From One Medium to Another: Continuing the Narrative World on Twitter and Facebook

Using narrative processing approach and uses and gratification theory, this study examined the motivations behind audience members desire to use social networking sites (SNS) Facebook and Twitter to: 1)connect with entertainment shows, 2) engage with characters on those shows, and 3)to engage with the actors that play those characters.After exposing participants to an episode of a show to gauge their engagement, they were asked to complete a questionnaire. The results suggest that levels of transportation predict likelihood of connecting with the show on social networking platforms. The levels of identification and parasocial interaction experienced during viewing predict the likelihood of following the character and the actor that played that role. Parasocial interaction with a character mediated the relationship between source attraction and connecting with the character and the actor on SNS. Theoretically, this research extends uses and gratification and narrative processing in the area of social networking research. Further theoretical and practical implications are discussed.


Uses and Gratifications andNarrative Involvement
Uses and gratification perspective can be applied to this research to understand motivations behind connecting with someone via social media. The traditional view of uses and gratification research posits that audiences are active media users who are free to choose from various forms of mass media to gratify certain social and psychological needs (Katz, Blumler, Gurevitch, 1973;Rubin, 2009). With the advent of so many media choices, this perspective becomes more relevant as today's media user is free to select from a multitude of media options (Rubin, 2009;Ruggiero, 2000). Naturally, there has been a surge in studies trying to understand motivations behind video game use (Reijmersdal, Jansz, Peters, &Noort, 2013), internet use (Papacharissi& Rubin, 2000) and use of SNS platforms (Pai& Arnott, 2013) that are framed under the uses and gratification perspective. We argue and test in this research that various forms of narrative involvement, elaborated in the next section, creates a need for connection that extends past viewing that is gratified through extended involvement with television shows viavarious social media platforms.

Narrative Involvement
Transportation or absorption in a narrative is a cognitively engaging activity that blurs the boundaries between narrative and real world (Green & Brock, 2000. Transportation is a "distinct mental process, an integrative melding of attention, imagery, and feelings, which may mediate the impact of narratives on beliefs" (Green & Brock, 2000, p. 701;2002, p. 324). Consequently, being transported means that people temporarily forget some aspects of reality and become immersed in the world of narrative (Green & Brock, 2000. That is, the message recipient becomes "cognitively and affectively engaged in the narrative" (Slater, 2002, p. 179). Transportation in a narrative could persuade message recipients since they are less critical of the narrative content and therefore less likely to counter-argue the story consistentclaims (Green & Brock, 2002;Moyer-Guse, 2008;Slater &Rouner,2002); are more likely to perceive the events in the narrative to be real, which can influence attitudes due to mimicry of direct experience (Green & Brock, 2002); and could develop affective feelings toward the characters which in turn may impact beliefs (Green & Brock, 2000;. Although research has provided enough evidence to suggest that narrative could still influence the message recipient once the transportation experience has ended (Bilandzic&Busselle, 2008;Green & Clark, 2012), what we don't know is how participants Online Journal of Communication and Media Technologies Volume: 6 -Issue: 4 October -2016 fulfill the need to maintain connection to the narrative, if they so desire, after the viewing is terminated.While repetitive watching is one option, social media provides an avenue for social involvement where people can still feel connected to the show without leaving their reality by engaging in conversations with fellow aficionados of the program.
Previous research suggests that some of the reasons for engaging in social media by undertaking activities such as sharing information, "liking" something, or "following" someone has to do with intrinsic and/or image-related utilities (Toubia & Stephen, 2013).
Intrinsic utility refers to engaging in social media activity such as sharing content that may satisfy intrinsic motivation and amassing followers may satisfy image-related utility that encompasses user's identity and sense of self-worth. Contextualizing this information to the current research, a viewer may decide to follow a particular show because they are so cognitively engaged in the show that they may want to continue the hedonistic experience by interacting with like-minded individuals. Alternatively, the viewers of a particular show may want to communicate a message to their social circle that they are fans of the program by clicking like or following the show on SNS.Some of the other motivations behind activities such assharing news (e.g. about a particular character) on SNS platforms may be to stir conversation with fellow users and to maintain a level of personal connection (Greer & Ferguson, 2011). Specifically, Wohn and Eun-Kyung (2011) found four prevailing message types when examining Tweets posted during television viewing, which include: 1) Attentionseeking (self-reference), 2) Information (program-reference), 3) Emotion (self-reference), and 4) Opinion (program-reference). Indeed, social media based brand communities are considered a way for individuals with similar interest to virtually congregate for various purposes such as self-expression and participation, which in turn, further intensifies the investment one experiences with a brandin the present context, a television show or it's character/actor (Habibi, Laroche, Richard, 2014).
Based on this we propose that: ended.Identification involves taking on the identity of the character by merging one's self with the character during the viewing experience (Cohen, 2001). In conceptualizing identification with media characters, Cohen (2001) notes: "When identifying with a character, an audience member imagines him-or herself being that character and replaces his or her personal identity and role as audience member with the identity and role of the character within the text" (p. 251). Identification has four dimensions-empathy, cognitive aspect, motivational aspect, absorption-and includes adopting both cognitive as well as the affective identity of the character and a complete immersion in the text (Cohen, 2001). While the dimension of absorption is common to identification and transportation, these are distinct concepts as identification goes beyond the involvement with the narrative-the primary focus of transportation-to the involvement with the characters in the narrative (Moyer-Guse,

2008).
While identification occurs during viewing and is perceived to end after the viewing session has concluded, parasocial interaction (PSI)-defined as considering the characters in the narrative as friends and developing a bond with them without ever meeting the characters in person-extends beyond the viewing experience (Cohen, 2006;Giles, 2002;Horton &Wohl, 1956). PSI is defined as an "extension of normal social activity by considering shared and different qualities of social and parasocial encounters" (Giles, 2002 p. 298). Viewers who form parasocial interaction with media characters are generally aware that these characters are not real (Cohen, 2006;Giles, 2002;Horton & Wohl, 1956;Moyer-Guse, 2008). To experience parasocial interaction, neither similarity, nor role-takingof the media character, is an absolute must although it might aid the "relationship" (Rubin & McHugh, 1987;Moyer-Guse, 2008). Because parasocial interaction extends beyond the viewing experience, message recipients could seek out the characters with whom they experience PSI, and the actor that played the role of that character, to gratify the need to connect. In other words, while identification is supposed to end with the viewing experience, PSI may continue even after the show has ended.Since SNS provide an avenue to the viewers to continue the "interaction" with the characters and actors that play those characters beyond the viewing situation and the recent calls to investigate how these processes unfold online (Cohen, 2014)  Liking-also referred to as social attraction to a character-refers to positive evaluationsleading to a generally favorable attitude toward them (Cohen, 2001;Giles, 2002; Moyer-Guse, 2008). Liking of media characters was a part of interpersonal attraction to physically thin media personalities and positively predicted eating disorders in one study (Harrison, 1997). Social attractionto the actress,Anne Hathaway, successfully predicted parasocial interaction with the character played by Hathaway, even when she was portrayed in a stigmatized role (Chung & Moyer-Guse, 2012;Moyer-Guse, 2008). Perhaps the level of attractiveness that a viewer feels toward a character would influence the level of engagement.
In otherwords, if following an actor and a character on SNS are ways to engage in parasocial interaction, then this should be influenced by how much the character is liked in the first place. Indeed past research has demonstrated a positive relationship between PSI and social attractiveness (Rubin & McHug, 1987). Therefore, it stands to reason that viewers who experience stronger levels of attraction would be more likely to follow the actor and character, which would be mediated via PSI as indicated in figure 1and 2. More formally, H3: PSI with the character would mediate the effect of social attraction on a) following the character on SNS and b) following the actor on SNS.

Method Participants and Procedure
Participants were 168 students recruited from various communication courses at a Midwestern university in exchange for course credit. The average age of the participants was 20years (SD = 2.8), and 80% (n = 107) were females. Given, the sample was biased in terms of gender, it is added as a control variable in all the analyses to alleviate any effects gender may have on the results. The study was approved by the institution's review board for human subjects research. The data was collected in a large hall-type setting during a regular class period, after seeking permission from the instructor. Participants were asked to switch off their cell phones and laptops while watching a 30-minute segment from the show Mad Men(Season 1, episode 1).The reason participants were exposed to a clip before watching the show was because we wanted to understand levels of engagement, i.e., transportation in and Online Journal of Communication and Media Technologies Volume: 6 -Issue: 4 October -2016 identification with the show. Hence, instead of implementing a survey only design, this current approach was used where participants were first exposed to the narrative following which theyresponded to a survey. They were instructed not to talk to anyone during the process. The video focused on Don Draper, the main protagonist from the show Mad Men, who is shown contemplating creative idea for a campaign for a cigarette company as the client. Identification. Identification with the protagonist Don Draper was measured using twelve items on a 7-point (1-strongly disagree to 7-strongly agree) scale (Cohen, 2001; α = .90; M = 4.16, SD = 1.07). A sample item included, "I think I have a good understanding of Don".

Measures
Transportation. Engagement with the narrative was measured with the 11-item scale developed by Green and Brock (2000). Response options were 1 (strongly disagree) to 7 (strongly agree) (α = .76; M = 4.34, SD = .87). A sample item is, "I wanted to learn how the story ended".

Discussion
In this study, factors that influence people's decision to connect with a show and the celebrity in that show on social networking platforms were explored. As hypothesized, narrative involvement was a strong predictor of maintaining a need for connection with the drama, the character, and the actor even after the viewing experience ended. Specifically, transportation experienced while viewing was predictive of intention to connect with the show after the viewing experience had ended. Parasocial interaction during viewing, and not identification, was predictive of connecting with both the character and the actor via SNS post viewing. PSI also mediated the effect of source attraction on post-viewing involvement with the character and the actor via connecting with them on SNS. Overall, the findings of the study suggest that involvement in a narrative is a strong predictor of connecting with the narrative world once the viewing experience has ended. In other words, participants seem to maintain this need for connection, which is gratified via participation in SNS. These findings are provocative at many levels given recent research finding that suggests a direct correlation between Twitter activity and television ratings (Nielsen, 2013). Though, no causal inferences could be drawn, this study found that increased activity on Twitter was directly associated with an increase in total rating points (TRP)for the show and, conversely, the show that had higher ratings evoked a higher level of engagement on Twitter. Yet another study found that interaction on SM during program viewing enhanced engagement amongst message recipients (Pynta et al., 2014). The current study adds to this research by suggesting involvement with the narrative and its characters as one of the processes driving this effect.
The findings of the study have several implications. First, these findings suggest that in order to maintain an engaged audience, producers and the networks on which these programs are aired need to make sure that social media channels involving the show and the protagonists are created in a manner that keeps the audience engaged (Pynta et al., 2012). Further, such online engagement with the narrative world may reinforce the desire for a continuous subsequent interaction with the narrative given media selection is influenced by attitudinal outcomes shaped by exposure to media in the first place (Slater, 2007(Slater, , 2014.This may be Online Journal of Communication and Media Technologies Volume: 6 -Issue: 4 October -2016 one of the explanations behind the relationship foundbetween Twitter activity and TV ratings (Nielsen, 2013;Pynta et al., 2014). Future research should examine this possibility.
Furthermore, perhaps message recipients, in the first place, seek out specific media due to an affinity towards an actor that they may have been exposed to in other places as suggested in the reinforcing spirals model that emphasizes that " the effects of media exposure are likely to reinforce the tendencies, preferences, and values that predict media choice in the first place" (Slater, 2007, pp. 298-299). Indeed, a recent research found that social attraction to the actress Anne Hathawaypredicted transportation into the narrative and parasocial interaction with the character in the storyline even when the character was portrayed as stigmatized (Chung & Moyer-Guse, 2012). Therefore, maintenance of online communities for fans of actors and characters could help in driving ratings and traffic, a possibility that needs more investigation.
The findings of this study are also important from a perspective of entertainment education that argues that deliberate and purposive embedding of educational messages in the form of among message recipients (Moyer-Guse, Jain, Chung, 2012).Future research should examine the influence of connecting with a narrative character on SNS on message recipient's longterm attitudes. Perhaps observational learning from the narrative and continued motivation provided via engagement in related online networks may drive people to adopt positive attitudes and expectancies-a possibility that remains to be explored (Bandura, 1986(Bandura, , 2002. Another implication of this study is that continued engagement post viewing could lead to attitudinal conformity. Cameron and Geidner's (2014) work suggests that dual viewing, that is engaging simultaneously with content on traditional medium such as TV and SNS such as Twitter, leads to opinion conformity in line with the majority opinion expressed on Twitter.
The authors in this study did not measure engagement variables but the mere effect of seeing a majority positive or negative opinion on Twitter influenced viewer's perceptions. Given, narrative engagement is, to some extent, an investment of self in the message and its characters, another possibility for future research may examine the subsequent effects on opinion conformity should be reified for those message recipients who are highly invested and engaged in the story and its characters. This again alludes to the potential of using these characters and shows for entertainment-education purposes.
At a theoretical level, this study raises the questions regarding the concept of parasocial breakup (Cohen, 2003;2004;Lather and Moyer-Guse, 2011) that suggests the negative affect experienced when a show ends or a favorite character in a program dies or is taken off the air.
Given the pervasiveness of social media and the presence of television shows and characters on these platforms, perhaps viewers don't feel as much distress when their favorite show is taken off the air, as suggested in previous research, because they know that continued interaction is a possibility. Maybe when the show or the character stops adding content to their social media platform or exhibits lack of engagement that is when the real breakup occurs? And even after that, the engagement provided via online community may act as a buffer against the distress experienced as a result of the breakup. All theses possibilities remain to be tested.
This research, though promising, was not without its limitations. Though utmost care was exercised during the study, errors in survey research due to subject-related issues such as fatigue and self-report cannot be ruled out. Further, behavioral intention to connect with the Online Journal of Communication and Media Technologies Volume: 6 -Issue: 4 October -2016 show and celebrities was measured. Though intention is a good indicator of behavior, assessment of actual behavior would have provided more confidence in findings. However, because it is not feasible to see if participants connected with the show without being given access to their social networks (e.g. by being participant's friend) a conscious decision to use intention measure was used as a proxy so as to not violate participants' privacy. Another limitation of the study is that the data is collected from a convenient sample of students and uses one viewing of a particular show. Therefore, future research should repeat the study with a more diverse sample and multipleshows. However, given just one show and one viewing had significant impact on outcome variables of interest, this limitation also demonstrates the strength of the study; if one time exposure could have such an impact, repetitive long term relationship with the show is only going to magnify the effects.
These results clearly indicate that narrative involvement process does not end with the viewing experience and one motivation behind connecting with the programs and celebrities after the show has ended could be related to the involvement during viewing. It is important to note that the majority of the participant sample in this study were not regular viewers of the show Mad Men. Despite that, however, participants who reported to be involved in the narrative or with the character, expressed intentions to connect with them via different SNS post viewing. This suggests the importance of narrative engagement and its long-term influence.These findings are especially striking given most media organizations are adopting Web 2.0 and striving to gain audience attention by enhancing online engagement. Recent research has pointed to the power of narratives and of SNS in influencing attitudes and behaviors. This study adds to the literature by bringing these two distinct concepts together and urges future research to examine how this framework could be utilized for prosocial messaging. In so doing, this study bridges previousresearch on narrative processing of entertainment shows, uses and gratification, and social media engagement.
Online Journal of Communication and Media Technologies Volume: 6 -Issue: 4 October -2016