THE PERSUASIVE APPEAL OF DISGUST: EXAMINING THE EFFECTS OF GRUESOME IMAGES AND REPULSIVE FEELINGS IN HEALTH AND PROSOCIAL COMMUNICATION A Dissertation Presented to the Faculty of the Graduate School of Cornell University in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy by Deena Gervonne Kemp August 2018 © 2018 Deena Gervonne Kemp THE PERSUASIVE APPEAL OF DISGUST: EXAMINING THE EFFECTS OF GRUESOME IMAGES AND REPULSIVE FEELINGS IN HEALTH AND PROSOCIAL COMMUNICATION Deena Gervonne Kemp, Ph. D. Cornell University 2018 Researchers have been fascinated by disgust’s power to influence thoughts and behaviors in undesirable ways—especially in circumstances where the emotion should have no influence at all. In contrast, the explicit use of disgust to promote positive outcomes is not as widely studied. Described as the emotion of unfettered rejection, disgust is considered helpful for promoting avoidance of harmful behaviors, such as smoking, but not for encouraging socially supportive behaviors, such as those that benefit the welfare of others. However, the health promotion literature has mainly considered disgust’s influence with select populations, and research on its use to promote other social issues is relatively dormant. This dissertation applies tenants from functional emotion and cognitive appraisal theories to three studies, extending research on disgust and health persuasion and initiating inquiry regarding disgust and charitable behavior. The first study explores how adolescents visually engage with disgust depictions in cigarette graphic warning labels and the relationship between attention, emotions, and risk beliefs. The second study also considers attentional processes, as well as message elaboration and behavioral intentions, in the context of a charitable appeal. Together, these studies suggest that disgust repels attention but effectively recruits other negative emotions that facilitate persuasive outcomes in both contexts. The final study tests the hypothesis that the prosocial persuasive context can shift processing patterns for disgust stimuli. Compared to disgust exposure independent of the prosocial context, the interaction of disgust exposure and a prosocial appeal reduced participants certainty appraisals and increased their empathetic responses. Results also show that feelings of disgust are not associated with behavior and suggest disgust appeals can promote helping behaviors similar to other emotional stimuli. Findings across the three studies have implications for how we conceptualize disgust and provide some considerations for the strategic use of disgust in persuasive communication. BIOGRAPHICAL SKETCH Deena Kemp completed her doctoral studies in the Department of Communication at Cornell University. Her research draws on persuasion, emotion, and behavioral economics theories to examine the impact of hard-hitting messages on decision making. She addresses the question of whether disturbing graphic content facilitates persuasive outcomes in different contexts including health and charity aid communication. During her studies, Kemp served as the lead research assistant on an NIH/FDA sponsored project examining the impact of cigarette graphic warning labels among low-income populations. She also enjoyed working with multilingual graduate students as a peer tutor with Cornell’s English Language Support Office. She was honored to receive the Anson E. Rowe Advanced Graduate Student Award from the faculty of the communication department in the Spring of 2018. Prior to her studies at Cornell, Kemp developed and implemented communication strategies for community and academic health organizations. Her professional achievements have been recognized by the Council for the Advancement and Support of Education. She received a CASE 2013 Award of Excellence for her work in fundraising and stewardship communications. Kemp completed both her master’s and bachelor’s degrees in mass communication and public relations at the University of South Florida where she was a University Graduate Fellow, Provost Scholar, and Latin American and Caribbean Scholar. She also completed training in Health Management and Leadership through the USF College of Public Health. Deena is a native of Nassau, Bahamas. v for Arianna and Mordecai vi ACKNOWLEDGMENTS When I left Tampa for Ithaca, a friend and colleague gave me a gift with the inscription “Embrace the Journey.” I have tried to do just that during my time at Cornell. It has been challenging, demanding, and exhausting, but I am so thankful for the incredible people who helped me embrace every moment. I first visited Cornell during the fall of 2013. It was my third or fourth visit of the six programs I was applying to. Yet, it was the first place I left with a sense of peace (despite the snow and frigid November air) that both my academic goals and my life needs would be met. Joanna Alario, Meghnaa Tallapragada, and Graham Dixon, you all provided an excellent first impression of the university. You were genuinely welcoming, you shared your experiences candidly, and personally connected me with other members of the department. I have to especially thank Joanna for being an exceptional program coordinator. You have been an advocate in so many ways. I will miss your clever e-mails, corralling us grad students to meet the latest impending deadline, and your creative vegan treats. Cornell has a unique way of recruiting students: a personal call from your soon-to-be advisor letting you know you’ve been accepted, a social at the old- fashioned bowling alley on campus, and potluck dinner at a faculty member’s home. I never knew whether the phone call was standard protocol or a just another testament to the wonderful mentor I found in Sahara Byrne. There are the things you expect advisors to do—give feedback on your research, meet with you frequently, approve the many forms that keep you moving forward in grad school. Sahara, you did all of vii this well. I think differently about theoretical constructs, am more sensitive to the needs of research participants, and strive to make my research accessible to a variety of audiences thanks to your guidance. But you have also done so many things that I would never have asked or expected of you, and I am so grateful. I could not have survived the intense schedule and negotiations of the job market without your support. Ari and I will miss you (and Charlotte and Chris) very much. My other committee members, Jeff Niederdeppe, Dawn Schrader, and Alan Mathios are top-notch advisors too. I have learned so much about the value of research in addressing social issues from each of you. You asked the toughest questions, challenged my methodological skills, encouraged me to explore new areas of research, and wrote what I’ve been told were phenomenal recommendation letters. Thank you for bringing your different perspectives to my work. Of course, my original research committee was my cohort. Franccesca, John, Patrick, and Sam, you questioned my assumptions, offered ideas, allowed me to question your ideas, and, most of all, provided encouragement. A lot has changed, but I am glad we had the opportunity to survive Advanced Theory together, and I am proud of how you are all thriving in your respective career paths. The most rewarding part of this experience has been working with the NIH/FDA Tobacco Grant team. Amelia, Leah, Sherri, Julie, Joe, Emma, Carolyn, Chris, Motasem, Mia and Mawuena, and Jeff and Sahara (co-PIs), thank you for all the fun we managed to have while learning how to design eye-tracking experiments and administer nicotine tests. I have good memories of traveling around New York and Ohio collecting data in the mobile lab with you. viii I am also thankful to the talented graduate students I worked with as a tutor in the English Language Support Office—I was there to help you, and in return, you introduced me to interesting and innovative research, sparked ideas about my own work in ways I never anticipated, and inspired me with your resilience in face of language and cultural barriers. Michelle, Nate and Melissa, you provide an invaluable service to Cornell students. Thank you for allowing me to be a part of it. My sincerest appreciation to Janna Lamey and her colleagues at the Graduate School for providing resources, social activities, and travel and childcare grants; and many thanks to the staff of the Cornell Statistical Consulting Unit and the Cornell Institute for Social and Economic Research for your help with data analysis. Florio Arguillas and Daniel Alexander generously included me in a data replication pilot program at CISER. They patiently reviewed the results chapters in this manuscript against my syntax and output. Thank you both for catching data entry issues and for helping me see my data files through the eyes of future reviewers. I am also grateful to the National Science Foundation for funding my research, which not only eased the burden of completing my degree on schedule but also allowed me to hire a student to create the graphics for the third study in this manuscript.i Rachael, you did a fantastic job with the stimulus manipulations—the study is so much better thanks to your creativity and technical savvy. One of the things I will miss most about Ithaca is the sense of community my daughter and I found here. Growing up on a university campus is something special. For the first time, Ari had a neighborhood, with friends from all around the world. Hasbrouck—with all of its quirks—will always be a special place for us. To my ix neighbors Samar, Barfiya, and Sarah, and all your girls, thank you for the fun outings, wonderful memories, and the delicious meals (especially the pumpkin bread). We will also miss fellowship and worship with our Ithaca SDA Church family. Your friendship extended far beyond Sabbath hours and the church walls. Rebecca and Steve, from our very first week here, you welcomed us into your home, offered to pinch hit with childcare, and treated us to amazing, home-cooked Indian food. Our kids have a lot of great memories together; I believe they will be friends for life. Liina and Tiit, you left Ithaca way too soon, but the creative pizza nights and Hot Fuzz outtakes will remind me of you always. Steve and Tiit, I miss making music with you guys. Too bad we couldn’t perform as the Three Professors—one day. Junie, you, and your mission mobile, have been a life saver so many times—GOVT 6242 would not have happened without you. Claudia, Theo, Marie and Leigh, the Tau, Castillo and Thomas families, Mardocheê and Lory, Barbara, Angel and Anny, thank you all for sharing your homes, hearts, and talents with us. Much love to my family, and friends who are family. Despite the distance, you helped me stay connected to life beyond my studies: my brother, Alexas, who would call once a week just to make sure I was good and never begrudged me for not doing the same; my mother, Greta, who was also my late night statistics coach in year one and personal EARS service throughout; my father, Demetrius, who would call randomly just to say “I love you”; Dr. P. and Denise, both of whom personally understand the challenges of (single) parenting in grad school—I appreciate you both checking in on me from time to time; Kimberly for that special gift and the great example you’ve set of embracing the journey; the Sterling and Munnings families, you x are wonderful friends who also inspire me in how you embrace life; and my sister friends Yakie, Bonnie, Debbie, Monique, Carolyn, Sofia, Mimi and Melissa—I always look forward to our reunions. Two special people know firsthand how challenging this journey has been because it challenged them to do things differently every day too. Ari, I am so proud of how you have grown and adapted to life’s changes. If you weren’t the amazing, independent, optimistic child that you are, this wouldn’t have been possible. Thank you for taking it easy on me. Mordecai, you got on planes, trains and buses, exchanged date nights for Skype writing sessions, proofread articles and debated emotion theory, and somehow always found a way to make me smile—that’s some kind of special. All in all, I have been blessed! i This research was supported by the National Science Foundation Doctoral Dissertation Research Improvement Grant No. 1757097, and by the National Institute of Child Health and Human Development and the FDA Center for Tobacco Products (R01-HD079612). Any opinions, findings, and conclusions or recommendations expressed in this material are my own and do not necessarily reflect the views of the National Science Foundation, the National Institute of Child Health and Human Development, or the FDA Center for Tobacco Products. xi TABLE OF CONTENTS Biographical Sketch ........................................................................................................ v Acknowledgments ........................................................................................................ vii Table of Contents ......................................................................................................... xii List of Figures ............................................................................................................... xv List of Tables ............................................................................................................... xvi List of Abbreviations .................................................................................................. xvii Chapter 1 Introduction .................................................................................................................... 1 Disgust Elicitation in Media Persuasion ......................................................... 1 Addressing Gaps in Disgust and Persuasion Research ................................... 3 Attention, Anti-smoking Messages, and Adolescent Reactions ..................... 5 Disgust and Charitable Behavior .................................................................... 6 Summary ......................................................................................................... 7 Chapter 2 Conceptual Definitions and Theoretical Framework ..................................................... 9 Disgust Defined .............................................................................................. 9 Discrete Emotion Theory: Propositions about Disgust .................................. 9 The Discrete versus Dimensional Perspective ............................................. 11 Implications for Persuasion .......................................................................... 12 Attention for Disgust-Eliciting Messages .................................................... 12 Disgust, Certainty, and Information Processing ........................................... 15 Disgust and Behavior ................................................................................... 18 Multiple Emotions in Persuasion .................................................................. 20 Summary ....................................................................................................... 21 Chapter 3 Disgust, Attention, and Emotion: Adolescent Reactions to Cigarette Graphic Warning Labels ....................................... 23 Introduction .................................................................................................. 23 Rationale ....................................................................................................... 25 Attentional Bias and Information Processing ............................................... 29 Attention and Emotional Reactions .............................................................. 31 Method .......................................................................................................... 32 Measures ....................................................................................................... 37 Results .......................................................................................................... 39 Discussion ..................................................................................................... 44 Limitations .................................................................................................... 46 Summary ....................................................................................................... 47 xii Chapter 4 The Effect of Attention and Discrete Emotions on Message Elaboration and Intentions in Response to a Disgust-Eliciting Charitable Appeal ................................ 48 Introduction .................................................................................................. 48 Attention for Disgust-Evoking Stimuli ........................................................ 50 Novelty and Emotional Forewarning ........................................................... 52 Cognitions and Prosocial Behavior .............................................................. 53 Multiple Emotions ........................................................................................ 55 Method .......................................................................................................... 57 Measures ....................................................................................................... 62 Results .......................................................................................................... 64 Discussion ..................................................................................................... 69 Limitations .................................................................................................... 73 Summary ....................................................................................................... 75 Chapter 5 Encountering Disgust in Prosocial Persuasion: How Context Shapes Appraisals, Emotions, and Behavior ......................................... 76 Introduction .................................................................................................. 76 Disgust and Moral Decision Making ............................................................ 78 Feeling and Thinking in Prosocial Decision Making ................................... 80 Certainty Appraisals and Information Processing ........................................ 80 Disgust Reappraisal, Distress and Empathy ................................................. 82 Predicting Prosocial Behavior ...................................................................... 84 Method .......................................................................................................... 86 Measures ....................................................................................................... 93 Results .......................................................................................................... 96 Discussion ................................................................................................... 103 Limitations .................................................................................................. 107 Summary ..................................................................................................... 108 Chapter 6 Conclusion .................................................................................................................. 110 Attention for Disgust-Eliciting Messages .................................................. 110 Multiple Emotions in Persuasion ................................................................ 112 Disgust Elicitation and Cognitive Processes .............................................. 114 Behavioral Responses to Disgust Appeals ................................................. 116 Theoretical and Empirical Contributions ................................................... 117 Practical and Ethical Implications .............................................................. 118 Limitations .................................................................................................. 121 Future Research .......................................................................................... 122 Summary ..................................................................................................... 125 References .................................................................................................................. 126 Chapter 1 .................................................................................................... 126 Chapter 2 .................................................................................................... 129 Chapter 3 .................................................................................................... 140 xiii Chapter 4 .................................................................................................... 147 Chapter 5 .................................................................................................... 157 Chapter 6 .................................................................................................... 164 Appendices ................................................................................................................. 169 Appendix 1: Research Questions, Hypotheses, and Operationalizations ... 169 Appendix 2: Chapter 3 Post-test Items ....................................................... 173 Appendix 3A: Chapter 4 Distractor Task ................................................... 176 Appendix 3B: Chapter 4 Forewarning Pretreatment .................................. 177 Appendix 3C: Chapter 4 Stimuli Manipulation .......................................... 178 Appendix 3D: Chapter 4 Post-test Items .................................................... 181 Appendix 4A: Chapter 5 Stimuli Manipulation ......................................... 183 Appendix 4B: Chapter 5 Post-test Items .................................................... 186 xiv LIST OF FIGURES Figure 3.1 Proposed FDA graphic warning labels with disgust imagery. .................... 34 Figure 3.2 Proposed FDA graphic warning labels without disgust imagery. ............... 34 Figure 3.3 Interaction effect of attention and living with a smoker ............................. 44 Figure 4.1 Sample Stimuli ............................................................................................ 61 Figure 5.1 Sample Stimuli ............................................................................................ 88 Figure 5.2 Main and interaction effects of disgust exposure and persuasive appeal . 101 xv LIST OF TABLES Table 3.1 Participant Demographics ............................................................................ 35 Table 3.2. Descriptive Statistics ................................................................................... 38 Table 3.3 Unstandardized Regression Coefficients with Standard Errors ................... 43 Table 4.1 Participant Demographics ............................................................................ 59 Table 4.2 Bivariate Correlations of Dependent Variables ............................................ 64 Table 4.3 Means and Confidence Intervals for Dependent Variables .......................... 65 Table 4.4 Main Effects on All Dependent Variables compared to Control ................. 66 Table 4.5 Indirect Effects of Attention and Emotion ................................................... 67 Table 4.6 Forewarning Effects on Attention and Disgust ............................................ 68 Table 5.1 Experimental Conditions .............................................................................. 86 Table 5.2 Participant Demographics ............................................................................ 91 Table 5.3 Bivariate Correlations of Dependent Variables ............................................ 96 Table 5.4 Descriptive Statistics by Condition with 95% Confidence Intervals ........... 97 Table 5.5 Main Effects Compared to Non-persuasive, Neutral Condition .................. 98 Table 5.6 Main and Interaction Effects of Persuasive Context and Disgust Imagery 100 Table 5.7 Indirect Effects on Behavior ....................................................................... 102 xvi LIST OF ABBREVIATIONS AFD Average Fixation Duration AOI Area of Interest ATF Appraisal-Tendency Framework ATFF Average Time to First Fixation ELM Elaboration Likelihood Model of Persuasion FC Fixation Count FDA U.S. Food and Drug Administration GWL Graphic Warning Label NTD Neglected Tropical Disease OLS Ordinary Least Squares SCM Stereotype Content Model TFD Total Fixation Duration TFF Time to First Fixation xvii CHAPTER 1 INTRODUCTION Scholarly interest in the emotion of disgust and its effect on human behavior dates back to early research on discrete emotions (Angyal, 1941), but the strategic use of disgust as a means of social influence has only gained attention in the emotion and media persuasion literature in recent years (Nabi, 1998). Indeed, scholars have examined the intentional use of disgust to incite violence and perpetuate injustice against marginalized social groups (i.e. referring to immigrants as rabid pests) for some time (Harris & Fiske, 2011; Stanton, 1998; Taylor, 2007). However, the present group of studies fits within a research tradition concerned with prosocial outcomes of employing disgust as a media persuasion strategy. Disgust Elicitation in Media Persuasion Media content can be defined as disgust-eliciting, independent of the emotions it actually evokes (O’Keefe, 2003), if the content includes depictions of objects known to reliably elicit disgust reactions. Disgust-inducing content, such as depictions of disease, death, and bodily decay, are often used to discourage people from engaging in unhealthy and dangerous behaviors such as smoking, binge drinking, or texting while driving. At the same time, this type of content has also been used to encourage people to support social causes such as fighting poverty or treating diseases. Scholars have opposing views on whether disgust is a more or less effective, or appropriate, tool to motivate positive behavior (Clark & Powell, 2017; Giner-Sorolla & Harris, 2017; Lupton, 2014; Strohminger, 2014). These debates center around 1 theoretical accounts of disgust as a feeling of aversion and rejection and ethical concerns regarding what, or who, becomes the object of rejection. Although “graphic imagery including disgust-inducing [visuals] can be effective in conveying the main messages of the campaigns” (Lupton, 2014, p. 4), some have argued that this strategy ultimately triggers message rejection due to the feelings of discomfort and revulsion it may elicit. For others, these feelings and the potential to associate revulsion with certain behaviors are what make disgust-elicitation an effective persuasive strategy (Morales et al., 2012; Judah et al., 2009). It may be that disgust is well suited for some issues and more effective with certain audiences but has negative consequences in other contexts. This dissertation examines these possibilities by addressing four overarching questions regarding the use of disgust-inducing content in persuasive communication. The first question asks how disgust imagery influences visual attention patterns for a message overall, as well as for specific components of the message, and how attention influences downstream persuasion outcomes such as emotion evocation. Closely related to this is the second question of how people attend to disgust-evoking content across different communication contexts—specifically when disgust depictions communicate threat to self versus when they depict the needs of another person. A third question regards the various emotions, in addition to disgust, that these messages evoke, and understanding how each contributes to persuasive outcomes such as information processing and behavioral responses. Given concerns about the negative influences of disgust as an emotion, the final question asks whether there are negative consequences of using disgust in persuasive communication about prosocial issues. 2 Addressing Gaps in Disgust and Persuasion Research Although there has been a rapid increase in studies examining how exposure to disgust content in persuasive appeals (and the feelings of disgust these messages may trigger) affects audience responses, there are still gaps in our understanding of how the emotion influences persuasion outcomes for different audiences across varied social contexts. Recent persuasion research has mainly focused on disgust elicitation in health decision making such as its influence on smoking behavior, handwashing hygiene, and seeking help for colorectal cancer symptoms (Leshner, Bolls, & Thomas, 2009; Morales, Wu, & Fitzsimons, 2012; Porzig-Drummond, Stevenson, Case, & Oaten, 2009; Reynolds, McCambridge, Bissett, & Consedine, 2014). Yet, examples abound of the use of disgust in everyday persuasive communication regarding political activism, environmental issues, social responsibility, and charitable giving appeals. Few studies have addressed these additional contexts. Disgust research in both persuasion and judgment and decision making relies heavily on college student samples, with the exception of anti-smoking research where non-student samples, primarily adult smokers, are well represented. Other audiences remain underrepresented in the disgust and persuasion literature. This includes adolescents, particularly those with risk factors for smoking initiation, who are often the target of both tobacco marketing and anti-tobacco campaigns. As research begins to explore other persuasive contexts, there is also a need to recruit participant samples with greater heterogeneity of sociodemographic factors that reflect likely target audiences for different persuasive appeals. 3 Finally, there are opportunities to expand the methodological tools used to examine outcomes in disgust and persuasion research. Attracting attention and modifying behavior are often the initial and ultimate goals, respectively, of persuasive communication. Although attentional mechanisms for disgust have been examined, persuasion research has not yet considered visual attention measures. Further, emotions cannot fully be understood without examining their influence on behavior (Teper, Zhong, & Inzlicht, 2015). Persuasion studies are often limited in their ability to measure actual behavior, relying on behavior intention measures instead. This limitation is significant in the disgust and persuasion literature where certain behavioral outcomes (i.e. charitable giving) have not yet been considered. The research reported here addresses these gaps in the literature. Three experiments examine attentional, affective, cognitive and behavioral responses to disgust-inducing persuasive appeals. Two social contexts are considered: the use of disgust visuals in tobacco graphic warning labels and disgust depictions in a charitable giving appeal. Participants include adolescents in the anti-tobacco context, and both a college student sample and a non-student sample in the charitable appeal context. Key features of these studies include using eye tracking to measure visual attention, a donation task to capture giving behavior, and a social judgment task to explore potential unintended consequences of disgust as a persuasive strategy. The following sections provide a more detailed overview of the context and objectives of each of the three studies. 4 Attention, Anti-smoking Messages, and Adolescent Reactions Within the health persuasion literature, a majority of studies consider the role of disgust in anti-smoking campaigns (Clayton, Leshner, Tomko, Trull, & Piasecki, 2017; Cochran, Kydd, Lee, Walker, & Consedine, 2017; Halkjelsvik & Rise, 2015; Jónsdóttir, Holm, Poltavski, & Vogeltanz-Holm, 2014; Leshner et al., 2009; Tugrul, 2013). One line of studies in particular has employed physiological measures such as cardiovascular deceleration to examine motivated processing in response to disgust- inducing anti-smoking messages (Clayton, Leshner, Bolls, & Thorson, 2016; Leshner et al., 2017; Leshner, Bolls, & Wise, 2011; Leshner, Vultee, Bolls, & Moore, 2010). In contrast, visual attention has not been considered as widely in this area despite functional emotion theories proposition that disgust reduces attentional activity. Additionally, it has been suggested that disgust may be a particularly effective strategy to discourage smoking initiation among adolescents (Pechmann & Reibling, 2006), but few studies consider the effect of disgust manipulation on how youth attend to anti-smoking messages and whether attention for this content in turn influences smoking beliefs and intentions. Chapter 3 of this dissertation uses eye tracking to explore adolescents’ attentional patterns toward tobacco graphic warning labels that include disgust imagery and how these viewing patterns relate to changes in emotions and beliefs about smoking. Adolescents from low socioeconomic social groups are disproportionately affected by smoking issues; therefore, understanding how they respond to graphically aversive anti-smoking messages is particularly important. 5 Disgust and Charitable Behavior Beyond health persuasion, few published studies address the strategic use of disgust to promote other social causes. This gap is particularly interesting in the prosocial domain given that much attention has been given to the influence of disgust on moral judgments. Although prosocial behaviors, such as charitable giving, are closely related to moral decision making, it is unclear whether research the effects of disgust on moral judgment also generalize to helping behavior. Chapters 4 and 5 of this dissertation look at the relationship between disgust exposure and elicitation and prosocial intentions and behaviors. These chapters draw on cognitive appraisal theories of emotion to consider the information processing implications of using disgust as a persuasive tool in prosocial communication. The study reported in Chapter 4 explores visual attention for disgust-eliciting messages in the prosocial persuasion context. Because disgust is not used in this context to enhance perceptions of threat to oneself, lack of personal relevance may shift attentional patterns for disgust-eliciting stimuli compared to situations where disgust augments threat. This study considers several negative emotions evoked compared to a neutral message condition and how attention patterns and discrete emotions relate to cognitive responses and behavioral intentions. The third and final study presented in Chapter 5 considers differences in message processing when disgust is integrated within a persuasive appeal compared to disgust elicitation without the persuasive context. This is important because the few studies that examine the relationship between disgust and behavior use incidental or context free disgust elicitors rather than explicit inductions relevant to the decision 6 context. This study extends the research presented in Chapter 4 to include a non- student sample and a donation behavior measure. It also considers the relationship between disgust and empathy, emotions which to this point are characterized as opposing in terms of generating concern for others. In keeping with studies on moral judgment, this final chapter also examines potential negative consequences of using disgust in response to concerns that disgust-eliciting charitable appeals may actually cause harm by reinforcing demeaning stereotypes of charity aid recipients. Summary The current chapter outlines the research objectives and methodological considerations for this series of studies. It provides an overview of the gaps in the literature on disgust and persuasion and how each of the three experiments conducted address these gaps. These studies draw on the discrete view of emotion applying tenants from functional emotion and cognitive appraisal theories to explore the effects of disgust-evoking appeals in the context of health and prosocial behavior. Chapter 2 provides a conceptual definition of disgust, outlines the theoretical framework concerning the influence of disgust on social processes as applied to the current studies, and discusses the implications of theory for persuasion outcomes including attention, cognitions, other affective reactions, and behavior, as well as potential unintended effects of disgust-elicitation. Chapters 3 through 5 present the results of three experiments testing theoretical predictions about the influence of disgust elicitation on adolescent attentional, emotional and cognitive responses to tobacco graphic warning labels (Chapter 3); the influence of attention and discrete negative emotional reactions on cognitions and 7 behavioral intentions in response to a disgust-eliciting charitable appeal (Chapter 4), and the interaction of disgust-elicitation and the persuasion context on certainty appraisals, empathetic reactions and actual donation behavior among a non-student sample (Chapter 5). Appendix 1A provides a summary of the research questions and hypotheses put forward in each study along with construct operationalization, analyses and results. Chapter 6 concludes with a discussion of the contributions of this research, including theoretical, practical, and ethical considerations, and outlines future research to address limitations and outstanding questions across the three studies. 8 CHAPTER 2 CONCEPTUAL DEFINITIONS AND THEORETICAL FRAMEWORK Although scholars largely agree on how to define disgust, there is variation in their perspectives on how it functions. This chapter provides a conceptual definition of the emotion of disgust, outlines the theoretical framework that has been applied to the current studies, and discusses the implications of these theories for persuasion outcomes including attention, cognitions, emotional responses, and behavior. Disgust Defined Disgust is characterized by a feeling of aversion or visceral repulsion in response to a noxious, organically or psychologically spoiled object (Nabi, 1999; Rozin, Haidt, & McCauley, 1993). The discrete emotion approach to persuasion posits that disgust is a distinct emotion state that influences message outcomes in ways different than other negatively valenced emotions such as fear, guilt or sadness (Nabi, 2002a). These differences are based on the unique thoughts, feelings, and behaviors associated with each emotion (Izard, 1977; Roseman, Wiest, & Swartz, 1994). This perspective draws on both functional and appraisal theories of emotion. Discrete Emotion Theory: Propositions about Disgust Functional theories posit that emotions serve as guides to solving problems or attaining goals by triggering subjective feelings, physiological reactions and behavioral drives in response to obstacles or events in our environment (Lazarus, 1991). As a discrete emotion, disgust’s subjective feeling is repulsion. For the purposes of the current research, this definition is distinguished from feelings of 9 contempt or the concept of disgust as a reaction to perceived wrong doing (i.e. sociomoral disgust; see Abitan & Krauth-Gruber, 2015). Underscoring this distinction, Nabi (2002b) argued that the term “grossed out” best describes audience reactions to message-induced disgust. This subjective feeling of disgust may also be accompanied by physiological responses such as nausea or a bad taste in the mouth (Rozin, Haidt, & McCauley, 2008). Disgust reactions are believed to be strongest for body-envelope violations, such as open wounds, orifices, and internal organs, where the skin or flesh is breached exposing the inner body (Haidt, McCauley, & Rozin, 1994; Strohminger, 2014). Other disgust elicitors include bodily excretions, death, certain animals and disease pathogens, and spoiled or exotic foods (Haidt et al., 1994). Just the thought of something repulsive can bring about feelings of disgust. Actions and appearances of people, and even ideas, can elicit disgust (Ekman & Friesen, 1975). Disgust elicitors are usually not personally relevant. Yet, they direct attention to one’s own thoughts and feelings, namely the desire to escape the emotional situation (Haidt, 2003; Smith & Ellsworth, 1985). Although emotions do not automatically produce behavior, the behavioral drives and action tendencies that emotions trigger are believed to make action more likely (Frijda, 1986). Disgust’s behavioral drive includes a compulsion to get rid of and away from the disgust object “as swiftly as possible” (Morales et al., 2012, p. 385). Its action tendency is to turn away from, to avoid taking in or to expel the disgust object (Lazarus, 1991). 10 The Discrete versus Dimensional Perspective Not all functional theories espouse the discrete view of emotion. For instance, theories of emotion as motivated processing predict that emotions trigger approach or avoidance behavior based on their valence, arousal and intensity qualities (Gable & Harmon-Jones, 2010; A. Lang, 2000). Although these dimensions provide a parsimonious account of emotion (Bolls, 2010), some scholars argue that the dimensional approach cannot distinguish when emotions of similar valence and intensity produce different effects on thoughts and behavior (Lerner, Li, Valdesolo, & Kassam, 2015). Appraisal theories of emotion differentiate individual emotions based on a range of evaluations or interpretations of events in our environment (Scherer, 1999; Smith & Ellsworth, 1985). The Appraisal-Tendency Framework (ATF; Lerner & Keltner, 2000) draws on classical appraisal theories to define six appraisals that underlie emotion: pleasantness, certainty, attentional activity, control, anticipated effort, and responsibility. The ATF posits that variations in appraisals central to one emotion distinguish it from other emotions and directly influence judgments and behavior linked to each emotion. According to the ATF, disgust’s central appraisals include unpleasantness, certainty, and low attentional activity. In addition to the intense negative feelings described earlier, experiencing disgust has been found to increase certainty about what is happing during an emotional situation—this includes certainty about its cause and confidence about how one should respond to it or what will happen next (Tiedens & Linton, 2001). Disgust is also associated with decreased 11 willingness to devote attention to the emotion-eliciting object or consider the emotional event further (Ellsworth & Smith, 1988). Implications for Persuasion Although the discrete emotion view has been usefully applied to explain the role of emotion in persuasion processes (Dillard & Nabi, 2006), few studies since Nabi’s (1998) seminal paper have applied functional emotion theory to examine the effects of disgust as a persuasive strategy. Likewise, appraisal theories have also not been applied as widely in persuasive communication research as it has been in research on judgment and decision making. The attentional activity appraisal clearly has direct implications for persuasion outcomes, as does the action-tendency component from functional theories. The current research also considers implications of the persuasion context for certainty appraisals and how emotional certainty may influence message processing. Attention for Disgust-Eliciting Messages Models of persuasion posit that in order for a message to have influence, it must first attract attention (Macguire, 1968; Lang, Park, Sanders-Jackson, Wilson, & Wang, 2007). Similarly, visual attention theories propose that patterns of attentional selection and engagement influence downstream processes including judgment and behavior (Wedel & Pieters, 2008). Attention research in persuasion considers how stimulus features and perceiver goals influence visual attention as well as downstream effects of attention—whether attention is predictive of persuasion outcomes. Gaining attention is difficult as persuasive messages compete for attention against proliferate other stimuli. Numerous studies demonstrate that distinct and 12 diverse characteristics of persuasive messages influence attention toward the message. In particular, including pictures have been found to attract greater attention to a message (Pieters & Wedel, 2004). Emotional pictures are especially attention getting, with more emotionally intense stimuli attracting attention more effectively than neutral or less intense stimuli (Zeelenberg, Wagenmakers, & Rotteveel, 2006). In this way, emotion serves as a bottom-up driver of attention in as much as it is as a feature of the stimulus. Emotion inducements also trigger motivational goals within the perceiver (Lang, Bradley, & Cuthbert, 1997; Lazarus, 1991). As such, emotional stimuli that initially gain attention may further hold attention via the emotions they trigger in the individual that result in top-down goals for processing. Whereas high attention emotions are expected to increase engagement, low attention emotions are expected to reduce engagement. Discrete emotion theories characterize disgust as a low attention emotion that triggers an unwillingness to attend to (Smith & Ellsworth, 1985) or a desire to stop perceiving the disgusting object (Shimp & Stuart, 2004). However, research provides mixed support for these propositions. Much of the research supporting a relationship between disgust exposure and visual avoidance can be found outside the persuasion literature primarily looking at the role of disgust in anxiety, phobia and sociopathy disorders (Armstrong, McClenahan, Kittle, & Olatunji, 2014). In contrast, there is evidence that disgust imagery captures attention (Charash & McKay, 2002; Rubenking & Lang, 2014) even to the exclusion of other aspects of the message (Miller & Leshner, 2007; Newhagen, 1998). 13 Likewise, research on disgust in health persuasion provides evidence that disgust increases the allocation of attentional resources to the message. Leshner and colleagues (Leshner, Bolls, & Wise, 2011; Leshner, Vultee, Bolls, & Moore, 2010) find that disgust increases attention to and memory for anti-smoking messages and positively influences message-advocated behavioral intentions. It is only when disgust is paired with additional aversive stimuli, such as content that elicits threat (Leshner et al., 2009) or anger (Leshner, Clayton, Bhandari, & Bolls, 2013), that evidence of avoidance is observed. These studies utilize a dimensional approach to emotion and findings are consistent with the view that aversive stimuli, including disgust elicitors, recruit attentional resources. However, this approach does not consider visual attention (overt attention), instead referring to attention as the allocation of cognitive resources, inferred from cardiac deceleration (covert attention), to encode information in short- term memory. Eye movements represent the selection of and engagement with stimuli. Therefore, metrics that capture eye movement can be used to infer overt attentional processes (Duchowski, 2002). For instance, initial selection of a stimulus can be measured in terms of the time it takes for the eyes to stop, or fixate, on an object after onset in one’s visual field. Engagement is often inferred from the amount of attention devoted to an object as a function of the number and overall duration of fixations (Wedel & Pieters, 2006). The common eye tracking metrics of (a) time to first fixation (TFF) and (b) both fixation count (FC) and total fixation duration (TFD) operationalize selection and engagement constructs, respectively. Eye tracking allows for more precise, objective measures of attention over self-report measures in 14 persuasion research (Bol, Boerman, Bergstrom, & Kruikemeier, 2016). These measures can also be used to examine if and how attentional patterns affect downstream outcomes of persuasion. It is possible, that for some groups, disgust promotes overt visual avoidance even if participants continue to devote cognitive resources to processing the stimulus beyond exposure (Hajcak & Olvet, 2008). Further, developmental stage has been previously found to influence how people respond to disgust-eliciting stimuli (Stevenson, Oaten, Case, Repacholi, & Wagland, 2010). Little is known about attention for disgust eliciting messages in persuasive domains beyond health communication. The role of disgust in social and cause-related marketing hasn’t been studied singularly. Instead it is often subsumed within studies incorporating a variety of shocking or distressing imagery from sexual taboos to depictions of child abuse (Cockrill & Parsonage, 2016; Dahl, Frankenberger, & Manchanda, 2003). Such stimuli may recruit a variety of responses other than or exclusive of physical repulsion. Additionally, these studies have not examined visual attention or its influence on downstream effects. Although a few studies have examined attention for emotional stimuli in non-profit advertising (Alonso Dos Santos, Lobos, Muñoz, Romero, & Sanhueza, 2017; Bebko, Sciulli, & Bhagat, 2014), these studies also have not applied a discrete view of emotion. Disgust, Certainty, and Information Processing The attentional activity appraisal can be interpreted as referring to both visual attention and cognitive elaboration, or a willingness to consider the emotional event further, as Smith and Ellsworth (1985) use both definitions to explicate attentional 15 activity. Cognitive responses are considered an important factor in determining whether exposure to a message will result in persuasion (Petty & Cacioppo, 1977). Although Loewenstein and colleagues (Loewenstein & Donoghue, 2015; Loewenstein & Lerner, 2003) refer to the affective versus deliberative routes to persuasion, scholars recognize that emotion can influence message processing in multiple ways. For instance, heuristic processing, which relies on easily accessible cues to determine an appropriate response to a social situation (Shah & Oppenheimer, 2008), may be influenced by feelings of liking or negative responses to factors such as the message source. As such, emotions are often associated with heuristic processing (Slovic, Finucane, Peters, & MacGregor, 2007). However, systematic or central processing, which involves evaluations of argument strength and message quality, may also be triggered by emotional reactions (Schwarz, 2000). For instance, fear may prompt more careful analysis of a fear appeal, such as counterarguing, in an effort to diminish the perceived threat or deliberate consideration of suggestions for how to overcome it. Similarly, appraisal theories position distinct emotions as triggering one processing mode over another based on the specific appraisals associated with that emotion (Lerner & Keltner, 2000; Smith & Ellsworth, 1985). The certainty appraisal criterion has been directly linked to information processing. Emotions typified by more certain appraisals of an emotion’s cause and/or consequent action tendency, such as disgust, are predicted to result in heuristic processing, whereas those typified by greater uncertainty, such as sadness, are expected to produce more systematic processing. Consistent with the theoretical predictions, studies show evidence of a reliance on heuristic cues following disgust inductions (Tiedens & Linton, 2001). 16 Message processing can be measured in several ways. The simplest method asks participants to self-report the level of consideration they give or intend to give to a message (Smith & Ellsworth, 1985). A less subjective method involves inferring information processing routes employed from the decision outcome. For instance, studies testing appraisal theories often present decision scenarios with both heuristic cues such as source expertise and specific claims. Decisions that are in line with the source cue rather than the claims provide evidence that participants engaged in heuristic rather than systematic processing (Tiedens & Linton, 2001). Processing can also be measured in terms of the amount of cognitions produced in response to the message. Consistent with other dual processing models (Chaiken, 1980), the Elaboration Likelihood Model of Persuasion (ELM; Petty & Cacioppo, 1986) posits that exposure to a persuasive message can result in high or low levels of thinking. Peripheral routes to persuasion should produce low-level thinking and central routes should stimulate high-level thinking. Petty and Cacioppo argue that the extent of thinking, and thus the processing route employed, can be indexed by the quantity of thoughts generated in response to the message. ELM research also considers the role emotion plays in processing in terms of the emotions receivers bring to the persuasion situation as well as the emotions a message may trigger. In addition to features of the message and characteristics of receivers, the ELM also specifies a role for the influence of forewarning of a persuasion attempt. Forewarning is expected to increase the extent of thinking via the central processing route. Forewarning may also affect how we respond to emotional events. The unexpected encounter with disgusting stimuli, has been suggested to increase the avoidance and rejection 17 responses associated with disgust (Herz, 2012). Conversely, anticipating disgust may attenuate both the emotional and cognitive reactions to a disgust-eliciting stimulus. The relationship between disgust and information processing has mainly been demonstrated when disgust was incidental or irrelevant to the decision context. Tiedens and Linton (2001) found that participants who recalled a past disgusting experience relied more on heuristic cues in a current decision task. Such studies provide compelling evidence that emotions have powerful influence on decision making in situations where they should be irrelevant. However, it is equally important, if not more so, to examine the influence of integral emotions—those relevant to the decision context—in order to understand how emotions are most likely to influence judgments and choice (Zeelenberg, Nelissen, Breugelmans, & Pieters, 2008). Given that persuasive messages explicitly evoke emotion to influence decisions, the persuasion context is ideal for examining the influence of integral emotions on message processing (Nabi, 2003). Although similar patterns of processing have been found for incidental and integral inducement of other emotions, like sadness (Bodenhausen, Sheppard, & Kramer, 1994; Yoo, Kreuter, Lai, & Fu, 2014), the relationship between disgust as an integral, message-relevant emotion and message processing has not been examined. Disgust and Behavior Although persuasive messages seek to gain attention and promote message processing, whether heuristic or systematic, many domains of persuasion ultimately aim to motivate behavior. Target behaviors may include those that benefit the message recipient, such as health behaviors, or behaviors that benefit close or distant others. 18 Because emotions are motivators of behavior that induce states of action readiness, emotional appeals may effectively stimulate behavioral outcomes depending on the match between the emotional strategies employed and the specific behavioral goals. When people come into contact with a disgust-eliciting object, “their first and only reaction is to move away from the object as swiftly as possible” (Morales et al., 2012, p. 385). As such, disgust may be particularly well suited for promoting avoidance behavior. Both discrete and dimensional views of emotion describe disgust as triggering avoidance and rejection behavior. A breadth of studies provide evidence of the potential positive effects of disgust cues in health persuasion messages which often advocate rejecting unhealthy behaviors, such as excessive drinking (Collymore & McDermott, 2015), or avoiding unhealthy consequences, such as the spread of disease due to improper handwashing (Judah et al., 2009). The evidence linking disgust to decreased desirability of smoking is particularly strong (Clayton, Leshner, Tomko, Trull, & Piasecki, 2017; Cochran, Kydd, Lee, Walker, & Consedine, 2017; Tugrul, 2013). Chapter 3 explores whether similar effects occur for adolescents experiencing disgust in response to tobacco graphic warning labels. The relationship between disgust and behaviors that benefit the wellbeing of others is much less understood. Research has shown that disgust can influence moral judgments—people’s one-shot evaluations of whether an action is right or wrong— often leading to severe or harsh judgments of others (Inbar & Pizarro, 2009; Wheatley & Haidt, 2005; cf. Landy & Goodwin, 2015). As such, it is reasonable that some scholars predict disgust also discourages behaving charitably towards others (Herz, 2012; Kelly & Morar, 2014; Sherman & Haidt, 2011). Yet others argue that disgust 19 portrayals can signal the severity of need experienced by others (Strohminger, 2014). From this perspective, disgust may also be useful for prompting behaviors such as charitable giving, but few studies examine the relationship between disgust and prosocial behavior. In general, studies examining the relationship between disgust and moral decision making utilize incidental rather than integral emotion inducements where participants are unaware of the emotion source or its influence on their choices. Although it has been suggested that disgust is inflexible—that is feeling disgust triggers a predictable response regardless of context (Lazarus, 1991), a recent study showed that the influence of incidental disgust on one-shot moral judgments does not generalize to more enduring moral attitudes (Wisneski & Skitka, 2017), only integral or decision-relevant disgust imagery increased moral convictions. Multiple Emotions in Persuasion So far this discussion has focused singularly on the emotion of disgust, but persuasive messages are known to trigger more than one emotion (James Price Dillard, Plotnick, Godbold, Freimuth, & Edgar, 1996). A persuasive message may evoke emotions consistent with its dominant theme while also aiming to recruit additional supporting emotions (James Price Dillard & Shen, 2018). In addition to these intended emotions, persuasive messages may also evoke unintended or concomitant emotions (Coulter, Cotte, & Moore, 1999; Nabi, 2002a). The discrete emotion view suggests, and prior research has shown, that each emotion evoked by a message would have distinct effects on cognitive and behavioral responses (James Price Dillard & Peck, 2000) and may exert oppositional influences on processing (Lerner et al., 2015). 20 It seems natural for health messages that mainly evoke disgust to also induce some amount of fear (Chapter 3). Disgust and anger are believed to overlap in certain contexts also (Rozin, Lowery, & Haidt, 1999). It is possible, though not well established, that messages evoking disgust would induce other negative emotions as well. Sadness and guilt are particularly relevant to the charitable giving context; their potential to co-occur with disgust is explored in Chapter 4. Although people may enjoy experiencing disgust in some settings (i.e. entertainment), few studies examine the possibility for disgust to co-occur with prosocial positive emotions. Empathy is in many ways the opposite of disgust (Haidt, 2003). If disgust has been defined as closing our moral circle, empathy could be defined as opening it to include socially distant others (Pizarro, Detweiler-Bedell, & Bloom, 2006). Research on prosocial behavior and altruism places empathetic concern as a key motivator of giving and helping behavior (Batson, Lishner, & Stocks, 2015). Summary The feeling of repulsion and the subsequent rejection motivation associated with disgust has implications for attention, information processing, and behavior in response to persuasive messages that depict disgust-eliciting imagery. The current chapter provides an overview of functional emotion and cognitive appraisal theories— the theoretical framework guiding the research presented in the next three chapters— as well as a review of visual attention theory. The following chapters report three studies conducted to examine attention for disgust-eliciting messages in health and prosocial persuasion contexts, consider the influence of disgust on information processing including cognitive appraisals associated with disgust, and explore the 21 relative influence of disgust and other emotions triggered by persuasive messages on behavioral outcomes. Taken together, these studies show that while disgust imagery may not effectively hold visual attention, the use of disgust in persuasive appeals can positively influence message effects, potentially shifting cognitive appraisals and recruiting additional, prosocial emotions that may in turn influence information processing and behavioral decision making. 22 CHAPTER 3 DISGUST, ATTENTION, AND EMOTION: ADOLESCENT REACTIONS TO CIGARETTE GRAPHIC WARNING LABELS This first study reported here examines adolescents’ attentional patterns toward tobacco graphic warning labels that include disgust imagery and how these viewing patterns relate to changes in emotions and beliefs about smoking. As a significant portion of the sample lives with a smoker, the study also considers the role of personal relevance and how it interacts with attention to influence message outcomes. Introduction Nearly 90% of smokers in the United States began smoking before age 18 (CDC, 2011), making adolescents a primary target of anti-smoking campaigns and messages. Prior research suggests that using hard-hitting imagery linking smoking with disgust may be an effective tool for making smoking less attractive to younger audiences (Pechmann & Reibling, 2006; Rozin & Singh, 1999; White, Webster, & Wakefield, 2008). Indeed, one key theme of many cigarette graphic warning labels (GWLs) is their visual depictions of death, decay, deformity, and disease to emphasize health risks and portray the physical effects of smoking on the body (Newman- Norlund et al., 2014). In 2009, the U.S. Food and Drug Administration proposed nine GWLs to appear on cigarette packages—four of these labels feature a corpse, diseased lungs, rotting teeth and open sores, and other types of bodily disfigurement (Figure 1). Such stimuli are referred to as disgust-elicitors because they usually evoke the emotional 23 feeling of disgust or visceral repulsion (Haidt, McCauley, & Rozin, 1994; Rozin & Singh, 1999). Exposure to disgust-elicitors is typically associated with avoidance, but research also suggests that disgust visuals can also attract attention (Rubenking & Lang, 2014). The current study examines how adolescents attend to cigarette packages when grotesque depictions of smoking consequences are portrayed in GWLs on the package. Although it has been suggested that disgust-eliciting messages may be an effective communication strategy with adolescents, studies have mainly looked at the reactions of adults, both smokers and nonsmokers, to disgust portrayals in anti-smoking messages. These studies have mostly examined the effects of exposure to such imagery embedded in televised anti-smoking advertisements (Clayton, Leshner, Bolls, & Thorson, 2016; Jónsdóttir, Holm, Poltavski, & Vogeltanz-Holm, 2014; Leshner, Bolls, & Thomas, 2009) rather than static warning labels. Youth responses to GWLs are important because, in addition to conveying risk information, warning labels on cigarette packages also compete for attention against a tobacco company’s marketing brand and package logo. Research suggests youth in particular are attracted to the sophisticated, colorful designs of cigarette brands (Germain, Wakefield, & Durkin, 2010) and that exposure to tobacco branding correlates with attitudes and intentions in favor of smoking (Pechmann & Knight, 2002; White et al., 2008). Effective warnings draw attention away from such stimuli (Krugman, Fox, Fletcher, Fischer, & Rojas, 1994; Wogalter & Leonard, 1999). Thus, warning labels that draw attention may curtail tobacco branding effects while at the same time fostering negative attitudes toward smoking (Tobacco Free Kids, 2003; 24 White, Williams, & Wakefield, 2015). Capturing attention is also important because attending to a message is a necessary condition for additional information processing (Meernik, Jarman, Wright, Klein, & Goldstein, 2016; Wogalter & Laughery, 1996). However, a distinction should be made between exposure to a message and visual attention to message features. Disgust depictions may have an effect because they attract visual attention or because they are avoided. This study also examines whether overall visual attention to the disgust imagery predicts risk beliefs and negative emotional reactions. Rationale As an emotional response, disgust’s primary behavioral tendency is avoidance or withdrawal from the disgusting stimuli (Nabi, 1998; Shimp & Stuart, 2004). It is presumed that disgust visuals in GWLs are used to promote avoidance or rejection of smoking behaviors (Cochran, Kydd, Lee, Walker, & Consedine, 2017). However, studies have also considered whether disgust visuals may lead to attention withdrawal from the anti-smoking message itself (Clayton et al., 2016), or whether disgusting images may promote both initial attention and subsequent avoidance. Attentional processes involve both initial orienting toward a stimulus as well as continued focusing of attention on the stimulus. The Defense Cascade Model of message processing (Bradley, Codispoti, Cuthbert, & Lang, 2001) posits an initial attentional bias for negative stimuli followed by defensive processing, or attention withdrawal, as resources are allocated to behavioral responses. Therefore, attention may increase at the onset of a potential threat, as viewers seek to make sense of the situation, but decrease as the perceived threat level intensifies. Although fear and disgust elicitors 25 are both negative stimuli that indicate a threat in one’s environment (Oaten, Stevenson, & Case, 2009), previous research has established they may trigger different attentional processes (Van Hooff, Devue, Vieweg, & Theeuwes, 2013). There is some evidence of a defensive cascade in attention for disgust-evoking stimuli with initial orienting attention toward the disgust object followed by decreases in attention (Leshner et al., 2017). However, research findings are not definitive. Studies have also found evidence of initial attentional avoidance followed by an increase in attention (Leshner, Bolls, & Wise, 2011), as well as sustained stimulus avoidance (Krusemark & Li, 2011). Other studies have found evidence of increased, sustained attention for disgust-evoking messages (Clayton, Leshner, Tomko, Trull, & Piasecki, 2017), suggesting that grotesque content allows for a faster visual orienting response towards a stimulus first, and then facilitates a longer duration of attention to the stimulus. This wide variation highlights the need for further research in this area. One reason for this discrepancy may be the diversity of methods used to gauge orienting and avoidance responses. Studies specifically focused on the use of disgust visuals in anti-smoking messages have mainly inferred attention using psychophysical measures such as heart rate, skin conductance, and facial motor muscle movements. These measures provide valuable insight regarding changes in physiological processes related to attention for dynamic content. However, they do not provide direct evidence of visual attention to disgust-eliciting warning messages. In addition, while a few studies have examined how adults attend to graphic warnings with aversive visuals, including images that could be conceptualized as grotesque (Brown & Richardson, 2012; Süssenbach, Niemeier, & Glock, 2013), little is known about how adolescents 26 allocate visual attention for GWLs when they contain disgust-eliciting imagery. As noted above, understanding adolescent responses to grotesque anti-smoking imagery in GWLs on cigarette packs may be particularly important in diverting attention from cigarette branded imagery. Given that exposure to disgust-evoking stimuli may affect attention by influencing both initial orienting and overall duration of attention towards an object, there are four potential patterns that account for a combination of slower versus faster orienting and shorter versus longer attention duration towards GWLs that contain grotesque images, compared to GWLs that do not. Based on the review above, theoretically it is possible that adolescents attend to GWLs with grotesque images: a) more quickly and for longer periods of time, b) more quickly but for shorter periods of time, c) more slowly but for longer periods of time, or d) more slowly for shorter periods of time. Thus, this study tests the following competing hypotheses: H1a: Warnings that contain disgust-eliciting images will draw participants’ attention more quickly and hold their attention for a longer time than warnings that do not contain disgust-eliciting images. H1b: Warnings that contain disgust-eliciting images will draw participants’ attention more quickly but hold their attention for a shorter time than warnings that do not contain disgust-eliciting images. H1c: Warnings that contain disgust-eliciting images will draw participants’ attention more slowly but hold their attention for a longer time than warnings that do not contain disgust-eliciting images. 27 H1d: Warnings that contain disgust-eliciting images will draw participants’ attention more slowly and hold their attention for a shorter time than warnings that do not contain disgust-eliciting images. When a health threating message is personally relevant individuals may engage in defensive processes, such as avoiding the message, as a means of minimizing the threat (Liberman & Chaiken, 1992). Despite a lack of evidence that personal relevance may decrease attention to the message as a whole, Liberman and Chaiken (1992) suggest that viewers may avoid more threatening parts of a message (i.e. gruesome GWLs). More recent work supports this view of attention disengagement for personally relevant health threatening messages among smokers and nonsmokers (Brown & Locker, 2009; Kessels, Ruiter, & Jansma, 2010; Kessels & Ruiter, 2012). In the current study, personal relevance is conceptualized in terms of a living in a home with someone who smokes (i.e. a parent or sibling). The GWLs proposed for use in the United States do not depict youth nor reference risk factors that are specific to this age group. However, for youth who live with a smoker, these messages represent a potential threat to important people in their lives—making GWLs particularly relevant to this group. Furthermore, living with a smoker is a key predictor of future smoking behavior among adolescents (Hill, Hawkins, Catalano, Abbott, & Guo, 2005). Given their potential for defensive processing and their increased risk of smoking initiation, it is important to understand whether youth who live with a smoker are more avoidant of labels that contain gruesome visuals than their peers who do not live with a smoker. 28 RQ1: Do participants who live with smokers attend to warnings with gruesome visuals differently than participants who do not? Attentional Bias and Information Processing When visual attention is directed at disgust-evoking messages, viewers may attend closely to disgust-evoking visuals to the exclusion of message features. For instance, Van Hooff, Devue, Vieweg and Theeuwes (2013) found that participants viewing disgust-eliciting images disengaged from attending to the stimulus more slowly than those viewing fear-eliciting or neutral stimuli. In the context of GWLs, an attentional bias for grotesque imagery may limit the processing of additional information, such as textual content within the warning. Prior research has established that aversive images in GWLs can create attentional biases that increase attention to the images and reduce attention to the text within the warning label (Brown & Richardson, 2012; Süssenbach et al., 2013). Sussenbach and colleagues (2013) found that adult smokers spent more time looking at images than the text when warnings contained aversive imagery. In addition, while exposure to the aversive warnings predicted negative emotional responses, only exposure to the textual part of the warnings predicted increased risk beliefs. Similarly, exposure to more vivid, distressing images in an anti-alcohol message reduced gaze time for textual elements of the warning among adult participants (Brown & Richardson, 2012). Based on these previous studies, it is expected that the time participants spend looking at the warning will mainly be focused on attending to the graphic image when disgust visuals are present. Conversely, when disgust visuals are not present, attention to the textual content is expected to be higher. 29 H2a: Participants will spend more time looking at the image when the warning contains a disgust visual than when the warning does not contain a disgust visual. H2b: Participants will spend less time looking at the text when the warning contains a disgust visual than they will when the warning does not contain a disgust visual. The pictorial content of GWLs is expected to influence audience risk perceptions about smoking. Although it has been suggested that graphic warnings may help convey risk information to individuals who are not motivated to process warning content (McQueen et al., 2016), research on exposure and attention to aversive imagery in anti-smoking messages has both supported and refuted this argument (Pechmann & Reibling, 2006; Süssenbach et al., 2013). Nevertheless, in light of some evidence to support this assertion, we test for the possibility that adolescents’ perceptions of smoking risks are directly associated with the level of attention they pay to the grotesque imagery when exposed to GWLs that contain these images. We predict: H3: There will be a significant, positive relationship between attention to disgust imagery and adolescents’ perceptions of risk. Again, on the basis of assumed greater personal relevance and defensive processing, the research question associated with these hypotheses addresses whether there are differences in risk perception outcomes for adolescents who live with a smoker, compared to those who do not, based on their attention to the grotesque warnings: 30 RQ2: Does attention predict differences in changes in risk beliefs between participants who live with smokers and those who do not live with a smoker? Attention and Emotional Reactions Eliciting strong negative emotional reactions is viewed as a key mechanism driving GWL effectiveness (Peters, Evans, Hemmerich, & Berman, 2012; Shi, Wang, Ma, Ba, & Romer, 2016). Emotions provide information about the importance or relevance of an issue (Slovic, Finucane, Peters, & MacGregor, 2007) and also motivate judgements and behavioral decisions, such as rejection of smoking, accordingly (Roseman, Wiest, & Swartz, 1994). Disgust-eliciting imagery has been found to evoke stronger aversive reactions than fear-evoking content (Jónsdóttir et al., 2014; Yartz & Hawk, 2002), and may be a convenient way to increase the emotional impact of graphic warning labels (Halkjelsvik & Rise, 2015). Research on cognitive reappraisal and emotion regulation show that duration of attention to emotional cues is a significant predictor of changes in emotional state (Manera, Samson, Pehrs, Lee, & Gross, 2014). The emotional outcomes of increased attention to the more aversive parts of warnings among adolescents have not yet been explored. Based on findings of a relationship between exposure to aversive warning content and negative emotional responses for adults, it is expected that greater attention to the disgust imagery will positively predict negative emotion. H4: Increased attention to disgust visuals will be associated with increased negative emotional reactions. Given that anti-smoking messages are expected to be more influential with adolescents at greater risk of smoking initiation, this study also examines whether 31 participants who live with a smoker have stronger emotional reactions as a result of attending to the grotesque warning labels. RQ3: Does attention predict differences in changes in negative emotional reactions between participants who live with smokers and those who do not live with a smoker? Method The study uses eye tracking methodology and a within-subjects experimental design to examine how adolescents view GWLs that contain disgust-eliciting images, compared to how the same participants view GWLs that do not contain such images. Eye tracking techniques capture visual attention to specific areas of a stimulus based on objective metrics that record where participants look during message viewing (Duchowski, 2002). These data come from three larger randomized experiments originally analyzing between-subject effects for variations in the design and execution of nine FDA graphic warnings proposed in response to the Family Smoking Prevention and Tobacco Control Act (2009). The primary results of these between- subject analyses are described in detail elsewhere (Byrne et al., 2017; Skurka et al., 2017). The current study presents an analysis of responses from participants across the three experiments who were assigned to view only the full-color warnings appearing on the top 50% of cigarette packages. Cornell University’s Institutional Review Board (IRB) approved all three studies in accordance with the University’s Policy & Procedures for conducting research involving children. The primary question for this study is whether grotesque warnings attract greater attention than do warnings that do not contain grotesque depictions of 32 smoking. The “warning” is defined as the entire warning area composed of both imagery and text. The “image” is defined as the pictorial visual element of the warning and can contain either a disgust visual (i.e. a diseased lung) or a non-disgust visual (i.e. a child exposed to smoke). “Text” are the words associated with the warning (i.e. “Smoking can kill you.”), and the “brand” is the lower 50% of a cigarette box depicting standard brand packaging (i.e. the tobacco company logo). Across the nine graphic health warnings, four images contain disgust-eliciting content based on characteristics of disgust objects as described in literature on discrete emotion elicitors (Figure 1). The images of the lip sore and the blackened lungs correspond with the decay and disease disgust categories, the cadaver image represents death, and the picture of the man with a tracheostomy (hole in the throat) depicts deformity. These images were grouped together to form the disgust-eliciting warning within-subject exposure. The remaining 5 images (Figure 2) were grouped to form the non-disgust-eliciting warnings—the comparison exposure. For both warning groups, areas of interest (AOIs) were marked to capture participants gaze activity within the brand and warning regions of the cigarette packages and within the image and text regions of the warning portion of each package. Gaze scores were averaged across each of the images within a particular group (disgust or not) to account for the difference in the number of packs for each exposure type. 33 Figure 3.1 Illustration of the proposed FDA graphic warning labels with disgust imagery. Figure 3.2 Illustration of the proposed FDA graphic warning labels without disgust imagery. Participants. Study participants were youth attending middle schools in urban and rural communities in the northeastern United States (see Table 3.1 for demographic details). Nearly half of all participants reported living with a smoker (47%). Participants were enrolled in 6th (14.7%), 7th (44.7%), or 8th (40.6%) grade and ranged in age from 11 to 15 years (M = 12.91); 50.5% identified as male and 47% identified as female (the remainder either selected “prefer not to answer” or did not respond to the gender item). Participants identified as white (48.9%), black (40.6%), Hispanic (11.5%) or another ethnicity (20.2%; numbers do not add to 100% because respondents could choose, and were thus reported in, more than one race and/or ethnic category). Based on site selection criteria, 40% to 100% of students at participating 34 schools receive free or reduced lunch—an indicator that the sample is primarily students from low-income households. A small subgroup of participants had ever tried a cigarette (8.9%). Participants were excluded from the analysis if their responses to the post-test questionnaire were wholly missing (n=4) or if the eye tracking technology captured 30% or less of their viewing time as gaze data, including those who did not calibrate (n=30; usually attributable to respondents wearing thick glasses, heavy mascara, or having a form of amblyopia). These participants have substantial amounts of missing data across all AOI groups. After removal of the 34 cases, the final sample for analysis contained 436 responses. Table 3.1 Participant Demographics Means (SD) or n’s (valid %) Age M = 12.91 (0.90) Gender Male 219 (50.5%) Female 204 (47%) Prefer not to say 11 (2.5%) Hispanic 50 (11.5%) Race Black 177 (40.6%) White 213 (48.9%) Other 88 (20.2%) Smoking variables Live with a smoker 205 (47%) Tried a cigarette 39 (8.9%) Colorblind 28 (6.5%) Procedure. School administrators sent parents IRB approved opt-out consent forms in the weeks before the study was conducted at each school. Students were recruited during physical and health education classes using convenience sampling. 35 Interested students, whose parents did not return the form opting them out of the study, read and signed assent forms indicating their participation was voluntary, and that they could discontinue the study at any time. The study was conducted in a mobile laboratory parked on-site at participating middle schools. Computers equipped with TobiiStudio 3.4.4 eye tracking software and Tobii T60XL 24” LCD monitors (1920 x 1200 resolution) recorded participants’ gaze data. Research assistants escorted participants to one of the five private viewing stations in the mobile lab. Once seated, participants completed 9-point calibration and then viewed the study instructions on screen as a research assistant also read the instructions. Participants viewed the images seated approximately 64 centimeters from the computer monitor. Stimuli. The stimuli for the study were slightly modified versions of the 9 GWLs the FDA proposed in 2009 (Figure 1). Each GWL appeared in color covering the top 50% of a cigarette package appearing against a white background. One of three different cigarette brand logos appeared in the bottom half of each image. Warnings were rotated across brands so that participants did not always see particular warnings on the same brand. The cigarette pack images measured 312 pixels x 482 pixels and subtended approximately 7.5° x 11.5° in visual angle (Shankleman, Sykes, Mandeville, Di Costa, & Yarrow, 2015). The images were counterbalanced and appeared on the screen for 10 seconds each followed by an “X” randomly placed in different locations on the outer edge of the screen in between images to reset participants’ gaze. Participants were instructed to keep their eyes on the screen for the 36 duration of all images and to look at the “X” when they saw it; the images advanced automatically. After viewing all 9 images, participants also completed a post-exposure questionnaire that contained outcome variable measures including negative emotions and risk beliefs. Finally, participants were debriefed and received a gift card for $5 to $10 when school policies permitted this incentive. For schools that would not allow direct compensation to participants, we donated funds to the school in proportion to the number of students participating. Measures Eye tracking. The time to first fixation (TFF) gaze statistic measures the initial amount of time (in seconds) participants take to fixate on a given AOI for the first time. Because there was an unequal number of packages in each warning group, the average time to first fixation (ATFF) scores for each AOI (brand, warning, image, text) is used in analyzes. The summed TFF scores across all images within each set were divided by 4 and 5 respectively to create the ATFF score for the disgust-eliciting and non-disgust eliciting warning groups. The total fixation duration (TFD) gaze statistic measures the total amount of time a participant spends looking at a particular AOI group across all images. Similar to calculations for ATFF, average fixation duration (AFD) scores were also created for AOIs in each warning group. Risk Beliefs. Participants also responded to items measuring perception that cigarette smoking is related to heart disease, lung disease, cancer, stroke, a hole in the throat, asthma and problems in babies whose moms smoke. Participants responded to each of these items on a 4-point scale ranging from 1 (Definitely not) to 4 (Definitely 37 yes). These items were averaged to create an index of risk beliefs (M=3.73; Cronbach’s α=0.91). Negative Emotion. In keeping with the literature reviewed, the analysis focuses on measures of fear and disgust as indicators of negative emotional reactions to the graphic warnings. After viewing the images, participants indicated how afraid, scared, disturbed and grossed out they felt on a 5-point scale ranging from 1 (Not at all) to 5 (Extremely) for each of the 4 items (adapted from PANAS; Watson, Clark, & Tellegen, 1988). These items were averaged to create an index of negative emotion (M = 2.81; Cronbach’s α=0.81). The mean and standard deviations for all the outcome measures, along with the scale alphas, are reported in Table 3.2. Table 3.2. Descriptive Statistics N items α Mean (SD) ATFF Gross Warning -- -- 0.18 (0.53) AFD Gross Warning -- -- 4.52 (1.62) AFD Gross Image -- -- 3.14 (1.35) AFD Gross Text -- -- 1.21 (0.78) AFD Gross Brand -- -- 2.09 (1.15) ATFF Not Gross Warning -- -- 0.10 (0.16) AFD Not Gross Warning -- -- 4.99 (1.52) AFD Not Gross Image -- -- 2.31 (0.94) AFD Not Gross Text -- -- 2.52 (1.04) AFD Not Gross Brand -- -- 1.67 (0.91) Risk Beliefs 7 .91 3.73 (0.43) Negative Emotion 4 .81 2.81 (1.16) Note: ATFF = Average Time to First Fixation, AFD = Average Fixation Duration. 38 Results Hypothesis 1a and 1b predicted that warnings with grotesque visuals would attract participants’ attention more quickly than warnings without such visuals, whereas hypotheses 1c and 1d predicted that warnings with grotesque visuals would draw attention more slowly than warnings without. A paired samples t-test compared warnings with and without disgust visuals to determine whether there were differences in the average time to first fixation to the graphic warning label between the two groups. Results show that participants fixated more slowly to the warnings with disgust visuals (M = 0.18 seconds/warning, SD = 0.53) than to the warnings without disgust visuals (M = 0.10 seconds/warning, SD = 0.16; t(435) = 3.31, p < 0.01, d = 0.16). Therefore, with regard to initial attention, the results do not support hypotheses 1a and 1b. Although hypotheses 1c and 1d both predict that the presence of grotesque visuals will predict slower orienting toward the warning, H1c predicted that participants would spend more time attending to warnings that contain grotesque images than to warnings that do not. Alternatively, H1d predicted participants would spend less time attending to warnings that contain grotesque images than to warnings that do not. To test these hypotheses, an additional paired samples t-test compared the AFD time for each type of warning. Participants viewed the warnings with disgust visuals for a shorter time (M = 4.52 seconds/warning, SD = 1.62) than they did the warnings without disgust visuals (M = 4.99 seconds/warning, SD = 1.52; t(436) = - 8.09, p < 0.01, d = 0.39). Thus, the combined results of these tests support hypothesis 39 1d which predicted that warnings with grotesque images would attract attention more slowly and for a shorter time than would warnings without such images. RQ1 asked whether there is any evidence that adolescents who live with a smoker are more attentive or avoidant of warnings with grotesque images than are adolescents who do not live with a smoker. To answer this question, a set of independent samples t-tests was run using the same variables to measure initial attention and duration of attention as in the preceding hypotheses tests. In terms of initial attention to the grotesque warnings, the results show no significant difference in ATFF between adolescents who live with a smoker (M = 0.18 seconds/warning, SD = 0.66) and those who do not (M = 0.18 seconds/warning, SD = 0.37; t(433) = -0.09, p = 0.93, d = .01). In terms of duration of attention to the warnings with grotesque imagery, the results show a marginally significant difference in AFD between adolescents who live with a smoker (M = 4.68 seconds/warning, SD = 1.52) and those who do not (M = 4.39 seconds/warning, SD = 1.69; t(434) = -1.86, p = 0.06, d = .18), with the trend suggesting adolescents who live with a smoker fixate on the grotesque warnings a bit longer. Hypotheses 2a and 2b concern attentional differences toward the pictorial and textual elements within the GWL for warnings with disgust-eliciting images versus warnings without such imagery. Within the warning, we predicted participants would allocate greater attention to the image (H2a) and lesser attention to the text (H2b) when viewing warnings with disgust visuals than when viewing warnings without. To test these hypotheses, we compared AFD for the image and the text between both types of warnings. Participants fixated on images longer when viewing warnings with 40 disgust visuals (M = 3.14 seconds/warning, SD = 1.35) than when viewing warnings without (M = 2.31 seconds/warning, SD = 0.94; t(435) = 16.87, p < 0.01, d = 0.81). Participants fixated for a shorter time on the text when viewing warnings with disgust visuals (M = 1.21 seconds/warning, SD = 0.78) than when viewing warnings without (M = 2.52 seconds/ warning, 1.04; t(435) =-30.59, p < 0.01, d = 1.46). Therefore, hypotheses 2a and 2b were both supported. H3 proposed that attention to the grotesque images within the warnings would predict greater risk perceptions. To test this hypothesis, the AFD variable for the disgust image AOI was entered into a multiple linear regression model with risk beliefs as the outcome variable. The AFD for the non-disgust image AOI variable and the text AOI variables were also entered as predictors in the model. Results do not support H3. Risk beliefs were not predicted by average fixation duration to the disgust imagery, B = -0.01, SE = .02, p = 0.66; 95% CI [-0.05, 0.03], and the overall regression model was not significant, F(14, 414) = .69, Adj. r2 = -.01, p = .79. RQ2 asked whether attention to the grotesque imagery predicted differences in changes in risk beliefs for adolescents who live with a smoker. To answer this question, four interaction variables multiplying the categorical variable for living with a smoker by the AFD for the image and text AOIs within each type of warning were added to the previous model. Similar to the previous hypotheses tests, none of the interaction terms were significant predictors of these participants’ risk beliefs in this model, F(15, 413) = .64, Adj. r2 = -.01, p = .84. The model reported in Table 3 shows the regression coefficient for just the interaction variable that accounts for participants’ living with a smoker status by AFD for the disgust images, B < -.01, SE 41 = .03, p = .97; 95% CI [-0.07, .06]. The results do not provide evidence of differences in changes in risk beliefs based on attention to the disgust images for participants who live with a smoker compared to those who do not. Hypothesis 4 predicted that increased attention to the grotesque images would increase negative emotional responses. To test this hypothesis, the average fixation duration variable for the disgust image AOI was entered into a multiple linear regression model with negative emotion as the outcome variable. As with our analysis for H3a and H3b, the non-disgust imagery variables and the text AOI variables were also included as predictors in the model. The regression model was significant, F(14, 408) = 1.80, Adj. r2 = .03, p = 0.04. AFD for the disgust images positively predicted negative emotion, B = 0.14, SE = 0.06, p = 0.02; 95% CI [0.03, 0.25]. Therefore, H4 was supported. Table 3 column 4 shows the result of this model including coefficients for the other warning AOIs as well as control variables such as participant demographics. RQ3 asked whether increased attention to the disgust visuals strengthened emotional responses for adolescents who live with a smoker more so than for those who don’t. To answer this question, the four interaction variables previously used for RQ2 were added to the model with negative emotion as the outcome variable. The interaction term for attention to disgust-eliciting images by adolescents’ status of living with a smoker was a marginally significant predictor of negative emotional reactions in this model, B = .19, SE = 0.11, p = .10, 95% CI [-0.03, 0.41]; the overall regression model was significant, F(18, 404) = 1.71, Adj. r2 = .03, p = 0.04. The remaining interaction terms were not significant predictors of negative emotional 42 reactions. These interaction terms were removed from the model, resulting in the second negative emotion model1 reported in Table 3, which includes the interaction term for attention to disgust-eliciting images by adolescents’ status of living with a smoker B = .19, SE = 0.09, p = .03, 95% CI [0.02, 0.35]; F(15, 407) = 2.02, Adj. r2 = .04, p = 0.01. Figure 3.3 presents a graphical depiction of the interaction effect of attention to the disgust imagery and living with a smoker on negative emotion. Table 3.3 Unstandardized Regression Coefficients with Standard Errors. Risk Risk Negative Negative Beliefs Beliefs1 Emotion Emotion1 Constant 3.67 (0.33)*** 3.67 (0.34)*** 3.33 (0.86)*** 3.75 (0.87)*** AFD Gross Image -0.01 (0.02) -0.01 (0.03) 0.14 (0.06)* 0.05 (0.07) AFD*Live w/Smoker - <-0.01 (0.03) - 0.19 (0.09)* AFD Not Gross Image 0.05 (0.03)# 0.05 (0.03)# -0.14 (0.08)# -0.13 (0.08)# AFD Gross Text 0.01 (0.03) 0.01 (0.03) 0.04 (0.09) 0.05 (0.09) AFD Not Gross Text -0.01 (0.03) -0.01 (0.03) -0.01 (0.07) -0.01 (0.07) Covariates Age <0.01 (0.03) <0.01 (0.03) -0.02 (0.06) -0.04 (0.06) Male (vs. female) -0.01 (0.04) -0.01 (0.04) -0.26 (0.12)* -0.25 (0.12)* Other sex 0.04 (0.14) 0.04 (0.14) -0.04 (0.36) -0.08 (0.35) Hispanic 0.02 (0.07) 0.02 (0.07) -0.13 (0.18) -0.14 (0.18) Black <0.01 (0.06) <0.01 (0.06) 0.27 (0.16)# 0.27 (0.16)# White 0.08 (0.06) 0.08 (0.06) 0.15 (0.16) 0.17 (0.16) Previous smoking -0.07 (0.08) -0.07 (0.08) -0.16 (0.22) -0.12 (0.22) Sensation seeking -0.03 (0.03) -0.03 (0.03) -0.20 (0.08)* -0.21 (0.08)** Colorblind 0.01 (0.09) 0.01 (0.09) 0.02 (0.23) <0.01 (0.23) Live w/Smoker -0.04 (0.04) -0.04 (0.11) 0.05 (0.12) -0.54 (0.29)# Adjusted R2 -0.01 -0.01 0.03 0.04 N 429 429 423 423 Note: 1regression model includes coefficient for interaction of attention to disgust image and living with a smoker. #p ≤ .10, *p ≤ .05, **p ≤ .01, ***p ≤ .001 43 5 4 3 2 1 0 2 4 6 8 10 Total Fixation Duration on Disgust Imagery Participants Living with a Smoker Others Figure 3.3 Interaction effect of attention and living with a smoker on negative emotions reported Discussion The results of this study provide evidence that disgust visuals in cigarette GWLs induce greater levels of overall avoidance than do graphic warning labels that do not contain disgust imagery. Consistent with the proposed hypotheses, however, when participants did look at the warning labels, they paid greater attention to the image over the text when the label contained a disgust visual. Participants also paid greater attention to the image when the warning label contained a disgust visual than they did when the warning did not. Attentional patterns were similar both for participants who live with smokers and for participants who do not, although trends suggest those who live with a smoker may attend to the grotesque warnings slightly longer than those who do not. Combined, these results support theoretical arguments 44 Negative Emotions that disgust objects both repel and attract attention, although in more subtle and nuanced ways than originally understood. Although participants paid less attention to the entire warning when disgust visuals were present, only their attention to the disgust-containing imagery predicted negative emotional responses. Additionally, attention to disgust imagery increased negative emotion at a greater rate for participants who reported living with a smoker. This finding is important because participants who live with a smoker are at a greater risk of smoking initiation (Hill et al., 2005). If the FDA proposed warnings (or an alternate version) were to be implemented, youth who live with a smoker would be more likely to see GWLs on cigarette packs belonging to smokers in their homes. Prior research has found a connection between exposure to GWLs that contain disgust elicitors and adolescents’ perceptions of smoking risk. The present study goes beyond exposure to examine how attention to disgust visuals relates to risk beliefs. In contrast to theories that disgust imagery can influence negative smoking beliefs, the reported findings do not show direct effects of attention to disgust visuals on smoking risk beliefs. This finding is consistent with previous results among adult smokers and nonsmokers. However, unlike these previous studies, the current results show no evidence of a relationship between attention to the textual content and smoking risk beliefs either (Brown & Richardson, 2012; Süssenbach et al., 2013). This finding calls into question the role of text in informing adolescent smoking beliefs. One possibility is that exposure to the GWLs induced defensive processing, as has been suggested in previous graphic warning research with adolescents (Ruiter & Kok, 2005). However, in both groups of participants, levels of risk beliefs about smoking were already 45 relatively high. This study was primarily concerned with attentional viewing patterns for disgust stimuli in GWLs and the impact of attention on negative emotional outcomes and risk beliefs among adolescents. Future research might consider whether negative emotion in response to GWLs could help to develop stronger risk beliefs over time, particularly for youth who live with a smoker. The current study is unable to speak to these questions because longitudinal data was not available. One strength of the current study is the use of eye tracking methodology among a large sample of adolescents from low-income communities, many of whom also live with a smoker. Graphic warnings have been shown to have positive downstream persuasive effects for adolescents (White et al., 2008). The current results suggest disgust imagery in the warnings may play a unique role in this process as only attention to these images predict increased negative emotional reactions, particularly for youth who live with a smoker. Limitations There are also several limitations to the study’s design. Earlier research identified a relationship between disgust exposure and risk beliefs among adolescents (Pechmann & Reibling, 2006). However, their study used images featuring young victims. The FDA-proposed GWLs do not depict consequences to young smokers, only very young children being exposed to secondhand smoke. This is a drawback of using the actual proposed images rather than those designed according to message considerations specific to the population; but it is also an important finding for future policy consideration. Further, GWLs in other countries and anti-smoking television ads may induce strong emotions due to higher vividness of the disgust visuals 46 employed. Future research should utilize eye tracking to measure youth attention to more vivid messages and examining the relationship between attention and emotions and risk beliefs. Other study limitations include selecting disease outcomes as measures of risk beliefs. A variety of beliefs have been identified including perceptions of message effectiveness, concerns about cosmetic and beauty risks, and normative smoking beliefs. Future studies should consider how attention to disgust visuals predict changes in additional message outcome variables. Finally, only a small percentage of participants reported prior smoking behavior. Given the adolescent smoking rates in the United States, research involving youth smokers, rather than nonsmokers, is also important and should also consider whether they exhibit any attentional biases in viewing disgust-eliciting warning labels, and how attention relates to emotions and risk beliefs as well as to quit intentions. Summary This study provides evidence that adolescents do attend to disgust imagery in graphic warning labels, despite some degree of overall warning avoidance, and supports prior arguments in favor of utilizing disgust visuals. The findings also point to additional questions about the impact of attention to such visuals, particularly negative emotional responses among youth at risk of smoking initiation. Future studies and pending policy decisions should consider extending this research to examine the role of attention to disgust-elicitors when featured victims are younger or when visuals are presented in more vivid formats, and how attention relates to other outcomes including recall and defensive processing. 47 CHAPTER 4 THE EFFECT OF ATTENTION AND DISCRETE EMOTIONS ON MESSAGE ELABORATION AND INTENTIONS IN RESPONSE TO A DISGUST-ELICITING CHARITABLE APPEAL This second study explores visual attention for disgust-eliciting messages compared to a neutral message condition in the context of prosocial persuasion. It considers differences in the vividness or intensity of disgust imagery, the negative emotions evoked in response to such stimuli, and how attention patterns and discrete emotions relate to cognitive responses and behavioral intentions. Introduction The factors that motivate charitable behavior have increasingly gained scholarly attention across many disciplines over the last several decades (Bekkers & Wiepking, 2011). Research related to prosocial persuasive messaging has been concerned with how the representation of social issues impact perceptions of need, helping intentions, and behavior. Studies examining factors such as the presentation of size and scope (Slovic, 2010), the identifiability of portrayed victims (Kogut & Ritov, 2005a, b; Small & Loewenstein, 2003), and the inclusion of statistical information (Dickert, Kleber, Peters, & Slovic, 2011) have deepened what is known about the process of charitable behavior. In contrast, the effect of emotionally-evocative images has been noted as an understudied aspect of this line of research (Hudson, VanHeerde-Hudson, Dasandi, & 48 Gaines, 2016). Recently, scholars have investigated the use of positive images (Perrine & Heather, 2000), negative images (Albouy, 2017), including shock tactics (Cockrill & Parsonage, 2016), and comparisons of these strategies (Burt & Strongman, 2005; Hudson et al., 2016). However, despite acknowledgment that persuasive messages give rise to a variety of distinct emotional responses, each of which motivate behavior differently (Dillard & Nabi, 2006), prosocial persuasion research has not fully incorporated a discrete emotions perspective toward imagery manipulation, emotion measurement, and outcome predictions. In the realm of health-related charity appeals, the emotion of disgust is particularly relevant as the depiction of infirmity often involves imagery that may evoke disgust (i.e. portraying people with gruesome diseases and/or deformities). As charity organizations face increasing competition for individual donor support (Ranganathan & Henley, 2008; Sargeant, 1999), it is important that their fundraising strategies distinguish them from other organizations, help their messages stand out to capture the attention of potential donors and motivate giving among the general populace. Intuitively, then, charities that use potentially disgust-evoking images may aim to draw attention to their messages, induce additional emotional reactions, increase awareness of the issue (i.e. get people to think about it more) and motivate charitable intentions. In contrast, emotion theories describe disgust as a low attention, low prosocial emotion (Haidt, 2003; Lerner, Li, Valdesolo, & Kassam, 2015). However, disgust’s ability to capture attention and motivate prosocial intentions in response to a charitable appeal has not been examined directly. The current study examines the relationship between attention to disgust-eliciting content, message 49 processing and behavioral intentions and considers the role of discrete emotions experienced in response to a charitable giving appeal. Attention for Disgust-Evoking Stimuli Our reactions to depictions of deformity and gore, extreme sickness, and disease pathogens may range from feeling “grossed out” (Nabi, 1998) to experiencing acute nausea (Ekman & Friesen, 1975). These subjective and physiological responses are unique to the emotion of disgust, and discrete emotion theories suggest that they influence our thoughts and behaviors in ways that can be distinguished from other negative emotions such as guilt or sadness (Izard, 1977). Functional theories of discrete emotion further propose that different emotions help us accomplish different goals (Lazarus, 1991). In the case of disgust, the goal is to reduce contact with the source of disgust. Its action tendency involves getting-rid-of or getting-away-from behavior including taking steps to stop perceiving whatever is eliciting the disgust response (Ekman & Friesen, 1975; Roseman, Wiest, & Swartz, 1994). Scholars have referred to disgust as triggering a look-away reaction (Joffe, 2008). Similarly, cognitive appraisal theories of emotion characterize disgust as generating low attentional activity both in terms of a willingness to attend to disgust stimuli and likelihood to give further thought or consideration to disgust eliciting events (Lerner et al., 2015; Smith & Ellsworth, 1985). Persuasion research has not yet applied discrete emotion theory to examine the relationship between disgust and attention, but in general, support for attentional avoidance of disgust-evoking stimuli is limited (Armstrong, McClenahan, Kittle, & Olatunji, 2014; Yartz & Hawk, 2002). 50 The dimensional view of emotion also describes disgust’s function as promoting avoidance behavior in that it triggers the aversive rather than appetitive system (Bolls, 2010). This avoidance goal is presumed to increase attentional vigilance, rather than withdrawal of attention, in order to identify a potential threat and facilitate avoidance (Clayton, Leshner, Tomko, Trull, & Piasecki, 2017). An attentional bias for disgust is only expected to give way to attentional avoidance if the intensity of the disgust-eliciting event breaches a certain threshold (Leshner et al., 2017). Studies applying the dimensional view provide substantial evidence that disgust depictions can increase attention toward a persuasive message (Clayton, Leshner, Bolls, & Thorson, 2016; Leshner, Bolls, & Wise, 2011; Leshner, Vultee, Bolls, & Moore, 2010). However, in these studies, attention is defined in terms of the allocation of cognitive resources to processing a message—often measured via cardiac deceleration. Additionally, this line of research has mainly considered the role of disgust in health appeals, anti-tobacco advertisements specifically, where disgust serves to augment fear (Morales, Wu, & Fitzsimons, 2012). Message attention is a key component of the persuasion process (McGuire, 1968). Thus, it is important to understand how the use of disgust imagery in various contexts actually influences attention toward a persuasive message, and how attention in turn influences persuasion outcomes. One thing that distinguishes health messages from other forms of persuasive messages, such as charity appeals, is the implication of a personally relevant potential threat. It is possible that attention is employed differently toward disgust-evoking stimuli in other persuasive domains. Regardless of research domain (i.e. health persuasion, nonprofit marketing, picture processing, 51 anxiety and emotion), few studies consider visual attention, (i.e. recording eye movements), in response to disgust stimuli (Schienle, Übel, Gremsl, Schöngassner, & Körner, 2016). Given the opposing theories, conflicting evidence, and gaps in the literature, the following research questions are posed regarding visual attention for a charitable appeal that employs disgust depictions: RQ1: Does exposure to disgust stimuli decrease or increase visual attention toward a charitable appeal? RQ2: Is there evidence of an attentional bias for the disgust-evoking imagery? RQ3: Does intensity of the disgust-evoking imagery exacerbate or attenuate these attentional patterns? RQ4: Does attention to disgust-evoking imagery predict subsequent cognitions and intentions in response to the message? Novelty and Emotional Forewarning Appraisal theories recognize the role of novelty, the extent to which an emotional event is unexpected or violates certain expectations, on emotional experiences (Scherer, 2009). Novelty appraisals are closely related to attentional activity in that novelty can influence whether an emotional event demands attention or not (Smith & Ellsworth, 1985). In the case of disgust, novelty may heighten the subjective experience and intensify the action tendency components of the emotional response (Herz, 2012). Accordingly, visceral repulsion and avoidance responses should be greater when disgust stimuli are encountered unexpectedly. Conversely, Herz (2012) suggests that anticipating exposure to a disgust-eliciting object may 52 function to reduce levels of disgust reactions because the actual disgust-eliciting experience is perceived as less disgusting than what was anticipated. Forewarning has classically been used to increase resistance to persuasive appeals (McGuire & Papageorgis, 1962) and more recently to reduce resistance to persuasion (Richards & Banas, 2014). Inoculation theory (McGuire, 1961) provides a framework for the use of forewarning as a persuasive strategy. An inoculation pretreatment includes both forewarning of exposure to a persuasive message and refutational preemption of one’s likely response to that message in an effort to encourage an alternative response. The present study applies Richards and Banas’ (2014) modification of the inoculation pretreatment to forewarn participants that the message “may evoke intense, unpleasant feelings” (see Appendix 3A). The forewarning also prompted participants that the message was important and worth considering to the end. The current study asks whether forewarning increased attentional activity and predicts that forewarning will attenuate the level of disgust participants report experiencing. Thus, RQ5: Does forewarning increase or reduce visual attention toward the message? H1: Message conditions preceded by the forewarning pretreatment will reduce levels of disgust experienced in comparison to message conditions without the forewarning pretreatment Cognitions and Prosocial Behavior The attentional activity appraisal also refers to willingness to consider the emotional event further (Smith & Ellsworth, 1985). Similar to predictions related to 53 visual attention, appraisal theories also characterize disgust in terms of low cognitive processing. Reactions to core disgust elicitors (i.e. depictions of disease and decay), in particular, have been described as fast and less thoughtful (Rubenking & Lang, 2014). However, cognitive processing has usually been measured using biophysical indicators as discussed previously or inferred from participants decisions in certain contexts. For instance, making a judgment in line with a stereotype rather than given evidence provides evidence of heuristic processing. The Elaboration Likelihood Model of Persuasion (ELM; Petty & Cacioppo, 1986) suggests that quantity of thoughts is also an indicator of whether participants engage in high- or low-level thinking in response to a persuasive message. Based on the logic of appraisal theories and the ELM, it is predicted that degree of elaboration will be negatively associated with level of disgust experienced. H2: Higher levels of disgust experienced will be negatively associated with quantity of thoughts expressed in response to the message. Pfister & Böhm (2008) propose that emotions not only direct our attention to relevant aspects of a decision context but also serve to commit us to morally or socially significant decisions. Prosocial behavior, including charitable giving, is a form of moral decision making that gives weight to the needs of others or benefits society as a whole (Batson, 2006). Although the relationship between disgust exposure, feelings of disgust and prosocial behavior have not been investigated directly, research on the implications of disgust for moral and social judgments suggest disgust is an emotion with low prosocial drive (Haidt, 2003; Sherman & 54 Haidt, 2011). Persons experiencing disgust demonstrate dehumanizing thoughts toward outgroups (Buckels & Trapnell, 2013) and render more harsh judgments of their choices and behavior (Wheatley & Haidt, 2005). Extending these reactions to behavior suggests disgust “commits” us to acting uncharitable towards others or to not enacting a prosocial behavior. Thus, there should be an inverse relationship between level of disgust experienced and behavioral intentions. H3: Higher levels of disgust experienced will be negatively associated with behavioral intentions in response to the message. Multiple Emotions Appraisal theories posit that emotion is a dynamic process (Moors, Ellsworth, Scherer, & Frijda, 2013). Thus, initial appraisals in response to an emotional appeal may lead to adjustments in how the situation is evaluated, which in turn may trigger new appraisals and additional emotional states (Smith & Ellsworth, 1985). Although a persuasive appeal may have a dominant emotional theme (Nabi, 1998), research has demonstrated that persuasive appeals induce multiple emotional states (Dillard & Peck, 2000; Dillard, Plotnick, Godbold, Freimuth, & Edgar, 1996), which may be recruited in a supporting role or may arise as unintended responses to the message (Coulter, Cotte, & Moore, 1999; Dillard & Shen, 2018). Regardless of whether additional emotions aroused are supportive or oppositional to the persuasive appeal, in keeping with the discrete view of emotion, each aroused emotion can influence subsequent thoughts and actions in diffuse ways according to their unique appraisal and functional components (Dillard & Nabi, 2006). 55 Previous research has linked guilt (Basil, Ridgway, & Basil, 2008) and sadness (Fultz, Schaller, & Cialdini, 1988; Small & Verrochi, 2009) with charitable behavior. Guilt in particular has received considerable attention in this vein (Basil, Ridgway, & Basil, 2006; Basil et al., 2008; O’Keefe, 2000, 2002). Guilt is described as occurring when one has violated a social or moral code (Lazarus, 1991). It can occur as the result of an action that has caused harm or in anticipation of inaction that may result in harm (Massi Lindsey, Yun, & Hill, 2007). In addition to its subjective component, “a gnawing feeling that one has done something wrong” (Nabi, 1999, p. 299), guilt involves a desire to remedy the situation; its action tendency involves reparation and compliance behavior (O’Keefe, 2000). Guilt is frequently solicited by prosocial appeals. Both anticipated guilt and existential guilt in response to a charitable appeal have been found to influence message cognitions and behavioral intentions (Massi Lindsey, 2005). However, there is also the potential for guilt appeals to back fire when guilt is evoked at higher levels (Coulter & Pinto, 1995), particularly when it regards a past violation that one be able to remedy (Renner, Lindenmeier, Tscheulin, & Drevs, 2013). Sadness is a response to loss or failure, real or imagined, in relation to oneself or someone else (Nabi, 1999). Whereas guilt ascribes personal responsibility for another’s misfortune, sadness appraises misfortune as being caused by situational factors beyond the other person’s control (Lerner et al., 2015). In theory, sadness motivates behavior to change the situation (Lerner, Small, & Loewenstein, 2004) and may promote prosocial action as a result (Fultz et al., 1988). Findings that sadness increases donations in response to a charity appeal are tenuous (Burt & Strongman, 56 2005; Small & Verrochi, 2009). However, there is substantial evidence that sadness promotes greater processing effort, most likely in an effort to reduce the uncertainty associated with events that evoke sadness (Bodenhausen, Sheppard, & Kramer, 1994; Xing, 2014). As a result, sadness should increase further consideration of the emotion situation. Based on this previous research and the expectation that disgust, sadness and guilt will have differential effects on persuasion outcomes, it is hypothesized that to the extent that participants experience sadness and guilt in response to the message, sadness and guilt will positively predict both elaboration of the message and behavioral intention. H4: There will be a positive association between a) level of guilt and b) level of sadness experienced and cognitive elaboration in response to the message. H5: There will be a positive association between a) level of guilt and b) level of sadness experienced and behavioral intentions in response to the message Method This study employed a 2 (disgust vividness: high/low) x 2 (forewarning pretreatment: present/not) + 1 (control) factorial design. The vividness manipulation was designed to create variation in emotional intensity of the treatment stimuli (high versus low). The control group received neither the emotion treatment nor the forewarning message; instead these participants saw pictures of maps that did not contain disgust elicitors. Thus, there were five conditions: high disgust/forewarning, high disgust/no forewarning, low disgust/forewarning, low disgust/no forewarning and the control condition. Participants were randomly assigned to a condition. 57 Participants.One hundred eighty-eight undergraduates at a large university in the northeastern United States participated in the study. Participants took the study for partial credit towards a communication or information science course for which they were enrolled during the semester they participated in the study. Data collection occurred between November 2015 and April 2016. The university’s IRB approved the study and each participant provided informed consent. All participants were 18 years or older; 87.8% of participants were between the ages of 19 and 22. Female students comprised 71.3% of the sample; 60.6% identified as Caucasian, 23.9% as Asian, and 12.2% as Black; 78.2% were born in the United States. The majority had never heard about NTDs before (N = 116), had no personal experience with the issue (N = 168), and did not feel at risk for being exposed to an NTD (N = 182). Table 4.1 provides descriptive statistics of the sample demographics. Procedure. The study was conducted in a research lab equipped with computers running the TobiiStudio 3.4.4 eye tracking software. The stimuli were displayed on Tobii T60XL 24” 1080p monitors which contain sensors to track eye movements. After completing informed consent, participants were seated at a computer, approximately 65 cm away from the monitor. The eye trackers’ sensors were adjusted to each participants gaze using 9-point calibration. In the rare case that calibration was not successful (N = 1), participants viewed the stimuli on screen, but no gaze data was collected. 58 Table 4.1 Participant Demographics Means (SD) or n’s (valid %) Age M = 20.87 (3.03) Sex Male 54 (28.7%) Female 134 (71.3%) Race White 114 (60.6%) Asian 45 (23.9%) Black 23 (12.2%) Other 18 (9.6%) Born in the United States 147 (78.2%) Participants were told they were participating in a study to evaluate messages about a global health issue. The on-screen study instructions were followed by a distractor task in which all participants read a news article on eating fruits and vegetables. The article was neutral in tone, and participants were not told it was unrelated to the main task. Participants controlled the duration of their exposure to the distractor material and the stimuli in that they could move on to the next item at any time. Following viewing the stimuli, participants completed a post-test questionnaire. Stimuli. Following the distractor material, participants in the two forewarning conditions saw a message explaining that the subsequent global health message, which they were about to see, “may evoke intense, unpleasant feelings” and that they “may feel the urge to stop viewing the message” (See Appendix 3A). The forewarning message also encouraged participants to consider the realities of the health issue and suggestions for how to address it. In the control group and the high and low disgust conditions without the forewarning pretreatment, participants saw a blank page in place of the forewarning message. The study instructions told all participants to click 59 through any blank pages that might appear to continue viewing the content. Clicking through the forewarning message or blank stand-in page brought participants to the first of three content pages for each condition. The content for this study is based on an actual campaign about Neglected Tropical Diseases (NTDs). The stimuli for each condition consisted of three static pages each with text and an image. The textual content included brief facts about the impact of NTDs, which affect more than one billion people worldwide. NTDs are easy and cost-effective to treat, but until recently have been a low priority in terms of health aid funding. In addition to a donation request, the message also included efficacy assurance that the advocated behavior, a 50-cent donation, was both possible for viewers to perform and an effective solution to the problem. The text and page format were consistent across all conditions. The image content was manipulated so that different pictures accompanied the text, varying by condition. The highly vivid image manipulation (high disgust) utilized three images, one per page, showing detailed, close-up depictions of people suffering from a tropical disease (i.e. hookworm, river blindness, elephantiasis). These images were designed to elicit greater disgust reactions than the less-vivid image manipulation (low disgust) which utilized three images that were less detailed and shot further away than those in the high disgust manipulation. Low disgust depictions also featured milder disease cases than high disgust depictions. The control message replaced the disease imagery with three pictures of maps, one for each continent where NTDs are most prevalent, which should not elicit feelings of disgust. Figure 4.1 a, b and c show samples of the stimuli used in the experiment. 60 Image Selection and Pretest. In order to obtain images that differed in the degree of disgust depicted, a pretest was conducted with 20 images selected from the NTD campaign as well as from the Open Access Biomedical Image Search Engine (Open-i). Eight undergraduate research assistants in a communication research lab completed the pre-test as part of their research credit responsibilities. IRB approval and informed consent was obtained for their participation, and they were not made aware of the purpose of the study. a. b. c. Figure 4.1 Sample Stimuli: a. high disgust b. low disgust c. control/neutral Pre-test participants viewed each of the images and responded on a scale of 1 to 7 to indicate “the extent to which you feel grossed out right now.” Participants also responded to 5 statements about physiological indicators of disgust responses such as “made my skin crawl” and “made my stomach turn” from “strongly disagree” (1) to “strongly agree” (5). High disgust images were defined as those that received a mean rating of 5 or greater on the grossed-out item (whose midpoint was 4) and a mean greater than 3 (the midpoint) on the physiological disgust matrix (Nimages = 5). Low disgust images were defined as those that received a mean rating of 4 or lower on the grossed-out item and a mean rating lower than 3 on the physiological disgust items (Nimages = 7). In making the final image selections, we featured different diseases while 61 keeping image type (i.e. faces or feet) consistent within disgust conditions. Of the 20 possibilities, the 3 images selected for the highly vivid disgust stimuli had a mean rating ranging from 5.38 to 6.0 on the grossed-out item and a mean rating of 3.58 to 3.86 on the physiological disgust matrix. The 3 images selected for the low vivid disgust stimuli had a mean rating ranging from 2.75 to 2.88 on the grossed-out item and a mean rating of 2.73 to 2.90 on the physiological disgust matrix. Measures Objective attention measures were collected via eye tracking, and a posttest questionnaire captured subjective responses including disgust, guilt, sadness, elaboration and behavioral intention. Visual Attention. Total fixation duration (TFD) is a sum of overall viewing time across the three static messages comprising each stimulus. TFD measures, reported in seconds, indicate how long participants looked at three areas of interest (AOIs): the overall message (M = 16.73, SD = 8.55), the image (M = 5.87, SD = 4.38) and the text (M = 9.79, SD = 5.45). Disgust. Five items measured the extent to which participants felt disgust in response to the stimuli. Measures included items reflecting both the subjective and physiological components of disgust (Lazarus, 1991). Participants indicated the extent to which they felt disgusted and grossed out (Nabi, 2002) on a scale of 1 (not at all) to 7 (extremely) and the extent to which they felt nauseous, felt their skin crawl, and their stomach turn on a scale of 1 (strongly agree) to 5 (strongly disagree). The average of these items forms a single index of disgust reactions to the stimuli (Cronbach’s α = .89; M = 2.73, SD = 1.23, range 1-5.4). 62 Guilt and Sadness. Guilty and sad reactions to the stimuli were each obtained using single-item measures rated on a scale of 1 (not at all) to 7 (extremely). After viewing the content, participants indicated how guilty (M = 2.56, SD = 1.54) and sad (M = 3.23, SD = 1.8) they felt. Elaboration. A thought-listing task prompted participants to describe “all the thoughts and feelings” they experienced while viewing the message. Total word count in response to this item was taken as an indicator of the extent to which each participant engaged in processing the message (M = 40.6, SD = 29.01). Behavioral Intention. Participants indicated their intention to donate by responding to the following three items, all rated on 5-point scales: “After seeing the ads I want to make a donation.” (strongly disagree to strongly agree); “How likely are you to make a donation to the organization?” (not very likely to very likely); and “If you were to make a donation, how soon would you do so?” (not at all soon to very soon). The average of these items forms a single index of behavior intention (Cronbach’s α = .73; M = 2.84, SD = .91). Tables 4.2 reports the bivariate correlations between all the dependent variables. Covariate. Gender was included as a covariate given substantial evidence that women are often more sensitive to disgust evocation than men (Judah et al., 2009; Schienle et al., 2016). 63 Table 4.2 Bivariate Correlations of Dependent Variables 1 2 3 4 5 6 7 8 1. Attention to Message - 2. Attention to Image .80*** - 3. Attention to Text .86*** .41*** - 4. Disgust -0.07 0.06 -.17* - 5. Sadness 0.07 0.19** -.03 .59*** - 6. Guilt 0.09 0.10 0.05 .53*** .60*** - 7. Elaboration .22** .23*** .16* 0.04 .25*** 0.09 - 8. Behavioral Intention 0.11 .01 0.16* .20* .27*** .37*** 0.13 - ***p ≤ .001; **p ≤ .01; *p≤.05 Results Table 4.3 reports the mean and confidence intervals for each of the dependent variables by condition. Attention. To answer RQ1 through RQ3, multiple linear regression models using ordinary least squares (OLS) estimation were run with dummy variables for each of the disgust-evoking treatment conditions leaving the control group as the comparison condition. RQ1 asked whether exposure to disgust stimuli decrease or increase visual attention toward a charitable appeal relative to the control condition. Data show no significant differences in attention between any of the treatment conditions and the control group (see Table 4.4). Although total fixation duration (TFD) on the message as a whole was lower for both high disgust conditions and higher for both low disgust conditions these differences were not significant in comparison to TFD for the control condition. This 64 result also suggests that intensity of the disgust-evoking stimuli neither exacerbated nor attenuated attention or avoidance (RQ2). Table 4.3 Means and Confidence Intervals for Dependent Variables TFD Message TFD Image TFD Text Disgust N Control 16.40 3.39 12.19 1.59 38 [14.12, 18.69] [2.70, 4.08] [10.43, 13.94] [1.41, 1.76] No Forewarning Low Disgust 17.84 6.89 9.56 2.60 38 [14.97, 20.71] [5.36, 8.41] [7.76, 11.35] [2.24, 2.96] High Disgust 16.22 5.83 9.32 3.76 37 [13.50, 18.93] [4.51, 7.15] [7.52, 11.11] [3.36, 4.15] Forewarning Low Disgust 17.61 6.51 10.03 2.54 37 [15.09, 20.13] [5.15, 7.86] [8.46,11.59] [2.18, 2.89] High Disgust 15.58 6.68 7.93 3.22 38 [11.96, 19.20] [4.84, 8.51] [6.04, 9.81] [2.88, 3.55] Sadness Guilt Elaboration Intention N Control 1.84 1.68 31.97 2.58 38 [1.49, 2.18] [1.33, 2.02] [23.33, 40.62] [2.25, 2.90] No Forewarning Low Disgust 3.08 2.61 37.95 2.88 38 [2.50, 3.66] [2.14, 3.07] [30.36, 45.54] [2.60, 3.16] High Disgust 4.00 3.22 41.49 2.83 37 [3.49, 4.51] [2.74, 3.70] [31.81, 51.16] [2.54, 3.12] Forewarning Low Disgust 3.51 2.51 44.51 2.90 37 [2.90, 4.13] [1.95, 3.08] [33.47, 55.56] [2.59, 3.22] High Disgust 3.76 2.82 47.74 3.03 38 [3.14, 4.38] [2.26, 3.37] [37.29, 58.18] [2.73, 3.32] Note: TFD = Total Fixation Duration in seconds Findings indicate increased attention toward the image in all disgust-evoking conditions compared to control (RQ3). Although this viewing pattern also decreased attention toward the text, overall attention toward the text (M = 9.79 seconds, SD = 5.44) remained greater than attention for the image (M = 5.87 seconds, SD = 4.38; 65 t(186) = 9.96, p < 0.001) in all conditions (see AOI means by condition reported in Table 4.3). Table 4.4 Main Effects of Exposure on All Dependent Variables compared to Control TFD Message TFD Image TFD Text Disgust No Forewarning Low Disgust 1.45 (1.99) 3.53 (.98) *** -2.67 (1.23)* 1.00 (.23)*** High Disgust -.17 (2.01) 2.49 (.99) ** -2.93 (1.24)** 2.15 (.23)*** Forewarning Low Disgust 1.22 (2.01) 3.16 (.99) ** -2.22 (1.24)# .93 (.23)*** High Disgust -.81 (1.99) 3.34 (.98) *** -4.32 (1.23)*** 1.60 (.23)*** Gender .19 (1.40) .59 (.69) -.69 (.86) -.32 (.16)* Constant 16.34 (1.50)*** 3.18 (.74) *** 12.43 (.93)*** 1.69 (.17)*** Adj. R -.02 .07 .04 .35 N 187 187 187 188 Sadness Guilt Elaboration Intention No Forewarning Low Disgust 1.26 (.38)*** .95 (.34)** 6.35 (6.62) .28 (.21) High Disgust 2.17 (.38)*** 1.56 (.34)*** 9.86 (6.67) .23 (.21) Forewarning Low Disgust 1.69 (.38)*** .85 (.34)** 12.88 (6.67)# .30 (.21) High Disgust 1.94 (.38)*** 1.15 (.34)*** 16.09 (6.63)* .42 (.21)* Gender -.14 (.27) -.04 (.24) -1.83 (4.66) -.32 (.14)* Constant 1.86 (.28)* 1.67 (.25)*** 32.13 (4.94)*** 2.69 (.15)*** Adj. R .17 .09 .01 .03 N 188 188 188 188 ***p ≤ .001; **p ≤ .01; *p≤.05; # p≤.10 RQ4 asked whether attention to the disgust-evoking imagery predicts subsequent cognitions and intentions in response to the message. To address this question, two regression models were run with elaboration and intention as the dependent variables and the TFD for image AOI as a predictor variable, along with the emotion variables (H2 – H5). The model also controls for TFD on the text AOI (see Table 4.6). Data from the control group were excluded from these analyses as the 66 control group did not see disgust-evoking imagery. Results show that attention to the disgust-evoking imagery is not significantly associated with elaboration or behavioral intentions. Table 4.5 Indirect Effects of Attention and Emotion Elaboration Intention TFD Image .60 (.61) -.03 (.02) TFD Text .56 (.52) .03 (.02)* Disgust -4.68 (2.40)* -.07 (.07) Sadness 4.94 (1.66)** .06 (.05) Guilt -1.27 (1.85) .16 (.05)** Gender -1.99 (5.34) -.31 (.15)* Constant 34.47 (8.46)*** 2.42 (.24)*** Adj. R .07 .15 N 150 150 ***p ≤ .001, **p ≤ .01, *p≤.05; #p≤.10 Novelty. RQ5 asked whether reducing novelty via forewarning increases visual attention toward the message. H1 predicted that forewarning would reduce levels of disgust experienced in comparison to message conditions without the forewarning pretreatment. Two multiple regression models were run for each dependent variable, message attention and disgust, alternating the high disgust and low disgust conditions as the comparison group. Results show no significant effects of forewarning on overall message attention for either the high or low disgust conditions. Forewarning significantly decreased feelings of disgust in response to the high disgust message, but forewarning did not decrease feelings of disgust in response to the low disgust condition. Thus, H1 is only partially supported. 67 Cognitions and behavior. Table 4.4 shows that participants in each of the treatment groups not only experienced increased levels of disgust compared to the control condition, they also reported increased levels of sadness and guilt in response. H2 through H5 predict that each emotion will uniquely predict cognitions and behavioral intentions in distinct ways. Specifically, H2 and H3 predict that reported feelings of disgust will be negatively associated with elaboration and behavior. Table 4.6 Forewarning Effects on Attention and Disgust TFD Messagea TFD Messageb Disgusta Disgustb No Forewarning Low Disgust 1.62 (1.99) -- -1.15 (.23)*** -- High Disgust -- -1.62 (1.99) -- 1.15 (.23)*** Forewarning Low Disgust 1.40 (2.00) -.22 (1.99) -1.22 (.23)*** -.07 (.23) High Disgust -.64 (1.99) -2.26 (.1.98) -.54 (.23)** .61 (.23)** Control .17 (2.00) -1.45 (1.99) -2.15 (.23)*** -1.00 (.23)*** Gender -.19 (1.39) .19 (1.39) -.32 (.16)* -.32 (.16)* Constant 16.17 (1.47)*** 17.79 (1.46)*** 3.84 (.17)*** 2.69 (.17)*** Adj. R -.02 -.02 .35 .35 N 187 187 187 187 Note: aHigh Disgust is comparison bLow Disgust is comparison. ***p ≤ .001, **p ≤ .01, *p≤.05; #p≤.10 H4 and H5 predict that both reported sadness and guilt will be positively associated with elaboration and behavior. To test these hypotheses, two regression models were run with elaboration and behavioral intention as the dependent variables (see Table 4.5). As previously described, these models include attention to the text and image as predictor variables along with disgust, sadness and guilt. Results show a significant, negative association between disgust and elaboration as predicted. Thus, H2 is supported. The association between disgust and 68 behavioral intention is not significant. Thus, H3 is not supported. H4a is also not supported as the association between guilt and elaboration is not significant. However, there is a significant, positive relationship between sadness and elaboration as predicted, thus supporting H4b. In contrast, H5a is supported—there is a significant, positive association between guilt and behavioral intention; but H5b is not supported—the association between sadness and behavioral intention is not significant. Discussion This study examined the use of disgust-eliciting imagery in the context of a charitable appeal. It asked whether such imagery increases visual attention to the persuasive message as a whole and to the disgust imagery in particular. Compared to the control condition, mean trends suggest an increase in visual attention for messages with low vivid disgust imagery, consistent with the dimensional view that disgust triggers attentional vigilance, and a decrease in attention for messages with highly vivid disgust imagery, consistent with the functional emotion view that disgust triggers attentional avoidance. However, these differences were not significant, thus we can only conclude that disgust depictions did not significantly increase or decrease attention for the message overall. Further, including a forewarning pretreatment had no apparent effect on duration of attention to the overall message compared to treatment conditions without forewarning. One interpretation of this finding is that novelty or unexpected exposure to a disgust-eliciting event is not related to attentional activity, and only influences the magnitude of emotion experienced in response to the 69 message, as was found in differences in disgust reported between the highly vivid conditions with and without forewarning. The inclusion of disgust imagery did increase attention to the image AOI in all treatment conditions compared to control. This finding provides some support for the dimensional view that disgust, as a negatively valenced emotion, increases attentional activity toward the stimulus. Previous research has found that disgust elicitors attract and hold attention to the exclusion of other stimuli (Miller & Leshner, 2007; Van Hooff, Devue, Vieweg, & Theeuwes, 2013); however, the shift in attention observed here does not clearly indicate an attentional bias for disgust. Although there was an increase in attention toward the image when it contained a disgust elicitor, and thus a decrease in attention to the text, overall attention to the text remained higher than attention to the image across all conditions. There are at least two possible explanations for this finding. Disgust imagery may initially increase attentional activity relative to neutral content, which in turn triggers feelings of disgust and high motivation to avoid the content. Alternatively, participants may have been less willing to devote further attention to the images once they determined the images did not reference a personal threat, perhaps moving on to the text to understand the purpose of the message. The latter explanation is also consistent with the discrete view in terms of the low personal relevance and low attention appraisals associated with disgust. In addition to identifying visual attention patterns, this study also examined the relationship between attention and other message outcomes. Attention to the image did not predict elaboration or behavior. Although there were no research questions regarding effects of attention to the text, attention to the text was associated with 70 behavioral intention. Given that participants spent more time looking at the text, and the text contained the call to action, this finding seems logical. The lack of effects for attention to the image and the relationship between attention to the text and behavioral intentions suggests the influence of disgust elicitors may not be as powerful when presented as part of complex social stimuli such as a prosocial persuasive message. The persuasion context and the juxtaposition of disgust with rational arguments may alter gut reactions to disgust stimuli influencing the way disgust eliciting events are processed. Future research can directly manipulate the inclusion of text or other message features to clarify the influence of their interaction with the disgust-elicitors. It was predicted that disgust-eliciting messages would evoke multiple emotions in addition to disgust including sadness and guilt. The research hypotheses predicted each emotion would have differential associations with message processing and behavioral intentions. This general hypothesis was supported. As predicted, disgust reduced elaboration whereas sadness increased elaboration, and guilt increased behavioral intentions. However, sadness was not associated with behavior and guilt was not associated with elaboration. Previous research has found a relationship between guilt and cognitive measures such as perceived effectiveness, but such measures do not presume that people engage in higher levels of thinking. Although previous studies have connected sadness with both giving and taking, functional emotion theories describe sadness as immobilizing or creating a state of inaction. It could be that deeper thinking was sufficient to address the goal state of sadness, namely the desire to change one’s situation, thereby alleviating or overriding its potential to motivate prosocial behavior. Thus, these findings, although seemingly 71 inconsistent with previous research are not wholly inconsistent with theoretical predictions. The relationship between disgust elicitation and prosocial behavior has not been examined previously. This study provides initial evidence that in the context of a persuasive charitable appeal, eliciting disgust may not have the deleterious effects that have been observed in the moral judgment domain. Feelings of disgust did not predict behavioral intention in contrast to the negative effects hypothesized. To the extent that disgust appeals recruit other prosocial emotions, it appears they can have positive influences on message relevant cognitions and behaviors. Of course, other unintended consequences, such as reinforcing demeaning stereotypes or reducing empathetic concern, remain to be examined in future studies. The current findings are consistent with previous research showing that men report lower levels of disgust than women. This difference appears to be unique for the emotion of disgust as gender did not predict differences in reported levels of guilt or sadness. Gender was also a significant predictor of behavioral intention in both the main effects and indirect effects models. From a practical perspective, the finding that disgust-evoking stimuli are no more attention-getting than neutral stimuli is disconcerting given the presumed importance of message attention in the persuasion process. Although it has been argued that disgust objects “stick in the mind” (Joffe, 2008, p. 87), and by extension may not require increased visual attention to have influence, there is no evidence in the current study that exposure to disgust evoking-stimuli or levels of disgust experienced increase cognitive processing of the message either. However, to the 72 extent that disgust-evoking charitable appeals also evoke sadness and guilt, the current study suggests they can increase message relevant cognitions and promote behavioral intentions. Alerting audiences of impending exposure to a disgust depiction appears to allow them to cope with more vivid disgust portrayals. Given the potential for feelings of disgust to manifest unintended consequences, such as reducing issue elaboration, forewarning may help to attenuate potential negative effects of high disgust elicitation. However, forewarning also appears to reduce other emotional responses to the message as the trends for sadness and guilt suggest (see Table 4.3). As these co- occurring emotions may facilitate positive message outcomes, message planners should be aware of the potential benefits and drawbacks of using strategies that alter the novelty of persuasive disgust appeals. Limitations The current findings should be interpreted within the context of the study’s limitations. Participants saw the message stimuli at a single point in time void of the larger media and social context in which charitable appeals are often encountered. Therefore, analyses cannot account for the complementary or competing influences of other content that participants were or could potentially be exposed to. The study sample consisted of undergraduate students whose sociodemographic profiles, economic considerations and message reactions may not reflect the typical target audience of charitable appeals (i.e. persons employed fulltime). Additionally, the sample is predominantly female which may have biased estimates of the influence of gender on message outcomes. 73 Although using a neutral comparison image allows for an absolute measure of the effects of the disgust imagery, it does not address effects compared to other emotion strategies. Given the real-world nature of the stimuli, experimental control of the image manipulation was limited. It is possible that differences in low-level features of the images, such as color, contrast and luminance, influenced the eye tracking results (Wedel & Pieters, 2006), but both the pilot and main study show that emotional reactions to the two groups of images varied significantly. The study focused on differential effects between negative emotions; the potential for disgust-evoking appeals to evoke positive emotions that in turn influence message outcomes was not considered. Thus, potential predictors are unobserved in the analyses. Likewise, potential unintended effects were not explored. Single items were used to measure sadness and guilt which may raise concerns about their predictive validity. However, single-item measures can be effective when they represent homogenous constructs and additional item measures are semantically redundant (Diamantopoulos, Sarstedt, Fuchs, Wilczynski, & Kaiser, 2012). This study treats the attention-emotion relationship as unidirectional using a collective measure of attention, rather than a cyclical process with feedback loops. It is possible that changes in attention occurred overtime as additional emotional appraisals were made. Future research should employ methods suitable for examining this process. Finally, the results of this study provide no guidance regarding potential unintended consequences of disgust elicitation, particularly as it relates to ethical concerns that scholars and fundraising organizations have raised (Dochas, 2007; Lupton, 2014). It is also not known how this strategy compares to appeals that evoke 74 sadness or guilt directly and without also evoking disgust. Sadness and guilt inductions may also involve unintended effects (Coulter & Pinto, 1995; Polman & Kim, 2013). One interesting question is whether evoking these emotions implicitly, such as with disgust-evoking imagery, attenuates their potential drawbacks. Overall, message designers should use caution in employing disgust for prosocial persuasive outcomes; although positive effects seem likely, negative effects remain unclear. Summary The current study examined attention for disgust-eliciting charitable appeals that vary in emotional intensity. Results show that regardless of vividness, the disgust- eliciting appeals do not attract or repel attention significantly more than appeals that use a relatively neutral image. Although a forewarning message that alerted participants they were about to view emotionally disturbing content reduced feelings of disgust in response to the highly vivid appeal, no such effect occurred for the low vivid appeal and forewarning did not appear to influence attentional patterns either. Feelings of disgust negatively predicted message elaboration but was unrelated to behavioral intentions. Findings suggest, notably, that disgust-eliciting appeals can positively influence message outcomes by recruiting more prosocial emotions (sadness and guilt) that have significant, positive effects on elaboration and behavior intentions. Future research should consider whether disgust and positive prosocial emotions such as empathy can cooccur and whether there are unintended consequences of utilizing disgust in prosocial persuasion. 75 CHAPTER 5 ENCOUNTERING DISGUST IN PROSOCIAL PERSUASION: HOW CONTEXT SHAPES APPRAISALS, EMOTIONS, AND BEHAVIOR Disgust has been described as an unreasoning emotion—insensitive to the context in which it is elicited. The current study aims to test this assumption by examining changes in cognitive appraisals, emotional reactions and behavioral responses that may occur when disgust elicitors are encountered in a persuasive prosocial appeal. Specifically, it asks whether the prosocial context influences the certainty appraisal often associated with disgust. The previous chapter established that a disgust-eliciting appeal can evoke additional negative emotions and impact donation intentions. The current chapter considers the possibility that a disgust-eliciting appeal may also elicit empathy and examines its influence on donation behavior relative to disgust elicitors without a persuasive appeal and relative to persuasive appeals that do not evoke disgust. Introduction Several years ago, the Sabin Vaccine Institute launched the End7 campaign to raise awareness of and funding for the treatment of seven neglected tropical diseases. One campaign video featured explicitly grotesque imagery of people with parasites like the roundworm crawling out of their noses and mouths, or with enormous feet deformed by elephantiasis, and other such imagery (Sabin Vaccine Institute, 2013). Likewise, during the recent Zika crisis, Metropolitan Ministries included pictures of children covered in sores in their appeals for funding to provide mosquito nets and 76 repellent to homeless families within the United States in areas deemed to be at risk of an outbreak (Metropolitan Ministries, 2016). With the rise of crowdfunding social platforms (Gierczak, M. M., Bretschneider, U., Haas, P., Blohm, I., & Leimeister, 2016), examples of grotesque imagery accompanying requests for help are not limited to communication from non- profit fundraising organizations. Intuitively, there may be a notion that if you show people how bad an issue is, they will be motivated to provide a solution for the problem. This may explain the more recent phenomenon of a cancer patient who willingly posted a picture of her puss-ridden, inflamed breast and a young man who circulated images of his mangled face after he was severely injured in a motorcycle accident, both in hopes of raising money for medical expenses.1 All of these examples fit the description of disgust-inducing stimuli as they include depictions of deformity and gore, extreme sickness, bodily fluids and secretions, and disease pathogens (Haidt, McCauley, & Rozin, 1994). Exposure to such stimuli has reliably been shown to elicit the emotion of disgust (Rozin, Haidt, & McCauley, 2008). Because disgust content is often not personally relevant, particularly when it depicts another’s need, the emotion’s rejection response or instinct to “shut off” contact with the disgust-elicitor may be particularly strong. The question then is whether exposure to disgust depictions can “draw one closer” to the depicted other and what factors allow for this to occur. 1 For examples, see https://www.facebook.com/valderine.mckinney/posts/1118778278259077 and https://www.gofundme.com/motorcycle-accident-recovery-2vnaxyk 77 Disgust and Moral Decision Making Charity associations warn against the use of shocking imagery, such as those that may elicit disgust, in fundraising appeals citing concerns that such imagery may demean the persons portrayed and perpetuate stereotypical and discriminatory beliefs (Dochas, 2007). From a psychological perspective, disgust has been described as symbolically protecting against contamination including social contamination (Haidt, Rozin, Mccauley, & Imada, 1997). Thus, scholars have raised similar ethical concerns about disgust appeals and their potential to stigmatize those portrayed as social contaminants (Lupton, 2014). There is some empirical evidence supporting this perspective. Buckels and Trapnell (2013) found that disgust-inducements contribute to the dehumanization of an outgroup. Similarly, Dasgupta and colleagues (2009) found disgust-inducements activated implicit bias towards different outgroups. Research on moral judgment also shows that disgust-inducements increase the severity of moral condemnation (Wheatley & Haidt, 2005), and this effect appears to be amplified for stigmatized social groups (Inbar, Pizarro, & Bloom, 2012; cf. Landy & Goodwin, 2015). Based on these findings, scholars in this field have surmised that disgust shrinks our moral circle and persuades us that certain groups are not worthy of concern (Pizarro, Detweiler-Bedell, & Bloom, 2006; Sherman & Haidt, 2011). One challenge to translating this research to issues beyond moral condemnation is that disgust-inducements in these studies are incidental to the judgment task. For instance, Buckels and Trapnell (2013) showed participants images of open heart surgery, a dirty toilet, and a cockroach on a plate before conducting an implicit attitude test measuring beliefs about an arbitrarily created outgroup. Other 78 studies have triggered disgust by exposing participants to noxious odors or having them sit at a dirty desk (Schnall, Haidt, Clore, & Jordan, 2008), giving them a bitter beverage to drink (Eskine, Kacinik, & Prinz, 2011), or having them listen to the sound of vomiting (Seidel & Prinz, 2013). All of these examples of disgust-eliciting stimuli were external and irrelevant to the social context. In contrast, integral emotion inductions are triggered by the decision situation and are relevant to the decision task and may have different effects on message processing (Nabi, 2003). Examining integral emotions may reveal how emotions are most likely to influence prosocial judgments and behavior (Zeelenberg, Nelissen, Breugelmans, & Pieters, 2008). To my knowledge, neither the persuasion or moral psychology literature has addressed disgust evocation as an integral function when it is being used strategically to induce prosocial or charitable behavior. This gap in knowledge may be a result of the assumption that disgust is insensitive to context (unreasoning) and that it manifests consistently to stimulus inductions regardless if those inductions are external or integral (Russell & Giner-Sorolla, 2013). The current study aims to test this assumption by examining responses when disgust cues are used in a prosocial persuasive context compared to when disgust cues are not connected to the decision-making context (i.e. there is no expectation of prosocial responses). The Heuristic-Systematic Model (Chaiken, 1980), Appraisal- Tendency Framework (Lerner & Keltner, 2000) and the Social Intuitionist Model of Moral Judgment (Haidt, 2001) each consider the role of emotion in information processing and decision making. Next, I briefly discuss these models and the intersection of key concepts from each that prove useful for framing my hypotheses. 79 Feeling and Thinking in Prosocial Decision Making The Heuristic-Systematic Processing Model (Chaiken, 1980) accounts for the influence of both affective elements and cognitive arguments on information processing. Heuristic processing is more effortless and automatic compared to the deliberative systematic processing mode, relying on easily accessible cues to determine an appropriate response to a social situation (Shah & Oppenheimer, 2008). Emotions are often associated with heuristic processing (Slovic, Finucane, Peters, & MacGregor, 2007) in as much as they provide visceral cues or rules of thumb that compel behavior. For instance, the inflexible avoidance response ascribed to disgust could serve as a cue that the appropriate response is to reject or distance oneself from an object. This judgment could be made without the need for deliberation. Certainty Appraisals and Information Processing However, recent research has demonstrated that other qualities of an emotion, beyond action tendencies, may lead to systematic processing instead (Inbar & Gilovich, 2011; Lerner & Tiedens, 2006). According to cognitive appraisal theories (Smith & Ellsworth, 1985) emotions can be categorized along various dimensions such as pleasantness, control or certainty. The certainty dimension refers to the degree to which an emotion is accompanied by a sense that future events are predictable and confidence in how one should respond. The Appraisal-Tendency Framework (ATF) (Lerner, Li, Valdesolo, & Kassam, 2015) posits that certainty increases reliance on easily accessible cues, such as an emotion’s behavioral tendency, and thus initiates heuristic processing. In contrast, emotions with uncertainty appraisals, such as sadness, should initiate systematic processing. 80 The ATF framework characterizes disgust as a high-certainty emotion (i.e. certainty that the rejection response is appropriate). Tiedens and Linton (2001) demonstrate that incidental disgust is associated with heuristic processing. Participants who experienced a disgust induction not only expressed greater certainty, they also appeared to engage in heuristic processing relying more on a stereotype cue in forming their judgment of a moral event. Similarly, it would be consistent with ATF to predict that when participants are exposed to disgust eliciting content alone (i.e. without the persuasive context) they will express greater certainty regarding the emotional experience and how they should respond to it. If so, they may also exhibit signs of heuristic processing. Prior research examined participants reliance on stereotype cues as evidence of heuristic processing. The goal of the current study is to examine the influence of disgust on making or reinforcing stereotype judgments. Thus, reliance on heuristic processing as a result of the disgust elicitation may be evinced by evaluations of various target groups that conform to respective stereotypes about each group (Cuddy, Fiske, & Glick, 2008). H1a: Exposure to disgust cues unaccompanied by a persuasive prosocial context will result in higher certainty appraisal ratings. H2a: Exposure to disgust cues unaccompanied by a persuasive prosocial context will result in outcomes consistent with heuristic processing on a subsequent social judgment task (higher stereotypes). It is possible to alter an emotion’s certainty appraisal (Tiedens & Linton, 2001). Although this has not previously been shown for disgust, the Social Intuitionist Model suggests that people rely on reasoning more when they have conflicting 81 intuitions in response to a decision situation. This experience of conflicting intuitions may be particularly true for charitable appeals that seek to evoke empathetic concern while employing disgust-eliciting visuals. If so, this conflict between feelings of disgust and contextual norms of caring may produce uncertainty about what one should do next. This uncertainty should lead to systematic processing of future decision tasks. H1b: Exposure to disgust cues accompanied by a persuasive prosocial context will result in lower certainty appraisal ratings. H2b: Exposure to disgust cues accompanied by a persuasive prosocial context will result in outcomes consistent with systematic processing on a subsequent social judgment task (lower stereotypes). Disgust Reappraisal, Distress and Empathy Although specific emotions are associated with certain core appraisals, changes or shifts in these appraisal patterns may lead to the recruitment of new emotions in addition to the emotion most proximally evoked by the emotion situation (Moors, Ellsworth, Scherer, & Frijda, 2013). As seen in the previous chapter, other negatively-valenced emotions may co-occur with disgust. An open question, unexamined in the previous chapter, is whether positively-valenced emotions may be evoked by disgust stimuli. The possibility for empathy to co-occur with disgust is particularly interesting because empathy and disgust are often characterized as opposite emotions in terms of their moral inclinations and prosocial motivations (Pizarro et al., 2006). Empathetic concern—the desire to protect the welfare of another—is considered a key motivator of prosocial behavior (Batson, Lishner, & 82 Stocks, 2015; Shelton & Rogers, 1981). It is distinguished from personal distress— defined as a self-focused emotional reaction to the circumstances of another person in need (Fultz, Schaller, & Cialdini, 1988)—which has also been found to motivate prosocial behavior. Scholars have suggested that empathy and distress have different implications for helping behavior (which I discuss in further detail later). By definition, disgust- eliciting events are likely to increase feelings of personal distress in as much as they focus attention on one’s need to escape the disgust-eliciting event (Haidt, 2003; Smith & Ellsworth, 1985). But, these events need not elicit empathy as prior research suggests exposure to disgust elicitors diminishes moral concern for others. Therefore, finding increases in empathetic concern in response to a persuasive appeal provides additional evidence that context may shift appraisals of disgust stimuli. Alternatively, finding similar disgust and empathy levels regardless of context would suggest disgust experiences have limited flexibility. If contextual expectations and persuasive arguments shift appraisals and alter emotional experiences, both disgust and distress reactions should be diminished. Conversely, I predict feelings of empathy will be induced to a greater degree in the persuasive context than by the disgust-imagery alone. H3a: Participants exposed to the disgust-eliciting images without the persuasive context will experience greater levels of disgust. H3b: Participants exposed to the disgust-eliciting images without the persuasive context will also report greater personal distress. 83 H3c: Participants exposed to the disgust-eliciting images in the persuasive context will experience greater empathy than participants exposed to disgust imagery without the persuasive context. Predicting Prosocial Behavior Whereas disgust reactions are expected to reduce prosocial behavior, both empathy and distress should increase prosocial behavior. Considering these different outcomes is also important because research suggests self-focused helping may reduce the amount of aid provided compared to empathy-driven charitable behavior (Dickert, Sagara, & Slovic, 2011). Thus, persuasive messages that succeed in raising empathetic responses may be more effective than those that merely induce personal distress. Individuals may make an initial giving decision to alleviate feelings of distress if other options for alleviating distress are unavailable or deemed ineffective. This would be the case when escape from the helping situation, whether physical or psychological, is not easy (Stocks, Lishner, & Decker, 2009). Physical escape involves terminating one’s exposure to the other’s need; psychological escape refers to the extent one expects to forget or continue to be aware of the other’s need even after exposure to the helping situation has ended. In contrast, persons experiencing empathetic concern are not expected to be affected by ease or difficulty of escape—only actions that alleviate the other’s need are viable responses (Stocks et al., 2009). Thus, other-focused respondents are not only more likely to help, they also tend to provide more help. Additionally, participants experiencing greater distress should have a stronger motivation to maximize personal payoff thereby reducing gift amount (Camerer & Fehr, 2004). Due 84 to these differing motivations, I expect that self-focused responses (i.e. distress) versus other-focused emotions (i.e. empathy) will predict differences in giving. Predictions regarding giving behavior also take in to account the effect of initial disgust reactions. H4: Participants exposed to the disgust-eliciting images without persuasive context will be less likely to make a gift than participants exposed to the persuasive appeal. H5a: Disgust responses will be negatively associated with donation behavior (whether participants make a gift of any amount or not). H5b: Distress and empathy will positively predict donation behavior mediating the relationship between exposure and giving. H5c: Distress will negatively predict donation amount among participants who make a gift, reducing gift amount. H5d: Empathy experienced will positively predict donation amount among participants who make a gift, increasing gift amount. An irony of disgust is that it has been described as stifling motivation to consider the emotional event further, while at the same time producing involuntary recall of the disgust elicitor (Joffe, 2008). Thus, psychological escape may be more difficult when exposed to disgust-inducing stimuli. As such participants should report greater expectation to continue thinking about the issue. As discussed, greater difficulty in escaping the situation psychological should influence giving in relations to distress responses but should have no influence in regard to emotional reactions. H6a: Participants exposed to disgust stimuli will report greater difficulty of psychological escape. 85 H6b: Psychological escape will moderate the relationship between personal distress and donation behavior. Increasing difficulty of escape should increase likelihood of giving. Method To test these hypotheses, responses to disgust-stimuli with and without a persuasive context are compared to similar emotional stimuli, as well as neutral stimuli. This results in a three (emotional imagery: disgust present/disgust removed/neutral) by (context: persuasive/non-persuasive) two factorial design. Table 5.1 Experimental Conditions Imagery Disgust present Disgust removed Neutral Context Non-persuasive The following content was taken from a medical database that provides information for tracking and diagnosing tropical diseases. Images of children Images of children Maps of regions with disease with disease with highest depictions depictions prevalence of removed tropical diseases Persuasive The following content was taken from a charity campaign that raises money for treating and preventing tropical diseases. Images of children Images of children Maps of regions with disease with disease with highest depictions depictions prevalence of removed tropical diseases Accompanied by Accompanied by Accompanied by text text text The persuasive context manipulation was achieved via a message prompt before the emotional imagery manipulation as well as via persuasive arguments accompanying the image stimuli. Participants in the persuasive context treatment 86 condition were told the content they were about to see was taken from a campaign that raises money for treating and preventing tropical diseases. These participants saw images along with text that constituted the persuasive appeal for help. Participants in the non-persuasive context treatment were told the content they were about to see was taken from a medical database that provides information for tracking and diagnosing tropical diseases. These participants only saw images without the accompanying text (see Table 5.1). Next, participants in each context condition were randomly assigned to an emotional imagery condition. Participants saw one of three sets of images, each set contained five images. Images in the disgust-present condition depicted children with signs or symptoms of various tropical diseases (see Figure 5.1 left column). Besides cropping the images to achieve uniform dimensions across all conditions, images in the disgust present condition were not otherwise altered. Given that these images are expected to elicit disgust reactions, this condition will be referred to as the disgust condition hereafter. Images in the disgust-removed condition were the same images used in the disgust condition except these images were altered to remove their respective disgust elicitors (see Figure 5.1 middle column). Images were manipulated with the Pixelmator Pro image editing software by a paid graphic designer. Hereafter the disgust-removed condition will be referred to as the no disgust condition for simplicity. Participants in the neutral condition saw grayscale maps of five regions in the world where tropical diseases are most prevalent (see Figure 5.1 right column). 87 Figure 5.1 Sample Stimuli without (top) and with text (bottom); disgust-elicitor present (left), disgust elicitor removed (middle) neutral map images (right). Thus, participants were randomly assigned to one of six conditions: disgust, no disgust, or neutral images appearing with and without the persuasive appeal. Except for the differences described, participants saw a consistent layout across conditions. Cornell University’s IRB approved the study design and procedure. Appendix 5A provides a chart of all the stimulus materials. Image selection and pretesting. A wide search was conducted to find images for the disgust condition using both the lay and scientific names of various tropical diseases as the search terms (i.e. roundworm or Ascariasis). More than 60 potential images of people with various tropical diseases were collected. A review of these images took several factors into consideration including the age, gender, and ethnicity of the persons portrayed, the number of individuals shown, the part of the body affected, and technical qualities of the image such as orientation and composition. One 88 goal of the selection process was to keep consistent certain factors known to influence charitable behavior such as the depiction of faces, children and single individuals (Jenni & Loewenstein, 1997). Thus, images that only depicted body parts or would only depict body parts after cropping were excluded, and only images showing a single child were used. Gender and ethnicity were varied to minimize potential effects due to these factors. Based on these criteria, a subset of seven images were selected for pretesting prior to finalizing the study design. The edited versions of four of these images were also included in the pilot test to ensure that removing the disgust elicitor significantly reduced the level of disgust evoked in comparison to the original images. The five map images and the text element were also included in the pretest to ensure that they too evoked significantly lower levels of disgust. Pretest participants (N = 30) were recruited using the Prolific Academic online survey recruitment platform in accordance with the eligibility criteria for the main study (see Participants section). All pretest participants saw one randomly selected map image, to provide a baseline level of emotion, before they were randomized to see either all seven images for the disgust condition or the text and all four images for the no disgust condition. Participants rated each image, and the text if they saw it, separately before viewing the next stimulus. The order of the map and text were fixed, at the beginning of the survey, while the disgust and no disgust images appeared in random order. The items used to measure disgust were the same as those used in the main study (see Measures section). Independent samples t-tests compared ratings of the 89 disgust and no disgust images; paired samples t-tests compared the disgust image ratings to those of the map images. The mean disgust ratings for the disgust images (M = 2.81) was significantly higher than the ratings for the no disgust images (M = 1.50; t(28) = 3.20, p = .003). Mean disgust ratings were also significantly higher for the disgust images compared to the maps (M = 1.22, t(14) = 5.47, p < .001) and the text (M = 1.57, t(28) = 2.98, p = .01). Mean disgust ratings for the maps, text and no disgust images were not significantly different (p’s > .14). The individual mean ratings for each disgust image were than rank ordered (range: 2.39-3.56), and the image with the lowest mean was removed from the stimulus set. To maintain variation in the ethnicity of children depicted, the image with the next lowest rating was retained in the stimulus set, and the image ranked just above it was removed. The mean ratings for these two images were not significantly different (t(14) = .80; p = .44). The five selected disgust images were edited to create the full set of images for the no disgust condition. All images were formatted and combined with the text to create the final stimuli for the study. Participants. Participants for the main study (N = 364), also recruited via Prolific Academic2, comprise a sample of adults over the age of 25 not currently enrolled in a full-time undergraduate degree program. Use of a non-student sample is an important consideration in charitable giving research because student samples are 2 Originally, Qualtrics was contracted to recruit participants. The company reneged citing concerns that the content would take participants “down a dark path.” Their decision is illustrative of some of the objections to the strategic use of disgust in persuasion. Chapter 6 provides a discussion of the ethical concerns as well as the potential benefits of conducting this research. 90 often more homogenous in terms of age, level of education, and income—all of which have significant associations with giving behavior (Bekkers, 2010). Table 5.2 Participant Demographics N’s (valid %) Age 25 – 34 189 (51.9%) 35 – 44 104 (28.6%) 45 – 54 39 (10.7%) 55 – 64 21 (5.8%) 65 and older 11 (3%) Sex Male 206 (56.6%) Female 152 (41.8%) Hispanic 31 (8.5%) Race African American/Black 19 (5.2%) Asian/Asian American 25 (6.9%) Caucasian/White 321 (88.2%) Other 9 (2.5%) Relevance Previously aware 186 (51.1%) Perceived susceptibility 24 (6.6%) Past experience 47 (12.9%) Education Highschool diploma 29 (8%) Some college 61 (16.8%) Vocational or Associate degree 53 (14.6%) Bachelor’s degree 135 (37.1%) Graduate degree 83 (22.8%) Employed Full Time 225 (62.2%) Household Income Less than $20,000 42 (12.3%) $20,000 – 34,999 60 (17.6%) $35,000 – 49,999 59 (17.3%) $50,000 – 74,999 67 (19.6%) $75,000 – 99,999 45 (13.2%) $100,000 or more 68 (19.9%) 91 Participants were predominantly White (88.2%), Asian (6.9%), or Black (5.2%); the majority were between the ages of 25 and 34 (51.9%) with a substantial minority between 35 and 44 (28.6%); 56.6% were male. There was a diverse range of educational attainment among participants: 24.8% reported a high school diploma as their highest level of education completed, 37.1% had completed a bachelor’s degree, and 22.8% achieved a graduate degree including professional and doctorate degrees. Participants were mainly employed full time (61.8%) and their household incomes ranged from less than $20,000 per year (12.3%) to more than $100,000 per year (19.9%). A majority (51.1%) reported being aware of NTDs before participating in the study; 25.3% indicated that they or someone they knew had previously had an NTD, and 12.9% reported feeling personally at risk of contracting an NTD in the past year. Table 5.2 provides full demographic details of the study sample. Participants received $1.85 for taking the study, in addition to a $5 incentive that was used to measure their behavioral decision. Procedure. The experiment was administered using the Qualtrics online survey tool via a link embedded in Prolific Academics’ study invitation. Informed consent was administered electronically. Participants were told they would view, and provide their reactions to, content about a health issue. After giving consent, all participants were told they had, ostensibly, been selected to complete an extended version of the study and would be compensated an additional $5 for the added time. This step was included to prompt participants expectations that thy were earning the incentive and minimize the potential that they would give simply because they received money unexpectedly (windfall giving; Cherry, Frykblom, & Shogren, 2002). 92 Participants were randomly assigned to view one of the six stimulus conditions. The five images in each condition were not randomized nor rotated across text but appeared in the same order in each respective condition. Viewing time was not restricted—participants determined how long they would look at each image. Following exposure to the stimuli, participants completed the post-test measures including the behavioral measure (donation task) described below. Finally, participants were debriefed on the purpose of the study and given the option to change their responses to the donation task. Measures Disgust. The level of disgust participants experienced in response to each of the treatment messages was measured using the English-Language version of the Ekel- State-Fragebogen state disgust scale (Bates & Chadwick, 2015; Ihme & Mitte, 2009). The scale includes 13 items, one of which was not used in the current study to reduce redundancy. Items included measures of both the subjective (i.e. I felt repulsed, I was grossed out) and physiological (i.e. My stomach churned, I felt a gag reflex) components of disgust. Participants indicated the extent to which they experienced each item on 5-point scales ranging from not at all to a great deal. Items were summed and averaged to create a single disgust index (α = .96; M = 2.34, SD = 1.23). Certainty. Certainty appraisals were measured with three items. First, participants were prompted to give a one-word response describing their reaction to what they saw. Participants then rated their level of certainty that about this reaction and level of certainty that it was an appropriate response to the material. A third item asked “how certain were you of what you wanted to do in response to what you saw” 93 with response options ranging from 1, not at all certain, to 5, very certain (adapted from Tiedens & Linton, 2001). Items were summed and averaged to compute a certainty scale (α = .67 M = 4.18, SD = .77). Social Judgment. Two semantic differential scales measured participants evaluation of 12 target groups, including charity aid recipients, in terms of the competency and warmth dimensions from the Stereotype Content Model (Fiske, Cuddy, Glick, & Xu, 2002). Participants indicated how competent and warm they perceived each target group on 5-point scales ranging from extremely incompetent to extremely competent and extremely cold to extremely warm. The competent and warmth blocks appeared in random order with target groups randomized within each block. The product of the mean ratings for charity recipients on the competency (M = 3.21, SD = .86) and warmth (M = 3.40, SD = 1.01; r(362) = .34, p ≤ .01) dimensions was calculated to create a single stereotype rating for this group (M = 11.20, SD = 5.33). Distress. Six items measured participants distress reactions to each treatment message (Fultz et al., 1988). Participants indicated how upset, worried, disturbed, distressed, troubled, and grieved they felt while viewing the content on 7-point scales ranging from not at all to extremely. Items were summed and averaged to create a single distress index (α = .96; M = 4.39, SD = 1.95). Empathetic Concern. Six items measured empathetic concern (Baton, 1991). Participants indicated how sympathetic, softhearted, compassionate, moved, tender, and warm they felt while viewing the content on 7-point scales ranging from not at all 94 to extremely. Items were summed and averaged to create a single empathetic concern index (α = .95; M = 4.65, SD = 1.72). Psychological escape. Perceptions of difficulty of escape were operationalized in terms of memorability of the content and expectations that participants would continue thinking about it for specific time periods beyond exposure. Four items asked participants how memorable the content was in comparison to other information they had encountered that day (7-point scale: not at all memorable to very memorable), how much they anticipated thinking about it over the next 15 minutes and the next 60 minutes (9-point scale: 0/not at all to 9/very much) and the number of times they anticipated thinking about it over the next two days (6 items, see Appendix 5B; adapted from Stocks et al., 2009). Items were summed and averaged to create a psychological escape index (α = .85 M = 5.18 SD = 1.68) Giving behavior. Giving behavior was measured in two-parts (Dickert et al., 2011). Participants were asked if they would like to donate some of their earnings from this study to The END Fund to provide help for persons with neglected tropical diseases. First, participants responded by answering yes (N = 150, 41.2%) or no to this item. Those who answered yes were then asked to indicate how much they would like to donate, up to a maximum of $5 (M = 2.26, SD = 1.46). This amount was deducted from the additional incentive paid to participants. Thus, there are two measures of giving: donation behavior, whether the participant made any gift or not, and the donation amount for each participant who gave. Except for the giving behavior items, item order was randomized within each respective block. Additionally, the disgust, distress, and empathy blocks appeared in 95 randomized order. Table 5.3 shows the bivariate correlations between each of the dependent variables. Table 5.3 Bivariate Correlations of Dependent Variables 1 2 3 4 5 6 7 8 1. Disgust Response - 2. Certainty .25** - 3. Social Judgment -.04 .15** - 4. Distress .71** .29** .08 - 5. Empathy .36** .27** .20** .66** - 6. Psychological Escape .50** .25** .10 .60** .58** - 7. Donation Behavior .19** .11* .12* .33** .27** .24** - 8. Donation Amount .06 -.02 .08 .17* .17* .03 -- - Covariates. Gender was included as a covariate based on established between- sex differences in sensitivity to disgust (Rohrmann, Hopp, & Quirin, 2008). Personal relevance was measured with two items asking participants if they, or anyone they knew, had previously had a tropical disease or if they have felt personally at risk of contracting a tropical disease in the past year. Participants were counted as having personal relevance for the issue if they answered yes to one or both of these questions (N = 56, 15.4%). Results Data were analyzed using IBM SPSS v. 25. A series of cross-tab analyses confirmed that participant demographic variables generally did not vary significantly by condition. Therefore, no demographic variables other than the covariates identified will be considered in the regression models. 96 Table 5.4 Descriptive Statistics by Condition with 95% Confidence Intervals Disgust Certainty Social Distress N Judgment Non-Persuasive Neutral 1.39 3.76 11.39 2.07 56 [1.16, 1.6] [3.52, 4.01] [10.07, 12.71] [1.64, 2.49] No Disgust 2.08 3.89 11.26 4.61 62 [1.84, 2.32] [3.66, 4.11] [9.96, 12.56] [4.21, 5.01] Disgust 3.39 4.46 12.67 5.67 63 [3.10, 3.67] [4.29, 4.63] [11.06, 14.28] [5.28, 6.07] Persuasive Neutral 1.90 4.25 9.95 4.09 61 [1.66, 2.14] [4.08, 4.43] [8.84, 11.06] [3.63, 4.54] No Disgust 2.02 4.26 10.61 4.21 61 [1.77, 2.27] [4.12, 4.40] [9.30, 11.91] [3.80, 4.63] Disgust 3.17 4.39 11.30 5.44 61 [2.89, 3.44] [4.23, 4.56] [9.85, 12.73] [5.08, 5.80] Empathy Psychological Donation Donation N Escape Behavior Amount Non-Persuasive Neutral 2.99 3.86 1.85 10 (17.9%) 56 [2.50, 3.49] [3.35, 4.40] [.55, 3.15] No Disgust 5.44 5.15 2.35 26 (41.9%) 62 [5.10, 5.79] [4.76, 5.54] [1.82, 2.88] Disgust 4.94 6.10 2.26 28 (44.4%) 63 [4.50, 5.38] [5.74, 6.45] [1.63, 2.89] Persuasive Neutral 4.52 4.97 2.22 25 (41%) 61 [4.13, 4.90] [4.57, 5.37] [1.67, 2.77] No Disgust 4.70 5.07 2.00 28 (45.9%) 61 [4.36, 5.04] [4.69, 5.45] [1.56, 2.44] Disgust 5.13 5.76 2.56 33 (54.1%) 61 [4.76, 5.50] [5.42, 6.11] [1.97, 3.15] Note: Donation Behavior is a categorical variable indicating whether a participant gave (or not) N’s and valid percent reported. The main effects of each treatment condition on all dependent variables were first examined in comparison to the neutral non-persuasive condition (map image unaccompanied by text). Table 5.4 provides the descriptive statistics by condition for each dependent variable. Multiple linear regression models using ordinary least 97 squares (OLS) estimation were run with dummy variables for each of the six conditions excluding the neutral non-persuasive comparison group. Based on these analyses, the disgust manipulation appears effective as both the persuasive disgust and non-persuasive disgust conditions significantly increased reported levels of disgust. Table 5.5 Main Effects Compared to Non-persuasive, Neutral Condition Social Disgust Certainty Distress Judgment Non-Persuasive No Disgust 0.75 (0.18)*** 0.13 (0.14) -0.28 (0.98) 2.53 (0.29)*** Disgust 2.03 (0.18)*** 0.70 (0.14)*** 1.12 (0.97) 3.57 (0.29)*** Persuasive Neutral 0.55 (0.18)** 0.49 (0.14)*** -1.47 (0.98) 2.04 (0.29)*** No Disgust 0.68 (0.19)*** 0.50 (0.14)*** -0.74 (0.98) 2.2 (0.29)*** Disgust 1.81 (0.18)*** 0.63 (0.14)*** -0.06 (0.98) 3.41 (0.29)*** Covariates Gender 0.22 (0.11)* 0.04 (0.08) -1.35 (0.56)* -0.38 (0.17)* Relevance 0.36 (0.15)** <-0.01 (0.11) 0.99 (0.77) 0.61 (0.23)** Constant 1.18 (0.15)*** 3.74 (0.11)*** 12.05 (0.80)*** 2.18 (0.24)*** Adj. R sq 0.34 0.09 0.03 0.36 N 364 364 364 364 Psychological Donation Donation Empathy Escape Behavior Amount Non-Persuasive No Disgust 2.42 (0.28)*** 1.34 (0.28)*** 1.29 (0.44)** 0.09 (0.52) Disgust 1.9 (0.28)*** 2.25 (0.28)*** 1.36 (0.44)** 0.21 (0.51) Persuasive Neutral 1.53 (0.28)*** 1.15 (0.28)*** 1.25 (0.44)** 0.19 (0.52) No Disgust 1.75 (0.28)*** 1.3 (0.28)*** 1.5 (0.44)*** -0.08 (0.51) Disgust 2.16 (0.28)*** 1.96 (0.28)*** 1.81 (0.44)*** 0.41 (0.5) Covariates Gender -0.4 (0.16)* 0.05 (0.16) <0.01 (0.22) -1.06 (0.23)*** Relevance 0.53 (0.22)* 0.87 (0.22)*** 0.84 (0.31)** 0.39 (0.28) Constant 3.14 (0.23)*** 3.66 (0.23)*** -1.75 (0.39)*** 2.62 (0.47)*** Adj. R sq 0.21 0.19 -- 0.11 N 364 364 364 150 ***p ≤ .001, *p ≤ .05, **p ≤ .01, #p ≤ .10 98 Results show the addition of emotional imagery, with and without disgust, produced significant differences on all other dependent variables except for the social judgment task and the donation amount (see Table 5.5). Given the significant effects of the emotional imagery, with and without disgust, over the neutral imagery on the majority of the outcome variables, hypotheses testing will focus solely on examining differences in effects between the disgust and no disgust conditions. The presence of a disgust elicitor is the only element that varies between these conditions. Therefore, excluding the neutral conditions will isolate the effect of the disgust manipulation. Analyses consider main and interaction effects of the persuasive context and disgust stimuli in comparison to the non-persuasive no disgust condition (see Table 5.6). Multiple linear regression models using ordinary least squares (OLS) estimation were run with a dummy variable for the persuasive context conditions and one for the disgust conditions. An interaction term was computed for the disgust persuasive condition leaving the no disgust, non-persuasive condition as the comparison group. As predicted, the presence of a disgust elicitor significantly increased certainty appraisals; whereas, the interaction of disgust and the persuasive context produced a significant decrease in participants certainty about how they reacted or thought they should react to the message. Therefore, H1a and H1b are supported. However, results of the social judgment task show there are no significant main or interaction effects of the disgust or the persuasive context manipulations on participants’ evaluations of warmth and competence of charity aid recipients. H2a and H2b are not supported. 99 Table 5.6 Main and Interaction Effects of Persuasive Context and Disgust Imagery Disgust Certainty Social Distress Judgment Persuasive -0.07 (0.19) 0.38 (0.13)** -0.43 (1.00) -0.35 (0.28) Disgust 1.3 (0.19)*** 0.57 (0.12)*** 1.35 (0.99) 1.05 (0.27)*** Interaction -0.16 (0.27) -0.44 (0.18)** -0.7 (1.41) 0.17 (0.39) Covariates Gender 0.19 (0.13) 0.05 (0.09) -1.23 (0.71)# -0.37 (0.2)# Relevance 0.12 (0.19) 0.11 (0.13) 2.06 (1.02)* 0.38 (0.28) Constant 1.97 (0.15)*** 3.85 (0.1)*** 11.55 4.73 (0.22)*** (0.79)*** Adj. R sq 0.26 0.08 0.03 13 N 247 247 247 247 Empathy Psychological Donation Donation Escape Behavior Amount Persuasive -0.67 (0.26)** -0.02 (0.26) 0.23 (0.37) -0.14 (0.37) Disgust -0.52 (0.26)* 0.91 (0.25)*** 0.07 (0.37) 0.13 (0.37) Interaction 0.94 (0.37) ** -0.25 (0.36) 0.24 (0.52) 0.33 (0.51) Covariates Gender -0.46 (0.19)* -0.08 (0.18) -0.2 (0.26) -1.13 (0.26)*** Relevance 0.58 (0.27)* 0.91 (0.26)*** 0.96 (0.39)** 0.52 (0.32) Constant 5.58 (0.21)*** 5.06 (0.2)*** -0.37 (0.29) 2.7 (0.29) Adj. R sq 0.05 11 -- 14 N 247 247 247 115 ***p ≤ .001, *p ≤ .05, **p ≤ .01, #p ≤ .10 Treatment conditions with the disgust elicitor significantly increased disgust reactions, but the interaction of the persuasive context manipulation with disgust image produced disgust reactions similar to the no disgust, non-persuasive condition supporting H3a. Similarly, in support of H3b, the presence of disgust elicitors significantly increased personal distress; the interaction of the persuasive context manipulation with disgust image did not produced significantly different levels of distress compared to the disgust image alone. Both the disgust manipulation (as 100 predicted) and the persuasive context (for which there was no a priori hypothesis) significantly decreased empathetic concern relative to the comparison condition. However, the interaction of the disgust manipulation with the persuasive context appears to mitigate these effects. Empathetic responses are higher when disgust imagery is used in a persuasive context than when disgust imagery appears alone or when the persuasive appeal is accompanied by imagery that does not include disgust elicitors (see Figure 5.2). Thus, H3c is supported. 6 5 4 3 2 1 0 Non Persuasive Persuasive Context Condition No Disgust Disgust Figure 5.2 Main and interaction effects of disgust exposure and persuasive appeal on empathy reported. H4 predicted lower levels of donation behavior among participants exposed to the disgust-eliciting images without persuasive context. This hypothesis was not 101 Empathy supported as there were no significant differences in donation behavior between the experimental conditions. To test hypotheses H5a through d, as well as H6b, I ran two regression models with disgust, distress and empathy as the predictor variables (see Table 5.7). To test the potential moderating role of psychological escape on personal distress, both the psychological escape variable and an interaction term for psychological escape and personal distress were also added to the model. Results do not support the hypotheses. Neither disgust nor empathetic concern influenced donation behavior, or the amount donated. Perceived difficulty of psychological escape was greatest in the disgust conditions (supporting H6a), however, when this variable and its interaction with personal distress are added to the model, the effect of personal distress on donation behavior is only marginally significant and the interaction of distress and perceived psychological escape does not predict giving. Table 5.7 Indirect Effects on Behavior Donation Donation Behavior Amount Disgust -0.16 (0.16) -0.02 (0.13) Distress 0.59 (0.33)# 0.24 (0.50) Empathy 0.11 (0.12) 0.13 (0.12) Psych. Escape 0.25 (0.31) -0.12 (0.51) Escape*Distress -0.05 (0.06) <0.01 (0.09) Gender 0.05 (0.28) -0.96 (0.26)*** Relevance 0.80 (0.40)* 0.45 (0.33) Constant -3.33 (1.60)* 1.43 (2.69) Adj. R sq -- .18 N 247 115 ***p ≤ .001, *p ≤ .05, **p ≤ .01, #p ≤ .10 102 Discussion The emotion of disgust is characterized as eliciting an unfettered rejection response. Core disgust elicitors in particular have been described as triggering fast and largely unthinking responses (Rubenking & Lang, 2014). According to appraisal theories, these low-level cognitive responses to disgust stimuli can be explained in terms of the appraisal of certainty—the feeling that one knows how to respond to an emotional event—which limits the need for deeper processing. Although discrete emotions are associated with specific appraisals, appraisal patterns are not fixed— consistent with the view that emotional experiences are dynamic (Moors et al., 2013). Thus, someone may experience an emotion even as a specific appraisal connected to that emotion shifts. The ability to shift certainty appraisals has been demonstrated for sadness, and the current study provides initial evidence that the certainty appraisal associated with disgust may also be malleable. Given the modest reliability of the certainty measure, this result should be interpreted with caution, however. In their initial appraisal theory explication, Smith & Ellsworth (1985) note that the certainty dimension is difficult to measure. This may explain why studies have manipulated or inferred certainty instead. In the rare case where a subjective measure has been used, it served to confirm the certainty manipulation. In contrast, the present study examined certainty as an outcome of an emotion eliciting event. Understanding how certainty is implicated is important because appraisals in response to an emotional event will vary in actuality. In the context of persuasion, shifts in certainty have implications for message processing including how participants evaluate portrayed message targets. 103 Researchers and practitioners alike have warned against the use of disgust elicitors in prosocial appeals citing concerns that such imagery diminishes our concern for others and reinforces demeaning stereotypes of aid recipients (Buckels & Trapnell, 2013; Wheatley & Haidt, 2005). As noted, experimental evidence supporting this dehumanizing view of disgust is based on the use of incidental disgust primes rather than disgust-elicitation that is relevant to the decision context. The current study examined the possibility that disgust elicitation as part of a prosocial message may improve social judgments and increase empathetic concern for others. According to the Stereotype Content Model (SCM), dehumanization rests on an unwillingness to ascribe competency (agency) and warmth (friendliness) to socially distant others (Fiske et al., 2002). The warmth and competency dimensions are also the basis of stereotypes that guide our judgments of certain social groups. Charity aid recipients are often seen as having moderate (to high) warmth and low competence (Kennedy & Hill, 2010). Emotional certainty, particularly certainty due to disgust, should reinforce, or even exacerbate, this stereotype, but increasing uncertainty should lead to more thoughtful evaluations of the target group. By this logic, participants exposed to the persuasive disgust condition should have exhibited less reliance on the stereotype in their judgments. In contrast, average ratings for charity aid recipients were moderately competent and moderately warm across all study conditions. Perhaps the pervasiveness of a stereotype resists the influence of cognitive and emotional factors. However, the results show more ambivalent evaluations rather than the expected paternalistic view of charity recipients. 104 It is also possible that social judgments are more influenced by implicit emotion primes. Although the context participants viewed the images in was manipulated, participants in the non-persuasive condition may have connected the social judgment task to the images they saw. In other words, the portrayal of people, rather than exposure to a scene or scent for instance, may have made the social judgment task explicit for all participants. This explanation does not account for similar ratings in the neutral non-persuasive condition, however, which did not show humans or provide context. I assumed that relying on stereotypes in decision making involved similar processes as making a stereotype judgment itself. The results of this study may simply be an indicator that they do not. Thus, although results show that the context of disgust exposure can influence certainty appraisals, this study does not provide evidence that it also influences, or improves, social judgments. Those who caution against disgust not only worry that it may reinforce negative stereotypes, they also express concern that eliciting disgust diminishes the capacity to feel empathy for others. This is a valid concern given the evidence linking empathetic concern to prosocial behavior; reducing empathy would reasonably be expected to reduce aid. Indeed, participants reported significantly lower levels of empathy in the non-persuasive disgust condition than the non-persuasive, no disgust condition. Interestingly, this pattern is reversed in the persuasive conditions. The persuasive context appears to reduce empathy when paired with images that do not contain disgust elicitors, but empathy levels increase significantly when the disgust eliciting images are encountered in the prosocial persuasive context (Figure 5.2). 105 These shifts in empathetic concern may seem particularly promising when one also considers that giving was highest in the persuasive disgust condition—in terms of the number of donors and the average gift amount. But between condition differences in giving were not significant in the two-by-two design (excluding the neutral conditions). Further, increased empathy did not predict changes in giving as would be expected given the body of evidence in support of the Empathy-Altruism Hypothesis (Batson et al., 2015). A post-hoc analysis showed that participants in the persuasive disgust condition reported similar levels of empathy and distress. Distress appeared to be a significant predictor of giving before controlling for participants’ expectation that they would continue to recall the message for several minutes, hours or days after completing the study (psychological escape). It is not clear why neither empathy nor distress significantly predicted increased giving or why participants experiencing greater empathy would not also give significantly more. Although research participants are often treated as experiencing either distress or empathy, in everyday experiences these helping motives may cooccur. Disgust elicitation appears to be one context that triggers both responses. The remaining question is whether empathy and distress complement or oppose (neutralize) each other (Stocks et al., 2009). This study offers limited guidance for the design of prosocial campaigns. Message planners may choose to use disgust imagery to illustrate the severe realities of an issues. Their intuitions that disgust can increase empathy, perhaps more than other emotional imagery, when used to promote a social cause may be well founded. While the cooccurrence of disgust and empathy is theoretically interesting, the use of disgust does not appear to benefit behavioral outcomes significantly compared to other 106 emotional imagery. For those who argue that potentially shocking visuals should only be used when their benefits (increasing support) outweigh potential consequences (offending people’s sensibilities, reinforcing demeaning stereotypes), eliciting disgust may not be ethically justifiable (Brown & Whiting, 2014). Alternatively, disgust eliciting appeals do not appear to worsen social judgments nor reduce support. Campaigns often utilize multiple messages that fulfill different objectives. The psychological escape measure showed people expected to think about the persuasive disgust imagery more than the other stimuli. If disgust eliciting messages can effectively shift people’s appraisals in response to depictions of human suffering, increase empathy for affected individuals and get people to think more about the issue, these outcomes may be effective for achieving longer-term results. Limitations This study contained several methodological improvements on some of the limitations to the second study in this dissertation including tighter experimental control via the disgust-removed condition, measures of empathetic concern and donation behavior, and use of a non-student sample. There are also several limitations to the design and opportunities for further research. The study’s main goal was to examine emotion appraisal and downstream effects when disgust is elicited as an integral, context relevant emotion. The persuasive disgust condition represented integral disgust elicitation. However, its comparison condition, the non-persuasive disgust condition, may not be a strong manipulation of incidental disgust given the portrayal of people. Although the lack of a prosocial decision context affected both certainty and empathy measures compared 107 to the persuasive disgust condition, future research should directly compare effects of the types of images used in this study with the more typical incidental (non-human) disgust primes. This may be particularly useful in clarifying the results of the social judgment task in this study as well as the general discussion in the literature concerning when and how disgust exerts influence on social cognition. Developing more reliable measures of uncertainty is another opportunity for future research. The certainty construct is complex incorporating evaluations of the situation, appropriate reactions to the situation, and predictions about future outcomes. One or all of these components may be of interest to the context of a given study. Participants may have reported higher levels of empathy, and lower levels of disgust, because the images portrayed children. Although this doesn’t account for the between condition differences for these variables, future research should consider a variety of contexts including depictions of adults, animals and groups known to elicit low levels of empathy. Finally, future research should consider the influence of disgust with new audiences, different issues and alternative prosocial behaviors. Summary Understanding the factors that motivate, or diminish, prosocial behavior has both practical and theoretical implications. Vivid images that concretely depict need and evoke negative emotion are often an integral part of prosocial appeals (Small & Verrochi, 2009). Images that elicit disgust however are often cautioned against as disgust is characterized as a low prosocial emotion. The current study provides mixed evidence in response to these concerns. It shows that disgust-eliciting prosocial messages may shift certainty appraisals, increase empathy and allow for more 108 prolonged consideration of a social issue. While such messages do not appear to affect social judgment either negatively or positively, they also do not appear to enhance helping behavior significantly over other emotional stimuli. Future research should address additional contexts of disgust elicitation including variations in the portrayed target groups, audiences and behavioral outcomes. 109 CHAPTER 6 CONCLUSION This series of studies applied functional and appraisal theories of emotion, dual-processing models of persuasion, and moral judgment theories to investigate the interaction of disgust-evoking imagery and persuasive arguments on decision making and subsequent action in the context of health and prosocial behavior. These studies addressed questions of attention, the influence of discrete emotions, cognitive processes, and behavioral responses. The current chapter begins with a summary of findings for each of these aspects across the three studies, continues with a discussion of the theoretical contributions and practical implications of this research including limitations common to the three studies, and concludes by outlining future research to address limitations, answer outstanding questions from the present group of studies, and explore new lines of research. Attention for Disgust-Eliciting Messages Functional and appraisal theories of emotion associate the emotion of disgust with attention withdrawal. In contrast, research using measures of attention allocation, such as cardiac deceleration, have repeatedly found increased attention for disgust- eliciting persuasive messages (Leshner, Bolls, & Thomas, 2009; Leshner, Bolls, & Wise, 2011; Leshner, Vultee, Bolls, & Moore, 2010). But few studies utilize visual attention measures to examine the influence of disgust on attentional processes. The first two studies of this dissertation consider visual attention for disgust-eliciting messages among adolescents exposed to cigarette graphic warning labels (GWLs) and 110 a college student sample exposed to a charitable giving appeal. Analyses considered effects on overall message attention as well as attention for the image versus text components of the message. Both studies provide evidence of reduced attention to the message overall. Study 1 showed significant decreases in attention compared to other emotional stimuli, both in terms of the time it took participants to orient to warning labels with disgust depictions and their total fixation time on these warnings. Study 2 did not show any differences between both the low and high disgust conditions and the neutral control condition. Given that emotion-evoking stimuli are generally considered more attention getting than neutral stimuli (Weinberg & Hajcak, 2010), finding that the disgust-evoking imagery was no more attention getting than the neutral imagery suggests disgust does not hold visual attention as well as other emotional stimuli. Thus, the results of studies 1 and 2 are consistent with propositions of functional and appraisal theories of emotion. Different patterns of attention emerge between study 1 and study 2 when considering attention for the image versus text components of the message. Although overall attention to the message was reduced in study 1, participants paid significantly more attention to the image over the text when looking at warning labels that contained disgust content. This shift is indicative of an attention bias (Charash & McKay, 2002; Van Hooff, Devue, Vieweg, & Theeuwes, 2013) for disgust stimuli despite overall reduced attention for the disgust-eliciting warnings as a whole. These findings are also consistent with the notion that disgust simultaneously attracts and repels attention (Joffe, 2008). 111 However, the context of exposure may influence attention for the disgust- eliciting imagery itself compared to other components of the message. In study 2, attention toward the image also increased in the disgust conditions compared to the neutral control, but the shift is not as dramatic. Attention to the text remained higher than attention to the image in all disgust treatment groups, suggesting limited willingness to attend to the disgust stimuli. Whereas in study 1 disgust eliciting images may also convey threat, the images for study 2 were intended to convey the severity of need of another person. This difference in relevance may account for evidence of an attention bias in study 1, whereas study 2 does not provide such evidence. Notably these findings differ from other persuasion studies that find overall increased attentional allocation to messages that contain disgust content. Yet, the current results may not be inconsistent with previous research given the differences in attentional measures employed (visual attention versus cardiac deceleration). Eye movements indicate what elements of a social situation people select for further processing (Orquin & Mueller Loose, 2013). However, people may continue to process a message component even after visual exposure has ended; disgust imagery may linger in the mind even as one is unwilling to visually attend to it (Hajcak & Olvet, 2008; Joffe, 2008). Future research should directly consider differences in visual attention versus other attentional processes (i.e. encoding and storage) and the time course of these different effects in response to disgust-eliciting content. Multiple Emotions in Persuasion Both exposure to treatment messages and level of attention paid to specific parts of the message (study 2 and 3) were predictive of participants emotional 112 responses across studies. Emotion measures included items to gauge disgust as well as other reactions including general negative affect, sadness, guilt, distress and empathy. In study 1, only attention to the disgust image predicted increased negative affect in response to warning labels, and this effect was only significant for participants who lived with a smoker. Arousing negative affect is an important function of anti-tobacco messages as exposure alone often does not influence risk perceptions or behavioral intentions (Peters, Evans, Hemmerich, & Berman, 2012). This finding provides empirical evidence that disgust imagery may be an effective strategy for anti-tobacco messaging targeting adolescents (Pechmann & Reibling, 2006; Rozin & Singh, 1999), filling a gap as previous studies have focused on adult reactions to disgust-eliciting anti-tobacco messages. Adolescents who live with a smoker are at an increased risk of smoking initiation; the potential for disgust eliciting images to singularly influence negative responses among this group has significant practical implications. Study 2 found that charitable appeals that utilized disgust visuals elicit more than disgust responses. Participants also responded with sadness and guilt that often equaled or exceeded the level of disgust participants felt. This study considered the relative influence of each of these emotions on subsequent thoughts and behavioral intentions. The results of these hypothesis tests support the discrete emotions perspective as disgust, sadness and guilt were differentially associated with both cognitions and emotions (discussed in the sections below). Whereas study 2 did not consider the potential for a disgust-eliciting charitable appeal to evoke positive emotions, study 3 measured empathetic concern in addition to disgust and distress responses. Relative to a neutral image in a non-persuasive context, all treatment 113 conditions increased empathetic responses. However, compared to emotional conditions that did not include disgust-elicitors, the non-persuasive disgust condition significantly reduced empathetic concern while the persuasive disgust condition significantly increased empathetic concern. Although disgust and empathy are considered opposing emotions in terms of their ability to generate moral consideration or concern for the welfare of others, these results suggest that in the context of prosocial persuasion the two emotions can be activated simultaneously. This finding has implications for behaviors people enact in response to disgust-eliciting appeals compared to responses to incidental or irrelevant disgust primes. Disgust Elicitation and Cognitive Processes Cognitive responses to treatment stimuli were examined across the three studies in a variety of ways that included measuring beliefs about smoking, a thought listing task, cognitive appraisal, and a social judgment task. It is important to note that emotion and cognition are not viewed as independent or separate processes. Cognitive appraisal theories suggest our evaluations of an emotional event shape the specific feelings we have in response. These evaluations are often rapid, occurring outside our consciousness, and tend to manifest consistently for specific emotional triggers (Smith & Ellsworth, 1985). However, appraisal patterns, such as the degree of certainty about how to respond, may shift based on the context of emotion elicitation. This was illustrated in study 3. Although disgust is typically associated with certainty appraisals, participants who encountered disgust-eliciting content in the context of a prosocial persuasive appeal expressed increased uncertainty about how to respond. 114 These participants also reported lower levels of disgust experienced than participants who viewed the same images without the prosocial persuasive context. Likewise, emotions are not solely feeling states that drive our behavior in relatively unthinking ways; there is considerable evidence that emotions can influence how we think—whether we rely on message heuristics or choose to process information systematically (Nabi, 2002). Information processing in response to core disgust stimuli is thought to be quick and fairly limited (Rubenking & Lang, 2014; Tiedens & Linton, 2001). Consistent with this perspective, the results of study 2 show that feelings of disgust significantly reduced cognitive elaboration in response to the charitable appeal. In contrast, sadness responses predicted an increase in the number of thoughts (elaboration) participants expressed in response to the message (enhanced information processing). Sadness is characterized as promoting systematic processing in an effort to reduce uncertainty about an emotional event’s cause or consequence and how one should respond to it. Despite the evidence from these studies that exposure to disgust content and feelings of disgust can influence how we process information, findings pertaining to the explicit judgments participants made as a result did not align with stated hypotheses. In study 1, adolescents risk beliefs were neither negatively nor positively associated with participants attention to disgust imagery. Similarly, elaboration did not influence behavioral intentions in study 2 and, in study 3, participants did not differ across conditions in their evaluations of various target groups. There may be a variety of reasons for these results including already strong beliefs about the risks of smoking, or persistent and seemingly ambivalent stereotypes of aid recipients. Other forms of 115 judgment, including implicit cognitions, may be more influenced by the disgust and persuasion context and should be considered in future research. Behavioral Responses to Disgust Appeals Studies 2 and 3 considered the relationship between exposure to disgust- eliciting messages, emotional reactions to those messages, and behavioral intention (study 2) and behavior (study 3). Based on accounts of disgust as a low prosocial emotion, it was hypothesized that increased feelings of disgust would be negatively associated with each behavioral measure. Instead, there was only a marginally significant, negative relationship between levels of disgust experienced and intentions to make a donation to the organization identified in the appeal in study 2. However, there was a significant positive association between guilt experienced and behavioral intentions as predicted. Study 3 expanded on this methodology to include a giving task as a direct measure of behavior. More than 40% of participants donated some amount and the average gift was approximately 45% of their additional study earnings. But the treatment groups were only different from the control, and not each other, on these measures. There was also no relationship between feelings of disgust and giving behavior, and distress was only a marginally significant predictor once perceived psychological escape was accounted. These findings raise more questions than answers regarding the relationship between disgust-eliciting content, feelings of disgust and behavior. Participants should be motivated to increase payouts to themselves (Camerer & Fehr, 2004), and, from the perspective of moral emotion theory, disgust should reduce motivation to help others 116 resulting in even greater payouts to self (Liljenquist, Zhong, & Galinsky, 2010). However, both the rate of giving and average gift in study 3 exceed typical payouts (approximately 34% and 20% respectively) for dictator games (Camerer, 2003). Contrary to study hypotheses, empathetic concern and distress, two widely studied motivators of helping behavior, do not appear to account for the results of the giving task either. Both studies demonstrate that people can respond with helping when exposed to disgust content in a prosocial persuasive context. While feelings of disgust do not diminish such behavior, the mechanisms that allow for and increase giving require further investigation. Theoretical and Empirical Contributions Lazarus (1991) characterized disgust as a primitive and simple emotion, restricted in content and rigid in elicitation. More recent discussions classify disgust as inflexible (Russell & Giner-Sorolla, 2013), while others question whether disgust reactions are homogenous regardless of how the emotion is elicited (Simpson, Carter, Anthony, & Overton, 2006). The current studies provide evidence that disgust can be sensitive to contextual assessments (Clark & Powell, 2017). These studies extended the Social Intuitionist Model of moral judgment (Haidt, 2001) to prosocial behavior suggesting that competing emotional goals, such as those aroused by prosocial, disgust-eliciting message, increase uncertainty about how to respond and elicit prosocial emotions allowing for positive responses. In addition to confirming these hypotheses, the current research also affirms the utility of the discrete emotion perspective for explaining cognitive and behavioral reactions to persuasive communication. Results show attentional processes differ for 117 various types of emotional stimuli and that distinct emotions influence persuasive outcomes in different ways that are not explained by arousal or valence alone. From a methodological perspective, these studies diversified the sample of participants. Studies on disgust and anti-smoking messages often recruit adult smokers and moral judgment and prosocial persuasion studies often rely on student samples or predominantly European non-student samples limiting knowledge about the persuasive influence of disgust for other social groups. The design of study 3 also represents a methodological innovation—the digital removal of disgust from persuasive appeals allowed for strict isolation of the effect of disgust elicitors rather than comparing disgust images to stimuli that differed in content and composition. Finally, these studies measure emotional and cognitive responses to content participants might experience in their daily lives rather than presenting contrived hypothetical situations or manipulating their emotional and cognitive states. As such, results are usefully for understanding how audiences are most likely to respond to persuasive messages that employ disgust elicitors relevant to the decision context and behavioral responses such messages seek to motivate. Practical and Ethical Implications The current study provides some guidance regarding the strategic benefits and potential consequences of using disgust to achieve desirable ends such as promoting health and prosocial behavior. Campaigns often use strong emotional appeals to attract attention over the myriad of other stimuli, including other persuasive messages, that compete for attention. In terms of visual attention, however, message designers should be mindful that audiences may be less willing to look at messages that contain disgust 118 elicitors. In natural viewing environments this may mean that audiences discontinue exposure to the message altogether. On the other hand, visual attention appears to be distinct from other indicators of processing such that audiences may be more likely to remember the issue despite reduced viewing time. Further, the results of the current group of studies suggest that despite reducing visual attention, disgust elicitors, in a persuasive context, effectively increase additional emotional reactions including empathetic concern. Given the limited amount of attention audiences may devote to messages regardless of emotional tone, this tradeoff between reduced attention to disgust-eliciting messages, and memorability and emotional responses may be an advantage of employing disgust in persuasive appeals. Although disgust stimuli evoke the feeling of disgust, the additional emotions that accompany disgust in response to a persuasive appeal may temper the potentially negative consequences of disgust feelings allowing for positive outcomes such as increased cognitive processing. The potential for emotional indifference or ambivalence due to the mixed emotions audiences experience in response to disgust- eliciting appeals (i.e. both distress and empathy) maybe a challenge for designing these messages, however. Message planners should also be aware that disgust eliciting strategies maybe more or less effective with certain audiences and consider the role of personal relevance in segmenting potential audiences. This discussion of practical implications is not oblivious to the ethical tensions regarding disgust and persuasion. Although the current group of studies do not suggest that disgust uniquely reinforces demeaning stereotypes, as has been suggested by previous research, there are still questions about the audience’s right not to see 119 graphically explicit messages and the right of affected groups not to be portrayed in ways that elicit revulsion. These concerns apply as much to research as they do to media campaigns. The intent of the current research was not to violate these ethical principles, but rather to examine intuitive assumptions about the consequences of disgust for different issues and audiences. For instance, the results of study 2 and 3 show that persuasive disgust appeals can generate sadness and empathy, and do not inhibit helping behavior. These findings, which contradict previously untested claims that disgust is not suitable for achieving these goals (Herz, 2012; Sherman & Haidt, 2011), could not have been made clear without exposing participants to disgust- eliciting depictions of people in need and represent one benefit of this research. Likewise, disgust-eliciting content shouldn’t be used in persuasive messages for shock value alone. Consideration should be given to whether the benefits outweigh the consequences (Brown & Whiting, 2014). Results of the current studies show that disgust imagery are effective at increasing negative emotions toward smoking when adolescents live with a smoker. From an ethical perspective, then, the use of disgust imagery with other young audiences may not be justifiable. Similarly, if the goal of a prosocial appeal is simply to raise more money, the present study suggests other forms of emotional images may do this just as well. However, if the goal is to generate prolonged salience of the issue, the benefits of disgust in the long term likely outweigh its short-term drawbacks. Further, images like those used in this research portray the realities of social issues. While using them risks offending people’s senses (Angyal, 1941), avoiding such graphic representations may risk the symbolic annihilation of certain issues from social consciousness allowing atrocities to persist. 120 Limitations A few overarching limitations are relevant to all three studies in this dissertation. Given the practical implications of this research, there were tradeoffs and tensions between ecological validity and tight experimental control. For example, study 1 utilized actual GWLs proposed by the FDA and were thus not systematical matched in terms of limiting differences in stimuli composition between the treatment and comparison groups. While the disgust-eliciting treatment group was selected according to theoretical definitions of disgust, the emotional nature of the comparison group was more diffuse. As such, other factors beyond the presence of disgust elicitors may account for differences in outcomes. In contrast, study 3 employed a clean experimental manipulation using the exact images across the two-by-two treatment conditions that only differed based on the presence of disgust elicitors. Although this represented a methodological improvement over studies 1 and 2, it did not allow for a direct comparison of incidental disgust primes, which typically do not depict people, and the kinds of images that are relevant to charitable appeals. Beyond the measurement issues previously discussed for each study, there were differences between studies in how disgust responses were measured that may affect the reported relationships between disgust and other outcomes. For instance, study 1 did not include measures of physiological disgust reactions, and more measurement items were included in study 3 over study 2. Although no comparisons between studies are referenced in this regard, readers should bear these differences in mind as they evaluate the collective findings. The one comparison that can be made regards the level of disgust participants report experiencing—across studies only 121 moderate levels of disgust are reported despite qualitative differences in disgust depictions. Emotion inductions are typically lower in response to images as opposed to video stimuli, both of which evoke milder reactions than coming into contact with actual, or ostensibly real, disgust objects (Grühn & Sharifian, 2016). Thus, the use of pictorial stimuli is another limitation of these studies. Eye tracking and behavioral measures employed in these studies help to overcome the biases associated with self-report measures. Yet, even these objective measures are explicit as were the stimuli and other measures in each study. Consciously presented information may still have influence without people being aware. As an example, differences in social judgment (study 3) may have been found if measured with a more implicit social cognition task. Although the social judgment task and other measures are similar to those commonly used in research on persuasion or judgment and decision-making, they cannot account for residual effects of disgust that operate below thresholds of consciousness. All three studies expose participants to messages at a single point in time and measured reactions immediately following exposure. Therefore, results do not generalize to long-term effects of disgust as a persuasive strategy, nor do they account for the influence of multiple exposures or the influence of other messages or social interactions on message outcomes. Future Research Future research plans can be conceptualized according to three lines of inquiry that consider: i) the relative effectiveness of disgust compared to other emotional strategies including longitudinal effects, ii) issues related to the social groups that are 122 portrayed in different types of disgust-evoking messages, and iii) advances in methodologies used to examine emotion inducing communication processes. Few persuasion studies compare disgust stimuli to stimuli evoking emotions other than fear or anger. The comparison conditions in the current studies, while more diverse, are not explicitly sad, guilt or empathy inducing. Further these stimuli are predominantly negative in tone. Future research will consider the relative effectiveness of disgust-eliciting messages compared to specific types of emotional stimuli including images intended to induce feelings of hope and happiness. The long-term effects of disgust are also important for this line of research to address. Results showed that people expected to continue thinking about the disgust content to a greater extent than other emotional stimuli. A follow up study to this dissertation will examine how responses to disgust-evoking stimuli change overtime and whether disgust is an effective means of prolonging the perceived saliency of need compared to other emotion inducing message strategies. Potential unintended consequences have been identified for prosocial appeals employing guilt and sadness. One outstanding question from the current research is whether inducing these emotions via a message with a predominant disgust theme, rather than an explicit guilt or sadness appeal, circumvents or exacerbates these consequences. The finding that empathy was reduced when non-disgust images were presented in a persuasive context is particularly interesting especially since the reverse occurred for the disgust eliciting images. Prior research has suggested that deliberation reduces empathetic concern (Loewenstein & Donoghue, 2015), but this unintended effect of a persuasive appeal may only apply to certain types of emotional experiences. 123 Future research will have to explore this hypothesis as it is difficult to separate the contextual expectations manipulated in the current study from the rational processes they may have triggered. In designing the prosocial persuasion studies, it was difficult to find quality images that portrayed a diversity of people suffering from NTDs despite their prevalence across the world. This led me to question the relationship between ethnicity and content themes in non-profit imagery. For instance, are persons of Asian ethnicity more likely to be portrayed as sufferers or survivors in NTD aid materials? What are the cultural, organizational and political factors that influence these portrayals? How do these differences contribute to the consequences often ascribed to shocking or disgust-eliciting content? Similar questions regarding the types of people portrayed in persuasive messages are also relevant to the health persuasion context. The GWLs in study 1 did not picture adolescents, although for the purpose of the study adolescents were the target audience. One goal of future research should be to understand how messages tailored for youth audiences shift attentional patterns and cognitive reactions among adolescent participants. The current studies treat the attention-emotion relationship as unidirectional, rather than a dynamic process with feedback loops. Attention is measured using single, averaged items rather than trend data overtime. However, appraisal theories position emotion experience as a dynamic process such that exposure and attention influence emotion which in turn influence attention and so on. Using dynamic content and dynamic measures is helpful for examining attention and emotional flow in 124 response to messages (Nabi & Green, 2015). One promising measurement strategy involves syncing visual and biophysical indicators of emotion and tracking these data overtime to determine where they coincide or diverge for different emotional stimuli. Shifts in uncertainty appraisals found in study 3, while interesting, require more investigation including developing more reliable measures of uncertainty that account for the diverse evaluations that comprise the construct and determining which components of uncertainty are most relevant to various persuasive appeals. Finally, future studies should continue to examine potential mechanisms that facilitate positive behavioral responses to disgust-eliciting appeals. Summary Results of three studies show that disgust-eliciting content can have positive effects on cognitions and behaviors despite reducing attention for persuasive content. These effects appear to be the result of the additional emotions evoked by such appeals rather than the disgust responses induced. Feelings of disgust negatively predicted elaboration but were generally unrelated to other persuasive outcomes. Disgust elicitation may be particularly effective for certain audiences while having significant unintended effects among other groups. The prosocial persuasive context appears to alter emotion appraisals, but further research is needed to understand the implications of these cognitive shifts and the specific mechanisms that explain behavioral responses to disgust-eliciting appeals. 125 REFERENCES Chapter 1 Angyal, A. (1941). Disgust and related aversions. Journal of Abnormal and Social Psychology, 36(3), 393–412. Clark, J. A., & Powell, P. A. (2017). Defending disgust: Why disgust is morally beneficial. In C. Price & E. Walle (Eds.), Emotion Researcher, ISRE’s Sourcebook for Research on Emotion and Affect. Retrieved from http://emotionresearcher.com/why-disgust-is-morally-beneficial/ Clayton, R. B., Leshner, G., Bolls, P. D., & Thorson, E. (2016). Discard the smoking cues—Keep the disgust: An investigation of tobacco smokers’ motivated processing of anti-tobacco commercials. Health Communication, 1–12. http://doi.org/10.1080/10410236.2016.1220042 Clayton, R. B., Leshner, G., Tomko, R. L., Trull, T. J., & Piasecki, T. M. (2017). Countering craving with disgust images: Examining nicotine withdrawn smokers’ motivated message processing of anti-tobacco public service announcements. Journal of Health Communication, 22(3), 254–261. http://doi.org/10.1080/10810730.2016.1268222 Cochran, J., Kydd, R., Lee, J. M., Walker, N., & Consedine, N. S. (2017). Disgust but not health anxiety graphic warning labels reduce motivated attention in smokers: A study of P300 and late positive potential responses. Nicotine & Tobacco Research. Giner-Sorolla, R., & Harris, L. T. (2017). The negative side of disgust. In C. Price & E. Walle (Eds.), Emotion Researcher, ISRE’s Sourcebook for Research on 126 Emotion and Affect. Retrieved from http://emotionresearcher.com/the-negative- side-of-disgust/ Halkjelsvik, T., & Rise, J. (2015). Disgust in fear appeal anti-smoking advertisements: The effects on attitudes and abstinence motivation. Drugs: Education, Prevention & Policy, 22(4), 362–369. http://doi.org/10.3109/09687637.2015.1015491 Harris, L. T., & Fiske, S. T. (2011). Dehumanized perception: A psychological means to facilitate atrocities, torture, and genocide? Zeitschrift Für Psychologie, 219(3), 175–181. http://doi.org/https://doi.org/10.1027/2151-2604/a000065 Jónsdóttir, H. L., Holm, J. E., Poltavski, D., & Vogeltanz-Holm, N. (2014). The role of fear and disgust in predicting the effectiveness of television advertisements that graphically depict the health harms of smoking. Preventing Chronic Disease, 11, E218. http://doi.org/10.5888/pcd11.140326 Leshner, G., Bolls, P. D., & Thomas, E. (2009). Scare’ em or disgust ’em: The effects of graphic health promotion messages. Health Communication, 24(5), 447–458. http://doi.org/10.1080/10410230903023493 Leshner, G., Bolls, P. D., & Wise, K. (2011). Motivated processing of fear appeal and disgust images in televised anti-tobacco ads. Journal of Media Psychology, 23(2), 77–89. http://doi.org/10.1027/1864-1105/a000037 Leshner, G., Clayton, R. B., Bolls, P. D., Bhandari, M., Leshner, G., Clayton, R. B., … Clayton, R. B. (2017). Deceived, disgusted, and defensive: Motivated processing of anti-tobacco advertisements. Health Communication, 1–10. http://doi.org/10.1080/10410236.2017.1350908 Leshner, G., Vultee, F., Bolls, P. D., & Moore, J. (2010). When a fear appeal isn’t just 127 a fear appeal: The effects of graphic anti-tobacco messages. Journal of Broadcasting & Electronic Media, 54(3), 485–507. http://doi.org/10.1080/08838151.2010.498850 Lupton, D. (2014). The pedagogy of disgust: The ethical, moral and political implications of using disgust in public health campaigns. Critical Public Health, (February), 1–14. http://doi.org/10.1080/09581596.2014.885115 Morales, A. C., Wu, E. C., & Fitzsimons, G. J. (2012). How disgust enhances the effectiveness of fear appeals. Journal of Marketing Research, 49(3), 383–393. http://doi.org/10.1509/jmr.07.0364 Nabi, R. L. (1998). The effect of disgust-eliciting visuals on attitudes toward animal experimentation. Communication Quarterly, 46(4), 472–484. http://doi.org/10.1080/01463379809370116 O’Keefe, D. J. (2003). Message properties, mediating states, and manipulation checks: Claims, evidence, and data analysis in experimental persuasive message effects research. Communication Theory, 13(3), 251–274. Pechmann, C., & Reibling, E. T. (2006). Antismoking advertisements for youths: An independent evaluation of health, counter-industry, and industry approaches. American Journal of Public Health, 96(5), 906–913. http://doi.org/10.2105/AJPH.2004.057273 Porzig-Drummond, R., Stevenson, R., Case, T., & Oaten, M. (2009). Can the emotion of disgust be harnessed to promote hand hygiene? Experimental and field-based tests. Social Science & Medicine, 68, 1006–1012. http://doi.org/10.1016/j.socscimed.2009.01.013 128 Reynolds, L. M., McCambridge, S. A., Bissett, I. P., & Consedine, N. S. (2014). Trait and state disgust: An experimental investigation of disgust and avoidance in colorectal cancer decision scenarios. Health Psychol, 33(12), 1495–1506. http://doi.org/10.1037/hea0000023 Stanton, G. (1998). The seven stages of genocide. In Genocide Program Seminar, Yale Center for International and Area Studies (Vol. 12). New Haven, CT. Strohminger, N. (2014). Disgust talked about. In Philosophy Compass (Vol. 9, pp. 478–493). http://doi.org/10.1111/phc3.12137 Taylor, K. (2007). Disgust is a factor in extreme prejudice. British Journal of Social Psychology, 46(3), 597–617. http://doi.org/10.1348/014466606X156546 Teper, R., Zhong, C., & Inzlicht, M. (2015). How emotions shape moral behavior: Some answers (and questions) for the field of moral psychology. Social and Personality Psychology Compass, 9(1), 1–14. Tugrul, T. O. (2013). The impacts of fear and disgust on the perceived effectiveness of smoking warning labels: A study on Turkish university students. Asia-Pacific Journal of Public Health, 20(10), 1–7. http://doi.org/10.1177/1010539513475647 Chapter 2 Abitan, A., & Krauth-Gruber, S. (2015). The two sides of disgust: A lexical and thematic content analysis of narratives of personally experienced physical and moral disgust. Social Science Information, 54(4), 470–496. http://doi.org/10.1177/0539018415597316 Alonso Dos Santos, M., Lobos, C., Muñoz, N., Romero, D., & Sanhueza, R. (2017). 129 The influence of image valence on the attention paid to charity advertising. Journal of Nonprofit & Public Sector Marketing, 29(3), 346–363. http://doi.org/10.1080/10495142.2017.1326355 Armstrong, T., McClenahan, L., Kittle, J., & Olatunji, B. O. (2014). Don’t look now! Oculomotor avoidance as a conditioned disgust response. Emotion, 14(1), 95– 104. http://doi.org/10.1037/a0034558 Batson, C. D., Lishner, D. A., & Stocks, E. L. (2015). The empathy–altruism hypothesis. In The Oxford Handbook of Prosocial Behavior (p. 259). http://doi.org/10.1093/oxfordhb/9780195399813.013.023 Bebko, C., Sciulli, L. M., & Bhagat, P. (2014). Using eye tracking to assess the impact of advertising appeals on donor behavior. Journal of Nonprofit & Public Sector Marketing, 26(4), 354–371. http://doi.org/10.1080/10495142.2014.965073 Bodenhausen, G. V, Sheppard, L. A., & Kramer, G. P. (1994). Negative affect and social judgment: The differential impact of anger and sadness. European Journal of Social Psychology, 24, 45–62. http://doi.org/10.1002/ejsp.2420240104 Bol, N., Boerman, S. C., Bergstrom, J. C. R., & Kruikemeier, S. (2016). An overview of how eye tracking is used in communication research. In Universal Access in Human-Computer Interaction (pp. 421–429). Switzerland: Springer International Publishing. http://doi.org/10.1007/978-3-319-40250-5 Bolls, P. D. (2010). Understanding emotion from a superordinate dimensional perspective: A productive way forward for communication processes and effects studies. Communication Monographs, 77(2), 146–152. http://doi.org/10.1080/03637751003790477 130 Chaiken, S. (1980). Heuristic versus systematic information processing and the use of source versus message cues in persuasion. Journal of Personality and Social Psychology, 39(5), 752–766. http://doi.org/10.1037/0022-3514.39.5.752 Charash, M., & McKay, D. (2002). Attention bias for disgust. Journal of Anxiety Disorders, 16(5), 529–541. http://doi.org/10.1016/S0887-6185(02)00171-8 Clayton, R. B., Leshner, G., Tomko, R. L., Trull, T. J., & Piasecki, T. M. (2017). Countering craving with disgust images: Examining nicotine withdrawn smokers’ motivated message processing of anti-tobacco public service announcements. Journal of Health Communication, 22(3), 254–261. http://doi.org/10.1080/10810730.2016.1268222 Cochran, J., Kydd, R., Lee, J. M., Walker, N., & Consedine, N. S. (2017). Disgust but not health anxiety graphic warning labels reduce motivated attention in smokers: A study of P300 and late positive potential responses. Nicotine & Tobacco Research. Cockrill, A., & Parsonage, I. (2016). Shocking people into action: Does it still work? An empirical analysis of emotional appeals in charity advertising. Journal of Advertising, 56(4), 401–413. http://doi.org/10.2501/JAR-2016-045 Collymore, N. N., & McDermott, M. R. (2015). Evaluating the effects of six alcohol- related message frames on emotions and intentions: The neglected role of disgust. Journal of Health Psychology. http://doi.org/10.1177/1359105314567910 Coulter, R. H., Cotte, J., & Moore, M. L. (1999). Believe it or not: Persuasion, manipulation and credibility of guilt appeals. Advances in Consumer Research, 26, 288–294. Retrieved from 131 http://www.acrwebsite.org/volumes/display.asp?id=8264 Dahl, D. W., Frankenberger, K. D., & Manchanda, R. V. (2003). Does it pay to shock? Reactions to shocking and nonshocking advertising content among university students. Journal of Advertising Research, 43(3), 268–280. http://doi.org/10.1017/S0021849903030332 Dillard, J. P., & Nabi, R. L. (2006). The persuasive influence of emotion in cancer prevention and detection messages. Journal of Communication, 56(s1), S123– S139. http://doi.org/10.1111/j.1460-2466.2006.00286.x Dillard, J. P., & Peck, E. (2000). Affect and persuasion: Emotional responses to public service announcements. Communication Research, 27(4), 461–495. http://doi.org/10.1177/009365000027004003 Dillard, J. P., Plotnick, C. A., Godbold, L. C., Freimuth, V. S., & Edgar, T. (1996). The multiple affective outcomes of AIDS PSAs: Fear appeals do more than scare people. Communication Research, 23(1), 44–72. Dillard, J. P., & Shen, L. (2018). Threat appeals as multi-emotion messages: An argument structure model of fear and disgust. Human Communication Research, 44, 103–126. http://doi.org/10.1093/hcr/hqx002 Duchowski, A. T. (2002). A breadth-first survey of eye-tracking applications. Behavior Research Methods, Instruments, & Computers, 34(4), 455–470. http://doi.org/10.3758/BF03195475 Ellsworth, P. C., & Smith, C. A. (1988). From appraisal to emotion: Differences among unpleasant feelings. Motivation & Emotion, 12(3), 271–302. Frijda, N. (1986). The emotions: Studies in emotion and social interaction. Paris: 132 Maison de Sciences de l’Homme. Gable, P., & Harmon-Jones, E. (2010). The motivational dimensional model of affect: Implications for breadth of attention, memory, and cognitive categorisation. Cognition and Emotion, 24(2), 322–337. http://doi.org/10.1080/02699930903378305 Haidt, J. (2003). The moral emotions. In Handbook of Affective Sciences (11th ed., pp. 852–570). Haidt, J., McCauley, C. R., & Rozin, P. (1994). Individual differences in sensitivity to disgust: A scale sampling seven domains of disgust elicitors. Personality and Individual Differences, 16, 701–713. http://doi.org/10.1016/0191- 8869(94)90212-7 Hajcak, G., & Olvet, D. M. (2008). The persistence of attention to emotion: Brain potentials during and after picture presentation. Emotion, 8(2), 250–255. http://doi.org/10.1037/1528-3542.8.2.250 Herz, R. S. (2012). That’s disgusting: Unraveling the mysteries of repulsion (First). New York, NY: W. W. Norton & Company, Inc. Inbar, Y., & Pizarro, D. (2009). Grime and punishment: How disgust influences moral, social, and legal judgments. The Jury Expert, 21(March), 12–24. Izard, C. E. (1977). Human emotions. New York, NY: Springer Science & Business Media. Judah, G., Aunger, R., Schmidt, W. P., Michie, S., Granger, S., & Curtis, V. (2009). Experimental pretesting of hand-washing interventions in a natural setting. American Journal of Public Health, 99, 405–411. 133 http://doi.org/10.2105/AJPH.2009.164160 Kelly, D., & Morar, N. (2014). Against the yuck factor: On the ideal role of disgust in society. Utilitas, 26(2), 153–177. http://doi.org/10.1017/S0953820813000290 Landy, J. F., & Goodwin, G. P. (2015). Does incidental disgust amplify moral judgment? A meta-analytic review of experimental evidence. Perspectives on Psychological Science, 10(4), 518–536. http://doi.org/10.1177/1745691615583128 Lang, A. (2000). The limited capacity model of mediated message processing. Journal of Communication, 50(1), 46–70. Lang, A., Park, B., Sanders-Jackson, A. N., Wilson, B. D., & Wang, Z. (2007). Cognition and emotion in TV message processing: How valence, arousing content, structural complexity, and information density affect the availability of cognitive resources. Media Psychology, 10(3), 317–338. http://doi.org/10.1080/15213260701532880 Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1997). Motivated attention: Affect, activation, and action. In Attention and orienting: Sensory and motivational processes. (pp. 97–135). Lazarus, R. S. (1991). Emotion and Adaptation. Oxford University Press. Lerner, J. S., & Keltner, D. (2000). Beyond valence: Toward a model of emotion- specific influences on judgement and choice. Cognition & Emotion, 14(4), 473– 493. Lerner, J. S., Li, Y., Valdesolo, P., & Kassam, K. S. (2015). Emotion and decision making. Annual Review of Psychology, 66, 799–823. 134 http://doi.org/10.1146/annurev-psych-010213-115043 Leshner, G., Bolls, P. D., & Thomas, E. (2009). Scare’ em or disgust ’em: The effects of graphic health promotion messages. Health Communication, 24(5), 447–458. http://doi.org/10.1080/10410230903023493 Leshner, G., Bolls, P. D., & Wise, K. (2011). Motivated processing of fear appeal and disgust images in televised anti-tobacco ads. Journal of Media Psychology, 23(2), 77–89. http://doi.org/10.1027/1864-1105/a000037 Leshner, G., Clayton, R. B., Bhandari, M., & Bolls, P. D. (2013). The impact of anger and disgusting images in anti-tobacco ads on viewers’ message processing. Psychophysiology, 50, S45–S45. Leshner, G., Vultee, F., Bolls, P. D., & Moore, J. (2010). When a fear appeal isn’t just a fear appeal: The effects of graphic anti-tobacco messages. Journal of Broadcasting & Electronic Media, 54(3), 485–507. http://doi.org/10.1080/08838151.2010.498850 Loewenstein, G., & Donoghue, T. O. (2015). Modeling the interplay between affect and deliberation. Decision, 2(2), 55–81. Loewenstein, G., & Lerner, J. S. (2003). The role of affect in decision making. In Handbook of affective science (3rd ed., pp. 619–642). Miller, A., & Leshner, G. (2007). How viewers process live, breaking, and emotional television news. Media Psychology, 10, 23–40. Morales, A. C., Wu, E. C., & Fitzsimons, G. J. (2012). How disgust enhances the effectiveness of fear appeals. Journal of Marketing Research, 49(3), 383–393. http://doi.org/10.1509/jmr.07.0364 135 Nabi, R. L. (1998). The effect of disgust-eliciting visuals on attitudes toward animal experimentation. Communication Quarterly, 46(4), 472–484. http://doi.org/10.1080/01463379809370116 Nabi, R. L. (1999). A cognitive-functional model for the effects of discrete negative emotions on information processing, attitude change, and recall. Communication Theory, 9(3), 292–320. http://doi.org/10.1111/j.1468-2885.1999.tb00172.x Nabi, R. L. (2002a). Discrete emotions and persuasion. In J. P. Dillard & M. Pfau (Eds.), The Persuasion Handbook: Developments in Theory and Practice (pp. 289–308). Nabi, R. L. (2002b). The theoretical versus the lay meaning of disgust: Implications for emotion research. Cognition & Emotion, 16(5), 695–703. http://doi.org/10.1080/02699930143000437 Nabi, R. L. (2003). “Feeling” resistance: Exploring the role of emotionally evocative visuals in inducing inoculation. Media Psychology, 5(2), 199–223. http://doi.org/10.1207/S1532785XMEP0502_4 Newhagen, J. E. (1998). TV news images that induce anger, fear, and disgust: Effects on approach‐avoidance and memory. Journal of Broadcasting & Electronic Media, 42(2), 265–276. http://doi.org/10.1080/08838159809364448 Petty, R., & Cacioppo, J. (1986). The elaboration likelihood model of persuasion. In Advances in Experimental Social Psychology (pp. 123–205). http://doi.org/10.1016/S0065-2601(08)60214-2 Petty, R. E., & Cacioppo, J. T. (1977). Forewarning, cognitive responding, and resistance to persuasion. Journal of Personality and Social Psychology. 136 http://doi.org/10.1037/0022-3514.35.9.645 Petty, R. E., Cacioppo, J. T., & Schumann, D. (1983). Central and peripheral routes to advertising effectiveness: The moderating role of involvement. Journal of Consumer Research, 10(2), 135–146. Pieters, R., & Wedel, M. (2004). Attention capture and transfer in advertising: Brand, pictorial, and text-size effects. Journal of Marketing, 68(April), 36–50. Pizarro, D. A., Detweiler-Bedell, B., & Bloom, P. (2006). The creativity of everyday moral reasoning. In J. C. Kaufman & J. Baer (Eds.), Creativity and Reason in Cognitive Development (pp. 81–98). New York, NY: Cambridge University Press. Roseman, I., Wiest, C., & Swartz, T. (1994). Phenomenology, behaviors, and goals differentiate discrete emotions. Journal of Personality and Social Psychology, 67(2), 206–221. http://doi.org/10.1037/0022-3514.67.2.206 Rozin, P., Haidt, J., & McCauley, C. R. (1993). Disgust. In M. Lewis & M. Haviland (Eds.), Handbook of Emotions. Rozin, P., Haidt, J., & McCauley, C. R. (2008). Disgust. In Handbook of Emotions (pp. 757–776). Rozin, P., Lowery, L., & Haidt, J. (1999). The CAD Triad Hypothesis: A mapping between three moral emotions (contempt, anger, disgust) and three moral codes (community, autonomy, divinity). Journal of Personality and Social Psychology, 76(4), 574–586. Rubenking, B., & Lang, A. (2014). Captivated and grossed out: An examination of processing core and sociomoral disgusts in entertainment media. Journal of 137 Communication, 64(3), 543–565. http://doi.org/10.1111/jcom.12094 Scherer, K. R. (1999). Appraisal theory. In Handbook of cognition and emotion (pp. 637–663). Schwarz, N. (2000). Emotion, cognition, and decision making. Cognition and Emotion, 14(4), 433–440. Shah, A. K., & Oppenheimer, D. M. (2008). Heuristics made easy: An effort-reduction framework. Psychological Bulletin, 134(2), 207–222. http://doi.org/10.1037/0033-2909.134.2.207 Sherman, G. D., & Haidt, J. (2011). Cuteness and disgust: The humanizing and dehumanizing effects of emotion. In Emotion Review (Vol. 3, pp. 245–251). http://doi.org/10.1177/1754073911402396 Shimp, T. A., & Stuart, E. W. (2004). The role of disgust as an emotional mediator of advertising effects. Journal of Advertising, 33(1), 43–53. Retrieved from http://www.jstor.org/stable/4189245 Slovic, P., Finucane, M. L., Peters, E., & MacGregor, D. G. (2007). The affect heuristic. European Journal of Operational Research, 177(3), 1333–1352. Smith, C. A., & Ellsworth, P. C. (1985). Patterns of cognitive appraisal in emotion. Journal of Personality and Social Psychology, 48(4), 813–838. Stevenson, R. J., Oaten, M. J., Case, T. I., Repacholi, B. M., & Wagland, P. (2010). Children’s response to adult disgust elicitors: Development and acquisition. Developmental Psychology, 46(1), 165–177. http://doi.org/10.1037/a0016692 Strohminger, N. (2014). Disgust talked about. In Philosophy Compass (Vol. 9, pp. 478–493). http://doi.org/10.1111/phc3.12137 138 Tiedens, L. Z., & Linton, S. (2001). Judgment under emotional certainty and uncertainty: The effects of specific emotions on information processing. Journal of Personality and Social Psychology, 81(6), 973–988. Tugrul, T. O. (2013). The impacts of fear and disgust on the perceived effectiveness of smoking warning labels: A study on Turkish university students. Asia-Pacific Journal of Public Health, 20(10), 1–7. http://doi.org/10.1177/1010539513475647 Wedel, M., & Pieters, R. (2006). Eye tracking for visual marketing. Foundations and Trends® in Marketing, 1(4), 231–320. http://doi.org/10.1561/1700000011 Wheatley, T., & Haidt, J. (2005). Hypnotic disgust makes moral judgments more severe. Psychological Science, 16(10), 780–784. Wisneski, D. C., & Skitka, L. J. (2017). Moralization through moral shock: Exploring emotional antecedents to moral conviction. Personality and Social Psychology Bulletin, 43(2), 139–150. http://doi.org/10.1177/0146167216676479 Yoo, J. H., Kreuter, M. W., Lai, C., & Fu, Q. (2014). Understanding narrative effects: The role of discrete negative emotions on message processing and attitudes among low-income African American women. Health Communication, 29(5), 494–504. http://doi.org/10.1080/10410236.2013.776001 Zeelenberg, M., Nelissen, R. M. A., Breugelmans, S. M., & Pieters, R. (2008). On emotion specificity in decision making: Why feeling is for doing. In Judgment and Decision Making (Vol. 3, pp. 18–27). Retrieved from https://search.proquest.com/docview/1011218372?accountid=10267 Zeelenberg, R., Wagenmakers, E.-J., & Rotteveel, M. (2006). The impact of emotion on perception: Bias or enhanced processing? Psychological Science, 17(4), 287– 139 291. http://doi.org/10.1111/j.1467-9280.2006.01700.x Chapter 3 Bradley, M. M., Codispoti, M., Cuthbert, B. N., & Lang, P. J. (2001). Emotion and motivation I: Defensive and appetitive reactions in picture processing. Emotion, 1(3), 276–298. http://doi.org/10.1037//1528-3542.1.3.276 Brown, S. L., & Richardson, M. (2012). The effect of distressing imagery on attention to and persuasiveness of an antialcohol message: A gaze-tracking approach. Health Education & Behavior, 39(1), 8–17. http://doi.org/10.1177/1090198111404411 Brown, S., & Locker, E. (2009). Defensive responses to an emotive anti-alcohol message. Psychology and Health, 24(5), 517–528. http://doi.org/10.1080/08870440801911130 Byrne, S., Safi, A. G., Kemp, D., Skurka, C., Davydova, Y., Scolere, L., … Niederdeppe, J. (2017). Effects of varying color, imagery and text of cigarette package warning labels among socioeconomically disadvantaged middle school youth and adult smokers. Health Communication, 1–11. http://doi.org/10.1080/10410236.2017.1407228 Clayton, R. B., Leshner, G., Bolls, P. D., & Thorson, E. (2016). Discard the smoking cues—Keep the disgust: An investigation of tobacco smokers’ motivated processing of anti-tobacco commercials. Health Communication, 1–12. http://doi.org/10.1080/10410236.2016.1220042 Clayton, R. B., Leshner, G., Tomko, R. L., Trull, T. J., & Piasecki, T. M. (2017). 140 Countering craving with disgust images: Examining nicotine withdrawn smokers’ motivated message processing of anti-tobacco public service announcements. Journal of Health Communication, 22(3), 254–261. http://doi.org/10.1080/10810730.2016.1268222 Cochran, J., Kydd, R., Lee, J. M., Walker, N., & Consedine, N. S. (2017). Disgust but not health anxiety graphic warning labels reduce motivated attention in smokers: A study of P300 and late positive potential responses. Nicotine & Tobacco Research. Duchowski, A. T. (2002). A breadth-first survey of eye-tracking applications. Behavior Research Methods, Instruments, & Computers, 34(4), 455–470. http://doi.org/10.3758/BF03195475 Family Smoking Prevention and Tobacco Control Act, Pub. L. No. 11131, H.R. 1256 (2009). Germain, D., Wakefield, M. A., & Durkin, S. J. (2010). Adolescents’ perceptions of cigarette brand image: Does plain packaging make a difference? Journal of Adolescent Health, 46(4), 385–392. http://doi.org/10.1016/j.jadohealth.2009.08.009 Haidt, J., McCauley, C. R., & Rozin, P. (1994). Individual differences in sensitivity to disgust: A scale sampling seven domains of disgust elicitors. Personality and Individual Differences, 16, 701–713. http://doi.org/10.1016/0191- 8869(94)90212-7 Halkjelsvik, T., & Rise, J. (2015). Disgust in fear appeal anti-smoking advertisements: The effects on attitudes and abstinence motivation. Drugs: Education, Prevention 141 & Policy, 22(4), 362–369. http://doi.org/10.3109/09687637.2015.1015491 Hill, K. G., Hawkins, J. D., Catalano, R. F., Abbott, R. D., & Guo, J. (2005). Family influences on the risk of daily smoking initiation. Journal of Adolescent Health, 37(3), 202–210. http://doi.org/10.1016/j.jadohealth.2004.08.014 Jónsdóttir, H. L., Holm, J. E., Poltavski, D., & Vogeltanz-Holm, N. (2014). The role of fear and disgust in predicting the effectiveness of television advertisements that graphically depict the health harms of smoking. Preventing Chronic Disease, 11, E218. http://doi.org/10.5888/pcd11.140326 Kessels, L. T. E., & Ruiter, R. A. C. (2012). Eye movement responses to health messages on cigarette packages. BMC Public Health, 12(1), 352. http://doi.org/10.1186/1471-2458-12-352 Kessels, L. T. E., Ruiter, R. A. C., & Jansma, B. M. (2010). Increased attention but more efficient disengagement: Neuroscientific evidence for defensive processing of threatening health information. Health Psychology, 29(4), 346–354. http://doi.org/10.1037/a0019372 Krugman, D. M., Fox, R. J., Fletcher, J. E., Fischer, P. M., & Rojas, T. H. (1994). Do adolescents attend to warnings in cigarette advertising? An eye-tracking approach. Journal of Advertising Research, 34, 39–52. Krusemark, E. A., & Li, W. (2011). Do all threats work the same way? Divergent effects of fear and disgust on sensory perception and attention. Journal of Neuroscience, 31(9), 3429–3434. http://doi.org/10.1523/JNEUROSCI.4394- 10.2011 Leshner, G., Bolls, P. D., & Thomas, E. (2009). Scare’ em or disgust ’em: The effects 142 of graphic health promotion messages. Health Communication, 24(5), 447–458. http://doi.org/10.1080/10410230903023493 Leshner, G., Bolls, P. D., & Wise, K. (2011). Motivated processing of fear appeal and disgust images in televised anti-tobacco ads. Journal of Media Psychology, 23(2), 77–89. http://doi.org/10.1027/1864-1105/a000037 Leshner, G., Clayton, R. B., Bolls, P. D., Bhandari, M., Leshner, G., Clayton, R. B., … Clayton, R. B. (2017). Deceived, disgusted, and defensive: Motivated processing of anti-tobacco advertisements. Health Communication, 1–10. http://doi.org/10.1080/10410236.2017.1350908 Liberman, A., & Chaiken, S. (1992). Defensive processing of personally relevant health messages. Personality and Social Psychology Bulletin, 18(6), 669–679. Manera, V., Samson, A. C., Pehrs, C., Lee, I. A., & Gross, J. J. (2014). The eyes have it: The role of attention in cognitive reappraisal of social stimuli. Emotion, 14(5), 833. McQueen, A., Waters, E. A., Kaphingst, K. A., Caburnay, C. A., Sanders Thompson, V. L., Boyum, S., & Kreuter, M. W. (2016). Examining interpretations of graphic cigarette warning labels among U.S. youth and adults. Journal of Health Communication, 21(8), 855–867. http://doi.org/10.1080/10810730.2016.1177142 Meernik, C., Jarman, K., Wright, S. T., Klein, E. G., & Goldstein, A. O. (2016). Eye tracking outcomes in tobacco control regulation and communication: A systematic review. Tobacco Regulatory Science, 2(4), 377–403. Nabi, R. L. (1998). The effect of disgust-eliciting visuals on attitudes toward animal experimentation. Communication Quarterly, 46(4), 472–484. 143 http://doi.org/10.1080/01463379809370116 Nabi, R. L. (2003). “Feeling” resistance: Exploring the role of emotionally evocative visuals in inducing inoculation. Media Psychology, 5(2), 199–223. http://doi.org/10.1207/S1532785XMEP0502_4 Newman-Norlund, R. D., Thrasher, J. F., Fridriksson, J., Brixius, W., Froeliger, B., Hammond, D., & Cummings, M. K. (2014). Neural biomarkers for assessing different types of imagery in pictorial health warning labels for cigarette packaging: A cross-sectional study. BMJ Open, 4(12). http://doi.org/10.1136/bmjopen-2014-006411. PMCID: PMC4281542 Oaten, M., Stevenson, R. J., & Case, T. I. (2009). Disgust as a disease-avoidance mechanism. Psychological Bulletin, 135(2), 303–321. http://doi.org/10.1037/a0014823 Pechmann, C., & Knight, S. J. (2002). An experimental investigation of the joint effects of advertising and peers on adolescents’ beliefs and intentions about cigarette consumption. Journal of Consumer Research, 29(1), 5–19. http://doi.org/10.1086/339918 Pechmann, C., & Reibling, E. T. (2006). Antismoking advertisements for youths: An independent evaluation of health, counter-industry, and industry approaches. American Journal of Public Health, 96(5), 906–913. http://doi.org/10.2105/AJPH.2004.057273 Peters, E., Evans, A. T., Hemmerich, N., & Berman, M. (2012). Emotion in the law and the lab: The case of graphic cigarette warnings. Tobacco Regulatory Science, 2(4), 404–413. 144 Roseman, I., Wiest, C., & Swartz, T. (1994). Phenomenology, behaviors, and goals differentiate discrete emotions. Journal of Personality and Social Psychology, 67(2), 206–221. http://doi.org/10.1037/0022-3514.67.2.206 Rozin, P., & Singh, L. (1999). The moralization of cigarette smoking in the United States. Journal of Consumer Psychology, 8(3), 321–337. http://doi.org/10.1207/s15327663jcp0803_07 Rubenking, B., & Lang, A. (2014). Captivated and grossed out: An examination of processing core and sociomoral disgusts in entertainment media. Journal of Communication, 64(3), 543–565. http://doi.org/10.1111/jcom.12094 Ruiter, R. A. C., & Kok, G. (2005). Saying is not (always) doing: Cigarette warning labels are useless. European Journal of Public Health, 15(3), 329–330. http://doi.org/10.1093/eurpub/cki095 Shankleman, M., Sykes, C., Mandeville, K. L., Di Costa, S., & Yarrow, K. (2015). Standardised (plain) cigarette packaging increases attention to both text-based and graphical health warnings: Experimental evidence. Public Health, 129(1), 37–42. http://doi.org/10.1016/j.puhe.2014.10.019 Shi, Z., Wang, A., Ma, L. F. E., Ba, K. M. S., & Romer, D. (2016). The importance of relevant emotional arousal in the efficacy of pctorial health warnings for cigarettes. Nicotine & Tobacco Research, 19(6), 1–6. http://doi.org/10.1093/ntr/ntw322 Shimp, T. A., & Stuart, E. W. (2004). The role of disgust as an emotional mediator of advertising effects. Journal of Advertising, 33(1), 43–53. Retrieved from http://www.jstor.org/stable/4189245 145 Skurka, C., Kemp, D., Davydova, Y., Thrasher, J. J., Byrne, S., Safi, A. G., … Niederdeppe, J. (2017). Effects of 30% and 50% cigarette pack graphic warning labels on visual attention, negative affect, quit intentions, and smoking susceptibility among disadvantaged populations in the United States. Nicotine & Tobacco Research, 20(7), 859–866. Slovic, P., Finucane, M. L., Peters, E., & MacGregor, D. G. (2007). The affect heuristic. European Journal of Operational Research, 177(3), 1333–1352. Süssenbach, P., Niemeier, S., & Glock, S. (2013). Effects of and attention to graphic warning labels on cigarette packages. Psychology & Health, 28(10), 1192–1206. Van Hooff, J. C., Devue, C., Vieweg, P. E., & Theeuwes, J. (2013). Disgust- and not fear-evoking images hold our attention. Acta Psychologica, 143(1), 1–6. http://doi.org/10.1016/j.actpsy.2013.02.001 Watson, D., Clark, L. a, & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54(6), 1063–1070. http://doi.org/10.1037/0022-3514.54.6.1063 White, V., Webster, B., & Wakefield, M. (2008). Do graphic health warning labels have an impact on adolescents’ smoking-related beliefs and behaviours? Addiction, 103(9), 1562–1571. http://doi.org/10.1111/j.1360-0443.2008.02294.x Wogalter, M. S., & Laughery, K. R. (1996). Warning! Sign and label effectiveness. Current Directions in Psychological Science, 5(2), 33–37. Wogalter, M. S., & Leonard, S. D. (1999). Attention capture and maintenance. Warnings and Risk Communication, 123–148. 146 Yartz, A. R., & Hawk, L. W. (2002). Addressing the specificity of affective startle modulation: Fear versus disgust. Biological Psychology, 59(1), 55–68. http://doi.org/10.1016/S0301-0511(01)00121-1 Chapter 4 Albouy, J. (2017). Emotions and prosocial behaviours: A study of the effectiveness of shocking charity campaigns. Recherche et Applications En Marketing, 32(2), 4– 25. http://doi.org/10.1177/2051570716689241 Armstrong, T., McClenahan, L., Kittle, J., & Olatunji, B. O. (2014). Don’t look now! Oculomotor avoidance as a conditioned disgust response. Emotion, 14(1), 95– 104. http://doi.org/10.1037/a0034558 Basil, D. Z., Ridgway, N. M., & Basil, M. D. (2006). Guilt appeals: The mediating effect of responsibility. Psychology & Marketing, 23(12), 1–20. http://doi.org/10.1002/mar.20145 Basil, D. Z., Ridgway, N. M., & Basil, M. D. (2008). Guilt and giving: A process model of empathy and efficacy. Psychology & Marketing, 25(1), 1–23. http://doi.org/10.1002/mar.20200 Batson, C. D. (2006). Orchestrating prosocial motives. In Moral leadership: The theory and practice of power, judgment, and policy (pp. 197–212). Bekkers, R., & Wiepking, P. (2011). A literature review of empirical studies of philanthropy: Eight mechanisms that drive charitable giving. Nonprofit and Voluntary Sector Quarterly, 40(5), 924–973. http://doi.org/10.1177/0899764010380927 147 Bodenhausen, G. V, Sheppard, L. A., & Kramer, G. P. (1994). Negative affect and social judgment: The differential impact of anger and sadness. European Journal of Social Psychology, 24, 45–62. http://doi.org/10.1002/ejsp.2420240104 Bolls, P. D. (2010). Understanding emotion from a superordinate dimensional perspective: A productive way forward for communication processes and effects studies. Communication Monographs, 77(2), 146–152. http://doi.org/10.1080/03637751003790477 Buckels, E. E., & Trapnell, P. D. (2013). Disgust facilitates outgroup dehumanization. Group Processes & Intergroup Relations, 20(10), 1–10. http://doi.org/10.1177/1368430212471738 Burt, C., & Strongman, K. (2005). Use of images in charity advertising: Improving donations and compliance rates. International Journal of Organizational Behavior, 8(8), 571–580. Clayton, R. B., Leshner, G., Bolls, P. D., & Thorson, E. (2016). Discard the smoking cues—Keep the disgust: An investigation of tobacco smokers’ motivated processing of anti-tobacco commercials. Health Communication, 1–12. http://doi.org/10.1080/10410236.2016.1220042 Clayton, R. B., Leshner, G., Tomko, R. L., Trull, T. J., & Piasecki, T. M. (2017). Countering craving with disgust images: Examining nicotine withdrawn smokers’ motivated message processing of anti-tobacco public service announcements. Journal of Health Communication, 22(3), 254–261. http://doi.org/10.1080/10810730.2016.1268222 148 Cockrill, A., & Parsonage, I. (2016). Shocking people into action: Does it still work? An empirical analysis of emotional appeals in charity advertising. Journal of Advertising, 56(4), 401–413. http://doi.org/10.2501/JAR-2016-045 Coulter, R. H., Cotte, J., & Moore, M. L. (1999). Believe it or not: Persuasion, manipulation and credibility of guilt appeals. Advances in Consumer Research, 26, 288–294. Retrieved from http://www.acrwebsite.org/volumes/display.asp?id=8264 Coulter, R. H., & Pinto, M. B. (1995). Guilt appeals in advertising: What are their effects? Journal of Applied Psychology, 80(6), 697–705. Diamantopoulos, A., Sarstedt, M., Fuchs, C., Wilczynski, P., & Kaiser, S. (2012). Guidelines for choosing between multi-item and single-item scales for construct measurement: A predictive validity perspective. Journal of the Academy of Marketing Science, 40(3), 434–449. http://doi.org/10.1007/s11747-011-0300-3 Dickert, S., Kleber, J., Peters, E., & Slovic, P. (2011). Numeracy as a precursor to pro- social behavior: The impact of numeracy and presentation format on the cognitive mechanisms underlying donation decisions. Judgment and Decision Making, 6(7), 638–650. Dillard, J. P., & Nabi, R. L. (2006). The persuasive influence of emotion in cancer prevention and detection messages. Journal of Communication, 56(s1), S123– S139. http://doi.org/10.1111/j.1460-2466.2006.00286.x Dillard, J. P., & Peck, E. (2000). Affect and persuasion: Emotional responses to public service announcements. Communication Research, 27(4), 461–495. http://doi.org/10.1177/009365000027004003 149 Dillard, J. P., Plotnick, C. A., Godbold, L. C., Freimuth, V. S., & Edgar, T. (1996). The multiple affective outcomes of AIDS PSAs: Fear appeals do more than scare people. Communication Research, 23(1), 44–72. Dillard, J. P., & Shen, L. (2018). Threat appeals as multi-emotion messages: An argument structure model of fear and disgust. Human Communication Research, 44, 103–126. http://doi.org/10.1093/hcr/hqx002 Dochas. (2007). The Dochas Code of Conduct on Images and Messages. Irish Association of Non-governmental Development Organizations. Ekman, P., & Friesen, W. V. (1975). Unmasking the face: A guide to recognizing emotions from facial clues. Englewood Cliffs, NJ: Prentice Hall, Inc. Fultz, J., Schaller, M., & Cialdini, R. B. (1988). Empathy, sadness, and distress: Three related but distinct vicarious affective responses to another’s suffering. Personality and Social Psychology Bulletin, 14(2), 312–325. Haidt, J. (2003). The moral emotions. In Handbook of Affective Sciences (11th ed., pp. 852–570). Herz, R. S. (2012). That’s disgusting: Unraveling the mysteries of repulsion (First). New York, NY: W. W. Norton & Company, Inc. Hudson, D., VanHeerde-Hudson, J., Dasandi, N., & Gaines, N. S. (2016). Emotional pathways to engagement with global poverty: An experimental analysis. In Public Opinion & Foreign Aid Workshop, University of Essex. Izard, C. E. (1977). Human emotions. New York, NY: Springer Science & Business Media. 150 Joffe, H. (2008). The power of visual material: Persuasion, emotion and identification. Diogenes, (2006), 84–93. http://doi.org/10.1177/0392192107087919 Judah, G., Aunger, R., Schmidt, W. P., Michie, S., Granger, S., & Curtis, V. (2009). Experimental pretesting of hand-washing interventions in a natural setting. American Journal of Public Health, 99, 405–411. http://doi.org/10.2105/AJPH.2009.164160 Kogut, T., & Ritov, I. (2005a). The “identified victim” effect: An identified group, or just a single individual? Journal of Behavioral Decision Making, 18(3), 157–167. http://doi.org/10.1002/bdm.492 Kogut, T., & Ritov, I. (2005b). The singularity effect of identified victims in separate and joint evaluations. Organizational Behavior and Human Decision Processes, 97(2), 106–116. http://doi.org/10.1016/j.obhdp.2005.02.003 Lazarus, R. S. (1991). Emotion and Adaptation. Oxford University Press. Lerner, J. S., Li, Y., Valdesolo, P., & Kassam, K. S. (2015). Emotion and decision making. Annual Review of Psychology, 66, 799–823. http://doi.org/10.1146/annurev-psych-010213-115043 Lerner, J. S., Small, D. A., & Loewenstein, G. (2004). Heart strings and purse strings: Carryover effects of emotions on economic decisions. Psychological Science, 15(5), 337–341. http://doi.org/10.1111/j.0956-7976.2004.00679.x Leshner, G., Bolls, P. D., & Wise, K. (2011). Motivated processing of fear appeal and disgust images in televised anti-tobacco ads. Journal of Media Psychology, 23(2), 77–89. http://doi.org/10.1027/1864-1105/a000037 151 Leshner, G., Clayton, R. B., Bolls, P. D., Bhandari, M., Leshner, G., Clayton, R. B., … Clayton, R. B. (2017). Deceived, disgusted, and defensive: Motivated processing of anti-tobacco advertisements. Health Communication, 1–10. http://doi.org/10.1080/10410236.2017.1350908 Leshner, G., Vultee, F., Bolls, P. D., & Moore, J. (2010). When a fear appeal isn’t just a fear appeal: The effects of graphic anti-tobacco messages. Journal of Broadcasting & Electronic Media, 54(3), 485–507. http://doi.org/10.1080/08838151.2010.498850 Lupton, D. (2014). The pedagogy of disgust: The ethical, moral and political implications of using disgust in public health campaigns. Critical Public Health, (February), 1–14. http://doi.org/10.1080/09581596.2014.885115 Massi Lindsey, L. L. (2005). Anticipated guilt as behavioral motivation: An examination of appeals to help unknown others through bone marrow donation. Human Communication Research, 31(4), 453–481. http://doi.org/10.1111/j.1468- 2958.2005.tb00879.x Massi Lindsey, L. L., Yun, K. A., & Hill, J. B. (2007). Anticipated guilt as motivation to help unknown others: An examination of empathy as a moderator. Communication Research, 34(4), 468–480. http://doi.org/10.1177/0093650207302789 McGuire, W. J. (1961). The effectiveness of supportive and refutational defenses in immunizing and restoring beliefs against persuasion. Sociometry, 24(2), 184–197. http://doi.org/10.2307/2786067 152 McGuire, W. J. (1968). Personality and attitude change: An information-processing theory. In Psychological Foundations of Attitudes (pp. 171–196). http://doi.org/10.1016/B978-1-4832-3071-9.50013-1 McGuire, W. J., & Papageorgis, D. (1962). Effectiveness of forewarning in developing resistance to persuasion. Public Opinion Quarterly, 26(1), 24–34. http://doi.org/10.1086/267068 Miller, A., & Leshner, G. (2007). How viewers process live, breaking, and emotional television news. Media Psychology, 10, 23–40. Moors, A., Ellsworth, P. C., Scherer, K. R., & Frijda, N. H. (2013). Appraisal theories of emotion: State of the art and future development. Emotion Review, 5(2), 119– 124. Morales, A. C., Wu, E. C., & Fitzsimons, G. J. (2012). How disgust enhances the effectiveness of fear appeals. Journal of Marketing Research, 49(3), 383–393. http://doi.org/10.1509/jmr.07.0364 Nabi, R. L. (1998). The effect of disgust-eliciting visuals on attitudes toward animal experimentation. Communication Quarterly, 46(4), 472–484. http://doi.org/10.1080/01463379809370116 Nabi, R. L. (1999). A cognitive-functional model for the effects of discrete negative emotions on information processing, attitude change, and recall. Communication Theory, 9(3), 292–320. http://doi.org/10.1111/j.1468-2885.1999.tb00172.x Nabi, R. L. (2002). The theoretical versus the lay meaning of disgust: Implications for emotion research. Cognition & Emotion, 16(5), 695–703. http://doi.org/10.1080/02699930143000437 153 O’Keefe, D. J. (2000). Guilt and social influence. Annals of the International Communication Association, 23(1), 67–101. http://doi.org/10.1080/23808985.2000.11678970 O’Keefe, D. J. (2002). Guilt as a mechanism of persuasion. In The Persuasion Handbook: Developments in Theory and Practice (pp. 329–344). Perrine, R. M., & Heather, S. (2000). Effects of picture and even-a-penny-will-help- appeals on anonymous donations to charity. Psychological Reports, 86, 551–559. Petty, R., & Cacioppo, J. (1986). The elaboration likelihood model of persuasion. In Advances in Experimental Social Psychology (pp. 123–205). http://doi.org/10.1016/S0065-2601(08)60214-2 Pfister, H.-R., & Böhm, G. (2008). The multiplicity of emotions: A framework of emotional functions in decision making. Judgment and Decision Making, 3(1), 5– 17. Polman, E., & Kim, S. H. (2013). Effects of anger, disgust, and sadness on sharing with others. Personality & Social Psychology Bulletin, 39(12), 1683–92. http://doi.org/10.1177/0146167213500998 Ranganathan, S. K., & Henley, W. H. (2008). Determinants of charitable donation intentions: A structural equation model. International Journal of Nonprofit & Voluntary Sector Marketing, 13(1), 1–11. http://doi.org/10.1002/nvsm.297 Renner, S., Lindenmeier, J., Tscheulin, D. K., & Drevs, F. (2013). Guilt appeals and prosocial behavior: An experimental analysis of the effects of anticipatory versus reactive guilt appeals on the effectiveness of blood donor appeals. In Journal of 154 Nonprofit & Public Sector Marketing (Vol. 25, pp. 237–255). http://doi.org/10.1080/10495142.2013.816595 Richards, A. S., & Banas, J. A. (2014). Inoculating against reactance to persuasive health messages. Health Communication, 0(0), 1–10. http://doi.org/10.1080/10410236.2013.867005 Roseman, I., Wiest, C., & Swartz, T. (1994). Phenomenology, behaviors, and goals differentiate discrete emotions. Journal of Personality and Social Psychology, 67(2), 206–221. http://doi.org/10.1037/0022-3514.67.2.206 Rubenking, B., & Lang, A. (2014). Captivated and grossed out: An examination of processing core and sociomoral disgusts in entertainment media. Journal of Communication, 64(3), 543–565. http://doi.org/10.1111/jcom.12094 Sargeant, A. (1999). Charitable giving: Towards a model of donor behaviour. Journal of Marketing Management, 15, 215–238. Scherer, K. R. (2009). Emotions are emergent processes: They require a dynamic computational architecture. Philosophical Transactions of the Royal Society B: Biological Sciences, 364(1535), 3459–3474. http://doi.org/10.1098/rstb.2009.0141 Schienle, A., Übel, S., Gremsl, A., Schöngassner, F., & Körner, C. (2016). Disgust proneness and the perception of disgust-evoking pictures. Journal of Psychophysiology, 30(3), 124–129. Sherman, G. D., & Haidt, J. (2011). Cuteness and disgust: The humanizing and dehumanizing effects of emotion. In Emotion Review (Vol. 3, pp. 245–251). http://doi.org/10.1177/1754073911402396 155 Slovic, P. (2010). The more who die, the less we care. In E. Michel-Kerjan & P. Slovic (Eds.), The Irrational Economist: Making Decisions in a Dangerous World (pp. 30–40). New York: Public Affairs Press. Small, D. A., & Loewenstein, G. (2003). Helping a victim or helping the victim: Altrusim and identifiability. Journal of Risk and Uncertainty, 26(1), 5–16. http://doi.org/10.1023/A:1022299422219 Small, D. A., & Verrochi, N. M. (2009). The face of need: Facial emotion expression on charity advertisements. Journal of Marketing Research. http://doi.org/10.1509/jmkr.46.6.777 Smith, C. A., & Ellsworth, P. C. (1985). Patterns of cognitive appraisal in emotion. Journal of Personality and Social Psychology, 48(4), 813–838. Van Hooff, J. C., Devue, C., Vieweg, P. E., & Theeuwes, J. (2013). Disgust- and not fear-evoking images hold our attention. Acta Psychologica, 143(1), 1–6. http://doi.org/10.1016/j.actpsy.2013.02.001 Wedel, M., & Pieters, R. (2006). Eye tracking for visual marketing. Foundations and Trends® in Marketing, 1(4), 231–320. http://doi.org/10.1561/1700000011 Wheatley, T., & Haidt, J. (2005). Hypnotic disgust makes moral judgments more severe. Psychological Science, 16(10), 780–784. Xing, C. (2014). Effects of anger and sadness on attentional patterns in decision making: An eye-tracking study. Psychological Reports, 114(1), 50–67. http://doi.org/10.2466/01.04.PR0.114k14w3 156 Yartz, A. R., & Hawk, L. W. (2002). Addressing the specificity of affective startle modulation: Fear versus disgust. Biological Psychology, 59(1), 55–68. http://doi.org/10.1016/S0301-0511(01)00121-1 Chapter 5 Bates, B. R., & Chadwick, A. E. (2015). Measuring state disgust: Initial evidence for the reliability and validity of an English-Language version of the Ekel-State- Fragebogen. Communication Research Reports, 32(1), 73–82. http://doi.org/10.1080/08824096.2014.989978 Batson, C. D., Lishner, D. A., & Stocks, E. L. (2015). The empathy–altruism hypothesis. In The Oxford Handbook of Prosocial Behavior (p. 259). http://doi.org/10.1093/oxfordhb/9780195399813.013.023 Bekkers, R. (2010). Who gives what and when? A scenario study of intentions to give time and money. Social Science Research, 39(3), 369–381. http://doi.org/10.1016/j.ssresearch.2009.08.008 Brown, S. L., & Whiting, D. (2014). The ethics of distress: Toward a framework for determining the ethical acceptability of distressing health promotion advertising. International Journal of Psychology, 49(2), 89–97. http://doi.org/10.1002/ijop.12002 Buckels, E. E., & Trapnell, P. D. (2013). Disgust facilitates outgroup dehumanization. Group Processes & Intergroup Relations, 20(10), 1–10. http://doi.org/10.1177/1368430212471738 Camerer, C. F., & Fehr, E. (2004). Measuring social norms and preferences using 157 experimental games: A guide for social scientists. In Foundations of human sociality: Economic experiments and ethnographic evidence from fifteen small- scale societies (pp. 55–95). http://doi.org/10.1093/0199262055.003.0003 Chaiken, S. (1980). Heuristic versus systematic information processing and the use of source versus message cues in persuasion. Journal of Personality and Social Psychology, 39(5), 752–766. http://doi.org/10.1037/0022-3514.39.5.752 Cherry, T. L., Frykblom, P., & Shogren, J. F. (2002). Hardnose the dictator. The American Economic Review, 92(4), 1218–1221. Retrieved from http://www.jstor.org/stable/3083309 Cuddy, A. J. C., Fiske, S. T., & Glick, P. (2008). Warmth and competence as universal dimensions of social perception: The Stereotype Content Model and the BIAS Map. In Advances in Experimental Social Psychology (Vol. 40, pp. 61–137). http://doi.org/10.1016/S0065-2601(07)00002-0 Dasgupta, N., Desteno, D., Williams, L. A., & Hunsinger, M. (2009). Fanning the flames of prejudice: The influence of specific incidental emotions on implicit prejudice. Emotion, 9(4), 585–591. http://doi.org/10.1037/a0015961 Dickert, S., Sagara, N., & Slovic, P. (2011). Affective motivations to help others: A two-stage model of donation decisions. Journal of Behavioral Decision Making, 24(4), 361–376. http://doi.org/10.1002/bdm.697 Dochas. (2007). The Dochas Code of Conduct on Images and Messages. Irish Association of Non-governmental Development Organizations. Eskine, K. J., Kacinik, N. A., & Prinz, J. J. (2011). A bad taste in the mouth: Gustatory disgust influences moral judgment. Psychological Science, 22(3), 295– 158 299. http://doi.org/10.1177/0956797611398497 Fiske, S. T., Cuddy, A. J. C., Glick, P., & Xu, J. (2002). A model of (often mixed) stereotype content: Competence and warmth respectively follow from perceived status and competition. Journal of Personality and Social Psychology, 82(6), 878–902. http://doi.org/10.1037//0022-3514.82.6.878 Fultz, J., Schaller, M., & Cialdini, R. B. (1988). Empathy, sadness, and distress: Three related but distinct vicarious affective responses to another’s suffering. Personality and Social Psychology Bulletin, 14(2), 312–325. Gierczak, M. M., Bretschneider, U., Haas, P., Blohm, I., & Leimeister, J. M. (2016). Crowdfunding: outlining the new era of fundraising. In Crowdfunding in Europe (pp. 7–23). Springer, Cham. Haidt, J. (2001). The emotional dog and its rational tail: A social intuitionist approach to moral judgment. Psychological Review, 108(4), 814–834. http://doi.org/10.1037/0033-295X.108.4.814 Haidt, J. (2003). The moral emotions. In Handbook of Affective Sciences (11th ed., pp. 852–570). Haidt, J., McCauley, C. R., & Rozin, P. (1994). Individual differences in sensitivity to disgust: A scale sampling seven domains of disgust elicitors. Personality and Individual Differences, 16, 701–713. http://doi.org/10.1016/0191- 8869(94)90212-7 Haidt, J., Rozin, P., Mccauley, C., & Imada, S. (1997). Body, psyche, and culture: The relationship between disgust and morality. Psychology and Developing Societies, 9(1), 107–131. 159 Ihme, J. M., & Mitte, K. (2009). Measuring state disgust: Evaluation of a German state disgust scale. European Journal of Psychological Assessment, 25(3), 150–156. http://doi.org/10.1027/1015-5759.25.3.150 Inbar, Y., & Gilovich, T. (2011). Angry (or disgusted), but adjusting? The effect of specific emotions on adjustment from self-generated anchors. Social Psychological and Personality Science, 2(6), 563–569. http://doi.org/10.1177/1948550611401426 Inbar, Y., Pizarro, D. A., & Bloom, P. (2012). Disgusting smells cause decreased liking of gay men. Emotion, 12(1), 23–27. http://doi.org/10.1037/a0023984 Jenni, K. E., & Loewenstein, G. (1997). Explaining the “Identifiable Victim Effect.” Journal of Risk and Uncertainty, 14, 235–257. http://doi.org/10.1023/A:1007740225484 Joffe, H. (2008). The power of visual material: Persuasion, emotion and identification. Diogenes, (2006), 84–93. http://doi.org/10.1177/0392192107087919 Kennedy, S., & Hill, S. (2010). Global poverty, aid advertisements, and cognition: Do media images of the developing world lead to positive or negative responses in viewers. New Zealand Journal of Psychology, 39(2), 56–66. Landy, J. F., & Goodwin, G. P. (2015). Does incidental disgust amplify moral judgment? A meta-analytic review of experimental evidence. Perspectives on Psychological Science, 10(4), 518–536. http://doi.org/10.1177/1745691615583128 Lerner, J. S., & Keltner, D. (2000). Beyond valence: Toward a model of emotion- specific influences on judgement and choice. Cognition & Emotion, 14(4), 473– 160 493. Lerner, J. S., Li, Y., Valdesolo, P., & Kassam, K. S. (2015). Emotion and decision making. Annual Review of Psychology, 66, 799–823. http://doi.org/10.1146/annurev-psych-010213-115043 Lerner, J. S., & Tiedens, L. Z. (2006). Portrait of the angry decision maker: How appraisal tendencies shape anger’s influence on cognition. Journal of Behavioral Decision Making, 19, 115–137. Lupton, D. (2014). The pedagogy of disgust: The ethical, moral and political implications of using disgust in public health campaigns. Critical Public Health, (February), 1–14. http://doi.org/10.1080/09581596.2014.885115 Metropolitan Ministries. (2016). Zika virus kit for the homeless. Retrieved May 24, 2018, from https://secure.metromin.org/site/Ecommerce/1702732264?VIEW_PRODUCT=tr ue&product_id=1501&store_id=1301 Moors, A., Ellsworth, P. C., Scherer, K. R., & Frijda, N. H. (2013). Appraisal theories of emotion: State of the art and future development. Emotion Review, 5(2), 119– 124. Nabi, R. L. (2003). Exploring the framing effects of emotion: Do discrete emotions differentially influence information accessibility, information seeking, and policy preference? Communication Research, 30(2), 224–247. http://doi.org/10.1177/0093650202250881 Pizarro, D. A., Detweiler-Bedell, B., & Bloom, P. (2006). The creativity of everyday moral reasoning. In J. C. Kaufman & J. Baer (Eds.), Creativity and Reason in 161 Cognitive Development (pp. 81–98). New York, NY: Cambridge University Press. Rohrmann, S., Hopp, H., & Quirin, M. (2008). Gender differences in psychophysiological responses to disgust. Journal of Psychology, 22(2004), 65– 75. http://doi.org/10.1027/0269-8803.22.2.65 Rozin, P., Haidt, J., & McCauley, C. R. (2008). Disgust. In Handbook of Emotions (pp. 757–776). Rubenking, B., & Lang, A. (2014). Captivated and grossed out: An examination of processing core and sociomoral disgusts in entertainment media. Journal of Communication, 64(3), 543–565. http://doi.org/10.1111/jcom.12094 Russell, P. S., & Giner-Sorolla, R. (2013). Bodily moral disgust: What it is, how it is different from anger, and why it is an unreasoned emotion. Psychological Bulletin, 139(2), 328–351. http://doi.org/10.1037/a0029319 Sabin Vaccine Institute. (2013). How to shock a celebrity. Retrieved May 24, 2016, from https://www.youtube.com/watch?v=sYimJKg9QiE Schnall, S., Haidt, J., Clore, G. L., & Jordan, A. H. (2008). Disgust as embodied moral judgment. In Personality and Social Psychology Bulletin (Vol. 34, pp. 1096– 1109). http://doi.org/10.1177/0146167208317771 Seidel, A., & Prinz, J. (2013). Sound morality: Irritating and icky noises amplify judgments in divergent moral domains. Cognition, 127(1), 1–5. http://doi.org/10.1016/j.cognition.2012.11.004 Shah, A. K., & Oppenheimer, D. M. (2008). Heuristics made easy: An effort-reduction framework. Psychological Bulletin, 134(2), 207–222. 162 http://doi.org/10.1037/0033-2909.134.2.207 Shelton, M. Lou, & Rogers, R. W. (1981). Fear-arousing and empathy-arousing appeals to help: The pathos of persuasion. In Journal of Applied Social Psychology (Vol. 11, pp. 366–378). http://doi.org/10.1111/j.1559- 1816.1981.tb00829.x Sherman, G. D., & Haidt, J. (2011). Cuteness and disgust: The humanizing and dehumanizing effects of emotion. In Emotion Review (Vol. 3, pp. 245–251). http://doi.org/10.1177/1754073911402396 Slovic, P., Finucane, M. L., Peters, E., & MacGregor, D. G. (2007). The affect heuristic. European Journal of Operational Research, 177(3), 1333–1352. Small, D. A., & Verrochi, N. M. (2009). The face of need: Facial emotion expression on charity advertisements. Journal of Marketing Research. http://doi.org/10.1509/jmkr.46.6.777 Smith, C. A., & Ellsworth, P. C. (1985). Patterns of cognitive appraisal in emotion. Journal of Personality and Social Psychology, 48(4), 813–838. Stocks, E. L., Lishner, D. A., & Decker, S. K. (2009). Altruism or psychological escape: Why does empathy promote prosocial behavior? European Journal of Social Psychology, 39(5), 649–665. Tiedens, L. Z., & Linton, S. (2001). Judgment under emotional certainty and uncertainty: The effects of specific emotions on information processing. Journal of Personality and Social Psychology, 81(6), 973–988. Wheatley, T., & Haidt, J. (2005). Hypnotic disgust makes moral judgments more severe. Psychological Science, 16(10), 780–784. 163 Zeelenberg, M., Nelissen, R. M. A., Breugelmans, S. M., & Pieters, R. (2008). On emotion specificity in decision making: Why feeling is for doing. In Judgment and Decision Making (Vol. 3, pp. 18–27). Retrieved from https://search.proquest.com/docview/1011218372?accountid=10267 Chapter 6 Angyal, A. (1941). Disgust and related aversions. Journal of Abnormal and Social Psychology, 36(3), 393–412. Brown, S. L., & Whiting, D. (2014). The ethics of distress: Toward a framework for determining the ethical acceptability of distressing health promotion advertising. International Journal of Psychology, 49(2), 89–97. http://doi.org/10.1002/ijop.12002 Camerer, C. F. (2003). Behavioral game theory: Experiments in strategic interaction. New York: Russell Sage Foundation. Camerer, C. F., & Fehr, E. (2004). Measuring social norms and preferences using experimental games: A guide for social scientists. In Foundations of human sociality: Economic experiments and ethnographic evidence from fifteen small- scale societies (pp. 55–95). http://doi.org/10.1093/0199262055.003.0003 Charash, M., & McKay, D. (2002). Attention bias for disgust. Journal of Anxiety Disorders, 16(5), 529–541. http://doi.org/10.1016/S0887-6185(02)00171-8 Clark, J. A., & Powell, P. A. (2017). Defending disgust: Why disgust is morally beneficial. In C. Price & E. Walle (Eds.), Emotion Researcher, ISRE’s Sourcebook for Research on Emotion and Affect. Retrieved from 164 http://emotionresearcher.com/why-disgust-is-morally-beneficial/ Grühn, D., & Sharifian, N. (2016). Lists of emotional stimuli. In Emotion Measurement (pp. 145–164). Elsevier Ltd. http://doi.org/10.1016/B978-0-08- 100508-8.00007-2 Haidt, J. (2001). The emotional dog and its rational tail: A social intuitionist approach to moral judgment. Psychological Review, 108(4), 814–834. http://doi.org/http://dx.doi.org/10.1037/0033-295X.108.4.814 Hajcak, G., & Olvet, D. M. (2008). The persistence of attention to emotion: Brain potentials during and after picture presentation. Emotion, 8(2), 250–255. http://doi.org/10.1037/1528-3542.8.2.250 Herz, R. S. (2012). That’s disgusting: Unraveling the mysteries of repulsion (First). New York, NY: W. W. Norton & Company, Inc. Joffe, H. (2008). The power of visual material: Persuasion, emotion and identification. Diogenes, (2006), 84–93. http://doi.org/10.1177/0392192107087919 Lazarus, R. S. (1991). Emotion and Adaptation. Oxford University Press. Leshner, G., Bolls, P. D., & Thomas, E. (2009). Scare’ em or disgust ’em: The effects of graphic health promotion messages. Health Communication, 24(5), 447–458. http://doi.org/10.1080/10410230903023493 Leshner, G., Bolls, P. D., & Wise, K. (2011). Motivated processing of fear appeal and disgust images in televised anti-tobacco ads. Journal of Media Psychology, 23(2), 77–89. http://doi.org/10.1027/1864-1105/a000037 Leshner, G., Vultee, F., Bolls, P. D., & Moore, J. (2010). When a fear appeal isn’t just a fear appeal: The effects of graphic anti-tobacco messages. Journal of 165 Broadcasting & Electronic Media, 54(3), 485–507. http://doi.org/10.1080/08838151.2010.498850 Liljenquist, K., Zhong, C.-B., & Galinsky, A. D. (2010). The smell of virtue: Clean scents promote reciprocity and charity. Psychological Science, 21(3), 381–383. http://doi.org/10.1177/0956797610361426 Loewenstein, G., & Donoghue, T. O. (2015). Modeling the interplay between affect and deliberation. Decision, 2(2), 55–81. Nabi, R. L. (2002). Anger, fear, uncertainty, and attitudes: A test of the cognitive- functional model. Communication Monographs, 69(3), 204–216. http://doi.org/10.1080/03637750216541 Nabi, R. L., & Green, M. C. (2015). The role of a narrative’s emotional flow in promoting persuasive outcomes. Media Psychology, 18, 137–162. http://doi.org/10.1080/15213269.2014.912585 Orquin, J. L., & Mueller Loose, S. (2013). Attention and choice: A review on eye movements in decision making. Acta Psychologica, 144(1), 190–206. http://doi.org/10.1016/j.actpsy.2013.06.003 Pechmann, C., & Reibling, E. T. (2006). Antismoking advertisements for youths: An independent evaluation of health, counter-industry, and industry approaches. American Journal of Public Health, 96(5), 906–913. http://doi.org/10.2105/AJPH.2004.057273 Peters, E., Evans, A. T., Hemmerich, N., & Berman, M. (2012). Emotion in the law and the lab: The case of graphic cigarette warnings. Tobacco Regulatory Science, 2(4), 404–413. 166 Rozin, P., & Singh, L. (1999). The moralization of cigarette smoking in the United States. Journal of Consumer Psychology, 8(3), 321–337. http://doi.org/10.1207/s15327663jcp0803_07 Rubenking, B., & Lang, A. (2014). Captivated and grossed out: An examination of processing core and sociomoral disgusts in entertainment media. Journal of Communication, 64(3), 543–565. http://doi.org/10.1111/jcom.12094 Russell, P. S., & Giner-Sorolla, R. (2013). Bodily moral disgust: What it is, how it is different from anger, and why it is an unreasoned emotion. Psychological Bulletin, 139(2), 328–351. http://doi.org/10.1037/a0029319 Sherman, G. D., & Haidt, J. (2011). Cuteness and disgust: The humanizing and dehumanizing effects of emotion. In Emotion Review (Vol. 3, pp. 245–251). http://doi.org/10.1177/1754073911402396 Simpson, J., Carter, S., Anthony, S. H., & Overton, P. G. (2006). Is disgust a homogeneous emotion ? Motivation & Emotion, 30(1), 31–41. http://doi.org/10.1007/s11031-006-9005-1 Smith, C. A., & Ellsworth, P. C. (1985). Patterns of cognitive appraisal in emotion. Journal of Personality and Social Psychology, 48(4), 813–838. Tiedens, L. Z., & Linton, S. (2001). Judgment under emotional certainty and uncertainty: The effects of specific emotions on information processing. Journal of Personality and Social Psychology, 81(6), 973–988. Van Hooff, J. C., Devue, C., Vieweg, P. E., & Theeuwes, J. (2013). Disgust- and not fear-evoking images hold our attention. Acta Psychologica, 143(1), 1–6. http://doi.org/10.1016/j.actpsy.2013.02.001 167 Weinberg, A., & Hajcak, G. (2010). Beyond good and evil: The time-course of neural activity elicited by specific picture content. Emotion, 10(6), 767–782. http://doi.org/10.1037/a0020242 168 APPENDICES Appendix 1: Research Questions, Hypotheses, and Operationalizations Disgust, Attention, and Emotion: Adolescent Reactions to Cigarette Graphic Warning Labels (Study 1) RESEARCH QUESTIONS & HYPOTHESES OPERATIONALIZATION H1a: Warnings that contain disgust-eliciting images H1a: t-test will draw participants’ attention more quickly and hold IV: Disgust vs. No Disgust GWLs their attention for a longer time than warnings that do DV: Warning ATFF & AFD not contain disgust-eliciting images. Result: Not Supported H1b: Warnings that contain disgust-eliciting images H1b: t-test will draw participants’ attention more quickly but hold IV: Disgust vs. No Disgust GWLs their attention for a shorter time than warnings that do DV: Warning ATFF & AFD not contain disgust-eliciting images. Result: Not Supported H1c: Warnings that contain disgust-eliciting images H1c: t-test will draw participants’ attention more slowly but hold IV: Disgust vs. No Disgust GWLs their attention for a longer time than warnings that do DV: Warning ATFF & AFD not contain disgust-eliciting images. Result: Not Supported H1d: Warnings that contain disgust-eliciting images H1d: t-test will draw participants’ attention more slowly and hold IV: Disgust vs. No Disgust GWLs their attention for a shorter time than warnings that do DV: Warning ATFF & AFD not contain disgust-eliciting images. Result: Supported RQ1: Do participants who live with smokers attend to RQ1: t-test warnings with gruesome visuals differently than IV: Disgust vs. No Disgust GWLs participants who do not? DV: Warning ATFF & AFD Result: No differences between groups H2a: Participants will spend more time looking at the H2a: t-test image when the warning contains a disgust visual than IV: Disgust vs. No Disgust GWLs when the warning does not contain a disgust visual. DV: Image AFD Result: Supported H2b: Participants will spend less time looking at the H2b: t-test text when the warning contains a disgust visual than IV: Disgust vs. No Disgust GWLs they will when the warning does not contain a disgust DV: Text AFD visual. Result: Supported H3: There will be a significant, positive relationship H3: OLS Regression between attention to disgust imagery and adolescents’ IV: Disgust Image AFD perceptions of risk. DV: Risk Beliefs Result: Not Supported 169 RQ2: Does attention predict differences in risk beliefs RQ2: OLS Regression between participants who live with smokers and those IV: AFD * Lives with Smoker who do not? DV: Risk Beliefs Result: No interaction effect H4: Increased attention to disgust visuals will be H4: OLS Regression associated with increased negative emotional reactions. IV: Disgust Image AFD DV: Negative Emotion Result: Supported RQ3: Does attention predict differences in negative RQ3: OLS Regression emotional reactions between participants who live with IV: AFD * Lives with Smoker smokers and those who do not? DV: Negative Emotion Result: Significant interaction effect The Effect of Attention and Discrete Emotions on Message Elaboration and Intentions in Response to a Disgust-Eliciting Charitable Appeal (Study 2) RESEARCH QUESTIONS & HYPOTHESES OPERATIONALIZATION RQ1: Does exposure to disgust stimuli decrease or RQ1: OLS Regression increase visual attention toward a charitable appeal? IV: Treatment Conditions DV: Message AFD Result: No differences by condition RQ2: Is there evidence of an attentional bias for the RQ2: OLS Regression, t-test disgust-evoking imagery? IV: Treatment Conditions DV: Image TFD Result: Increased attention to disgust image over neutral image, but greater attention to text RQ3: Does intensity of the disgust-evoking imagery RQ3: OLS Regression exacerbate or attenuate these attentional patterns? IV: Treatment Conditions DV: Image AFD Result: No differences based on intensity/vividness of disgust content RQ4: Does attention to disgust-evoking imagery RQ4: OLS Regression predict subsequent cognitions and intentions in IV: Image AFD response to the message? DV: Elaboration, Intention Result: No effect of attention to disgust image on cognitive or intention measures RQ5: Does forewarning increase or reduce visual RQ5: OLS Regression attention toward the message? IV: Treatment Conditions DV: Message AFD Result: No differences by condition 170 H1: Message conditions preceded by the forewarning H1: OLS Regression pre-treatment will reduce levels of disgust experienced IV: Treatment Conditions in comparison to message conditions without the DV: Disgust Reactions forewarning pre-treatment Result: Supported for highly vivid conditions only H2: Higher levels of disgust experienced will be H2: OLS Regression negatively associated with quantity of thoughts IV: Disgust Reactions expressed in response to the message. DV: Elaboration Result: Supported H3: Higher levels of disgust experienced will be H3: OLS Regression negatively associated with behavioral intentions in IV: Disgust Reactions response to the message. DV: Intention Result: Not supported H4: There will be a positive association between a) H4: OLS Regression level of guilt and b) level of sadness experienced and IV: Sadness, Guilt cognitive elaboration in response to the message. DV: Elaboration Result: H4b supported for sadness; H4a not supported for guilt H5: There will be a positive association between a) H5: OLS Regression level of guilt and b) level of sadness experienced and IV: Sadness, Guilt behavioral intentions in response to the message DV: Intention Result: H5a supported for guilt; H5b not supported for sadness Encountering Disgust in Prosocial Persuasion: How Context Shapes Appraisals, Emotions, and Behavior (Study 3) RESEARCH QUESTIONS & HYPOTHESES OPERATIONALIZATION H1a/b: Disgust cues unaccompanied (accompanied) by H1: OLS Regression a persuasive prosocial context will result in greater IV: 2x2 Treatment Groups (lower) certainty about the emotional event. DV: Certainty Result: Supported H2a/b: Disgust cues unaccompanied by a persuasive H2: OLS Regression prosocial context will result in outcomes consistent IV: 2x2 Treatment Groups with heuristic processing on a subsequent social DV: Certainty*Warmth ratings (aid judgment task (higher stereotypes) recipients) Result: Not Supported H3a: Participants exposed to the disgust-eliciting H3a: OLS Regression images without the persuasive context will experience IV: 2x2 Treatment Groups greater levels of disgust. DV: Disgust Result: Supported 171 H3b: Participants exposed to the disgust-eliciting H3b: OLS Regression images without the persuasive context will also report IV: 2x2 Treatment Groups greater personal distress. DV: Personal Distress Result: Supported H3c: Participants exposed to the disgust-eliciting H3c: OLS Regression images in the persuasive context will experience IV: 2x2 Treatment Groups greater empathy than participants exposed to disgust DV: Empathetic Concern imagery without the persuasive context. Result: Supported H4: Participants exposed to the disgust-eliciting images H4: OLS Regression without persuasive context will be less likely to make a IV: 2x2 Treatment Groups gift than participants exposed to the persuasive appeal. DV: Donation Behavior Result: Not supported H5a: Disgust responses will be negatively associated H5a: OLS Regression with donation behavior (whether participants make a IV: Disgust gift of any amount or not). DV: Donation Behavior Result: Not supported H5b: Distress and empathy will positively predict H5b: OLS Regression donation behavior mediating the relationship between IV: Personal Distress, Empathetic exposure and giving. Concern DV: Donation Behavior Result: Not supported H5c: Distress will negatively predict donation amount H5d: OLS Regression among participants who make a gift, reducing gift IV: Personal Distress amount. DV: Donation Amount Result: Not supported H5d: Empathy experienced will positively predict H5d: OLS Regression donation amount among participants who make a gift, IV: Empathetic Concern increasing gift amount. DV: Donation Amount Result: Not supported H6a: Participants exposed to disgust stimuli will report H64: OLS Regression greater difficulty of psychological escape IV: 2x2 Treatment Groups DV: Psychological Escape Result: Supported H6b: Psychological escape will moderate the H5d: OLS Regression relationship between personal distress and donation IV: Personal Distress*Psychological behavior. Increasing difficulty of escape should Escape increase likelihood of giving. DV: Donation Behavior Result: Not supported 172 Appendix 2: Chapter 3 Post-test Items Now we will ask you questions about the pictures of cigarette packs you just saw. Please give us your best answers. When you have finished answering a set of questions, click "NEXT" to move on. REMEMBER: this isn’t a test or anything, so just answer honestly. We don’t connect your name to the answers. Click "NEXT" below to begin. Negative Emotion After looking at the pictures of cigarette packs, I felt... Not at all A little Somewhat Quite a bit Extremely Afraid Scared Disturbed Grossed out Risk Beliefs Do you believe cigarette smoking is related to... Definitely not Probably not Probably yes Definitely yes Cancer Heart disease Lung disease Stroke Hole in the throat Asthma Problems in babies whose moms smoke 173 Live with Smoker Does anyone who lives with you smoke cigarettes? Yes No Gender Which of the following best describes you? Male Female Prefer not to answer Age How old are you today? 10 years old 11 years old 12 years old 13 years old 14 years old 15 years old 16 years old 17 years old 18 years old or older Ethnicity Which of the following best describes you? Please select 1 or more. White Black or African American Hispanic/Latino American Indian or Alaska Native Asian/Asian American 174 Previous Smoking Have you ever tried cigarette smoking, even one or two puffs? Yes No Sensation Seeking How much do you agree or disagree with the following statements? Strongly agree Agree Disagree Strongly disagree I like to do frightening things. I like new and exciting experiences, even if I have to break the rules. I prefer friends who are exciting and unpredictable. Colorblind Have you ever been told you are colorblind? Yes No Uncertain 175 Appendix 3A: Chapter 4 Distractor Task 176 Appendix 3B: Chapter 4 Forewarning Pretreatment The following message about a global health issue may evoke intense, unpleasant feelings. You may feel the urge to stop viewing the message. However, the realities of these health conditions are pretty compelling when you think about them, and suggestions at the end for how to address these issues are important to consider. 177 Appendix 3C: Chapter 4 Stimuli Manipulation Low Disgust Treatment 178 High Disgust Treatment 179 Control 180 Appendix 3D: Chapter 4 Post-test Items Negative Emotion On a scale of 1 to 7, where 1 indicates not at all and 7 indicates very strongly, select the extent to which you feel this way right now: Not at all/Very Strongly Sad Guilty Grossed out Disgusted Physiological Disgust Items Thinking back to the messages about tropical diseases you just saw, indicate the extent to which you agree with the following statements. Strongly disagree/Strongly agree Viewing the message made me feel nauseous. Viewing the message made my stomach turn. Viewing the message made my skin crawl. Thought Listing Task (90 seconds) Describe all the feelings you experienced and thoughts that came to mind when you were exposed to the information about neglected tropical diseases. Behavioral Intention After seeing the ads, I want to make a donation. Strongly disagree/Strongly agree How likely are you to make a donation to the organization? Not very likely/Very likely If you were to make a donation, how soon would you do so? Not at all soon/Very soon 181 Demographic Variables What is your age? Enter a number. Which gender do you identify with? Male Female Other What is your ethnic background? Select all that apply. Black Caucasian Hispanic Asian Other Personal Relevance Were you aware of the issue before participating in this study? Yes No Unsure Do you feel personally at risk for tropical diseases? Yes No Unsure Have you or anyone you know ever suffered a tropical disease? Yes No Unsure What country were you born in? 182 Appendix 4A: Chapter 5 Stimuli Manipulation Image credit: Aatma, Bravo. 2016. Flies taking food from around a baby’s mouth. https://www.idealismprevails.at/en/my-first-medical-camp-frightened-me/ Image credit: Centers for Disease Control/Castro, Martins & Georg, Lucille. 1965. https://phil.cdc.gov/details.aspx?pid=12156 183 Image credit: Al-Kamel, M. A. (2016). Impact of leishmaniasis in women: a practical review with an update on my ISD-supported initiative to combat leishmaniasis in Yemen (ELYP). International Journal of Women's Dermatology, 2(3), 93-101. (Fig. 1). https://www.ncbi.nlm.nih.gov/pmc/?term=28492018%5BPMID%5D&report=imagesdocsum 184 185 Appendix 4B: Chapter 5 Post-test Items Certainty Measures Using only ONE word, describe your response to what you just saw? How certain are you about the answer you just provided? Not at all certain Somewhat uncertain Neither certain nor uncertain Somewhat certain Completely certain How certain are you that this is an appropriate response to what you saw? Not at all certain Somewhat uncertain Neither certain nor uncertain Somewhat certain Completely certain Thinking back to your experience viewing the content, how certain were you of what you wanted to do in response to what you saw? Not at all certain Somewhat uncertain Neither certain nor uncertain Somewhat certain Completely certain 186 Social Judgment Task In terms of group traits, how COMPETENT (versus incompetent) would you rate each of the following groups compared to the average population. Extremely incompetent (1) to Extremely competent (5) In terms of group traits, how WARM (versus cold) would you rate each of the following groups compared to the average population. Extremely cold (1) to Extremely warm (5) elderly people children disabled people middle-class people educated people business executives rich people drug addicts welfare recipients refugees charity aid recipients people from 3rd world countries uneducated people Disgust Measures On a scale of 0 to 4, where 0 is "not at all" and 4 is "a great deal," indicate the extent to which you experienced the following while viewing the content. I felt sick I felt disturbed I felt repulsed I felt disgusted 187 It disgusted me I felt wretched I was grossed out I wanted to escape the situation I wanted to turn away I had a bad taste in my mouth A wave of nausea swept over me I felt a gag reflex My stomach churned Empathetic Concern On a scale of 0 to 6, where 0 is "not at all" and 6 is "extremely," indicate the extent to which you experienced the following while viewing the content. Sympathetic Softhearted Compassionate Moved Tender Warm Personal Distress On a scale of 0 to 6, where 0 is "not at all" and 6 is "extremely," indicate the extent to which you experienced the following while viewing the content. Upset Worried Disturbed Distressed Troubled Grieve 188 Psychological Escape Compared to other information you encountered today, how memorable is the content you just saw? Not at all memorable (1) to very memorable (7) To what extent do you anticipate thinking about the content you saw during the next 15 minutes? Not at all (1) to very much (9) To what extent do you anticipate thinking about the content you saw during the next 60 minutes? Not at all (1) to very much (9) How frequently do you anticipate thinking about this content over the next two days? 0 times; 1 – 3 times; 4 – 6 times; 7 – 10 times; 11 – 13 times; 14 or more times Which of the following statements best describes your experience? I did not have any negative reactions while viewing the content; Any negative feelings I experienced have already passed; Negative feelings should pass once I finish the study; I anticipate having some negative feelings about this for a few hours; I anticipate have some negative feelings about this for at least a day or more; I anticipate having some negative feelings about this for quite a while 189 Behavioral Measure The END Fund is a global network partnership dedicated to eradicating the most common tropical diseases in low-income countries. The END Fund is verified by GiveWell as a top-rated charity. Would you like to donate some of your earnings from this study to The END Fund to provide help for persons with neglected tropical diseases? All donations will go directly to the organization. Yes, I would like to donate part of the additional $5 I will receive for taking this study. No, I would like to complete the study without making a donation. IF YES… Please indicate how much of your additional $5 incentive you would like to donate. This amount will be deducted from your additional incentive earned. Thank you for making a donation to the END Fund! There are just a few more questions about you. Personal Relevance Were you aware of the issue of tropical diseases before participating in this study? Yes No In the past year, have you felt personally at risk for contracting a tropical disease? Yes No 190 Have you or anyone you know ever suffered from a tropical disease? Yes No Demographic Items Which gender do you identify with? Male Female Transgender Other Are you Hispanic, Latino/a/x or of Spanish origin? No, not of Hispanic, Latino/a/x or Spanish origin Yes, Mexican American, Chicano/a/x Yes, Puerto Rican Yes, Cuban Yes, of another Hispanic, Latino/a/x or Spanish origin What is your ethnic background? Select all that apply African American/Black American Indian/Alaska Native Asian/Asian American Caucasian/White Other How old are you today? 18 – 24 25 – 34 35 – 44 45 – 54 55 – 64 65+ What is the highest level of school you have completed? No schooling; Nursery to 8th grade; Some high school, no diploma; High school graduate, diploma or the equivalent; 191 Some college, no degree; College degree What is the highest degree you have obtained? Associate degree in college - Occupational/vocational/technical program; Associate degree in college -- Academic program; Bachelor's degree (BA, AB, BS, etc.); Graduate degree (MA, MS, MEng, MEd, MSW, MBA, etc.); Professional School Degree (MD, DDS, DVM, LLB, JD, etc.); Doctorate degree (PhD, EdD, etc.) Which one of these options best describes your employment situation last year? Working full time Working part time Temporarily not working Unemployed, laid off Retired Attending school full time Not formally employed (keeping house, raising kids, etc.) Other No answer What is the annual average income for your household (including all persons who live at your permanent address)? less than $20,000 $20,000 and $34,999 $35,000 - $49,999 $50,000 - $74,999 $75,000 - $99,999 $100,000 or more 192