The Media Sequence Effect

Published Date: 02 Nov 2017

By

LINWAN WU

A THESIS PRESENTED TO THE GRADUATE SCHOOL

OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT

OF THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF ADVERTISING

UNIVERSITY OF FLORIDA

2013

© 2013 Linwan Wu

To my family, the source of my inspiration and energy

ACKNOWLEDGMENTS

I would like to acknowledge the incredible support and guidance provided by my committee. To my chair, Dr. Lu Zheng, and committee members, Dr. Wayne Wanta and Dr. Churchill Roberts, the quality of this thesis was largely improved by your insight.

I thank my parents for their unreserved encouragement and support, which motivated me to complete my study. I also thank Dr. Hui Fang at the Huangshan College, and Dr. Huajun Lee at the Huazhong University of Science and Technology, who offered me great help in my experiment in China. Last but not least, I thank my fiancée, Li. You have supported and encouraged me with full love during my tough time in the United States.

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS 4

INTRODUCTION 8

REVIEW OF LITERATURE 11

Media Difference 11

Comprehension and Retention 11

Persuasion 11

Modality 12

Medium Control 13

Narrative Advertising 14

Transportation 15

Media Effect on Transportation 16

Processing Multiple Media 20

Repetition Effect 24

Cross-Media Effect 26

Forward Encoding 26

Image Transfer 27

Multiple Source Perception 28

Media Sequence Effect 29

Sequence Effect on Persuasion 30

Media Sequence Effect on Advertising 31

METHODS 34

Design 34

Subjects 34

Procedure 34

Materials 35

RESULTS 37

DISCUSSION 40

Discussion 40

Research Limitation 44

Conclusion 45

TRANSPORTATION SCALE 46

AD ATTITUDE SCALE 48

BRAND ATTITUDE SCALE 49

PURCHASE INTENTION SCALE 50

LIST OF REFERENCES 51

BIOGRAPHICAL SKETCH 60

Abstract of Thesis Presented to the Graduate School

of the University of Florida in Partial Fulfillment of the

Requirements for the Degree of Master of Advertising

THE MEDIA SEQUENCE EFFECT ON THE EFFECTIVENESS OF NARRATIVE ADVERTISING

By

Linwan Wu

May 2013

Chair: Lu Zheng

Major: Advertising

The current study investigates the media sequence effect on the effectiveness of narrative advertising. In particular, the current study tries to answer the question that which sequence of the combination of print and video could make narrative advertising the most effective. Ninety Chinese College students with the average age of 20 participated in a two-stage experiment, in which they read or watched a narrative advertisement in both stages. Advertising presenting sequence was manipulated, resulting in four media sequences: print-video, video-print, print-print, and video-video. Subjects’ transportation scores, attitude towards the ad, attitude towards the brand and purchase intention were measured. The result turned to be that the media sequence of print-video caused the most transportation to the narrative ad, but there were no differences of attitude and purchase intention caused by the four media sequences. Such a result implicates that, in advertising industry, print-video may be the most effective way to present narrative advertising. Research limitations and directions for future research are also mentioned.

CHAPTER 1

INTRODUCTION

A narrative advertisement is a commercial telling a story about the "experiences or consequences of product use" (Chang, 2009a, p.52). A narrative ad should include the main characteristics of a story, like protagonists, events, and story-lines, etc. These special characteristics of a narrative ad determine the fact that narrative advertising persuades consumers in a different way from what is described in the traditional persuasion models (like Elaboration Likelihood Model, Petty & Cacioppo, 1986; Heuristic-Systematic Model, Chaiken, 1980) (Green & Brock, 2000; Green, 2004; Chang, 2009a, 2009b; Zheng, 2010). In particular, narrative advertising persuades by transporting consumers into the narrative world of the commercials (Green & Brock, 2000; Green, 2004; Chang, 2009a, 2009b; Zheng, 2010).

Transportation is considered as a highly focused process in which audiences post their most attention to the narratives they read or watch (Green, 1996). Green and Brock (2000) put forward that transportation meant audiences of a narrative material lost themselves in it and felt like they got the experience similar to the characters'. Therefore, the process of transportation can be regarded as "an integrative melding of attention, imagery, and feelings" (Green & Brock, 2000, p. 701). Moreover, the consequences of transportation experience are ignoring the real word physically or psychologically, feeling strong emotions, and behavior change (Green & Brocl, 2000).

Transportation into narrative advertising has been investigated from several perspectives. Previous research has considered individual factors (like mental image ability & need for cognition, Zheng & Phelps, 2011a), and message factors (like narrative focus & vividness of product depiction, Zheng & Phelps, 2011b) that could influence audiences’ transportation into narrative advertising. In addition to individual and message factors, "factors external to the ads themselves" may also significantly influence people’s reactions to advertising (Durkin & Wakefield, 2008, p. 667). According to Green et al. (2008), one external factor, media, is an important issue in narrative processing, and they believed the nature of transportation would be understood better if media factor was taken into consideration.

However, previous research about media factor in narrative advertising largely focused on the effect of media context. For example, Wang and Calder (2006) proposed that high transportation to media content could facilitate the effectiveness of ads inserted, but such a phenomenon could only be observed when ad exposure didn't cause intrusion to people's transportation experience into media content. On the contrary, Chang (2009b) contended that people’s involvement into narrative media content would negatively influence the effectiveness of subsequent advertisements. However, media context effect is not the whole picture of investigating media factor in narrative advertising, and not so many studies have explored the impacts of different media per se on the effectiveness of narrative advertising.

In the meantime, previous studies in the media research field have confirmed that cross-media setting would make advertising more effective than single medium, leading to more favorable attitudes and more behavior intention (Edell & Keller, 1989, 1999; Dijkstra 2002; Naik & Raman, 2003; Chang & Thorson, 2004; Dijkstra, Buijtels, & Van Raaij, 2005; Bronner, 2006; Havlena, Cardarelli, & De Montigny 2007; Voorveld, Neijens, & Smit 2011). Moreover, in the advertising industry, the number of cross-media campaigns is on the increase (Darby, 2007). And actually, one of the "top trends in the branding and communications industry for 2010 are cross-media advertising campaigns" (Al Bawaba, 2010). It is believed that different media platforms in a cross-media campaign reinforce each other’s message, providing consumers with a deeper connection with the brand (Marketing Week, 2006). However, previous research about cross-media effects on advertising effectiveness mainly focused on the comparison between multiple media and single medium, and there was no enough concern about media sequence effect in cross-media campaigns, even though media sequence effect was critical for advertisers. (Voorveld, Neijens, & Smit, 2012).

Therefore, according to what has been described, the current research tries to fill the gaps mentioned above. In particular, this study aims to investigate the media sequence effect on the effectiveness of narrative advertising.

CHAPTER 2

REVIEW OF LITERATURE

Media Difference

Previous research has examined media difference from distinctive perspectives, such as comprehension, retention, and persuasion (eg. Jacoby, Hoyer, & Zimmer, 1983; Hayes, Kelly, & Mandel, 1986; McGinnies, 1965; Chaiken & Eagly, 1983). However, these studies reached no agreements.

Comprehension and Retention

When considering message comprehension and retention, print is believed by some researchers to be the most effective media type. Because the only stimulation is visual in print media, the distraction caused by vocal cues in the audiovisual counterparts can be avoided (Baron, Baron, Miller, 1973; Haaland & Venkatesan, 1968; Insko, Turnbull, & Yandell, 1974), and readers have entire control of information delivery. This provides more opportunities for them to elaborate the message content (Jacoby, Hoyer, & Zimmer, 1983). By contrast, the opposite viewpoint states that comprehension and retention should be better when people are exposed to audiovisual presentation, because such media type activates more than one sensory mode, which makes messages more comprehensible and memorable (Jacoby, Hoyer, & Zimmer, 1983; Hayes, Kelly, & Mandel, 1986; Greenfield, 1984; Pezdek, Lehrer, & Simon, 1984).

Persuasion

Similarly, different results came from various studies about the comparison of persuasion between visual media and auditory media. Wilke (1934) and Carver (1935) both supported the idea that better persuasion occurred in vocal media than print. When studying how opinions change, Hovland (1954) proposed that oral presentations seemed to be better than visual ones. However, the contradictory argument, claiming greater opinion change in written media, was also proposed by other scholars (McGinnies, 1965; Tannenbaum & Kerrick, 1954; Werner, 1978; Werner & Latane, 1976).

Although research about media difference generates contradictory results, the detection of media effect on message comprehension, retention and persuasion implies that "the medium of communication is capable of altering the nature of communication between senders and receivers" (Dijkstra, 2002, p. 11). Accordingly, comparing different media based on audiences’ responses is a "receiver-centered perspective" which focuses on examining the psychological dimensions within different media. Previous research discovered two significant dimensions in which media differ from each other: modality and medium control (Dijkstra, Buijtels, & Van Raaij, 2005).

Modality

Modality has been investigated for a long time in the persuasion research field (Liu & Stout, 1987). In the 1970s, Chaiken and Eagly (1976) found out modality (video, audio, or written modes) had a different impact on individuals’ learning and attitude. Salomon (1970) differentiated media by each medium’s unique presenting attributes which had particular psychological functions. In recent studies, modality referred to the forms of message delivery associated with human senses which were responsible for information processing (Dijkstra, Buijtels, & Van Raaij, 2005). Similarly, Moreno (2006) defined modality as "the sensory channel that is initially used by learners when they process information" (p. 149). These definitions of modality are in line with the working memory theory which claims that visual and auditory presentations are processed by independent sensory channels (Baddeley, 1992).

Examples of modality are pictures, which correspond to visual sense; audio to auditory sense; as well as video to audiovisual senses. When we adopt the perspective of modality, it is easy to understand that different media present information depend on stimulating distinctive sensory modes (Edell, 1988). For instance, the visual and auditory senses would be simultaneously motivated when people watch television. The modality perspective was supported by Edell’s (1988) discussion of nonverbal elements in advertising, which included sound, music and pictures. These nonverbal elements, which are presented by some media and absent in others, influence individuals’ information processing and mood induction (Edell, 1988). For example, pictorial stimuli are more likely to elicit better recall than their verbal counterparts (Shepard, 1967; Bower, 1972), and nonverbal elements, such as music and pictures, increase affective arousal which facilitates individuals’ information evaluation (Zajonc, 1980).

Medium Control

Medium control is defined as senders or receivers dominating the space and flow of information delivery from various media (Dijkstra, Buijtels, & Van Raaij, 2005). Based on this concept, Van Raaij (1998) came up with two categories of media: delivery and retrieval. Television is a delivery medium. The speed of presentation and the sequence of messages are both decided by information senders, or advertisers in the case of advertising (Dijkstra, Buijtels, & Van Raaij, 2005). As a consequence, audience cannot contemplate the information exhibited by TV because they are not able to keep it. Conversely, print media are labeled as retrieval. Information receivers can make their own decisions about how fast and in what order they read newspapers or magazines. Jacoby et al. (1983) believed that when consumers had control over the medium, they were more likely to digest the information they received.

Narrative Advertising

The current research focuses on the impact of media on the effectiveness of narrative advertising. A narrative ad can be defined as a commercial story presenting the "experiences or consequences of product use" (Chang, 2009a). Previous research has demonstrated that stories in narrative advertising can establish an emotional connection to audiences and those connections will make them more likely to express positive attitudes to the ads (Escalas, Moore, & Britton 2004).

Fiske (1993) proposed that when individuals were reading a narrative, new information would be associated with their previous "narrative knowledge structure" or "event prototype". These structures and prototypes are formed based on individuals’ personal life and can facilitate their comprehension of the new information from the narratives (Schank & Abelson 1995; Schank & Berman 2002). Their dependence on these prototypes will be more obvious if the characters of the narratives are not real (Wyer, Adaval, & Colcombe 2002). More to the point, previous knowledge or prototypes are not enough for people to understand the narratives. In most instances, mental models of what is going on in narratives will be established in people’s mind (Wyer 2004; Wyer & Radvansky 1999; Wyer, Adaval, & Colcombe 2002). Furthermore, if an individual is really into a story, he may imagine himself in the very event unfolded in the narrative and such a phenomenon is called mental simulation (Taylor & Schneider, 1989). Escalas (2004) proposed that mental simulation can bring up a favorable evaluation as well as a positive attitude, and this effect will be amplified if such simulation is self-related.

Transportation

Green proposed the Transportation-Imagery Model which was considered as a suitable model to interpret narrative persuasion (Green 1996, Green & Brock, 2000). Green’s (1996) conceptualization of transportation comes from Gerrig’s (1993) study. Gerrig (1993) considered readers of narratives as "travelers". These "travelers" are taken from their original world in the process of involvement in the reading materials, and when they return, their beliefs change (Gerrig, 1993).

Transportation, therefore, is considered as a highly focused process in which audience post their most attention to the unfolded stories in the narratives (Green, 1996). Green and Brock (2000) put forward that transportation meant that audiences of a narrative material lost themselves in it and felt like they had an experience similar to the characters'. Therefore, the process of transportation can be regarded as "an integrative melding of attention, imagery, and feelings" (Green & Brock, 2000, p. 701). Green and Brock (2000) also proposed some consequences of transportation, such as ignoring the real word physically or psychologically, feeling strong emotions, and behavior change.

Moreover, Transportation has been demonstrated as the underlying mechanism of narrative persuasion (Green, 1996; Green & Brock, 2000; Green, 2004) and has been applied successfully to explain the effectiveness of narrative advertising (Escalas, 2004; Chang, 2009a, 2009b; Zheng, 2010). Thus, it is reasonable to measure transportation elicited by narrative ads when exploring the effect of media on how people process narrative advertising.

Media Effect on Transportation

Involvement is a widely examined concept in communication research (Salmon, 1986; Roser, 1990; Vorderer, 1993; Wirth, 2006). Previous research explored media involvement from the perspectives of modality and medium control (Krugman, 1965; McLuhan, 1964). Krugman (1965) considered television as a low-involving medium, because television "bombards us with enough trivia" (p. 353) which might cause us to overlearn some information. In other words, audiences get no control of information delivery on television. On the contrary, McLuhan (1964) thought television should be more involving than print due to the stimulation of both visual and auditory sensory modes. Other scholars tried to reconcile such contradictory conclusions by adding message content as a moderating factor (Wright, 1974), or dividing involvement into two categories: rational involvement and emotional involvement (Chaudhuri & Buck, 1995). All the studies mentioned above admitted the roles of modality and medium control in deciding individuals’ involvement to different media.

The concept of involvement shares some similar elements with the concept of transportation. They both refer to the experience that individuals get intensely engaged into stimuli. Moreover, it is believed that there are cognitive dimension (like attention) (Batra & Ray, 1985; Roser, 1990) and affective dimension (like emotion) (Roser, 1990) in the concept of involvement. Similarly, when Green and Brock (2000) introduced the Transportation-Imagery Model, they emphasized cognitive attention and emotional involvement in individuals’ experience of transportation. However, although there are some overlaps, involvement and transportation are different. Involvement "is regarded as a motivational meta-concept that includes various forms of intense interactions with a mediated stimulus" (Wirth et al., 2007, p. 521). Such a concept stresses that people connect what they see to what they believe and care. A high involving individual would consider the stimulus as being of "intrinsic importance" (Sherif & Hovland, 1961, p. 197), or "personal meaning" (Sherif et al., 1973, p. 311). In media research, involvement also accentuates connections between media content and audiences’ values, experiences and beliefs (Krugman, 1965). Highly involving audiences tend to explain media content in their own ways (Batra & Ray, 1985; Perse, 1988). However, a highly transported individual would be deeply engaged into the narrative world without taking their own values and beliefs into consideration (Green et al., 2008). In transportation experience, audiences focus on the narratives with their full cognitive capacity and feel the emotions of the characters (Green & Brock, 2000; Green, 2004). Unlike people with high involvement, a highly transported individual wouldn’t consider themselves and the real world. In other words, "transportation refers to immersion in the narrative world rather than drawing links between the narrative world and the real world" (Green et al., 2008, p. 513). However, in spite of the difference between involvement and transportation, as we discussed before, these two concepts shared some similar elements. In addition to the fact that people would feel different levels of involvement into different media, it is reasonable to claim that different media could also give rise to different levels of transportation.

Green et al. (2008) thought vividness and effort were the two important factors that should be taken into consideration when comparing the potential of eliciting transportation between print and film. Movies are thought to be more vivid than print, so it is supposed that individuals are more likely to be transported into movie stories (Green et al., 2008). However, when scholars investigated the efforts needed to process different media, contradictory conclusions were made by previous research. On the one hand, processing print messages is believed to require more efforts, because people need to imagine pictures in their minds (Singer, 1980; Salomon, 1984). For example, in Salomon’s (1984) study, subjects were exposed to either TV or print messages, with their perceptions of self-efficiency measured. The results showed that compared to TV, print was found to demand more mental efforts, which was positively correlated with self-efficiency in print media (Salomon, 1984). On the other hand, watching films or television might need more efforts, because characters’ thoughts and feelings wouldn’t be expressed as explicitly as in the print. Therefore, an audience has to imagine characters’ thoughts and feelings from any cues in a movie or a TV show (Green et al., 2008).

So when we take the factors of vividness and effort into consideration to compare the potential of triggering transportation between print and film, several outcomes may occur. One claims that "film could increase transportation relative to print because all viewers are given a vivid, concrete set of images" (Green et al., 2008, p. 517), which is beneficial for the fluency of processing of narratives. The visual presentation in films makes it easy for the individuals who are not good at mental imagery to form scenarios in their mind; the sounds and music in films provide opportunities for audience to have deeper insight into the characters and make audiences emotionally involved into the stories (Green et al., 2008). Nevertheless, the "vivid, concrete set of images" (Green et al., 2008, p. 517) in films may hinder viewers’ own imaginations, leading to less transportation, whereas print narratives make readers free to form mental images of the unfolded stories, and allow them to deal with the narrative information on their own space. As a result, transportation is more likely to be elicited by print media rather than films. A third line of opinion about transportation in film and print suggests that these two types of media are equal in inducing transportation. They both share the factors that are crucial for transportation, like attracting plots as well as making individuals feel sympathy for the characters in the stories (Green et al., 2008; Dal Cin et al., 2004). Furthermore, Green et al. (2008) believed that it was possible that print and video elicited transportation in different ways, with video using audiovisual motion pictures and print providing more chances to elaborate information. Such a viewpoint was supported by Dal Chin et al. (2004) who found no media difference between movie and print stories. In conclusion, previous research provided little evidence to support any medium’s superiority on eliciting transportation, indicating that more research should be conducted on this issue. Since the current study concerns about the effect of media on the effectiveness of narrative transportation, we begin with the following research question:

RQ1: Which medium (print or video) would make individuals more transported to narrative ads?

This research question aims to compare the effects of different media on the effectiveness of narrative advertising. As it is mentioned above, the more individuals are transported into a narrative ad, the more likely they would change their attitudes according to the story in the ad, and the more effective the ad would be (Green & Brock, 2000; Escalas, 2004; Chang, 2009a, 2009b; Zheng, 2010). Therefore, comparing subjects’ transportation experience to the ads with the same content presented by print and video would tell us which medium (print or video) could make narrative advertising more effective.

Processing Multiple Media

The current study does not just compare the effects of different media on eliciting transportation. We are more interested in media sequence effect on how people process narrative advertising and how people experience transportation. The discussion of media sequence effects will be established on the elaboration of multiple media effects. As discussed before, modality and medium control "have an impact on the way consumers attend to, process, respond to, remember, and learn from media" (Dijkstra, 2002, p.11). In other words, modality and medium control influence people’s choices of information-processing approaches when they are exposed to different media. The following section will inspect the effect of multiple media on advertising effectiveness by adopting the perspective of information processing, expecting to lead to the discussion of media sequence effect on the effectiveness of narrative advertising.

It has been demonstrated that intense cognitive responses are associated with multiple information sources (Chang & Thorson, 2004). In one study, subjects were exposed to several similar ads, which shared the same basic arguments, but differed in the expression and sequence of presenting these arguments. After viewing each ad, subjects were asked to provide their thoughts, which were used to analyze their cognitive responses. The results of the study showed that the overall cognitive response scores were positively related to the number of ads they were exposed, indicating that multiple ads could generate more cognitive activities (McCullough & Ostrom, 1974). Harkins and Petty (1981a, 1981b, 1987) conducted a study about the effect of multiple information resources. In one such study, subjects viewed a videotape of one or three speakers presenting one or three arguments, and after that their thoughts were recorded. It turned out that the subjects who were exposed to three arguments in a video generated more thoughts about the messages than those who watched the video including only one argument (Harkins & Petty, 1981a). In a follow-up study, Harkins and Petty (1981b) added a distraction task to the similar experimental design, and found that the multiple information resources could make subjects recall more about the messages than single information resource, even in the presence of distraction, indicating that during the exposure of messages, multiple information resources made subjects think more about the messages than a single information resource. Harkins and Petty’s (1987) third study explored why multiple information resources could stimulate more cognitive thinking. In the experiments, subjects’ responses to messages were impacted by the manipulation of telling them the messages came from a dependent committee or an independent one, or informing them that the members of the committee had similar perspectives or dissimilar perspectives. The results implied that multiple information resources caused more cognitive thinking by making people perceive that the resources were independent with different perspectives, which needed to be contemplated (Harkins & Petty, 1987). In addition, Edell and Keller (1999) investigated the combination of TV and print advertising, claiming that multiple sources gave rise to more evaluation by subjects. In particular, they detected that the subjects who watched the a TV ad first, would process the subsequent print ad more carefully, and recalled more information than those who didn’t watch the TV ad (Edell & Keller, 1999). Similarly, when subjects saw a print ad first and then watched a TV ad, they would also process the TV ad to a greater extent by generating more recall than those who just watched the TV ad without former exposure to the print ad (Edell &Keller, 1999).

Accordingly, Chang and Thorson (2004) adopted the classical Elaboration Likelihood Model (ELM, Petty & Cacioppo, 1986) to compare the different routes of information-processing that individuals used when they were exposed to multiple advertising media (labeled as media synergy) versus single advertising medium (labeled as medium repetition). In line with previous research, media synergy motivates people to be more concerned about the advertisements than media repetition (Allen, Kania, & Yaeckel, 1998; Blackwell, Miniard, & Engel, 2001; Brock, Albert, & Becker, 1970; Grass & Wallace, 1969; Putrevu & Lord, 2003; Rossiter & Bellman, 1999). Moreover, individuals exposed to media synergy are more likely to contemplate the advertising messages than those exposed to media repetition (Edell & Keller, 1999; Harkins & Petty, 1981a, 1981b, 1987; McCullough & Ostrom, 1974). Such experience is labeled as the central processing route in ELM, which refers to the process in which individuals base their judgments on the elaboration of the main arguments in the advertising messages as well as the crucial characteristics of the products (Petty & Cacioppo, 1986, 1996a). On the contrary, message repetition is more likely to be processed in the peripheral processing route which means individuals formed their attitude based on peripheral cues (Petty & Cacioppo, 1986, 1996a), since repeated exposure to the same advertising message makes audiences less motivated to deliberate the arguments.

The fact that advertisements presented by multiple media are processed differently from repeated presentation by a single medium indicates the possibility that multiple media and single medium would have a different impact on advertising effectiveness. Actually, previous research about cross-media campaign has already substantiated that multiple-media campaigns are more persuasive than those using single medium, with more favorable attitudes and more behavior intention associated with cross-media campaigns (Edell & Keller, 1989, 1999; Dijkstra 2002; Naik & Raman, 2003; Chang & Thorson, 2004; Dijkstra, Buijtels, & Van Raaij, 2005; Bronner, 2006; Havlena, Cardarelli, & De Montigny 2007; Voorveld, Neijens, & Smit 2011). For example, Edell and Keller (1989) examined the synergy of TV and radio ads. Subjects were randomly assigned to six different conditions (TV-radio, radio-TV, TV-TV, radio-radio, TV, radio). The analysis of subjects’ thoughts found that a radio replay, which meant listening to a radio ad in the second exposure, generated less critical thoughts, a sign of less evaluative processing (Edell & Keller, 1989). Besides, Edell and Keller (1989) detected better recall of ads in media synergy conditions than in single medium ones. Moreover, Chang and Thorson (2004) also explored the media synergy effect by considering television and the internet. In particular, they detected that compared to medium repetition, television-web synergy made subjects pay more attention to the target ads, and made them more likely to believe the advertising messages. The analysis of the subjects’ thoughts also revealed that synergy stimulated more positive attitudes than single medium repetition (Chang & Thorson, 2004).

However, knowing the fact that cross-media can make advertising more effective than single-medium still leaves an incomplete picture of the process. According to Voorveld, Neijens and Smit (2012), cross-media effects should be further investigated by examining the effect of media sequence on advertising effectiveness. It is necessary to take media sequence into consideration when investigating cross-media effects on advertising effectiveness, because it has been substantiated that individuals’ judgments are influenced by the order of information presentation (Haugtvedt & Wegener 1994; Loda & Coleman 2005) In fact, the topic of media sequence has been more or less covered by previous research about cross-media campaigns (Ephron, 2000; Bronner, 2006; Wang & Nelson, 2006; Havlena, Kalluf, & Cardarelli, 2008). The current study will also make some contribution to this line of research by investigating media sequence effect on the effectiveness of narrative advertising. In the following sections, we start with the discussion of repetition effect. Then we consider message repetition across different media, exploring why cross-media presentation makes advertisements more effective. Afterwards, we discuss how sequence of information presentation influences persuasion, leading to our final concern of media sequence effect on advertising effectiveness.

Repetition Effect

Advertising presentation in a cross-media campaign can be regarded as a particular type of message repetition, because those ads in the same campaign "have the same valence (i.e. favorable to the product and the brand)" (Voorveld, Neijens, & Smit, 2012, p. 204). Thus, we start our discussion about media sequence effect from the repetition effect.

During the 1980s, based on several field studies and lab experiments, Tannenbaum (1985) proposed that there was "an appetite among the public for repetition and redundancy for the familiar fare of popular culture and entertainment as represented on television" (p. 232). Tannenbaum (1985) explained such phenomenon by adopting Schachter’s (1964) theory of a two-stage model. In particular, with a second exposure, the emotional reactions would not be as strong as audiences experienced in the first exposure, but audiences needed less efforts to search for explanations for their feelings, due to their familiarity with the program. "The motivation is not in the reexposure but in the reexperiencing" (Tannenbaum, 1985, p. 238).

Different theories, however, predicted different outcomes as a consequence of repetitive reading (Brewer, 1996). Uncertainty theories (Ortony, Clore, & Collins, 1988; Kintsch, 1980) suggested rereading was unattractive, because there was no uncertainty. Some theories held an opposite viewpoint by adopting an "amnesia effect," which meant when people read stories for a second time, they might forget the information they learned in a previous reading (Brewer, 1996). These theories claimed that rereading was supposed to be as enjoyable as the first reading, because there was no emotional change (Walton, 1978; De Beaugrande & Colby, 1979, Gerrig, 1989). Other scholars with the same opinion contended that rereading might be appreciated because people would forget part of the stories (Brewer, 1996) or they just didn’t realized some content due to their limited cognitive capacity (Boulton, 1975).

Therefore, although there was no agreement about what effects repetition imposed on message persuasion, one consensus was confirmed that message repetition effect did exist. In the media research field, repetition effect has been popularly investigated by using single medium. Message repetition within the same medium was believed by some scholars to facilitate message effectiveness, due to fluency feeling (Vaughn et al., 2007) as well as deeper understanding of the message (Bortolussi & Dixon, 1996). Other scholars emphasized the negative effect of repetition within the same medium, claiming that curiosity and suspense tended to be reduced (Brewer, 1996). Furthermore, in accordance with Green et al. (2008), repetition exposure should be further studied in the setting of cross-media. Such a suggestion is consistent with actual practice in the advertising industry. It was reported that in the year of 2001 nearly 85% of large advertising campaigns adopted a multiple advertising strategy (Bronner, Neijens, & Van Raaij, 2003). Therefore, message repetition presented by different media deserves more scrutiny in advertising research.

Cross-Media Effect

As we discussed before, individuals adopted different information-processing approaches to deal with synergy of different media versus several exposures of the same single medium, which contributed to different message effectiveness. Media synergy seems to make communication more effective. However, why could cross-media setting facilitate advertising effectiveness? Voorveld, Neijens, and Smit (2011) came up with three psychological processes that might occur when consumers processed advertisements in a cross-media setting: forward encoding, image transfer, and multiple source perception. We believe this deeper insight of information processing would explain why multiple media can positively influence advertising effectiveness.

Forward Encoding

Forward encoding is defined as the process in which the first ad exposure motivates an audience to process the second ad (Voorveld, Neijens, & Smit, 2011). By arousing the audience’s interest and attention (Bronner, Neijens, & Van Raaij 2003; Edell & Keller 1989, 1999; Dijkstra 2002), the initial exhibition of a certain advertisement would lead to deeper processing and easier encoding of a latter one (Dijkstra 2002). In other words, during the experience of forward encoding, the first ad makes individuals more willing and more able to process the second ad.

Voorveld, Neijens, and Smit (2011) also explained the process of forward encoding from the perspective of memory storage. Actually, during the exposure of the first ad, part of the ad would be stored in people’s memory (Keller, 1987). Thus later, when people deal with a similar ad in a different medium, the memory which was previously stored would be activated to help process the second ad in more depth (Voorveld, Neijens, & Smit, 2011).

Furthermore, it should be emphasized that forward encoding is more likely to occur within the cross-media circumstance. According to Dijkstra (2002), people will not be encouraged to process an exact duplicate of a certain ad. Such a phenomenon can be explained by the Differential Attention Theory (Unnava & Burnkrant, 1991), which supposes that both motivation and interest would be reduced when people deal with the same message repeatedly. Notwithstanding, the use of different media can avoid the problem of pure repetition by delivering the same information with different presentation forms (Voorveld, Neijens, & Smit, 2011).

Some previous research has mentioned forward encoding in cross-media advertising campaigns. By asking subjects to report their interest in the second ad, Dijkstra (2002) detected higher forward encoding in the cross-media condition. Edell and Keller (1989, 1999) also discovered the appearance of forward encoding in some cross-media conditions in their studies, by checking the results of thought listing. Voorveld, Neijens, and Smit (2011) listed forward encoding as one of the approaches of information-processing that people would adopt in a cross-media campaign, and stated that forward encoding enhanced campaign results.

Image Transfer

Image transfer happens when people are processing the second ad after seeing the first one. It refers to audiences’ experience of mentally reiterating what they have watched in the first ad when they are exposed to the second one (Voorveld, Neijens, & Smit, 2011). The Encoding Specificity Principle claimed that individuals would have better recall if they were exposed to the clues which were used initially to store information into their memory (Tulving & Thomson, 1973). By adopting this principle, Dijkstra (2002) interpreted image transfer as the process in which "the elements in the second ad may function as retrieval cues to the ad memory trace from the first exposure" (p.66), due to the similarity of the two ads.

Moreover, in the same way as forward encoding, image transfer is also believed to be easily triggered in a cross-media campaign (Edell & Keller 1989), but not single-medium. It is said that message repetition presented by a single medium makes it unnecessary to get information from people’s memory about the first exposure, because the exactly same ad can be directly processed with no difficulty (Dijkstra, 2002). Previous research also confirmed the appearance of image transfer in a cross-media campaign (Dijkstra, 2002; Edell & Keller, 1989), and its contributing role of facilitating advertising effectiveness (Voorveld, Neijens, & Smit, 2011).

Multiple Source Perception

The so-called multiple source perception talks about the fact that people are more likely to trust the message delivered by multiple independent sources (Harkins & Petty, 1981a, 1987; Bronner , Neijens, & Van Raaij, 2003; Gotlieb & Sarel, 1991). Such a phenomenon can be illustrated by the Economic Signaling Theory (Nelson, 1974). In the context of advertising based on this theory, repeated advertising exposure indicates high expenditures, which in turn implies more credible brand and product quality (Kirmani, 1997). Moreover, Voorveld, Neijens, and Smit (2011) further contended that a cross-media campaign would magnify such an effect, because multiple media were considered more costly than using one single medium.

When individuals are exposed to advertisements presented by multiple media, they would go through the processes of forward encoding, image transfer and multiple sources perception. Therefore, influenced by these three psychological processes, individuals would be more likely to express favorable responses to the advertisements. Consequently, it is reasonable to claim that a cross-media effect does exist, and has positive influence on advertising effectiveness. As discussed before, the effectiveness of narrative advertising could be measured by an audience’s transportation experience (Green, 1996; Green & Brock, 2000; Green, 2004; Chang, 2009a, 2009b; Zheng, 2010). Therefore, we assume that the cross-media effect on narrative advertising can be detected by investigating the impact of multiple media on eliciting transportation:

H1: Subjects would be more transported into the second narrative ad than the first one.

This hypothesis is proposed to demonstrate that cross-media setting would make narrative advertising more effective. The cross-media setting is created by providing subjects with a second exposure to the ads. If subjects are more transported to the ads in the second exposure than they were in the first one, it is reasonable to believe that cross-media setting would positively influence advertising effectiveness.

Media Sequence Effect

A deeper insight into a cross-media effect can be obtained by investigating the media sequence effect, which refers to the impact of different orders of multiple media on the effectiveness of the advertisements presented by these media (Voorveld, Neijens, & Smit, 2012). Apparently, a media sequence effect can only happen in a cross-media setting because of the premise of multiple media. It has been substantiated that cross-media can make advertising more effective, but what remains unclear is whether the sequence of media presentation can make a difference (Voorveld, Neijens, & Smit, 2012).

Sequence Effect on Persuasion

The investigation of a media sequence effect can benefit from research in other fields about information persuasion being influenced by presentation sequence (Haugtvedt & Wegener, 1994; Loda & Coleman, 2005). According to Hovland, Campbell, and Brock (1957), message order tends to have two types of effects on how individuals process two opposite messages within a similar topic. One is the primacy effect which happens when the initial message dominates people’s evaluation of the topic. The other is the recency effect, which means judgments are produced mainly based on the last message.

Previous research about this issue mainly focused on the conditions in which either primacy or recency effect would take place. Lana (1961) found that people’s familiarity to the messages played a part, with high familiarity causing primacy effect, and low familiarity leading to a recency effect. Later, Lana (1963b) also discovered that when audiences were dealing with something controversial, the primacy effect occurred. Furthermore, individual interest in the message’ topics was also influential (Lana, 1963a). A highly interested subject was more likely to adopt the primacy effect, while his less interested counterpart was more likely to show the recency effect. Haugtvedt and Wegener (1994) noted that a primacy effect would happen when people expressed strong attitudes about first message and pondered it further, suggesting that they were motivated to process the first message in a central way, which refers to the process of individuals basing their judgments on the elaboration of the main arguments in the messages (Petty & Cacioppo, 1986, 1996a); while for people who considered the first message lightly, the recency effect was more likely to happen (Haugtvedt & Wegener, 1994). Although the discussion about a primacy/recency effects can be difficult to apply to the research on media sequence effects, due to the fact that primacy/recency effects focus on contradictory messages, while in a cross-media setting, messages are always in tune with each other (Voorveld, Neijens, & Smit, 2012), the proposal of a primacy/recency effects at least proves that sequence does influence communication effectiveness.

Media Sequence Effect on Advertising

It has already been demonstrated that a cross media campaign can facilitate advertising effectiveness by making consumers go through forward encoding, image transfer and multiple source perception during their exposure to advertisements (Voorveld, Neijens, & Smit, 2011). Research in other fields also proved that presentation sequence can influence the effectiveness of communication (Lana, 1961, 1963a, 1963b; Haugtvedt & Wegener, 1994). Therefore, based on these two lines of research, it is logical to assume that the presentation sequence of advertisements delivered by different media would influence individuals’ responses to the ads. In other words, a media sequence effect on advertising effectiveness does exist.

So far, media sequence effects have received little attention in the advertising research field (Voorveld, Neijens & Smit, 2012). Most studies about multiple media emphasize the comparison between the effectiveness of a cross-media campaign and that of a single medium campaign (Edell & Keller, 1989, 1999; Dijkstra, 2002; Chang & Thorson, 2004). In fact, based on our knowledge, only one study, conducted by Voorveld, Neijens and Smit (2012), focuses on media sequence effects on advertising effectiveness.

Voorveld, Neijens and Smit (2012) adopted the concept of product involvement to investigate the sequence effect of the combination of television and website advertising (TV-web vs web-TV). Specifically, they believed that a TV-website sequence would be more effective in presenting a low involvement product. Their rationale was that TV advertising would be better at attracting consumers to low involvement products (Krugman, 1965; Buchholz & Smith, 1991; Dijkstra, Buijtels, & Van Raaij, 2005). Therefore, a TV ad would evoke consumers’ interests and motivations to process the subsequent web ads (Yoo & Stout, 2001; Levy & Nebenzahl, 2008; Liu & Shrum, 2009). If such a sequence was reversed, which means the web ads would be presented first followed by the TV ads, the former would fail to trigger enough interest or attention, because consumers would be less involved in the products. As a result, they wouldn’t be motivated to process the later TV ads. On the contrary, it is posited that both TV-web and web-TV sequences are equally effective for high involvement products. When consumers are highly involved, they are motivated to get any information that is available (Beaty & Smith, 1987; Buchholz & Smith, 1991; Volk & Kraft 2005), so "they might appreciate both websites and TV commercials" (Voorveld, Neijens and Smit, 2012, p. 208).

The focus of the current research is whether the sequence of combination of print and video would influence the effectiveness of narrative advertising. Voorveld, Neijens and Smit’s (2012) research provides a promising starting point to investigate media sequence effect by considering two media types: television and internet, as well as adding product involvement as a moderating factor. However, the result of Voorveld, Neijens and Smit’s (2012) research should be carefully scrutinized. The website ads used in the research just included texts and photos, making the Internet advertising more like print ads, ignoring the fact that the internet platform could contain almost every other medium. Furthermore, they also adopted the theories of print media processing to explain how website communication worked (Voorveld, Neijens, & Smit’s 2012). Although Voorveld, Neijens and Smit (2012) considered the internet medium, their study was still based on previous research about traditional media.

Building upon the work of Voorveld, Neijens and Smit’s (2012) work, the current study examines the media sequence effect of print and video on the effectiveness of narrative advertising:

RQ2: What combination of print and video media (print-video; video-print; print-print; video-video) would cause the most transportation to the second narrative ad?

RQ3: What combination of print and video media (print-video; video-print; print-print; video-video) would cause the most positive attitude toward the ad (RQ3a), the most positive attitude toward the brand (RQ3b), and the most purchase intentions (RQ3c).

The last two research questions are proposed to test media sequence effect on the effectiveness of narrative advertising. In particular, we expect to find out which one of the four sequences (print-video; video-print; print-print; video-video) can make narrative advertising the most effective. We first test the effectiveness of narrative advertising by measuring transportation, because it has been demonstrated that transportation is positively related to the effectiveness of narrative advertising (Green & Brock, 2000; Escalas, 2004; Chang, 2009a, 2009b; Zheng, 2010). Then we also test the effectiveness of narrative advertising by measuring subjects’ attitude towards the ad, attitude towards the brand, purchase intention. These three variables were popularly used in previous research about advertising effectiveness.

CHAPTER 3

METHODS

Design

The experiment was separated into two stages. There was a three-day interval between the two stages. Subjects attended both stages. In each stage, they were randomly assigned to read a print narrative ad or watch the same ad in video version, resulting in four conditions: print-video, video-print, print-print, and video-video. The print ad was made into hard copies and was handed out to subjects during the experiment. The video ad was presented by projectors.

Subjects

Ninety first year undergraduate students at a southeastern Chinese university were recruited (37 males, 53 females, M age = 20). They were told that the experiment was about individuals’ responses to some advertisements. After the first stage 20-minute experiment, subjects signed up for the second stage of the experiment three days later. They received extra course credit for participating in the experiment.

Procedure

In each stage, subjects in a certain condition were tested at the same time. They were assembled in a small classroom with a projector. At the beginning, they were given a brief explanation of the experiment, including what they would do in the experiment, what rights they had during the experiment, how they could contact the researchers if they wanted more information about the experiment, and to whom they could complain if they thought their rights were violated during the experiment. After that, subjects either watched a video narrative ad named "why people alive", or read the corresponding print version of this ad. The video ad was presented by a projector, and the print ad was handed out as hard copies. After reading or watching the ad, subjects completed a questionnaire (described below).

Materials

Why people alive. This ad is selected from hundreds of ads in several video websites in China, and lasts for three minutes. This ad is aimed to present the brand image of a bank in Taiwan (Ta Chong Bank). This ad depicts five old men in Taiwan riding their motorcycles traveling around the whole island trying to find out the essential meaning of being alive; all of them suffer some serious diseases and the death of a close friend prompts them to make this trip. The connection of this story to the bank is the name of the bank. "Ta Chong" means "people" in Chinese. The characters in the story are all ordinary people, but their behaviors are unique. Accordingly, the ad tries to communicate the idea that Ta Chong Bank is the bank for ordinary people, but ordinary people can be extraordinary. The print version of this ad was made by the researchers. We selected the six most important scenes, which could present the outline of the story, from this ad. We snapshot the pictures from the video, put them in the print ad, and included introductory text for the story.

Demographics. Subjects gave their age, gender, and major.

Transportation. Subjects completed a 12-item transportation scale (Green & Brock, 2000), consisting 11 general questions measuring cognitive attention, mental imagery, and emotional involvement, and one specific questions about subjects’ mental images of the main characters (old men riding motorcycles). Example items are "While I was reading/watching the narrative, I could easily picture the events in it taking place", "While I was reading/watching the narrative, activity going on in the room around me was on my mind", and "When I read or watched the story, the vivid images of those old men came into my mind." All the items were answered by choosing from "1=strongly disagree" to "7=strongly agree." The scale was translated into Chinese. Three coders individually translated the Chinese version back to English. The retranslated versions were almost completely identical to the original one.

Ad attitude, brand attitude, and purchase intention. Subjects also completed the scales measuring their attitude towards the ad, attitude towards the brand, and purchase intention. Although lots of previous research measuring advertising effectiveness used the variables of attitude towards the ad, attitude towards the brand, and purchase intention, there was no consensus about which items were the most appropriate to be include in the scales. In the current research, we reviewed nearly 100 articles in which the three variable were measured, and came up with our own scales of attitude towards the ad, attitude towards the brand, and purchase intention. Attitude towards the ad was measured by six bipolar adjectives: good/bad; positive/negative; favorable/unfavorable; like/dislike; pleasant/unpleasant; and informative/uninformative. Attitude towards the brand was measured by five bipolar adjectives: good/bad; pleasant/unpleasant; favorable/unfavorable; positive/negative; and appealing/unappealing. Purchase intention was measured by the following questions: I am more likely to remember the brand name after seeing the ad; I would recommend this bank to my friends who are interested in bank service; I am more likely to purchase the product after seeing the ad; I am more likely to use a free trial of this service after seeing this ad; I am more likely to request additional information of the service after seeing this ad. All of the items were measured along a seven-point scale and were translated into Chinese. Three coders individually translated the Chinese version back to English. The retranslated versions were again almost identical to the original one.

CHAPTER 4

RESULTS

First stage transportation. We compared the first stage transportation scores between the print version and the video version. The results of Independent-Sample T Test showed that, subjects’ transportation scores to video version (M=4.97, SD=0.55) were significantly larger than those to print version, (M=4.52, SD=0.60), t (88) =3.66, p< .05, d=0.77. Therefore, the answer to RQ1 was that video made individuals more transported into the narrative advertising than print did.

Second stage transportation. First we conducted Paired-Sample T Test to compare the first stage transportation scores and the second stage transportation scores. The results showed that the second stage transportation scores (M=4.93, SD=0.53) were significantly larger than the first stage transportation scores (M=4.36, SD=0.58) only in print-video condition, t (25) = 5.16, p< .05, d=1.01. However, there were no significant differences between first and second stage transportation for the print-print condition (t (17) = 2.05, p> .05, d=0.48) and video-video condition (t (23) = 1.96, p> .05, d=0.40). And the second stage transportation (M=4.37, SD=0.49) was significantly lower than the first stage transportation (M=4.91, SD=0.55) in the video-print condition, t (21) = 4.47, p< .05, d=0.95. Thus, H1 (greater transportation to the second ad exposure) was only supported when the narrative ad was presented in the sequence of print-video, but not in other sequences.

Then we conducted an ANOVA on second stage transportation with media sequence (print-video, video-print, print-print, and video-video) as the independent variable. There was a significant main effect of second stage transportation (F (3, 86) = 5.33, p< .05, = .16). The second stage transportation in print-video condition (M=4.93, SD=0.53) was significantly larger than that in video-print condition (M=4.37, SD=0.49) and print-print condition (M=4.49, SD=0.62), but was insignificantly larger than that in video-video condition (M=4.81, SD=0.55). In conclusion, the answer to RQ2 was that the narrative ad presented in print-video sequence caused the most transportation among the four sequences (print-video, video-print, print-print, and video-video).

Attitude towards the ad. We also conducted an ANOVA on subjects’ attitude towards the second stage ad with media sequence (print-video, video-print, print-print, and video-video) as the independent variable. However, we detected no significant main effect of subjects’ attitude towards the second stage ad, F (3, 86) = 1.52, p> .05, = .05. Therefore, the answer to RQ3a was that there were no differences of attitude toward the ad caused by different media sequences.

Attitude towards the brand. We also conducted an ANOVA on subjects’ attitude towards the brand in the second stage with media sequence (print-video, video-print, print-print, and video-video) as the independent variable. We detected no significant main effect of subjects’ attitude towards the brand in the second stage, F (3, 86) = 1.51, p> .05, = .05. Therefore, the answer to RQ3b was that there were no differences of attitude toward the brand caused by different media sequences.

Purchase intention. We also conducted an ANOVA on subjects’ purchase intention in the second stage with media sequence (print-video, video-print, print-print, and video-video) as the independent variable. We detected no significant main effect of subjects’ purchase intention in the second stage, F (3, 86) = 2.89, p> .05, = .01. Therefore, the answer to RQ3c was that there were no differences of purchase intention caused by different media sequences.

CHAPTER 5

DISCUSSION

Discussion

The first research question is to explore which medium (print or video) would make individuals more transported to narrative advertising. According to the results of our experiment, subjects who watched the narrative ad in video version were more transported into the ad than those who read the print version. As we have discussed before, a video not only provides audiences with motion pictures, but also lets them hear the sound and music. Compared to print media, video media stimulate multiple sensory modes of human beings, making the narratives more vivid than those written in words do. Furthermore, a video presents stories directly in front of audiences. Individuals can see characters, backgrounds, and almost everything in the stories. They don’t have to take effort to build all the stories elements in their minds. It is much easier for audiences to get engaged into a story presented in video. In the research conducted by Green and her colleagues (Green et al., 2008), they emphasized individual difference in need for cognition. They claimed that people low in need for cognition would be more transported into film narratives than into print ones, because a film was "perceived to be an ‘easier’ medium" and a print might "require more mental effort" (Green et al., 2008, p. 530). In the contrary, those who were high in need for cognition would be more transported into prints than into films. However, previous research in psychology field has claimed that people are "cognitive misers", which means people tend to not make too much cognitive effort to deal with any mental task they encounter (Fiske & Taylor, 1991). Accordingly, it is reasonable to assume that the division of low and high need for cognition in Green’s research (Green et al., 2008) is not absolute. In a more general way, people are reluctant to take effort to process media content, making them like video media more than print media. In addition, in real life, few people are willing to take full effort to process advertisements. When exposed to ads, people are more likely to be relaxed, and they wouldn’t take ads seriously. Thus, in our experiment, when subjects were told that they were going to watch or read an ad, they would probably get prepared to take little effort to process the stimuli, because advertisements were ordinarily for entertainment purpose. In this case, all the subjects in our experiment were like those low in need for cognition in Green’s study to some extent, which explained the seemly contradictory results of the two studies.

We also hypothesized that multiple media could facilitate transportation into narrative advertising. However, this hypothesis was only supported with the media sequence of print-video. This result was in parallel with the answer to our second research question: narrative advertising presented in print-video sequence caused the most transportation among the four sequences (print-video, video-print, print-print, and video-video). Green and her colleagues gained the similar result that "reading followed by watching is the most effective way of inducing transportation into a narrative" (Green et al., 2008, p. 530). They believed such a combination of media was the best way in making individuals transported into the narrative world. In particular, when individuals first read the narrative, they can control their own reading pace, and have enough time to mentally establish the narrative world in their minds. Then "on the second exposure, they are able to relax and see that narrative world created for them on the screen" (Green et al., 2008, p. 530). However, when subjects watch the narrative in video first, they would accept the images created for them and store these images into their memory. The next time when they read the same story in print, they can easily retrieve the pictures from their memory and don’t have to mentally imagine the story in their mind, so they would be less transported. We have already talked about forward encoding (the process in which the first ad exposure motivates an audience to process the second ad) and image transfer (audiences’ experience of mentally reiterating what they have watched in the first ad when they are exposed to the second one) in