SOCIAL NORM CUES AND THE NARRATIVE ENJOYMENT AND APPRECIATION RATIONALE (NEAR) By Kevin Kryston A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Communication—Doctor of Philosophy 2021 ABSTRACT SOCIAL NORM CUES AND THE NARRATIVE ENJOYMENT AND APPRECIATION RATIONALE (NEAR) By Kevin Kryston Narrative content and social cues (e.g., others’ opinions, context) are powerful determinants of audiences’ appraisals of entertainment, such as enjoyment and appreciation (Nabi & Krcmar, 2004; Tamborini, 2013). However, few studies examine the interaction effect of content and social cues on audience appraisals. The model of intuitive morality and exemplars (Tamborini, 2013) and its narrative enjoyment and appreciation rationale (Lewis et al., 2014) suggest that social cues could influence appraisals by either reinforcing or conflicting with content appraisals. In this project, I tested the extent that social cues, e.g., social norm messages (Lapinski & Rimal, 2005) moderate narrative appraisals. Across two studies, participants read or listened to stories with either positive or mixed endings. The endings were accompanied by norm cues that indicated others’ overwhelming dis/liking of the story. In Study 1, reinforcing social cues (others’ liking of positive endings) solidified audiences’ enjoyment of the narrative content, whereas conflicting social cues (others’ disliking of positive endings) decreased enjoyment and lengthened evaluation times. For mixed endings, norm cues led audiences to dis/like content in line with others’ opinions. Study 2 explored how moderators of normative influence affect appraisals and social sharing. Few hypotheses received support, but the findings overall suggest that social norm cues increased participants’ enjoyment and social sharing of narrative conflict. Taken together, social norm cues affect enjoyment because these cues provide social information that satisfies intuitive needs that reinforce, conflict with, or resolve initial appraisals. ACKNOWLEDGEMENTS Though I wish I had more space to recognize others, I would like to extend a few special acknowledgments to people who played a special role in my graduate studies and other fantastic individuals who I am blessed to have in my life. First, Teri Thompson, who was among the first to encourage me to attend graduate school and would eventually advise my Master’s thesis and push me toward Michigan State. Second, Jean and John Henninger, who gave me a loving home while I pursued my Master’s degree. And finally, the amazing people in the Communication Department at Michigan State, especially Allison Eden, Ron Tamborini, Brandon van der Heide, Marge Barkman, Thomi Chrisinske, Lindsay Hahn, and Eric Novotny, who helped me survive and thrive over my five years in East Lansing. My close friends and family also deserve extra praise. I would like to thank Patrick Firme, Alex Schock, Katrina Staker, Mary Wray, and Kevin Sullivan, and many others for their friendship. Special thanks to my mother, Dawn Raspanti, for being the most amazing, wonderful, fabulous mom I could ask for, and my father, the late David Kryston, for being a sweet, caring and loving dad, may he live on forever in memory. Last, but certainly not least, thank you to the love of my life, Sunyoung Park who brings me infinite happiness in good times and bad. Thank you—and everyone not listed who also deserves my gratitude—for helping me get to where I am now. iii TABLE OF CONTENTS LIST OF TABLES vi LIST OF FIGURES viii INTRODUCTION 1 LITERATURE REVIEW 4 The NEAR and Media Appraisals 4 Integrating Social Cues into the NEAR 6 Social Effects on Entertainment Appraisals 7 Social Perceptions and Entertainment 8 STUDY 1 13 Study 1 Method 13 Participants 13 Procedure 13 Stimuli 14 Stories 14 Norm Cues 14 Measures 14 Enjoyment and Appreciation 14 Liking and Evaluation Times 14 Self-reported Enjoyment and Appreciation 15 Thought-listing 15 Demographics 15 Results 16 Descriptive Statistics 16 Confirmatory Factor Analysis 16 Randomization and Covariate Tests 17 Hypothesis Testing 17 Study 1 Discussion 20 STUDY 2 21 Novelty 21 Publicness 23 Social Sharing 23 Study 2 Method 25 Participants 25 Procedure 25 Stimuli 26 Video Pretest 27 Measures 29 iv Perceived Norms 29 Perceived Publicness 30 Social Sharing 30 Group Identity 30 Self-presentation Motivation 31 Results 31 Confirmatory Factor Analysis 31 Descriptive Analysis and Covariate Tests 31 Manipulation Checks 32 Hypothesis Testing 33 RQ1 and RQ2: Differences from Study 1 34 Mediation Analyses 35 Alternative Tests of Hypotheses 35 Publicness, Group Identity, and Self-presentation Motivation 36 Social Sharing 38 DISCUSSION 39 Study 1 39 Study 2 41 General Discussion 42 Limitations 45 Conclusion and Future Directions 47 APPENDICES 51 APPENDIX A: Narrative stories 52 APPENDIX B: Sample story stem screen 55 APPENDIX C: Sample manipulations 56 APPENDIX D: All items for Study 1 57 APPENDIX E: All items for Study 2 59 APPENDIX F: Summary of significant covariate tests in Study 2 62 APPENDIX G: Tables 63 APPENDIX H: Figures 82 REFERENCES 90 v LIST OF TABLES Table 1. Story randomization in Study 1. 63 Table 2. Story appraisals by story stem, ending, and norm cue. 64 Table 3. Means and standard deviations by condition and repeated-measure ANOVA test statistics. 65 Table 4. Correlation matrix for Study 1 across conditions. 66 Table 5. Confirmatory factor analyses (CFA) model statistics for Study 2. 66 Table 6. Appraisal means by condition in Study 2. 67 Table 7. Social sharing means by condition in Study 2. 68 Table 8. Correlation matrix for Study 2. 69 Table 9. PROCESS Step 1: Norm cue condition predicting perceived descriptive norms. 70 Table 10. Effect of perceived norms and endings on appraisals and social sharing. 71 Table 11. Conditional indirect effects of norm condition on appraisals through perceived descriptive norms. 71 Table 12. Conditional indirect effects of norm cues on social sharing through descriptive norms. 72 Table 13. PROCESS Model 18 Step 2: Moderating effects of ending and publicness. 73 Table 14. Conditional indirect effects of norm cues on appraisals through descriptive norms by ending and publicness. 74 Table 15. Conditional indirect effects of norm cues on sharing through descriptive norms by ending and publicness. 75 Table 16. PROCESS Model 18 Step 2: Moderating effects of ending and identity. 76 Table 17. Conditional indirect effects of norm condition on appraisals through perceived norms by ending and identity. 77 Table 18. Conditional indirect effects of norm cues on social sharing through descriptive norms by ending and identity. 78 vi Table 19. PROCESS Model 21 Step 1: Moderating role of self-presentation on norms. 79 Table 20. Conditional effects of norm cues at different levels of self-presentation motivation. 79 Table 21. Conditional indirect effects of norm condition and self-presentation on appraisals. 80 Table 22. Conditional effects of norm condition and self-presentation on social sharing. 81 vii LIST OF FIGURES Figure 1. The model of intuitive morality and exemplars (MIME) and its narrative enjoyment and appreciation rationale (NEAR; Eden et al.,2021). 82 Figure 2. Enjoyment by condition. 83 Figure 3. Appreciation by condition. 83 Figure 4. Liking by condition. 84 Figure 5. Raw evaluation times by condition. 84 Figure 6. Log-transformed evaluation times. 85 Figure 7. Path model for alternative tests of hypotheses and research questions (Study 2). 85 Figure 8. Effect of perceived descriptive norms on enjoyment by story ending (Study 2). 86 Figure 9. Effect of perceived descriptive norms on the probability of liking by publicness. 87 Figure 10. Effect of perceived descriptive norms on the probability of liking by identity. 88 Figure 11. Effect of norm condition on perceived descriptive norms by self-presentation. 89 viii INTRODUCTION Narrative features and social context are two of the most prominent factors that influence media appraisals such as enjoyment and appreciation (Bryant & Oliver, 2009; Oliver & Krakowiak, 2009). Research examining enjoyment responses to narratives suggests that content cues (i.e., contained within the narrative, such as character behaviors, emotional displays, the resolution of a narrative) are the most important factor affecting enjoyment (e.g., Tamborini et al., 2010; Vorderer, 2003; Zillmann, 2000). However, external influences, namely social cues (e.g., the physical context of viewing, others’ opinions, social norms), can alter individuals’ appraisals such that people are more likely to like what other people like (e.g., Bourdieu, 1984; Sundar, 2008). Therefore, research is needed to better understand the individual and combined roles of narrative features and social cues in entertainment appraisals. The model of intuitive morality and exemplars (MIME; Tamborini, 2013) is one of the few theories that explicates the way content and social factors interact to influence appraisals. According to the MIME, narrative appeal is driven by the salience and upholding of intuitive motivations portrayed in media content. With increased exposure, motivations become more salient, and thus more appealing to individual audience members. These processes also happen at the social level: the relative salience of intuition is shared among groups of individuals (e.g., nations, friends) who consume and prefer similar media content. The MIME suggests that narrative appeal is a product of both narrative appraisal processes and social factors that affect intuition salience. MIME research, however, mostly focuses on individual narrative appraisals and their effects on media appeal and moral judgments (Eden et al., 2021). Studies focusing on the MIME’s claims regarding social-level effects are 1 limited to content analyses of cultural content preferences (Prabhu et al., 2020) rather than the way social factors impact appraisal processes (cf. Eden & Tamborini, 2017). This gap in the literature should be addressed because social and content cues interact to determine judgments not only about what is right, appropriate, and normative, but also what is fun and meaningful. The current paper applied the logic of the MIME (Tamborini, 2013) to examine how social cues can affect media appraisals by either (a) reinforcing content cues or (b) creating conflict between social and content cues. Across two studies, I relied specifically on the subsection of the MIME known as the narrative enjoyment and appreciation rationale (NEAR; Lewis et al., 2014; Tamborini et al., 2021; see Figure 1). These studies address key issues in determining the role of social cues in content appraisals. First, I tested how social cues (e.g., norms) influence the appraisals of content (Study 1). However, the MIME does not explicate how social cues relative to narrative appraisals are made more or less salient to viewers. To examine social cue salience and its effects, I drew from logic proposed by the motivation and opportunity as determinants (MODE; Ewoldsen et al., 2015) model. Broadly, the MODE model argues that highly accessible constructs (attitudes, norms) exert more influence on behavior than less accessible constructs. According to the MODE model, constructs like social norms, a group’s predominant behavior, or cognition (Cialdini et al., 1990; Lapinski & Rimal, 2005) can be made accessible by social and contextual cues. For example, one such cue is the presence or absence of fellow group members. However, constructs like social norms may become increasingly salient (chronically accessible) with repeated exposure and thus exert greater influence on judgments than novel or minimally accessible norms. The MODE model therefore suggests that social cues, specifically social norm cues, may play a more important role in narrative processing when they are (1) committed to 2 memory and (2) in contexts that increase norm salience (e.g., public). I examined these claims in Study 2, in which I tested the influence of situational moderators of social norms on entertainment appraisals. Additionally, Study 2 took advantage of its public context to study the role of social sharing in developing and maintaining social norms around entertainment content. The results from these studies form the basis to better understand how social and content cues interact in a media context to shape appraisals of content, thereby filling in an important gap in the MIME, i.e., how, when, and to what extent social context shapes responses to content. 3 LITERATURE REVIEW The NEAR and Media Appraisals The MIME is multifaceted, but one portion of the MIME, the narrative enjoyment and appreciation rationale (NEAR; Lewis et al., 2014; Tamborini et al., 2021; see Figure 1), can help explain the role of social cues in media appraisals. The NEAR focuses on short-term appraisals of content, and is based on a dual-process framework that argues audience response is shaped by intuitive or deliberative processing systems. The NEAR argues that (1) narrative appraisals are formed based on the way narratives affect the salience of intuitive needs in viewers, and (2) the satisfaction of these needs in content—and by extension, in audiences—influences appraisals. Intuitive needs, as used in the NEAR, are psychological motivators that drive self- and social- oriented behaviors (Ryan & Deci, 2000; Haidt & Joseph, 2007). The NEAR argues that content cues make needs more or less salient (Lewis et al., 2014; Tamborini et al., 2019; Tamborini et al., 2021). Need salience can vary as a product of a wide array of content cues, such as the frequency a need is discussed in a narrative, which needs are valued by protagonists, dialogue, or cinematography that emphasizes a given need, and so on. When all salient intuitive needs are satisfied by entertainment experience, positive affect produced by intuitive processes is labeled enjoyment. By comparison, when the entertainment experience satisfies the most salient intuitive needs, but not all salient intuitive needs, positive affect is produced by deliberative processing and labeled as appreciation. By taking a functional, process-oriented approach to explain audience responses to narratives, the NEAR makes clear predictions about how narrative cues affect intuitive (enjoyment) or deliberative (appreciation) appraisals of the content. Empirical findings support 4 the NEAR’s claims regarding how need satisfaction in narrative content affects enjoyment and appreciation (Lewis et al., 2014). To date, NEAR research has exclusively examined how content cues affect appraisals, but the NEAR’s guiding framework, the MIME (Tamborini, 2013), suggests a role of social cues in appraisal processes. According to the MIME, the salience of intuitive needs affects media appeal at both the individual and social level. For an individual, exposure to content depicting an intuitive need increases that need’s salience, and the most salient needs are the most appealing. Importantly, the MIME also argues for social-level effects. The relative salience of intuitive needs is shared within groups with a similar culture and/or media diet. Membership to these “morality subcultures” (Zillmann, 2000, p. 61) affects media appeal such that the needs which are salient to a given subculture become more appealing to individual members, thereby altering individual members’ appraisals. Although the MIME describes both individual- and social-level effects, most MIME research has focused on individual-level effects (Eden et al., 2021). Research of social-level effects is limited to examinations of the aggregate of individuals’ appraisals and selection behaviors (e.g., content analyses examining the representation of intuitive needs across cultures; Prabhu et al., 2020) or how membership to morality subcultures moderates character judgments (Eden & Tamborini, 2017). On the other hand, the role of social cues (artifacts and messages conveying societal-level information) in individuals’ appraisal processes has not been studied. However, in his explication of the MIME, Tamborini (2013) noted that the salience of social cues may affect audience appraisals of entertainment in ways that are consistent with the NEAR. Although never explicitly mentioning social cues in the explication of appraisal processes, Tamborini suggests that environmental cues (e.g., context of consumption, presence 5 of other people) may affect intuition salience, and as a result, appraisals. Tamborini’s argument that social cues play a role in appraisal processes is apparent when discussing how intuition salience would influence the experience of watching a football game at a crowded stadium: As a case in point, if I were at a football game pitting my favorite team against its arch rival, how would I feel near the end of the game as I watched my team routing the detested foe? In many instances, I might watch gleefully, and consider it a just and moral act of recompense. But if I was sitting next to my daughter, who often brings to my mind the parable of the Good Samaritan, I might be saddened by the lack of compassion displayed, even toward a bitter enemy. (p. 6) Although this example features during a discussion of moral judgments, it is important to note that (1) according to the MIME, moral judgments and narrative appeal are governed by the same intuition salience logic, and (2) this example mentions the way that salient social cues (the crowded stadium, the daughter) affect the response to the football game. As such, this example— specifically the crowd reinforcing the satisfaction and enjoyment of seeing his team win or his daughter’s presence conflicting with this satisfaction and evoking negative emotions—points to the potential for experiences outside the narrative—in this case, social cues—to affect an audience member’s appraisal of entertainment. Integrating Social Cues into the NEAR Salient social cues may play two roles in the NEAR. First, social cues can reinforce an individual’s appraisal of content cues. Reinforcement occurs when the social cue reaffirms the individual’s appraisal of content (e.g., most people enjoy content that satisfies needs and dislike content that thwarts needs). Put differently, the social cue should bolster initial reactions to content (e.g., greater or solidified enjoyment of liked content or less enjoyment of disliked content). Second, social cues can conflict with content cues (Bartsch et al., 2008; Tamborini, 2013). Conflict occurs when the social cue is the opposite of an individual’s appraisal of content (e.g., most people enjoy content that thwarts needs or dislike content that satisfies needs). 6 Conflict between needs fulfilled by content and a drive to adhere to the majority opinion should elicit controlled reappraisal of the content, resulting in appreciation (Lewis et al., 2014). In the following section, I review literature that shows salient social cues, specifically those that communicate social norms, might influence media appraisals, and argue that applying logic from the norms literature can explicate the role of social cues in the NEAR. Social Effects on Entertainment Appraisals Research examining the effects of social cues on media appraisal has shown that consensus cues such as reports of box office success and critical acclaim (Basuroy et al., 2003; Eliashberg & Shugan, 1997) and homogenous dis/approving comments from others can affect enjoyment (Möller et al., 2019; Shedlosky-Shoemaker et al., 2011; Waddell & Sundar, 2017; Waddell et al., 2020; Winter et al., 2018). Broadly, the literature supports the claim that individuals’ enjoyment usually aligns with collective opinion (i.e., if others like it, I like it as well). As such, the literature on social cues and media appraisal echoes seminal social influence research showing that people have a propensity to conform to the consensus behaviors and opinions of those around them to make a correct decision or maintain group relationships (Asch & Guetzkow, 1951; Deutsch & Gerard, 1955; Heider, 1946; Sherif, 1973; see Shulman et al., 2017 for review). Similarly, perceptions of consensus can also lead people to select (Rubin, 1983), enjoy (Roe, 1980), and endorse (Johnson & Ranzini, 2018) media content that they believe other people like, even when social cues are absent. These findings align with research showing that people tend to conform to the perceived consensus of those around them (i.e., the perceived norm; Carey et al., 2010; Park & Smith, 2007). Taken together, social norm cues and perceived social norms both affect individuals’ appraisals of content. 7 In a media context, consensus cue research mostly ignores the reasons why these cues affect enjoyment, whereas social perception research ignores the reasons these perceptions arise (i.e., exposure to social cues). Social norms, a well-established theoretical perspective in the social influence literature, clarifies the process by which social cues affect social perceptions, which subsequently influence outcomes. Additionally, given the prevalence of viewer scores on entertainment websites (e.g., IMDb, RottenTomatoes), the popularity of social viewing apps (e.g., Twitch, TeleParty), and the relevance of social cues in in-person viewing contexts (e.g., co- present others in theaters or at home), social norm cues are likely to be salient throughout audiences’ viewing experiences. As such, social norm cues are an important moderator of entertainment appraisals. Social Perceptions and Entertainment Social norms refer to a group’s predominant behavior or cognition (Cialdini et al., 1990; Lapinski & Rimal, 2005). Although there are various theories of normative influence, almost all norms theories claim that individuals generally adopt behaviors that they perceive as normative within a group to which they belong (Rhodes et al., 2020). Additionally, most theories of normative influence argue that norms can be grouped into two categories, each with specific ways of inspiring conformity to the norm. Descriptive norms are the prevalent behavior in a group (Cialdini et al. 1990). People adhere to descriptive norms because descriptive norms provide a decision-making shortcut that affords quick action. Injunctive norms are the group’s collective dis/approval of a given behavior (Lapinski & Rimal, 2005). People adhere to injunctive norms to maintain relationships with their group and/or out of fear of social sanctions. Social norms and cues communicating norms have been studied for decades, and normative effects have been observed in many behavioral domains, including health (Lapinski et 8 al., 2013), environmental (Goldstein et al., 2008), commerce (Bobek et al., 2007), and various other behaviors (Rhodes et al., 2020; Shulman et al., 2017). Additionally, recent work showed that norms affect media enjoyment (Kryston & Eden, 2021). Yet, research has not examined how social cues interact with content cues to influence appraisals. Doing so would (1) help develop a cohesive framework by which to study content and social influence on appraisals and (2) explicate how the mechanisms of audience appraisal work in different social contexts. Although various social factors (e.g., interpersonal communication) may influence NEAR processes, I suggest that the social norms literature can act as a guiding framework to understand the role of social cues in media appraisal theories such as the NEAR, the focus of the current manuscript. Norms theories can help explain the role of social cues in NEAR processes because the perspectives share two compatible mechanisms. First, both the NEAR and norms theories suggest that implicit social motivations can generate pleasure in individuals who either witness or adhere to the motivation (Cialdini et al., 1990; Tamborini et al., 2021). Second, both are based on dual-process logic. The norms literature argues that intuitive reactions to norms occur when the decision to adhere to a norm is made quickly with little to no cognitive processing (e.g., Ciadini et al., 1990), while deliberative reactions occur when an individual consciously processes a norm cue or normative perceptions (Carey et al., 2010). The NEAR is based on the same logic, suggesting enjoyment stems from unconflicted intuition presentation in content, and appreciation from conflicted intuition presentation (Tamborini et al., 2021). The conflict inspires contemplation and thus leads to longer processing times and increased deliberation. Inherent to these claims is the assumption that both narrative satisfaction and salient norms will have positive main effects on enjoyment (Kryston & Eden, 2021; Lewis et al., 2014). 9 However, the role of norms in affecting appraisals of narrative conflict is less clear. The first possibility is that social cues increase or reduce the salience of narrative conflict. Some media scholars argue that the main function of social cues in affecting media appraisals is that they moderate the way audiences process content (e.g., by altering expectations; Shedlosky- Shoemaker et al., 2011). If so, when social cues indicate an overwhelming preference for narrative conflict, an individual should spend more time processing that conflict than they would if social cues were absent, thereby appreciating said narrative conflict more than when social cues are absent. Conversely, processing time should decrease when narrative conflict is not normative. If this first hypothesis is supported, appreciation—both in terms of processing time and explicit responses—should increase when a norm cue indicates the popularity of a narrative featuring content wherein needs conflict since the norm calls attention to the conflict between needs rather than satisfying (or indicating the potential satisfaction of) a salient need. The second possibility is that social cues moderate appraisal of conflict the same way that they do for narratives that do not feature narrative conflict. Some research has shown that social cues moderate appraisals of content synonymous with appreciation responses. For example, supportive comments posted on elevating videos made the experience more elevating than when comments were absent (Krämer et al., 2019). Similarly, past work showed that chronically accessible norm perceptions can reduce the extent to which counter-attitudinal media messages are processed, suggesting a role of norms in resolving cognitive conflict (Rhodes et al., 2009). If social cues affect the appraisal of narrative conflict the same way they affect non-conflict narratives, then normative information should alter the extent to which conflict is enjoyed. As such, information that indicates the popularity of the content would increase the enjoyment of narrative conflict—both by increasing audience liking and by eliciting faster response times. 10 Liking scores would decrease when content is not normative, and the judgment would be made more quickly than if the norm cue was absent. Therefore, I pose the following hypotheses: H1: When social cues reinforce positive narrative endings, enjoyment will increase compared to when norms conflict with endings. H2: When social cues conflict with positive narrative endings, appreciation will increase compared to when norms reinforce endings. There are two possible ways that social cues could moderate narrative conflict, leading to two competing hypotheses. The first is that social cues alter the salience of needs in conflict, leading to differences in appreciation, which is represented in the following hypothesis: H3: Social cues will moderate appraisals of narrative conflict such that (a) social norm cues indicating the content’s popularity will increase appreciation and (b) cues indicating the content is not normative will reduce appreciation. The second possibility is that social cues moderate appraisals similarly to the way they affect narratives wherein intuitions are not in conflict, thereby affecting enjoyment. This leads to the following hypothesis: H4: Social cues will affect appraisals of narrative conflict such that (a) social norm cues indicating the content’s popularity will increase enjoyment and (b) cues indicating the content is not normative will decrease enjoyment. Additionally, measuring if and when social cues are explicitly salient while appraising stories could shed light on the processes by which narrative and social cues affect enjoyment and appreciation. Therefore, participants listed salient thoughts during their evaluation of each narrative to examine whether social cues play a role in their sub/conscious appraisal of each 11 story (Cacioppo et al., 1979). Thought-listing procedures have been employed to determine whether certain concepts or stimuli were overtly salient while individuals made judgments (Cacioppo et al., 1979; Kuppens et al., 2013), and may offer support for the processes explained in the NEAR if linked to appreciation responses. 12 STUDY 1 Study 1 Method Participants Participants were recruited from the Communication Department research pool at Michigan State. Participants provided informed consent before participating, and all procedures were approved by the university’s institutional review board. Two participants were removed for failing an attention check, resulting in a sample of n = 115 (76% female; Mage = 19.88, SDage = 1.42; nwhite = 90, nAsian = 13, nblack =11, nHispanic = 4, nMid. Eastern = 3, nNat. Am. = 2, nIndian = 11). Procedure This study was a 2 (narrative ending: positive, mixed) x 3 (norm cue: universal acclaim, overwhelming dislike, control) within-subjects design. In a modified version of the procedures from Lewis et al. (2014), participants were instructed to read six stories billed as film synopses that varied in terms of their endings and norm cues. As in Lewis et al. (2014), participants were randomly assigned to one of six groups of stories. Within each group, each story was assigned to one of the six potential norm-ending combinations, resulting in 36 unique combinations of story stems, endings, and norms (see Table 1). After reading and appraising six stories (one for each norm-ending pair), participants completed socio-demographic measures. Then, they were debriefed about the true purpose of the study, thanked, and dismissed. 1 Participants could select all races/ethnicities that applied to them, write-in their race/ethnicity if it was missing, or select “prefer not to answer.” 13 Stimuli Stories Six stories from Lewis et al. (2014) were used in the current study. The stories have two parts: a stem and an ending. Each stem is a paragraph that introduces the characters and presents a conflict between two of the protagonists’ salient needs (e.g., protecting a baby versus saving a group). After reading the stem, participants proceeded to read either a positive (all needs satisfied) or mixed (one need satisfied at the cost of the other) ending. Over the course of six stories, each participant saw three positive and three mixed endings. Norm Cues Norm cues were presented after the story stem in tandem with narrative endings. Each story was randomly assigned to either the high percentage norm cue (92% or 94% liked), low percentage norm cue (9% or 12% liked), or no normative information (control). For clarity, I refer to these conditions as universal acclaim (92%/94% liked), overwhelming dislike, (9%/12% liked), and control based on labels from Metacritic.com. Since participants see each norm cue condition twice, I did not repeat percentages so that participants would not be suspicious about the manipulation (this procedure is in line with Kryston et al., 2020). Measures Enjoyment and Appreciation Story appraisals were measured using Lewis et al.’s (2014) method, which measured enjoyment and appreciation using (1) liking and evaluation times and (2) self-report scales. Liking and Evaluation Times. First, evaluation times and simple dis/like responses (Like = 1, Dislike = 0) were measured on the ending screen. After advancing to the end screen, participants had three seconds to read before the buttons appeared. Time was recorded as soon as 14 they made a selection, at which time they advanced to the next screen. See Table 2 and Table 3 for scores and distributions. Following Lewis et al., (2014), evaluation times were non-normal and right-skewed across conditions, so both raw and log-transformed scores are reported in Table 3. Per Lewis et al., fast liking scores indicate enjoyment, whereas slow responses and variable liking judgments (e.g., group mean of ~.50 on a scale from zero to one) indicate appreciation. Self-reported Enjoyment and Appreciation. Second, after the ending screen, participants completed Oliver and Bartsch’s (2010) enjoyment and appreciation scales (e.g., “I had fun reading this story” for enjoyment, “The story was thought-provoking” for appreciation). Both scales have three items and were measured on a seven-point scale (1 = Strongly disagree, 7 = Strongly agree). High scores indicate greater enjoyment or appreciation. The scales showed strong internal consistency across conditions (.91 ≥ α ≥ .83). Thought-listing A thought-listing procedure adapted from Cacioppo et al. (1979) was used to determine what participants were thinking about while rating the story. Participants were asked to briefly state what they were thinking about while rating and could list up to eight words or phrases (K = 1,748 thoughts). Responses were coded by searching for social words (e.g., participants, people), specifically those in the message (students). Social thoughts were coded as 1, and if a social word did not feature, the unit was coded as 0. Social thoughts were added to create a score for each participant, which indicated the total number of social thoughts (M = 0.07; SD = 0.26). However, I should note that no participant referenced social cues multiple times. Demographics Participants reported their age, race/ethnicity, and gender at the end of the study. 15 Results Descriptive Statistics Means and standard deviations are reported in Table 3 and a correlation matrix can be found in Table 4. Confirmatory factor analysis First, confirmatory factor analysis (CFA) was conducted to determine the validity of the three-item enjoyment and appreciation scales. CFI, TLI, RMSEA, and SRMR were used as the primary indicators of model fit, with factor loadings and patterns used to identify potentially invalid items (Hunter & Gerbing, 1982). First, across all measures (K = 690 across 115 participants) the scales displayed acceptable fit, χ2(8) = 62.76, p < .001, RMSEA = .10, CFI = .98, TLI = .96, SRMR = .05 (Browne & Cudeck, 1993). Removing an item from the appreciation scale (“The story was thought-provoking”) improved fit slightly, χ2(4) = 22.02, p < .001, RMSEA = .08, CFI = .99, TLI = .98, SRMR = .03, but not enough to warrant altering the scale given its frequent use and validation in other work (Weinmann et al., 2016). I then conducted CFAs within each condition using the model specification (n = 115). Model fit in four of the six conditions was average to above-average (CFI ≥ .98, TLI ≥ .97, RMSEA ≤ .09, SRMR ≤ .04) but fit was poor in the mixed ending condition combined with universal acclaim norm cues (CFI = .91, TLI = .83, RMSEA = .19, SRMR = .10) and when norms were absent (control; CFI = .94, TLI = .89, RMSEA = .16, SRMR = .08). In both instances, multiple items in both the enjoyment and appreciation scales loaded poorly with their factors. However, given the results of the CFA across observations, I proceeded using Oliver and Bartsch’s (2010) scales as planned, and discuss reasons for caution in the limitations. 16 Randomization and Covariate Tests First, I assessed whether there were any systematic differences in the dependent variables by randomization group, story stem, and gender. I used one-way ANOVAs to assess differences in average liking, evaluation times, enjoyment, and appreciation across conditions by randomization groups. These differences were not significant for any measure (p > .18), indicating randomization was successful. Next, to assess differences by story stem, the ANOVA was repeated on means by story stem. As expected, some stories were enjoyed and appreciated more than others for certain ending/norm combinations, but these differences were not systematic (see Table 2). Therefore, story stems within condition were collapsed for analysis. After these tests, a series of multiple regressions explored whether stem reading time was associated with story evaluations in any condition. These tests produced non-significant results; none of the reading times for the story stems significantly predicted liking, enjoyment, and appreciation. Last, t-tests probed potential differences in dependent measures by gender. No significant differences were observed (p > .05). Hypothesis Testing Hypotheses were tested using 2 (ending: positive, negative) x 3 (norm cues: universal acclaim, overwhelming dislike, control) repeated measures ANOVAs, which were repeated for each dependent measure (liking, raw and log-transformed evaluation times, enjoyment, appreciation). These tests produced the anticipated main effects of norm cues and endings, and a significant interaction of norms and endings on enjoyment, appreciation, and log-transformed evaluation times. All significant main effects of norms and endings were in the predicted direction. No significant main effects of ending were observed for raw or transformed evaluation 17 times.2 The interaction effect on liking (p = .08) and raw evaluation times (p = .06) approached significance (Table 3). Results are summarized in relation to their corresponding hypothesis. H1 predicted that norm cues that reinforce positive endings (i.e., those indicating universal acclaim) would increase enjoyment compared to norm cues conflicting with endings (i.e., those indicating overwhelming dislike). This increased enjoyment should be indicated by increased liking and self-reported enjoyment and faster evaluation times. H2 stated that norm cues that conflict with positive endings (i.e., those indicating overwhelming dislike) would increase appreciation compared to norm cues reinforcing endings. This increased appreciation should be indicated by slower evaluation times and increased self-reported appreciation. Supporting H1, self-reported enjoyment (M = 4.53, SD = 1.37) and liking (M = 0.78, SD = 0.41) were greater and reaction times were faster (Figure 2, 4, and 5) when norms reinforced positive endings (positive ending—universal acclaim) compared to when norms conflicted with the ending (positive ending—overwhelming dislike; Menjoy = 3.89, SDenjoy = 1.50; Mlike = 0.54, SDlike = 0.50), Fenjoy(2, 228) = 3.17, p = .04, η2 = .02, Flike(2, 228) = 2.52, p = .08, η2 = .02. Contrary to H2, appreciation was higher when norms reinforced positive endings (M = 4.25, SD = 1.26) compared to when they conflicted with endings (M = 3.64, SD = 1.40; Figure 3), F(2, 228) = 11.96, p = .05, η2 = .02. Additionally, positive endings with no normative information (control) were liked (M = 0.79, SD = 0.41), enjoyed (M = 4.43, SD = 1.22), and appreciated (M = 4.19, SD = 1.28) significantly more than when associated with cues indicating overwhelming dislike. There were no significant differences between the universal acclaim and control 2 Outliers were identified before analysis using a stem and leaf plot. Extreme values (eight of a possible 36 cells across six participants) were replaced with mean evaluation time values for analysis purposes). I also ran analyses on evaluation times with these outlier cases removed, but all main and interaction effects to raw and/or log- transformed scores reflected the significance of the tests reported in text and in Table 1. 18 conditions. The results support H1 and H2 for liking and evaluation times, but do not support H2 for self-reported appreciation. H3 and H4 were competing hypotheses that suggested that norms increase appreciation (H3) or enjoyment (H4) of mixed endings. Mixed endings associated with universal acclaim norm cues were enjoyed (M = 4.50, SD = 1.26) and appreciated (M = 4.51, SD = 1.26) significantly more than mixed endings associated with overwhelming dislike norm cues (Menjoy = 3.73, SDenjoy = 1.45; Mappr = 4.08, SDappr = 1.36) or no norm cues (Menjoy = 3.90, SDenjoy = 1.34; Mappr = 4.15, SDappr = 1.28). The results support both H3 and H4. However, liking scores for mixed endings were significantly higher when associated with universal acclaim cues (M = 0.67, SD = 0.47) compared to the control condition (M = 0.50, SD = 0.50), and significantly lower in the overwhelming dislike norm cues (M = 0.37, SD = 0.49) condition than the control condition (Figure 4). There were no significant differences in raw and transformed evaluation times, but the pattern of means suggests that both the universal acclaim and overwhelming dislike norm cues induced longer evaluation times than no-norm controls. The positive relationship between explicit normative information and mixed-ending liking supports H4. Lastly, participants’ thoughts were analyzed by searching for the words “other,” “people,” “participant,” and “student” and their plural forms (i.e., social words mentioned in the experimental procedure and their synonyms) among the responses. Of the 1,748 thoughts listed across 115 participants, only eight (0.0045%) mentioned one of these words, indicating that very few participants were explicitly or overtly thinking about social cues when making their assessments of the content. 19 Study 1 Discussion The current study applied the logic from social norms literature to the NEAR to explicate the process by which social cues moderate narrative appraisals. In line with the hypotheses, norm cues indicating universal acclaim increased liking and enjoyment for unconflicted content compared to norm cues indicating dislike or when cues were absent. Audiences liked, enjoyed, and had faster evaluations of stories accompanied by norm cues reinforcing positive endings compared to norm cues conflicting with those endings. Conflicting norm cues decreased liking and enjoyment. The results support the notion that social cues play a similar role to narrative cues in determining audience response: social cues suggesting the opportunity to satisfy needs associated with norm adherence increase enjoyment, whereas social cues indicating that the content is not normative decrease enjoyment. But, norm cues appeared to affect audiences’ enjoyment (or disliking) of mixed endings directly, rather than by reinforcing appreciation responses. There were no differences in liking, enjoyment, and evaluation times of stories with positive endings, whereas norm cues affected enjoyment and liking of mixed endings. The effect of social norm cues on mixed-ending enjoyment supports the potential to integrate social cues into the NEAR given their complementary mechanisms and association with enjoyment. However, some questions remained after Study 1, including whether the cues in the study were truly norm cues, or some other social construct or moderating variable that was not measured in the experiment. Therefore, some of these issues were addressed by conducting a second study in which potential moderators were added to the design. 20 STUDY 2 Study 1 focused on how content and social cues interact to influence appraisals, but the social norms literature highlights two other important considerations regarding the way social cues may affect NEAR processes: norm novelty and publicness. Along with retesting H1–H4 from Study 1, Study 2 posed four research questions to address these two considerations. Additionally, Study 2 used a different operationalization of norms to explore the extent to which norms are inferred from group behaviors and reactions and better ensure that the salient social concept in these studies is norms rather than some other construct. Replicating the findings of Study 1 and observing significant moderations would offer confidence and nuance to the underlying mechanisms of social cue effects. Novelty First, novel social norms (those formed based on new information) and chronic social norms (those stored in memory based on experience) can have different mechanisms and effects (Cialdini, 1993; Ewoldsen et al., 2015; Rhodes & Ewoldsen, 2009). According to the motivation and opportunity as determinants (MODE) model (Ewoldsen et al., 2015), norms committed to long-term memory are more likely to become salient than norms that are less chronic. The MODE model refers to frequently activated norms stored in long-term memory as chronic norms. Chronic norms are formed when norm cues are committed to memory and become increasingly likely to be activated with repeated exposure (Higgins & Brendl, 1995). Given their predominance in conscious processing, the MODE model argues that chronic norms have a stronger influence on behaviors than less-salient norms, especially when an individual strongly identifies with the group communicating the chronic norm (Ewoldsen et al., 2015) or when the 21 group communicating a chronic norm is present or salient (Baldwin & Holmes, 1987). By contrast, single-exposure norm messages are less likely to be committed to memory and less likely to affect behaviors compared to chronic norms. Novel or single-exposure norm messages still affect individuals’ decision-making (Rhodes et al., 2020), but since they are less likely to be activated and are often less salient than chronic norms, their influence on outcomes is argued to be weaker (Ewoldsen et al., 2015). In sum, the MODE model suggests that chronic norms can either moderate or override the influence of novel cues. Thus, chronic norms may have a stronger influence on narrative appraisals than novel single-exposure norms. However, research on chronic norms is limited (Ewoldsen et al., 2015) and has not yet examined how chronic norms affect entertainment appraisal processes. Alternatively, chronic norms may motivate additional processing about both social and narrative cues. Making attitudes or norms more salient may increase the likelihood of deliberate processing, so long as an individual possesses the cognitive and temporal resources to control automatic response and is motivated to devote said resources to their evaluation (Ewoldsen et al., 2015). In this case, appraisals of both conflicting and unconflicted social and narrative cues would be derived via deliberation. Study 2 fills this gap in the literature by establishing norms before participants consume the narrative to see how existing norms affect subsequent narrative processing. That norm is reintroduced concurrent with the story ending (as in Study 1). Since research in chronic norms is limited, especially for minimally chronic norms, I pose the following research question: RQ1: How do established social cues moderate narrative appraisals? 22 Publicness The second consideration from the social influence literature regards the way that environmental cues moderate normative influence. According to the norms literature, social norm cues are more likely to exert influence in contexts where social rewards and/or punishments are more salient (e.g., in public settings, when decisions affect others or the individual’s public image; Lapinski & Rimal, 2005; Manning, 2011). Past studies have found that public contexts amplify normative effects (Bagozzi et al., 2000). Regarding the role of publicness in normative media influence, past research found that norms affect entertainment appraisals even in private situations with minimal social motivation (Kryston & Eden, 2021; Study 1). The results suggest that although publicness may amplify normative effects for health and environmental behaviors, perhaps publicness is less important to the influence of norms on media appraisals. Thus, I ask: RQ2: How do social and narrative cues interact to influence appraisals in a public setting? Social Sharing Finally, a study that creates and then measures the effect of norms in a public setting allows for a test of the extent to which norms affect subsequent sharing of content. Social sharing behavior occurs when mass media (e.g., TV/film content, music, news) is disseminated or discussed through interpersonal or mediated channels (Johnson & Ranzini, 2018; O’Sullivan & Carr, 2017). Research has identified two main reasons that people share and discuss media content: (1) to elaborate about content (e.g., discuss plot events, narrative themes, or special effects) or (2) for self-presentation/relationship-building (e.g., talk about a movie with a member of your in-group to strengthen your relationship with others; French & Bazarova, 2017). 23 Relevant to this research line, content that elicits appreciation is more likely to be shared (Bartsch, 2012; Oliver, 2008) for two reasons. First, appreciated content is more likely to have a complex narrative (e.g., multiple interwoven plotlines) or inspire complex emotions (e.g., mixed affect; Oliver & Hartmann, 2010). Therefore, people may turn to others in an attempt to make sense of the narrative or discuss complex feelings evoked by narrative conflict. Second, the narrative themes of appreciated content often highlight relational connections or human excellence (Oliver & Hartmann, 2010; Oliver & Raney, 2011). As such, relationships are likely more salient to the viewer after exposure, meaning a viewer may seek opportunities to talk with others. In either case, narrative content may evoke social sharing based on audience members’ desire to elaborate about the content, and viewers may be more likely to elaborate when social cues indicate that either (1) content is normative, or (2) when social cues conflict with narrative cues. Given the lack of research in this area, I pose the following research question. RQ3: How do narrative and social cues interact to affect social sharing in the form of written story reviews? Along with elaborating with group members, self-presentation motives can elicit social sharing (French & Bazarova, 2017). When self-presentation motivation drives social sharing, individuals share what they think their social network likes rather than their favorite entertainment (Johnson & Ranzini, 2018). Thus, those for whom self-presentation is more salient should be affected by social cues more than those who are less concerned with self-presentation. Likewise, individuals should be more prone to share with valued in-groups compared to dissimilar out-group members, pointing to a moderating role of group identity on sharing (Chung & Rimal, 2016; Johnson & Ranzini, 2018). However, again, research examining the role of group identity and social motivation effects on the extent of social sharing is limited, and scant 24 attention has been paid to the way group identity and social motivation moderate narrative and social cue effects. Thus, I ask: RQ4: How do (a) self-presentation motivation and (b) group identity moderate the effects of narrative and social cues on appraisals and social sharing? Study 2 Method Participants Participants were recruited from Amazon’s Mechanical Turk (MTurk) paid participant pool and were paid 2.50 USD for completing the study. In all, 262 passed the necessary attention check, and after removing participants who wrote in unrelated (e.g., a description of an Edgar Allan Poe story, information about resume writing; n = 25) or identical responses to data points submitted at earlier times (n = 5) in open-ended responses, the final sample consisted of N = 232 responses. Unlike the student sample in Study 1, Study 2’s sample had more men (n = 155, 66.8%) than women (n = 74, 31.9%; non-binary or transgender: n = 2, 0.9%; did not disclose: n = 1, 0.4%) and were older (Mage = 35.01, SDage = 9.11, range: 22-76). However, just like Study 1, the sample was mostly white (n = 188, 81%; nblack = 27, nHispanic = 10, nAsian = 4, nNat. Am. = 4). Two participants did not disclose their race/ethnicity. Procedure To test H1–H4 and RQ1–RQ4, Study 2 was designed as a 2 (ending: positive, mixed) x 3 (norm cue: universal acclaim, overwhelming dislike, control) between-subjects design. After providing consent, participants were told they would watch a pre-recorded webinar wherein a group listened to a short story. Participants were asked to act as if they were present in the live webinar with the others even though the session was pre-recorded. After receiving these instructions, participants completed a short practice session to familiarize themselves with the 25 evaluation procedure. During the practice, participants were told that “Like” or “Dislike” buttons would appear when they heard the story’s ending, and clicking one of these buttons would auto- advance them to the next screen, but given no other instructions about how and when to respond. After the practice, participants were randomly assigned to one of the six conditions and watched the video webinar. The webinar video started automatically when they arrived at the screen. As in Study 1, the dis/liking buttons appeared three seconds after the story ending, and pressing the button automatically advanced the participant to the next screen. After indicating their dis/liking of the story, participants completed measures of perceived norms, enjoyment, and appreciation. Then, participants wrote a review of the story. Participants were told that their review would be shared with future participants. Next, participants completed measures of perceived publicness, self-presentation motivation, group identity, and socio-demographic measures, as well as a simple attention check.3 Participants were then debriefed and thanked for their time. Stimuli Videos were created to reflect the experimental conditions. During the video, the story (Saving Comrades) was read aloud and the participant saw an array of five other people (confederates) listening to the story (see https://osf.io/37akv/?view_only=58e2b61c078d4b6d829e1dcd9b0b4285 for stimuli). The confederates demographics reflected those of MTurk participants (~50% fe/male; ~50% white; age 30-60; Moss & Litman, 2020). As mentioned, dis/liking buttons appeared three seconds after the story ending to give participants time to see the confederates react positively (e.g., smiles, nods; universal acclaim), negatively (e.g., frowns, looks of disgust; overwhelming dislike), or 3 If the attention check was answered incorrectly, participants were automatically removed. A non-essential indicator of how much the participant focused on the video during the webinar was measured at this time, as well. 26 blankly (control) to the ending. Additionally, in the universal acclaim and overwhelming dislike conditions, the social norm cue was presented before participants were exposed to the story and then again in tandem with narrative endings. Though the decision to pre-record videos rather than conduct live sessions may seem to limit realism, imagined situations can induce similar responses to real situations, and most differences in real versus imagined reactions involve overestimations about the extent of cooperation and connectedness among a group (Vlaev, 2012). Given Study 2’s research questions, these overestimations were desirable. Video Pretest Before employing the videos in Study 2, they were pretested to ensure they successfully induced perceived norms and confederates’ affective responses. I recruited 37 participants (N = 37) from social media and an undergraduate communication course (18 male, 19 female; Mage = 34.48, SDage = 16.55; nwhite = 32, nAsian = 3, nblack = 2, nHispanic = 1, nIndian = 1). More than half of the participants (n = 21) were personal acquaintances with at least one person in the experimental videos. All procedures were approved by Michigan State’s IRB. Participants in the communication course received extra credit for completion. After providing informed consent, participants were randomly assigned to one of the six experimental conditions from Study 1 and watched the video matching that condition. After watching, participants indicated how positive/negative the audience’s reaction was (M = 3.81, SD = 1.22; 1 = Extremely negative, 7 = Extremely positive) using a single item, and completed a three-item measure of perceived descriptive norms (full description below; α = .97, M = 3.57, SD = 1.63). After completing demographic measures, participants were thanked and dismissed. First, I analyzed differences in norms and reactions using a 3 (norm cue: universal acclaim, overwhelming dislike, control) x 2 (ending: positive, mixed) ANOVA. There was a 27 significant main effect of norms on audience reaction, F(2, 31) = 12.44, p < .001, η2 = .41, and perceived descriptive norms, F(2, 31) = 18.17, p < .001, η2 = .52. A post hoc Bonferroni determined that reactions and norm perceptions followed the expected pattern: reactions were perceived as significantly more positive (M = 4.82, SD = 1.17) and normative (M = 5.18, SD = 1.40) in the universal acclaim condition (n = 11) compared to the overwhelming dislike condition (n = 14; Mrxn = 2.93, SDrxn = 1.07; MDN = 2.29, SDDN = 1.08) and the control (n = 12; Mrxn = 3.92, SDrxn = 0.52; MDN = 3.58, SDDN = 0.90). Also, perceived norm scores were significantly lower in the overwhelming dislike condition compared to the control despite the fact that confederates’ reactions were not perceived as more negative. There were no main effects of story ending, nor a significant interaction between story ending and norm condition on confederate reaction valence or perceived norms (p > .17). Since so many participants knew people in the video, I used a 3 (norm cue) x 3 (knew audience: no, one member, multiple members) ANOVA to check for experimenter bias effects. No main effects of knowing the audience were observed (p > .36), nor did knowing audience members interact with norms in predicting descriptive norms (p > .38). However, there was a significant interaction between knowledge of the audience and norm condition in predicting audience reactions, F(4, 28) = 3.83, p = .01, η2 = .21. The interaction is driven by the fact that those who knew one (n = 12; M = 2.25, SD = 0.96) or more people in the video (n = 9, M = 2.50, SD = 0.58) perceived the reaction as more negative than those who knew no one (n = 16, M = 3.67, SD = 1.03) in the negative norm condition. Likewise, in the overwhelming acclaim condition, those who knew one (M = 5.50, SD = 1.54) or more people in the video (M = 5.00, SD = 1.73) perceived reactions as more positive compared to those who knew no one (M = 4.00, SD = 0.82). Regardless, the analyses indicated that the videos successfully manipulated perceived 28 norms. Thus, the six videos (one for each norm-ending combination) were used in their original form in Study 2 due to the successful manipulation of norms which was not moderated by ending type. Measures Oliver and Batsch’s (2010) enjoyment (M = 4.65, SD = 1.42; α = .79) and appreciation scales (M = 4.89, SD = 1.31; α = .72), the dichotomous measure of liking (M = .80, SD = .40), evaluation times (Mraw = 73.50, SDraw = 115.87; Mlog = 1.51, SDlog = 0.53), and demographic measures from Study 1 were repeated in Study 2. Scales new to Study 2 are described below. All but one (see below) were presented in counterbalanced order after the review was completed. Items were measured on a scale from 1 (“Strongly disagree”) to 7 (“Strongly agree”) and scales were computed by taking the average of all scale items. Perceived Norms Perceived descriptive norms (three items; e.g., “Most other people in my webinar session like this story”; M = 4.66, SD = 1.64; α = .89) and injunctive norms (four items; e.g., “I believed most people in my webinar session would approve of me liking this story”; M = 4.96, SD = 1.17; α = .76) were measured using adapted versions of Park and Smith’s (2007) scales. Additionally, perceived social sanctions were measured using an adapted version of Liu’s (2017) six-item measure (e.g., “I worried that if I didn’t like this story, other people in my session would be disappointed; M = 3.47, SD = 1.77; α = .94). Higher scores indicate a stronger perception of norms or sanctions. Descriptive norms were measured after the manipulation, whereas injunctive norms and sanctions were measured after the review. 29 Perceived Publicness Perceived publicness of the situation (M = 4.55, SD = 1.46; α = .82) was measured using an adapted version of Chung and Lapinski’s (2018) scale. The scale contains four items (e.g., “I can see whether other people like the story”), with high scores indicating that the behavior was perceived as more public. I also measured whether participants focused on the audience (7) or story content (1) with a one-item bipolar scale. The midpoint was “Focused on both equally” (M = 3.89, SD = 1.68). Social Sharing After completing both measures of enjoyment and appreciation, participants rated the story from 1 to 10 and wrote a review of the story. All participants were told that the review and score would be shared with future participants. Reviews were coded for length (number of words; M = 19.66, SD = 18.49; Min. = 1, Max. = 117; Eden et al., 2017) and the valence of the rating (1-10; M = 7.41, SD = 2.16). Group Identity Group identity (M = 4.88, SD = 1.10; α = .89) was measured using Chung’s (2020) scale, which adapts measures from previously established measures (e.g., Crocker & Luhtanen, 1990; Rimal & Real, 2005). The scale contains nine items that capture the extent to which participants identify with other people in their webinar session (e.g., “I feel strong ties with other people in the session”). High scores indicate a stronger identification with the referent group. Though identification with strangers may seem farfetched, minimal group paradigms elicited normative and identity-consistent effects in the past (Deutsch & Gerard, 1955; Gagnon & Bourhis, 1996). 30 Self-presentation Motivation In addition to group identity, participants completed an 11-item scale measuring self- presentation motivation developed based on the own-ideal self condition used by Johnson and Ranzini (2018; e.g., “When rating the story, I wanted to make a good impression on other participants”) and extant impression management scales (Dillard et al., 1989; Neubaum & Krämer, 2017; e.g., “I was not willing to risk possible damage to my relationships with others based on the way I rated the story”). High scores indicate greater self-presentation motivation. An eight-item version of the measure was used (described below; M = 3.83, SD = 1.63, α = .93). Results Confirmatory Factor Analysis Confirmatory factor analyses (CFA) were conducted to ensure the validity of the data (Hunter & Gerbing, 1982). Five models were tested, one each analyzing: (1) enjoyment and appreciation, (2) perceived descriptive norms, injunctive norms, and social sanctions, (3) perceived publicness, (4) group identity, and (5) self-presentation motivation. Factors were specified using the scale structures listed in the measures and alternative models were probed. As evidenced by the fit indices (Table 5), all scales except the social motivation scale displayed average or better fit for the data. For the self-presentation scale, the eight-item solution omitting items 3, 7, and 8 was the best fit for the data. Otherwise, though certain items’ factor loadings were below 0.80, scales were computed as planned. Descriptive Analysis and Covariate Tests Chi-square analysis (ethnicity/race, gender) and ANOVA (age) were used to determine if demographics were equally distributed across conditions. Those self-identifying as Hispanic/Latinx were significantly more likely to be assigned to the negative norm condition (n 31 = 7, adj. std. R = 2.4) compared to the control (n = 2, adj. std. R = 0.8) or universal acclaim condition (n = 1, adj. std. R = -1.7; χ2(2) = 6.05, p = .049, Cramer’s V = .16).4 T-tests (summary in Appendix F) and correlations (Table 8) were used to determine significant relationships between measured variables and demographic data. Liking, enjoyment, and descriptive norms did not differ by gender or race. However, differences in dependent and explanatory variables were observed by gender, race/ethnicity (black, Hispanic/Latinx), and age. Also, attentional focus (i.e., whether participants were paying attention to the story content or the confederates) was significantly associated with many variables. Given the results, attentional focus, gender,5 and dummy variables indicating self-identification as Black/African American and Hispanic/Latinx were noteworthy covariates. Focus was used as a covariate in all analyses, and models featuring control variables are noted below. Manipulation Checks The success of the manipulation was evaluated using a 3 (norm cue: universal acclaim, overwhelming dislike, control) x 2 (ending: positive, mixed) ANOVA predicting perceived descriptive norms. There was a significant main effect of norm cues on perceived descriptive norms, F(2, 226) = 12.64, p < .001, η2 = .10. A post hoc Bonferroni revealed that perceived norms in the universal acclaim (M = 5.21, SD = 1.43) and control condition (M = 4.76, SD = 1.49) were significantly higher than the overwhelming dislike condition (M = 4.01, SD = 1.75), but the universal acclaim and control conditions did not significantly differ from one another. There was no significant main effect of endings, nor did endings and norms interact to predict 4 The result may be driven by the low number of Hispanic/Lantix participants, but since this was the only significant failed randomization based on race/ethnicity, I retained the Hispanic/Latinx variable as a covariate. 5 For analysis purposes and given the small number of participants not identifying as fe/male, analyses with gender as a covariate were analyzed first with non-binary/transgender and those who chose not to identify grouped with women and significant findings were reconducted with non-binary/transgender and non-disclosers as distinct groups. 32 perceived descriptive norms (p > .19). Adding relevant covariates did not affect significance levels. The results point to a successful norm manipulation. The analysis was repeated with perceived injunctive norms, social sanctions, publicness, group identity, and self-presentation motivation as dependent variables. Ending was the only significant predictor of perceived injunctive norms, F(1, 226) = 6.86, p = .01, η2 = .03, such that participants perceived more approval for liking the positive (M = 5.17, SD = 1.11) than the mixed ending (M = 4.77, SD = 1.19). The result remained significant after applying controls. No other significant effects were observed. Hypothesis Testing H1, H2, H3, and H4, which regarded the interaction effects of norms and endings on enjoyment and appreciation, were tested using a series of 2 (ending: positive, negative) x 3 (norm cues: universal acclaim, overwhelming dislike, control) ANOVAs, which were repeated for each dependent measure (liking, raw and log-transformed evaluation times, enjoyment, appreciation). Differences in means and effects between Study 1 and Study 2, which may have been caused by making norms salient before the story and conducting the study in a quasi-public setting, are discussed to answer RQ1 and partially answer RQ2, respectively (Table 6, Table 7). Unlike Study 1, few significant main effects of ending or norms on dependent measures were observed in Study 2. Endings significantly affected enjoyment and liking such that positive endings were enjoyed and liked more than mixed endings. Endings did not significantly affect evaluation times or appreciation. Neither norms nor the interaction of conditions affected any dependent measures (Table 6).6 Thus, H1, H2, and H4 all failed to replicate in Study 2. Similar to Study 1, H3 was not supported by the data. 6 Tests with evaluation times (raw, log-transformed) and word count were repeated with variables that replaced outlier scores with mean values for each scale, but the test statistics’ significance was unchanged. 33 To answer RQ3, the same ANOVA was repeated with review score and number of words as dependent measures. There was a significant main effect of ending in predicting review scores such that positive endings scored higher than mixed endings. However, neither norms nor the interaction of norms and endings influenced review scores. There were no significant effects of conditions on the number of words written (Table 7). Thus, the results suggest that endings lead people to share higher ratings of stories, but norms have no effect on their reviews. RQ1 and RQ2: Differences from Study 1 There was one key difference in effects and means between the studies: all dependent variables of interest—liking, enjoyment, appreciation, and evaluation times—had higher values in Study 2 than Study 1, even after controlling for associated covariates (Table 3, Table 6). Furthermore, the starkest differences in appraisal variables across the two studies manifested in cells wherein norms and/or narrative cues were in conflict (i.e., overwhelming dislike—positive ending condition and all mixed ending conditions), whereas the differences in non-conflict conditions were less pronounced. Evaluation time differences were even more extreme: evaluation times across conditions in Study 2 were 68.08 seconds longer than in Study 1. These findings have major implications for interpreting and analyzing the results of Study 2. Evaluation times suggested that intuitive reactions in Study 2 were subsumed under deliberate processes. Also, Study 2 participants evaluated conflict conditions considerably more favorably than Study 1 participants, whereas differences in non-conflict conditions are less pronounced. Together, the results indicate that presenting a norm cue before story exposure and the public setting motivated processing about all elements of the experimental session, not just the narrative. However, in line with Carey et al. (2010) and Kryston and Eden (2021), perceived norms could be modeled as a mediator of the norm cue–evaluation relationship. Thus, I 34 conducted alternative analyses of H1 through H4 and RQ3 with perceived descriptive norms as a mediator and ending condition moderating the effect of perceived norms on evaluations and social sharing. Tests of RQ2 and RQ4 replicated the models but also examined the moderating role of perceived publicness, group identity, and self-presentation motivation. Mediation Analyses Alternative Tests of Hypotheses H1 and H2 posited that norms reinforce the enjoyment when matching, and evoke appreciation when conflicting, with positive endings. H3 and H4 were competing hypotheses suggesting that norms affect mixed ending appreciation (H3) or enjoyment (H4). Alternative tests of these hypotheses were conducted using a moderated-mediation analysis (PROCESS Model 14; Hayes, 2017) with 10,000 bootstrap-corrected samples. Norm condition (dummy coded as 1 = overwhelming dislike, 2 = control, 3 = universal acclaim) was the multicategorical indicator exogenous variable, perceived descriptive norms was the mediator, and ending condition (dummy coded as 1 = mixed, 2 = positive) was the moderator of the relationship between perceived descriptive norms and appraisals (see Figure 7). All relevant control variables (focus, age, race/ethnicity=black, race/ethnicity = Hispanic/Latinx, gender) were modeled as covariates. The test was repeated with enjoyment, appreciation, liking, and log-transformed evaluation times as the dependent variable. At step 1, norm condition significantly predicted perceived descriptive norms when comparing the control and universal acclaim condition to the overwhelming dislike condition, respectively (Table 9). The results reflect a successful manipulation of perceived norms. Regarding H1 and H4, the results at step 2 revealed that perceived descriptive norms were a significant predictor of enjoyment and liking. However, ending condition had neither 35 main (p > .06) nor interaction effects (p > .19) on enjoyment or liking. Likewise, there were no main effects of norm condition on outcomes. The indirect effect of norm condition on enjoyment and liking through perceived norms was significant for both positive and mixed endings (Table 10). Although the index of moderated mediation was not significant in either case (p > .32), the indirect effect on enjoyment was stronger for mixed versus positive endings (Figure 8), whereas the effects on liking were largely similar across endings (Table 11). Turning to H2 and H4, perceived norms significantly predicted appreciation and log- transformed evaluation times at step 2. Again, ending condition had neither main (p > .21) nor moderating effects (p > .17) on these outcomes, nor did norm condition directly predict appreciation (p = .40). However, there was a significant main effect of norm condition on evaluation times such that those in the overwhelming dislike condition reacted significantly faster than those in the universal acclaim condition. Perceived descriptive norms mediated the norm condition–appreciation relationship similarly across positive and mixed endings, but indirect effects on evaluation times were only significant among mixed endings (Table 11). The latter results offer further credence to H4 by showing that perceived norms increased reaction times to mixed-ending stories (i.e., resolved narrative conflict) but not positive endings. To summarize these findings, alternative analyses offered no support for H1, H2, or H3. However, H4 received additional support: norm cues affect audiences’ enjoyment versus appreciation of mixed endings (narrative conflict). Publicness, Group Identity, and Self-presentation Motivation To address RQ2 and RQ4b, which concern the moderating role of perceived publicness and social identity, I conducted mediated-moderation analysis (PROCESS Model 18), which replicated the model testing H1–H4 and RQ3, but added perceived publicness as a moderator of 36 perceived norms and endings (Figure 7). Again, the results were repeated on each appraisal type and the results at step 1 were the same as observed for tests of H1–H4. Three noteworthy interactions were observed at step 2 across these tests: three-way interactions between perceived descriptive norms, endings, and (1) publicness in predicting liking and review scores (Table 13) and (2) identity predicting liking (Table 16). The conditional direct and indirect effects indicate that perceived publicness (Table 14; Figure 9) and social identity (Table 16; Figure 10) amplified the effect of perceived norms on liking, but only among mixed ending stories. Just as in Study 1, in the absence of narrative conflict (i.e., positive endings), neither norms nor identity were capable of increasing enjoyment, suggesting a ceiling to which content can be liked. No significant three-way interactions were observed in models predicting enjoyment, appreciation, evaluation times, or number of words in the review. Turning to RQ4a, which was tested using a moderated-mediation model where self- presentation motivation moderated the a path and ending moderated the b path (PROCESS Model 21; Figure 7),7 self-presentation motivation had a significant main effect on perceived norms and significantly moderated the effect of norm condition on perceived descriptive norms (Table 19; Table 20; Figure 11). At step 2, the effects are the same as those reported for H1–H4 and RQ2 across dependent variables (Table 9). The conditional indirect effects show that the effect of norm cues on all outcomes except evaluation times was only significant if self- presentation motivation was at or below the mean (Table 21, Table 22). For evaluation times, the aforementioned interaction manifested only in the positive ending condition. In sum, high levels 7 A series of moderation analyses were conducted to determine if self-presentation is more appropriate to model as a moderator to the a or b path. Although this sort of exploratory analysis is usually frowned upon, given the variable’s novelty and the changes to the analyses, I felt it important to present the best possible model including self- presentation motivation to help guide future work. The results of RQ4a (like any, but especially here) should be questioned and challenged by attempted replications in future work. 37 of self-presentation motivation made participants insensitive to norm cue manipulations, thereby negating downstream effects on appraisals. Social Sharing Lastly, I answered RQ3, which asked how norms and ending interact to predict social sharing, by repeating the analysis of H1 through H4 with review score and length as the dependent variables. Model statistics are identical to tests of H1–H4 at step 1. Perceived descriptive norms were the only significant predictor of review scores at step 2 (Table 10). As with enjoyment, the indirect effect of norm condition on review scores was stronger among those in the mixed ending condition, but the index of moderated mediation was not significant. Perceived norms, ending, or their interaction did not predict word count. The indirect effect of norm cues on words through perceived norms was only significant among those in the mixed ending condition, suggesting that perceived norms explained the preponderance of those in the overwhelming dislike condition to write longer reviews for mixed-ending stories (Table 12). 38 DISCUSSION This manuscript presented the results of two studies that tested the role of social cues, specifically those communicating social norms, in the narrative enjoyment and appreciation rationale (NEAR). Study 1 found that norm cues indicating universal acclaim increased liking and enjoyment of positive endings compared to norm cues indicating dislike or the absence of norm cues. Examination of how these norm cues affected audiences’ appraisals of narrative conflict (mixed endings) revealed that social cues influenced audiences’ enjoyment (or disliking) but did not affect appreciation responses. Study 2 conceptually replicated Study 1 but made norms salient before narrative consumption and tested these effects in a simulated public environment. Under these conditions, few effects from Study 1 were replicated. However, Study 2 did find that (1) social norms seem more relevant to enjoyment than appreciation of narrative conflict, (2) the procedural differences largely affected processing times, and (3) publicness, social identity, and self-presentation motivation moderated normative influence in ways consistent with past literature. The results of Study 1 and Study 2 are discussed in terms of their individual and shared impact on a broader understanding of the effects of social cues on content appraisals. Study 1 According to the NEAR, intuitive enjoyment responses are the product of the satisfaction of all salient needs. Regardless if one or many needs are satisfied, the NEAR argues that audiences experience enjoyment so long as need conflict is absent. Indeed, in Study 1, when conflict was absent because norms reinforced positive endings, audiences experienced 39 enjoyment. This finding replicates past work (Kryston & Eden, 2021) and supports the NEAR’s perspective on narrative appeal (Tamborini et al., 2021). Social cues appear to act similarly to content cues in determining audiences’ appraisals of mixed-ending narratives. The presence of norm cues indicating universal acclaim increased enjoyment and liking, whereas cues indicating overwhelming disliking decreased enjoyment and liking beyond the control. The results largely mirror the NEAR’s distinction between dominant and overriding salience (Tamborini et al., 2021). Overriding salience refers to when satisfied and thwarted needs are both salient (as was the case with mixed endings in the norms control condition) and elicit appreciation. On the other hand, dominant salience refers to when needs are in conflict, but only one is salient, thereby producing enjoyment or disliking. Tamborini et al. (2021) argue that narratives give satisfied needs dominant salience through various means, such as by diminishing the importance or salience of the thwarted intuition, using cues to alter perceived attribution of cause, or having the protagonist (and by extension, the audience) satisfy additional needs. Given that the salience of narrative and social cues should have been similar in this study, I argue that norm cues indicating universal acclaim shifted the salience of satisfied needs from overriding towards dominant salience. The effect of social norm cues on mixed-ending enjoyment indicates that, similar to narrative cues, norm cues generate intuitive, positive responses to entertainment. As such, the results support the integration of social cues into the NEAR to explain social influences on entertainment appraisals given their complementary mechanisms related to human motivation, salience, and need satisfaction. 40 Study 2 However, many of the takeaways from Study 1 are undercut by the failure to replicate most findings in Study 2. Besides H4, no hypotheses received support. However, Study 2’s procedures may have been ill-suited to examine the NEAR’s specific claims about narrative appeal. Notably, the results indicate an inability to differentiate between intuitive and deliberative responses; evaluation times suggested that all appraisals were made deliberately, or at least with ample time for deliberative processing to occur. Under these conditions, evaluations were more positive than in Study 1, even when conflict between narrative and/or social cues was present. It appears that increasing participants’ motivation and opportunity to process narrative and social cues led to a failure to elicit the expected reactions in participants. Furthermore, other procedural changes, including incentives (discussed below) and the relative ease of processing cues across different modalities (e.g., audio versus visual) versus the same modality (e.g., two visual cues; Fisher et al., 2018), may have exacerbated deliberation. Regardless, since procedural changes mostly increased processing times, Study 2’s design seems less suited than Study 1’s to detect the desired outcomes. Yet, various results of Study 2 still point to the utility of integrating logic from the social norm literature to understand narrative appeal broadly and the NEAR’s claims specifically. First, Study 2 used a different social cue, but conceptually replicated both an effect of social norm cues on perceived descriptive norms and the indirect effect of these cues (through perceived norms) on appraisals (see Kryston & Eden, 2021). Further highlighting the utility of applying social norms literature to understand narrative appeal, three constructs identified in the social influence literature—publicness, social identity, and self-presentation (Rimal & Lapinski, 2005; Chung & Rimal, 2016)—moderated the indirect effect of norm cues on narrative appraisals. Unlike Study 41 1, where the social norm cue could easily be interchanged with other social influences (e.g., critic credibility, interpersonal influence), Study 2 suggests that the construct affecting narrative enjoyment was indeed social norms, specifically descriptive norms. Additionally, two findings were consistent across both studies. First, social cues had a greater effect on enjoyment (rather than appreciation) of mixed endings, suggesting yet again that social cues generate intuitive, positive affect in audiences, just like narrative cues. Second, social norm cues played less of a role in the enjoyment of positive endings. Furthermore, though publicness, identity, and self-presentation motivation increased norm effects on enjoyment of mixed endings, these positive moderators nullified norm effects of positive endings. Taken together, the results suggest that when conflict is absent, social norm and other contextual cues cannot increase audiences’ enjoyment of narratives in which all salient needs are already satisfied. I interpret the successful manipulation of norm effects, the replication of past literature, and suggestions of parallel mechanisms underlying social and narrative cue effects as reasons to further pursue the integration of social norms into the NEAR. General Discussion Although various studies indicate a role of social cues on entertainment (e.g., Shedlosky- Shoemaker et al., 2011; Waddell et al., 2020), to my knowledge, scholars lack a general understanding of the mechanisms and processes by which social cues affect said appraisals. This study addressed this gap by exploring how social cues alter audiences’ appraisals of enjoyable and meaningful entertainment. Findings suggest that the processes by which social cues moderate narrative appraisals are similar to the NEAR’s claims describing how the interaction of multiple narrative cues in a given story affect appraisals (Lewis et al., 2014; Tamborini et al., 2021). Given that social influences, including norms, interpersonal opinions, and expert 42 consensus, are guided by individuals’ motivations to maintain relationships and/or make correct decisions (Cialdini et al., 1990), social cues might influence enjoyment because adhering to others’ opinions offers the opportunity to satisfy intuitive needs. Together, these studies also shed light on the psychological mechanisms of the NEAR and its governing model, the MIME. Assuming that narrative and social cues were equally salient given the texts’ similar length and sample demographics, in Study 1, I found that the addition of salient needs incrementally influenced audience response. In line with recent calls to model communication processes computationally (e.g., Huskey et al., 2020), I suggest that modeling the effect of narrative and social cues on appraisals mathematically may be possible. For example, in both this study and Lewis et al. (2014), one could model the product of cue salience (both within individuals and as affected by the stimulus) multiplied by a dummy factor where need satisfaction produces a positive score (+1) and thwarting produces a negative score (- 1) to determine the effect of an individual narrative or social cue on audience appraisals (represented as enjoyment divided by evaluation time). The sum of the ith cues determines the audience’s overall response. Additionally, given the results of Study 2, motivation and opportunity should be considered as moderators or necessary conditions for any computational model of narrative appeal. While refining the equation and identifying valid operationalizations of cue salience may prove difficult, testing computational models can help develop a verifiable, unifying framework to study social and/or narrative cue effects across a wide range of contexts, ranging from finite examinations of brain response to stimuli to large-scale text analyses. Relevant to both media and social influence scholars, the results in the mixed-ending condition across both studies suggest that intuitive processes govern social cue effects. This claim is reinforced by the fact that, in Study 1, despite the influence of norms on participants’ 43 enjoyment and appreciation, few of them mentioned norm cues in the thought-list task. The suggestion of norms’ intuitive influence is consistent with social influence work (Cialdini et al., 1990; Ewoldsen et al., 2015). But, as evidenced by Study 2 and other social influence research (e.g., Carey et al., 2010), normative perceptions also influence judgments and social cues are deliberated on to influence outcomes such that peoples’ perceptions and cognitive thoughts about norms can affect their appraisals of stimuli and behavioral outcomes. Purposeful examinations of how and when intuitive versus deliberative judgments occur would be a good follow-up, perhaps by analyzing think-aloud data while participants make appraisals, or employing the thought listing procedure in a replication of Study 1. Additionally, these studies also suggest a more comprehensive way of indexing motivations in social norm theories. Although most theories of normative influence suggest that individuals’ motivations, the motivation to adhere to norms, and needs fulfilled by performing certain behaviors impact the persuasive outcomes of normative messages and perceived norms (e.g., Cialdini et al., 1990; Chung & Rimal, 2016), research showing how and why these motivations guide normative effects remains limited. I suggest that norm adherence should tap into people’s motivation to satisfy intrinsic needs, thereby generating positive affect. However, need salience and satisfaction may differ based on the nature of normative information. For example, the descriptive norm manipulations featured in these studies likely serve an informational role (Deutsch & Gerard, 1955); thus, their role in affecting enjoyment might be related to competence concerns: audiences may consider liking a universally acclaimed film to be the “right decision” (Cialdini et al., 1990). By contrast, in certain situations, norm messages may affect appraisals based on a need to maintain relationships. In these instances, norm messages may heighten the salience of intuitive needs which must be upheld to reify in-group 44 bonds and/or avoid potential ostracization. For example, when the group communicating social norms greatly values fairness, fairness may play a greater role in determining the appeal of narrative cues versus other intuitions. To explore these possibilities, future work can explore (1) which motivations are made salient by norm messages and (2) to what extent adhering to norm satisfies salient needs. These studies may add more nuance and detail to norm theory claims about motivation and outcome expectations. Limitations These studies do have various limitations. First, although Study 1 replicated Lewis et al.’s (2014) findings regarding differences in liking and evaluation times between positive and mixed endings, Study 1 did not replicate findings for self-reported appreciation when two cues were in conflict (i.e., when norms conflicted with positive endings and lack of differences between the positive- and mixed-ending story in the norm control condition). In Lewis et al. (2014), appreciation was greater for mixed endings compared to positive endings, but in these studies, appreciation and enjoyment were higher for positive endings compared to mixed endings when norm cues were absent. Also, in Lewis et al. (2014), evaluation times and self-reported appreciation varied together between positive and mixed endings, whereas in both studies presented in this manuscript, liking, self-reported enjoyment, and appreciation followed the same pattern. I see two potential explanations. First, past work validated Oliver and Bartsch’s (2010) enjoyment and appreciation scales across various media (Weinmann et al., 2016), but not the relationship between the scales when considering social contexts. When social cues are presented with content these scales show strong positive correlations (r > .60; Kryston & Eden, 2021), and the scales’ fit in Study 1 deteriorated for mixed endings associated with universal acclaim. 45 Therefore, social context may moderate how these scales are interpreted, especially the extent to which a socially consumed story is “thought-provoking” and, by extension, appreciated. This claim is further supported by the inflated appreciation scores in Study 2’s “public” setting. On the other hand, enjoyment and appreciation correlate differently across entertainment studies (e.g., positively in Grizzard et al., 2019; Yoshimura et al., 2017; negatively in Bartsch & Hartmann, 2017). Though Oliver and Bartsch’s (2010) appreciation scale has proven useful in various contexts and helped bridge the gap between conceptual disagreements (see Lewis et al., 2014), I propose that further efforts are warranted to ensure measures of enjoyment and appreciation are sufficiently distinct, especially given the poor measurement model structure under certain conditions. Focusing on the “thought-provoking” nature of appreciated films versus the extent to which they are moving or meaningful may be a good first step given the results of the CFA within each condition (Study 1). Second, although evaluation times had a similar distribution and mean pattern by condition to those observed in Lewis et al. (2014), mean evaluation times were much longer in Study 1 than in Lewis et al. (2014), and even longer in Study 2. Additionally, evaluation time outliers were more frequent and had more extreme values than expected in Study 1. I suspect that these issues were a product of both the setting and instruments used in the studies. Lewis et al.’s study collected data in a lab, whereas the current study was moved online due to COVID-19 restrictions. As such, participants may have been less focused. For this reason and because evaluation time measures are correlated with participants’ internet connection and processor quality, I caution the interpretation of the evaluation time results in either study and suggest a replication of the current study in a laboratory setting. 46 Third, our narrative stories are short (six to seven sentences). While this helped control the relative salience of social versus content cues, I cannot be certain about how the results generalize to explain norm-influenced appraisals of long-form media (e.g., feature-length films). Speaking to this concern, in Study 2, when there was greater incentive (financial compensation, vivid social cues, potential experimenter bias since I narrated the story), motivation (participants were reminded to pay attention multiple times), and time to process (participants knew they would consume one versus six stories), liking, enjoyment, and appreciation all greatly increased compared to Study 1. Perhaps when they had time to think, participants found more reasons to positively appraise the story. The same might be true for many long-form media, especially in social settings. Short stories may afford experimental control, but tests of the NEAR across various kinds of narratives and social environments are warranted. Fourth, beyond procedural differences, the differences in results between Study 1 and Study 2 may also be a product of the different samples. Study 1 should be replicated among a non-student sample, whereas Study 2 should be replicated among students to fully disentangle theoretical implications from socio-demographic idiosyncrasies. Conclusion and Future Directions In sum, social cues play an important role in moderating narrative appraisals, likely because adhering to social cues (e.g., norm cues) can satisfy intuitive needs. Although seminal media research has implicated social gratifications as important determinants of media preference (e.g., Herzog, 1944; Rubin, 1983) and bandwagon effects on entertainment appeal and processing have been observed in countless studies (e.g., Basuroy et al., 2003; Möller et al., 2019; Waddell et al., 2020), the mechanisms by which social cues affect media appraisals had not been fully explicated. Thus, Study 1 serves as the first step towards a cohesive theory of how 47 social norms and other social cues influence entertainment processing. However, more work should be done to (1) disentangle the mechanisms of various social cue effects (e.g., Study 2) and (2) explore whether social cues help explicate other macro- and micro-level media processes. Future work can identify which social cues moderate NEAR processes by (1) performing conceptual replications or indexing how various social cues are conceptually interpreted, (2) manipulating and/or measuring known moderators of a given social influence (e.g., in-group identity for norms, credibility and trust for expert consensus), and (3) testing computational models of appraisals against observed appraisals. For example, this manuscript focused on social norms as a social cue moderating NEAR processes, but I expect that had a close friend (interpersonal influence) or most critics (expert consensus) dis/liked the film, the results would be largely similar to those observed in Study 1. Likewise, while entertainment studies have used various text and audiovisual norm manipulations (e.g., Kryston & Eden, 2021; Kryston et al., 2019; Park et al., 2020), these operationalizations represent only a small segment of the cornucopia of social cues in common viewing environments. Regardless of the type of social influence driving the effect of a social cue, exploring how adherence to these social cues satisfies salient needs requires attention. Entertainment’s ability to satisfy needs is well-documented (Tamborini et al., 2010), but the same is not true for social cues (cf. Grayson et al., 2019). Thus, future work should consider the extent to which social cues make specific psychological needs salient, and whether these needs are satisfied by adhering to or defying the social cue. Doing so would not only confirm the assumption that social cues influence narrative processing by offering an avenue by which to satisfy intuitive needs, but may also help explain how adhering to (or defying) social cues affects subsequent need salience (Eden et al., 2014) and behavioral coping strategies (Bowman & Tamborini, 2012). 48 Additionally, exploring the role of social norms in narrative processing might further clarify media’s role in shaping “morality subcultures” and other groups (Zillmann, 2000). The results point to two ways of examining this phenomenon. First, in Study 2, endings, but not norm cues, affected perceived injunctive norms (others’ approval of liking the story) such that positive endings were associated with a greater extent of perceived approval than mixed endings. Thus, it seems that whether salient altruistic intuitions are upheld (e g.., the positive ending) or violated (e.g., the mixed ending) may drive the perception that liking said content will result in social rewards or punishments. Second, normative perceptions, norm cues, and endings were associated with social sharing, indicating that narrative and social cues may perpetuate (or stymie) with in- group and/or out-group members (also discussed in Bartsch et al., 2008). Examining the covariance of perceived injunctive norms, consumption trends, and sharing behavior surrounding content that upholds or violates different intuitions may help explain how morality subcultures and their norms are formed, managed, and maintained via media consumption. Lastly, these studies borrowed from the logic of the MIME (Tamborini, 2013), and integrating social cues into the model might help shed light on social elements of the MIME that have received less empirical attention (Eden et al., 2021). For example, the MIME suggests that individuals’ appraisals (explicated by the NEAR) lead to selection behaviors (Stage 4). These selection behaviors subsequently influence cultural production trends (Stage 5), which in turn affect exemplar exposure and the “cultural environment” (Stage 6; see Figure 1). Despite precisely laying out many MIME processes, the way individuals’ selection leads to cultural trends and what constitutes the “cultural environment” garner far less attention. I suspect that studying social cues, specifically social norm (i.e., aggregated selection behaviors) and contextual cues (e.g., signs, ads, and other environmental artifacts), could simultaneously 49 provide additional nuance to the MIME and other models of media influence and/or develop an understanding of social cues’ role not only in narrative processing, but also in selection behaviors, production trends, and morality subcultures. 50 APPENDICES 51 APPENDIX A: Narrative stories Note: All stories and endings are from Lewis et al. (2014). Intuitive needs named in the MIME (Eden et al., 2021; Tamborini, 2013) made salient in each story are noted in parentheses afer the story’s title. Saving Civilians (Care vs. In-group Loyalty) Enemy soldiers have taken over Ava’s village. They have orders to kill all remaining civilians. Ava and some of her townspeople have sought refuge in the cellar of a large house. Outside she hears the voices of soldiers who have come to search the house for valuables. Her baby begins to cry loudly. She covers his mouth to block the sound. If she removes her hand from his mouth his crying will summon the attention of the soldiers who will kill her, her child, and the others hiding out in the cellar. Positive: Ava comforts and quiets the baby so the group is saved. Mixed-positive: Ava suffocates the baby to save the group. Mad Dog (Care vs Fairness) Bobby's sister, Sarah, was killed by Damon, a psychopathic troublemaker who lives in the outskirts of town. Damon mercilessly beat Bobby's sister to death and boasted about it in a drunken stupor. While Damon evades the law, Bobby seeks revenge for his sister. They encounter each other on a bridge and find themselves in a situation where Damon is about to fall on the rocks below. Bobby holds Damon's hands as he is dangling off the bridge. He is in complete control over whether Damon lives or dies. Positive: When Damon boasts about his murder, he slips and falls to his death. Mixed-positive: Bobby really wants revenge, but is too virtuous to kill Damon. 52 Vacation’s Over (Authority vs Fairness) David and his family are on vacation in Bali. While on a train, some armed officers with a sniffer dog become suspicious of David's teenage son and his backpack. The officer discovers a small amount of marijuana in the backpack, which David was unaware of. The officer asks who the marijuana belongs to but David hesitates out of confusion. In Bali, he would receive life in prison or the death sentence for just a small amount of marijuana. David and his wife both realize this, and his son looks like he is about to speak up and turn himself in. Positive: Luckily, the armed officers are distracted by a petty thief and leave David's family alone. Mixed-positive: David takes the blame for the marijuana and is left not knowing whether he will spend his life in prison. Safari (In-group Loyalty vs Authority) Julie is part of a group of archeologists who live in a remote stretch of jungle. The entire group, which includes eight children, has been taken hostage by a group of paramilitary terrorists. One of the terrorists takes a liking to Julie. He informs her that his leader intends to kill her and the rest of the hostages the following morning. He is willing to help Julie and the children escape, but as an act of good faith he wants Julie to kill one of her fellow hostages whom he does not like. Positive: Julie refuses. She kills the terrorists and the children escape. Mixed-positive: Julie agrees. She sacrifices her friend to let the children escape. Saving Comrades (Care vs in-group Loyalty) John is the leader of a small group of soldiers. He is on his way back from a completed mission deep in enemy territory when one of his men has stepped into a trap that has been set by the 53 enemy and is badly injured. The trap is connected to a radio device that by now has alerted the enemy to their presence. They will soon be on their way. If the enemy finds John’s injured man, they will torture him and kill him. He begs John not to leave him behind, but if John tries to take him along his entire group will be captured. Positive: John orders his men to carry the injured man, and the whole group luckily survives. Mixed-positive: John shoots the injured man to save him from torture. The others survive. Keller’s Firm (Fairness vs In-group Loyalty) Mr. Keller is in charge of filling a position in his firm. His friend, Joey, is qualified for the position, but another applicant is more qualified. Mr. Keller feels guilty that he wants to hire his friend. He feels that he should be fair and impartial in his decision despite his loyalty to his friend. Yet if he doesn't hire his friend Joey, he knows that Joey will be angry. Positive: At the last minute, the other applicant drops out and Joey gets the job anyway. Mixed-positive: Mr. Keller cannot let himself hire Joey, and they lose their friendship. 54 APPENDIX B: Sample story stem screen 55 APPENDIX C: Sample manipulations Note: Top image is the positive ending—norm combination (reinforcing) for story stem “Saving Civilians” and the bottom image is the positive ending—no norm combination (conflicting). Endings and norms were displayed simultaneously after reading the story stem, response time was recorded based on the time it took for participants to select “like” or “dislike” to the resolution screen. The buttons appeared three seconds after advancing to the ending screen. 56 APPENDIX D: All items for Study 1 All items measured on a scale from 1 (Strongly disagree) to 7 (Strongly agree) unless noted otherwise. Story liking 1. Did you like the story? 1 (Like) – 0 (Dislike) Enjoyment (Oliver & Bartsch, 2010) Please indicate your agreement with the following statements. 1. It was fun for me to read the story. 2. I had a good time reading the story. 3. The story was entertaining. Appreciation (Oliver & Bartsch, 2010) Please indicate your agreement with the following statements. 1. I found this story to be very meaningful. 2. I was moved by the story. 3. The story was thought-provoking. Thought listing (Cacioppo, Glass, & Merluzzi, 1979). “We are now interested in everything that went through your mind when you rated the story. Please list these thoughts, whether they were about yourself, the story, others, or anything else you were thinking about; whether they were positive, neutral, and/or negative. Ignore spelling, grammar, and punctuation. Please list at least one thought, and you can write as many thoughts as you like until you are done.” Responses will be coded for whether the participant’s focus was on social cues or not. Demographic information What gender do you identify as? Male Female Transgender Non-binary My gender identity is not listed (please specify) Prefer not to answer What race/ethnicity do you identify as? Select all that apply. 57 Asian/Asian American Black/African American Hispanic/Latinx Indian Native American Pacific Islander/Hawaiian Middle Eastern/Arabic White/Caucasian Other (please specify) Prefer not to answer How old are you (in years)? (open-ended numeric response) 58 APPENDIX E: All items for Study 2 Story liking 1. Did you like the story? 1 (Like) – 0 (Dislike) Note: Response times are also measured. Enjoyment (Oliver & Bartsch, 2010) Please indicate your agreement with the following statements. 1. It was fun for me to listen to the story. 2. I had a good time listening to the story. 3. The story was entertaining. Appreciation (Oliver & Bartsch, 2010) Please indicate your agreement with the following statements. 1. I found this story to be very meaningful. 2. I was moved by the story. 3. The story was thought-provoking. Social sharing Thank you for answering those questions. Now, we would like you to write a short review of the story. We will post this review to a message board so that other participants can read what you wrote. Be sure to say whether you liked the story and give a reason why. (open-ended response) Responses will be coded for (1) length by recording the number of words and (2) valence by independent coders using two seven-point scales that assess how positive the review is (“Based on the review, the reviewer has positive feelings about the story”) and now negative the review is (“Based on the review, the reviewer has negative feelings about the story”). Perceived norms The post on the prior screen was the way that other people felt about one of the stories you read. With this in mind, please indicate your agreement with the following statements. Descriptive norms (adapted from Park & Smith, 2007) 1. Typical people in my webinar session like this story. 2. Most other people in my webinar session like this story. 3. The majority of other people in my webinar session like this story. Injunctive norms (adapted from Kryston, Park, & Eden, 2020; Liu, 2017) Approval 59 4. I believed most people in my webinar session would approve of me liking this story. 5. I felt like most other people in the session would endorse me liking this story. 6. Most other people in my webinar session would not oppose me liking this story. 7. The majority of people in the session would think it’s acceptable to like this story. Sanction 8. I thought other people in my webinar session would think less of me if I liked this story. 9. I was worried that other people in my webinar session would form a negative impression of me if I didn’t like this story. 10. I worried that If I don’t like this story, other people in my session would be disappointed. 11. I was concerned that if I don’t like this story, other people in my webinar session would judge me negatively. 12. I feared that I will be disliked by other people in my session if they found out that I don’t like this story. 13. I feared that I will be rejected and abandoned if other people in my webinar session discovered that I don’t like this story. Perceived publicness (adapted from Chung & Lapinski, 2019) Please think back to when you were listening to the story in the webinar and indicate your agreement with the following statements. 1. I could see whether or not other people liked the story. 2. I felt like I could see whether I liked the story. 3. I noticed if other people liked the story. 4. I felt like people noticed if I liked the story. Group Identity (Chung, 2020) Please indicate your agreement with the following statements. 1. I identify with other people in my webinar session. 2. I feel strong ties with other people in my webinar session. 3. I think most other people in my webinar session are similar to me intellectually. 4. I think most other people in my webinar session are similar to me in the way they think. 5. I think most other people in my webinar session have similar values to my own. 6. I think most other people in my webinar session behave similarly to me. 7. I see myself as belonging to the group of people in my webinar session. 8. I see myself as a part of the group of people in my webinar session. 9. I am pleased to be a part of the group of people in my webinar session. Self-presentation motivation (Dillard et al., 1989; Johnson & Ranzini, 2018) Please think back to being in the webinar and about the way you rated the story (e.g., your review for future participants) and indicate your agreement with the following statements. 1. I was concerned with making a good impression while rating the story. 60 2. When I was rating the story, I was careful to avoid acting in a way which was socially inappropriate. 3. I was very conscious of what was appropriate and inappropriate when I was rating the story.** 4. I was concerned with putting myself in a “bad light” when I was rating the story. 5. I didn’t want to look stupid based on the way I rated the story. 6. I was not willing to risk possible damage to my relationships with others based on the way I rated the story. 7. Rating the story honestly was more important to me than giving a rating that other people would agree with. (R)** 8. I didn’t really care if my rating and review made other people mad. (R) 9. When rating the story, I wanted to look as good as possible. ** 10. When rating the story, I wanted to make a good impression on other participants. 11. When rating the story, I wanted to show other people that I had good taste. (R) = reverse-coded item; ** = removed from final measure. Note: Items 1-5 were taken from Dillard et al.’s interaction subscale, 6-8 were taken from Dillard et al’s relationship resource scale, and items 9-11 were adapted from Johnson and Ranzini’s own-ideal self manipulation. 61 APPENDIX F: Summary of significant covariate tests in Study 2 Women (n = 74, M = 5.17, SD = 1.06) appreciated all stories significantly more than men (n = 155, M = 4.89, SD = 1.23; t(179.74) = 2.32, p = .02). Also, there were also significant differences in group identity such that women (M = 5.21, SD = 0.87) identified more with the confederates than men (M = 4.75, SD = 1.14; t(182.60) = 3.35, p = .001). The result is perhaps unsurprising given that there were three female and two male confederates. Additionally, self- identified Hispanic/Latinx participants also more strongly identified with the confederates (n = 10, M = 5.60, SD = 0.60) compared to non-Hispanic/Latinx participants (n = 222, M = 4.85, SD = 1.10; t(230) = 2.13, p = .03). Some differences were also observed between those identifying and black/African American and others. Specifically, those self-identifying as black/African American (n = 27) perceived significantly more sanction for disliking the story (M = 4.30, SD = 1.93) compared to those who did not identify as black/African American (n = 205, M = 3.36, SD = 1.72; t(230) = 2.63, p = .01). Also, black/African American participants had greater self-presentation motivation (M = 4.47, SD = 1.77) than non-black/African American participants (M = 3.74, SD = 1.59; t(230) = 2.21, p = .03). Otherwise, no other race or gender differences were observed. Lastly, age was significantly correlated with appreciation, r(231) = .18. p = .01, and group identity, r(231) = .13, p = .04, and attentional focus (i.e., whether participants were paying attention to the story content or the audience) was correlated with many variables, including ending condition and descriptive norms. Other correlations were not significant (Table 8). 62 APPENDIX G: TABLES Table 1. Story randomization in Study 1. Positive end Mixed ending Univ. Acclaim Control Over. Dislike Univ. Acclaim Control Over. Dislike Saving Vacation’s Saving RG1 Mad Dog Safari Keller’s Firm Civilians Over Comrades Vacation’s Saving Saving RG2 Mad Dog Keller’s Firm Safari Over Civilians Comrades Vacation’s Saving Saving RG3 Mad Dog Keller’s Firm Safari Over Civilians Comrades Saving Saving Vacation’s RG4 Safari Keller’s Firm Mad Dog Comrades Civilians Over Saving Keller’s Vacation’s Saving RG5 Safari Mad Dog Comrades Firm Over Civilians Saving Vacation’s Saving RG6 Keller’s Firm Safari Mad Dog Comrades Over Civilians Note: All participants were randomly assigned to groups. Stories were presented in random order within each group. RG = Randomization group. 63 Table 2. Story appraisals by story stem, ending, and norm cue. Enjoyment Saving Vacation’s Saving Keller’s Mad Dog Safari Civilians Over Comrades Firm End Norm M SD M SD M SD M SD M SD M SD Acc. 4.83 1.02 4.63 1.61 4.72 1.26 4.52 1.42 5.08 1.23 3.37 1.09 Pos. Ctrl 4.33 1.16 4.45 1.67 3.80 1.74 4.23 1.53 3.52 1.38 3.05 1.07 Dis. 4.44 0.85 4.46 0.78 4.23 1.46 4.95 1.43 4.77 1.07 3.72 1.22 Acc. 3.58 1.27 5.45 1.02 4.46 1.04 4.68 1.22 4.69 1.36 4.13 0.82 Mix Ctrl 3.65 1.11 4.18 1.58 4.53 1.41 4.02 1.27 3.13 1.35 2.81 1.34 . Dis. 3.95 1.47 4.19 1.34 4.28 1.39 3.76 1.45 3.96 0.91 3.25 1.28 Appreciation Saving Vacation’s Saving Keller’s Mad Dog Safari Civilians Over Comrades Firm End Norm M SD M SD M SD M SD M SD M SD Acc. 4.48 1.13 4.13 1.32 4.02 1.43 4.37 1.20 5.15 1.02 3.25 1.05 Pos. Ctrl 3.93 1.46 3.85 1.40 3.20 1.39 4.19 1.48 3.85 1.34 2.85 1.00 Dis. 4.74 1.06 4.00 1.13 3.42 1.26 4.78 1.09 4.88 1.05 3.37 1.16 Acc. 3.85 1.36 5.30 0.88 4.30 1.29 4.40 1.24 5.13 1.10 4.11 1.10 Mix Ctrl 4.37 1.26 3.95 1.50 4.48 1.36 3.81 1.42 4.03 1.25 3.76 1.38 . Dis. 4.73 1.20 4.16 1.19 3.97 1.15 4.17 1.43 4.33 1.31 3.57 1.25 Liking Saving Vacation’s Saving Keller’s Mad Dog Safari Civilians Over Comrades Firm End Norm M SD M SD M SD M SD M SD M SD Acc. .90 0.31 .67 0.49 .78 0.43 .75 0.44 .95 0.22 .63 0.50 Pos. Ctrl .72 0.46 .65 0.49 .44 0.51 .68 0.51 .45 0.51 .40 0.50 Dis. .94 0.24 .72 0.46 .70 0.47 .95 0.22 .89 0.32 .55 0.51 Acc. .45 0.51 .85 0.37 .63 0.50 .75 0.44 .72 0.46 .61 0.50 Mix Ctrl .32 0.48 .50 0.51 .60 0.50 .28 0.46 .30 0.47 .22 0.43 . Dis. .45 0.51 .68 0.48 .50 0.51 .44 0.51 .44 0.51 .45 0.51 Notes. Acc. = Universal acclaim, Ctrl = Control, Dis. = Overwhelming dislike. 64 Table 3. Means and standard deviations by condition and repeated-measure ANOVA test statistics. Enjoyment Appreciation Liking Evaluation times (raw) Evaluation times (log) Ending Ending Ending Ending Ending Positive Mixed Positive Mixed Positive Mixed Positive Mixed Positive Mixed Norm cue M SD M SD M SD M SD M SD M SD M SD M SD M SD M SD Univ. Acc. 4.53 1.37 4.50 1.26 4.25 1.30 4.51 1.26 .78 .41 .67 .47 4.59 4.25 5.89 4.82 0.51 0.38 0.64 0.36 Ov. Dislike 3.89 1.50 3.73 1.45 3.64 1.40 4.08 1.36 .54 .50 .37 .49 6.31 5.40 5.79 5.04 0.64 0.41 0.59 0.40 Control 4.43 1.22 3.90 1.34 4.19 1.28 4.15 1.28 .79 .41 .50 .50 4.89 4.63 5.04 4.27 0.52 0.41 0.56 0.37 Total 4.28 1.36 4.04 1.35 4.03 1.33 4.25 1.30 .70 .44 .51 .49 5.26 4.76 5.57 4.71 0.56 0.40 0.60 0.38 Fnorms 16.36*** 11.96*** 21.03*** 4.31* 4.45*A η2norms .15 .10 .18 .04 .04 Fending 7.13** 6.98** 30.38*** 0.77 2.83 η2ending .06 .06 .21 .01 .02 Fnorm*ending 3.17* 3.03* 2.52 2.91 4.71** η2norm*ending .02 .02 .02 .02 .04 dfnorms = 2, 228, dfending = 1, 228, dfnorms*ending = 2, 228 * p ≤ .05, ** p ≤ .01, *** p ≤ .001. A: Test violated sphericity assumption, so I reported the Greenhouse-Geisser statistic (df = 1.87, 213.41). 65 Table 4. Correlation matrix for Study 1 across conditions. 1 2 3 4 5 6 7 8 9 10 1 Enjoyment -- 2 Appreciation .50** -- 3 Liking .47** .42** -- 4 Ev. Time (raw) -.10 -.05 .08 -- 5 Ev. Time (log) -.09 -.02 .09 .89** -- 6 Gender .04 .06 .02 .05 .18 -- 7 Age .08 .01 .05 .01 -.05 -.36** -- 8 White -.01 .07 .09 -.03 -.02 .17 -.01 -- 9 Black -.07 -.08 -.19* -.01 .00 -.03 -.20* -.47** -- 10 Asian -.06 -.12 -.05 .04 .03 -.13 -.01 -.54** -.02 -- * p ≤ .05, ** p ≤ .01 Note. Only races/ethnicities with greater than five cases were added in this analysis (1 = self- identified as a member of that race/ethnicity, 0 = did not identify as this race/ethnicity). See https://osf.io/37akv/?view_only=58e2b61c078d4b6d829e1dcd9b0b4285 for correlations between conditions. Table 5. Confirmatory factor analyses (CFA) model statistics for Study 2. χ2 χ2 df χ2 p CFI TLI RMSEA [90% CI] SRMR Enj. & Appr. 14.83 8 0.06 0.99 0.97 0.061 [0.000, 0.108] 0.03 DN, IN, & Sanction 150.26 62 <.001 0.96 0.95 0.078 [0.062, 0.094] 0.05 Publicness 2.15 2 0.34 1.00 1.00 0.018 [0.000, 0.133] 0.02 Group identity 50.60 27 0.004 0.97 0.96 0.061 [0.034, 0.087] 0.04 Self-pres. Motivation 64.34 20 <.001 0.97 0.95 0.098 [0.072, 0.125] 0.03 All measures 1448.01 712 <.001 0.88 0.87 0.067 [0.062, 0.072] 0.06 N = 232. Note. Since many scales contained only three items, CFA was conducted incorporating conceptually related scales to generate model fit indices. 66 Table 6. Appraisal means by condition in Study 2. Enjoyment Appreciation Liking ET (raw) ET (log) Ending Ending Ending Ending Ending Positive Mixed Positive Mixed Positive Mixed Positive Mixed Positive Mixed Norm cue M SE M SE M SE M SE M SE M SE M SE M SE M SE M SE Univ. Acc. 5.20 0.23 4.56 0.22 4.91 0.21 5.00 0.20 0.98 0.07 0.71 0.06 58.94 18.95 76.06 17.782 1.43 0.09 1.47 0.08 Ov. Dis. 4.70 0.23 4.26 0.22 4.81 0.21 4.55 0.20 0.81 0.06 0.71 0.06 98.98 18.66 57.76 18.118 1.62 0.08 1.57 0.08 Control 4.71 0.24 4.55 0.24 5.14 0.22 4.93 0.21 0.84 0.07 0.78 0.07 58.99 19.25 91.38 19.158 1.48 0.09 1.48 0.09 Total 4.87 0.13 4.46 0.13 4.96 0.13 4.83 0.12 0.87 0.04 0.73 0.04 72.30 10.88 75.07 10.45 1.51 0.05 1.50 0.05 Fnorms 1.53 1.55 0.80 0.18 1.63 η 2 norms .01 .01 .01 <.01 .01 Fending 4.93* 0.96 7.83** 0.03 0.01 η 2 end .02 <.01 .03 <.01 <.01 Fnorms*end 0.52 0.69 1.54 2.20 0.88 η 2 norms*end .01 <.01 .01 .02 <.01 * p ≤ .05, ** p ≤ .01, *** p ≤ .001. 67 Table 7. Social sharing means by condition in Study 2. Review Score Number of Words Ending Ending Positive Mixed Positive Mixed Norm cue M SE M SE M SE M SE Univ. Acc. 8.10 0.35 7.42 0.33 16.53 2.79 20.66 2.62 Ov. Dis. 7.58 0.35 6.69 0.34 18.38 2.75 19.55 2.67 Control 7.47 0.36 7.26 0.36 20.81 2.84 22.04 2.82 Total 7.72 0.20 7.13 0.19 18.57 1.60 20.75 1.54 Fnorms 1.64 0.62 η2norms .01 <.01 Fending 4.44* 0.96 η end 2 .02 <.01 Fnorms*end 0.5 0.19 η norms*end 2 <.01 <.01 * p ≤ .05, ** p ≤ .01, *** p ≤ .001. 68 Table 8. Correlation matrix for Study 2. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 Norm -- 2 End -.01 -- 3 Like .04 .18** -- 4 Enj. .10 .14* .52** -- 5 Appr. .09 .04 .32** .62** -- 6 ET (raw) -.05 .00 .19** .01 -.01 -- 7 ET (log) -.12 .01 .26** .06 .02 .84** -- 8 Score .10 .13* .62** .57** .48** .21** .25** -- 9 Words .04 -.06 -.30** -.16* .01 -.23** -.32** -.19** -- 10 DN .31** .07 .51** .57** .43** .08 .14* .50** -.24** -- 11 IN .13 .17** .38** .58** .43** -.04 -.01 .37** -.13* .66** -- 12 Sanc. .01 .07 .39** .33** .28** .28** .36** .39** -.53** .39** .26** -- 13 Pub. .05 .02 .37** .57** .46** .10 .19** .45** -.31** .48** .43** .50** -- 14 ID -.02 .07 .30** .54** .64** .00 .06 .41** -.12 .34** .50** .30** .49** -- 15 SPM .00 .06 .44** .42** .39** .23** .36** .44** -.48** .51** .38** .79** .63** .40** -- 16 Focus -.04 -.02 .15* .10 .05 .10 .19** .18** -.34** .25** .11 .40** .19** .07 .33** -- 17 Gend. -.07 -.05 .03 .02 .04 -.08 -.10 .01 .01 .02 .02 -.13* .05 .09 -.08 -.03 18 Age .22** -.02 -.15* .05 .18** -.15* -.16* -.01 .26** .06 .12 -.08 .03 .13* -.02 -.05 19 White .08 .02 .08 -.03 -.12 .03 .06 .00 -.04 -.01 -.03 -.08 -.03 -.10 -.07 -.10 20 Black -.05 -.03 -.02 .02 .13 .05 -.01 .04 -.02 .05 .07 .17** .04 .07 .14* .12 21 Nat. Am. .00 .01 -.10 -.06 -.02 -.02 .04 -.09 -.08 -.03 -.05 -.07 .02 -.03 -.03 .07 22 Asian .08 .01 -.10 .05 .07 -.07 -.09 .01 .18** -.04 -.07 -.07 .01 .01 -.08 .03 23 Hisp. -.15* -.03 -.05 .04 .05 -.09 -.09 -.02 .13* -.05 .08 -.01 .00 .14* .02 .00 Notes: * p ≤ .05, ** p ≤ .01. Age and gender were not significantly correlated with any ethnicity and were removed from the matrix for space considerations. Norm (Norm condition): 1 = Overwhelming dislike, 2 = Control, 3 = Universal acclaim. End (ending condition): 1 = Mixed, 2 = Positive. For all race/ethnicity variables, 1 = self-identified member of race/ethnicity, 0 = does not self-identify. Gender: 1 = Male, 2 = Female, 3 = Nonbinariy/Transgender, 4 = Prefer not to answer. 69 Table 9. PROCESS Step 1: Norm cue condition predicting perceived descriptive norms. b SE p Constant 2.59 0.54 <.001 Control vs. Over. Dislike 0.82 0.25 <.001 Univ. Acc. vs Over. Dislike 1.28 0.25 <.001 Focus 0.26 0.06 <.001 Age <.01 0.01 .86 Eth.=Black 0.20 0.31 .51 Eth=Hisp. 0.08 0.50 .87 Gender 0.32 0.21 .14 F 6.79*** adj. R2 .15 df = 7, 224 70 Table 10. Effect of perceived norms and endings on appraisals and social sharing. Enjoyment Appreciation Liking ET log Review Score Number of Words b SE p b SE p b SE p b SE p b SE p b SE p Constant 1.04 0.82 .21 2.46 0.82 <.01 -1.57 1.83 .39 1.25 0.36 <.01 2.72 1.32 .04 23.62 11.63 .04 N2 vs. N1 -0.27 0.20 .17 0.06 0.20 .75 -0.43 0.53 .41 -0.16 0.09 .07 -0.29 0.32 .37 4.24 2.80 .13 N3 vs. N1 -0.26 0.20 .21 -0.17 0.20 .40 -0.74 0.55 .18 -0.20 0.09 .02 -0.19 0.33 .57 2.50 2.90 .39 DN 0.71 0.15 <.01 0.32 0.15 .03 0.89 0.39 .02 0.14 0.06 .04 0.84 0.24 <.01 -2.28 2.11 .28 Ending 0.87 0.47 .07 -0.07 0.47 .89 0.90 1.05 .39 0.25 0.21 .22 1.04 0.76 .17 -1.89 6.67 .78 DN*End -0.13 0.10 .19 0.02 0.10 .83 0.05 0.27 .86 -0.06 0.04 .18 -0.13 0.15 .40 0.06 1.35 .97 Focus -0.05 0.05 .33 -0.05 0.05 .29 -0.05 0.14 .72 0.04 0.02 .04 0.06 0.08 .43 -3.09 0.68 <.01 Age 0.01 0.01 .54 0.02 0.01 .01 -0.06 0.02 .02 -0.01 0.01 .07 -0.01 0.01 .70 0.51 0.12 <.01 Eth.=Bl. 0.03 0.24 .92 0.47 0.24 .05 -0.33 0.62 .59 -0.07 0.11 .50 0.12 0.39 .76 1.85 3.45 .59 Eth=Hisp. 0.38 0.38 .33 0.56 0.38 .15 -0.71 0.95 .45 -0.32 0.17 .05 -0.04 0.62 .95 13.63 5.45 .01 Gender 0.03 0.17 .84 0.19 0.17 .24 0.32 0.46 .49 -0.11 0.07 .13 0.17 0.27 .52 0.24 2.35 .92 F 12.28 6.93 76.95 A 2.87 8.29 6.76 2 adj. R .33 .14 .28B .07 .24 .20 Notes: df = 10, 221. All models significant at p < .01. A: Model LL. B: Cox & Snell R. N1 = Overwhelming Dislike, N2 = Control, N3 = Universal Acclaim. DN = Perceived descriptive norms. Gender: 1 = Men, 2 = Women, Transgender, Non-binary. Table 11. Conditional indirect effects of norm condition on appraisals through perceived descriptive norms. Enjoyment Appreciation Liking ET log Norm End b SE CI LL CI UL b SE CI LL CI UL b SE CI LL CI UL b SE CI LL CI UL Mixed 0.48 0.16 0.19 0.81 0.28 0.11 0.09 0.53 0.77 0.33 0.30 1.58 0.07 0.03 0.02 0.13 N2 vs. N1 Pos. 0.38 0.14 0.14 0.67 0.30 0.12 0.10 0.56 0.81 0.48 0.27 1.97 0.02 0.03 -0.03 0.08 Mixed 0.75 0.18 0.42 1.13 0.44 0.14 0.19 0.75 1.21 0.40 0.66 2.22 0.10 0.04 0.03 0.18 N3 vs. N1 Pos. 0.59 0.17 0.29 0.97 0.47 0.15 0.20 0.79 1.27 0.68 0.55 2.87 0.03 0.04 -0.05 0.12 Note: 95% CI presented, where LL refers to the lower limit and UL refers to the upper limit. 71 Table 12. Conditional indirect effects of norm cues on social sharing through descriptive norms. Review Score Number of Words Norm End b SE CI LL CI UL b SE CI LL CI UL Mixed 0.59 0.22 0.20 1.05 -1.83 0.99 -4.02 -0.13 N2 vs. N1 Positive 0.48 0.18 0.16 0.86 -1.78 1.22 -4.53 0.25 Mixed 0.92 0.27 0.44 1.47 -2.85 1.43 -5.92 -0.23 N3 vs. N1 Positive 0.75 0.25 0.31 1.27 -2.77 1.72 -6.31 0.42 Notes: N1 = Overwhelming Dislike, N2 = Control, N3 = Universal Acclaim. DN = Perceived descriptive norms. 95% CI presented, where LL refers to the lower limit and UL refers to the upper limit. 72 Table 13. PROCESS Model 18 Step 2: Moderating effects of ending and publicness. Enjoyment Appreciation Liking ET log Review Score Number of Words b SE p b SE p b SE p b SE p b SE p b SE p Constant 2.24 1.95 .25 4.43 1.99 .03 7.23 5.85 .22 1.61 0.93 .08 10.18 3.28 <.01 0.49 29.47 .99 N2 vs. N1 -0.05 0.18 .79 0.24 0.19 .20 -0.01 0.57 .99 -0.12 0.09 .16 -0.02 0.31 .95 2.64 2.79 .34 N3 vs. N1 -0.14 0.19 .47 -0.07 0.19 .72 -0.82 0.59 .16 -0.19 0.09 .04 -0.08 0.32 .80 1.50 2.85 .60 DN 0.01 0.45 .99 -0.39 0.46 .40 -1.60 1.51 .29 0.03 0.21 .90 -1.30 0.76 .09 4.71 6.81 .49 Ending -0.08 1.21 .95 -1.06 1.24 .39 -6.51 3.86 .09 -0.09 0.58 .88 -3.86 2.04 .06 7.57 18.28 .68 DN*Ending 0.13 0.28 .65 0.22 0.29 .45 2.08 1.12 .06 0.01 0.14 .94 1.03 0.48 .03 -1.35 4.29 .75 Pub. -0.05 0.45 .91 -0.32 0.46 .49 -1.84 1.31 .16 -0.06 0.22 .76 -1.57 0.76 .04 4.42 6.84 .52 Pub.*DN 0.11 0.10 .27 0.13 0.10 .19 0.56 0.33 .09 0.02 0.05 .65 0.44 0.16 .01 -1.36 1.46 .35 Pub.*End 0.19 0.29 .52 0.23 0.30 .43 1.79 0.88 .04 0.08 0.14 .54 1.17 0.49 .02 -2.42 4.36 .58 Pub.*End*DN -0.05 0.06 .43 -0.04 0.06 .49 -0.46 0.23 .05 -0.02 0.03 .59 -0.26 0.10 .01 0.36 0.92 .70 Focus -0.06 0.05 .19 -0.06 0.05 .22 -0.19 0.16 .25 0.04 0.02 .07 0.04 0.08 .61 -3.08 0.68 <.01 Age 0.00 0.01 .78 0.02 0.01 .03 -0.07 0.03 .01 -0.01 0.01 .06 -0.01 0.01 .39 0.55 0.12 <.01 Eth.=Bl. -0.02 0.23 .93 0.39 0.23 .10 -0.07 0.66 .92 -0.06 0.11 .55 0.13 0.38 .74 2.94 3.44 .39 Eth=Hisp. 0.26 0.36 .47 0.42 0.37 .26 -1.20 1.09 .27 -0.34 0.17 .05 -0.41 0.61 .50 15.15 5.45 .01 Gender 0.00 0.15 .98 0.18 0.16 .25 0.10 0.49 .84 -0.12 0.07 .11 0.13 0.26 .62 0.32 2.30 .89 A F 13.75 8.04 89.56 2.32 8.24 6.12 2 adj. R .44 .30 .32B .07 .30 .24 df = 14, 217. A: Model LL. B: Cox & Snell R. N1 = Overwhelming Dislike, N2 = Control, N3 = Universal Acclaim. DN = Perceived descriptive norms. Gender: 1 = Men, 2 = Women, Transgender, Non-binary. 73 Table 14. Conditional indirect effects of norm cues on appraisals through descriptive norms by ending and publicness. Enjoyment Appreciation Liking ET log Norm End. Pub. b SE CI LL CI UL b SE CI LL CI UL b SE CI LL CI UL b SE CI LL CI UL -1 SD 0.26 0.13 0.06 0.56 0.07 0.12 -0.14 0.34 0.65 0.39 0.19 1.70 0.04 0.03 -0.01 0.12 Mix. M 0.33 0.13 0.12 0.62 0.18 0.10 0.01 0.42 0.77 0.37 0.29 1.72 0.05 0.03 0.00 0.11 +1 SD 0.40 0.16 0.15 0.75 0.28 0.13 0.07 0.57 0.90 0.47 0.27 2.07 0.06 0.03 0.00 0.13 N2 vs. N1 -1 SD 0.25 0.15 0.02 0.59 0.14 0.14 -0.08 0.47 1.19 22.91 0.33 3.91 0.01 0.04 -0.06 0.10 Pos. M 0.26 0.12 0.07 0.54 0.19 0.11 0.03 0.45 0.76 8.67 0.08 2.47 0.00 0.03 -0.06 0.06 +1 SD 0.27 0.15 0.05 0.61 0.24 0.14 0.04 0.57 0.33 6.96 -0.56 1.54 -0.01 0.04 -0.10 0.07 -1 SD 0.41 0.17 0.12 0.80 0.12 0.18 -0.22 0.50 1.01 0.52 0.40 2.39 0.07 0.05 -0.01 0.17 Mix. M 0.52 0.16 0.24 0.88 0.28 0.15 0.02 0.59 1.20 0.47 0.61 2.44 0.08 0.04 0.01 0.16 +1 SD 0.63 0.21 0.29 1.09 0.44 0.17 0.15 0.82 1.40 0.63 0.56 2.97 0.09 0.05 -0.01 0.19 N3 vs. N1 -1 SD 0.38 0.22 0.03 0.88 0.22 0.21 -0.11 0.69 1.86 30.44 0.66 5.53 0.02 0.06 -0.10 0.14 Pos. M 0.41 0.16 0.13 0.77 0.30 0.15 0.06 0.64 1.19 11.73 0.16 3.53 0.00 0.05 -0.10 0.09 +1 SD 0.43 0.19 0.11 0.87 0.38 0.18 0.07 0.79 0.52 9.59 -0.91 2.22 -0.02 0.06 -0.16 0.11 Note: 95% CI presented, where LL refers to the lower limit and UL refers to the upper limit. 74 Table 15. Conditional indirect effects of norm cues on sharing through descriptive norms by ending and publicness. Review Score Number of Words Norms End. Pub. b SE CI LL CI UL b SE CI LL CI UL -1 SD 0.24 0.21 -0.14 0.72 0.21 1.45 -2.55 3.39 Mixed M 0.46 0.21 0.11 0.93 -1.00 1.01 -3.14 0.93 +1 SD 0.68 0.28 0.20 1.30 -2.21 1.32 -5.13 0.01 N2 vs. N1 -1 SD 0.43 0.23 0.07 0.96 0.00 1.31 -2.84 2.52 Positive M 0.34 0.17 0.06 0.73 -0.78 1.19 -3.36 1.46 +1 SD 0.24 0.20 -0.07 0.70 -1.56 1.62 -5.08 1.36 -1 SD 0.38 0.31 -0.23 1.02 0.33 2.20 -4.00 4.78 Mixed M 0.72 0.28 0.23 1.32 -1.55 1.52 -4.65 1.39 +1 SD 1.06 0.38 0.40 1.87 -3.44 1.86 -7.29 -0.01 N3 vs. N1 -1 SD 0.67 0.33 0.13 1.40 0.01 2.01 -4.06 4.09 Positive M 0.52 0.24 0.13 1.05 -1.21 1.80 -4.71 2.53 +1 SD 0.38 0.28 -0.12 0.99 -2.43 2.44 -7.31 2.29 Note: 95% CI presented, where LL refers to the lower limit and UL refers to the upper limit. 75 Table 16. PROCESS Model 18 Step 2: Moderating effects of ending and identity. Enjoyment Appreciation Liking ET log Review Score Number of Words b SE p b SE p b SE p b SE p b SE p b SE p Constant 4.96 2.86 .08 2.01 2.64 .45 21.57 8.15 .01 3.39 1.39 .02 13.84 4.85 <.01 -4.47 45.18 .92 N2 vs. N1 -0.16 0.18 .37 0.19 0.16 .25 -0.48 0.58 .41 -0.15 0.09 .09 -0.16 0.30 .60 3.99 2.81 .16 N3 vs. N1 -0.05 0.19 .77 0.06 0.17 .73 -0.58 0.61 .34 -0.19 0.09 .04 0.04 0.31 .91 1.67 2.93 .57 DN -0.24 0.57 .67 -0.27 0.53 .61 -4.99 1.89 .01 -0.28 0.28 .32 -1.82 0.97 .06 5.62 9.03 .53 Ending -2.38 1.72 .17 -1.32 1.59 .41 -12.17 4.90 .01 -1.29 0.84 .12 -7.16 2.92 .01 10.14 27.19 .71 DN*End 0.36 0.36 .33 0.38 0.34 .26 3.35 1.30 .01 0.24 0.18 .17 1.54 0.62 .01 -2.58 5.75 .65 ID -0.73 0.60 .22 0.30 0.55 .59 -4.82 1.74 .01 -0.45 0.29 .12 -2.22 1.02 .03 5.64 9.49 .55 ID*DN 0.20 0.12 .09 0.10 0.11 .35 1.37 0.42 <.01 0.09 0.06 .12 0.55 0.20 .01 -1.59 1.83 .38 ID*End 0.76 0.36 .04 0.29 0.34 .38 3.01 1.07 .01 0.34 0.18 .06 1.80 0.62 <.01 -2.68 5.74 .64 ID*End*DN -0.12 0.07 .10 -0.08 0.07 .22 -0.77 0.27 .00 -0.06 0.04 .07 -0.36 0.12 <.01 0.58 1.15 .61 Focus -0.04 0.04 .41 -0.04 0.04 .37 -0.19 0.17 .27 0.04 0.02 .04 0.08 0.07 .29 -3.26 0.69 <.01 Age 0.00 0.01 .56 0.01 0.01 .11 -0.08 0.03 <.01 -0.01 0.00 .04 -0.02 0.01 .21 0.53 0.13 <.01 Eth.=Bl. -0.14 0.23 .55 0.38 0.21 .07 -1.07 0.73 .14 -0.08 0.11 .46 -0.06 0.38 .89 3.15 3.56 .38 Eth=Hisp. -0.03 0.35 .94 -0.03 0.32 .93 -1.83 1.18 .12 -0.36 0.17 .04 -0.54 0.60 .37 14.76 5.55 .01 Gender -0.14 0.15 .34 -0.02 0.14 .90 -0.37 0.54 .49 -0.13 0.07 .07 -0.06 0.26 .81 0.73 2.38 .76 A F 15.09 14.74 103.04 2.40 8.94 5.17 2 adj. R .46 .45 .36B .08 .32 .20 df = 14, 217. A: Model LL. B: Cox & Snell R. N1 = Overwhelming Dislike, N2 = Control, N3 = Universal Acclaim. DN = Perceived descriptive norms. Gender: 1 = Men, 2 = Women, Transgender, Non-binary. 76 Table 17. Conditional indirect effects of norm condition on appraisals through perceived norms by ending and identity. Enjoyment Appreciation Liking ET log Norm End. ID b SE CI LL CI UL b SE CI LL CI UL b SE CI LL CI UL b SE CI LL CI UL -1 SD 0.35 0.16 0.05 0.66 0.15 0.10 0.00 0.40 0.55 0.64 -0.29 1.42 0.05 0.04 -0.02 0.13 Mix. M 0.42 0.14 0.16 0.72 0.17 0.08 0.04 0.36 1.10 1.57 0.48 2.62 0.07 0.03 0.02 0.13 +1 SD 0.49 0.17 0.20 0.86 0.18 0.11 0.01 0.43 1.65 3.36 0.74 4.52 0.09 0.04 0.03 0.18 N2 vs. N1 -1 SD 0.27 0.12 0.04 0.52 0.21 0.13 -0.02 0.50 0.93 5.51 0.21 3.71 0.05 0.04 -0.03 0.14 Pos. M 0.23 0.10 0.06 0.45 0.15 0.09 0.00 0.36 0.79 4.88 0.23 2.94 0.01 0.03 -0.05 0.08 +1 SD 0.20 0.11 0.02 0.44 0.09 0.08 -0.05 0.29 0.65 5.23 0.05 2.49 -0.03 0.04 -0.12 0.05 -1 SD 0.54 0.22 0.09 0.94 0.23 0.15 0.00 0.57 0.86 0.89 -0.48 2.04 0.07 0.06 -0.03 0.19 Mix. M 0.65 0.17 0.35 1.00 0.26 0.11 0.07 0.51 1.72 1.79 1.07 3.72 0.11 0.04 0.04 0.19 +1 SD 0.76 0.19 0.44 1.20 0.29 0.16 0.03 0.64 2.57 3.64 1.66 6.51 0.14 0.05 0.06 0.25 N3 vs. N1 -1 SD 0.41 0.17 0.08 0.75 0.32 0.19 -0.03 0.72 1.45 9.98 0.40 5.50 0.07 0.06 -0.04 0.19 Pos. M 0.36 0.14 0.11 0.66 0.23 0.13 0.01 0.50 1.23 9.92 0.46 4.19 0.02 0.05 -0.08 0.11 +1 SD 0.31 0.15 0.04 0.65 0.14 0.12 -0.08 0.41 1.01 11.60 0.11 3.70 -0.04 0.06 -0.18 0.07 Note: 95% CI presented, where LL refers to the lower limit and UL refers to the upper limit. 77 Table 18. Conditional indirect effects of norm cues on social sharing through descriptive norms by ending and identity. Review Score Number of Words Norm End ID b SE CI LL CI UL b SE CI LL CI UL -1 SD 0.35 0.23 -0.07 0.83 -0.65 1.19 -3.15 1.66 Mixed M 0.52 0.20 0.17 0.95 -1.56 1.00 -3.75 0.15 +1 SD 0.69 0.28 0.23 1.32 -2.48 1.30 -5.29 -0.22 N2 vs. N1 -1 SD 0.49 0.21 0.14 0.95 -0.95 1.56 -3.99 2.28 Positive M 0.34 0.15 0.09 0.67 -1.34 1.24 -3.97 1.01 +1 SD 0.18 0.14 -0.06 0.50 -1.73 1.84 -5.52 1.81 -1 SD 0.55 0.33 -0.12 1.19 -1.01 1.80 -4.55 2.54 Mixed M 0.81 0.25 0.37 1.34 -2.43 1.49 -5.65 0.24 +1 SD 1.08 0.35 0.51 1.87 -3.86 1.97 -8.27 -0.40 N3 vs. N1 -1 SD 0.77 0.29 0.27 1.40 -1.48 2.36 -5.60 3.74 Positive M 0.52 0.21 0.16 1.00 -2.09 1.83 -5.70 1.60 +1 SD 0.28 0.22 -0.09 0.76 -2.69 2.77 -8.12 2.81 Note: 95% CI presented, where LL refers to the lower limit and UL refers to the upper limit. 78 Table 19. PROCESS Model 21 Step 1: Moderating role of self-presentation on norms. b SE p Constant 0.40 0.52 .45 N2 vs. N1 1.89 0.54 <.01 N3 vs. N1 3.40 0.52 <.01 SPM 0.79 0.09 <.01 N2 vs. N1*SPM -0.31 0.13 .02 N3 vs. N1*SPM -0.57 0.13 <.01 Focus 0.10 0.05 .05 Age 0.00 0.01 .62 Eth.=Black -0.10 0.27 .71 Eth=Hisp. -0.13 0.42 .75 Gender 0.28 0.18 .12 F 16.80 adj. R2 .41 Notes: df = 10, 221. DV = Perceived descriptive norms. Model significant at p < .001. Table 20. Conditional effects of norm cues at different levels of self-presentation motivation. SPM Norm b SE p CI LL CI UL N2. vs N1 1.22 0.30 <.01 0.63 1.81 -1 SD N3 vs. N1 2.15 0.29 <.01 1.58 2.72 N2. vs N1 0.72 0.21 <.01 0.31 1.13 M N3 vs. N1 1.22 0.21 <.01 0.81 1.63 N2. vs N1 0.22 0.30 .46 -0.37 0.81 +1 SD N3 vs. N1 0.29 0.29 .33 -0.29 0.87 Note: 95% CI presented, where LL refers to the lower limit and UL refers to the upper limit. 79 Table 21. Conditional indirect effects of norm condition and self-presentation on appraisals. Enjoyment Appreciation Liking ET log Norms SPM End b SE CI LL CI UL b SE CI LL CI UL b SE CI LL CI UL b SE CI LL CI UL Mix. 0.71 0.21 0.34 1.14 0.42 0.15 0.16 0.74 1.15 0.44 0.52 2.24 0.10 0.04 0.03 0.18 -1 SD Pos. 0.56 0.17 0.26 0.92 0.44 0.15 0.18 0.77 1.20 0.92 0.47 2.90 0.03 0.04 -0.05 0.12 Mix. 0.42 0.13 0.18 0.69 0.25 0.09 0.09 0.44 0.68 0.27 0.30 1.35 0.06 0.02 0.02 0.11 N2 vs. N1 M Pos. 0.33 0.11 0.14 0.56 0.26 0.09 0.10 0.46 0.71 0.54 0.27 1.74 0.02 0.02 -0.03 0.07 Mix. 0.13 0.13 -0.13 0.39 0.08 0.08 -0.08 0.25 0.21 0.25 -0.23 0.76 0.02 0.02 -0.02 0.06 +1 SD Pos. 0.10 0.11 -0.10 0.33 0.08 0.09 -0.08 0.27 0.22 0.31 -0.25 0.91 0.01 0.01 -0.01 0.03 Mix. 1.26 0.26 0.78 1.78 0.73 0.22 0.35 1.19 2.02 0.57 1.27 3.49 0.17 0.06 0.06 0.29 -1 SD Pos. 0.99 0.26 0.53 1.54 0.78 0.24 0.36 1.29 2.12 1.52 1.00 4.80 0.05 0.07 -0.09 0.19 Mix. 0.71 0.15 0.43 1.02 0.42 0.13 0.19 0.68 1.14 0.35 0.68 2.04 0.10 0.03 0.03 0.17 N3 vs. N1 M Pos. 0.56 0.15 0.29 0.89 0.44 0.14 0.20 0.74 1.20 0.88 0.57 2.75 0.03 0.04 -0.05 0.11 Mix. 0.17 0.13 -0.08 0.43 0.10 0.08 -0.05 0.27 0.27 0.25 -0.14 0.83 0.02 0.02 -0.01 0.07 +1 SD Pos. 0.13 0.11 -0.07 0.35 0.10 0.08 -0.05 0.29 0.28 0.35 -0.17 0.98 0.01 0.01 -0.01 0.04 Note: 95% CI presented, where LL refers to the lower limit and UL refers to the upper limit. 80 Table 22. Conditional effects of norm condition and self-presentation on social sharing. Review Scores Number of Words Norms SPM Ending b SE CI LL CI UL b SE CI LL CI UL Mixed 0.87 0.29 0.35 1.48 -2.70 1.39 -5.66 -0.21 -1 SD Positive 0.71 0.24 0.28 1.22 -2.63 1.72 -6.35 0.35 Mixed 0.51 0.18 0.19 0.89 -1.60 0.85 -3.47 -0.13 N2 vs. N1 M Positive 0.42 0.15 0.15 0.73 -1.56 1.01 -3.75 0.20 Mixed 0.16 0.17 -0.16 0.50 -0.49 0.59 -1.86 0.50 +1 SD Positive 0.13 0.14 -0.13 0.42 -0.48 0.59 -1.83 0.56 Mixed 1.53 0.38 0.83 2.33 -4.76 2.34 -9.65 -0.43 -1 SD Positive 1.25 0.39 0.56 2.09 -4.64 2.80 -10.43 0.67 Mixed 0.87 0.23 0.45 1.35 -2.70 1.34 -5.47 -0.24 N3 vs. N1 M Positive 0.71 0.22 0.31 1.19 -2.63 1.59 -5.91 0.38 Mixed 0.21 0.16 -0.10 0.56 -0.64 0.61 -2.02 0.33 +1 SD Positive 0.17 0.14 -0.08 0.47 -0.62 0.62 -2.08 0.36 Note: 95% CI presented, where LL refers to the lower limit and UL refers to the upper limit. 81 APPENDIX H: Figures Figure 1. The model of intuitive morality and exemplars (MIME) and its narrative enjoyment and appreciation rationale (NEAR; Eden et al.,2021). 82 Figure 2. Enjoyment by condition. Note: Error bar = 95% CI. Figure 3. Appreciation by condition. Note: Error bar = 95% CI. 83 Figure 4. Liking by condition. Note: Error bar = 95% CI. Figure 5. Raw evaluation times by condition. Note: Error bar = 95% CI. 84 Figure 6. Log-transformed evaluation times. Evaluation times (log) Note: Error bar = 95% CI. Figure 7. Path model for alternative tests of hypotheses and research questions (Study 2). Publicness Self presentation Ending condition motivation Social identity Perceived descriptive norms a path b path Enjoyment Norm cue condition Appreciation Social sharing Note. Moderators tested individually for each dependent variable. Enjoyment and appreciation include evaluation times (log) and liking, social sharing includes rating and number of words. 85 Figure 8. Effect of perceived descriptive norms on enjoyment by story ending (Study 2). Note: Error bar = 95% CI. 86 Figure 9. Effect of perceived descriptive norms on the probability of liking by publicness. Note: Error bar = 95% CI. 87 Figure 10. Effect of perceived descriptive norms on the probability of liking by identity. Note: Error bar = 95% CI. 88 Figure 11. Effect of norm condition on perceived descriptive norms by self-presentation. Note: Error bar = 95% CI. 89 REFERENCES 90 REFERENCES Asch, S. E., & Guetzkow, H. (1951). Effects of group pressure upon the modification and distortion of judgments. Organizational Influence Processes, 295–303. Bagozzi, R., Wong, N., Abe, S., & Bergami, M. (2000). Cultural and situational contingencies and the theory of reasoned action: Application to fast food restaurant consumption. Journal of Consumer Psychology, 9(2), 97–106. https://doi.org/10.1207/S15327663JCP0902_4 Baldwin, M. W., & Holmes, J. G. (1987). Salient private audiences and awareness of the self. Journal of Personality and Social Psychology, 52(6), 1087–1098. https://doi.org/10.1037/0022-3514.52.6.1087 Bartsch, A. (2012). Emotional gratification in entertainment experience. Why viewers of movies and television series find it rewarding to experience emotions. Media Psychology, 15(3), 267–302. https://doi.org/10.1080/15213269.2012.693811 Bartsch, A., & Hartmann, T. (2017). The role of cognitive and affective challenge in entertainment experience. Communication Research, 44(1), 29– 53. https://doi.org/10.1177/0093650214565921 Bartsch, A., Vorderer, P., Mangold, R., & Viehoff, R. (2008). Appraisal of emotions in media use: Toward a process model of meta-emotion and emotion regulation. Media Psychology, 11(1), 7–27. https://doi.org/10.1080/15213260701813447 Basuroy, S., Chatterjee, S., & Ravid, S. A. (2003). How critical are critical reviews? The box office effects of film critics, star power, and budgets. Journal of Marketing, 67(4), 103– 117. https://doi.org/10.1509/jmkg.67.4.103.18692 Bobek, D. D., Roberts, R. W., & Sweeney, J. T. (2007). The social norms of tax compliance: Evidence from Australia, Singapore, and the United States. Journal of Business Ethics, 74(1), 49–64. https://doi.org/10.1007/s10551-006-9219-x Bourdieu, P. (1984). Distinction: A social critique of the judgement of taste. Harvard University Press. Bowman, N. D., & Tamborini, R. (2012). Task demand and mood repair: The intervention potential of computer games. New Media & Society, 14(8), 1339– 1357. https://doi.org/10.1177/1461444812450426 Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen & J. S. Long (Eds.), Testing structural equation models (pp. 136–162). Sage. Bryant, J., & Oliver, M. B. (2009). Media effects: Advances in theory and research (3rd ed.). Routledge. 91 Carey, K. B., Henson, J. M., Carey, M. P., & Maisto, S. A. (2010). Perceived norms mediate effects of a brief motivational intervention for sanctioned college drinkers. Clinical Psychology: Science and Practice, 17(1), 58–71. https://doi.org/10.1111/j.1468- 2850.2009.01194.x Cacioppo, J. T., Glass, C. R., & Merluzzi, T. V. (1979). Self-statements and self-evaluations: A cognitive response analysis of heterosexual social anxiety. Cognitive Therapy and Research, 3, 249–262. https://doi.org/10.1007/BF01185965 Chung, A., & Rimal, R. N. (2016). Social norms: A review. Review of Communication Research, 4, 1–28. https://doi.org/10.12840/issn.2255-4165.2016.04.01.008 Chung, M. (2020). The role of trending norms and group identity on pro-environmental behaviors [Doctoral dissertation, Michigan State University]. ProQuest Dissertations and Theses Global. https://www-proquest- com.proxy1.cl.msu.edu/docview/2433253971?accountid=12598 Chung, M., & Lapinski, M. K. (2018). Extending the theory of normative social behavior to predict hand-washing among Koreans. Health Communication, 34(10), 1120-1129. https://doi.org/10.1080/10410236.2018.1461586 Cialdini, R. B. (1993). Social proof: Truths are us. In Influence: The psychology of persuasion (pp. 114–116). Harper Business. Cialdini, R. B., Reno, R. R., & Kallgren, C. A. (1990). A focus theory of normative conduct: Recycling the concept of norms to reduce littering in public places. Journal of Personality and Social Psychology, 58(6), 1015–1026. Crocker, J., & Luhtanen, R. (1990). Collective self-esteem and ingroup bias. Journal of Personality and Social Psychology, 58(1), 60–67. https://doi.org/10.1037/0022- 3514.58.1.60 Deutsch, M., & Gerard, H. B. (1955). A study of normative and informational social influences upon individual judgment. Journal of Abnormal and Social Psychology, 51(3), 629–636. https://doi.org/10.1037/h0046408 Dillard, J. P., Segrin, C., & Harden, J. M. (1989). Primary and secondary goals in the production of interpersonal influence messages. Communication Monographs, 56(1), 19–38. https://doi.org/10.1080/03637758909390247 Eden, A., Tamborini, R., Grizzard, M., Lewis, R., Weber, R., & Prabhu, S. (2014). Repeated exposure to narrative entertainment and the salience of moral intuitions. Journal of Communication, 64(3), 501–520. https://doi.org/10.1111/jcom.12098 Eden, A., Daalmans, S., Ommen, M. Van, & Weljers, A. (2017). Melfi’s choice: Morally conflicted content leads to moral rumination in viewers. Journal of Media Ethics, 32(3), 142–153. https://doi.org/10.1080/23736992.2017.1329019 92 Eden, A., & Tamborini, R. (2017). Moral intuitions: Morality subcultures in disposition formation. Journal of Media Psychology, 29(4), 198–207. https://doi.org/10.1027/1864- 1105/a000173 Eden, A., Tamborini, R., Aley, M. & Goble, H. (2021). Advances in research on the model of intuitive morality and exemplars (MIME). In P. Vorderer and C. Klimmt (Eds.) The Oxford Handbook of Entertainment Theory. Oxford University Press. Eliashberg, J., & Shugan, S. M. (1997). Film critics: Influencers or predictors? Journal of Marketing, 61(2), 68–78. Ewoldsen, D. R., Rhodes, N., & Fazio, R. H. (2015). The MODE model and its implications for studying the media. Media Psychology, 18(3), 312–337. https://doi.org/10.1080/15213269.2014.937440 Fisher, J. T., Huskey, R., Keene, J. R., & Weber, R. (2018). The limited capacity model of motivated mediated message processing: Looking to the future. Annals of the International Communication Association, 42(4), 291– 315. https://doi.org/10.1080/23808985.2018.1534551 French, M., & Bazarova, N. N. (2017). Is anybody out there?: Understanding masspersonal communication through expectations for response across social media platforms. Journal of Computer-Mediated Communication, 22(6), 303-319. https://doi.org/10.1111/jcc4.12197 Gagnon, A., & Bourhis, Ri. Y. (1996). Discrimination in the minimal group paradigm: Social identity or self-interest. Personality and Social Psychology Bulletin, 22(12), 1289–1301. https:/doi.org/10.1177/01461672962212009 Goldstein, N. J., Cialdini, R. B., & Griskevicius, V. (2008). A room with a viewpoint: Using social norms to motivate environmental conservation in hotels. The Journal of Consumer Research, 35(3), 472–482. https://doi.org/10.1086/586910 Grayson, A., Totzkay, D. S., Walling, B. M., Ingalls, J., Viken, G., Smith, S. W., & Silk, K. J. (2019). Formative research identifying message strategies for a campus bicycle safety campaign using self-determination theory and the social norms approach. Accident Analysis and Prevention, 133, 105295– 105295. https://doi.org/10.1016/j.aap.2019.105295 Grizzard, M., Fitzgerald, K., Francemone, C. J., Ahn, C., Huang, J., Walton, J., … Lewis, R. J. (2019). Narrative retribution and cognitive processing. Communication Research, 1–23. https://doi.org/10.1177/0093650219886512 Haidt, J., & Joseph, C. (2007). The moral mind: How 5 sets of innate intuitions guide the development of many culture-specific virtues, and perhaps even modules. In P. Carruthers, S. Laurence, & S. Stich (Eds.), The innate mind (Vol. 3, pp. 367–391). Oxford University Press. 93 Hayes, A. F. (2017). Introduction to mediation, moderation, and conditional process analysis methodology: A regression-based approach. Guilford Press. Heider, F. (1946). Attitudes and cognitive organization. The Journal of Psychology, 21, 107– 112. Higgins, E. T., & Brendl, C. M. (1995). Accessibility and applicability: Some "activation rules" influencing judgment. Journal of Experimental Social Psychology, 31(3), 218–243. https://doi.org/10.1006/jesp.1995.1011 Hunter, J. E., & Gerbing, D. W. (1982). Unidimensional measurement, second order factor analysis, and factor models. Research in Organizational Behavior, 4, 267–320. Huskey, R., Bue, A. C., Eden, A., Grall, C., Meshi, D., Prena, K., ... & Wilcox, S. (2020). Marr’s trilevel framework integrates biological explanation across communication subfields. Journal of Communication, 70, 356–378. https://doi.org/10.1093/joc/jqaa007 Johnson, B. K., & Ranzini, G. (2018). Click here to look clever: Self-presentation via selective sharing of music and film on social media. Computers in Human Behavior, 82, 148–158. https://doi.org/10.1016/j.chb.2018.01.008 Krämer, N. C., Neubaum, G., Winter, S., Schaewitz, L., Eimler, S., & Oliver, M. B. (2019). I feel what they say: The effect of social media comments on viewers’ affective reactions toward elevating online videos. Media Psychology, 1–27. https://doi.org/10.1080/15213269.2019.1692669 Kryston. K., & Eden, A. (2021). Social norms and media enjoyment. Mass Communication and Society. Advance online publication. https://doi.org/10.1080/15205436.2021.1934703 Kryston, K., Park, S., & Eden, A. (2020, May). Do social sanctions explain entertainment avoidance? Paper presented at the 70th annual meeting of the International Communication Association, Gold Coast, Australia. Kuppens, T., Yzerbyt, V. Y., Dandache, S., Fischer, A. H., & van der Schalk, J. (2013). Social identity salience shapes group-based emotions through group-based appraisals. Cognition & Emotion, 27(8), 1359–1377. https://doi.org/10.1080/02699931.2013.785387 Lapinski, M. K., Maloney, E. K., Braz, M., & Shulman, H. C. (2013). Testing the effects of social norms and behavioral privacy on hand washing: A field experiment. Human Communication Research, 39(1), 21–46. https://doi.org/10.1111/j.1468- 2958.2012.01441.x Lapinski, M. K., & Rimal, R. N. (2005). An explication of social norms. Communication Theory, 15(2), 127–147. https://doi.org/10.1111/j.1468-2885.2005.tb00329.x Lewis, R. J., Tamborini, R., & Weber, R. (2014). Testing a dual-process model of media enjoyment and appreciation. Journal of Communication, 64(3), 397–416. https://doi.org/10.1111/jcom.12101 94 Manning, M. (2011). When we do what we see: The moderating role of social motivation on the relation between subjective norms and behavior in the theory of planned behavior. Basic and Applied Social Psychology, 33, 351–364. https://doi.org/10.1080/01973533.2011.589304 Möller, A. M., Baumgartner, S. E., Kühne, R., & Peter, J. (2019). The effects of social information on the enjoyment of online videos: An eye tracking study on the role of attention. Media Psychology, 1–22. https://doi.org/10.1080/15213269.2019.1679647 Moss, A. & Litman, L. (2020). Demographics of People on Amazon Mechanical Turk. CloudResearch. https://www.cloudresearch.com/resources/blog/who-uses-amazon- mturk-2020-demographics/ Nabi, R. L., & Krcmar, M. (2004). Conceptualizing media enjoyment as attitude: Implications for mass media effects research. Communication Theory, 14(4), 288–310. https://doi.org/10.1111/j.1468-2885.2004.tb00316.x Neubaum, G., & Krämer, N. C. (2018). What do we fear? Expected sanctions for expressing minority opinions in offline and online communication. Communication Research, 45(2), 139–164. http://doi.org/10.1177/0093650215623837 O’Sullivan, P. B., & Carr, C. T. (2018). Masspersonal communication: A model bridging the mass-interpersonal divide. New Media & Society, 20(3), 1161-1180. https://doi.org/10.1177/1461444816686104 Oliver, M. B. (2008). Tender affective states as predictors of entertainment preference. Journal of Communication, 58(1), 40–61. https://doi.org/10.1111/j.1460-2466.2007.00373.x Oliver, M. B., & Bartsch, A. (2010). Appreciation as audience response: Exploring entertainment gratifications beyond hedonism. Human Communication Research, 36(1), 53–81. https://doi.org/10.1111/j.1468-2958.2009.01368.x Oliver, M. B., & Hartmann, T. (2010). Exploring the role of meaningful experiences in users’ appreciation of “good movies.” Projections, 4(2), 128–150. https://doi.org/10.3167/proj.2010.040208 Oliver, M. B., & Krakowiak, K. M. (2009). Individual differences in media effects. In J. Bryant & M. B. Oliver (Eds.), Media effects: Advances in theory and research (pp. 517–531). Routledge. Oliver, M. B., & Raney, A. A. (2011). Entertainment as pleasurable and meaningful: Identifying hedonic and eudaimonic motivations for entertainment consumption. Journal of Communication, 61, 984–1004. https://doi.org/10.1111/j.1460-2466.2011.01585.x Park, H. S., & Smith, S. W. (2007). Distinctiveness and influence of subjective norms, personal descriptive and injunctive norms, and societal descriptive and injunctive norms on behavioral intent: A case of two behaviors critical to organ donation. Human 95 Communication Research, 33(2), 194–218. https://doi.org/10.1111/j.1468- 2958.2007.00296.x Prabhu, S., Hahn, L., Tamborini, R., & Grizzard, M. (2020). Do morals featured in media content correspond with moral intuitions in media users?: A test of the MIME in two cultures. Journal of Broadcasting & Electronic Media, 64(2), 255-276. https://doi.org/10.1080/08838151.2020.1757364 Rhodes, N., & Ewoldsen, D. R. (2009). Attitude and norm accessibility and cigarette smoking. Journal of Applied Social Psychology, 39(10), 2355-2372. https://doi.org/10.1111/j.1559-1816.2009.00529.x Rhodes, N., Roskos-Ewoldsen, D. R., Edison, A., & Bradford, M. B. (2008). Attitude and norm accessibility affect processing of anti-smoking messages. Health Psychology, 27(3, Suppl), S224. https://doi.org/10.1037/0278-6133.27.3(Suppl.). S224 Rhodes, N., Shulman, H., & McClaran, N. (2020). Changing norms: A meta-analytic integration of research on social norms appeals. Human Communication Research, 46, 161–191. https://doi.org/10.1093/hcr/hqz023 Rimal, R. N., & Real, K. (2005). How behaviors are influenced by perceived norms: A test of the theory of normative social behavior. Communication Research, 32(3), 389–414. https://doi.org/10.1177/0093650205275385 Roe, K. (1983). Mass media and adolescent schooling: Conflict or co-existence? Almqvist & Wilksell International. Rubin, A. M. (1983). Television uses and gratifications: The interactions of viewing patterns and motivations. Journal of Broadcasting & Electronic Media, 27, 37–51. Ryan, R., & Deci, E. (2000). Self-determination theory and the facilitation of intrinsic motivation. American Psychologist, 55(1), 68–78. http://doi.org/10.1037/0003- 066X.55.1.68 Shedlosky-Shoemaker, R., Costabile, K. A., DeLuca, H. K., & Arkin, R. M. (2011). The social experience of entertainment media: Effects of others’ evaluations on our experience. Journal of Media Psychology, 23(3), 111–121. https://doi.org/10.1027/1864- 1105/a000042 Sherif, M. (1973). The formation of a norm in a group situation. In The psychology of social norms (pp. 89–113). Octagon Books. Shulman, H. C., Rhodes, N., Davidson, E., Ralston, R., Borghetti, L., & Morr, L. (2017). The state of the field of social norms research. International Journal of Communication, 11, 1192–1213. https://ijoc.org/index.php/ijoc/article/view/6055/1966. 96 Sundar, S. (2008). The MAIN model: A heuristic approach to understanding technology effects on credibility. In M. Metzger & A. Flanagin (Eds.), Digital Media, Youth, and Credibility (pp. 73–100). MIT Press. https://doi.org/10.1162/dmal.9780262562324.073 Tamborini, R. (2013). Model of intuitive morality and exemplars. In R. Tamborini (Ed.), Media and the moral mind (pp. 43–74). Routledge. Tamborini, R., Baldwin, J., Hahn, L., Aley, M., Sethi, N., Goble, H., Grady, S., Prabhu, S., & Kryston, K. (2019, May). Moral Intuition Salience as a Mediator of Approbation in ADT. Paper presented at the 69th annual meeting of the International Communication Association, Washington D. C. Tamborini, R., Bowman, N. D., Eden, A., Grizzard, M., & Organ, A. (2010). Defining media enjoyment as the satisfaction of intrinsic needs. Journal of Communication, 60, 758–777. https://doi.org/10.1111/j.1460-2466.2010.01513.x Tamborini, R., Grady, S. Baldwin, J. McClaran, N., & Lewis, R. (2021). The narrative enjoyment and appreciation rationale. In P. Vorderer and C. Klimmt (Eds.) The Oxford Handbook of Entertainment Theory. Oxford University Press. Tamborini, R., Grizzard, M., Bowman, N. D., Reinecke, L., Lewis, R. J., & Eden, A. (2011). Media enjoyment as need satisfaction: The contribution of hedonic and nonhedonic needs. Journal of Communication, 61(6), 1025–1042. https://doi.org/10.1111/j.1460- 2466.2011.01593.x Vlaev, I. (2012). How different are real and hypothetical decisions? overestimation, contrast and assimilation in social interaction. Journal of Economic Psychology, 33(5), 963– 972. https://doi.org/10.1016/j.joep.2012.05.005 Vorderer, P. (2003). Entertainment theory. In J. Bryant, D. Roskos-Ewoldsen, & J. Cantor (Eds.), LEA's communication series. Communication and emotion: Essays in honor of Dolf Zillmann (p. 131–153). Lawrence Erlbaum Associates Publishers. Waddell, T. F., & Sundar, S. (2017). #thisshowsucks! The overpowering influence of negative social media comments on television viewers. Journal of Broadcasting & Electronic Media, 61(2), 393–409. https://doi.org/10.1080/08838151.2017.1309414 Waddell, T. F., Sundar, S. S., & Jimirro, J. P. (2020). Bandwagon effects in social television: How audience metrics related to size and opinion affect the enjoyment of digital media. Computers in Human Behavior, 1–27. Weinmann, C., Schneider, F. M., Roth, F. S., Bindl, M. J., & Vorderer, P. (2016). Testing measurement invariance of hedonic and eudaimonic entertainment experiences across media formats. Communication Methods and Measures, 10(4), 248-257. https://doi.org/10.1080/19312458.2016.1227773 Winter, S., Krämer, N. C., Benninghoff, B., & Gallus, C. (2018). Shared entertainment, shared opinions: The influence of social TV comments on the evaluation of talent 97 shows. Journal of Broadcasting & Electronic Media, 62(1), 21–37. https://doi.org/10.1080/08838151.2017.1402903 Yoshimura, K., Bowman, N. D., Cohen, E. L., & Banks, J. (2021). Character morality, enjoyment, and appreciation: a replication of Eden, Daalmans, and Johnson (2017). Media Psychology, https://doi.org/10.1080/15213269.2021.1884096 Zillmann, D. (2000). Basal morality in drama appreciation. In I. Bondebjerg (Ed.), Moving images, culture, and the mind (pp. 53–63). University of Luton Press. 98