine with either a commercial (branded) or editorial (nonbranded) frame and read a story quoting either a peer or a corporate source. Readers rated the nonbranded magazine higherincredibility,butsourcecueshadnodirecteffectsoncredibilityratings. Sourcedidmatterwhencombinedwithconsumerproductinvolvement.Highly involved consumers had stronger brand attitudes and purchase intent after reading advice from a peer source; low-involved consumers responded more favorably to a corporate source.Keywordscommunication effects, magazines, framing, communication theory, advertisingBrand journalism, also called custom content or custom publishing, or named for themediumofdelivery(i.e.,custommagazines),allowsbusinessestotargetcus-tomers with useful, tailored editorial content while promoting their brand, values, and products. In 2011, U.S. companies poured 26% of their media budgets, a total of $40.2 billion,intocustomcontent.1Despitethedigitalpublishingrevolution,four-color magazine-styleprintpublicationsremainthepreferreddeliverymethod,accounting 1Red Barn Media Group/Advanced Content, Birmingham, AL, USA2University of Alabama, Tuscaloosa, USACorresponding Author:Jennifer D. Greer, Department of Journalism, University of Alabama, 490 Reese Phifer Hall, Box 870172, Tuscaloosa, AL, 35487-0172, USA. Email: [email protected]/1077699013503160Journalism & Mass Communication QuarterlyCole and Greerresearch-article2013 by guest on November 20, 2014 jmq.sagepub.com Downloaded from 674Journalism & Mass Communication Quarterly 90(4)for about $24 billion of those expenditures. Even with rapid growth in electronic deliv-ery, videos, and other new forms of reaching customers, 87% of companies still use print as a custom content platform. About 88,500 unique custom print magazines were produced in 2011.2 In a marketing environment, companies hope to build trust using the relative power and credibility of editorial content, often seen as more pure than advertising.3AsJohnCarrollofBostonUniversityputit,Editorialisthenew advertising.4Todate,littleacademicresearchexistsonhowaudiencesreacttocustommaga-zines. What works best in these publications when companies seek to promote credi-bility, positive attitudes toward the brand, and purchase intent (PI)? Van Reijmersdal, Neijens, and Smit,5 in one of the few published studies on effects of commerciality in custom publications, found that as audiences perceive a publication as more com-mercial, credibility ratings dropped.6Reactions to this mix of information and product promotion may indeed depend on whetherconsumersperceiveacustommagazinetobethecreationofasponsor(a commercialframe)orasthecreationofanindependentpublisher(aneditorial frame). The source of the information within the content might also affect audience reaction. In custom magazines, customers and company experts are commonly used as quotedsourcesinarticles.Finally,custommagazinesoftenaretargetedtoexisting customers, or those highly involved with a company or product category. Involvement could interact with frame and source cues in shaping audience attitudes.To investigate these concepts, this study uses an experimental design to examine the effect of frame and attributed story source in a custom magazine on audience rat-ings of content credibility, attitude toward the brand, and PI. Furthermore, this study examinedtheinteractionofthesevariableswithconsumersinvolvementwiththe product category. Understanding which type of sources and which type of publications work best for which type of consumers provides valuable insight into those working in brand journalism, and custom magazines in particular. The study also could provide valuable insight into all those working in persuasive communication, including adver-tising and public relations. The questions also are vital in digital publishing environ-mentsthatincreasinglyblendeditorialandpromotionalcontent.Howmuch commercialism is accepted by readers in these contexts?Literature ReviewThis study is guided by scholarly literature on framing, source credibility, and product involvement.Framing TheoryFraming, first formally named by Goffman7 in 1974, suggests that all individual per-ceptions occur within certain frames of reference, that is, accessible and preexisting schema. Information is processed differently, given the interpretive schema an indi-vidualapplies.8Thesamemessagecanbeframedindifferentways,invokingan individualtousedifferentschemastointerpretthemessage.9Framesarecriticalto by guest on November 20, 2014 jmq.sagepub.com Downloaded from Cole and Greer675communication research, as they are important tools for individuals processing infor-mation and for message creators when tailoring a message.10 Entman argues that fram-ingessentiallyinvolvesselectionandsaliencebymakingsomeinformationmore noticeableandsalient.11Framescanalsoobscurecertaininformation,reducingits salience.Marketers, noting the demonstrated power of framing, see editorial content as per-haps the most valuable form of media, because it is perceived as more believable and trustworthythanthehardsellofpureadvertising.12Outsideofcustomcontent research, past studies have suggested that presenting an article as commercial (i.e., sponsored) content versus editorial (i.e., a news story or service story) can influence how audiences interpret information. The placement and prominence of commercial cues may trigger an advertising schema, thereby creating skepticism regarding the content.13Evenadvertorials,inwhichinherentlycommercialmessagestakeontheappear-ance of editorial, have an advantage over pure advertising messages.14 Audiences find pure editorial content more credible than advertising.15 In several experimental stud-ies, researchers have found that audiences do not recognize advertiser-sponsored con-tent when it is packaged to look like a news story.16 Even the label advertorial does littletohelpreadersrecognizetheinformationasanadvertisement.Therefore, advertiser-sponsored content presented as news stories is seen as more credible than the same information presented as a traditional advertisement.17Commercial entities often look to custom magazines to use the authority of edito-rial to shore up brand values.18 The intent, then, is not only to sell the product, but to position the overall brand in the minds of the target audience.19 Attitude toward the brand (Ab), or the audiences internal, individual evaluation of a brand, also is linked to evaluations of promotional content. The audience sees the ad itself as a brand attri-bute. A likeable ad can contribute to the likeability of the brand.20Custommagazinesalsomaybeusedasawaytoimpactorevendrivesales leads.21 When product information is presented in an editorial format, or a news-like presentation,22 perceived risk is decreased, and consumers show a tendency to trust the information to draw a conclusion for their purchase intention.23Literature on framing and perceived relative purity of editorial content led to the creation of the following hypotheses:H1a:Participantswillrateinformationasmorecrediblewhenitcomesfroman editorial/non-branded frame, rather than a commercial/branded frame.H1b:Participantswillhaveamorepositiveattitudetowardthecompanywhen information comes from an editorial frame, rather than a commercial frame.H1c: Purchase intent toward the featured product will increase when information comes from an editorial frame, rather than a commercial frame.Source CredibilityMessages are filtered through the receivers perception of the message source.24 Early communication studies in source credibility set the construct forth as a component of by guest on November 20, 2014 jmq.sagepub.com Downloaded from 676Journalism & Mass Communication Quarterly 90(4)persuasion.25Thesourcesintention,expertise,andtrustworthinessaffecttheaudi-ences perception of the credibility of the message.Inpaststudies,messagesfromahigh-credibilitysourcehavebeenshownto increase the amount of opinion change, at least short-term.26 To see a source as credi-ble, the audience must believe the communicators opinions are unbiased and perceive the source as knowledgeable in the content area.27Companiesthatwanttospreadideasthroughcustommagazinesfacenumerous formatchoicesforpresentingideasandnumeroussourcechoicestorepresentthose ideas.28 Of particular interest to a discussion of brand journalism is the intention of the attributed source, referred to by McCroskey and Teven as the lost dimension of ethos/credibility.29 This has been confirmed in editorial copy, where research indicates that a competent (i.e., expert) neutral source is more credible than a competent source that hasavestedinterest.30Otherstudiessuggestthatinformationprovidedbypeersis perceived as credible as information from experts.31Theeffectofattributedsourceinproductinformationalsohasbeenexploredin relation to brand-building and PI. Although celebrity endorsers often are featured in advertising, expert and peer endorsements are more relevant in custom magazines, as theyaregenerallyperceivedasmorecredibleandmorerelatabletothetarget audience.32Framinghasbeenusedtoexaminetheeffectsofsourcecredibility.Druckman33 suggested that audiences rely on credible sources to sort through frames.34 Although Druckmans work was concerned with the moderating effects of source credibility on framing success, it can be helpful in the context of custom magazines, where the audi-encesdefensesmayalreadybeheightenedtothecommercialityofthecontent. Perceptionofthesourceascommercial(i.e.,anadvertiserservingitsownbenefit) leads audiences to see this as an attempt to persuade. They, therefore, find the content lesscredible.35VanReijmersdal,Neijens,andSmitclarifiedthatpersuasiveknowl-edge has a negative effect on message credibility.36The literature on the importance of source in audience perceptions led to the fol-lowing hypotheses:H2a: Participants will rate information as more credible when it is attributed to a peer customer, rather than a corporate source.H2b: Participants will have a more positive attitude toward the company when the information is attributed to a peer customer, rather than a corporate source.H2c: Purchase intent toward the product will increase when participants are pre-sentedwithinformationattributedtoapeercustomer,ratherthanacorporate source.InvolvementProduct involvement has been a key variable in numerous advertising studies, where ithasbeenusedasanexplanatoryconceptcontributingtoconsumerattitudesand behaviors. Research has shown that the level of product involvement an individual has with a product or product category can influence the complexity and extensiveness of by guest on November 20, 2014 jmq.sagepub.com Downloaded from Cole and Greer677thecognitiveandbehavioraldecision-makingprocesses.37Whenindividualshavea greaterperceptionofproductrelevance,theirinvolvementishigh,andtheytendto engage in more effortful information processing on product evaluation.38 Researchers have suggested that product involvement is a relatively stable and long-lasting attri-butecomparedwithothervariablesinfluencingtheconsumerdecision-making process.39Park and Lee found that involvement with a product directly influences the degree of purchase intention.40 In another study, researchers found that product involvement played a key moderating role in purchase decisions that audiences made after reading onlinereviews,whicharealsoaformofpersuasivecontent.Althoughqualityand number of online reviews both increased purchase intention, high-involved consumers attended more to review quality, while low-involved consumers were more attuned to the number of reviews.41 Other advertising studies have established links between the perceived credibility of commercial messages and the audiences knowledge or expe-rience with a product or brand.42Although product involvement has not been tested in a custom magazine context, similareffectsmayemerge.Forexample,Wang43foundthatwhileexpertandpeer endorsementshaveapositiveeffectonconsumerattitudes,consumerendorsements areperceivedasmorecredibleandhavemoreinfluenceonPI,especiallywhenan audience is interested in the endorsed product. To explore the effect of involvement with the product and its interaction with source and frame cues, the following hypoth-eses are posed:H3a:Greaterinvolvementwiththeproductwillleadtohigherratingsof credibility.H3b:Greaterinvolvementwiththeproductwillleadtomorepositiveattitudes toward the brand.H3c: Greater involvement with the product will lead to greater intentions to pur-chase the product.Although the literature was clear that involvement would be positively associated with the dependent variables, no past studies have examined effects of the interaction ofinvolvementwithlevelsofcommercialityinbrandjournalism.Therefore,the studys sole research question is posed:RQ1: Does involvement interact with commercial frame and attributed source on audiencesperceptionsofmessagecredibility,attitudetothebrand,andpur-chase intent?Interaction Between Frame and SourceThe limited research in this area related to custom magazines shows that commercial-ity has a negative effect on credibility. Van Reijmersdal and colleagues explored the format credibility44 of custom magazines, testing custom attitudes toward the maga-zineitself.Theresearchteammanipulatedtheamountofcommercialityby by guest on November 20, 2014 jmq.sagepub.com Downloaded from 678Journalism & Mass Communication Quarterly 90(4)manipulating levels of brand integration (0%, 50%, or 100%) present in a magazine. The researchers found that greater product integration had a negative effect on read-ers perceived credibility of the magazine. They suggested that future studies examine smaller differences in brand integration to find the tipping point where commercial-ity begins to affect credibility.45 To test interaction between frame and source, the final hypotheses are posed:H4a:Themorecommercialtheframeandsourcecombination,thelesscredible participants will rate the content.H4b: The more commercial the frame and source combination, the lower partici-pants will rate attitude toward the brand.H4c: The more commercial the frame and source combination, the lower partici-pants will rate their purchase intention.MethodAn experimental study was designed to examine the effects of three independent vari-ables(frame,attributedsource,andproductinvolvement)onthreedependentvari-ables (message credibility, PI, and Ab).ParticipantsParticipants in the study were 512 undergraduate students enrolled in an introductory grammar class at a large southern university. The mean age was 19.78 (mdn = 20, SD = 1.89, range = 18 to 40), and 97.7% were 22 or younger. Women were 78.7% (n = 403) of the participant pool, a percentage slightly higher than the 71.2% of women enrolled in the college offering the class. Most participants (n = 423, 82.6%), identi-fiedthemselvesasCaucasian,withsmallerpercentagesidentifyingasAfrican American (n = 51, 10.0%), Asian or Asian American (n = 27, 5.3%), Hispanic/Latino (n = 16, 3.1%), Native American (n = 5, 1.0%), or other (n = 5, 1.0%).46ProcedureAfterapprovalbytheuniversitysInstitutionalReviewBoard,participantswere recruited and promised extra credit. Four versions of a print magazine excerpt (a cover and a one-page) were created. These versions were randomly ordered and distributed in a large lecture hall during five data collection sessions, effectively randomly assign-ing participants to treatment groups. Participants were relatively equally assigned to group, with between 124 and 133 seeing each of the four conditions.Participants received a packet of stimulus material consisting of a cover letter, one of the four magazine excerpts, and a questionnaire. The first page of the questionnaire measuredinvolvementwiththeproductcategory(ahometheatersystem).After answeringthesequestions,participantswereinstructedtopickuptheirmagazine excerpt and carefully review the cover and the article. After five minutes, researchers by guest on November 20, 2014 jmq.sagepub.com Downloaded from Cole and Greer679allowed participants to complete the last section of the questionnaire, which included dependent variable measures, demographic questions, and manipulation checks.The magazine excerpt was printed in full color on slick, heavyweight paper to rep-licate a custom magazine. Because product advice is common in custom content,47 the article focused on providing tips to first-time buyers of home theater equipment. All versions of the one-page article featured a sidebar on a product, a relatively inexpen-sive($399)all-in-onehometheatersystemsoldbyafictionalcompany,NextAV. Audio-visual equipment was selected because it is often featured in custom magazines and because college students are a key market for this product category. Participants indicated high involvement with the product, with 43.3% indicating either they or their familiesownedahometheatersystem.Furthermore,participantsperceivedthem-selves as coming from affluent backgrounds (M = 3.62 on a scale estimating familys economic status between 1 = lower class and 5 = upper class). Before reading the materials,about80%ofparticipantsindicatedtheywouldconsiderbuyingahome theater system priced less than $400.Independent VariablesFrameThemagazineinthisstudywasframedaseithercommercial/brandedoreditorial/nonbranded. To make this frame apparent, the cover letter told participants they were abouttoreadanarticlefromamagazineproducedbyeitherNextAV,Inc.,ahome theater retailer, or Sight & Sound, an independent journalistic magazine. On the first pageofthequestionnaire,participantsreadthatthemagazinetheywereaboutto review was either designed to promote the companys products (commercial frame) or to help provide objective information so consumers could make wise choices in select-ing AV products (editorial frame). In addition, seven text-based and visual frame cues wereincludedinthemagazineexcerpt.Forexample,thetitleofthemagazinewas prominentlydisplayedaseitherNextAVInc.,Youraudioandvideoequipment superstoreorSight&Sound,Yoursourceforhomeentertainmentnews.Visual cuesincludedalargepictureofastoreonthecommercialcoverandapictureofa home theater on the editorial cover. These frame cues were carried over onto the arti-cle page, with visual and verbal cues reinforcing NextAV Inc. or Sight & Sound.Manipulationchecksshowedthat89.95%ofparticipantscorrectlynamedthepro-ducer of the magazine based on their treatment group. However, when asked what type of company produced the magazine (a manufacturer/superstore or an independent jour-nalistic outlet), participants were less attuned to the cues. Only 67.6% correctly identi-fied this item, with participants in the commercial frame conditions (71.3%) more likely to correctly answer this question than those in the editorial frame conditions (63.7%).Attributed SourceAll information in the article was attributed to a NextAV product engineer (corpo-ratesource)orarecentcollegegraduatewhorecentlypurchasedahometheater by guest on November 20, 2014 jmq.sagepub.com Downloaded from 680Journalism & Mass Communication Quarterly 90(4)(peer/consumer source). A gender-neutral name, Pat Cramer, was used, and no gen-deredpronounswereincluded.Thephotographs,designedtofurtherhighlightthe source differences, used for a man and a woman. In the corporate source version, the photodepictedayoungmanwearingaNextAVshirtandwomancarryingaclip-board.Inthepeercustomerversion,thephotodepictedayoungmanandwoman seated in front of a home entertainment center.Acrossconditions,79.0%oftheparticipantscorrectlyidentifiedthesourcein manipulation checks. Perhaps because of closer identification with the peer customer source, 87.7% in those conditions correctly identified the source, compared with only 70.3% who correctly identified the source in the corporate source groups.InvolvementParticipantsratedtheirinvolvementwithhometheatersystemsonaversionof Zaichkowskys personal involvement inventory.48 Attitudes toward home theater sys-temsweremeasuredona7-pointcontinuumwithninebi-polaradjectivepairsas anchors (e.g., boring/interesting and worthless/valuable). These items all worked as a reliable scale ( = 0.87) and, therefore, were averaged to create an involvement score with higher scores indicating greater involvement with the product category.Scores ranged from 1 to 7 (M = 5.00, SD = 0.96). The high mean was expected giventhattheproductwaschosenbecauseofitsinterestamongtheparticipant pool. Participants were split into three product involvement categories for analy-ses: low, including those with scores lower than 4.63 (n = 162, 31.9%); medium, with scores from 4.63 to 5.54 (n = 165, 32.5%); and high, with scores higher than 5.55 (n = 181, 35.6%).Dependent VariablesCredibilityMessagecredibilitywasmeasuredwithashortenedversionofascaledesignedby Flanagin and Metzger.49 Participants rated the information in the article they read on three bi-polar adjective pairs (unbelievable/believable, inaccurate/accurate, not trustworthy/trustworthy) along a 7-point continuum. The items created one reliable scale ( = 0.89) and were averaged to create an article credibility score where higher scores indicated stronger ratings of message credibility.Scores again ranged from 1 to 7 (M = 5.17, SD = 1.24). The high mean likely was because the article was written by a professional journalist to mirror high-quality pro-fessional media content. A histogram revealed that the scores followed a normal, bell-shaped curve skewed to the right, reflecting the high mean.Ab and PI. Ab and PI often are highly correlated.50 In some studies, they are mea-sured together as a single construct.51 However, because communicators use custom contentforbuildingrelationshipwiththebrandand,increasingly,forgenerationof sales leads,52 this study measures the two as separate constructs, using scales for each construct developed by Spears and Singh.53 by guest on November 20, 2014 jmq.sagepub.com Downloaded from Cole and Greer681The anchors for the Ab scale were unappealing/appealing, bad/good, unpleasant/pleasant, unfavorable/favorable, and unlikable/likable. All five items produced a reliable scale ( = 0.94) and were averaged to create an Ab score where higher values indicated more positive attitudes toward the brand. Scores ranged from 1 to 7 (M = 5.25, SD = 1.12).The anchors for the PI scale were never buy/definitely buy, very low purchase interest/very high purchase interest, and probably not buy it/probably buy it. The items produced a reliable scale ( = 0.91) and were averaged to create a PI score with higher scores indicating greater intention to purchase the product. Scores ranged from 1to7(M=4.1,SD=1.42).Ahistogramrevealedthatthescoresonbothofthese measures again followed a normal, bell-shaped curve, with a right skew.ResultsTo test most of the hypotheses, a multiple analysis of covariance (MANCOVA) was run. The model included three independent variables run on three dependent variables. Tocontrolforparticipantsabilitytopurchasetheproduct,aself-reportedscoreof their perceived affluence (ranging from 1 to 5) was used as a covariate. Gender, using the dummy variable 0/1, was also loaded as a covariate because of the large percentage of women in the participant pool. Significant models emerged for all three dependent variables (see Table 1), suggesting the model had at least moderate predictive values on scores for all of the measured attitudes.Table 1.MANCOVA Model Testing Main and Interactive Effects of Frame, Source, and Product Involvement on the Three Dependent Variables.Credibility AbPIF Part.2F Part.2F Part.2Corrected model 4.32*** .102 6.54*** .147 6.51*** .1471. CovariatesPerceived affluence 3.78 .008 6.61* .013 1.21 .002Gender 0.001 .000 3.81 .008 0.01 .0002. Main effectsCommercial frame 16.81*** .033 3.87* .008 2.46 .005Attributed source 0.142 .000 4.23* .009 0.01 .000Product involvement 14.73*** .056 29.71*** .108 38.48*** .1353. InteractionsFrame Source 0.04 .000 0.05 .000 0.01 .000Product involvement Frame 0.40 .002 0.82 .003 0.08 .000Product involvement Source 2.18 .009 7.06*** .028 1.50 .006Product involvement Frame Source2.05 .008 0.94 .004 0.24 .001Error, Total number 492, 506 492, 506 492, 506Note. Ab = Attitude toward Brand; PI = Purchase Intent.*p < .05. **p < .01. ***p < .001. by guest on November 20, 2014 jmq.sagepub.com Downloaded from 682Journalism & Mass Communication Quarterly 90(4)The first set of hypotheses examined the effect of frame. As Table 1 shows, a signifi-cant main effect for frame emerged for message credibility and Ab. Participants in the two editorial frame conditions rated message credibility on average at 5.36 (SD = 1.17), compared with 4.98 (SD = 1.29) for participants in the commercial frame conditions. For Ab, average scores were higher in the editorial conditions (M = 5.31, SD = 1.07) than in the commercial conditions (M = 5.20, SD = 1.18). In both cases, the more independent frame resulted in more positive attitudes, meaning H1a and H1b were supported. No effect was found for frame on purchase intention, meaning H1c was not supported.The second set of hypotheses focused on the effects of attributed source. As Table 1 shows, a statistically significant main effect on Ab emerged, based on source. Those in the two corporate source conditions rated the brand as more favorable, 5.31 (SD = 1.01), than those in the peer source conditions, 5.20 (SD = 1.24). This relationship was the opposite of what H2b predicted, so H2b is rejected. No other direct source effects werefound,meaningH2aandH2calsowererejected.Sourceonitsownwasnot related to variation in perceptions of credibility or purchase intention.The third set of hypotheses examined involvement. As Table 1 shows, involvement accounted for most of the predictive power of all the models, with effect sizes dwarf-ing those of the manipulated independent variables. For all three dependent variables, the most involved participants had the highest scores (see Table 2). These strong and statistically significant effects for involvement were consistent across all four condi-tions, indicating that consumers involved with the product category had more favor-ablefeelingsaboutthearticle,thebrand,andbuyingtheproduct,regardlessof experimental condition.The studys only research question investigated how product involvement, which wassuspectedtobeamajorpredictorforaudienceattitudes,interactedwiththe manipulated independent variables of frame and attributed source. As Table 1 shows, a significant interactive effect was found between product involvement and attributed source cues for Ab. No statistically significant interaction was found for message cred-ibility or purchase intention.As the means for credibility in Table 3 show, those highly involved with the prod-uct rated the information as highly credible. The significant effects appear to be driven bydifferencesinthewaymedium-andlow-involvementparticipantsresponded. Moderately involved customers rated message credibility lower when information was attributed to a corporate source rather than a peer customer source, regardless of frame. For those with low involvement with the product, the inverse was true.Table 2.Means for High, Medium, and Low Involvement Participants in All Conditions Combined by Dependent Variable.Low, N = 162 Medium, N = 165 High, N = 181Dependent variable M (SD) M (SD) M (SD)Message credibility 5.01 (1.26) 4.93 (1.27) 5.53 (1.13)Attitude toward brand 4.92 (1.08) 5.11 (1.09) 5.69 (1.12)Purchase intent 3.35 (1.44) 3.91 (1.22) 4.58 (1.33) by guest on November 20, 2014 jmq.sagepub.com Downloaded from Cole and Greer683Thefinalhypothesesinvestigatedtheoveralleffectofcommerciality(frameand source)onreaderattitudes.Althoughnoframeandsourceinteractiveeffectswere detectedinTable1,theMANCOVAmodelexaminedtheseasdistinctconstructs, rather than along a continuum of highest commerciality (corporate frame and source) tolowestcommerciality(editorialframeandpeersource).Therefore,aMANOVA wasrun,testingtheamountofcommercialityonthethreedependentvariablescol-lectively.Asignificantmodelemergedforarticlecredibilityonly,supportingH4a. Those in the commercial frame/corporate source condition rated credibility the lowest (M = 4.90, SD = 1.30), followed by those in the commercial frame/peer source (M = 5.05, SD = 1.28), and those with the editorial frame/corporate source (M = 5.29, SD = 1.13).Thoseintheconditionwiththelowestcommerciality,incontrast,ratedthe article as most credible (M = 5.41, SD = 1.13). No significant effects were found on Ab or PI; thus, H4b and H4c were not supported.DiscussionFraming a custom magazine as editorial can positively affect readers perception of mes-sage credibility and increase positive attitudes toward a brand. Attributing information to a corporate source also affects brand attitudes positively, a finding opposite to what was hypothesized. Furthermore, increased commerciality through combining frame and source can significantly depress ratings of message credibility. Finally, product involve-ment plays a significant role in explaining variation in all attitudes examined. Involvement alsoproducesaninterestinginteractiveeffectwhenexaminedinlightofsource.For low-involvedconsumers,corporatesourcesledtothehighestcredibilityratings;for medium involvement consumers, peer sources made content seem most credible. High involvement participants rated content from both sources as highly credible.Although two factors controlled by custom magazine producers (frame and source) were related to the participants perceptions of message credibility and Ab, only prod-uctinvolvementwaslinkedtoPI.Thisstudyindicatesthatproductinvolvement should be examined closely as custom magazine producers make decisions on target-ing their messages.Effect of FrameFraming theory suggests that presentation is critical to the audiences interpretation of a message. Marketers rely on the perception of editorial purity in brand journalism to Table 3.Mean Scores for the Message Credibility Based on Product Involvements Interaction with Source Cues.Involvement Corporate source Peer sourceLow 5.11 (N = 85) 4.90 (N = 77)Medium 4.79 (N = 98) 5.10 (N = 76)High 5.49 (N = 80) 5.56 (N = 101) by guest on November 20, 2014 jmq.sagepub.com Downloaded from 684Journalism & Mass Communication Quarterly 90(4)enhancetheappearanceofneutrality,eveniftheunderlyingintentisanythingbut neutral.54Itisnotsurprising,then,thatparticipantsinthisstudyfoundinformation more credible when it was presented in an editorial frame. Moreover, this finding sup-ports past studies that suggest commercial cues may create skepticism regarding con-tent.55 The findings provide support for the conclusions of Van Reijmersdal, Neijens, and Smit that increased commerciality in custom magazines is linked to lower credi-bility ratings.56Althoughconsumershavetroubleclearlyidentifyingadvertorialcontentinsome cases, when they do perceive a message as commercial, they respond negatively to it. Advertisersareincreasinglyembracingeditorial-stylestorytellingtosupplementor replace traditional advertising to combat credibility challenges.57 This study suggests that these efforts are in the right direction for marketers who aim to build credibility among current or potential customers.However, if the goal is to build Ab and PI for their products, this practice might not be proven. An editorial frame was linked to more positive attitudes toward the brand. But this effect, while statistically significant, accounted for only a small portion of the explanatory power in the model examining brand attitudes. Furthermore, no positive effectforeditorialframeemergedforPI.Perhapseditorialneutralityalonehasno effect on PI without specific calls to action, such as business reply prompts.Because sponsors of custom magazines are increasingly concerned with return on investment,investigatingtheinstancesinwhichhigherarticlecredibilityandbrand attitudes ratings are linked to greater PI would be valuable. It should also be noted that although those who invest in custom magazines may hope to drive PI, experts empha-size that journalism, not marketing, should be the end result of custom content, as the goalsofthetwodisciplinesaredifferent.Marketingseekstopushwaresatthe momentoftruth,whiletheeditorialapproachseekstopresentinsightsthatadd value to customers lives.58Effect of SourceStraughan,Bleske,andZhao59emphasizedtheimportanceofselectingeffective sources to represent a companys ideas. Of particular interest is the concept of persua-siveintent60anditspotentialnegativeeffectonattitudestowardthecontent.61 Persuasiveintentofcorporatesources,forinstance,hasbeenviewedinprevious research as causing a lost dimension of ethos/credibility.62 The hard persuasive sell coming from a source clearly vested in a product often is rejected.In this study, however, the attributed source exhibited an effect only on Ab. In con-trast to our hypothesis, participants had more positive attitudes toward the brand when reading information from a company source. This makes sense, as even in the editorial frame condition, participants reading the information from the corporate source (and seeing the man with the NextAV Inc., logo on his shirt) did receive more exposure to the brand. In this way, the company source may seem to be a good choice, at least for brand-building. However, analyses of interaction with involvement suggest that com-pany sources may be more effective when communicating to those with low product by guest on November 20, 2014 jmq.sagepub.com Downloaded from Cole and Greer685involvement. Using a clear corporate source might work best, therefore, in building new clientele rather than speaking to existing customers.No main effects for source were found on perceptions of article credibility and PI. Expert and peer endorsements are often used in custom content, but a testimonial from anotherbuyerisgenerallyperceivedasrelatingmoretothetargetaudience.63The failure to find effects on PI are not surprising, however, as the article did not include a specific call to action or a special offer that might clearly prompt PI.What is surprising is the lack of effect of source on perceived message credibility. Theinformationpresentedinthearticlecarriedanauthoritativetone,regardlessof source cited (Heres information to help you). Perhaps this tone, coupled with the professional presentation and writing style, led participants to perceive a high level of expertise with both source types. Appearance of expertise has been identified as one of the main tenets of credibility.64 The corporate source and the peer source offered onlygeneraladviceratherthanadirectendorsementofthebrandortheproduct. Therefore, bias might not have influenced how the participants evaluated credibility. Again, the findings for involvement must be considered, as corporate sources may be better used for communicating with lower-involved customers.Overall CommercialityLooking at combined effects of frame and source, this study found that increased com-mercialityincustommagazinesleadstolowerratingsofmessagecredibility.The article in the editorial frame with the peer source was rated as the most credible. This supports conclusions reached by Van Reijmersdal, Neijens, and Smit65 that credibility was rated higher when commerciality was lower. Because commerciality in the cur-rentstudywasoperationalizedasacontinuumbasedonthecombinationofacom-mercial frame and corporate source, the findings shed new light on the tipping point of what consumers will accept.Creators of custom magazines might want a more direct return on investment than message credibility, but credibility includes trustworthiness. Trust between customer and company is a key component of brand-building, a desirable outcome sought when investing in brand journalism. A lack of bias, or at least a perceived one, is what com-municatorsbelievemakescustommagazinesworkasamarketingtool.Takenwith previous studies that show negative effects of commerciality and product placement,66 this study suggests that content creators might consider the editorial approach as an important component of building trust with customers.Effect of Involvement with the ProductOne of the key findings is that product involvement produces the same effects in cus-tommagazinesasitdoesinmoreovertpersuasivecontexts,namelyadvertising. Product involvement emerged as a significant predictor for all three attitudes. Those withthehighestinvolvementwiththeproductcategoryratedcredibility,thebrand, and PI significantly higher than those with moderate and low involvement, regardless of manipulations of frame and source. by guest on November 20, 2014 jmq.sagepub.com Downloaded from 686Journalism & Mass Communication Quarterly 90(4)This finding is consistent with past studies, particularly those examining persuasive content in general67 and advertising in particular. 68 This study supported the positive relationship,documentedinpaststudies,betweeninvolvementandPIaswellas between involvement and brand attitudes. Furthermore, it established a link between product involvement and perceived credibility of information about that product.Consequently, product involvement cannot be overlooked as a variable in studies regardingcustommagazines,oranyformofbrandjournalism.Brandjournalismis typically targeted at either prospects or existing customers.69 The results of this study suggest that current customers, those most highly involved with the product, are the most receptive audience for custom magazines.Targeting brand journalism at existing customers likely is more effective than tar-geting prospects through custom content. In fact, given the return on investment pres-sures, marketers should examine spending on traditional magazine advertising versus custom magazines, especially when targeting highly involved audiences. The effect of commercial cues on credibility, combined with the involvement findings here, could indicatetosomethateditorialcontentmaybeawiserinvestmentthantraditional advertisingbuys.Thatconclusionshouldnotbebasedsolelyonthesedata;future researchshouldcomparetraditionaladvertisingwithcustommagazinesandother forms of brand journalism.An interesting interaction occurred for source and involvement. Participants with thelowestproductinvolvementratedAbashigherandthearticleasmorecredible when a corporate source was quoted. Moderately interested readers responded better to the peer source. In light of these results and those from Poorisat and colleagues,70 thecustomerespeciallyforthosemoreinvolvedmaybetheauthoritativevoice. Alternatively, when introducing a product or brand values to prospects with little or no brand involvement, brand journalists may be best served by using corporate sources.Limitations and Future ResearchAs with any experimental design, this study was limited by the artificiality of the labo-ratory setting, as participants were given a portion of a magazine and told to focus on it. In addition, cumulative effects could not be studied. Experimental research, in gen-eral, is limited by the available participants, in this case, college students. To mitigate this, we recruited freshmen with no formal instruction in advertising. Furthermore, we chose a product within economic reach and of high interest to the participants (indeed, 43.3% of participants already had direct experience with a home theater system). The magazine created for this study mimicked those targeted toward younger adults. Still, effects might differ among other types of customers.The study also was limited by the frame manipulation. Although participants could name the producer of the magazine, more than a third struggled to identify whether the magazine was produced by a product manufacturer in an independent journalistic out-let. Lack of recognition between commercial and editorial frames is a common refrain in brand journalism, as studies have shown. Participants did better recognizing sources, but both manipulations could have been stronger. by guest on November 20, 2014 jmq.sagepub.com Downloaded from Cole and Greer687The model tested in this study contains a great deal of unexplained variance. This suggests other unexplored factors at work in shaping attitudes. Participants recogni-tion and understanding of custom magazines, general perceptions of the trustworthi-ness of advertising,71 and trust of media in general72 all might be related to attitudes about brand journalism. Furthermore, these attributes tend to become more pronounced with age, again suggesting research with nonstudent participants.Although some interesting differences emerged by involvement level in tolerance of commerciality, the experimental design does not allow us to decipher why this is the case. Focus groups or depth interviews with low, medium, and high involvement consumers might provide valuable insight.ConclusionThemoreneutralapproachofeditorialandtargetingofhighlyinvolvedconsumers appear to be strong choices for custom magazine producers, findings useful beyond the field of brand journalism. This study does support the efficacy of lower-cost alter-natives. For instance, the finding that an editorial frame can enhance brand attitudes and credibility supports results of previous studies regarding effectiveness of using an editorialframeinadvertorials73placedinmainstreammagazines.Theinvolvement findings lend credence to practices such as placing advertisements in highly targeted trade publications to reach highly involved audiences.All forms of brand journalism likely will rise to prominence as consumers become more desensitized to traditional advertising. Companies are beginning to recognize the advantage of owning their own media to communicate with customers directly. This study provides clues that may help them make the most of these opportunities, espe-cially for developing content in custom magazines. They must be careful about fram-ing their publications as too commercial, at least if they want to be seen as credible. And they should investigate using different types of sources to target customers with different levels of motivation (especially low and moderate involvement consumers) if they want to improve brand attitudes. This study did little to shed light on how to improvePIusingcontent.Still,theseresultsmaybevaluabletocontentcreatorsin building trust and loyalty with a brand, a primary goal of brand journalism.Declaration of Conflicting InterestsThe author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.FundingThe author(s) received no financial support for the research, authorship, and/or publication of this article.Notes1.Stefanie Bothelo, Report: Custom Content Market Spend Up to $40.2 Billion, FolioMag.com, March 22, 2012, http://www.foliomag.com/2012/report-custom-content-market-spend-40-2-billion (accessed March 23, 2012). by guest on November 20, 2014 jmq.sagepub.com Downloaded from 688Journalism & Mass Communication Quarterly 90(4)2.Bothelo, Report: Custom Content Market Spend Up to $40.2 Billion.3.LyndaDyson,CustomerMagazines:TheRiseofGlossiesasBrandExtensions, Journalism Studies 8 (4, 2007): 634-41.4.Robin Young, Super Bowls Pre-Pre-Game Show Brought to You by Rent-A-Center (audio podcast),February4,2010,http://hereandnow.wbur.org/2011/02/04/super-bowl-ads (accessed June 13, 2012).5.EvaA.VanReijmersdal,PeterC.Neijens,andEdithG.Smit,CustomerMagazines: Effects of Commerciality on Readers Reactions, Journal of Current Issues & Research in Advertising 32 (1, 2010): 59-67.6.Van Reijmersdal, Neijens, and Smit, Customer Magazines: Effects of Commerciality.7.Erving Goffman, Frame Analysis: An Essay on the Organization of Experience (New York: Harper & Row, 1974).8.DavidTewksburyandDietramA.Scheufele,NewsFramingTheoryandResearch,in Media Effects: Advances in Theory and Research, ed. Jennings Bryant and Mary B. Oliver (New York: Routledge, 2009), 17-33.9.DietramA.Scheufele,FramingTheory,inTheInternationalEncyclopediaof Communication,ed.WolfgangDonsbach(Malden,MA:BlackwellPublishing,2008), 1862-68.10.Tewksbury and Scheufele, News Framing Theory and Research.11.Robert M. Entman, Framing: Toward Clarification of a Fractured Paradigm, Journal of Communication 44 (4, 1993): 51-58.12.Dyson, Customer Magazines: The Rise of Glossies.13.Marian Friestad and Peter Wright, Persuasion Knowledge: Lay Peoples and Researchers Beliefs about the Psychology of Advertising, Journal of Consumer Research 22 (1, 1995): 62-74.14.Denis Wu and Ben Wasike, Perceived Credibility of Dietary Supplement Advertisements: AComparisonbetweenMediumandFormat(paperpresentedatthemeetingofthe International Communication Association, Chicago, IL, 2006).15.Chanthika Pornpitakpan, The Persuasiveness of Source Credibility: A Critical Review of Five Decades Evidence, Journal of Applied Social Psychology 34 (2, 2004): 243-81.16.GlenT.CameronandPatriciaA.Curtin,TracingSourcesofInformationPollution:A Survey and Experimental Test of Print Medias Labeling Policy for Feature Advertising, Journalism & Mass Communication Quarterly 72 (1, 1995): 178-89.17.Glen T. Cameron, Does Publicity Outperform Advertising: An Experimental Test of the Third Party Endorsement, Journal of Public Relations Research 6 (3, 1994): 185-207.18.Dyson, Customer Magazines: The Rise of Glossies, 635.19.AdamArvidsson,Brands:ACriticalPerspective,JournalofConsumerCulture5(2, 2005): 235-58.20.Andrew A. Mitchell and Jerry C. Olson, Are Product Attribute Beliefs the Only Mediator of Advertising Effects on Brand Attitude? Journal of Marketing Research 18 (3, 1981): 318-32.21.VanessaVoltolina,HowContentImpactsLeadGen,FolioMag.com,September23, 2009,http://www.foliomag.com/2009/how-content-impacts-lead-gen(accessedFebruary 26, 2012).22.AlexWang,IntegratingandComparingOthersOpinions:TheEffectsofThird-Party Endorsements on Online Purchasing, Journal of Website Promotion 1 (1, 2005): 105-29.23.Wang, Integrating and Comparing Others Opinions, 125.24.James C. McCroskey and Jason J. Teven, Goodwill: A Reexamination of the Construct and Its Measurement, Communication Monographs 66 (1, 1999): 90-103. by guest on November 20, 2014 jmq.sagepub.com Downloaded from Cole and Greer68925.CarlI.Hovland,IrvingL.Janis,andHaroldH.Kelley,CommunicationandPersuasion (New Haven, CT: Yale University Press, 1953), 10.26.Hovland, Janis, and Kelley, Communication and Persuasion.27.Hovland,Janis,andKelley,CommunicationandPersuasion;ArthurLupia,WhoCan PersuadeWhom?ImplicationsfromtheNexusofPsychologyandRationalChoice Theory,inThinkingaboutPoliticalPsychology,ed.JamesH.Kuklinski(Cambridge, UK: Cambridge University Press, 2002), 51-88.28.Dulcie L. Straughan, Glen L. Bleske, and Xinshu Zhao, Modeling Format and Source EffectsofanAdvocacyMessage,Journalism&MassCommunicationQuarterly73 (1, 1996): 135-46.29.McCroskeyandTeven,Goodwill:AReexaminationoftheConstructandIts Measurement, 91.30.MichaelSalwen,TheInfluenceofSourceIntent:CredibilityofaNewsMediaHealth Story, World Communication 21 (2, 1992): 63-68.31.ThanomwongPoorisat,BenjaminDetenber,VaniViswanathan,andHelenNofrina, PerceptionsofCredibility:AComparisonofUser-GeneratedandExpert-Generated Websites (paper presented at the meeting of the International Communication Association, Chicago, IL, May 2009).32.ElizabethJ.WilsonandDanielL.Sherrell,SourceEffectsinCommunicationand PersuasionResearch:AMeta-AnalysisofEffectSize,JournaloftheAcademyof Marketing Science 21 (2, 1993): 101-12.33.James N. Druckman, On the Limits of Framing Effects, The Journal of Politics 63 (4, 2001): 1041-66.34.Druckman, On the Limits of Framing Effects, 1045.35.Van Reijmersdal, Neijens, and Smit, Customer Magazines: Effects of Commerciality.36.Van Reijmersdal, Neijens, and Smit, Customer Magazines: Effects of Commerciality.37.AmitavChakravartiandChrisJaniszewski,TheInfluenceofMacro-LevelMotiveson ConsiderationSetCompositioninNovelPurchaseSituations,JournalofConsumer Research 30 (2, 2003): 244-58.38.Beverly A. Browne and Dennis O. Kaldenberg, Conceptualizing Self-Monitoring: Links to Materialism and Product Involvement, Journal of Consumer Marketing 14 (1, 1997): 31-44.39.MarkE.HavitzandDennisR.Howard,HowEnduringIsEnduringInvolvement?A Seasonal Examination of Three Recreational Activities, Journal of Consumer Psychology 4 (3, 1995): 255-76.40.Do-Hyung Park and Jumin Lee, eWom Overload and Its Effect on Consumer Behavioral IntentionDependingonConsumerInvolvement,ElectronicCommerceResearchand Applications 7 (4, 2008): 386-98.41.Du-HyungPark,JuminLee,andIngooHan,TheEffectofOnlineConsumerReviews on Consumer Purchasing Intention: The Moderating Role of Involvement, International Journal of Electronic Commerce 11 (4, 2007): 125-48.42.JohnDeighton,TheInteractionofAdvertisingandEvidence,JournalofConsumer Research 11 (3, 1984): 763-70.43.Alex Wang, The Effects of Expert and Consumer Endorsements on Audience Response, Journal of Advertising Research 45 (4, 2005): 402-12.44.VanReijmersdal,Neijens,andSmit,CustomerMagazines:EffectsofCommerciality, 63.45.Van Reijmersdal, Neijens, and Smit, Customer Magazines: Effects of Commerciality. by guest on November 20, 2014 jmq.sagepub.com Downloaded from 690Journalism & Mass Communication Quarterly 90(4)46.Two participants did not identify ethnicity. Percentages for ethnicity total more than 100% because participants could select multiple ethnicities.47.Dyson, Customer Magazines: The Rise of Glossies.48.JudithL.Zaichkowski,ThePersonalInvolvementInventory:Reduction,Revision,and Application to Advertising, Journal of Advertising 23 (4, 1994): 59-70.49.AndrewJ.FlanaginandMiriamJ.Metzger,PerceptionsofInternetInformation Credibility, Journalism & Mass Communication Quarterly 77 (3, 2000): 515-40.50.Nancy Spears and Surendra N. Singh, Measuring Attitude toward the Brand and Purchase Intentions, Journal of Current Issues & Research in Advertising 26 (2, 2004): 53-66.51.Spears and Singh, Measuring Attitude toward the Brand.52.Voltolina, How Content Impacts Lead Gen.53.Spears and Singh, Measuring Attitude toward the Brand.54.Dyson,CustomerMagazines:TheRiseofGlossies;Young,SuperBowlsPre-Pre-Game Show.55.Pornpitakpan,ThePersuasivenessofSourceCredibility;WuandWasike,Perceived Credibility of Dietary Supplement Advertisements: A Comparison between Medium and Format.56.Van Reijmersdal, Neijens, and Smit, Customer Magazines: Effects of Commerciality.57.KirkCheyfitz,EveryonesaStoryteller.Not,Storyworldwide.com,June6,2011,http://www.storyworldwide.com/our-view/everyones-a-storyteller-not (accessed March 27, 2013).58.Mitch Joel, Will a Brands Next Big Move Be a Journalism Department? Twistimage.com,May14,2011,http://www.twistimage.com/blog/archives/will-a-brands-next-big-move-be-a-journalism-department (accessed March 27, 2013).59.Straughan, Bleske, and Zhao, Modeling Format and Source Effects.60.McCroskey and Teven, Goodwill: A Reexamination of the Construct and Its Measurement.61.Van Reijmersdal, Neijens, and Smit, Customer Magazines: Effects of Commerciality.62.McCroskeyandTeven,Goodwill:AReexaminationoftheConstructandIts Measurement, 91.63.Wilson and Sherrell, Source Effects in Communication and Persuasion Research.64.Hovland, Janis, and Kelley, Communication and Persuasion.65.Van Reijmersdal, Neijens, and Smit, Customer Magazines: Effects of Commerciality.66.EvaA.VanReijmersdal,BrandPlacementProminence:GoodforMemory!Badfor Attitudes? Journal of Advertising Research 49 (2, 2009): 151-53.67.Richard E. Petty, John T. Caccioppo, and David Schumann, Central and Peripheral Routes to Advertising Effectiveness: The Moderating Role of Involvement, Journal of Consumer Research 10 (2, 1983): 135-46.68.ParkandLee,eWomOverloadandItsEffectonConsumerBehavioralIntention DependingonConsumerInvolvement;Park,Lee,andHan,TheEffectofOnline Consumer Reviews on Consumer Purchasing Intention.69.Arvidsson, Brands: A Critical Perspective.70.Poorisat et al., Perceptions of Credibility.71.HyeonjinSoh,MeasuringTrustinAdvertising,JournalofAdvertising38(2,2009): 83-103.72.Matthias Kohring and Jorg Matthes, Trust in News Media, Communication Research 34 (2, 2007): 231-52.73.WuandWasike,PerceivedCredibilityofDietarySupplementAdvertisements:A Comparison between Medium and Format.