THE ATTITUDINAL AND BEHAVIORAL EFFECTS OF PICTORIAL METAPHORS IN ADVERTISING: CONSIDERING NEED FOR COGNITION AND THE MEDIATING EFFECT OF EMOTIONAL RESPONSE By S OOJIN KIM A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERTISY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2016
2016 S oojin Kim
To God, my beloved family and friends f or all their love, care, and support
4 ACKNOWLEDGMENTS First and foremost, I must thank God and my family. I am indebted to many people in my completion of this dissertation. In particular, I would like to acknowledge my deepest gratitude to my advisor and chair for this dissertation, Dr. Jon D. Morris for his ongoing support, guidance and, mostly, his patience at the University of Florida. Whenever I faced any problem, h is mentorship encouraged me to complete this dissertation during my doctoral graduate journey. In addition, I deeply appreciate my dissertation committee members, Dr. Tom Kelleher, Dr. Moon Lee, and Dr. Chris Janiszewski for their insights great support and valuable inputs as I developed this dissertation. It was their keen insights and further guidance that enabled me to complete this research. I would also like to extend thanks to my friends who were very encouraging and helpful to me during this process. A special thanks to my parents and sisters who consistently support me with their endless love and faith.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 11 Overview ................................ ................................ ................................ ................. 11 Purpose of Current S tudy ................................ ................................ ....................... 19 Significance of Current Study ................................ ................................ ................. 20 2 LITERATURE REVIEW ................................ ................................ .......................... 23 Definition and Characteristics of Metaphors ................................ ........................... 23 Metaphors in Advertising ................................ ................................ ........................ 25 Verbal and Visual Elements of Ads ................................ ................................ ......... 27 P ictorial Metaphor and Headline ................................ ................................ ............. 30 Theories of Metaphor Comprehension ................................ ................................ .... 32 Individual Differences in Cognitive Motivation ................................ ......................... 34 Cognitive and Affective Responses on Metaphors ................................ ................. 39 Affective Perspectives on Metaphors ................................ ................................ ...... 41 Affective Perspectives on Persuasion ................................ ................................ ..... 42 Affective Priming Effect ................................ ................................ ........................... 45 Mediated E ffect by E motional R esponses ................................ .............................. 48 3 METHODOLOGY ................................ ................................ ................................ ... 52 Experiment 1 ................................ ................................ ................................ ........... 52 Study Design ................................ ................................ ................................ .... 52 Recruitment and Participants ................................ ................................ ........... 52 Development of Stimuli ................................ ................................ ..................... 54 Proced ure ................................ ................................ ................................ ......... 59 Pretest Procedure ................................ ................................ ............................ 59 Measurement Instruments ................................ ................................ ................ 61 Results ................................ ................................ ................................ ............. 63 Manipulation check ................................ ................................ .................... 63 Tests of hypothesis 1 ................................ ................................ ................. 64 Discussion ................................ ................................ ................................ ........ 65 Experiment 2 ................................ ................................ ................................ ........... 65
6 Study Design ................................ ................................ ................................ .... 65 Recruitment and Participants ................................ ................................ ........... 67 Development for Stimuli ................................ ................................ ................... 69 Procedure ................................ ................................ ................................ ......... 72 Pretest procedure ................................ ................................ ...................... 72 Main test procedure ................................ ................................ ................... 74 Measurement Instruments ................................ ................................ ................ 76 Pretest measures for manipulations ................................ ........................... 77 Dependent variables ................................ ................................ .................. 78 Moderator ................................ ................................ ................................ ... 79 Potential moderated mediator ................................ ................................ .... 80 Analytic Strategy ................................ ................................ .............................. 80 Results ................................ ................................ ................................ ............. 81 Pretest ................................ ................................ ................................ ....... 81 Reliability checks ................................ ................................ ....................... 81 Manipulation checks ................................ ................................ .................. 82 Testing for confounding effects ................................ ................................ .. 83 Tests of hypotheses ................................ ................................ ................... 84 Moderated moderation analysis ................................ ................................ 87 Moderated mediation analysis ................................ ................................ ... 92 Moderated mediating effect of pleasure ................................ ..................... 94 Ad attitude ................................ ................................ ................................ .. 95 Brand attitude ................................ ................................ ............................. 96 Purchase intention ................................ ................................ ..................... 98 Discussion ................................ ................................ ................................ ........ 99 4 GENERAL DISCUSSION ................................ ................................ ..................... 102 Concl usion ................................ ................................ ................................ ........... 102 Theoretical Implications ................................ ................................ ........................ 105 Practical Implications ................................ ................................ ............................ 110 Limitations and Suggestions for Future Research ................................ ................ 112 APPENDIX A QUESTIONNAIRES ................................ ................................ .............................. 114 B EXPERIMENTAL AD STIMULI ................................ ................................ ............. 117 LI ST OF REFERENCES ................................ ................................ ............................. 132 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 140
7 LIST OF TABLES Table page 3 1 Description of Participants for Experiment 1. ................................ ...................... 54 3 2 Product Types of the ( Pictorial M etaphor s) Ads in the Experiment 1 total 12 ad s ................................ ................................ ................................ ...................... 58 3 3 Product Types of the ( Non M etaphor ) Ads in the Experiment 1 total 12 ad s ... 58 3 4 Simple Main Effects of Means and Standard Deviations for Dependent Variables in the Experiment 1 ................................ ................................ ............. 64 3 5 2 x 2 x 2 F actorial D esign ................................ ................................ ................... 66 3 6 Description of Participants for Experiment 2. ................................ ...................... 68 3 7 Product Types of the ( Implicit M etaphor s) Ads in the Experiment 2 total 12 ads ................................ ................................ ................................ ...................... 70 3 8 Product Types of the ( Explicit M etaphor s) Ads in the Experiment 2 total 12 ads ................................ ................................ ................................ ...................... 70 3 9 Pretests for Each Dimension ................................ ................................ .............. 73 3 10 Measurement Instrument ................................ ................................ .................... 76 3 11 Multivariate Tests of Two Way Interaction ................................ .......................... 85 3 12 Univariate Results for the Two way Interaction Effect Between Type of Metaphors and Headline Copy on Dependent Variables ................................ .... 8 6 3 13 Adjusted and Unadjusted Group Means for Ad Attitude, Brand Attitude, and Purchase Intention ................................ ................................ .............................. 87 3 14 PROCESS Model 3 Results on Purchase Intention ................................ ............ 92 3 15 Moderated Moderation Analysis Results: Effects of Type of Visual Metaphors on Purchase Intention Responding by Copy/Non copy and NFC level. .............. 92
8 LIST OF FIGURES Figure page 3 1 The FCB grid. ( Vaughn 1980) ................................ ................................ ............ 56 3 2 PROCESS Model 3 (Hayes, 2012) Moderated Moderation Analysis ............... 89 3 3 Three way Interaction Effect on Purchase Intention ................................ ........... 91 3 4 Three way Interaction Effect on Purchase Intention ................................ ........... 91 3 5 PROCESS Model 11 (Hayes, 2012) Moderated Mediation Analysis ................ 94 3 6 Conditional Indirect Effect on Attitude Toward the Ad ................................ ........ 96 3 7 Conditional Indirect Effect on Attitude Toward the Brand ................................ ... 97 3 8 Conditional Indirect Effect on Purchase Intention ................................ ............... 99
9 Abstract of Dissertation Presented to the Graduate School of the University of Florida Requirements for the Degree of Doctor of Philosophy THE ATTITUDINAL AND BEHAVIORAL EFFECTS OF PICTORIAL METAPHORS IN ADVERTISING: CONSIDERING NEED FOR COGNITION AND THE MEDIATING EFFECT OF EMOTIONAL RESPONSE By Soojin Kim August 2016 Chair: Jon D. Morrison Major: Mass Communication (Sopory tropes in ads provoked greater elaboration and positive ad appreciation than did verbal schemes. It is also necessary to conduct research on how persuasive responses resulting from the emotional processing of metaphorical ads attitude and behavioral intention. Thus, the current study provides the opportunity to consider the effect of the pictorial metaphor sty le in ads as a m ethod of visual thinking with verbal support, and to examine the persuasive impact of ads with visual metaphors on emotional responses as well as ad attitudes brand attitudes and purchase intention s An experimental design is used to inve stigate the effect of the interrelation between pictorial metaphor and headline copy in ads Need for Cognition ( NFC ) on the attitude toward the ad, brand, and purchase intention, while considering emotions as a mediator.
10 Two experiments were conducted: Experimental study 1 compares the differences between non visual metaphor ad and visual metaphor ad all across the types of product. An e xperimental study 2 between subjects randomized factorial design 2 (Pictorial metaphor type: implicit vs. explicit) x 2 (Headline type: directly illustrated copy vs. non copy) x 2 (Need for cognition : High vs. Low ) is proposed to test the hypotheses. The study also consider s that cognitive influences on the effectiveness of visual metaphors and emotional responses in metaphorical ads may have a crucial impact on ad effectiveness, while verbal elements in the metaphorical ads play a role in facilitating the interpretation of pictorial metaphors by consumers. Even though, with the assistance of headline copy, consumers completely understand the metaphorical meaning in the ad, if the ad with metaphors cannot finally evoke positive emotional responses, there is no positive effect from using such metaphors. he importance of the match between pictorial metaphor and headlines for ad of consumers considering the affe ctive responses of target consumers. In addition, the findings of this research provide better understanding or theoretical knowledge about the use of pictorial metaphor in ads and the effects they have on the attitude and behavior intentions of consumers.
11 CHAPTER 1 INTRODUCTION Overview Metaphors have become common in modern forms of communication, especially in the advertisements of today (Chang & Yen, 2013; Geary, 2011; Malkewitz et al., 2003; Phillips & McQuarrie 2004; Rossiter, 2008). Pictorial metaphors are often employed in print advertising and advertisers expend a great deal of time and energy in selecting appropriate visual/verbal expression s for product presentation to convey the main benefit of the adver tised product (Bergkvist, Eiderback, & Palombo, 2012) Merriam Webster defines metaphor as a word or phrase for one thing that is used to refer to another thing in order to show or suggest that they are similar (Merriam Webster, 2016). In previous studies A metaphor depends on cross domain comparisons and can be defined as a type of rhetorical figure or as an artful deviation from common expectat ion (McQuarrie & Mick, 1996). A metaphor compares two objects through analogy by suggesting that one object is figuratively or semantically like another, even though on the surface they appear to be quite different (Ward & Gaidis, 1990). In literature, one can find innumerable uses of metaphor. Shakespeare is widely from Romeo and Juliet (Shakespeare 1873). Although the use of metaphor in literature may l ead to the conclusion that metaphor is typically used only by writers, in fact it is used in everyday communication as well. Common examples of metaphor in daily use you are my sunshine marathon
12 various thoughts, emotions, or concepts, if metaphors are used effectively we do not need to have numerous modifiers or a good number of illustrative phrases. There would be a need to depict a sentence or a figure with metaphor. When we describe abstract things such as concepts ideas, and feelings, we very often and instinctively draw on metaphor (Geary, 2011). The influence s on every aspect are found, from daily communication and advertising to news headl psychology. In this light, metaphors work in all fields of our lives, even in imagery. V arious ad elements such as verbal or pictorial metaphor s, effectively convey the core benefits of products or services to consumers, persuading them to make purchases. Most of all, visual elements in ads can positively impact the effects ads achieve such as changing brand attitudes (Rossiter & Percy, 1980) and ad recall or recognition ( Jeong, 2008; Lutz & Lutz, 1977 ; McQuarrie & Mick, 1999, 2003; Zaltman, 2003 ). In particular, pictorial m etaphors as visual elements in ads can help receivers engage in messages and motivate changes in their attitudes toward ads a nd brands ( McQuarrie & Mick, 2003, 2009; Rossiter & Percy, 1980). The use of metaphor s in ads continues to rise consistently (Kim, Baek, & Choi, 201 2; Phillips & McQuarrie, 2003); moreover, t he use and proportion of metaphorical pictures or images in prin ted ads grew dramatically and rapidly during the twentieth century, while the proportion of verbal copy in ads steadily decreased (Phillips & McQuarrie, 2003 ; visual tropes in ads provoked grea ter elaboration and positive ad appreciation (ad liking) than did verbal schemes. Pictorial metaphors are an artful deviation (McQuarrie & Mick,
13 1996) and under this notion, the metaphors could have advantages to attract attention and add interest to an ad (McQuarrie & Mick, 1999). The use of pictorial metaphor s in advertising has benefits that can create additional entertainment or enjoyment (McQ uarrie & Phillips, 2005), and ad engagement (McQuarrie & Mick, 2009). Metaphors may influence various consumer responses regarding credibility (Jeong, 2008), consumer belief (Philips & McQuarrie, 2009), persuasion (Tom & Eves, 1999) and recall (Jeong, 2008 ; McQuarrie & Mick, 1999; Zaltman, 2003). Advertisements with pictorial metaphors are expected to foster a pleasurable experience (Forceville 199 4; McQuarrie & Phillips, 2005 ; Sopory & Dillard, 2002 ). The previous research only suggested a conceptual appr oach on the pleasurable experience from pictorial metaphors without any conditional experiment. Pictorial metaphors sometimes combine two mismatched images or contrasting meanings together, often without accompanying verbal explanations. Pictorial metaphor s, therefore, tend to be more inherent and complex than verbal metaphors, leading viewers to describe several possible interpretations (Jeong 2008) Along with this complexity comes images or meanings in ad vertisements If some metaphorical ads have less direct metaphors in the complex figurative form, receivers require more effort and knowledge to interpret them Chang and Yen (2013) classified pictorial metaphors as being composed of two st yles of metaphor: implicit and explicit. According to the study (Chang & Yen, 2013), an explicit metaphor refers to the metaphor in the ad that is clear about the subject with the product image. The advertised product is presented to the viewer clearly wit h a visual metaphoric
14 display in explicit metaphors (e.g., a MacBook Air coming out of an envelope). In the current study, an explicit metaphor refers to the pictorial metaphor presented with the advertised product in the ad, and the advertised product is used with other objects or with a visual metaphoric display in the ad. Visual information in the explicit metaphors is synthesized or juxtaposed. An implicit metaphor does not show the product directly in the metaphoric illustration of ads but occasionall y the advertised product could be shown at the bottom corner of the ad (Chang & Yen, 2013). Pictorial metaphor ads have sometimes been discussed as though they have one clear or unambiguous meaning and effect. However, do these ads always result in ad eff ect iveness (e.g., attitude toward the ad, attitude toward the brand, and behavioral intention) ? A lot of the positive effects of ads containing metaphors are not free of comprehension costs. Consumers do not always comprehend all metaphorical meaning corre ctly and sometimes do not understand what the ad wants to convey to consumers M etaphorical persuasion often concerns marketers and advertisers who worry about the possibility of misleading ads, or ads that result in misunderstanding. Ads that utilize pict orial metaphors may not always work under various conditions because consumers have to interpret the metaphors by themselves via their own abilities or through their own experiences. Consumers understand the ad messages communicated by pictorial metaphors through their own interpretive (or cognitive) process, and the interpreted meaning of the ad may differ among individuals or under various ad conditions. As one of the conditions, the type s of headline, including no headline, may persuasion ( Bergkvist, Eiderback, & Palombo, 2012; Philips, 2000) in metaphorical ads
15 For example, some ads with pictorial metaphors employ a headline to help the recipien ts understand what the ad is trying to convey, whereas other ads do not use any headline copy in the ads (Bergkvist, Eiderback, & Palombo, 2012; Philips, 2000). Phillips (2000) and attitud e toward the ad with complex images The study illustrated three levels of completeness of headlines: no headline (no copy), a moderate headline, and a complete headline, which are based on different amounts of verbal anchoring (including no copy) to a met aphorical image (Phillips, 2000). According to Philips (2000), a p. 18). In the case of the ad employed by complex images headlines may increase comprehension of an ad because those might help explain the meaning of the ad by providing a clue to the meaning of the ad image (Philips, 2000). H owever, the headlines decrea se attitude toward the ad : that is, complete verbal anchoring causes a lower attitude toward the ad than non headlines in the study (Philips, 2000). The researcher just inferred that the reason might come from decreasing meaning of the ad employed by complex images responses. When consumers are exposed to an ad employ ing different kinds of pictorial metaphors, with or without headlines what are the ch anges in cognitive, affective, and intended behavioral responses? Do these lead to positive attitudes and behavior al intention in all conditions ? In terms of the cognitive aspect, Need for Cognition (NFC)
16 illustrate s the difference of the individual tenden cy to engage in and enjoy cognitive endeavors when people deal with tasks and social information ( Cacioppo, Petty, & Kao, 1984) Therefore, in the current research, individual differences in NFC are further considered. In addition to the individual differences in cognitive tendency the current study considers the affective responses and the impact on the effect of ads as well According to traditional studies and theories (Klauer, 1998; Klauer & Musch, 2001; Murphy & Zajonc, 1993; Zaj onc, 1980) affect is considered to be post cognitive and to occur only after considerable cognitive operations. At the same time a number of experimental studies on preferences, attitudes, impression formation, and decision making, as well as some clinical phenomena, showed that affective judgments may be fairly independent or may take precedence over perceptual and cognitive operations (Morris, Woo, & Singh, 2005). However, in the case of the ad with a pictorial metaphor, particularly an implicit m etaphor, the cognitive process for the interpretation occurs first and then the emotional responses, such as pleasure or enjoyment, may be increased through creativity or novelty of the ad, by understanding (Sopory & Dillard, 2002). Furthermore, accordin g to the affective priming theory ( Klauer, 1998 ), the affective responses may be transferred to the evaluation of the ad itself or the attitude toward the advertised product and the brand, which are the target of the metaphor. This unintentional influence of a first evaluative response to the metaphor ad on subsequent processing of the target (i.e., the attitude toward the ad or the brand) can be substantiated by affective priming theory ( Klauer, 1998 ). Affective priming theory posits that the positive or n egative evaluation of the previous stimulus transfers to the next
17 evaluation (Klauer, 1998). The affective responses influence the evaluation conditioning of the subsequent processing. It could be said that this leads to a more favorable attitude toward th e ad and the brand or that it is more likely to change the attitude or behavior intention positively if the metaphorical ad results in positive emotional responses. Researchers have discovered extensive information that predicts consumer resp onse to different types of verbal presentations, as well as how effective pictorial metaphor s can be in several situations (DeRosia 2008; Phillips 2003; Rossiter & Percy, 1980; Zaltman 2003). A lthough few empirical studies exist using a visual metaphor perspective and the interest in metaphor is increasing, the majority of research has been limited to linguistic metaphors, specifically focusing on verbal metaphors (Forceville, 1998). In addition, even empirical studies on metaphors in ads that do exist e xplore the effects of pictorial metaphors without examining the impact of the verbal copy apart from the visual image (McQuarrie & Mick, 1992). For example, McQuarrie and Mick (1992; 1999) found that ads with rhetorical images generate rich elaboration and more positive responses than similar ads without these images. However, they did not consider the differences in experimental ads with verbal headlines and the effects of verbal copy with visual metaphors simultaneously. P ictorial metaphors are a f avorit e communication technique used by marketers and advertisers and there is a steadily increas ing usage in advertising ( Chang & Yen, 2013; Phillips & McQuarrie, 2004 ). However, limited research has been conducted to identify the effect of typology of pictorial metaphors in advertising, including verbal aid s individual difference s in cognitive responses and the role of emotional responses. O nly
18 limited empirical research has been conducted to understand the extent to which the advertising variables in (Chang & Yen, 2013; Bergkvist, Eiderback, & Palombo, 2012). Unfortunately, there is still little consumer or marketing theory available for differentiating the benefits of pictorial metaphors. There is also little research on organizing the variety of imagery strategies on display in advertising (Malkewitz, Wright, & Friestad, 2003). Whereas extensive research has been carried out on other techniques and their importance to overall ad effects, r elatively little is known about ads with pictorial metaphors (Delbaere, McQuarrie, & Phillips, 2011; Jeong, 2008) Under what conditions are those ads that are the verbal/visual elements of metaphor most effective at reaching consumers ? H ow do me taphorica l ads affect their in affective and cognitive aspects under various conditions? In fact, the role of visual metaphors in ads is so widely accepted that researchers have lar gely neglected the possibility that moderating or mediating factors may exist that render visual metaphors more or less influential in the positive ad effectiveness. To answer those questions and to identify the appropriate and effective ad conditions wi th metaphors research proposes to examine the effect of pictorial metaphors in ads with headlines in terms of cognitive and affective aspects The current study will also consider the influences o f the headline in ads and discuss one of the important motivational variables, such as need for cognition as individual differences (Brennan & Bahn, 2006), which could influence audience comprehension of metaphors in ads (Morgan &
19 Reichert, 1999; Pawlowsk i, Badziski, & Mitchell, 1998), and the impact of emotions. Purpose of Current S tudy Given the potential value of pictorial metaphors for advertisers and marketers, the current study tests and develops a theoretical explanation for the effects of such metaphors on consumer attitude towards the ads and the brands, and the purchase intention created by cognitive and affective perspectives considering the interrelationship between pictorial metaphors and headlines. Using mediation analysis, the stu responses to the metaphorical ads on ad effectiveness that employed the pictorial metaphors with or without headline support. The purposes of th is study are as follows: (1) to explore the impact of using pictorial metaphors with verbal elements in ads intention considering the level of need for cognition (NFC); and (2) to investigat e how to metaphor ads can mediate ad e ffectiveness (i.e., attitude toward the ad and the brand, and purchase intention) where metaphor types (e.g., implicit, explicit, and non metaphor) and the types of headline copy (e.g., no headline and complete headline) are manipulated. This enables the r esearcher to better understand and explain how pictorial metaphors in ads influence attitude formation and behavioral intention in consumers while considering other variables. Consideration is given to the cognitive and affective responses of consumers wit h various pictorial metaphors and headlines for all types of products.
20 Significance of Current Study Amid a deluge of advertising and information, features of ads need to be differentiated from similar ads with which consumers are annoyed especially beca use the market environment is gradually becoming more complicated and diverse. Accordingly, marketing managers and advertisers need advertisements that feature arresting images, create a totally new experience, are more impactful, and are easier to remembe r. They are always looking for good advertisements that stylishly assault a a brand and enhance their intention to purchase. R hetorical figures, such as those encountered in metaphor, are often used to draw the attention of consumers and to improve their memory (Lutz & Lutz, 1977; Zaltman, 2003). Ads that use pictorial metaphors are more likely to stand out from the normal, cluttered viewing conditions (McQuarrie & Mick, 20 02). Some studies, such as that by Jeong (2008), have found that metaphors (e.g., visual metaphors) may be more effective than non metaphor or literal verbal arguments, but there has been no rigorous attention given to exploring whether these persuasive in fluences are the results of the pictorial metaphor styles (e.g., implicit/explicit metaphors) or argument modality (i.e., only with visual or with verbal). Research also needs to be conducted on how persuasive responses, resulting from the affective proces and behavioral intention. Th us, the present study provides the opportunity to consider using different types of pictorial metaphors in ads as a method of visual thinking with a verbal element (i.e., headlines) Also examined is the persuasive impact of ads with visual metaphors on
21 intention. The investigation of the impact of pictorial metaphors on ad effec ts under various conditions, which has received attention but has had limited empirical testing, would provide both theoretical and practical implications for advertising scholars and practitioners. From a theoretical perspective, application of pictorial metaphors would be theoretically extended by considering both the cognitive and affective aspects and by examining the interrelationship between visual metaphors and verbal elements in ads. The study attempts to show that visual metaphors enhance understa nding of advertising messages, elicit and thereby positively affect the attitude toward ads, attitude toward brands, and purchase intention. The study also considers cognitive influences on the effectiveness of visual metaph ors. Affective responses in metaphorical ads may have a crucial impact on ad effectiveness, while verbal elements in the metaphorical ads play a role in facilitating the interpretation of pictorial metaphors by consumers. Even though, with the assistance o f headline copy, consumers completely understand the metaphorical meaning in the ad, if the ad with metaphors cannot finally evoke positive emotional responses (e.g., pleasure), there is no effect from using such metaphors. From a managerial perspective, the results of this study enable advertisers to gain important practical insights into the use of visual (pictorial) metaphors with verbal elements. The study also helps marketers or advertisers decide whether a certain type of pi ctorial metaphor with verbal elements (i.e., headlines) creates more positive ad effects (e.g., understanding messages and more favorable emotional reactions), compared to other conditions (e.g., visual metaphors only or even non metaphor). This
22 study also contributes to providing direction for the use of pictorial metaphors in ads and for suggesting target group s of consumers for whom it would be appropriate to use pictorial metaphor with complex meaning in ads The result would be the development of impo rtant empirical research not only for researchers but also for the practitioners who must decide if using such pictorial metaphors in ads would be effective pictorial metaphor and headlines for ad effects by consumers cognitive tendenc ies considering the affective responses of target consumers. In addition, t he findings of this research will provide a better understanding or knowledge of pictorial metaphor u sage in ads and the effect s they have on the attitude and behavior al intentions of consumers.
23 CHAPTER 2 LITERATURE REVIEW Definition and Characteristics of Metaphors A metaphor is one type of rhetorical figure or trope a category that also includes irony, puns, and other literary devices (Phillips, 2003). McQuarrie and Mick (1999) have but nonetheless illuminating equation of two d A metaphor is from an expected concept and an invitation to compare two ideas or objects and infer what they have in common (Kim, Baek, & Choi, 201 2 p. 77 ). De la Rosa (2009) defined metaphor as with abstract domains of experience by understanding and experiencing one thing in In addition, Lakoff and Johnson (2003) described metaphors hor is a rhetorical style comparing two dissimilar objects; the feature s of one object is transferred onto the other through the comparison (Sopory & Dillard 2002) and interpreted as different from the original. M etaphor uses one object to represent anoth er concept and helps the recipients understand compl icated and unfamiliar notions (ideas) by using simple and familiar ones for illustration (Lakoff & Johnson, 1980 ; 1999 ). Author James Geary (2011) explained that metaphorical thinking is an important way to understand ourselves and others, and the way that allows us to communicate, learn, and even create new ideas M etaphor is one way to communicate no matter what words or image s exist ( Geary 2011 ) Because metaphors are used in many kinds of field and sneak into our purchasing decision s Geary (2011) asserted that voters, consumers, investors, and all should know the role of metaphors in our communications
24 and the effects that well designed metaphors can have on decision making According to Geary (2011 ), people experience pleasure during the process of decoding metaphors, a process he refers to as "cognitive gymnastics This phrase also allows us to comprehend the parallels between two dissimilar objects (i.e., the process of decoding metaphors and gym nastics). p.444). When recipients fulfill the gap between two things, they probably need some minimu m level of cognitive effort ( Brennan & Bahn, 2006) The features of metaphors require recipients to be active interpreters of meaning. That is, recipients exert cognitive effort only when they are motivated and able to adequately interpret meaning. A kind of chain of related abductive or cognitive processes (Brennan & Bahn, 2006) may be necessary to interpret the ad concepts or strategies. The cognitive processes of interpreting such metaphorical meaning can be expected to require devot ed cognitive effort. Several researche r s (e.g., Brennan & Bahn, 2006) have asserted that motivation is a necessary antecedent of cognitive effort to interpret meaning. According to the study by Brennan and Bahn (2006), decod ing metaphor meaning or extended visual/verbal metap hor s requires the use of extensive cognition. The r ecipient has a high need for cognition in order to interpret a metaphor ical meaning, and individuals with a high need for cognition are better able to interpret it. The equation format for metaphor s is ver y simple. That is A is B.
25 two different ideas and are inferred by w hat they have in common (Kim, Baek, & Choi, 2012). Examples of basic conceptual metaphors includ e ( Lakoff & Johnson 2003 ) : Life is a journey Home is paradise Her lips are a volcano Social organizations are plants Love is magic For example, a piece of literature might use a metaphor by stating that life is a journey; it compares life to a journey to infer that the two concepts have something in common. Metaphors in Advertising Metaphor can be visual or verbal. We can find innumerable usages of verbal me taphors in literatures and lyrics. As mentioned before, Shakespeare is considered one of the most famous masters of metaphor in literature. Metaphors are frequently used in such areas as modern pop music We can easily access examples of metaphors in music such as those of Elvis Presley Presley was so successful at this endeavor that he could be considered a metaphor himself. Previous research on metaphors has primarily focused on the verbal component. Researchers have seldom attempted to stu dy metaphors in relation to their non verbal expressions. Among those who have explored this area, Forcevill (1998) has done extensive research into the use of pictorial metaphors in advertising, comparing them with non metaphors. In external appearance, o ne of the typical characteristics in modern advertising is the reliance on visuals for attractive attention and persuasive communication (Phillips & McQuarrie 2003). Furthermore, the level of dependence on pictorial metaphors in modern advertising has incr eased (Phillips & McQuarrie 2003).
26 can have a further advantage because such inferences are more likely to be generated spontaneously at the time of ad exposure ( McQua rrie & Phillips, 2005 ) The current study focuses on visual (i.e., pictorial) metaphors in ads. Gkiouzepas and Hogg (2011 ) presented a conceptual framework for pictorial metaphors in advertising, based on whether visual information in the ad is synthesize d or juxtaposed. The framework for pictorial metaphors included a structural and conceptual investigation that systematically categorizes the different patterns of representational discordance found in pictorial metaphors. In their study Gkiouzepas and Hogg ( 2011 ) categorize the forms of visual metaphors with reference to t wo mode of representation (i.e., juxtaposition and synthesis), and second, the visual scenarios (i.e., realistic symbiosis, replacement, and artificial symbiosis). Chang and Yen (2013) have specified that visual metaphors exist in two categories: implicit and explicit metaphors, depending on whether the advertised product is included in the metaphorical illustration of the ad or not. An advertised produ ct is not expressed or integrated into the main (metaphoric) images in an ad consisting of implicit visual metaphors. In contrast, an explicit metaphor shows the product directly in the metaphoric illustration of ads. The current study uses the conceptual implicit and explicit pictorial metaphors) The construct of distinguishing implicit or explicit metaphors is the ability to make an inference. It is the ability to infer from a metap hor what has been communicated i.e., what the exact meaning is.
27 Since metaphors are often used in all kinds of communications, it can easily be surmised that the impact of metaphor is powerful and effective, particularly in communication. Metaphors have been applied to ads for a long time to incorporate more meaning and to convey complex messages in advertising and marketing communication (Boozer, Wyld, & Grant, 1991; Phillips, 2003). Metaphors in advertising obviously draw attention away from the clutte r of typical advertisements and persuade consumers in unique ways (Phillips, 2003). The positive impact of metaphors can be illustrated with respect to attitude change and memory outcomes (McQuarrie & Mick, 2003a) study, if this study simply compares and juxtaposes metaphorical objects with ad visuals that synthesize conceptually similar metaphorical objects, the ad visual elements provoke greater elaboration and elicit more favorable consumer attitudes toward both the ad and the brand. P revious research leads to the following hypothesis: H1: The effect of using visual metaphors in an ad will have a more favorable influence on (a) attitude towards the ad, (b) attitude towards the brand, and (c) purchase intention, t han will a non metaphor ad. Verbal and Visual Elements of Ads Ads generally consist of two elements images as visual elements and headline copy as verbal. One of the interesting studies on the effect of verbal and visual stimuli in communication is ad effects in terms of responses to a visual, a verbal, and a combination of visual/verbal stimulus. Their exploratory study shows that individuals who have high affective processes respond more favorably to a visual ad than the other groups, and individuals who are high in both affect and cognition ( i.e., individuals with combined processes) respond more favorably to a combination visual/verbal ad (Sojka & Giese, 2006).
28 According to Lutz and Lutz (1977, p the potential to be more generated by their own mental processes (L utz & Lutz, 1977). Imagery has multiple channels of information processing and therefore a greater possibility to enhance advertising effectiveness (Lutz & Lutz, 1977). Affect and cognition, however, do work together as well as independently in ads Empirical and theoretical research ha s shown that individuals who appear to be high in affective processing styles would neglect the verbal elements in the ad and pay more attention to the visual elements, whereas individuals who tend to have high cognitiv e processing would overlook the visual elements in the ad and focus more on the verbal (Sojka & Giese, 2006). Sojka and Giese (2006) explored the processing of two different advertising stimuli visual and verbal They investigated the interrelationship between visual and verbal stimuli, comparing these ads to different groups identified as having affect ive or cogniti ve disposition Interestingly, Sojka and Giese (2006) wove individuals who have an affective tendency and visual information about advertis ing in to their study. However, not all advertising is produced with strictly visual content; therefore they investigated individuals who have a tendency to respond to cognitive and verbal information as well. A lot of researchers have investigated the in terwoven studies with the predisposition for affect ive tendency and a preference for visual stimuli, and with the cognitive tendency and a preference for verbal stimuli (Martin, Sherrard, & Wentzel, 2005; Sojka & Giese, 2006, 2001; Venkatraman, Marlino, Ka rdes, & Sklar, 1990 ). In
29 addition, consistent with the study by Cacioppo and Petty (1982), individuals with a high need for cognition and low affect ive tendency generally prefer to respond to verbal, factual information, while individuals with a high affect ive tendency and low cognition usually rely on emotion to make decisions and respond more positively to visual information (Childers, Houston, & Heckler, 1985; Richardson, 1977; Sojka & Giese 2006). In addition, a goal directed paradigm (Ressler, 2004) illustrates that humans are motivated by pleasure rewards. For example, dealing with pictures is more pleasurable than with words for people with predisposition f or affec t. Consequently, those people will process the visual elements and overlook the verbal elements in the ad. Various researchers have established the concept that some individuals prefer visual information and other individuals prefer verbal informa tion (Childers, Houston, & Heckler, 1985; Richardson, 1977). Furthermore, other researchers found connections between people who have a more affective tendency and a stated preference for visual stimuli, and between people who have a more cognitive tendenc y and a preference for verbal stimuli (Martin, Sherrard, & Wentzel, 2005; Sojka & Giese, 2001; Venkatraman, Marlino, Kardes, & Sklar, 1990). The study of Sojka and Giese (2006) suggests that images differ from words in the theoretical conceptualizations o f information processing. In addition, they assert that in the processing of visual stimuli and affect, individuals who tend to be more affective would respond more positively to visual information than individuals who tend to be more cognitive (Sojka & Gi ese, 2006). Visual thinking, therefore, may handle abstract concepts, including metaphor, more easily than verbal thinking for individuals who tend to be affective
30 P ictorial Metaphor and Headline Visual (i.e., pictorial) metaphors place or combine two co ntrasting images together, often without accompanying verbal explanations. The pictorial metaphors, therefore, tend to be more implicit and complex than verbal metaphors and can lead viewers to imagine several possible interpretations (Jeong 2008; McQuarri e & Mick 1996; Phillips 2000). Because metaphors are persuasion based on messages, they need conceptual interpretation. If an ad employs less related metaphors in the visual image and has no headlines, consumers require more cognitive effort and knowledge of the context to interpret and understand the ad s Mitchell and Olson (1981) explained that pictorial metaphors might be more effective than literal, verbal arguments in terms of persuasive outcomes. Philips and McQuarrie (2009) asserted that metaphors are used to communicate concrete information on consumer beliefs, and th at similar contexts can help people understand more abstract and different domains. In addition, verbal metaphorical claims in ads create further enjoyable advantages, and the effect of the metaphor may influence the attitude toward the ad or brand (Philips & McQuarrie 2009). Visual rhetorical figures such as trope (e.g., visual metaphor and pun) and scheme (e.g., visual rhyme and ant ithesis) generate a large amount of cognitive activity ( i.e., elaboration) and result in a positive impact on consumer response (McQuarrie & Mick 1999) without being any more difficult to understand If rhetorical figures appear insightful, they may insti ll a pleasurable response (Ward and Gaidis 1990). Some researchers classified the types of headlines in ads depending on the levels of completeness of those headlines (i.e., the amount of verbal anchoring added to an image) such as no headline, a moderate headline, or a complete headline (Bergkvist,
31 Eiderback, & Palombo, 2012 (2000), the effect of a higher level of headline completeness has a positive impact on the comprehension of the ads because headlines provide a clue to the meaning of a pictorial metaphor and this facilitates increasing comprehension. When ads with pictorial metaphors are designed with a headline, the types of headline s lead to different results in emotional effects depending on the degree of understanding (Philips, 2000). The higher the level of completeness a headline has the more comprehension it generates, even though this high level of completeness results in less pleasure (Philips, 2000). Furthermore, Bergkvist, Eiderba ck, and Palombo (2012) suggest that the use of a complete headline in metaphorical ads results in more positive effect on brand communications such as brand beliefs about the key benefit. In line with results of the study by Phillips (2000), the idea that headlines have a positive effect on attitude toward ads and brands is also supported (Bergkvist, Eiderback, & Palombo, 2012). As previously mentioned, some researchers (e.g., McQuarrie & Phillips, 2005; Phillips, 2000) examined the effects of headlines on ads with pictorial metaphors or on literal (straight) ads. However, they did not examine the effect that adding headlines to different types of pictorial metaphors would have on perception. When headlines are added, there will be differences of effort in understanding the ad messages according to the types of pictorial metaphors, and different effects will be generated from the ads. In addition, Bergkvist, Eiderback and Palombo (2012) examined the effects of headlines in metaphorical ads using one type of metaphor as the stimuli in their study. They theorized that the use of headlines in metaphorical ads brings about a positive effect on brand communications such as brand beliefs.
32 The research question on the interaction of pictorial metaphors and verbal cop y in advertisements leads to the following hypothesis: H2: Under conditions in which implicit metaphorical images are employed in ads, directly illustrated headlines (=completeness) will have a more positive impact on the ad effect s (i.e., (a) attitude toward the ad, (b) the brand, and (c) purchase inten t ion) than no headline, while no significant difference is expected in explicit metaphorical ads Theories of Metaphor Comprehension In order to explain the way in which metaphors in ads stimulate posit ive influences on consumers, it is necessary to describe the theories that increase understanding about the perspective impact of metaphor. Sopory and Dillard (2002) applied the literal primacy view as a theory that is especially relevant. Even though the claims in the study are from the perspective of verbal metaphorical ads, such theories may explain how the responses develop when metaphors are decoded and interpreted. Among them, the literal primacy view (Beardsley, 1962, 1976; MacCormac, 1985) regards m etaphor as literally false or exceptional language. Metaphor are therefore seen as semantic anomalies and they can easily be described by definitions such as According to the literal primacy view, the process of understanding metaphors req uires three stages: (1) Development of a method to find the literal meaning of a metaphor ; (2) testing whether the literal meaning makes sense and finding an anomaly or a violation of semantic rules ; and (3) seeking an alternati ve meaning (i.e., the metaphorical meaning) if the literal meaning does not make sense (Gibbs, 1994; Sopory & Dillard, 2002). For example when the consumer finds that there is an anomaly or an expression that is literally false, cognitive tension is creat ed, along with a desire to reduce it. When the meaning of the literally false
33 the interpreter, and the tension is reduced or disappears. When consumers confront the vi sually false expression (i.e., anomaly) in ads, the could be indicated by relevance theory ( Sperber & Wilson, 1995, 2004). Relevance theory, proposed by figurative langua ge scholars Sperber and Wilson ( 1995), argues that the receivers seek and find the meaning that fits their expectations of relevance. According to scholars of language, there are two ways to explain how ideas or abstract concepts are communicated. First, people c ommunicate by the process of coding and decoding. In this way, the author or advertiser encodes their intended ideas and conveys them to their receivers or targets. The receivers take the encoded message and decode it to get the meaning the author or advertiser intended. In this approach, the receivers almost always employ context in communication. Other factors considered by the receivers include contemporary issues, and cation are transferred to receivers by the author or advertiser. In this case, they may convey only as much information as needed for the given context, and then the final steps depend on the receivers. Receivers can interpret the intention or meaning from the communication itself as well as from the context and implications. This conceptual process shows how the author or communicator uses the context of the communication and demonstrates how the cognitive environment between the author/advertiser and the consumer develops.
34 According to Sperber & Wilson (2004), the communicative principle of relevance is the core of relevance theory. The relevance theory states that when a speaker says what he/she wants to convey, the act of making an utterance will provi de cognitive effects of processing effort to find the meaning. In this way, when speakers make an utterance, every ostensive act of communication, such as the lexical clues or visual clues, shows something. Sperber & Wilson (2004) illustrated the process to find the meaning with two general processes. First, the speaker inte ntionally gives a clue to the audience in order to convey exactly what he/she wants to communicate. That would be an ostensive act. Second, the audience or the receiver infers the intention from the clue and other context information. Then, the audience seeks meaning and interprets the meaning with the clue and context in any given communication situation. According to relevance theory, in explicit metaph or in advertising, the advertised product could be the ostensible stimulus, which provides the relevance with the ad in metaphorical images. If a metaphorical ad does not provide the ostensible stimulus (or a clue) to receivers, it needs more effort to fin d the meaning. P eople with low NFC (that is, individuals who have low cognitive effort tendency) might fail to decode and understand the meaning/ intention of the ad. Without the ability to convey the meaning of the ad to receivers or consumers, the ad is not successful. Individual Differences in Cognitive Motivation R esearch on p ersuasion has demonstrated the importance of understanding such cognitive processes (e.g., Petty & Cacioppo 1986) The level of cognitive elaboration is a critical variable in pro ducing attitude change s for consumers (Batra & Stayman, 1990), which leads to the ad effects. The c ognitive tendency of receivers has received more consideration in metaphor research than any other type of antecedents
35 because receivers need such cognitive efforts to interpret and comprehend the message of the ads, which then leads to the ad effects According to the study of Pawlowski, Badzinski, and Mitchell (1998), when children were exposed to advertisements containing metaphors and were asked to interpr et them, the higher grade children were better at interpreting metaphors than children in lower grades It would be fair to say that the ability of interpretation on metaphors is related to the development of cognitive ability. Need for cognition is defin & Cacioppo 1986, p. 48). Researchers have long known that there are differences among individuals in their tendency to t hink and how they enjoy thinking. Studies of cognition have focused on the nature of knowledge and the character of the processes being understood and being used with the knowledge. Cohen Stoland, and Wolfe (195 7 ) need to structure relevant situations in meaningful, integrated ways own study (1957) reported that people with a high need for cognition are more likely to arrange, make detailed analyses and assess or criticize the information, as opposed to people with low NFC. individuals who have high NFC are easily motivated to think about communication, whereas individuals who have low NFC need to be motivated to think about communication by clearly establishing the problem. Cacioppo and Petty (1982) also asserted that individuals who have high NFC should
36 prefer the complex (that is, the c ognitively demanding) to the simple version, whereas individuals with low NFC should prefer the simple to the complex version of the cognitive task. In this respect, previous research has explored the processing of two different advertising stimuli (e.g., visual vs. verbal) and further investigated the role of affect and cognition in processing visual and verbal stimuli ( Sojka & Giese 2006) P eople with a high NFC who have strong cognitive tendencies to engage in effortful information processing are likely to prefer rational data and rely on logical thoughts when solving problems or making decisions (Cacioppo & Petty, 1982; Sojka & Giese, 2006). In the study of Sojka and Giese ( 2006 ), p eople with a high NFC prefer verbal stimuli more than affectiv e people, who have a more affective tendency and enjoy the problem solving process presented in verbal stimuli, unlike people with low NFC. Petty, Cacioppo, and Schumann ( 1983 ) suggested that i ndividuals with high NFC have been observed to perform better a t cognitive jobs such as calculation and solving anagrams than do individuals with low NFC. As opposed to those with high NFC, individuals with low NFC were influenced more by peripheral cues ( Petty, Cacioppo, & Schumann, 1983). Individuals with high NFC w ere more likely to be persuaded by the quality of arguments (Petty, Cacioppo, & Schumann, 1983). P eripheral cues which are well known such as color or celebrity endorsements may be more likely to be conveyed in a visual format than a verbal format (Petty & Cacioppo, 1986) The findings of Sojka and Giese 2006 ) showed that individuals with high NFC elicit more pleasure from processing words than images, and, as a result, they will pay more attention to the verbal elements in an ad. Considering th is particular study as
37 the basis for my research I could further explore the interaction between the types of metaphors as visual stimuli, and headlines as verbal stimuli in individual differences with NFC and affective processing Prior research has in vestigated whether adding a headline to ads explains the visual image results of a reduced favorable impression T he headline may actually take away from for receivers to interpret the ad by themselves (Phillips, 2000). However, the current study asserts that effects depending on the type of pictorial metaphors in ads an NFC. In the case of implicit visual metaphors, adding a headline to ads will lead to stronger impacts for the receivers which will cause them to develop more favorable attitudes toward the ad through understanding of the metaphori cal meaning. Under the ad conditions in which implicit metaphorical images without any headline are employed, the ad will have a more positive impact on the attitudes towards the ad, brand, and purchase intention of consumers with high NFC than those with low NFC. Based on explicit metaphors, under the condition where there is an explicit pictorial metaphor with out any headline in the ad, consumers with high NFC will show more positive responses than those with low NFC in their attitudes towards the ad, br and, and purchase inten t ion As a result, for consumers with high NFC, when the ad does not have any headline, there would likely be no significant difference in impact between the ad with implicit and explicit pictorial metaphors on ad effects: attitude toward the ad, attitude toward the brand, and purchase intention. For consumers with low NFC, when an ad
38 has a headline that is directly illustrated (i.e., a complete headline), there would likely be no statistically significant differen ce in impact betwee n the ad with implicit and explicit pictorial metaphors on ad effects: attitude toward the ad, attitude toward the brand, and purchase intention. For consumers with high NFC, u nder ad conditions where pictorial images have headlines, the ad employed with implicit metaphors is more likely to reflect a favorable impact on the attitudes towards the ad, the brand, and purchase inten t ion. In those conditions, consumers need more cognitive effort to elaborate the implicit metaphorical ad. According to a study b y Kim, Baek, and Choi (2012), the level of metaphor elicited cognitive elaboration had a significant effect on attitude s toward the advertiser. For the consumers with low NFC, if there was no headline copy in the ad, and the ad employed implicit metaphors the consumers w ould not be motivated to draw conclusions or meaning Based on the theoretical framework and the literature review, the foregoing discussion leads to the formulation of the following hypotheses for the current study: H3 1 : For consumers with high NFC, when a headline directly illustrated the ad ( i.e., a complete headline), those with implicit pictorial metaphors create more positive responses than the ads with explicit pictorial metaphors and a complete headline have on ad effects including: (a) attitude toward the ad, (b) attitude toward the brand, and (c) purchase intention. H3 2 : For consumers with low NFC, when a headline directly illustrated the ad ( i.e., a complete headline), the ad with explicit pictorial metaphors en able s a more favorable impact than the ad with implicit pictorial metaphors on ad effects including : (a) attitude toward the ad, (b) attitude toward the brand, and (c) purchase intention. H3 3 : For consumers with high NFC, when there is no headline, the a d with implicit pictorial metaphors enable s a more favorable impact than the ad with explicit pictorial metaphors on the following ad effects: (a) attitude toward the ad, (b) attitude toward the brand, and (c) purchase intention.
39 H3 4: For consumers with low NFC, when the ad does not have any headline, the ad with explicit pictorial metaphors enable s a more favorable impact than the ad with implicit pictorial metaphors on the following ad effects: (a) attitude toward the ad, (b) attitude toward the brand, and (c) purchase intention. Cogn itive and Affective Responses on Metaphors M etaphors inspire human emotions and promote central thought (Phillips, 2003). That is, pictorial metaphors require cognitive effort be understood and then, once understood, to evoke positive or negative emotional responses. People who have either more or less cognitive tendency will evaluat e the visual or verbal components of a metaphorical ad differently T he resulting effects from metaphorical ads may be that each group with different tendencies will express different attitudes ( Sojka & Giese 2006 ). T he use of pictorial metaphor s in ads is imperative and effective in persuading consumers to appreciate what t he ads want to convey (Jeong, 2008; McQuarrie & Phillips, 2005; Sopory & Dillard, 2002). Metaphors in communication generate their own story (Fairhurst & Sarr, 1996; Lakoff & Johnson, 2003), which should be interpreted by audiences, create uncertainty, eli cit pleasure from novelty, lead to credibility, and enhance engagement with the ad messages through the metaphorical content (Jeong, 2008; Sopory & Dillard, 2002 ; Ward & Gaidis 1990; Zaltman, 2003). Pictorial metaphors are considered most pleasurable if th ey are perceived as appropriate and insightful (Ward & Gaidis 1990). The reason for the impact of metaphors might be found in cognitive and affective processes and responses. Researchers found that using metaphors results in a powerful persuasive tool for developing
40 decision making and memory (Jeong, 2008; McQuarrie & Mick, 1999; Zaltman, 2003) M etaphors are also an effective way to help bring people unexpected thoughts and feelings that result in novelty experiences (Jeong, 2008; Zaltman, 2003). As a result of these processes the use of metaphors may result in pleasurable feelings or even in feelings of joy (Sopory & Dillard, 2002 ; Zaltman, 2003). Therefore, the use of metaphors in ads is effecti ve in creating more positive attitudes and behaviors and helping people to comprehend meaning more deeply by fostering engagement with the ad message. Furthermore, metaphors have a positive influence on memory (Jeong, 2008; McQuarrie & Mick, 1999; Zaltman, 2003). However, even though metaphors have positive attributes, they might not always persuade consumers. The effects of metaphor s might be varied by means of different contexts, product categories, product knowledge, brand familiarity (Sopory & Dillard, 2002), or cultural competency (McQuarrie & Mick, 1999). As c onsumers attempt to understand the ad message presented in metaphors they will use cognitive processing created by their own specific experiences and these interpretations may differ depending o n various conditions in the ads Compar ed to literal argument, metaphors in ads are more effective in persuasion (Mitchell & Olson, 1981). The effectiveness of metaphorical rhetoric, compar ed to literal arguments, was illustrated by three categories cog nitive, affective, and motivational processes (Jeong, 2003) The persuasive effects of metaphors are influenced by affective processes and motivational processes (Sopory & Dillard, 2002). Using metaphors may result in pleasure as a relief from cognitive te nsions, which may result in positive responses towards the ads and brands (Jeong, 2008; Sopory & Dillard, 2002 ).
41 The important role of affective or motivational processes in persuasion can be explained via the elaboration likelihood model (ELM) and informa tion processing (Petty & Cacioppo, 1986). Although images are considered heuristic cues and cause peripheral or heuristic modes of processing by the elaboration likelihood model (ELM), pictorial metaphors are central to the message argument and may cause systematic modes and cognitive processing (Jeong 2008; Petty, Unnava, & Strathman 1991). Metaphors, one of the essential tools for message based persuasion, play a central role in our communications (Lakoff & Johnson, 2003). Affective Perspectives on Met aphors In the process of being persuaded by metaphors, Sopory and Dillard (2002) described two strands of the pleasure and relief view. These two strands demonstrate that we can recognize a semantic anomaly in a metaphorical expression and this can lead to negative tension (Bowers & Osborn, 1966; Sopory & Dillard, 2002). When the meaning of metaphor is finally understood, the negative tension is relieved. The process has three stages: perception of error (or novelty), the use of conflict (or recoil), and solving) (Sopory & Dillard, 2002). The first strand is illustrated when people resolve the meaning of metaphors and thus leasurable experiences. In the second strand, when the recipients find the metaphorical meaning, negative tension is dissipated and this turns out to bring relief. The feeling of pleasure or relief encourages recipients in reinforcement of the metaphorical meaning and the evaluation (Sopory & Dillard, 2002). The feeling of pleasure or relief compensates for the negative tension. That is, the dissipation of negative tension used in the process of decoding metaphorical expression can be used to reinforce mean ing. This evaluation
42 not only encourages engagement, but also increases persuasion and recall (Sopory & Dillard, 2002). Affective Perspectives on Persuasion According to some previous research, feelings ( or emotion s ) occur immediately, and then, cognition arises (Murphy & Zajonc, 1993 ; Zajonc, 1980). Zajonc (1980) cognitive process. Klauer (1998) claimed that spontaneously affective evaluation precedes cognitive analysis o f the stimulus. Zajonc (1980) asserted that affective reactions are inescapable and primary processes in advertising and affective judgments are structured around the advertising context. Affective responses from incoming information pre categorize the sti mulus for the recipient before a decision is made about criteria features and affective reactions that influence discrimination among stimuli (Zajonc, 1980). In contrast, several contemporary scholars in psychology consider the process of responses as the move from cognition to affect ( i.e., emotion) (Lazarus, 1982) Before automatic and unconscious evaluation of any received stimulus is labeled as liked or disliked, people have to have some information, knowledge, or experience about it and they have to a t least have identified the features or the discriminated value. Therefore, some theorists believe that affect is a result of cognitive processing of the incoming information (Lazarus, 1982). They believe that affective responses occur as a result of prior evaluation for the information such as an event or stimulus that was encoded, discriminated and identified for its critical feature or value. In the case of an ad that uses pictorial metaphors without headlines, we would consider the framework one in whic h
43 cognitive processing occur s prior to affective reactions in order to for information about the ad to be de coded. Some controversy still remains about what is perceived as the primary process. It seems it depends on each event, situation, or the kind of stimuli. Researchers assert that affect and cognitive processing cannot be considered separately (Lerner & Keltner, 2000). In some cases, initial emotional responses produce ideas or perception, and then the ideas or the perception produce further af fect ( Lerner & Keltner, 2000). Therefore, i t is difficult to determine the dominance and primacy of these two processes (i.e., cognitive and affective processing) even though they are clearly two different forms of conscious or unconscious processes. The resear ch of Morris, Woo, Geason, and Kim (2002) shows that cognition and affect are interdependent. They found that affect dominates over cognition for predicting conative attitude and action, and the emotional reactions are strongly predictive of behavioral int ention (Morris et al., 2002). In addition, Morris et al. (2002) assert that affect accounts for almost twice the variance towards conative attitude, and emotional response is a powerful predictor of behavioral intention and brand attitude. Researchers hav e long known that consumers in positive moods assess incoming information more positively (Batra & Stayman, 1990 ; Isen & Simmonds, 1978). Some studies have even determined that such positive moods incline consumers to accept perceived messages (Mackie & Worth, 1989; Worth & Mackie, 1987). Many researchers who study persuasive habits have illustrated that positive emotion may be provoked by ads and a positive mood facilitate s a brand (Batra & Stayman, 1990; Ba tra & Ray, 1986; Edell & Burke, 1987). According to
44 Morris, Woo, and Singh (2005), content processing (i.e., content elaboration) gives rise to emotions and this leads to a longer lasting change in attitudes. E motion s ( or mood) in a person may work on the peripheral or heuristic processes (Forgas & Bower, 1987) because they are expected to require relatively less effort to evoke the process and because they produce a reaction in a straightforward fashion (Petty & Cacioppo, 1986). In addition to such periph eral processes, research suggests that positive emotions also influence message acceptance (Mackie & Worth, 1989). That is, individuals in positive moods perceive and evaluate stimuli more favorably than those in negative moods (Isen & Simmonds, 1978). Bat emotions (i.e., positive moods evoked by ads) can influence the amount of total cognitive elaboration under various conditions of argument quality and need for cognition. The research of Batra and Stayman (1990) has shown that positive moods indirectly influence attitudes through two cognitive processes. One is that a bias against the generation of negative evaluation or thoughts leads to a more favorable evaluation of message arguments. Second, cognitive process invo lves a reduction in total cognitive elaboration, making the processing more heuristic rather than systematic (Batra & Stayman, 1990). Research has demonstrated that attitudes can be modified by interaction between initially neutral objects and affect produ cing forms of stimuli such as unpleasant odors, high or low temperatures, pleasant pictures, elating versus objects, issues, or conditions is a primitive form of att itude change, but one which has proven effective.
45 Affective Priming Effect There are a number of priming studies that have been adapted from cognitive psychology to the study of evaluative processes. These priming studies in previous literature have empl oyed polarized stimuli with strong positive and negative evaluation s Because the role of the evaluative response is very important in the cognitive fiel d, many lines of research have explored the characteristics, antecedents, and environments of the evalu ative process. Emotions generally include positive or negative evaluations (Ortony & Turner, 1990), so theories of emotion have considered the evaluation as a central concept (Hermans, 1996; Klauer, 1998). In the same vein, cognitive appraisal models of em otion (Lazarus, 1991; Scherer, 1988; Klauer, 1998) postulate that the incoming stimuli become spontaneously positive or negative very quickly during the initial state of forming emotion. Affective priming effect is the one of the theories of af fective influences on judgement and choice (Klauer, 1998; Klauer & Musch, 2001; Murphy & Zajonc, 1993; Musch & Klauer, 2003; Steinb ei s & Koelsch, 2010). There are typically two stimuli in affective priming: prime and target stimulus. The prime stimulus tha t is an affectively valenced one is followed by an affectively valenced target stimulus In other words, an affectively valenced (i.e., pleasant or unpleasant) prime stimulus is presented, follow ed by an affectively valenced target stimulus. The affective priming paradigm can be explained as a function of the evaluative response. Affective priming illustrates whether the evaluation of the prime stimulus (i.e., a first stimulus) has an impact on the processing of subsequent stimuli (i.e., the targets) (Klaue r, 1998). The findings of the studies on affective priming effect shows that prior affective valences such as like or dislike, pleasure or dis pleasure from the preceding prime stimulus have influence on
46 affective target processing, either by assisting consistent target processing or suspending inconsistent target processing (Musch & Klauer, 2003; Steinbies & Koelsch, 2010). In the classical conditioning of evaluations and attitudes (so called attitude conditioning, Staats & Staats, 1957, 1958; evaluati ve conditioning, Martin & Levey, 1978), affective priming explains the acquisition of evaluations and attitudes. If a positive or negative stimulus (e.g., a positive or negative word) is evaluated spontaneously, then a neutral stimulus (e.g., a nonsense sy llable) should be able to be influenced by the evaluative response. More recently, researchers have shown that the affective priming effect results in a positive response when the stimulus is shown under conditions that inhibit recognition of their appeara nce (Murphy & Zajonc, 1993). Thus, research outlined by some authors revealed that awareness of the affect arousing stimulus is not necessary for evaluative conditioning (Klauer, 1998; Martin & Levey, 1978; Murphy & Zajonc, 1993). As mentioned, previous re search has already shown that consumers who process an ad in a more positive mood develop a more positive attitude toward a brand (Batra & Stayman, 1990). We can consider the affective priming from the ELM. People are often exposed to affect inducing posi tive or negative material (e.g., pictures) before they are directly faced with the target stimulus. This affective priming process has been successfully established as modifying attitudes. For example, several research studies have shown that if a particip ant is asked to rate a target person who is performing an ordinary activity of daily life after they have been subliminally presented with positive photos (e.g., a group of smiling friends), they rate the target person more positively than if they
47 are expo sed to negative photos (Krosnick, Betz, Jussim, & Lynn, 1992). According to Murphy and Zajonc (1993), the effectiveness of affective priming processing may be different in different cases where primes are presented both subliminally or visibly. In other wo rds, if positive and negative affective primes, such as smiling or frowning faces, w ere shown just before a particular target stimulus, attitude toward the target were influenced by the primes presented outside of conscious awareness. Yet, this does not ha ppen when these primes are presented visibly (Murphy, Monahan, & Zajonc 1995). Researchers believe that one of the possible reasons for such an effect might be that directly flagrant priming stimuli are perceived as definitely irrelevant to understanding t he target. Thus, a visibly presented priming might prevent the participant from developing responses based on the emotional primes (Martin et al., 1990; Petty & Wegener, 1993). I n this regard, metaphorical ads are indirect expressions of the key and abstra ct benefits of the advertised product. The research on affect association contends that such affective processes allow for an impact on attitudes toward objects, but this occurs only when the objects have low significance. When consumers have little or no knowledge about the objects, there are low opportunities for appreciating the importance affect (Zajonc, 1998). However, this does not create the conclusion that it is imperative that elaborating attitude relevant information remains low in order to influe nce attitudes in affective processes. Affect can also modify attitudes even when the likelihood of elaboration is high. Affective priming is a concept for the emotions that suggests the contagion of affects, which is an important character istic of contiguity theory that comes up in conditioning and association learning. According to Zillmann (1991), this additive
48 emotional reaction (i.e., excitation) is not recognized by the subject when it occurs. The excitation transfer paradigm (Bryant & Miro n, 2003) also illustrates the affective phenomenon that the feeling of arousal provoke s when previous stimulus is transferred to excitation that occurred by a consecutive stimulus. (Gilbert, Fiske, & Lindzey, 1998). As previously mentioned, when the metap horical meaning employed in ads is understood, the expected emotion that may occur is pleasurable or joyful. Among the emotional responses from metaphorical ads, the current study focuses on pleasure dimensions in PAD theory ( Mehrabian & Russell 1974) Re ceivers can sense those entertainments at the moment they decode or interpret the meaning that the metaphorical ad tries to convey. Therefore, the excitation transfer theory (Bryant & Miron, 2003) can also demonstrate the positive emotional transfer to the target or to the next stimuli (Bryant & Miron, 2003; Zillmann, 2006). Mediated E ffect by E motional R esponses Just as other studies focused on emotional responses to advertising, Holbrook and Batra (1987) investigated the role of emotions to attitude toward ad and brand. They asserted that emotional responses of consumers mediate the effects of advertising in their model of advertising effects by path analysis (Holbrook & Batra, 1987). The results i ndicated that emotions represented a mediating relationship between ad content and attitude toward the ad or the brand (Holbrook & Batra, 1987). Based on their study, the metaphorical ad would also be one kind of advertising content and there might be inte rvening roles of emotion in mediating the relationship between the metaphor content of the ad and the attitudes. In other words, the emotions would mediate the effects of ad content on attitudes.
49 Metaphors not only arouse cognitive processes, but they a lso provoke affective responses (Sopory & Dillard, 2002). The process of decoding metaphor can be a rewarding experience for consumers. Pleasure is the most commonly anticipated emotional response from rhetorical figures (DeRosia, 2008). On the other hand, finding metaphorical meaning in literal falseness and grasping for similarity between two disparate concepts in a metaphorical statement or image can elicit negative tension. This is caused by incongruity in literally false statements or artificial images Pleasure can occur when this tension is relieved (McQuarrie & Mick, 1992; Sopory & Dillard, 2002). When an ad has an implicit pictorial metaphor and an illustrated headline that is directly related ( i.e., a complete headline ), the completeness of the he adline facilitates After understanding the metaphorical meaning in the ad, a positive emotional response such as pleasure occurs. Therefore, this is more likely to have a positive impact on the attitude towards the a d and the brand, and purchase inten t ion. Furthermore, when the ad has an explicit pictorial metaphor and a directly illustrated headline ( i.e., a complete headline ), the completeness of the headline may positive emotional responses suc h as pleasure which can come from interpreting the ad messages by themselves. Therefore, a complete headline that exactly explains the metaphorical meaning in the ad also increases understanding but may reduce positive emotional resp onses (Phillips, 2000). The complete headline is more likely to prevent the pleasure and relief that comes from interpreting and decoding and to decrease a positive impact on the attitude towards the ad, the brand, and on purchase intent ion.
50 Previous researchers argued that t he emotional effect is stronger than the motivational effect (Batra & Stayman, 1990; Mackie & Worth, 1989). When an implicit pictorial metaphor and a directly illustrated headline ( i.e., a complete headline ) are employed in the ad, r egardless of individual differences in NFC, if the ad with the pictorial metaphor induces a positive mood ( or emotion) to the receivers positive emotional responses from the metaphorical ad will lead to more positive ad effects In addition, if the pictor ial metaphor s in ads induce less positive or even negative emotional responses they will lead to less positive or negative ad effects. For example, when the ad employs an implicit pictorial metaphor and no headline, even though the consumers have low NFC if the consumers just feel a positive emotion (or mood) from it, these positive emotional responses are more likely to evoke a more positive attitude toward the brand and purchase inten t ion. If positive emotions are not generated, the relationship betwee n the ad with pictorial metaphors and positive ad effectiveness disappears. The affective priming effect theory and the role of emotions as mediator leads to the following hypothesis: H4: E motional responses from metaphorical ads will mediate the effects of the optimal combination of pictorial metaphors and headline types on advertising consequences in consumers (i.e ., attitude toward the brand and purchase inten t ion) According to Isen and Levin (1972), when individuals are involved in detailed cognitive processing, this processing requires some effort and disrupts positive moods. Thus, mood protection tendencies that individuals have should lead to reduced effort in the cognitive processing of ads (Isen & Levin, 1972). T he researchers asserted that indiv iduals in positive moods process messages with less cognitive elaboration and positive moods decrease cognitive elaboration (Batra & Stayman, 1990). Therefore,
51 emotional responses from metaphorical ads, especially positive affect such as pleasure, make for positive priming. Regardless of individual differences in cognitive tendency, if the ad with the pictorial metaphors elicits positive emotional responses, the responses will lead to a more positive impact on ad effectiveness. The current study shows th e effect of visual metaphors with headlines on intention.
52 CHAPTER 3 METHODOLOGY This chapter describes the procedure is used to examine the interrelations brand attitude, and purchase intention while considering individual differences on cognitive and affective responses. Two separate experiments were conducted to test all hypotheses. Experiment 1 was used to examine and verify the effect of the ad employing pictorial met aphors all across product types. E xperiment 2 was conducted to investigate the int eraction effects among visual, verbal, and cognitive differences in individuals including the role of emotional effects. Experiment 1 Study Design An experiment was implemented with the univariate analysis of variance (ANOVA) to examine the effect of picto rial metaphors on the attitude toward the ad brand, and purchase inten tion (i.e., Hypothesis 1). E xperiment 1 was conducted on the dependent variables to determine whether their mean scores were significantly different between the participants. Each set o f twelve advertisements in one type of ad images (i.e., visual modality: pictorial metaphor vs. non metaphor) was rated by nearly 21 participants. Recruitment and Participants A total of fifty five (N = 55) participants for the E xperiment 1 were recruited online via Amazon Mechanical Turk (M Turk) and were randomly assigned to one of the experimental conditions. Amazon Mechanical Turk has become a popular vehicle for online experimental research in social science (Mason & Suri, 2012; Paolacci,
53 Chandler, & Ipeirotis, 2010). Research has shown that experimental participants using Amazon Mechanical Turk demonstrate no difference in exhibiting classic heuristics and biases, and in paying attention to experimental directions, when compared to participants from other traditional data collection methods such as those chosen through use of a lab subject pool (Paolacci et al., 2010). In M Turk, two criteria were used to collect data: (1) a HIT approval rate for the M Turk workers equal to or greater than 95%, and (2) the location was restricted to the United States. Among fifty five ( n = 55) parti cipants for E xperiment 1, the responses from 8 participants who failed to complete the questionnaire were excluded and 6 participants were ex cluded in the data analysis because they were identified as having already experienced an experiment with very similar ads. Thus, a total of 41 responses were use d for the data analyses in E xperiment 1. Specifically, the number of participants in the ads w ith pictorial metaphors and the ads with non metaphors was 21 and 20, respectively. In terms of demographic information, among 41 participants, the average age was 39.1 years ( SD = 15.25), and 24 were male (58.5 %) while 17 were female (41.5 %). In terms of race, 33 participants (80.5 %) were Anglo American/Caucasian, 3 participants (7.3 %) were m ultiracial, 2 participants (4.9 %) were African American, 2 participants (4.9 %) were Hispanic/Latino, and 1 participant (2.4 %) was Asian American. In terms of e ducation level, a majority of the highest level of education were College graduate (26.8 %) and those with Some college (26.8 %) respectively, followed by High school graduate (22.0%), Graduate degree (19.5%), Some high school (2.4%),
54 and Other (2.4%). All descriptions of participants for E xperiment 1 were summarized in Table 3 1. Table 3 1. Description of Participants for Experiment 1. Variables Description Frequency Percent Gender Male 24 58.5 % Female 17 41.5 % Race Anglo American/Caucasian 33 80.5 % Multiracial 3 7.3 % African American 2 4.9 % Hispanic/Latino 2 4.9 % Asian American 1 2.4 % L evel of education College graduate 11 26.8 % Some college 11 26.8 % High school graduate 9 22.0% Graduate degree 8 19.5% Some high school 1 2.4% Other 1 2.4% Average Age M = 39.12 ( SD = 15.25) Development of Stimuli It is generally expected that advertising for high involvement product categories is processed differently than advertising for low involvement categories (Rossiter & Percy, 1997). Across typical product types, studies usually manipulate the advertising st imuli by product involvement or product categories to determine how people respond
55 to th e se ad treatments, and then relate these inter individual differences in attitude towards the ad to cross sectional variations. According to the previous study by Ang ( 2002), the use of metaphoric headlines for a utilitarian product enhances attitude and behavioral intention more than the use of a non metaphoric headline does. However, for a symbolic product, metaphoric headlines lead to less favorable attitudes and beha vioral intentions than non metaphoric headlines do (Ang, 2002). In another study, metaphors in headlines had an influence on ad and brand personality perceptions by product type (Ang & Lim, 2006). Therefore, product type could be an influential factor for ad effectiveness in metaphorical ads. However, previous studies on metaphors did not consider various product types such as high/low involvement including the wide price range and did not cover all product types in the studies (e.g., Chang & Yen, 2013; Bergkvist et al., 2012). Because the product types chosen in metaphorical ads might be a limitation for the results, the current study employed all product categories into the stimuli and focused on the effect of using pictorial metaphors and headlines in affective responses. In the current study including E xperiment 2, choosing the advertised products for stimuli was considered based on the advertising planning model; Foote, Cone, and Belding ( FCB ) model ( Ratchford 1987; Vaughn, 1980) which includes a grid in which purchase decisions can be classified To distribute the advertised products evenly in all categories for the pre test, the products in the ad chosen in the current study were selected from all categorie s of product types that made up of the FCB grid (Vaughn, 1980). The basic dimensions of the FCB grid were developed through testing by
56 Vaughn (1980) and Ratchford (1987) in 6 separate studies These studies resulted in the development of the following quadrants on the FCB grid: High involvement/thinking, high involvement/feeling, low involvement/thinking, and low involvement/feeling are quadrants in the grid (See Figure 3 1) The use of the adve rtised products across a variety of product categories was expected to enhance the external validity and generalization of the study. The FCB grid was put into the product categories as a model. There are four quarters in the FCB grid model: (1) informativ e, (2) affective, (3) habit formative, and (4) self satisfaction. Thinking Feeling HIGH 1. INFORMATIVE 2. AFFECTIVE 3. HABIT FORMATION 4. SELF SATISFACTION LOW INVOLVEMENT Figure 3 1. The FCB grid. ( Vaughn 1980) According to Ratchford (1987) and Vaughn (1980), products falling into the high involvement and thinking quadrant Quadrant 1, include such items as car s house s furnishing, televisions exter ior house paint, headache remedies bike helmet s insurance, computer s home appliances an d so on (Ratchford, 1987). P roducts with characteristics that allow for the informative product type which may cause consumers to think deeper imply a larger need for information. They appeal to the rational side offering cognitively oriented benefits. In contrast, the affective products of Quadrant 2 for high involvement and feeling products appeal to the emotional aspect such as self
57 esteem induced by the desire for sensory experience, fantasy, and fun. Products designated as high involvement and feeling products include eye glasses, jewelry, cosmetics, fashion, apparel, underwear, and motorcycles (Ratchford, 1987). The products falling in to the low involvement and thinking quadrant include paper towels, liquid bleach, insect repellant, regul ar shampoo, insecticide, food, small household items, salad oil, suntan lotion, and diapers (Ratchford, 1987). The products for low involvement and feeling include liquor, candy, movies, peanut b u tter, deodorant, toothpaste, fast food restaurant s fruit, b arbecue sauce, diet soft drinks, regular soft drinks, gum, chocolate, beer, liquid hand soap and cleaner (Ratchford, 1987). Considering the FCB grid for the stimuli (Vaughn, 1980 ) this could be a way to integrate cognitive and affective responses for produ cts. As a result, all advertised products for ad stimuli in this study w ere chosen by following the literature reviews (Ratchford, 1987; Vaughn, 1980). In each quadrant, three advertised products were interspersed. Even though the more ad stimuli the stud y ha s the better the external validity will be the number of ad stimuli should be limited to ensure good internal validity as well T his study uses three ads for the product types in each quarter. D eveloping the se type s of visual modality is important bec ause three ad s in each quarter of the product types are expected to be balanced to avoid bias by one side in a quarter of the product types. Therefore, each quarter of the product types has three different ads for a total of twelve advertisements that fall into each type of visual modality (visual metaphor vs. n on metaphor) across all product types
58 The ads and advertised products used in E xperiment 1 were displayed in each quadrant following the FCB grid and the products (Ratchford, 1987; Vaughn, 1980) as shown in Table 3 2 and 3 3. T he t o tal number of each type of ads wa s the same across all types of product categories. Therefore, the ad stimuli in the current study are covered for all product types. Table 3 2 Product T ypes for ( Pictorial M etaphor s) A ds in Experiment 1 total 12 ad s Involvement Think Feel 1 Informative 2 Affective High involvement Car, Blender mixer, Grass trimmer Cos metic, Outdoor shoes, Underwear 3 Habit formation 4 Self satisfaction Low involvement Laundry detergent, Lint roller, Baby diapers Candy, H ot sauce, Beverage Table 3 3 Product T ypes for ( Non M etaphor ) A ds in E xperiment 1 total 12 ad s Involvement Think Feel 1 Informative 2 Affective High involvement T.V., Car, Vacuum cleaner Cosmetic (Lip Balm), Jean s, Shoes 3 Habit formation 4 Self satisfaction Low involvement Laundry dete rgent, Toothpaste, Baby diapers Butter, Beverage, Chocolate The advertisement stimuli used in this study were selected from existing ad images in the world market. The ads with famous brands were modified to fictitious Another reason is that the brand name fulfills the same funct ion as a headline because well known brands may play a role that evoke s the benefits associated with it (Bergkvist, Eiderback, & Palombo, 2012). In addition, some ads are not available in the U.S. and the se ads were unknown to the participants when the exp eriment was conducted. All tests in the experiments were conducted with fictitious or unknown
59 brands to eliminate pre existing brand attitudes. Given the use of fictitious brands, when participants were asked about the ads, they were informed of the product item in the ad. As a result, a total of 24 advertisements were developed for E xperiment 1. T welve advertisements fell into pictorial metaphor ads and twelve advertisements were in non metaphor ads as shown in Table 3 2 and 3 3. By conducting multiple pretests with various metaphorical ads, degree of visual metaphors, difficulty of the ad, the pe rceived congruence between the copy and the visual metaphor image in the ad, and the advertising overall message were assessed for choosing the ad stimuli Procedure Participants were randomly assigned to one of the two experimental conditions (pictorial metaphors vs. non metaphors) with twelve advertisements. O n the M Turk invitation page, participants are guided to a link for the study. P rospective participants were asked to read the informed consent and they were told that the purpose of the study wa s t o examine their responses to print ads. The instrument was self guided, and had no time limit. Participants were informed that they would be given twelve advertisements and be asked about each ad. In the first stage of each advertisement, participants were asked to rate for ad attitude, brand attitude, and purchase intention, followed by questions about the degree of metaphor in the ad image for the manipulation check. Then, participants answered demographic questions (e.g., gender, race, age, and the highe st education level). Pretest P rocedure To develop rigorous advertising stimuli for visual metaphor ad s and to confirm the advertised products interspersed with each quarter in the FCB grid model (Vaughn,
60 1980), the pretest was followed from previous literature ( Ratchford, 1987; Vaughn, 1980 ). M ultiple pretests were conducted in E xperiment 1. Participants for the pretests were recruited from Amazon Mechanical Turk The purpose of the pretests was to identify some dimensions for only metaphorical ads such as degree of visual metaphors, difficulty of the ad, the perceived congruence between the copy and the v isual metaphor image in the ad. In addition, the overall advertising message was tested to select the appropriate metaphorical ads for the studies and to ensure that the instructions for participants were clear. The pretests also verified whether the level of ad difficulty wa s too high among ad stimuli The pretest for the difficulty of the ad was tested with the ads without a headline copy because i t needs initially to measure only for vi sual metaphors excluding any assistance Initially 38 metaphorical ads were selected from the existing ads in the world covering all product types. Those include both implicit and explicit metaphor ads evenly distr ibuted for the ads with visual metaphors In each pretest, approximately 20 participants were presented and rated in each dimension with those metaphorical ads After the pretest, each visual metaphor ad or non metaphor ad of 12 ads were selected from the 3 8 ads and those are interspersed throughout the main study in E xperiment 1 In the pretest, product involvement for the ad stimuli was not questioned or measured and the product types of the ad stimuli were interspersed with the FCB model ( Ratchford, 198 7; Vaughn, 1980) because the perceived product involvement would be different for each individual even for the same product in the ad. In addition, advertisement stimuli for non metaphor ads were not tested in the pretests. However, the pretest in non metaphor ads yielded to the manipulation check.
61 All 24 advertisements for non metaphor stimuli and visual metaphor stimuli based on the pretest results were selected across all product types and the ad stimuli with visual metaphors were selected where the degree of metaphors was higher than the mean value ( M visual metaphor = 4.10) The mean of the difficulty of the ad was 3.88 for the metaphorical ads and the ads were chosen around the mean value The mean of the perceived congruence between the copy and the visual in the metaphorical ads was 4.68. The mean of the advertising overall message was 4.55 and the ad stimuli with visual metaphors were selected by higher value than the mean. As a resu lt of the pretest for ad stimuli selection, a total of 24 different advertisements integrating pictorial metaphors (e.g., 12 non metaphor s and 12 visual metaphor s ) with headlines were developed across all product involvement categories. T he ad stimuli used in this study were provided in Appendix B. Measurement Instruments Most of the measures employed in the study were adapted from relevant research and have been commonly used in the applicable literature. In the process of research design implementation, the measurement instrument was also designed. The measurement instrument for E xperiment 1 included the items for the pretests and those measuring dependent variables. In the pretest, participants received 3 8 ads and they were asked to verify the degree of metaphor and to check the deviation in difficulty of the ad, the perceived congruence between the copy and the visual metaphor image in the ad, and the quality of the overall advertising message P articipa nts rated the metaph oric level of ad images on six 7 point scales, which were anchored by bipolar adjectives such as of
62 9; McQuarrie & Phillips, 2005; Phillips & McQuarrie, 2009). Participants measured the perceived congruence between the copy and the visual metaphor image in the ad on three 7 point scales, which were anchored by bipolar adjectives: Not compatible/compatibl e, a good fit/ not a good fit, and congruent/ not congruent ( Rifon, Choi, Trimble, & Li, 2004 ). Participants were also asked about the degree of difficulty of the ads on three 7 po int scales that best describe their impression toward the advertisement the y have seen: How much do you think it is difficult to get what the message is from this ad ? This ad must make you stop and think ; i t took some time to infer from a metaphor what has been communicated ( Bergkvist, Eiderback, & Palombo 2004 ; Phillips, 2000 ). They measured the overall advertising message on three 7 point scales, which were anchored by bipolar adjectives: not powerful/powerful, not convincing/convincing, and not strong/strong ( Hallahan, 1999; Wang, 2007 ). In addition, they were asked if they ha d ever seen the advertisement or heard the brand before in order to adjust ad familiarity and to confirm the fictitious brands For the dependent variables, participants were asked to measure their attitudes towards each ad on a three item, seven point se mantic differential scale (the extent to which they liked the advertisement). The bipolar adjective items were anchored by & The reli ability estimate for this measure was The participants were asked to rate what they thought of the advertised brand on a three item, seven point semantic differential scale
63 toward the b rand in the advertisement were assessed on the items, which were 2008). This was therefore a measure of how effective participants thought the brand wa s. A composite rating was created by calculating the mean of the items. The reliability item, seven point bipolar The three items were averaged to derive a Results Manipulation C heck The manipulation check questions measured on a 7 point scale were used to assess the degree of metaphor of the ad stimuli. The manipulation check question captured whether the ad with pictorial metaphors would be judged as artful/clever rather than plain/matter of fact (e.g., McQuarrie & Mick, 1996; 19 99 ). P articipants rated the metaphoric level of ad images on six 7 point scales, which were anchored by bipolar of ick, 1996; 1999; McQuarrie & Phillips, 2005; Phillips & McQuarrie, 2009). The reliability estimate for this measure was A one way ANOVA was performed to verify whether the manipulation of pictorial metaphor in ads differ ed significantly. The results suggest that participants consider ed the pictorial metaphor conditions significantly more metaphorical ( M metaphor = 5.12) than the non metaphorical (i.e., the control) ( M non metaphor = 4.25, F (1, 39) = 12.54, p < .001).
64 Tests of H ypothes i s 1 In support of H ypothesis 1, a separate Study 1 was conducted as part of this research. A series of one way ANOVA was conducted to examine mean differences in attitude toward the ad, attitude toward the brand, and purchase intention (DVs) f or visual types (IVs), namely ad image using visual metaphors or non metaphor in ads. The results of an ANOVA on attitude toward the ad, attitude toward the brand, and purchase intention indicated that the main effect of visual metaphors in ads on DVs was significant. With respect to attitude toward the ad, the results suggest a significant main effect ( F (1, 39) = 4.51, p < .05) for ad condition ( M metaphor = 5.12 vs. M non metaphor = 4.56). With respect to attitude toward the brand, the results show there is no significant differen ce ( F (1, 39) = 3.39, p = .073) for each ad condition ( M metaphor = 5.03 vs. M non metaphor significant main effect ( F (1, 39) = 5.49, p = .001) for the ad condition ( M metaphor = 4.53 vs. M non metaphor = 3.87). Therefore, these results provide partial support for H ypothesis 1. Simple main effects of means and standard deviations for all dependent variables are summarized in Table 3 4. Table 3 4 Sim ple Main Effects of Means and Standard Deviations for Dependent Variables in Experiment 1 Pictorial Metaphor vs. Non Metaphor Dependent Variables Mean (Standard Deviation) Ad Attitude 5.12 (.76) vs. 4.56 (.94) Brand Attitude 5.03 (.86) vs. 4.48 (1.06) Purchase Intention 4.53 (.64) vs. 3.87 (1.10) *** Note: *** p < .001; ** p < .01; p < .05.
65 Discussion attitude, brand attitude, and purchase intention for the selected twelve advertisements in each condition across all product categories ( Ratchford, 1987; Vaughn, 1980). As in the previous studies (Gkiouzepas & Hogg, 2011; Mitchell & Olson, 1981 ), the results revealed that consumers who were exposed to pictorial metaphor ads exhibited more favorable ad attitude and motivated a greater purchase intention than those who were exposed to non metaphor ads. The findings from Hypothesis 1 in E xperiment 1 provide distinct evidence that the ads which contain embedded visual metapho rs change ( Ratchford, 1987; Vaughn, 1980). In addition, the finding s ultimately verify that the ads employing visual metaphors provoke greater cognitive elaboration and this leads to greater effectiveness on ad attitude and purchase intention (ELM; Petty & Cacioppo, 1986 ) across different product categories than do the ads with non metaphor in the ad image. The results of Experiment 1 re established the effect of the ad emplo ying pictor ial metaphors in the ad image for persuasive advertising effects across all product types in the ad Experiment 2 Study Design An experimental design was used to investigate the effect of the interrelation between pictorial metaphor and headli ne copy NFC on the attitude toward the ad brand, and purchase intention, while considering emotion s as a mediator. The current study examine d whether the types of pictorial
66 metaphors with or without verbal support in ads (i.e., headline copy) ha ve an impact upon emotional responses understanding cognitive influences such as need for cognition and whether these ads are thereby able to affect behavioral intention An experiment al study between subjects randomized factorial design 2 ( Pictorial metaphor type: implicit vs. explicit) x 2 ( Verbal support : directly illustrated copy vs. non copy) x 2 (Need for cognition: high vs. low) was used to test the proposed H yp othes es 2, 3, and 4 (Table 3 5). E xperiment 2 included three independent variables: 1) type of pictorial metaphor s ; 2) headline copy ; and 3) level of n eed for cognition. Dependent variables include attitude toward s the ad, attitude toward s the brand, and purchase intention. E motiona l responses (i.e., pleasure) from participants were used as a mediator in the study. In Experiment 2, participants were randomly assigned to view twelve treatment stimuli for various product types (Table 3 7, 3 8) through several pretests T welve advertise ment s in each type of pictorial metaphor condition (i.e., implicit or explicit) were used as ad stimuli Table 3 5 2 x 2 x 2 F actorial D esign High NFC Low NFC Implicit pictorial metaphor Explicit pictorial metaphor Implicit pictorial metaphor Explicit pictorial metaphor H eadline copy I II V VI N on copy III IV VII VIII
67 The following sections show further information about the implementation of the research design Sections includ e information about recruitment of participants stimulus development, measurement instrument design steps taken to pretest the research design, and the procedure for the main study. Recruitment and Participants A total of 310 participants for the study were recruited through Amazon Mechanical Turk (M Turk) Consistent with Experiment 1, two criteria used to collect data in M Turk were : (1) a HIT approval rate for the M Turk workers greater than or equal to 95%, and (2) the location was restricted to the United States. Among the respondents, responses from 4 p articipants who failed to complete the ques tionnaire were excluded and 54 were excluded through use of attention questionnaires. Thus, a total of 252 responses were used for the data analyses in Experiment 2 and were randomly assigned to one of eight exper imental conditions. Specifically, the number of participants in the implicit and explicit metaphors was 129 and 123, respectively, and the number of subjects in the ad with copy and non copy conditions was 123 and 129. Among the 252 subjects, in terms of demographic information, 91 were male (36.1 %), while 161 were female (63.9 %). The average age of participants was 39.7 years ( SD the 18 to 25 years old range, 34 percent we re 26 to 34 years old, 27 percent were 35 to 44, 17 percent were 45 to 54, 11 percent were 55 to 64 and 3.6 percent were in the over 65 and under 80 years old category. Approximately 77.8 percent of participants were Anglo American/Caucasian, 7.9 percent A frican American, 6.7 percent Asian American, 2.8 percent Native American/American Indian, 2.4 percent Hispanic/Latino, and 2.4 percent Multiracial. A majority of the highest level of education (44 %) was College
68 graduate, followed by some college (28.7%), Graduate degree (13.5%), High school graduate (12.7%), Other (0.8%), and Some high school (0.4%). All demographic information for the participants i s summarized in Table 3 6. Table 3 6 Description of Participants for Experiment 2. Variables Frequency Percent Gender Male 91 36.1 % Female 161 63.9 % Age (years old) Average 39.7 ( SD = 12.59) 18 to 25 24 9.5 % 26 to 34 82 34 % 35 to 44 68 27 % 45 to 54 42 17 % 55 to 64 27 11 % 65 to 80 9 3.6 % Race Anglo American/Caucasian 196 77.8 % African American 20 7.9 % Asian American 17 6.7 % Native American/American Indian 7 2.8 % Hispanic/Latino 6 2.4 % Multiracial 6 2.4 %
69 Table 3 6. Continued Variables Frequency Percent Highest level of education College graduate 111 44 % Some College 72 28.7 % Graduate degree 34 13.5 % High school graduate 32 12.7 % Other 2 0.8 % Some high school 1 0.4 % Development for Stimuli Consistent with Experiment 1, the current study employed all product categories into the stimuli and focus on the effect of using pictorial metaphors and headlines in choosing the advertised p roducts for stimuli was considered based on the FCB grid ( Ratchford, 1987; Vaughn, 1980) as well. To distribute the advertised products evenly in all categories, the products in the ad chosen in the current study were selected from all categories of product types that made up the FCB grid ( Ratchford, 1987; Vaughn, 1980) similar to E xperiment 1 in order to enhance the validity and generalization of the study. Considering the FCB grid ( Ratchford, 1987; Vaughn, 1980) for the stimuli could be a way to int egrate cognitive and affective responses for products. The product types of the ads to use in Experiment 2 were displayed in each quadrant following the FCB grid and the products (Ratchford, 1987; Vaughn, 1980). The ads across the product types used in Exp eriment 2 are shown in Table 3 7 and 3 8.
70 The total number of ads in each product category is the same. Therefore, the ad stimuli in Experiment 2 were covered for all product types. Table 3 7 Product T ypes of ( Implicit M etaphor s) A ds in Experiment 2 to tal 12 ad s Involvement Think Feel 1 Informative 2 Affective High involvement Car, Blender mixer, Electric Iron. Eye wear (glasses), two Underwear ads. 3 Habit formation 4 Self satisfaction Low involvement Two Hand sanitizers ads, Salad dressing. Delivery Pizza, H ot sauce, Mouthwash. Table 3 8 Product T ypes of ( Explicit M etaphor s) A ds in Experiment 2 total 12 ad s Involvement Think Feel 1 Informative 2 Affective High involvement Ski Board helmet, Grass trimmer, Lawn mower. St oc king s Outdoor Shoes, Cosmetic. 3 Habit formation 4 Self satisfaction Low involvement Lint roller, Baby Diaper s Toothbrushes. Candy, Beverage, Gum. As previous ly outlined in Experiment 1, the advertisement stimuli were selected from existing ad images and the ads with real brands were modified to fictitious brands advertised products are not available in the U.S. and the se ads were unknown to the participants when the experiment was conducted. Therefore, Experiment 2 was conducted with fictitious or unknown brands to eliminate pre existing brand attitudes. Given the use of fictitious or unknown brands in the study, participants were informed of the advertised product item in order to identify what kind of product was advertised. Consequently, a total of 24 advertisements w as developed for Experiment 2 by two
71 types of pictorial metaphor (i.e., explicit/implicit) and whether or not a headline c opy was employed. Two types of headlines were provided for each ad in the study. One type of headline represented complete verbal anchoring (i.e., a headline that explicitly had no headline. Furthermore, beca use there was no statistically significant difference between using a moderate headline and a complete headline in the previous research (Bergkvist, Eiderback & Palombo, 2012; Phillips, 2000), the current study applied only complete headlines to the stimul i when using a headline in the metaphorical ads. The headlines in the study consisted of a literal description of what the ad represented. The headline cop y in each ad were modified under each condition. In the ad stimuli with headlines or no headlines, if t here was an ad with no headline versus headlines, the ad was exactly the same but headlines were placed in it. Headlines in each ad were used with the existing headlines when possible. But if there was no headline in the original ad, the appropriate head line copy in the ad stimuli was created or modified by an expert, an advertiser and a doctoral student with professional experience in advertising agencies. In addition, the created headline copies were assessed in the pretest in order to ensure the congru enc y with the metaphorical ad image and their apparent authenticity. All ad stimuli were presented to an advertising e xpert to review the ads to see whether they look ed realistic and pro f essional. Once it wa s agreed that the ads were plausible and represe ntative of the type that may be expected to be seen in magazines the ads were used to enable confidence in the generaliz ation of results.
72 Procedure Pretest p rocedure To develop a set of rigorous advertising stimuli, the stimulus development procedure f ollowed a multiple step process by conduct ing several pretest s. Participants for the pretests were also recruited from Amazon Mechanical Turk Respondents in the pretest did not participate in the main study. If participants were identified as those who were exposed to the same experiments or ad stimuli, the data of the participants were excluded in the data analysis. Several pretests for the me taphorical ads had been already conducted in E xperiment 1 and the pretest on the copy completeness was additionally assessed in E xperiment 2. The pretests enabl ed the researchers to manipulate ad stimuli and decreased the potential for confounding to occur The purpose of the pretests was to identify whether some dimensions such as degree of metaphors are in metaphorical ads in order to select the appropriate metaphorical ads for the main study and to ensure that the instructions for participants are clea r. The pretests also verified headline copy completeness and whether the level of ad difficulty is similar or not among ad stimuli. A d stimuli with headline copy were selected via use of the pretest on headline copy completeness. The mean of the copy compl eteness was 4.66 in 7 point scales and the ad stimuli were chosen with a higher value than the mean value. In each pretest, approximately 20 participants for each dimension were presented with 38 advertisements from the existing ads and answered the quest ions. Each participant reviewed all 38 ads with headline copy in two types of pictorial metaphor (i.e., implicit and explicit metaphor) ads. Of these, 24 kinds of metaphorical ads (12 implicit, 12 explicit) were selected covering all product types These ad stimuli
73 were developed using 12 implicit metaphor s and 12 explicit metaphors across all product categories based on the literatures of the FCB model The participants were asked to measure one dimension such as the degree of completeness for a headline copy in a related ad. Three dimensions (e.g., the degree of visual metaphors, difficulty of the ad, and the degree of the completeness of headline copy) for confounding checks were based on possible biases and the h ypotheses in the current study (Table 3 9 ). After the pretest, ads were selected from the 38 original ads for further review These ads were broken down into 12 ads for implicit metap hors and 12 for explicit metaphors Each of the series of 12 ads was interspersed for the main study. Table 3 9 Pretests for E ach D imension Pretest 1 Pretest 2 Pretest 3 Group 1(n=20) Average Group2(n=20) Average Group3(n=20) Average DIMENSIONS Implicit Metaphors Degree of Metaphor Headline Copy Completeness Difficulty of the ad Ad1 Ad2 Ad3 . Ad 12 Grand Average Value IG1 Value IG2 Value IG3 Explicit Metaphors Degree of metaphor Headline Completeness Difficulty Ad13 Ad14 Ad15 . Ad 24 Grand Average Value EG1 Value EG2 Value EG3 The pretests allowed us to select the ads by determining similar levels of the dimension such as difficulty of the ad and headline copy completeness for each type of
74 pictorial metaphor. All mean values of each dimension in implicit and explicit metaphor we re compared. The final 12 ads for each implicit/explicit pictorial metaphor ad were selected. Based on the results of the pretest, 12 ads in each type of pictorial representation (i.e., implicit and explicit metaphor) were determined as appropriate stimul i for the main study. After the pretest, a total of 24 ads confirmed from these results were used to prevent triggering difficulty of ad understanding, degree of pictorial metaphors, and As a result of the pretest for ad stimuli selection, a total of 48 differ ent advertisements integrating pictorial metaphors (e.g., 12 implicit metaphor and 12 explicit metaphor) and two types of headlines (i.e., 12 ads with directly illustrated headline copy and 12 ads with no headline copy) were developed across all product in volvement categories. After the pretests, the advertising stimuli used in this study were provided in Appendix B. Pictorial metaphors reflecting both types of pictorial (implicit/explicit) metaphors and headlines with the appropriate degree of completeness were interspersed in the experimental ads for Experiment 2. Main test p rocedure The main experiment of the current study was conducted online via Amazon Mechanical Turk When participants sign ed up to perform the experiment, the informed consent form was reviewed including information about their rights as study participants F or example, the purpose of the research was outlined, as well as the confidentiality of their responses to all information requests. Prospective participants were told that the purp ose of this study was to examine their responses to print ads using metaphorical images, which were woven from headline copy as well as their emotional responses.
75 After consenting to participate, the Web survey site (Qualtrics.com) led participants to the questionnaires. Participants were randomly assigned to one of the experimental advertisement conditions which clearly did implicit or explicit metaphor images and directly illustrated headline copy or no headline copy for a fictitious brand The instrumen t was self guided, and there was no time limit. Print ads were chosen because they afforded greater experimental control. For each condition, participants were exposed to 12 color ed magazine ad s containing a type of pictorial metaphor and a type of headline copy through a Web survey site (Qualtrics.com) on the screen of their individual computers. F irst, as an attention question participants were asked whether there was an advertised product image in the image of the ad in order to di stinguish types of visual metaphor An additional description of a small product image at the bottom of the ad was excluded. This was a way to ensure at the beginning of the questionnaire and it would also give the partic ipants incentive to think about what the ad intend to convey And then, participants were asked about their feeling, or emotional responses to each of the 12 ads right after seeing each ad ( Gilbert, Fiske, & Lindzey, 1998). In addition, participants were questioned about dependent variables (i.e., a d attitude, brand attitude, and purchase intention). Finally they were asked about the difficulty of the ad for the manipulation check. After the presentation of all 12 ads with questions i r Cognition was measured investigated by the level of cognitive motivational systems. P articipants were asked to answer questions intended to measure their level of Need for Cognition T heir level of cognitive elaboration, ency of and intrinsic
76 enjoyment Cacioppo 1986, p. 48 ) as a moderating factor of the effect of visual metaphors in the ad NFC was measured at the very end of the questionnaire in order to exclude the cognitive bias or influences that may come from the measure of NFC questions. Next participants were asked to answer questions about their demographics (e.g., gender, age, race, education level). Finally, participant s were informed that the ir classification information was used only to confirm that they participat ed in the research, and their identity was kept confidential to the extent provided by law. Participants were thanked for their participation, debriefed, and were directed to a page with the code number for their compensation Measurement Instruments The measurement instrument for E xperiment 2 included items measuring independent variables for the pretests, the manipulation checks, dependent variables, a covar iate, a moderating variable an d mediating variables (Table 3 10 ). Table 3 10 Measurement I nstrument Variable type Variable #of items Study Independent variable (Manipulation related variable) Degree of pictorial metaphor 6 McQuarrie & Mick, 1996; 1999 Difficulty of the ad 3 Phillips, 2000; Bergkvist, Eiderback, & Palombo, 2012 Headline completeness 3 Phillips, 2000
77 Table 3 10. Continued Variable type Variable #of items Study Dependent variable Ad attitude 3 MacKenzie & Lutz 1989 Brand attitude 3 Jeong, 2008 Purchase intention 3 Lee & Aaker, 2004 Moderator Need for cognition 5 Cacioppo, Petty, & Kao, 1984 Mediator Pleasure value of emotion 4 Gilbert, Fiske, & Lindzey, 1998 Pretest measures for m anipulations In the pretest, participants received 24 ads for consideration and they were then asked to verify the degree of metaphor and to check manipulation on type of metaphors (i.e., implicit or explicit metaphors) (Ang, 2002; Chang & Yen, 201 3). To make balanced stimuli and deviation in degree of metaphors, p articipants rated the metaphoric level of ad images on six 9 point scales, which were anchored by bipolar adjectives such as of McQuarrie & Mick, 1996; 1999; McQuarrie & Phillips, 2005; Phillips & McQuarrie, 2009). The stimuli of pictorial metaphor images were selected by measuring the degree of difficulty for the pictorial metaphor in the ads in order to have a similar degree of interpretation. Considering previous studies (Phillips, 2000; Bergkvist, Eiderback, & Palombo, 2012) that examined visual metaphors in ads, proper metaphors were identified by asking participants in the pretest to rate them The difficulty of the ad with
78 impression toward the ad they have seen. The difficulty of the ad was measured with a 9 point Likert type scale using three items. Participants were asked to rate the following three How much do you think this ad is difficult to understand what the message is ? (How difficult do you think the message in this ad is to understand?) some time to infer from a not at all 7 = v ery much) of pictorial metaphor ads were measured on the degree that would be fairly comprehended for the main study (Phillips, 2000; Bergkvist, Eiderback & Palombo, 2012). Headline copies in the ads were measured in the pretest for manipulation on the l evel of headline completeness by the amounts of verbal anchoring added to a metaphorical image. Participants were given the following questions to rate th e degree of headline completeness on a nine strongly Disagree 9 = strongly agree The pretest enabled us to select a simi lar degree of completeness about the headlines of the stimuli and facilitates the ability to manipulate the combination of visual metaphors and verbal copy. The higher the number, the more complete the headline in the ad was perceived (verbal anchoring hea dline). The current study selected the headlines that clearly spelled out the main message conveyed by the pictorial metaphor for the stimuli. Dependent variables Consistent with Experiment 1, p s toward s each ad were measured on a three item, nine point semantic differential scale (the extent to which
79 they liked the advertisement) in Experiment 2 The bipolar items were anchored by & Lutz 1989). s towards the advertised brand were assessed with a three item, nine point semantic differential scale in the Experiment 2 study and the items were 2 008). The three items were averaged to derive a composite rating. the phrase, advertisement, it may encourage me to purchase this product three item, nine point, semantic differen tial scale anchored by A composite rating was created by calculating the mean of the items. Moderator Participants were asked to fill out the questionnaire containing the measure for the Need for Cognition scale (Cacioppo, Petty, & Kao, 1984) in E xperiment 2 Based on the study by Cacioppo and Petty (1982), participants were asked to measure the degree of Need for Cognition scale. An 18 item Need for Cognitio n scale (Cacioppo, Petty, & Kao, 1984) was administered to participants in order to examine whether participants have high Need for Cognition or low Need for Cognition containing measures for the 18 item NFC scale. However, as previous studies have suggest ed (Epstein et al. 2013; 1996), the revised 5 item NFC scale was used in the current study. The Need for Cognition was tested and analyzed by using PROCESS macro in SPSS via mean centering regression with the NFC scores ( Belsley, Kuh, & Welsch 1980 ) inste ad of using a median split
80 Potential moderated m ediator The p leasure value of emotion represents a very intuitive response that consumers may feel on the spur of the moment after exposure to pictorial metaphor ad s Therefore, the measurement scales were examined right after the ad exposure in E xperiment 2. To rate the pleasure value of emotion in each ad, the measure scale of Gilbert, Fiske, and Lindzey ( 1998) was used with adjective words such as pleasure, enjoyment, and entertainment. T he participants use the bipolar pleasure value of emotion adjective pair on a nine point scale based on the following: Analytic Stra tegy E xperiment 2 tested the two way interactions between type of pictorial metaphor (implicit vs. explicit) and headline copy (headline copy or non copy) in predicting interaction effect, while controlling for difficulty of the ad. More importantly, E xperiment 2 tested the three way interaction among type s of pictorial m etaphor (implicit vs. explicit), headline copy (headline copy or non cognitive differences) in predicting a moderating effect, controlling for difficulty of the ad as a covariate. The current study employed the PROCESS macro via SPSS (Hayes, 2013) for moderated moderation analysis and moderated mediation analysis. To provide a more powerful statistical test of mediation effects, the data were also analyzed the standard errors of the indirect conditional effects on metaphorical ad effectiveness.
81 Results Pretest To create balanced stimuli for the difficulty of the ad and deviation in degree of pictorial metaphors, the current study conducted several pretests. For well adjusted manipulations, one way ANOVA was performed to assess the manipulations of the variables for E xper iment 2. The results of P retest 1 for the degree of metaphors in the ads demonstrated that there were no significant differences between implicit metaphor ( M = 4.42) and explicit metaphor for the degree of metaphor in the ad ( M = 4.52, F (1, 39) = .14, p = 0.712). I n addition, the results of P retest 2 showed that there was a significant manipulation effect for the difficulty of the ad with implicit metaphor ( M = 4.11) and in the ad with explicit metaphor ( M = 3.59, F (1, 37) = 1.97, p = .17). In the resu lts of Pretest 3, no significant differences between implicit metaphor ads ( M = 4.63) and explicit metaphor ads ( M = 4.96) were found f or headline copy completeness ( F (1, 41) =.96, p = .33). Thus, all manipulations for the variables were successfully con firmed. Reliability c hecks Reliability checks for variable measures with a multi item scale were conducted. Regarding the independent variables; Need for Cognition (NFC) was measured on five item, nine point semantic differential scales modified by Epstei n et al. (2013; 1996) and these consisted of two reversed questions. The reliability estimate for this measure was ards to the dependent variables, attitude toward the ad, attitude toward the brand, and purcha se intention were examined by a three item, nine point bipolar adjective scale developed by MacKenize and Lutz (1989).
82 very acceptable. evaluation of difficulty of the ad was measured as a covariate with a three item, nine point scale (Phillips, 2000) and the re mediator variable, the reliability estimate for the pleasure value of emotion scale Manipulation c hecks One way ANOVAs were performed to verify the degree of difficulty for metaphorical ads. The manipulation checks of the difficulty of the ads showed that there was a significant difference in difficulty. The mean score of the difficulty of the ads in the condition of implicit metaphor ( M implicit = 4.36) was significantly different from that of the explicit metaphor ad ( M explicit = 2.96, F = 47.60, p < 0.001). This is a different condition from the pretest. In the pretest, one participant rated all ads included as both implici t and explicit metaphor ads. However, in the manipulation check for the main test, one participant reviewed and assessed all the ads under the one condition with only implicit or explicit metaphor ads. In addition, 252 participants verified the type of vis ual metaphors for each stimulus of whom 252 (100 %) gave the correct response for the type of visual metaphors in ads In the attention question, participants were asked to verify the type of visual metaphor (i.e., implicit vs. explicit) and those who ans wered incorrectly were eliminated from the data analysis.
83 Testing for confounding e ffects Even though the pretest for the difficulty of the ad was conducted, there seemed a possibility to have an influence on cognitive responses in metaphorical ads to change the attitude and behavioral intention of consumers Therefore, the questionnaire included a control variable to ensure that the results would only be attributed by the effect of visual metaphors and headline copy ; a two way MANOVA (multivariate anal ysis of variance) was tested to explore whether there was an interaction effect between the type of pictorial metaphors and headline copy on ad attitude, brand attitude, and purchase intention. The results illustrated that the pictorial metaphor and headli ne copy did not significantly affect the combined DVs ( = .993, F (3, 246) = .574, p = .633, partial 2 = .007) and multivariate effect sizes are very small. Considering confounding factors for Hypothesis 2, a two way ANOVA (analysis of variances) was performed to examine the possible interaction effect of the type of visual metaphors and headline copy on the difficulty of the ad as a dependent variable. This tested whether headline copy had an influence on or whether such headline copy makes a dif ference on the difficulty of the ads. The results showed that the type of pictorial metaphor and headline copy have a significant impact on the difficulty of the ad ( F (1, 238) = 39.72, p < 0.01). The result of a simple effect test indicated that headline copy in an implicit metaphor ad significantly decreased the difficulty of the ad (Implicit M copy = 3.82 vs. Implicit M non copy = 4.96, F (1, 248) = 17.24, p < 0.001). Outcomes showed that headline copy reduced the level of difficulty of the metaphorical ad. However, in an explicit metaphor ad, there was no significant change demonstrated for the difficulty of the ad (Explicit M copy = 2.94 vs. Explicit M non copy = 2.98, F (1, 248) = .025, p = .87). Results indicated that the headline
84 copy more effectively reduced the difficulty of the ad in implicit pictorial metaphor condition than in the explicit metaphor condition. Thus, the difficulty of the ad was a confounding factor on D Vs. Tests of h ypotheses With respect to H ypothesis 2, a two way multivariate analysis of covariate variance (MANCOVA) was conducted to examine the effect of independent variables (i.e., pictorial metaphors and headline copy ) among experimental groups on c ombined dependent variable s such as Ad Attitude, Brand Attitude, and Purchase Intention while controlling for difficulty of the ads (i.e., covariate). The results of the MANCOVA indicated that the two way interaction effect between visual metaphors and he adline copies were not significant ( = .998, F (3, 245) = .150, p = .930, partial 2 = .002) and the main effects of headline copy = .991, F (3, 245) = .761, p = .517, partial 2 = .009) were not significant. However, significant main effects of visual metaphors (type of metaphors) were found ( = .887, F (3, 245) = 10.441, p < .001, partial 2 = .113). Thus, the absence of an interaction effect between visual metaphors and headline copies means that H2 was not confirmed The cov ariate (i.e., = .817, F (3, 245) = 18.239, p < .001, partial 2 = .183 (Table 3 10). Univ ariate ANOVA results (Table 3 12 ) indicate that only visual metaphors (type of metaphors) significantly exert a strong influence on all dependent variables including ad attitude ( F (1, 247) = 30.284, p < .001, partial 2 = .109), brand attitude ( F (1, 247) = 30.109, p < .001, partial 2 = .109), and purchase intention ( F (1, 247) = 13.932, p < .001, partial 2 = .053).
85 Table 3 11 Multivariate Tests of Two Way Interaction df Error df F 2 p Difficulty of Ads a .817 3 245 18.24*** .183 Head line Copy .991 3 245 .76 .009 Type of Metaphors .887 3 245 10.44*** .113 Copy Type of Metaphors .998 3 245 .15 .002 Note: a. Covariate; p < .05, ** p < .01, *** p < .001
86 Table 3 1 2 Univariate Results for the Two way Interaction Effect B etween Type of Metaphors and Headline Copy on Dependent Variables DVs F df p 2 p Difficulty of Ads AdAtt 50.702 1 .000 .170 BrdAtt 35.823 1 .000 .127 PI 28.833 1 .000 .105 Type of Metaphors AdAtt 30.284 1 .000 .109 BrdAtt 30.109 1 .000 .109 PI 13.932 1 .000 .053 Headline Copies AdAtt 1.852 1 .175 .007 BrdAtt 2.027 1 .156 .008 PI 1.948 1 .164 .008 Type of Metaphors x AdAtt .001 1 .984 .000 Headline Copies BrdAtt .026 1 .873 .000 PI .165 1 .685 .001 Note. DVs = Dependent Variable; AdAtt = Ad Attitude; BrdAtt = Brand Attitude; PI = Purchase Intention T able 3 13 presents the adjusted and unadjusted group means for ad attitude, brand attitude, and purchase intention. Comparison of the adjusted means of ad attitude, brand attitude, and purchase intention indicates that explicit metaphors produce a more favorable at titude and behavioral intention than implicit metaphors.
87 Table 3 13 Adjusted and Unadjusted Group Means for Ad Attitude, Brand Attitude, and Purchase Intention Dependent Variables Ad Attitude Brand Attitude Purchase Intention Adjusted M Unadjusted M Adjusted M Unadjusted M Adjusted M Unadjusted M Headline Copy Copy 5.59 5.633 5.75 5.77 4.72 4.76 Non Copy 5.80 5.742 5.96 5.92 4.99 4.93 Types of Metaphors Implicit 5.25 5.01 5.41 5.21 4.47 4.24 Explicit 6.15 6.40 6.3 6.52 5.24 5.49 M oderated m oderation analysis To assess the proposed H and purchase intention, controlling for the difficulty of the ads as a covariate. H ypothesis 3 states f Need for Cognition ( NFC) significantly influences ad effectiveness (i.e., Attitude toward ad, Attitude toward b rand, and Purchase intention) as a moderation hypothesis. To test the three way interaction; the types of pictorial metaphors x headline copy x specifically Model 3 (Shrout & Bolger, 2002; Preacher et al., 2007; Hayes, 2013), with 5,000 bootstrap samples and a 95% confidence level (CI) (Hayes, 2013) were used to conduct a moderated moderatio n analysis. analysis, intended to determine whether the effects of an independent variable interact
88 with the effects of two other variables (moderators). The decision to enter Copy and NFC as moderators and Ty pe of visual metaphors as the independent variable was theoretical and the three way interaction would be significant regardless of which variables were entered as moderators and which were entered as the IV (the independent variable) In short, the PROCESS macro uses ordinary least squares regression to estimate the coefficients for each predictor and their interactions. PROCESS model 3 is useful in determining the significance of the interactions between and among an independent varia ble and two moderators (see Figure 3 2) are and are centered around the means, t hat is, the mean and 1 SD from the mean for the IV and each moderator. Confidence intervals provided a basis for estimating whether, at each NFC level within each Copy level, type of visual metaphors (i.e., implicit and explicit) were significantly relat ed to composite scores. Effects were consi dered significant when confidence intervals did not contain zero (Hayes, 2012). In the results resented below, LLCI and ULCI refer to lower level and upper level confidence interval, respectively. To test H ypothes more positive responses when the ad has a headl ine copy that illustrated it the type of visual metaphors (i.e., implicit and explicit metaphors) was entered as an independent variable, head line copy was entered as one moderator, and NFC was entered as a second moderator (Figure 3 2). The difficulty of the ads was entered as a covariate.
89 Figure 3 2 PROCESS Model 3 (Hayes, 2012) Moderated Moderation Analysis The PROCESS (M odel 3) results indicate a marginally significant three way interaction among type s of visual metaphors, headline copies, and NFC (Need for Cognition) on only Purchase Intention among three DVs, B = .37 (SE = .20), t = 1.82, p = .07 (LLIC/ULCI = .0300/.7593), thus suppo rting H3 partially for the level of NFC on Purchase Intention (see Table 3 13). There were no significant effects or interaction ( p > .1) on Attitude toward the ad and Attitude toward the brand but marginally significant interaction effect s of type s of vi sual metaphors on Purchase Intention were found. The three way interaction on Purchase Intention indicated that the effects related to Type of visual metaphors differ ed by levels of Headline Copy and NFC. These findings, depicted in Figure 3 3 and 3 4 sup port Hypothesis 3 that Headline Copy and NFC have an influence on purchase intention in terms of the effects of visual metaphors
90 As one of the dependent variable s was Purchase Intention, regression analysis revealed that type of metaphor, headline copy, NFC, their interactions and the covariate, and difficulty of the ad explained 25% of the variance in composite scores, F (8, 243) = 10.04, p < .001. The main effec t of pictorial metaphor types was significant, B = .74 (SE = .21), t = 3.54, p < .001 (LLCI/ULCI = .3265/1.1466), but the effects of headline copy and NFC were not signi ficant (ps > .1). The covariate, difficulty of the ad had a significant effect on Purc hase Intention, B = .33 (SE = .06), t = 5.43, p < .001 (LLCI/ULIC = .4538/ .2120). All two way interactions were not significant (ps > .1), but the three way interaction between typ e of metaphor, copy, and NFC was marginally significant, B = .37 (SE = 20), t = 1.82, p = .07 (LLIC/ULCI = .0300/.7593) on purchase intention. As a marginal significant effect of three way interaction was found, PROCESS Model 3 revealed conditional effects of type of metaphor on Purchase Intention based on the pick a poi nt approach (Hayes, 2012). Under the low NFC, there was a significant difference between implicit and explicit metaphor when the ad involved headline copy, Effect = 1.17 ( SE = .38), t = 3.12, p < .01 (LLCI/ULCI = .4315/1.9166). More specifically, under the condition of the ad using headline copy, when people who had low N FC were exposed to explicit metaphor, they had higher level s of purchase intention ( M = 5.26) than when those who had low NFC exposed to implicit metaphor ( M = 4.08). In contrast, under the high NFC condition, there was a significant difference between implicit and explicit metaphor when no copy existed in the ad, Effect = .98 ( SE =.38), t = 2.61, p < .01 (LLCI/ULCI .2406/1.7202). More specific ally, under the condition of no copy in the ad when people who had high NFC were exposed to explicit metaphor ( M = 5.38), they
91 had a higher level of purchase intention than when those who had high NFC exposed to implicit metaphor ( M = 4.40). Figure 3 3 Three way Interaction Effect on Purchase Intention Figure 3 4 Three way Interaction Effect on Purchase Intention 3.8 4.3 4.8 5.3 5.8 Low NFC High NFC Copy Implicit Explicit 3.8 4.3 4.8 5.3 5.8 Low NFC High NFC Non Copy Implicit Explicit
92 Table 3 14 PROCESS Model 3 Results on Purchase Intention Variable B se t LLCI/ULCI Copy .3126 .1921 1.6272 .0658/.6909 Type of visual metaphor .7366 .2082 3.5384 .3265/1.1466 Type of metaphors x Copy .1785 .3819 .4675 .5737/.9307 NFC .0258 .0508 .5071 .1258/.0743 Type of metaphors x NFC .0947 .1001 .9458 .2918/.1025 Copy x NFC .0647 .1005 .6432 .2627/.1334 Type of metaphors x Copy x NFC .03647 .2003 1.8202 a .0300/.7593 Note: p < .05, a. marginally significant p =.07 Table 3 15 Moderated Moderation Analysis Results: Effects of Type of Visual Metaphors on Purchase Intention Responding by Copy/Non copy and NFC level. Effect se t LLCI/ULCI Copy NFC Low 1.1741 .3769 3.1149** .0021/.4316 High .1162 .4017 .2893 .6750/.9075 Non Copy NFC Low .6670 .3976 1.6774 .1162/1.4503 High .9804 .3756 2.6104** .2406/1.7202 Note: p < .05, ** p < .01, *** p < .001 Moderated mediation analysis To assess H ypothesis 4, mediation analysis was conducted using a temporal frame as the independent variable (temporal frame: 0 = implicit metaphor, 1 = explicit metaphor) and the PROCESS macro (Model 11) for SPSS for bias corrected
93 bootstrapping (Hayes 2012; Preacher and Hayes 2004). Bootstrapping was used to generate a 95% confidence interval (CI) around the indirect effect of mediator (i.e., pleasure value of emotion ). Successful mediation occurs when the CI does not contain zero (Preacher, Rucker, and Hayes 2007). For the purpose of the moderated mediation tests, the PROCESS macro (Hayes, 2012) was also employed because it verifies moderator whereby X is the independent variable, Y the dep endent variable, Mj are the mediator variables and W and Z are two moderator PROCESS technique that proposes that moderated mediation effects be ve rified was employed. N amely, there are relations between the indirect effe cts of independent variable X before the dependent variable Y, via the mediator variable M and moderator variables W and Z. To examine Hypothesis 4, Ad Attitude, Brand Attitude, and Purchase Intention were respectively defined as the dependent variable s ( DV), Type of visual metaphors as Headline Copy as moderator variable (W) and Need for Cognition was also set as a moderator variable (Z) ( see Figure 3 (2012) model 11 enables the specific examination of these moderation effects of Headline Copy (W) and of Need for Cognition (Z) via metaphors (X) and each dependent variable ad attitude, brand attitude, and purchase intention (Y). Thus, the intent is to verify the previously specified hypothesis, H 4.
94 Figure 3 5 PROCESS Model 11 (Hayes, 2012) Moderated Mediation Analysis Moderated m ediating effect of p leasure The current study predicted that pleasure in affective responses would mediate the effect of metaphor type, headline copy, and individual differences in NFC on Ad Attitude, Brand Attitude, and Purchase Intention. Since consumers who have high NFC may have less pleasure when they are presented with a pictorial metaphor ad with copy and consumers who have low NFC may also have less pleasure when they are exposed to pictorial metaphor ad without headline copy on the ad, the current study proposes moderated med iating effect of affective response on dependent variables. Thus, PROCESS model 11 was employed to test the moderated mediation. First, the current study employed three way interaction between type of metaphor, copy, and NFC on pleasure as the outcome vari able. The difficulty of the ad was added as a covariate. The results revealed that there were significant main effects from the ty pe of
95 metaphor, B = 1.90 ( SE = .72), t = 2.66, p < .01 (LLCI/ULCI = .4941/3.3127) and copy, B = 1.64 ( SE = .74), t = 2.22, p < .05 (LLCI/ULCI = .1824/3.0881). The covariate, difficulty of the ad also had a significant effect, B = .31 (SE =.05), t = 6.41, p < .001 (LLCI/ULCI = .4060/ .2152). Moreover, there was a significant two way interaction between copy and NFC, B = .23 ( SE = .11), t = 2.01, p < .05 (LLCI/ULCI = .4539/ .0042), but other two way interactions did not exist in any significant manner ( p s > .1). Three way interaction was marginally significant, B = .29 ( SE =.16), t = 1.80, p = .07 (LLCI/ULCI = .0267/.5961). Th erefore bootstrapping techniques were employed to test conditional indirect effects on each dependent variable (Hayes 2012; Preacher et al., 2007 ). Ad a ttitude First of all, the test of moderated mediating effect of pleasure was used by They created a pleasure value of emotion scale using adjective words such as pleasure, enjoyment, entertainment, and fun. Bootstrapping techniques of conditional indirect effects showed the mediating effect of pleasur e as a value of emotion depending on conditions. The conditional indirect effect of type of metaphor on Ad Attitude was significant when consumers who had low NFC were exposed to a headline copy (effect = 1.03) and no headline copy (effect = .57) because t he confidence interval for indirect effect excluded zero (CI for copy = .4866 to 1.5689, CI for non copy = .0413 to 1.1548). In addition, the conditional indirect effect of type of metaphor on Ad Attitude was significant when consumers who had high NFC wer e exposed to the ad with no copy on the pictorial metaphor ad (effect = .92) because the confidence interval for indirect effect also excluded zero (CI = .3464 to 1.4884). However, when consumers who had high NFC were exposed to the ad with a headline
96 copy the pleasure value of emotion did not mediate the effect of type of metaphor on Ad Attitude (effect = .4040, CI = .1958 to 1.0053) (see Figure 3 6). The results revealed that consumers who had high NFC did not feel pleasure or entertainment very much to influence results on ad attitude playing a role of mediator when they were exposed to the pictori al metaphor ad with copy. Therefore, the pleasure value of emotion (e.g., pleasure or entertainment) did not mediate the relationship between visu al metaphor Figure 3 6 Conditional I ndirect E ffect on Attitude Toward the Ad B rand a ttitude Bootstrapping techniques of conditional indirect effects proposed the mediating effect of the pleasure value of emotion depending on conditions. The conditional indirect effect of type of metaphor on Brand Attitude was significant when consumers who had low NFC were e xposed to the metaphor ad with headline copy (effect = .97) and no
97 headline copy (effect = .54) bec ause the confidence interval for indirect effect excluded zero (CI for copy = .4618 to 1.5001, CI for non copy = .0381 to 1.0844). Moreover, the conditional indirect effect of type of metaphor on Brand Attitude was significant when consumers who had high N FC were exposed to the ad with no copy in the pictorial metaphor ad (effect = .87) because the confidence interval for indirect effect also excluded zero (CI = .3346 to 1.4217). However, pleasure value of emotion did not mediate the effect of type of meta phor on Brand Attitude when consumers who had high NFC were presented with the ad with headline copy (effect = .38, CI = .1843 to .9625) (see Figure 3 7). The results revealed that consumers who had high NFC did not feel pleasure or entertainment when the y were exposed to the pictorial metaphor ad with copy so that pleasure value of emotion (e.g., pleasure or entertainment) did not mediate the relationship between metaphor and the Figure 3 7 Conditional I ndi rect E ffect on Attitude Toward the Brand
98 Purchase i ntention Bootstrapping techniques of conditional indirect effects showed the mediating effect of pleasure depending on conditions. The conditional indirect effect of type of metaphor on Purchase Intention was significant when consumers who had low NFC were exposed to headline copy (effect = 1.11) and no headline copy (effect = .62) because the confidence inte rval for indirect effect excluded zero (CI for copy = .5163 to 1.7086, CI for non copy = .0516 to 1.2559). Also, the conditional indirect effect of type of metaphor on Purchase Intention was significant when consumers who had high NFC were exposed to the a d with no copy on the pictorial metaphor ad (effect = .99) because the confidence interval for indirect effect also excluded zero (CI = .3858 to 1.6338). However, affective response did not mediate the effect of type of metaphor on Purchase Intention when consumers who had high NFC were shown the ad with headline copy (effect = .44, CI = .2158 to 1.0725) (see Figure 3 8). The results indicated that consumers who had high NFC did not feel pleasure when they were exposed to the pictorial metaphor ad with cop y so that the pleasure value of emotion did not mediate the relationship between metaphor and the purchase intention of consumers at this particular condition.
99 Figure 3 8 Conditional Indirect E ffect on Purchase Intention Discussion Experiment 2 exami ned the interaction effects of pictorial metaphors and mediation effects of pleasure value of emotion on each dependent variable (e.g., attitude toward the ad/the brand, purchase intention) with considering the moderation effect of headline copies and cogn itive individual differences (e.g., NFC). Based on the process of testing from Hypothesis 2 (i.e., headline copy in an implicit or explicit metaphorical ad has a significant impact on ad attitude, brand attitude, and purchase intention), the use of headli ne copy in metaphorical ads clearly has an influence on the difficulty of the ad. But interestingly, there was no significant effect created by headline copy in metaphorical ads for across all types of products on n. This is in contrast to the previous research that found headlines in metaphorical ad have a positive effect on attitude toward the ad
100 and toward the brand (e.g., Bergkvist, Eiderback, & Palombo, 2012) and that suggested the use of a complete headline in metaphorical ads results in more positive effects on brand communications (e.g., Phillips, 2000). Obviously, using headline copy in the ad containing visual metaphors was helpful to decrease the difficulty of understanding what the ad intend to communicat e. As partially expected in Hypothesis 3, the effects of the type of visual metaphor NFC. That is, if the metaphorical ad includes headline copy as a cue for underst anding the visual metaphor, explicit metaphors more positively influence purchase intention for low NFC consumers. Consumers with low NFC are more likely to have greater purchase intention when they are exposed to the ad embedded explicit metaphor with hea dline copy rather than implicit metaphor with headline copy. Furthermore, even though consumers are in the level of high NFC, if the metaphorical ad does not involve headline copy, explicit metaphors are more likely to elicit greater purchase int ention tha n implicit metaphors It seems because, even without a verbal cue (i.e., headline copy), the product image embedded in metaphorical ads increases pur chase intention even for the consumers in the high NFC. As anticipated by Hypotheses 4, the results support the role of mediator of ad attitude, brand attitude, and purchase intention regardless of whether there are ad elements (e.g., headline copy) to aid in understanding of what the ad intended in the metaphorical ad. The findings in the study also indicate that there would be another factor/construct to influence ad effectiveness (i.e., Ad Attitude, Brand ddition to
101 the use of copy as a cognitive cue when the ad is created by visual metaphors that require thinking process (i.e., cognitive elaboration) for the interpretation of such type of ad messages.
102 CHAPTER 4 GENERAL DISCUSSION Conclusion Metaphorical expression in visual modality (i.e., visual metaphors) is very common in ads and seem s to be very effective in cluttered viewing conditions. That is, visual metaphors apparently draw attention away from the clutter of typical ads and persuade targets in unique ways (Phillips, 2003). In addition, this rhetoric al visual expression is highly relevant to advertising creativity which is critical to successful advertising. Advertising creativity or novelty is an indispensable aspect in promoting sti muli for greater ad effectiveness (Smith, Chen & Yang, 2008). For example, ad creativity or novelty leads to increased ad recall, more favorable ad attitudes, and positive emotional response (Ang, Lee, & Leong, 2007). Advertisers or practitioners agree tha t the expression of ad elements related to ad creativity is a key component and plays a weighty role in effective ad design for persuasion. Yet, there has been little research on ad expression as a message factor in academic literature. Therefore, advertis ers or practitioners face a challenge to choose an appropriate type of creative stimuli to lead to greater ad effectiveness. The findings of the current study reveal that, compared with implicit metaphors, the ad employing explicit metaphor elicits greate r purchase intention at both high and low level s of NFC, but are contingent upon using headline copies in the ad. If consumers have high NFC when the ad with explicit metaphor is clear about the subject and the product image does not involve headline copy the ad engenders much higher purchase intention than an ad with implicit metaphor and non copy does. In
103 addition, if consumers have low NFC, when the ad with explicit metaphors involve headline copy, the ad provokes greater purchase intention than the ad with implicit metaphors involving headline copy does. However comparing the mean of the results, the condition of the former (that is, when consumers have high NFC and the ad has explicit metaphors without headline copy) has more positive effect s on achi eving great purchase intention than the condition of the latter (that is, when consumers have low NFC and the ad has explicit metaphors with headline copy). Furthermore, all findings illustrate that consumers with low NFC need to be actively assisted by th e advertised product image with explicit metaphors which show the advertised pr oduct in the metaphorical image. This leads to a more successful effect on purchase intention, regardless of the use of headline copy. In previous research, discussions have e laborated on the fact that imagery processing enhances affective responses to stimuli and behavior (Greenwald & Leavitt, 1984; MacInnis & Price, 1987). The findings of the current study are consistent with the se discussions in academia. In support of the d iscussions, the findings in this study note that an ad with explicit metaphors has a greater impact on consumers' behavioral intention than the ad with implicit metaphors even for the consumers with high NFC (Need for Cognition). In addition, in terms of t he effect of the headline copy in the metaphorical ad, there is no effect of copy even in the implicit metaphor ad which is not able to provide consumers any cue to decode the implicit metaphors. That is, for the consumers with low NFC, the ad with explici t visual metaphor has greater impact on purchase intention than the ad with implicit metaphors (regardless of the existence of headline copy ). For consumers with high NFC, when the ad has a headline copy, there
104 is no difference between the ad with explicit and implicit metaphors H owever, when the ad does not have headline copy, the ad with explicit metaphors has a much greater impact on purchase intention than the ad with implicit metaphors. According to the current study pleasure value of emotion in the ad employing visual metaphors plays an important role on the influence of persuasive ad effectiveness. Visual metaphors in the ad engender much deeper cognitive elaboration of messages, and visual metaphors may need mental imagery processing (MacInnis & Price, 1987; 1990) to contribute to the persuasive effects of the use of visua l metaphors besides cognitive motivation (e.g. NFC) and affective response (e.g. pleasure value of emotion) NFC and affective responses (especially, pleasure). For high NFC consumers, visual metaphors enhance purchase intentions when there is no headline copy but the behavioral int ention is not mediated by affective responses (i.e., pleasure) if there is headline copy in the ad. That is, under the condition of an ad with headline copy, for high NFC consumers visual metaphors may not provoke positive affective responses (i.e., pleas ure) and thus, they do not provoke greater purchase intention. When pleasure value of emotion occurs from visual metaphors in the ad and the affective response (i.e., pleasure) mediates the effect of visual metaphors visual metaphors in the ads enhance p ersuasive ad effectiveness for consumers with low level of NFC regardless of the use of headline copy in the ad. This illustrates the persuasive effects of metaphorical expression which may be due to the characteristic of visual metaphors themselves and ar ises out of cognitive verbal cues for the understanding of such type s of messages. The findings were relevant to previous research (e.g.,
105 Haugtvedt, Petty, & Cacioppo, 1992; Petty & Cacioppo, 1986) which indicated that the responses of low NFC consumers to visual metaphors were less influenced by the cognitive cues of the message (i.e., headline copy). The findings of the current study reveal that when using visual metaphors in the ad, it is important to evoke emotional responses, especially pleasure thro ugh tension and relief to decode in order to elicit more positive the brand, and purchase intention Early research suggested that affective priming of the ad context, such as affective tone of the article, significantly influenced advertising effectiveness (Batra & Stayman, 1990). The pleasure of metaphoric information processing provides sufficient interest to elicit great er ad effectiveness for the advertised products. Taken together, all findings indicate that there m ay be an advantage to using explicit metaphors with headline copy for low NFC target consumers or without headline copy for high NFC target consumers. However, only when consumers feel the pleasure value of emotion from the visual metaphor s in the ad does the use of such metaphors lead to greater ad effectiveness from even low NFC consumers on purchase intention even though a m etaphorical ad does not involve headline copy. Theoretical Implications visual arguments are more persuasive than verbal arguments. Furthermore, J (2008) study demonstrated that metaphorical image s without verbal argument in the ad were more eff ective than non metaphorical image s with verbal argument on cognitive elaboration and source cre dibility The researcher also found that there were significant differences between the ad with visual metaphor images and the ad with non metaphor
106 images with r egard to attitude toward the ad But there were no significant differences between the ad which employed metaphorical image with copy and the ad which employed a metaphorical image without accompanying verbal copy. However, the findings in the previous st udies related to metaphorical expression in visual modality (i.e., visual metaphors) seem ed to differ according to outcomes (e.g., dependent variables) and by advertised product types. The current study attempted to test which conditions associated with t he use of two types of visual metaphors work to positively influence s and purchase intention. The research also considered both cognitive and affective responses and determined the conditional impact of the ads by the type of visual metaphors and individual difference. The most important contribution of the present research is that the findings have confirmed the mediating effect of the pleasure value and mo tivational a spect (i.e., Need for Cognition ) and affective aspect, thereby verifying the effect of visual metaphors. Moreover, another contribution of the current study is to investigate the effects of two types of visual metaphors in ads, considering the ads applied to a ll product types (Vaughn, 1980). This could then improve the external validity of the study. The findings of E xperiment 1 are consistent with previous research ( Gkiouzepas & Hogg, 2011; Mitchell & Olson, 1981) and prove once again that vis ual metaphors in the ad are an effective persuasion strategy However, in contrast to the previous studies, this study clearly tested whether the effects of visual metaphors occur in the ads across all products. The current study confirms that the ad using metaphorical
107 images is more persuasive changing s toward the ad positively and increase s their purchase intention compared to the ad with non metaphorical images, especially, across all produc t types by FCB grid model (Vaughn, 1980). In addition to the support of E xperiment 1, E xperiment 2 suggests that a type of visual metaphor with verbal copy in the ad may lead to greater persuasive effects with regard to purchase intention, depending on the cy (e.g., NFC). The findings of E xperiment 2 also include why the effects of using visual metaphors and headline copies may vary The results of the current study are more specific than previous resear ch (e.g. Chang & Yen, 2013; Haugtvedt, Petty, & Cacioppo, 1992). According to the level of NFC, the results provided further evidence for the effect of using visual metaphors in ads when considering headline copy. The current study used ELM and NFC (Brenna n & Bahn, 2006; Chang & Yen, 2013; Haugtvedt, Petty, &Cacioppo, 1992) and relevance theory (Sperber & Wilson, 1995, 2004) to explain the relationship between headline the type of visual metaphors such as explicit exhibition of the product image in the metaphorical ad. When facing an advertised product image in visual metaphors such as explicit metaphor, consumers with high NFC may not need headline copy to be persuaded more positively. However, for con sumers with low NFC, the exhibition of the product image in the metaphorical ad is not enough to assist with persuasion of consumer s with low NFC in managing information efficiently (e.g., information processing). Therefore, consumers with low NFC need ano ther ad element, headline copy, to assist with their persuasion. When the ad does not involve headline copy, even if consumers have high
108 level s of NFC, implicit metaphors might be too difficult or complex to persuade consumers enough. Moreover, there was n o significant effect under the condition with both implicit metaphor and headline copy for consumers with high level s of NFC. This infers that response to implicit visual metaphor might be more influenced by other constructs such as imagery processing abil ities (MacInnis & Price, 1987), not only by cognitive elaboration or cognitive motivational factor s (e.g., NFC) Based on findings from testing of Hypothesis 2, 3 and 4, cognitive tendency (i.e., Need for Cognition) and emotional response might be importan t factors in interpreting visual metaphors in ads and persuading consumers of the positive effect of visual metaphors in the ad The findings suggest that there might be multiple criteria to explain the ad effect of visual metaphors and there might be other factors which should be considered when determining the effect of visual metaphors such as imagery processing abilities (MacInnis & Price, 1987). The findings in Hypotheses 2 and 3 reveal that the ads using implicit metaphors seem to provide more es sentials for imagination, thereby increasing cognitive elaboration of the ad message. Research on individual differences in information processing has illustrated a cognitive impact in consumer behavior (Batra & Stayman, 1990; Petty & Cacioppo, 1986). Info rmation processing consists of two factors: ability and preference (Childers, Huston, & Hecker, 1985). Individuals apply an information processing strategy based on their preferences and their cognitive ability. The study of individual differences in infor mation processing was extended to the one in mental imagery processing (Childers & Houston, 1983; Lutz & Lutz, 1978; MacInnis & Price,
109 visualization of a concept or relatio differences in imagery information processing can have significant influences on the range of cognitive performance such as memory, perception, learning, and problem solving (Ernest, 1977). Imagery processing i s evoked as internal sensory experiences in working memory (MacInnis & Price, 1987). Moreover imagery information processing involves perceptual representations of ideas, feelings, and memories within working memory and also allows for direct recovery of past experiences (MacInnis & Price, 1987). Imagery processing, therefore, involves intuitive and empirical support (MacInnis & Price, 1987). The literature about individual differences in imagery processing states that mental imagery can improve memory f or product related information (Childers & Houston, 1983; Lutz & Lutz, 1978; Rossiter, 1982) and mental imagery is also more likely to arouse favorable attitudes (MacInnis & Price, 1990). Additionally, imagery may play a significant role in forming cogniti ve elaboration (Greenwald & Leavitt, 1984) and thus, it contributes to the formation of consumer attitudes (Rossiter & Percy, 1978; 1980). According to MacInnis and Price (1987), mental imagery processing provides high elaboration processing and, consequen tly, this is considered a factor to improve communication effectiveness. In addition, it is influenced by individual differences in mental imagery processing ability Therefore, in addition to the level of NFC, it would be mental imagery ability to evaluate the extent to which visual metaphors are expressed in the ad image or which features should be included in a visual metaphor ad for successful persuasion.
110 From the point of view of imagery processing, visual metaphors h ave a story associated with the product benefits These stories are more helpful for consumers who have high imagery processing ability and who are able to generate their mental imagery well. In this light, visual metaphors seem to be more effective for co nsumers who can translate the meaning into vivid mental imagery. As a result, cognitive tendency to enjoy the thinking process (i.e., NFC) would only be one of the factors necessary to influence the effectiveness of the ad employing visual metaphors. Addin g to cognitive tendency (i.e., NFC), imagery processing ability should be considered as a critical factor that has behavioral intention. Practical I mplications The results of the current study provide several meaningful managerial implications for advertisers. First, the current investigation reveals that ads employing purchase intentio n. In particular explicit visual metaphors including headline copy are more likely to positively influence consumers who do not enjoy thinking deeply or solving problems. Therefore, in terms of creative strategy, if visual metaphors are employed in the pr intention could be expected considering target characteristics. However, even though consumers tend to have the characteristic of low NFC, if they feel pleasure value from visual metaphors this has a positive influence on purchase intention regardless of the use of headline copy in the ads. Generally, advertisers or marketers should consider visual modality (e.g., visual metaphor or non metaphor) and know when they should apply specific rhetoric styles (e.g., explicit or implicit visual metaphor) to the ad with a
111 headline copy, considering various product types which were interspersed for four product categories according to the FCB model (Vaughn, 1980). Second, advertisers or marketers may appl y visual metaphors to the ad according to targeted consumers while considering their level of NFC. For example, in terms of ad placement strategy, advertisers may apply visual metaphors to the different kinds of media placement contingent on targeting lite rate, well educated people who have a tendency to enjoy the thinking process with the hope that these consumers will notice, consider, recognize, and develop greater behavioral intention. This may cause greater purchase intention when using visual metapho rs in ads. One method of increasing advertising effectiveness is to choose an appropriate target media such as TIME or PEOPLE magazine whe n using metaphorical ad images or to employ unique ad related features ( Sierra, Heiser, & Torres, 2012) considering d ifferent types of visual metaphors within the ad. Third, when using visual metaphors in the ad, marketers and advertising practitioners should consider emotional responses (e.g., pleasure) as a critical antecedent, which is related to favorable advertisin g effectiveness. Feeling s of pleasure from the metaphorical ads may improve advertising results. The current investigation shows the important role of pleasure value of emotion between tension and relief while decoding the meaning of the metaphorical ads f or greater persuasive advertising effectiveness. The ad should then employ visual metaphors to elicit great pleasure feeling from decoding the metaphorical ads in order to develop these positive ad results. Fourth, the present study provides guidance in th e development of creative strategies on how to frame a visual approach using two types of metaphors (explicit or implicit)
112 with or without verbal elements by target consumers across all categories of product involvement. Limitations and Suggestions for Future Research Several limitations of this research must be acknowledged. First, the research shares common weaknesses of experimental studies, s uch as an artificial situation used when approaching participants with a lot of ads simultaneously. At the sam e time, an attempt was made to cover various product categories in the ads. Therefore, the internal validity such as experimental mortality or maturation is a concern Second, this study was conducted using real advertisements to verify the creative quali ty of the ads. Thus, the ad stimuli did not have one message feature such as a specific tone and manner. Therefore, another confounding factors could occur through the use of a lot of ad stimuli in each type of visual metaphor with various message s in tone and manner. Third, there could be a possibility of finding other confounding factors that influence consumers to change their attitudes or behavioral intention. These could include the degree of imagery processing ( Childers & Houston, 1983) ability, the degree of congruence of visual metaphors and product benefits, and product knowledge. One of these factors such as the level of product knowledge could be a possible moderator in future studies because it could facilitate the ability to decode visua l metaphors in the ads. In this regard, future research could consider individual differences in imagery processing ability (MacInnis & Price, 1987) and the level of advertised product knowledge for each advertised product as a confounding factor (e.g., c ovariate) or moderator. Moreover, f uture research needs to focus on identifying various factors that have an influence on the effectiveness of the ad involving visual metaphors, such as
113 imagery processing ability (MacInnis & Price, 1987 ). Future studies ma y consider creating the ad stimuli with the same tone and manner in both types of visual metaphors to gain better internal validity. Also, g iven the many positive effects of visual metaphors identified in the literature, it is important to consider target tendencies and whether or not to use headline copy when deciding which types of visual metaphors to employ in an ad. The current research extends visual communication study on to advertising effects in several aspects across all types o f advertised products. This study incorporates the analysis of visual metaphor in investigating the influences of two critical aspects of ad content: cognitive responses that provoke elaborated decoding processing, and affective responses that induce posit ive emotion such as pleasure, entertainment, or enjoyment.
114 APPENDIX A QUESTIONNAIRES For Experiment 2 We would appreciate it if you would take some time to complete this simple survey. This should be take about 15 minutes of your time and your results will be completely anonymous. Your responses are important to us so we know how we can better improve our efforts. The main objective of this research is to get your reactions to magazine advertisements and your behavior. These advertisements are for products or brands in the world market. Therefore, first, you will be asked about you rself, and then you wil l be given magazine advertisements. Before you evaluate or rate the advertisement, please take a moment and review the advertisements very carefully. The following advertisement is for an automobile product in the world market. Please take a moment and review the advertisement. [Place an implicit/explicit metaphor ad about here] Is the advertised product ( a car ) show ing or integrated into the main metaphoric image in this ad? (Excluding the small product image at the bottom of the ad) Yes, i t is. No, it is not. When you see the above advertisement and understand what the ad intends, how much do you feel one of each emotion below? Displeasure ____ : ____ : ____ : ____ : ____ : ____ : ____ Pleasure enjoyment ____ : ____ : ____ : ____ : ____ : ____ : ____ Not enjoyment Not entertainment ____ : ____ : ____ : ____ : ____ : ____ : ____ entertainment Fun ____ : ____ : ____ : ____ : ____ : ____ : ____ Not Fun The following questions are asking for your opinion on the ad which you have seen. Overall, I consider the ad above to be ______. Bad ____ : ____ : ____ : ____ : ____ : ____ : ____ Good Favorable ____ : ____ : ____ : ____ : ____ : ____ : ____ Unfavorable Unpleasant ____ : ____ : ____ : ____ : ____ : ____ : ____ Pleasant The following questions are asking for your opinion on the brand which you have seen. Overall, I consider the brand above to be ________.
115 Bad ___ : ___ : ___ : ___ : ___ : ___ : ___ Good Favorable ___ : ___ : ___ : ___ : ___ : ___ : ___ Unfavorabl e Unpleasa nt ___ : ___ : ___ : ___ : ___ : ___ : ___ Pleasant After seeing this advertisement, it may encourage me to purchase this product next time. very unlikely ___ : ___ : ___ : ___ : ___ : ___ : ___ very likely probable ___ : ___ : ___ : ___ : ___ : ___ : ___ improbabl e impossibl e ___ : ___ : ___ : ___ : ___ : ___ : ___ possible Need for Cognition (Cacioppo, Petty, & Kao, 1984) Next, you will be given a list of character traits which best represents you. Please consider the following lists of character traits, and click the button that describes you. Strongly Strongly Disagree Agree 1. I prefer complex to simple problems. _____:_____:_____:_____:_____:_____:_ ____ 1 2 3 4 5 6 7 2. I like to have to do a lot of thinking _____:_____:_____:_____:_____:_____:_ ____ 1 2 3 4 5 6 7 thinking in depth about something. _____:_____:_____:_____:_____:_____:_ ____ 1 2 3 4 5 6 7 4. I prefer to do something that challenges my thinking abilities rather than something that requires little thought. _____:_____:_____:_____:_____:_____:_ ____ 1 2 3 4 5 6 7
116 5. Thinking hard and for a long time about something gives me satisfaction. _____:_____:_____:_____:_____:_____:_ ____ 1 2 3 4 5 6 7 The following are demographic questions that will only be used for statistical analyses. Have you participated in this study before? (your response to this question will not influence your credit for participation) Yes No Maybe What is your gender? Fem ale ____ ___ M ale _______ What is your age? _____________ My highest level of education is: Some high school High school graduate Some college College graduate Graduate degree Other How would you classify yourself? Anglo American / Caucasian _________ African American _________ American Indian / Native American _________ Asian American _________ Hispanic / Latino _________ Multiracial _________ Other (Please specify ___________________)
117 APPENDIX B E XPERIMENTAL AD STIMULI All a d stimuli have been modified or re created and the original image resources come from google images. Ad s timuli used in E xperiment 1 by product types (Non Metaphor vs Metaphor) (Visual Metaphor) ad stimuli by product types total number of ad s : 12 Involvement Think Feel 1 Informative 2 Affective High involvement Blender mixer The original image: WALITA ad (2016). Retrieved from https://www.pinterest.com/pin/3084262 74455188663/ Car The original image: Jeep ad (2016). Retrieved from https://www. pinterest.com/pin/1130826 40616697457/ Cosmetic The original image: Olay UK ad (2016). Retrieved from http://www.wired.com/2007/04/cosmetic_com pan/ Outdoor shoes The original image: North Face poster (2016). Retrieved from http://www.doctordisruption.com/wp content/uploads/2012/09/designed_for_profes sionals2.jpg
118 Grass trimmer The original image: STIHL ad (2016). Retrieved from https://www.pinterest.com/pin/5621055 97213427256/ Underwear The original image: Wonderbra ad (2016). Retrieved from https://adsoftheworld.com/forum/2057 3 Habit formation 4 Self satisfaction Low involvement Laundry detergent The original image: Ariel ad (2016). Retrieved from https://www.coloribus.com/adsarchive/ prints/ariel m zim 5 bulb 10639855/ Candy The original image: Juice candy (2016). Retrieved from https://nickglauth.wordpress.com/advertising ads/#jp carousel 58
119 Lint roller The original image: 3M (2016). Retrieved from http://theinspirationroom.com/daily/201 2/3m lint roller pets/ Baby diapers The original image: Milette ad (2016). Retrieved from https://www.coloribus.com/adsarchive/ prints/unknownadvertiser milette 2533405/ Hot sauce The original image: Tabasco (2016). Retrieved from https://www.pinterest.com/pin/128071183130 068107/ Beverage The original image: Bramhults Swedish ad (2016). Retrieved from http://www.ateriet.com/clever juice ads for bramhults/
120 (Non Metaphor) ad stimuli by product types total number of ad s : 12 Involve ment Think Feel 1 Informative 2 Affective High involve ment T.V. The original image: Baumann Meyer (2016). Retrieved from http://www.baumannmeyer.com.au/ marketing.htm Car The original image: Honda ad (2016). Retrieved from http://indianautosblog.com/2013/04 /honda amaze garners over 6000 bookings 71729 Vacuum cleaner Cosmetic (Lip Balm) The original image: MAYBELLINE (2016). Retrieved from https://dejuanamwilliams.wordpress.com/2015/08/ Jean The original image: Google images (2016). Created and modified Retrieved from https://www.toovia.com/posts/2014/jul/03/0.13503.4 07356837261017098 Shoes The original image: Bata ad (2016). Retrieved from https://www.behance.net/gallery/7672579/Bata print ads
121 The original image: Electro lux ad (2016). Retrieved from https://www.choice.com.au/product s/home and living/laundry and cleaning/vacuum cleaners/electrolux ergorapido lithium 2in1 zb3012 3 Habit formation 4 Self satisfaction Low involve ment Laundry detergent The original image: Google images (2016). Created and modified Retrieved from https://www.masterfile.com/stock photo search/en/clean+clothes Toothpaste The original image: Aquafresh (2016). Created and modified Retrieved from https://www.buzzfeed.com/copyran ter/really gross print ads?utm_term=.hlAw1Wx6#.hj74V BjO Butter The original image: Google images (2016). Retri eved from http://8thgradeelapersuasiveunit.wikispaces.com/fil e/view/butter.jpg Beverage The original image: Glen fiddich (2016). Retrieved from https://www.pinterest.com/pin/31623720506107846 9/
122 Baby diapers The original image: Pampers (2016). Retrieved from http://strategyonline.ca/2010/12/01/ moypampers 20101201/ Chocolate The original image: Sunfeast Dark Fantasy chocolate (2016). Retrieved from http://nazninazeez.com/2012/07/18/review sunfeast dark fantasy chocolate biscuit/ Ad Stimuli used in E xperiment 2 by product types (Implicit vs Explicit Metaphor) (Implicit Metaphor) ad stimuli by product types total number of ad s : 12 Involvem ent Think Feel 1 Informative 2 Affective High involveme nt Car The original image: Jeep ad (2016). Retrieved from https://www.pinterest.com/pin/1130826406 16697457/ Eye wear (glasses)
123 Blender mixer The original image: WALITA ad (2016). Retrieved from https://www.pinterest.com/pin/3084262744 55188663/ Electric Iron The original image: Ozeal ad (2016). Retrieved from https://www.pinterest.com/pin/638245196 92271013/ Two Underwear ads The original image: Wonderbra ad (2016). Retrieved from https://adsoftheworld.com/forum/2057
124 The original image: F aultless spray starch (2016). It was re c reated and modified Underwear ads The original image: Wonderbra (2016). Retrieved from http://i1.wp.com/www.chrisrawlinson.com/ wp content/uploads/2010/02/wonderbrastairs .jpg 3 Habit formation 4 Self satisfaction Low involveme nt Hand sanitizers ads Delivery Pizza
125 The original image: Purell Sanitizer ad (2016). Retrieved from http://www.onlinemastersinpublichealth.co m/powerfully creative hand hygiene ads/ Sanitizer ad The original image: Lifebuoy Handwash ad (2016). Retrieved from http://justsomething.co/23 creatively funny print ads will make you giggle/ The original image: Pizza ad (2016). Retrieved from https://w ww.coloribus.com/adsarchive/pri nts outdoor/pizza hut home delivery service homeless 8073255/ Hot sauce The original image: Tabasco (2016). Retrieved from https://www.pinterest.com/pin/128071183 130068107/ Mouthwash
126 Salad dressing The original image: OSEM ad (2016). Retrieved from https://adsoftheworld.com/media/print/ose m_cucumber The original image: Even mouthwash ad (2016). Retrieved from https://www.pinterest.com/pin/478718635 368233854/ (Explicit Metaphor) ad stimuli by product types total number of ad s : 12 Involve ment Think Feel 1 Informative 2 Affective High involvem ent Ski Board helmet Stockings
127 The original image: Boeri ad (2016). Retrieved from https://forsenorse.com/tag/boeri ads/ Grass trimmer The original image: Nur Die ad (2016). Retrieved from http://adsoftheworld.com/media/print/yuc ca?s ize=original Outdoor Shoes The original image: North Face poster (2016). Retrieved from http://www.doctordisruption.com/wp content/uploads/2012/09/designed_for_p rofessionals2.jpg
128 The original image: STIHL ad (2016). Retrieved from https://www.pinterest.com/pin/562105597213 427256/ Lawn mower The original image: Gardena ad (2016). Retrieved from http://www.advertolog.com/gardena/print outdoor/quiet 11738805/ Cosmetic The original image: Olay UK ad (2016). Retrieved from http://www.wired.com/2007/04/cosmetic_ compan/ 3 Habit formation 4 Self satisfaction
129 Low involvem ent Lint roller The original image: 3M (2016). Retrieved from http://theinspirationroom.com/daily/2012/3m lint roller pets/ Baby Diapers Candy The original image: Juice candy (2016). Retrieved from https://nickgl auth.wordpress.com/adverti sing ads/#jp carousel 58
130 The original image: Milette ad (2016). Retrieved from https://www.coloribus.com/adsarchive/prints/ unknownadvertiser milette 2533405/ Toothbrushes The original image: Sensodyne ad (2016). Retrieved from https://www.coloribus.com/adsarchive/prints/ sensodyne toothpaste feathers 13444755/ Beverage The original image: Bramhults Swedish ad (2016). Retrieved from http://www.ateriet.com/clever juice ads for bramhults/
131 Gum The original image: Extra Gum ad (2016). Retrieved from https://www.coloribus.com/adsarchive/pri nts/extra toothbrush 3699505/
132 LIST OF REFERENCES Ang, S.H. & Lim, E. (2006). The influence of metaphors and product type on brand personality perceptions and attitudes. Journal of Advertising 35 (2), 39 53. Ang, S.H., Lee, Y.H., and Leong, S.M. (2007). The ad creativity cube: Conceptualization and initial validation. Journal of the Academy of Marketing Science, 35 (2), 220 232. Batra, R. & Stayman, D. M. (1990). The ro le of mood in advertising effectiveness. Journal of Consumer Research 17, 203 214. Batra, R. & Ray, M. L. (1986). Affective responses mediating acceptance of advertising. Journal of Consumer Research 13, 234 249. Bergkvist, L., Eiderback, D., & Palomb o, M. (2012). The brand communication effects of using a headline to prompt the key benefit in ads with pictorial metaphors. Journal of Advertising 41 (2), 67 75. Boozer, R. W., Wyld, D. C, & Grant, J. (1991) Using metaphor to create more effective sales messages Journal of Consumer Marketing, 8 (1), 59 67. Borghini, S., Visconti, L. M., Anderson, L., & Sherry, J.F. Jr. (2010). Symbiotic postures of commercial advertising and street art. Journal of Advertising 39 (3), 113 126. Bowers, J. W. & Osborn, M.M. (1966). Attitudinal effects of selected types of concluding metaphors in persuasive speeches. Speech Monographs 33 (2), 147 155. Brennan, I., & Bahn, K. D. (2006) Literal versus extended symbolic messages and advertising effectiveness: The moderat ing role of need for cognition Psychology & Marketing 23, 273 295. Bruner, G. C. & Hensel, P. J. (1992) Marketing Scales Handbook: A Compilation of Multi Item Measures 1. Chicago: American Marketing Association. Bryant, J., & Miron, D. (2003). Excita tion transfer theory. In J. Bryant, D. Roskos Ewoldsen, & J. Cantor (Eds.), Communication and emotion: Essays in honor of Dolf Zillmann (pp. 31 59). Mahwah, NJ: Erlbaum. Cacioppo, J. T., Petty, R. E., & Kao, C. F. (1984). The efficient assessment of need for cognition Journal of Personality Assessment 48 (3), 306 307. Cacioppo, J., Petty, R. (1982) The need for cognition Journal of Personality and Social Psychology 42, 116 131. Caine, R. N., & Caine, G. (1991). Making connections: Teaching and the hu man brain
133 Celsi, R i chard L. & Olson, Jerry C. (1988) The Role of involvement in attention and comprehension processes Journal of Consumer Research 15, 210 224. Chang, C., & Yen, C. (2013). Missing ingredients in metaphor advertising: The right formul a of metaphor type, product type, and need for cognition. Journal of Advertising 42(1), 80 94. Childers, T. & Huston M. (1983). Imagery paradigms for consumer research: alternative perspectives from cognitive psychology. Advances in Consumer Research, 10 59 64. Childers, T., Huston, M. & Hecker, S. (1985). Measurement of individual differences in visual vs verbal information processing, Journal of Consumer Research 12, 125 134. Cohen, A. R., Stoland, E., & Wolfe, D. M. (195 7 ) An experimental investigation of need for cognition Journal of Abnormal and Social Psychology (51), 291 294. Corbett, E. (1990) Classic rhetoric for the modern student New York: Oxford University Press. DeRosia, E. D. (2008). The effectiveness of non verbal symbolic signs and metaphors in advertisements: An experimental inquiry. Psychology & Marketing, 25 298 316. Edell, J. & Burke, M. C. (1987). The power of feelings in understanding advertising effects. Journal of Consumer Research 14, 421 433. Er nest, C. (1977). Imagery ability and cognition: A critical review. Journal of M ental Imagery 2, 181 216. Fairhurst, G. & Sarr, R. (1996). The Art of Framing: Managing the Language of Leadership Jossey Bass Feldman, J. & Narayanan, S. (2004). Embodie d meaning in a neural theory of language. Brain and Language 89(2):385 392 Forceville, C. (1998) Pictorial Metaphor in Advertising Psychology Press. Forgas, J. P. & Bower, G. H. (1987). Mood effects on person perception judgments. Journal of Personalit y and S ocial P sychology 53 (1), 53 60. Gardner, P. M (1985). Does Attitude toward the Ad Affect Brand Attitude under a Brand Evaluation Set? Journal of Marketing Research 22, 192 198. Geary, J. (2011). I is an other: the secret life of metaphor and how it shapes the way we see the world. New York: HarperCollins.
134 Gilbert, D., Fiske, S., & Lindzey, G. (1998). The handbook social psychology Oxford University Press. Gkiouzepas, L & Hogg, M. K. (2011). Articulatin g a new framework for visual metaphors in advertising: A s tructural, c onceptual, and p ragmatic i nvestigation Journal of Advertising 40 (1), 103 120. Greenwald, A.G. & Leavitt, C. (1984). Audience involvement in advertising: four levels, Journal of Consu mer Research 11, 581 592. Goatly, A. (2007). Washing the Brain Metaphor and Hidden Ideology (Discourse Approaches to Politics, Society and Culture). John Benjamins Publishing Company. Goldberg, Marvin E., & Hartwick, J. (1990) The effects of advertis er reputation and extremity of advertising claim on advertising effectiveness Journal of Consumer Research, 17. 172 179. Haugtvedt, C., Petty, R.E., & Cacioppo J.T. (1992). Need for cognition and advertising: Understanding the role of personality variables in consumer behavior. Journal of Consumer Psychology 1 (3), 239 260. Hayes, A. F. (2012). PROCESS: a Versatile Computational Tool for Observed Variable Med iation, Moderation, and Conditional Process Modeling. White Paper. The Ohio State University. Available on line at: http://www.afhayes.com/public/process2012.pdf (Accessed February, 2015). Hayes, A. F. (2013). SPSS, SAS, and M plus Macros and Code. Available online at: http://afhayes.com/introduction to mediation moderation and co nditional process analysis.html (Accessed: September 24, 2015) Holbrook, M.B. & Batra, R. (1987). Assessing the role of emotions as mediators of consumer responses to advertising. Journal of Consumer Research 14, 404 419. Hovland, C., Janis I., & Kelly, H. (1953) Communication and Persuasion New Haven, CT: Yale Univ. Press. Isen, A. M. & Levin, P.F. (1972). The effect of feeling good on helping: Cookies and Kindness. Journal of Personality and Social Psychology 21(3), 384 388. Isen, A. M. & Simmonds, S. (1978). The effect of feeling good on a task that is incompatible with good mood. Social Psychology Quarterly 41 (4), 346 349. Jeong, S. (2008) Visual metaphor in advertising: I s the persuasive effect attributable to visual argumentation or metaphor ical rhetoric? Journal of Marketing Communications 14, 59 73.
135 Kim, J., Baek, Y. & Choi, Y. H. (2013). The Structural Effects of Metaphor elicited Cognitive and Affective Elaboration Levels on Attitude Toward the Ad. Journal of Advertising 41(2), 77 96. Kim, J., & Morris, J. D. (2007). The power of affecitve response and cognitive structure in product trial attitude formation. Journal of Advertising 36(1), 95 106. Lakoff, G. & Johnson, M. (2003). Metaphors we live by Humans as Symbolic Creatures 2, 124 135 Lang, P. J. (1985). The cognitive psychophysiology of emotion: Anxiety and the anxiety disorders Hillsdale, NJ: Lawrence Erlbaum. Lang, P. J. (1995). The emotion probe Studies of motivation and attention. American Psychologist 50(5), 372 385. La ng, A. (2000). The limited capacity model of mediated message processing. Journal of Communication 50(1), 46 70. Lang, A. (2006). Using the Limited Capacity Model of Motivated Mediated Message Processing to Design Effective Cancer Communication Messages. Journal of Communication 56, 557 580. Leigh, J. H. (1994). The use of figures of speech in print ad headlines. Journal of Advertising 23(2), 17 33. Lutz, K. A. & Lutz, R. J. (1977). The effects of interactive imagery on learning: Application to adverti sing. Journal of Applied Psychology 62, 493 498. Lutz, K. A. & Lutz, R. J. (1978). Imagery Eliciting strategies: review and implications of research. Advances in Consumer Research 5, 611 620. MacInnis, D. J., & Price, L. L. (1987). The role if imagery in information processing: Review and extensions, Journal of Consumer Research 13, 473 491. MacInnis, D. J., & Price, L. L. (1990). An exploratory study of the effects of imagery processing and consumer experience on expectations and satisfactions. Advan ces in Consumer Research 17, 41 47. Malkewitz, K., Wright, P. & Friestad, M. (2003). Persuasion by Design: The State of Expertise on Visual Influence Tactics, in L. Scott and R. Batra (eds.) Persuasive Imagery: A Consumer Response Perspective pp. 3 15 Mahwah, NJ: Lawrence Erlbaum Associates. MacKenzie, S. B., Lutz, R. J., & Belch, G. (1986). The role of attitude toward the ad as a mediator of advertising effectiveness: A test of competing explanations. Journal of Marketing Research 23, 130 143.
136 Mack ie, D. M. & Worth, L. T. (1989). Processing deficits and the mediation of positive affect in persuasion. Journal of Personality and Social Psychology 57 (1), 27 40. McKelvie, S. J., Sano, E. K., & Stout, D. (2001) Effects of colored separate and interac tive pictures on cued recal l. The Journal of General Psychology 121, 241 251. McQuarrie, E. F. & Mick, D. G. (1996). Figures of Rhetoric in Advertising Language. Journal of Consumer Research 22, 424 438. McQ uarrie, E. F., & Mick, D. G. (1999). Visual rhetoric in advertising: Text interpretive, experimental, and reader response analyses. Journal of Consumer Research 26 (1), 37 54. http://doi.org/10.1086/209549 McQuarrie, E. F., & Mick, D. G. (2003) Visual and verbal rhetorical figures under directed processing versus incidental exposure to advertising Journal of Consumer Research 29 (March), 579 587. McQuarrie E. F. & Phillips, B. (2005). Indirect Persuasion in Advertising: How Consumers Process M etaphors Presented in Pictures and Words. Journal of Advertising 34(2), 7 20. Mehrabian, A (1978) Measures of individual differences in temperament Educational and Psychological Measurement 38, 1105 1117. Mehrabian, A. & Russell, J. A. (1974). An approach to environmental psychology Cambridge, MA: MIT Press. Mitchell, A. A. & Olson, J.C. (1981). Are product attribute b eliefs the only mediator of advertising effects on brand attitude? Journal of Marketing Research, 18(3), 318 332. Morris, J. D., Woo, C., & Singh, A. J. (2005). Elaboration likelihood model: A missing intrinsic emotional implication. Journal of Targeting, Measurement and Analysis for Marketing 14, 79 98. Morris, J. D., Woo, C ., Geason, J., & Kim, J. (2002). The powe r of affect: Predicting intention. Journal of Advertising Research 42(3), 7 17. revised creative product semantic scale. Creativity Research Journal 2(4), 267 278. P awlowski, D.R., Badzinski, D.M., & Mitchell N. (1998) Effects of metaphors on Journal of Advertising 27 (2), 83 98.
137 Petty, R.E., Cacioppo, J.T., (1986). Communication and persuasion: Centr al and peripheral routes to attitudes change New York: Spring Verlag. Petty, R. E. Unnava, H. R., & Strathman, A J. (1991). Theories of attitude change, in: T. S. Robertson & H. H. Kassarjian (Eds), Handbook of Consumer Behavior 241 280 (Englewood Cliff s, NJ: Prentice Hall). Petty, R. E., Cacioppo, J T., & Schumann, D (1983). Central and Peripheral Routes to Advertising Effectiveness: The Moderating Role of Involvement. Journal of Consumer Research 10, 135 146. Phillips, B. J. (200 0 ). The impact of ver bal anchoring on consumer response to image ads. Journal of Advertising 26 (1), 15 23. Phillips, B. J. (2003). Understanding Visual Metaphor in Advertising, in L. Scott and R. Batra (eds) Persuasive Imagery: A Consumer Response Perspective pp. 297 310. Ma hwah, NJ: Lawrence Erlbaum Associates. Phillips, B. J., & McQuarrie, E. F. (2003). The Development, Change, and Transformation of Rhetorical Style in Magazine Advertisements, 1954 1999. Journal of Advertising 21(4), 1 13. Philips, B.J. & McQuarrie, E.F. (2004). Beyond visual metaphor: A new typology of visual rhetoric in advertising. Marketing Theory, 4, 113 136. Philips, B.J. & McQuarrie, E.F. (2009). Impact of advertising metaphor in consumer belief. Journal of Advertising, 38, 49 61. Behavior Research Methods, Instruments, and Computers 36 (4), 717 31. Preacher, Kristopher J., Derek D. Rucker, & An Multivariate Behavioral Research 42 (1), 195 227. Ratchford, B. T. (1987). New insights about the FCB grid. Journal of Advertising Research 27, 24 38. Ressler, N. (2004). Rewards and punishment, goal directed behavior and consciousness. Neuroscience & Biobehavioral Reviews 28, 27 40. Rossiter, J. R. (2008). Defining the necessary components of creative, effective ads. Journal of Advertising 37 (4), 139 144. Rossiter, J. R. (1982). Visual imagery: applications to advertising. Advances in Consumers Research 5, 101 106.
138 Rossiter, J. R. & Percy, L. (1980). Attitude change through visual imagery in advertising. Journal of Advertising, 9, 10 16. Rossiter, J. R. & Percy, L. (1987). Advertising and Promotion Management Singapore: McGraw Hill. Rossiter, J. R. & Percy, L. (1978). Visual imaging ability as a mediator of advertising response. Advances in Consumer Research 5, 621 629. Russell, J A. & Mehrabi an A. (1977) Evidence for a Three Factor Theory of Emotions Journal of Research in Personality 11, 273 294. Russell N. Laczniak & Darrel D. Muehling (1993) The Relationship between experimental manipulations and tests of theory in an advertising message involvement context Journal of Advertising, 22(3), 59 74. Sierra, J. J., Heiser, R. S., & Torres, I. M. (2012). Text message copy in print ads: a shrewd communicative strat egy? Journal of Promotion Management, 18, 119 143. Scott, L. M. (1994) Images in Advertising: The need for a theory of visual rhetoric, Journal of Consumer Research, 21, 252 273. Scott, L. M., & Batra, R. (2003). Persuasive imagery: A consumer response perspective Mahwah, N.J: Lawrence Erlbaum Associates. Smith, R. E., Chen, J., & Yang, X. (2008). The impact of advertising creativity on the hierarchy of effects. Journal of Advertising 37(4), 47 61. Shrout P.E., & Bolger, B. (2002). Mediation in experimental and nonexperimental studies: new procedures and recommendations. Psychol. Methods 7, 422 445. Sullivan, D. M., & Ford, C. M. (2010). The alignment of measures and constructs in organizational research : The case of testing measurement models of creativity. Journal of Business and Psychology 25(3), 505 521. Sopory, P. & Dillard, J.P. (2002). The persuasive effects of metaphor a meta analysis. Human Communication Research 28 (3), 382 419. Sperber, D & Wilson, D. (1995). Relevance: Communication and Cognition Second Edition, Oxford/Cambridge: Blackwell Publishers, 2 9. Sperber, D & Wilson, D. (2004). Relevance Theory in G. Ward and L. Horn (eds.) Handbook of Pragmatics Oxford: Blackwell, 607 32. Steinbeis, N. & Koelsch,S. (2011). Affective priming effects of musical sounds on the processing of word meaning. Journal of Cognitive Neuroscience 23 (3), 604 21.
139 Vaughn, R. (1980). How advertising works: a planning model. Journal of Advertising Research 20 (5), 27 33. Ward, J. & Gaidis, W. (1990). Metaphor in Promotional Communication: A Review of Research on Metaphor Comprehension and Quality. Advances in Consumer Research 17, Marvin E. Goldberg, Gerald Gorn, and Richard W. Pollay, eds., Provo, UT: A ssociation for Consumer Research, 636 42. Wang, J. & Lee, A.Y. (2006). The role of regulatory focus in preference construction. Journal of Marketing Research 23, 28 38. Weber, R., Ritterfeld, U., & Kosygina, A. (2006). Aggression and hostility as effects of playing violent games. In P. Vorderer & J. Bryant (Eds.), Playing video games: Motives,responses, and consequence. NJ: Lawrence Erlbaum Associates, Inc. Weber, R., Ritterfeld, U., & Mathiak, K. (2006). Does playing violent video games induce aggressi on? Empirical evidence of a functional magnetic resonance imaging study. Media Psychology 8 (1), 39 60. Worth, L. T. & Mackie, D. M. (1987). Cognitive mediation of positive affect in persuasion. Social Cognition 5 (1), 76 94. Yi, Y. (1990). A critical review of consumer satisfaction. Review of Marketing 68 123. Yi, Y. (1990). Cognitive and affective e priming effects of the context for print advertisements. Journal of Advertising, 19(2) 40 48. Zaichkowsky, J. L. (1985). Measuring the involvement cons truct. Journal of Consumer Research 12, 341 352. Zaichkowsky, Judith Lynne. (1987). The emotional aspect of product involvement. Advances in Consumer Research 14 (1). 32 5. Zajonc, R. B. (1980). Feeling and thinking: Preferences need no inferences. Ame rican Psychologist 39, 117 123. Zaltman, G. (2003). How customers think: Essential insights into the mind of the market Boston, Mass: Harvard Business School Press. Zillmann, D. (2006). Dramaturgy for emotions from fictional narration. In J. Bryant & P. Vorderer (Eds.), Psychology of entertainment (pp. 215 238). Mahwah, NJ: Erlbaum.
140 BIOGRAPHICAL SKETCH Soojin Kim was born in Seoul, South Korea, and graduated from Chung Ang University, where she received her degree of Bachelor of Arts in Russian Language and Literature in February 1995. Subsequently, she earned a de gree of Master of Fine Arts in a dvertising d esign at Ewha Womans University in Seoul, Korea. After graduation, Soojin worked for J. Walter Thompson (JWT) Korea as an Advertising Creative Designer. Later, she worked for Neodigm, a marketing communication company, as a w eb d esign er, and subsequently was named online marketing m anager and an account executive (i.e., A .E.) for IT global companies such as Microsoft Korea, Intel Korea, Cisco Systems NCR HP etc. Later, while working for Neodigital Inc. as a marketing communication director, she managed advertising & marketing campaigns for IT global companies. Her ten years of professional experience include d two years as a CEO of her own company (marketing communication agency) in Seoul, Korea. After about ten years of professional experience, Soo jin earned a Master of Arts in a dvertising at the University of Texas at Austin in 2010. During her academic training in the Ph.D. program at the University of Florida, she received the assistantship fro m the College of Journalism and Communications, and taught the course, Introduction to Advertising Design and Graphics in the Department of Advertising for three semesters. For this dissertation project, she won the 2015 American Academy of Advertising Doc toral Dissertation Proposal Award. Soojin has accepted the position of a tenure track assistant professor at Louisiana State University and will teach undergraduate advertising and visual communication courses in the Fall of 2016.