PERCEPTIONS AND EMOTIONS ASSOCIATED WITH BROAD-BASED AUDIENCE CYBERBULLYING: IMPLICATIONS OF ONLINE COMMENTS FOR CYBERBULLIED  COPING  By   Jinsuk Kim                     A DISSERTATION  Submitted to  Michigan State University in partial fulfillment of the requirements  for the degree of   CommunicationDoctor of Philosophy  2016   ABSTRACT  PERCEPTIONS AND EMOTIONS ASSOCIATED WITH BROAD-BASED AUDIENCE CYBERBULLYING: IMPLICATIONS OF ONLINE COMMENTS FOR CYBERBULLIED  COPING  By   Jinsuk Kim  The presence of an invisible and broad-based audience is a unique feature of cyberbullying. This study focuses on perceptions and emotions associated with cyberbullying involving a broad-the es. This study examines (a) how cyberbullied e.g., the incident, the bully, or the audience), (b) perceived toward different targets, (c) audience judgments s about the victim, and (d) how empathy affects audience perceptions. Appraisal theories of emotion and the spotlight effect are used as theoretical frameworks. One hundred one young adults ages 18-25 years old participated in the online experimental survey with a 2 (perspective: victim vs. audience) × 2 (comment valence: positive vs. negative) x 2 (exposure to audience comments: pretest and posttest) mixed design. Participants read a hypothetical scenario of an outing incident, an involuntarily revealed sexual activity on YouTube, and responded to the questionnaire before and after reading other YouTube were either favorable or unfavorable about the victim. Results indicated that the bully and the audience evoked different negative emotional responses from victims of outing. Positive comments were associated with  and lessened the intensity of negative emotions perceived toward the audience. Negative comments were associated with negative reappraisals, but did not increase the intensity of negative emotions. A  test of the spotlight effect revealed that victims of outing tended to overestimate the negativity of audience judgments about the victim. Negative comments increased this perceptual discrepancy between the victim and the audience while positive comments minimized it. Empathy facilitated more positive audience perceptions of the victim. The current study discusses the important role of an audience and audience comments in cyberbullying coping.                            Copyright by JINSUK KIM 2016v             To Tyler Clementi            vi ACKNOWLEDGEMENTS  The completion of this dissertation would not have been possible without guidance, support, and encouragement of the following individuals. First of all, I give my deepest gratitude to my advisor, Dr. Kelly Morrison for being incredibly understanding and supportive throughout my doctoral studies. She has been an inspirational teacher, scholar, and mentor. I express my sincere gratitude to my dissertation committee members, Dr. Steve McCornack, Dr. Amanda Holmstrom, and Dr. Wei Peng for their insightful comments and suggestions that helped me improve my dissertation. I give thanks to Drs. Kathryn Dindia, Doshik Yun, Sang-Yeon Kim, Kiwon Seo, Nam Young Kim, Eunyoung Jang, Chang Lee, and soon-to-be Dr. Yumi Jung for their care, encouragement, and support that kept me motivated to finish my dissertation. I thank my writing support group members who filled me with positive energy every day. I give thanks to my family in Korea, especially my parents who have always believed in me and have supported me to pursue my passion. I give my special thanks to my husband, Dr. Jae-Wook Oh for his trust, love, encouragement, and support that helped me get through difficult times. Lastly, I thank my son, Joon for being a healthy distraction from my dissertation writing and for being the greatest motivation to complete this dissertation.     vii TABLE OF CONTENTS  .x  LIST OF FIGURESi  CHAPTER 1....1 INTRODUCTION..  CHAPTER 2....5   Characteristics, Clarifications, and Consequences of Cyberbullying Prevalence and characteristics of cyberbullying Definitional clarifications. Consequences of cyberbullying. Emotional Responses to Outing. Outing as a threat to privacy.. Appraisal theories of emotion. RQ1... The effects of online comments on emotional responses.. H1.19 RQ2... The Spotlight Effect in Outing.. The spotlight effect..............................22 H2.4 The moderation effect of online comments..4 H3.5 The effects of online comments on predicted audience judgments..... H4. Relationship between predicted audience judgments and negative emotions.. H5.8 The Influence of Online Comments and Empathy on Audience Perception... Peer influence on Audience judgments... H6.0 The role of empathy in audience judgments.....0 H7..  CHAPTER 3..34  METHOD.    Experimental Design. Procedure. Stimulus Materials....  Hypothetical scenario  Positive/Negative audience comments..7 viii Measures.  Predicted ... Actual audience judgment...8 Emotional responses.8 Reappraisal...9 Empathy...9  CHAPTER 4....  RESULTS. Preliminary Analysis... Induction Checks. .. RQ1: Emotional responses to the bully and the audience g and Emotions.. H1: The effects of comment valence on reappraisal.....46 RQ2: The effects of comment valence on negative emotions.. Composite negative emotion about the outing incident52 Composite negative emotion toward the bully...52 Composite negative emotion toward the audience........54 The Spotlight Effect in Outing.. H2: The spotlight effect in outing.... H3: The moderation effect of comment valence on the spotlight effect H4: The effects of comment valence on predicted audience judgments... H5: Relationship between predicted audience judgments and negative       emotions..59 The Influence of Online Comments and Empathy on Audience Perception... H6: The effects of comment valence on audience judgments H7: Relationship between empathy and audience judgments  CHAPTER 5....  DISCUSSION. Summary and Interpretation.  ... 6 The spotlight effect in outing.8 Peer influence on audience judgments...72 The role of empathy on audience judgments..73 Limitations Implications.. Future Research. Conclusion      ix APPENDICES  APPENDIX A: Hypothetical Scenario for Victim Condition83 APPENDIX B: Hypothetical Scenario for Audience Condition84 APPENDIX C: Stimulus Materials for Positive Comments Condition85  APPENDIX D: Stimulus Materials for Negative Comments Condition86  APPENDIX E: Measurement Items.....87  REFERENCES    x LIST OF TABLES  Table 1: Descriptive Statistics and t-test Results for Emotions toward the Bully and the Audience...  Table 2: Descriptive Statistics and t-test Results for Reappraisal by Comment Valence  Table 3: Descriptive Statistics and t-test Results for Perceived Anger, Fear, Sadness, and Disgust by Target in Positive Comment Condition.  Table 4: Descriptive Statistics and t-test Results for Perceived Anger, Fear, Sadness, and Disgust by Target in Negative Comment Condition.  Table 5: Means and Standard Deviations of Pre- and Post-Stimulus Judgment Ratings by   Table 6: Pearson Product-Moment Correlations for Predicted Negative Audience Judgment and Perceived Negative Emotions about the Outing Incident, toward the Bully, and toward   the Audience.....................................................................................................................60  Table 7: Pearson Product-Moment Correlations and Descriptive Statistics for Actual Audience Judgment and Empathy........................63      xi LIST OF FIGURES  Figure 1. Responses to Open-Ended Question about Emotions for Different Targets.  Emotional Responses by Target.  Figure 3. Effects of Comment Valence on Reapprai Figure 4. Changes in the Intensity of Perceived Anger, Fear, Sadness, and Disgust about the   Outing Incident by Comment Valence  Figure 5. Changes in the Intensity of Perceived Anger, Fear, Sadness, and Disgust toward the  Bully by Comment Valence  Figure 6. Changes in the Intensity of Perceived Anger, Fear, Sadness, and Disgust toward the  Audience by Comment Valence  Figure 7. Estimated Marginal Means of Perceived Negative Emotions about the Incident by   Comment Valence and Exposure to Audience Comments  Figure 8. Estimated Marginal Means of Perceived Negative Emotion toward the Bully by Comment Valence and Exposure to Audience Comments3  Figure 9. Estimated Marginal Means of Perceived Negative Emotion toward the Audience by Comment Valence and Exposure to Audience Comments  Figure 10. Estimated Marginal Means of Judgment Ratings by Perspective and Comment   Figure 11. Estimated Marginal Means of Predicted Audience Judgment Ratings by Comment   Valence and Exposure to Audience Comments.  Figure 12. Estimated Marginal Means of Actual Audience Judgment Ratings by Comment   Valence and Exposure to Audience Comments.  Figure 13. Hypothetical Scenario for Victim Condition...  Figure 14. Hypothetical Scenario for Audience Condition.  Figure 15. Stimulus Materials for Positive Comments Condition....  Figure 16. Stimulus Materials for Negative Comments Condition.  1 CHAPTER 1 INTRODUCTION On September 22, 2010, Tyler Clementi, an 18-year-old Rutgers University freshman, committed suicide after discovering that his roommate and another student had secretly filmed his same-sex sexual encounter and video-streamed it live on the Internet (Hu, 2010). On September 10, 2012, Audrie Pott, a 15-year-old high school girl, hanged herself after she was allegedly raped and pictures of the alleged assault were distributed online (Walsh, 2012). Days -year-old high school girl, committed suicide by hanging 17 months after being raped by four boys and a photo of the incident was widely spread online (Newton, 2013). These stories exemplify how the involuntary of this private information on the Internet can result in devastating consequences.  Many incidences of such involuntary disclosures by technological means have been reported; including an eighth-grade girl who e-mailed in confidence a video of herself engaging in sexual acts to a male classmate who then shared the video with a much broader audience on the Web; and a boyfriend forwarding a nude picture of his 15-year-old girlfriend, that had been taken with a camera phone, to his friends without the girlfri(Harmon, 2004). In 2014, over 100 celebrities, including famous film stars, became victims of the Internet by anonymous hackers (Duke, 2014). All of these examples constitute outing, involuntary disclosures of private information, which is a specific category of cyberbullying.  2 The bulk of previous cyberbullying research has primarily focused on the consequences of dyadic cyberbullying, when a bullying message is conveyed from one person directly, and privately, to a single, specific other person (e.g., a bully sending a threatening message to a target by email or text message). In contrast, very few studies on cyberbullying have examined a setting involving a broad-based, public audience. The presence of a broad-based and anonymous audience is a unique feature of cyberbullying, and the role of an audience in cyberbullying deserves more scholarly attention, especially given the devastating consequences described above. When private and potentially risky information is revealed to a broad-based audience Rehtaeh Parsons examples), the individual may experience a different set of perceptions and emotions associated with the incident because of the uncertainty related to characterizing and emotions associated with an audience because different perceptions and emotions may result in different psychological and behavioral consequences. Therefore, the first goal of this research is eptions and emotions associated with an audience that is perceived to be more invisible and broader based than a known dyadic bully.  As cyberbullying continues to be a serious social problem, both researchers and legislators continue to devote considerable attention to it. Much of the research has focused on identifying risk factors and establishing effective prevention and intervention strategies (Aboujaoude, Savage, Starcevic, & Salame, 2015; Guo, 2016; Kowalski, Giumetti, Schroeder, & Lattanner, 2014; Slonje, Smith, & Frisén, 2013; Tokunaga, 2010). For example, Kowalski et al. (2014) conducted a meta-analysis of 131 studies on cyberbullying and identified both risk and  3 protective factors. Risk factors for cyberbullying victimization included traditional bullying victimization, traditional bullying perpetration, anger, risky online behavior, frequency of Internet use, social anxiety, moral disengagement, and hyperactivity, while protective factors included school safety, school climate, social intelligence, perceived support, and parental monitoring. Also, legislative efforts have enforced laws against cyberbullying. As of January 2016, 48 states had laws against bullying using electronic means and 23 of those included the tchin, 2016). Such prevention-oriented research efforts and legislative movement may be effective in reducing incidences of cyberbullying and constitute important steps. However, given the preventative focus, these efforts are less useful in addressing outcomes, such as reducing the distress of those who already have been victims of cyberbullying. Thus, it is important for research to explore ways to help victims of cyberbullying cope with their psychological and emotional distress in order to prevent negative consequences.  The presence of an audience in cyberbullying plays an important role not only in intensifying these emotions, and in shaping the audience response. In an online world, the online commenting system facilitates audience involvement as critical commenter. The comments reveal a variety of responses to what has been witnessed, with some displaying empathy and others devoid of it. This audience response, in the form of online comments, may affect comments on cyberbullied victims because increased knowledge in this area should facilitate scholarly efforts to assist the coping of victims. Therefore, the second goal of this research is to and empathy for the victim.    4 In sum, the purpose of this study is twofold. First, this research aims to extend the current cyberbullying literature by examining a unique type of cyberbullying that occurs with an invisible, broad-associated with outing. Second, this research aims to provide the knowledge with which those who have been victims of cyberbullying can reduce their psychological and emotional distress by predicted audience judgments and emotional appments. In the following chapters, I first review the cyberbullying literature and describe characteristics and consequences of cyberbullying. I argue that the presence of a broad-based and anonymous audience is a unique feature of cyberbullying and discuss how broad-based audience ve well-being. I contend that judgments about the victim. The role of empathy on audience judgment also is discussed.      5 CHAPTER 2 REVIEW OF LITERATURE Characteristics, Clarifications, and Consequences of Cyberbullying In this chapter, I review the cyberbullying literature. Specifically, I will describe cyberbullying, its prevalence and the characteristics associated with it, highlight particular aspects of the definition that require clarification, distinguish between the differential effects associated with the bully versus the audience, and discuss consequences of cyberbullying. I begin the chapter by reviewing descriptions of cyberbullying.   Prevalence and characteristics of cyberbullying. Cyberbullying can be defined as  ). Cyberbullying is a pervasive social problem. A review of 51 journal articles published between 2002 and 2013 on prevalence rates for cyberbullying victimization found that cyberbullying victimization rates ranged widely from 2.3% to 72% across all of the studies with the average of 21.3% (Patchin & Hinduja, 2013). Patchin and Hinduja (2015) surveyed 15,347 teens for cyberbullying victimization rates from 2007 to 2015 in nine independent projects. The rates of lifetime cyberbullying victimization varied over the years ranging from 18.8% to 34.6%. On average, 29.4% of teens reported that they had experienced some form of cyberbullying, such as being threatened to hurt through a cell phone text message, having mean or hurtful comments/pictures/videos of themselves posted online, or having rumors about themselves spread online. Across all of the surveys, 9.4% of participants reported that they had been cyberbullied in the past 30 days. In 2015, 15% of participants reported that they had been cyberbullied in the past 30 days.   6 Several factors account for the pervasiveness of, and problems associated with, cyberbullying. With the prevalence of web-cameras, smartphones, and social media, an countless numbers of people on the Internet. Also, cyberbullying differs in fundamental ways from face-to-face bullying due to certain attributes of computer-mediated communication, such as availability and anonymity (McKenna & Bargh, 2000). Because the Web is accessible all the time, private information can be transferred 24 hours a day, seven days a week. Because the Web is characterized by anonymity, people can be extremely vicious without the fear of revenge (Suler, 2004). Thus, online channels, ranging from asynchronous text-based channels to synchronous media-rich channels, potentially facilitate cyberbullying.  Definitional clarifications. As the opening definition in this section illustrated, some researchers consider cyberbullying an extension of face-to-face bullying and simply employ the same characteristics of traditional forms of face-to-face bullying to define cyberbullying. From characteristics of traditional face-to-face bullying identified by Olweus (2003) as aggression or intentional harm-doing, repetitiveness, and the imbalance of power between bullies and victims.  Adopting the concept of face-to-face bullying can be useful in defining cyberbullying, but may not be comprehensive enough to capture the many ways that cyberbullying can occur. For example, verbal aggressions in the real world may occur in cyberspace via repeated e-mails or instant messaging. However, a one-time, rather than repeated, posting and forwarding of definition that requires repeated occurrences. And yet, this one-time event may inflict more harm  7 and lasting influence on the victim. Furthermore, the power imbalance is less pronounced in online settings than in face-to-face settings such that anyone can perform bullying acts to another person anonymously. Hence, bullying can occur between and across peers with equal levels of power. A more comprehensive list of cyberbullying actions has been suggested by Willard (2007). She identified eight categories of common cyberbullying actions (Willard, 2007, pp. 265-267). 1. Flaming  online fights using electronic messages with angry and vulgar language. 2. Harassment  repeatedly sending nasty, mean, and insulting messages. 3. Denigration  person to damage his or her reputation or friendships. 4. Impersonation  pretending to be someone else and sending or posting material to get  5. Outing  .  6. Trickery  talking someone into revealing secrets or embarrassing information, then sharing it online. 7. Exclusion  intentionally and cruelly excluding someone from an online group. 8. Cyberstalking  repeated, intense harassment and denigration that includes threats or creates significant fear.  All of these cyberbullying actions intend to insult or harm a target; however, some of these cyberbullying actions occur in dyads between a bully and a target while others involve third persons as an audience in the bullying process. Flaming, harassment, and cyberstalking occur between a bully and a target when a bullying action is directly conducted toward a target  8 and the target directly reacts to the bullying action. These bullying actions can be seen as an extension of face-to-face bullying in online settings.  Other types of cyberbullying actions involve third persons, in the form of an audience, in addition to a bully and a target. Denigration, impersonation, outing, trickery, and exclusion generate what I define as the bully effect and the audience effect. The bully effect refers to a perceived judgments of the victim by the observers. For example, if a person posts an embarrassing picture of another person to a webpage, an invisible and broad-based audience can about the self (the audience effect) and misbehavior (the bully effect). This type of cyberbullying may be more threatening and hurtful indirect threat, namely the fear of being judged by a broad-based, anonymous audience.  The presence of a broad-based audience is a unique feature of cyberbullying. Some may argue that there is an audience in face-to-face bullying; however, the size of the audience is limited in a face-to-face setting while the audience is infinite in cyberspace (Heirman & Walrave, 2008). Online communication typically is to multiple people whereas face-to-face communication typically occurs in dyads and small groups. Cyberbullying messages reach a great number of people as well as the target, and these hurtful messages can easily be copied, pasted, and spread in a very short period of time resulting in the audience effect. Thus, these broad-based and expedient bullying acts can be more enduring and hurtful to the victim when the threat is performed online rather than offline. For the purpose of this research, audience is defined as anyone who viewed a cyberbullying message whether they have intention to see the  9 message or not, including those who are exposed to the message by mindlessly clicking the accompanying link. Consequences of cyberbullying. A meta-analytic review on victimization in face-to-face settings shows that victimization is associated with psychological problems such as depression, loneliness, anxiety, and low self-esteem (Hawker & Boulton, 2000). Cyberbullying researchers argue that this research finding is replicated in online settings and that the consequences of cyberbullying are parallel to those of offline bullying (Kowalski, Limber, & Agatston, 2012). These claims are supported by empirical evidence. Ortega et al. (2012) found that anger was the most common emotion for both victims of traditional bullying and cyberbullying. Also, it has been reported that victims of cyberbullying experience emotional distress such as depression (Wang, Nansel, & Iannotti, 2011; Wright, 2016; Ybarra, 2004), social anxiety (Pabian & Vandebosch, 2016; Wang et al., 2011), low self-esteem (Patchin & Hinduja, 2010), and frustration, anger, and sadness (Patchin & Hinduja, 2006). While many researchers argue that consequences of cyberbullying are similar to those of face-to-face bullying; some researchers contend that cyberbullying can be more emotionally damaging to victims than face-to-face bullying because the characteristics of the Internet enable bullies to attack their targets anonymously and enable a broad-based audience to repeatedly view oung adolescents in grades six through ten found that victims of cyberbullying reported higher depression compared to victims of traditional bullying, including physical, verbal, and relational bullying. nalysis of a national census survey to high school students  10 revealed that victims of cyberbullying reported higher levels of depressive symptoms compared to victims of traditional bullying. In extreme cases, both cyberbullying and traditional bullying are associated with suicidal behaviors and cyberbullying is more strongly related to suicidal behaviors compared with traditional bullying (van Geel, Vedder, & Tanilon, 2014). Hinduja and Patchin (2010) conducted a survey of 1,963 sixth to eighth grade students to examine suicidal behaviors (i.e., ideation, attempts/experiences), among victims of traditional bullying and cyberbullying. They found that victims of cyberbullying tended to report more suicidal thoughts and were more likely to attempt suicide compared to those who did not have the experiences. Victims of traditional bullying were 1.7 times more likely to have attempted suicide than those who were not victims of traditional bullying. A similar tendency was found in cases of cyberbullying. Victims of cyberbullying were 1.9 times more likely to have attempted suicide than those who were not victims of that cyberbullying had modestly higher effects on self-harm and suicidal ideation than traditional bullying. Schneider et al. (2012) also found that cyberbullying victimization is even more strongly associated with suicidal ideation, suicide attempt, and self-injury compared with offline school victimization. The inchave made sharing and viewing personal images such as photographs and videos much easier tion (Rainie, Brenner, & Purcell, 2012). A national survey of 1,005 adults over the age of 18 conducted in 2012 reports that 46 % of Internet users posted original photos and videos that they had created online and 41 % of Internet users posted photos and videos that they found online on  11 imaging-sharing sites (Rainie et al., 2012). It is highly likely that the rate of photo sharing is much higher now since more photo sharing sites, such as Pinterest, Instagram, and Snapchat, have become more popular since the 2012 survey. Photos and videos that are created and distributed online often include private information such as displays of sexual behavior, substance use, and violence (Moreno, Parks, Zimmerman, Brito, & Christakis, 2009; Morgan, Snelson, & Elison-Bowers, 2010). This type of private and potentially risky information can be shared online and spread among online viewers psychological distress for the victim (Ybarra, Mitchell, Wolak, & Finkelhor, 2006). Research shows that cyberbullying via picture and video messaging has a greater of 360 Swedish adolescents examined which specific methods of cyberbullying had the greatest perceived impact on victims. They found that cyberbullying via picture/video messaging was perceived to have a greater negative impact on victims than cyberbullying by text message, email, or phone call. The reasons given by the participants were the large audience size, the concreteness effect (i.e., actually seeing the image), and the fear of not knowing who had seen British secondary school pupils found that photo/video clip bullying had a greater perceived negative impact on victims compared to other forms of cyberbullying via phone call, text message, email, instant messaging, Website, and chatroom.  Research on the efurther supports that some types of cyberbullying are more damaging than other types of cyberbullying. Straude-Müller, Hansen, and Voss (2012) studied emotional distress associated  12 with the most recent online victimization incidents of 4,498 survey participants. In their study, relational aggression (denigration, impersonation, outing, and trickery) and cyberstalking were less frequent than verbal and sexual harassment; however, relational aggression and cyberstalking were more emotionally distressing for victims than verbal and sexual harassment. Based on these research findings, it can be concluded that cyberbullying via picture and video messaging has a greater negative impact on victims than other forms of cyberbullying. Thus, the current study focuses on cyberbullying that involves a broader audience, specifically outing, spreading pictures and videos containing private information. Next, I examine the emotions and perceptions associated with this type of cyberbullying.  Emotional Responses to Outing In this section, I discuss how outing violates norms of self-disclosure and suggest emotional and cognitive responses to outing.  Outing as a threat to privacy. When an individual has intimate personal information self-image and self-concept are not made readily available to others (Altman & Taylor, 1973, p. 18). Every individual decides what information to disclose and how much he/she wants to disclose about himself/herself according to the setting and target. In fact, there is an optimal level of self-disclosure for a given situation and this optimal level of self-disclosure is correlated with mental health (Jourard, 1964, p. 15). Healthy disclosure varies in amount and depth from -disclosure is more intimate with a close friend than with a stranger. Sharing intimate information is appropriate between lovers while it normally would be inappropriate between strangers. For example, a person may  13 talk about intimate fear and sexual fantasy to his or her relational partner, but a person would not commonly share such intimate information with a stranger.  This norm of optimal self-disclosure in face-to-face settings holds true in cyberspace as well. It has been reported that MySpace and Facebook users adjust the visibility of their profile pages in order to avoid unwanted audiences (Tufekci, 2008). In social network sites, users can control privacy settings based on their preferences on what information to disclose and how much they want to disclose about themselves to whom. The privacy setting options ensure that the users achieve the optimal level of self-disclosure and stay within their comfort zone. The Facebook News Feed privacy outcry in 2006 illustrates how people react when they lose control over their personal information and the optimal level of self-disclosure cannot be maintained. On September 5, 2006, Facebook released a new News Feed feature which automatically displays new information that Facebook users post about themselves in their profile pages on the front page so that their in-network Facebook friends can see it on the front page. This new News Feed feature sparked a Facebook user privacy outcry and Facebook had to take down this feature until they were able to provide more privacy setting options to control visibility of personal information. Facebook users were upset because they lost control over their personal information and the information became more publicly visible than before without their consent (Hoadley, Xu, Lee, & Rosson, 2010). In outing, cyberbullies violate the norm of optimal self-disclosure by distributing their shared with strangers, on the Internet. This forced revealing of private information without experience negative emotions that could result in serious consequences such as depression and  14 suicide (Hinduja & Patchin, 2010). Given the emotional ramifications associated with outing, it makes sense to turn to the emotion literature as a guiding framework.   Appraisal theories of emotion. Appraisal theories of emotion may guide understanding of motional responses to outing. Appraisal theories contend that individuals experience certain emotions after they cognitively evaluate the impact of the situation in terms of how it would influence their personal well-being (Lazarus, 1991, 1993). Lazarus and Folkman environment that is appraised by the person as taxing or exceeding his or her resources and endangering his or her well-s depends on how people perceive a given event rather than the quality of the event. Emotional responses to a stressful event are the result of the interaction between stimulus and person. Lazarus and Folkman (1984) argued that people go through a two-stage process of cognitive appraisal: primary and secondary appraisals. The primary appraisal refers to an well-being) and motivational congruence (if an event is congruent with the person's desires and -being, no -being, either positive or negative  are -being and   15 The secondary appraisal refers to an evaluation of accountability (who or what is accountable for an event), perceived coping potential (if something can be done), and future expectancy (if the situation will change) (Smith & Lazarus, 1993). The secondary appraisal determines which specific emotion will be experienced. As a person evaluates who is accountable for a bad situation, what can be done to remedy the situation, and if the situation can change, different types of emotions may occur. For example, anger is felt when someone else is accountable for a bad situation while guilt is felt when the self is accountable for the situation (Smith & Lazarus, 1993). Fear is associated with danger or threat with low coping potential and sadness is associated with irrevocable loss with low coping potential and negative future expectations of a bad situation (Smith & Lazarus, 1993).  In the event of unwanted disclosure of private information, an individual first evaluates -being. In this primary appraisal stage, the person may experience a set of negative emotions such as anger, fear, and sadness. These emotions are felt as a result of appraisal, that is, cognitive evaluation of the event in terms of its relevance and congruence. Negative emotions should occur because of negative primary appraisals, namely, the outing is appraised to be personally relevant and incongruent with desired goals. However, in terms of secondary appraisals the negative emotion of anger also should be experienced because someone else is appraised to be accountable for the outing. These primary and secondary appraisals can happen simultaneously or in the opposite order. The secondary appraisal can happen first as a person figures out who is accountable for a bad situation and the person will feel anger if someone else is accountable for the bad situation. But, if the person -being as a result of the primary appraisal, the  16 person may not feel anger anymore. These primary and secondary cognitive appraisals should affect the level of stress, and emotions, that a victim of outing experiences.    When examining the consequences of cyberbullying, previous research tended to focus only on the bully effects on the victims (Kowalski et al., 2014; Tokunaga, 2010), but did not pay much attention to the influence of the invisible audience on the victims. According to appraisal audience (the audience effect) may not be the same because they should elicit different appraisals and emotions. Cyberbullying research identified feelings of frustration and anger as common emotions of the victim toward the bully (Ortega et al., 2012; Patchin & Hinduja, 2006). This follows from appraisal theories because the bully is appraised to be accountable for the cyberbullying. However, the victim may not feel frustrated or angry toward the audience because the audience is not accountable for the incident. Feelings of frustration and anger toward the bully could be intensified as the victim senses the presence of an audience. But, the victim may react differently toward the audience. The victim may fear social disapproval from the audience and loss of face (Goffman, 1958), and thus experience a different set of negative emotions or psychological responses, such as embarrassment, shame, or humiliation as well as sadness, due to what is seen as an irrevocable loss of public face, according to appraisal theories.  Although appraisal theories provide some theoretical foundation to explain different emotional responses of victims toward the bully versus the audience, it only explains certain types of emotions, not other types of emotional and psychological responses associated with the audience. Also, previous research does not provide a solid rationale to hypothesize emotional and psychological responses toward different targets. Therefore, this limitation in the previous research leads to the following research question.  17 RQ1. Arthe bully and the audience?  Thus far, I have discussed that emotions are results of cognitive appraisals and different emotions may occur toward different targets (e.g., the bully vs. the audience) in an outing incident, according to appraisal theories of emotion. Online social media often include spaces that allow an audience to comment on what they have read or observed online, essentially serving as critical commenters. In the following section, I consider the potential effects of this reappraisal of the outing incident and emotional responses associated with the outing incident, and further articulate how appraisal theories can be used to form hypotheses. The effects of online comments on emotional responses. Emotions are constantly changing as appraisals and reappraisals of the person-environment relationship change over time.  of new information from the environment, which may resist or nourish pressures on the person, and/or information from the that modifies an earlier appraisal of the coping.  specific external and/or internal demands that are appraised as taxing or exceeding the resources  18 resources that can be used to solve the problem. Coping strategies depend on the appraisal of coping potential, the belief whether something can be done to remedy the problem or not (Lazarus & Folkman, 1984). If the person thinks something can be done, the person is likely to use problem-focused coping strategies such as confrontation and rational argument to alter the person-environment relationship causing the distress. On the other hand, if the person thinks nothing can be done, the person is likely to use emotion-focused coping strategies such as self-controlling, seeking social support, and positive reappraisal to regulate the stressful emotions.  Emotion-focused coping can be an effective coping method for victims of outing. Once information has been shared online it cannot be permanently removed, even if the original posting is deleted from the host server, because it can be easily copied, saved, and transferred on the Internet. Also, even if the posting is removed, it can leave a strong, lasting impression on its viewers. Because outing is an irreversible and traumatic event for the victim, little can be done to fix the distressinregulating negative emotions rather than altering the situation. Previous research shows that positive reappraisal, among many other emotional coping strategies, can be a fa-being. Research found that positive appraisal is associated with positive outcomes. Folkman, Lazarus, Dunkel-Schetter, DeLongis, and Gruen (1986) found that careful problem solving and positive reappraisal are associated with satisfactory outcomes , while confrontative coping and distancing (e.g., refusing to get too serious about a bad situation) are associated with unsatisfactory outcomes (defined as . Martin and Dahlen (2005) examined the relationship between nine different cognitive coping strategies (self- 19 blame, other blame, rumination, catastrophizing, acceptance, putting into perspective, positive refocus, refocus on planning, and positive reappraisal) and negative emotions including depression, anxiety, stress, and anger in a negative event. They found that positive reappraisal was among the strongest predictors of anger, depression, anxiety, and stress. Specifically, positive reappraisal was inversely related to depression, anxiety, stress, and anger.  Positive or supportive messages may lead to positive reappraisal in a stressful event. Research shows that social support promotes cognitive appraisal and coping. Holt and Espelage (2007) found that perceived social support from friends and family helps victims of bullying cope. In their study, those who reported peer social support indicated the least anxiety and -bullying adjustment both behaviorally and psychologically. In particular, he found that positive reappraisal enhanced both -disclosure and long-term subjective well-being. Victims of outing may seek and utilize social support to cope with emotional and psychological distress. Feedback provided by an audience in response to what they have observed online may include encouraging or discouraging messages and this audience feedback positive or supportive comments to the victim, may lead to positive reappraisal, while negative audience responses to the victim may result in negative reappraisal. The next hypothesis is about cident. reappraisal such that positive audience comments are reappraised more positively than negative audience comments.  20 As previously discussed, the primary appraisal of an outing incident should elicit specifies which specific emotion is experienced in response to an outing incident based on perceived accountability, coping potential, and future expectancy. Therefore, discrete emotions associated with an outing incident can be identified. According to appraisal theories (Smith & Lazarus, 1993) and previous cyberbullying research (Ortega et al., 2012; Patchin & Hinduja, 2006), it is expected that victims of outing will experience anger, fear, and sadness. In addition to anger, fear, sadness, I include disgust in the analysis because these four emotions are among the six basic emotions suggested by Ekman, Friesen, and Ellsworth (1982). Because the current study focuses on negative emotions, the two other emotions of happiness and surprise are not included.  As victims of outing reappraise the outing incident influenced by the audience comments, eneral, negative audience comments are expected to intensify negative emotions while positive audience comments should weaken negative emotions. More specifically, positive audience comments are expected to lessen sadness and fear because these two emotions are associated with low coping potential and the positive audience feedback may increase perceived coping potential. It is also expected that anger may not be affected by the audience comments since perceived accountability is less likely to change (i.e., the person who committed the outing will be attributed to be the bully). However, it is not clear if this expected tendency will be observed toward different targets. As previously discussed, different targets should evoke different emotions such that the bully is responsible for the incident, thus eliciting anger, while the audience is unknown and amorphous, thus eliciting fear. Also, an outing incident as a target is an essential component to consider since how cyberbullied  21 victims feel about the outing incident overall should be different from how they feel about the bully and about the audience. Previous cyberbullying research has examined cyberbullied emotional responses to cyberbullying incidents in general rather than specifying the targets of their emotional responses (i.e., the incident, the bully, or the audience). Examining emotional responses to an outing incident is crucial since it provides a baseline to compare with the victims emotional responses to the bully and the audience and to other types of cyberbullying. Thus, the following research question is asked about the effects of the audience response to the outing incident. RQ2. How does the valence of audience comments affect the intensity of the negative emotions (anger, sadness, fear, and disgust) perceived (a) about the incident, (b) toward the bully, and (c) toward the audience? Thus far, I have discussed the phenomenon of cyberbullying, focusing on the incidence of outing, and offered a framework based on appraisal theories of emotion. I discussed that perception of audience judgments on the von the spotlight effect. The Spotlight Effect in Outing In the previous section, I discussed that emotions are felt as a result of cognitive about the cognitive appraisals associated with the audience can be explained by the spotlight effect.  22 The spotlight effect. their actions and appearance are more likely to be noticed, judged, and remembered by others because people are fundamentally egocentric and have difficulty going beyond their own perspectives when predicting how they will be noticed, judged, and remembered by others (Gilovich et al., 2000; Gilovich & Savitsky, 1999; Griffin & Ross, 1991). Because people focus egocentrically on the failure or the blunder itself and neglect other factors that might affect restimate how harshly they will be judged after a failure or a blunder (Savitsky, Epley, & Gilovich, 2001).  Experimental research shows how the spotlight effect takes place in various contexts regardless of whether it is a positive or negative experience. Gilovich et al. (2000) conducted two experiments to test the spotlight effect in a situation where participants wear a t-shirt depicting an embarrassing or non-embarrassing celebrity. In their study, participants were asked to wear an embarrassing t-shirt depicting a celebrity, Barry Manilow, in front of others. They were asked to estimate the percentage of people who could identify the celebrity on the t-shirt. Their prediction was significantly higher than the actual percentage of observers who were able to identify the celebrity on the t-shirt. A similar tendency was found for a non-embarrassing condition. In another study, participants were able to choose and wear a t-shirt depicting one of three celebrities, Bob Marley, Jerry Seinfeld, or Martin Luther King, Jr., with whom they felt pleased to be associated. Participants who were wearing the non-embarrassing t-shirt overestimated the salience of their attire as well. Gilovich et al. (2002) further tested the spotlight effect for physical appearance by cratings on their physical attractiveness in five seminars throughout the semester. The results  23 showed that students expected that their classmates would notice changes in their appearance more than actually was the case. The spotlight effect was found not only for attire and physical appearance, but also for behavioral acts (Gilovich et al., 2000; Gilovich et al., 2002). Gilovich et al. (2000) tested the spotlight effect in the context of group discussions. In this study, participants were asked to engage in a group discussion and estimate how other group members would rank them on positive (e.g., advancing the discussion) and negative (e.g., speech errors, offensive comments) dimensions. The results showed that participants overestimated the salience of their own behavior to the other members of the group for both positive and negative contributions during d in the contexts of athletic performance and video game performance (Gilovich et al., 2002).  The spotlight effect also exists when predicting judgments of others in an embarrassing ed accidentally tripping a security alarm in the library in the presence of others, they thought they would be judged more negatively than they actually were. The same results occurred for other embarrassing incidents, including showing up at a party without a gift, being spotted by classmates when carrying a plastic bag with a low-level brand name printed on it, and failing a particularly difficult test in front of others (Savitsky et al., 2001). These findings collectively  judgments are exaggerated in an embarrassing situation.  Research has also found that the level of social-evaluation influences the spotlight effect.  or a low social-evaluative condition. In the high socially evaluative condition, participants were informed that their performance would be videotaped and evaluated by a group of  24 communication experts. In the low social-evaluative condition, participants were informed that their performances would be audiotaped and evaluated by the experimenter. Results showed that the difference between the predicted and actual public self-awareness scores was greater in the high social-evaluative condition compared to the low social-evaluative condition. Haikal and participants in both conditions were informed that their introductory speech would be videotaped and they were presented with visual feedback of how they would be videotaped. Participants in --in images of themselves while participants in the low social evaluative condition were shown images of themselves zoomed-ous, this study also found that participants in the high social evaluation condition exhibited greater spotlight effect than those in the low social evaluation condition. The spotlight effect should be triggered by an outing experience. Outing is an embarrassing and highly socially-evaluative event for the victim. It is expected that when a ative judgments about the victim. Based on the previous reasoning, it can be hypothesized that when private information about self is revealed on the internet, the  is posed. H2: Victims of outing overestimate the negativity of audience judgments such that their predicted audience judgments are more negative than the actual audience judgments.   The moderation effect of online comments. As previously discussed, the spotlight effect  25 (Savitsky et al., 2001). Whether this tendency persists when they have certain references, such as retically intriguing question. Conceivably, audience predicted and the audimake the situation more embarrassing to the victim while positive comments should make the same situation as less embarrassing. This audience feedback is expected to have a greater influence on the victim than on the audience because the audience comments, including judgments about the victim, are more personally, or motivationally, relevant to the victim than to the audience. Therefore, when the victim and the audience of outing view negative audience cted audience judgments will be greater than the actual audience judgments. On the other hand, when they view positive audience comments, the difference between the predicted audience judgments and the actual audience judgments will be diminished. It is expected that the valence of audience comments moderates the relationship will be more pronounced in the negative comment condition compared to the positive comment condition. Based on this reasoning, the following hypothesis is posed. H3. The spotlight effect is greater in the negative comment condition compared to the positive comment condition. Thus far, I the victim based on the spotlight effect. I have also argued that the valence of audience comments affects the relationship between the predicted and the actual judgments in outing. In  26 order to fully understand how the valence of audience comments affects the relationship between the predicted and the actual judgments in outing, it is necessary to examine its effects on each of can be used to form a hypothesis about  The effect of online comments on predicted audience judgments. Audience comments play an important role for the victims of outing to learn about what others actually think of them, or perception-check, and they should judgments. This audience feedback provides new information that may lead victims of outing to reappraise the outing incident. According to appraisal theories (Lazarus, 1991, 1993), the primary appraisal evaluates the motivational relevance and congruence of information, in this case, the audience comments. Audience feedback that includes their judgments about the victim should be interpreted as -being. Whether this audience feedback is congruent with the victim in positive or negative perceptions. If the positive perceptions should occdesires or goals, negative perceptions should occur.  The valence of audience comments is expected to have a significant impact on the incident even more negatively  victim perceive the same situation less negatively. Furthermore, the  27 other hand, if the victims receive positive feedback from the audience and realize that others do not think as harshly as they expected, this should effects of audience comment valence on predicted audience judgments. audience judgments such that (a) negative comments increase the negativity of the predicted audience judgments about the victim while (b) positive comments increase the positivity of the predicted audience judgments about the victim. Relationship between predicted audience judgments and negative emotions. Misperception refers to the gap between misperceptions can be an effective way to promote positive appraisals for victims of outing, thus ultimately leading to more positive emotional outcomes and enhanced coping. In the case of outing, as previouslthe victim, according to the spotlight effect (Savitsky et al., 2001). If the victim corrects this s is rs would think and say about the victim. As previously discussed, emotional responses result from cognitive appraisals of an event. Negative appraisals of an incident lead to negative emotional responses while positive appraisals lead to positive emotional responses according to appraisal theories (Lazarus &  28 Folkman, 1984). Therefore, it is expected that negative perceptions of audience judgments produce negative emotional responses. Though much research has been conducted on the emotional consequences of cyberbullying victimization, no study has directly examined how s can be clarified by considering research on a closely-related construct, fear of negative evaluation, and how it is related to emotional responses.  over their negative evaluations, avoidance of evaluative situations, and the expectation that evaluation has been studied as a core feature of social anxiety. Research shows that individuals with high levels of fear of negative evaluation tend to have high levels of social anxiety (Haikal negative judgments leads to a negative emotional consequence. This pattern should apply to the current research. judgments are closely associated with negative emotional responses. Based on the previous reasoning, the next hypothesis is posed about the relationship between predicted audience judgments and emotions.  H5. Negative audience judgments perceived by victims of outing are positively correlated with their perceived negative emotions.  To this point, I have focused on the cognitive and emotional responses of the victim. In the following section, I turn my focus to the online audience, their comments, and the role of empathy.  29 The Influence of Online Comments and Empathy on Audience Perception In the previous section, I have discussed the effects of online comments on cyberbullied  processes, I discuss peer influence on audience judgments and consider the role of empathy. Peer influence on audience judgments. Audience judgment of a cyberbullied victim may be influenced by a yed in the section allowed for viewer comments below the online video. In particular, the valence of comments by others may affect audience perceptions about the victim portrayed in the video. Previous research demonstrated that peer influence was found in evaluating and making judgments of a target in a variety of online settings. Edwards, Edwards, Qing, and Wahl (2007) tested the effects of online professor evaluations on student perceptions and attitudes in an instructor rating site (RateMyProfessor.com). Participants in the study watched a videotaped lecture of a professor with either a positive or negative set of comments included about the professor. Those who read positive comments about the professor reported higher ratings on instructor attractiveness and credibility, and scored higher on attitudes about the course and motivation to learn compared to those who read negative comments. support for the influence of online peers on perceptions of a person in a social network site. They found that perceptions of Facebook profiles were influenced by the comments and pictures posted by others. In this affected ratings of physical, social, and task attractiveness of the Facebook profile owner. In  30 another study, Walther, DeAndrea, Kim, and Anthony (2010) found that the valence of comments about anti-marijuana public service announcement videos on YouTube significantly affected the evaluations of the videos. Reading positive comments about the video led the audience to form more favorable evaluations about the video while reading negative comments led them to more negative appraisals of the video. Ballantine, Lin, and Veer (2015) studied how the valence of comments appearing with a relationship status update on Facebook affected favorable attitudes toward the status update, while negative comments led to less favorable attitudes, regardless of the actual status update.  Based on the previous research findings, it can be concluded that the presence of audience comments has significant effects on judgments of a person. Particularly, the valence of seeing negative comments about the victim should lead an audience to generate more negative evaluations about the victim, whereas seeing positive comments should lead to less negative appraisals of the victim. Based on this reasoning, the following hypothesis is posed. H6. The valence of comments about a victim of outing affects the audience judgments such that (a) negative comments increase the negativity of the audience judgments about the victim while (b) positive comments increase the positivity of the audience judgments about the victim. The role of empathy in audience judgments. Empathy is a factor that also may affect the  a victim of outing. Empathy is defined as the ability to understand and share the feelings of others. Empathy is viewed as a multidimensional construct that has both cognitive and affective (emotional) components (Davis, 1994). Cognitive empathy, or  31 perspective whereas affective empathy, or empathic concern, refers to the ability to experience feelings of others (Davis, 1994).  Previous research has revealed a negative relationship between empathy and antisocial behavior. Miller and Eisenberg (1988) conducted a meta-analysis of 43 studies examining the relationship between affective empathy and aggression/antisocial behavior. They found a low-to-moderate negative relationship between affective empathy and aggression/antisocial behavior (z+ = -.06 to -.46). However, this study was limited by its definition of empathy such that it only included affective empathy and ignored cognitive empathy. A more recent meta-analysis by Jolliffe and Farrington (2004) included both cognitive and affective empathy in examining the relationship between empathy and criminal offending and confirmed the negative relationship between empathy and antisocial behavior. They summarized a total of 35 relevant studies, 21 studies on cognitive empathy and 14 studies on affective empathy. The results indicated an effect size of d = -.48 for cognitive empathy and an effect size of d = -.11 for affective empathy; meaning that cognitive empathy has a stronger negative relationship with aggression than affective empathy.  Any type of bullying, whether it is online or offline, is considered an aggressive and antisocial behavior; therefore, bully acts should be negatively associated with empathy. Gini, Albiero, Benelli, and Altoe (2007) found that empathy, defined as a combination of perspective SEM analysis showed that higher levels of empathy were associated with lower engagement in bullying behavior (standardized coefficient = - 32 behavior toward victimized school mates (standardized coefficient = .29) among Italian adolescent boys.  Research on cyberbullying also has found a negative relationship between empathy and empathic responsiveness characterized cyberbullies. In their study, cyberbullies scored higher on the lack of empathy scale than victims of cyberbullying and those who were non-involved (i.e., neither as a cyberbully nor as a victim). Ang and Goh (2010) found that cognitive and affective empathy were negatively associated with cyberbullying acts. They conducted a survey of Singaporean adolescents to study the interactions between cognitive and affective empathy, and gender on cyberbullying behavior. They found that at low affective empathy, those who had low cognitive empathy had higher scores on the prevalence and frequency of cyberbullying acts than those who had high cognitive empathy. The similar effect of cognitive empathy was found at high affective empathy for boys; however, cognitive empathy did not affect cyberbullying behavior for girls at high affective empathy. More recently, Del Rey et al. (2016) also found that both cognitive and affective empathy negatively predicted traditional bullying and cyberbullying perpetration, but the effect of empathy on bullying and cyberbullying was invariant across gender, age, and nationality.  Empathy has been found to be a strong predictor of audience behavior in cyberbullying. Van Cleemput, Vandebosch, and Pabian (2014) found that adolescents with high levels of empathic concern or affective empathy were more likely to help victims of cyberbullying, while adolescents with low levels of empathic concern were more likely to join in the bullying or remain passive. Shultz, Heilman, and Hart (2014) defined empathy more broadly and included other dimensions of empathy (perspective taking, fantasy, personal distress) in addition to  33 affective empathy. In their study, they showed research participants a fictitious Facebook conversation portraying several types of cyberbullying acts and analyzed their open-ended comments to the conversation. They found that individuals with higher levels of empathy more frequently identified with a cyberbullied victim and posted more supportive comments for the victim compared to individuals with lower levels of empathy.  People adjust their impressions of others when they can empathize with an actor or imagine being in his or her situation (Regan & Totten, 1975). Empathic persons are less likely to make harsh dispositional inferences when another person is in an embarrassing situation. Observers who can empathize with the cyberbullied victim should soften their assessments of the victim. Thus, it is expected that empathy is associated with positive audience judgments about the victim of outing. Based on the previous reasoning, the following hypothesis is posed. H7: Empathy is positively associated with positive audience judgments about the victim of outing.     34 CHAPTER 3 METHOD Participants Participants for this study were recruited from Turk), where participants receive monetary compensation for completing tasks. MTurk samples are more diverse and nationally representative than typical college student samples (Buhrmester, Kwang, & Gosling, 2011). U.S. residents between ages 18 and 25 were solicited for online survey research participation. This age group was chosen because young adolescents and young adults are at high risk for potential cyberbullying victimization. Participants received $1 for completing an online survey questionnaire. Data were collected and saved using Qualtrics.com, an online survey platform. For effect size 2 = .15, with a = 2 groups, a power of .80 is obtained with n = 24 according to the GPOWER, a computer program that calculates sample size (Erdfelder, Faul, & Buchner, 1996). Therefore, the target sample size was 96. A total of 128 participants completed the survey questionnaire for 10 weeks. Sixteen incomplete questionnaires were excluded when analyzing the data. Although the target age was specified in the advertisement for recruitment, some participants were outside of this age range. Ten questionnaires were excluded for not meeting the age requirement. An extreme outlier was excluded from the analysis. As a result, 101 young adults between ages 18 and 25 were included in the analysis. The sample of participants was 62.4% male and the average age was 22.60 years old, with a range from 18 to 25 years old (SD = 1.59). The majority of the participants were Caucasian (79.2 %), with 6.9 % Asian, 6.9% Latino/Hispanic, 5.9% African American, 2% Native American, and 1% other. About half of the participants were currently attending or enrolled in school (44.6%).  35 Experimental Design An online experimental survey was administered to test the hypotheses with a 2 (perspective: victim or audience) x 2 (valence of comments: positive or negative) x 2 (exposure to audience comments: pretest and posttest) design with the first two factors as between-subjects factors and the last factor a within-subjects factor. Participants were randomly assigned to one of ts condition (n = 24); (b) a n with positive comments condition (n comments condition (n = 28). Procedure Upon entering the survey website, participants were asked to consider a hypothetical scenario where sexual activity was involuntarily videotaped and posted online from either a ons (pretest). Then, participants were presented with either positive or negative comments regarding the person in the video and completed the measurements again (posttest). More specifically, in the victim condition, participants read a script about sexual intercourse being filmed and shared on YouTube. Participants were instructed to imagine how they would feel if this happened to them. After reading the script, participants completed measures for their predictions of audience judgments and their emotional responses to different targets. The emotional responses were measured for each of three different targets (the bully, the audience, and the incident) repeatedly. For possible order effects, the three sets of emotional response measures were presented using randomized order. After completing these measures, participants were shown screenshots of either positive or negative comments made by  36 anonymous others below the video on a YouTube page. After reading the comments, participants completed measures for predictions of audience judgments and again responded to the three sets of emotional responses toward three different targets. Additionally, emotional reappraisal was measured. In the audience condition, participants read a modified version of the script written from YouTube page. After reading the script, participants completed measures of judgment and empathy toward the person in the video. Then, they were shown screenshots of either positive or negative comments made by anonymous others below the video on a YouTube page. After reading the comments, they completed measures of judgment and empathy about the person in the scenario again. Stimulus Materials Hypothetical scenario. A scenario-based script was chosen instead of using an actual image (a photo or a video of the victim) because some attributes in an image, such as gender or elves in the situation. A hypothetical scenario of an outing incident was written by the researcher based on true stories. It describes a situation of receiving an e-mail that includes a hyperlink leading to a YouTube video portraying that someone was naked and having sex with a partner without noticing that the sexual intercourse was filmed and posted on YouTube.com, the largest video-sharing site on the Internet. The actor of the hypothetical scenario was modified according to the perspective of a victim scenario are displayed in Appendices A and B.  37 Positive/Negative audience comments. The researcher created positive and negative audience comments to the victim based on actual comments on cyberbullying-related videos on YouTube. Comments were written using gender neutral terms with similar word counts, and written to be equivalent but varying in positive versus negative valence g  nly one who hates this person. . Additionally, o in each set of comments to enhance realism. The comments are displayed in Appendices C and D. Measures Participants in the victim condition responded to measures of predicted audience judgment about the victim, perceived emotions, and reappraisal. Participants in the audience condition responded to measures of actual audience judgment about the victim and empathy toward the victim. A detailed description of each of these measures  is provided below. A complete list of measurement items is included in Appendix E.   Predicted audience judgment.  judgment of an audience about the victim was measured by nine items developed for this study. These items were developed based on previous research on the spotlight effect (Savistky et al., 2001). perspective indicated on a 7-point scale the extent to which viewers of the video depicted in the scenario would form each of the following impressions of them  that people who watched this video would form _ Predicted judgment items included the following semantic differential scales: negative-positive, unpleasant-pleasant,  38 bad-good, unattractive-attractive, disagreeable-agreeable, unfavorable-favorable, undesirable-desirable, worthless-worthy, and shameful-admirable    Actual audience judgment. Actual judgment of an audience was measured by the nine items used to measure  audience judgment. Participants in the audience perspective indicated on a 7-point scale the extent to which they form the following impressions of the person depicted in the scenario to the stat impressions of  The impression items included the following semantic differential scales: negative-positive, unpleasant-pleasant, bad-good, unattractive-attractive, disagreeable-agreeable, unfavorable-favorable, undesirable-desirable, worthless-worthy, and shameful-admirable   Emotional responses. Emotion items were adopted from previous research that tested and validated discrete emotion measures (Dillard & Peck, 2001; Shen & Dillard, 2007). Emotional reactions to different targets (the incident, the bully, and the audience) were assessed for each target using a 5-point response scale, ranging from 0 (none of this feeling) to 4 (a great deal of this feeling). The scales and their corresponding items for emotional reactions to the incident were: anger (irritated, fear (fearful, afraid, and scared; sadness disgust = .91), happiness An overall negative emotion index was created by combining all of these emotion measures but excluding the happiness items (81). In addition to these 16 items, participants were asked to indicate on the 5-point scale if they experienced any of the following seven emotional and psychological responses adopted from previous cyberbullying research (Ortega et al., 2012; Patchin & Hinduja, 2006): depressed, stressed, anxious, frustrated, ashamed, humiliated, and embarrassed. Thus, a total of 23  39 emotional and psychological response items were measured repeatedly for each of three different targets (toward the bully, the audience, and about the incident) before and after reading the audience comments. Along with these closed-ended emotions items, participants were asked an open-ended question to list any other emotions they had about the incident, the bully, and the audience.  Reappraisal. Reappraisal was assessed with  (2015) reappraisal scale. This measure consists of three dimensions: Emotion-focused coping potential, problem-focused coping potential, and future expectancy. It includes six emotion-focused coping potential items (help me to feel better view this situation more positively, ), five problem-focused coping potential items (I am capable of dealing with this situation I could do what is needed to manage this situation, 8), and three future expectancy items (This situation will improvethis situation will get better, 9) in response to the audience comments.   Empathy. To measure empathy toward the victim, a total of nine items (four perspective taking items and five empathic concern items) were adopted from the 28-item Interpersonal Reactivity Index (IRI; Davis, 1980). The IRI is a multi-dimensional measure that assesses perspective taking, empathic concern, fantasy, and personal distress. Some researchers argue that the perspective taking and empathic concern scales directly correspond to the conceptual definition of empathy whereas the fantasy and personal distress scales do not (Burkard & Knox, 2004). For this reason, in this study, the two dimensions of perspective taking and empathic concern were used to assess empathy. The perspective taking scale measures cognitive empathy measures affective empathy (the ability to experience feelings of others).   40  Participants were asked to indicate how well each item described them on a 5-point scale, from 0 (does not describe me well) to 4 (describes me very well) Two reverse-coded items were dropped because of low reliability. Four empathic concern items and three perspective taking items were used in the analysis. Affective empathy items included the : I have tender, concerned feelings for the person; I feel very sorry for the person; I feel kind of protective toward the person; I feel very much pity for the person (). Cognitive empathy items included: I  I try to   .     41 CHAPTER 4 RESULTS Preliminary Analysis Several measurement items were included to assess the perceptions of the outing scenario in terms of the sensitivity of the outed information, the realism of the scenario, and the intentionality of the disclosure. The perceptions of the sensitivity of the outed information were measured with six Likert-type items, ranging from 1 (strongly disagree) to 7 (strongly agree). Participants rated whether the information in the video was private, sensitive, risky, intimate, personal, and confidential. The data indicated that participants perceived the information revealed in the video to be highly sensitive (M = 6.42, SD            The perceived realism of the scenario also was measured with six Likert-type items. Participants were asked to rate the realism of the hypothetical scenario with the following statements on a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree): A situation like this could happen to someone; A scenario like this could potentially occur to anyone; This incident could realistically happen; It is possible for an event like this to happen to someone; It is probable that something like this could happen to someone; It is conceivable that an event like this could take place. The data indicated that participants perceived the hypothetical scenario to be realistic (M = 5.52, SD           The perceived intention of disclosure was measured with five Likert-type items. Participants were asked to rate if the video was shared voluntarily on a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree), with items that included the following The video was shared with your consent (the consent of the person in the video); You (the person in the video) voluntarily  42 shared the video; You (the person in the video) had intention to share the video; You (the person in the video) were aware that the video would be seen by other people; You (the person in the video) wanted other people to see the video. The data indicated that participants perceived that the person portrayed in the video did not intend for the video to be shared (M = 1.50, SD = .92). Induction Checks          To check the valence of audience comments induction, a scale was created consisting of seven 7-point semantic differential items. Participants were asked to assess the audience comments with the following word pairs: negative-positive, discouraging-encouraging, unsupportive-supportive, unkind-kind, not helpful-helpful, discomforting-comforting, and unfavorable-indicated more positive evaluations. The results indicated that participants who read positive comments reported significantly higher ratings of positivity (M = 5.61, SD = 1.43, n = 48) compared to the participants who read negative comments (M = 1.67, SD = .97, n = 53), t(99) = 16.30, p < .001, d = 3.26. This tendency was also found regardless of whether the participant was in the victim or audience perspective. Thus, the results indicated that the valence of audience comments induction was successful.  Additionally, the perceived realism of the audience comments was assessed with six 7-point semantic differential items. Participants were asked to make judgments about the realism of the comments with the following word pairs: unrealistic-realistic, impossible-possible, unlikely-likely, implausible-plausible, untruthful-truthful, and unbelievable-believable (M = 4.74, SD = indicated that the participants perceived the audience comments as more realistic. Examining the entire sample revealed that the participants who read positive comments reported significantly higher ratings of perceived realism (M = 5.12, SD = 1.30, n =  43 48) compared to participants who read negative comments (M = 4.35, SD = 1.59, n = 53), t(99) = 2.65, p = .009, d = .53separately, the perceived realism of comments differed in the audience condition, but not in the victim condition. In the audience condition, those who read positive comments reported higher ratings of perceived realism (M = 5.31, SD = 1.09, n = 24) than those who read negative comments (M = 3.99, SD = 1.45, n = 28), t(50) = 11.17, p < .001, d = 1.02. In the victim condition, no statistically significant difference was found in the perceived realism of the positive and negative comments, t(47) = .39, p = .70. Victim Responses to Outing RQ1: Emotional responses to the bully and the audience. The first research question asked if bully and the audience. The McNemar test and paired sample t-tests were used to answer this research question. The , a statistical procedure for analyzing related dichotomous variables, was used to test the null hypothesis that the probability of feeling a type of emotion toward the bully but not toward the audience is equal to the probability of feeling that emotion toward the audience but not toward the bully. Each emotion item originally measured with 5-point scales was recoded to dichotomous scales to indicate the presence or absence of each emotion. The original response of zero was coded as zero and responses one through four were recoded as one (0= no, 1= yes). A series of McNemar tests were performed to determine if there are differences in the frequencies of each of the 23 emotional and psychological responses that the victims experienced toward the audience and the bully. Results indicated that participants in the victim emotions toward the bully and the audience. Statistically significant differences were found in depressed (p = .039), humiliated (p  44 = .031), and embarrassed (p = .016). These psychological responses were felt more toward the audience than toward the bully. No statistically significant tendency was found for the other emotions. Additionally, participants were asked an open-ended question to list any other emotions they would have about the incident, the bully, and the audience. Responses from seven participants indicated that (a) when they think about the incident they would feel exploited, suicidal, nervous, paranoid, concerned, worried, hurt, betrayed, and violated; (b) when they think about the bully they would feel enraged, murderous, paranoid, vengeful, concerned, worried, violated, and hurt; and (c) when they think about the audience they would feel confused, vengeful, paranoid, concerned, worried, violated, invaded, and suicidal. Figure 1 illustrates -ended question about emotions perceived for different targets. Figure 1. Responses to Open-Ended Question about Emotions for Different Targets  Note. This graphical representation is based on responses from seven participants (n = 7). Two parti the bully. All other responses were reported once for each target.  45 Next, a series of paired-sample t-tests were computed to compare the intensity of each emotion perceived toward the bully and the audience. Results indicated that more anger was felt toward the bully (M = 3.49, SD = 0.89) than toward the audience (M = 2.84, SD = 1.20), t(48) = 4.14, p < .001, d = .59; more disgust toward the bully (M = 3.41, SD = 1.06) than toward the audience (M = 2.68, SD = 1.32), t(48) = 3.11, p = .003, d = .44; and more frustration toward the bully (M = 3.43, SD = 1.10) than toward the audience (M = 2.92, SD = 1.30), t(48) = 3.42, p = .001, d = .49. Participants reported that they felt more depressed toward the audience (M = 2.53, SD = 1.37) than toward the bully (M = 2.12, SD = 1.59), t(48) = -2.19, p = .034, d = -.31; embarrassed toward the audience (M = 3.18, SD = 1.13) than toward the bully (M = 2.73, SD =  1.55), t(48) = -2.29, p = .03, d = -.33. No statistically significant difference was found for other emotions. See Table 1 and Figure 2 for details. Table 1 Descriptive Statistics and t-test Results for Emotions toward the Bully and the Audience  Perceived Emotion Bully  Audience  95% CI for Mean Difference     M SD  M SD  Lower Upper  r t ANGER 3.49 0.89  2.84 1.20   0.33  0.96  .48***      4.14*** FEAR 2.07 1.53  2.25 1.49  -0.58  0.21  .59*** -.94 SADNESS 2.19 1.35  2.33 1.27  -0.46  0.19  .63*** -.85 HAPPINESS 0.29 0.67  0.29 0.68  -0.15  0.17  .65***  .09 DISGUST 3.20 0.98  2.68 1.32   0.18  0.85  .52***    3.11** Depressed 2.12 1.59  2.53 1.37  -0.78 -0.03  .62*** -2.19* Stressed 3.12 1.33  3.12 1.18  -0.42  0.42  .32*  .00 Anxious 2.69 1.47  2.76 1.41  -0.48  0.36  .49***       -.29 Frustrated 3.43 1.10  2.92 1.30   0.21  0.81  .64***     3.42** Ashamed 2.47 1.52  2.78 1.33  -0.67  0.06  .60***      -1.68 Humiliated 2.76 1.57  3.06 1.23  -0.69  0.08  .57***      -1.60 Embarrassed 2.73 1.55  3.18 1.13  -0.84 -0.06  .52***      -2.29* Note. The first five emotions were measured with multiple items while other emotions were measured with single items. n = 49. df = 48. * p < .05. ** p < .01. *** p < .001.    46 Figure 2. Emotional Responses by Target  Note. Emotion items with a statistically significant difference between emotions toward the bully and the audience were marked with asterisk (*).  To summarize, out of the 12 emotional responses examined, anger, disgust, and frustration were felt stronger toward the bully and depression and embarrassment were felt stronger toward the audience.  The Impact of Online Comments on Victim Coping and Emotions H1: The effects of comment valence on reappraisal. The first hypothesis predicted that such that positive audience comments are reappraised more positively than negative audience comments. An independent sample t-test was computed to compare problem-focused coping potential, emotion-focused coping potential, and future expectancy scores between positive and negative comments conditions from The results indicated that 0.00.51.01.52.02.53.03.54.0Degrees of IntensityBullyAudienceOuting 47 participants who read positive comments reported higher scores in emotion-focused coping potential (M = 5.12, SD = 1.15) than those who read negative comments (M = 2.07, SD = 1.03), t(47) = - 9.83, p < .001, d = -2.80. Participants who read positive comments reported higher scores in problem-focused coping potential (M = 5.61, SD = .94) than those who read negative comments (M = 3.13, SD = 1.62), t(38.88) = - 6.60, p < .001, d = -1.86. Participants who read positive comments reported higher scores in future expectancy (M = 5.22, SD = 1.54) than those who read negative comments (M = 2.88, SD = 1.81), t(47) = - 4.86, p < .001, d = -1.39. The data were consistent with the hypothesis, thus hypothesis one was supported. Table 2 presents the t-test results. Figure 3 illustrates the effects of comment valence on reappraisal measures. Table 2  Descriptive Statistics and t-test Results for Reappraisal by Comment Valence   Comment Valence 95% CI for Mean Difference    Negative  Positive    M SD n M SD n   Lower   Upper     t df Emotion-Focused Coping Potential  2.07 1.03 25 5.12 1.15 24   -3.68     -2.43       -9.83* 47 Problem-Focused Coping Potential  3.13 1.62 25 5.61 0.94 24   -3.25     -1.73       -6.60*   38.88a Future Expectancy 2.88 1.81 25 5.22 1.54 24   -3.31     -1.37       -4.86* 47 Note. aThe t and df were adjusted because variances were not equal. * p < .001.    48 Figure 3. Effects of Comment Valence on Reappraisal  RQ2: The effects of comment valence on negative emotions. The second research question asked how the valence of audience comments affects the intensity of the negative emotions (anger, sadness, fear, and disgust) perceived (a) about the incident, (b) toward the bully, and (c) toward the audience.  Anger, fear, sadness, and disgust. A series of paired-sample t-tests were computed to compare the intensity of anger, fear, sadness, and disgust perceived about the outing incident, toward the bully, and toward the audience before and after reading comments. Results indicated that positive comments reduced the intensity of anger, fear, sadness, and disgust. There was a statistically significant change in mean scores before and after reading positive comments in anger, MB = 2.49, SDB = 1.28, MA= 1.74, SDA = 1.23, t(23) = 3.69, p = .001, d = .75; fear, MB = 2.44, SDB = 1.39, MA = 1.62, SDA = 1.22, t(23) = 3.67, p = .001, d = .75; sadness, MB = 2.22, SDB = 1.35, MA = 1.68, SDA = 1.20, t(23) = 2.11, p = .046, d = .43; and disgust, MB = 2.53, SDB = 1.34, MA = 1.83, SDA = 1.19, t(23) = 2.84, p = .009, d = .58, toward the audience. However, -2.5-2-1.5-1-0.500.511.522.5Emotion-Focused CopingPotentialProblem-Focused CopingPotentialFuture ExpectancyReappraisalNegative CommentsPositive Comments 49 positive comments did not affect anger, fear, sadness, and disgust toward the bully and about the incident. Negative comments had no statistically significant effect on the intensity of anger, fear, sadness, and disgust toward the three different targets. These t-test results are presented in Tables 3 and 4. Figures 4, 5, and 6 illustrate the differences in the intensity of each emotion for the three different targets. Table 3 Descriptive Statistics and t-test Results for Perceived Anger, Fear, Sadness, and Disgust by Target in Positive Comment Condition  Perceived Emotion by Target Pre-Stimulus  Post-Stimulus  95% CI for Mean Difference     M SD  M SD  Lower Upper  r t Anger (O) 3.12 1.12  2.94 1.09  -0.09 0.47  .82***    1.39 Fear (O) 2.47 1.45  2.33 1.43  -0.33 0.61  .70***  .61 Sadness(O) 2.58 1.22  2.31 1.28  -0.14 0.70  .68***    1.37 Disgust (O) 3.01 1.14  2.81 1.06  -0.10 0.51  .79***    1.42 Anger (B) 3.26 1.11  3.25 0.94  -0.28 0.30  .78*** .07 Fear (B) 2.01 1.42  2.18 1.51  -0.54 0.21  .82***     -.93 Sadness (B) 2.29 1.36  2.17 1.33  -0.33 0.57  .69***      .57 Disgust (B) 2.91 1.12  2.81 1.01  -0.20 0.43  .76***      .76 Anger (A) 2.49 1.28  1.74 1.23   0.33 1.17  .69***    3.69** Fear (A) 2.44 1.39  1.62 1.22   0.36 1.28  .65***    3.67** Sadness (A) 2.22 1.35  1.68 1.20   0.01 1.07  .52**    2.11* Disgust (A) 2.53 1.34  1.83 1.19   0.18 1.20  .56**    2.84** Note. Range = 0 - 4. O = outing; B = bully; A = audience. n = 24. df = 23. * p < .05. ** p < .01. *** p < .001.      50 Table 4 Descriptive Statistics and t-test Results for Perceived Anger, Fear, Sadness, and Disgust by Target in Negative Comment Condition  Perceived Emotion by Target Pre-Stimulus  Post-Stimulus  95% CI for Mean Difference     M SD  M SD  Lower Upper  r t Anger (O) 3.53 0.79  3.35 0.93  -0.12 0.48  .65**    1.23 Fear (O) 2.31 1.55  2.16 1.55  -0.21 0.51  .85**  .84 Sadness(O) 2.29 1.40  2.31 1.33  -0.37 0.34  .81**     -.08 Disgust (O) 3.23 1.05  3.19 1.07  -0.20 0.28  .85**  .35 Anger (B) 3.71 0.57  3.68 0.63  -0.17 0.23  .69**  .32 Fear (B) 2.12 1.65  2.07 1.53  -0.42 0.53  .74**  .23 Sadness (B) 2.09 1.37  2.19 1.36  -0.55 0.37  .67**     -.42 Disgust (B) 3.47 0.76  3.48 0.78  -0.40 0.31  .47*     -.08 Anger (A) 3.18 1.03  3.33 0.85  -0.58 0.28  .41* -.73 Fear (A) 2.07 1.60  2.08 1.43  -0.39 0.36  .82** -.07 Sadness (A) 2.43 1.20  2.44 1.24  -0.38 0.35  .74** -.08 Disgust (A) 2.83 1.31  2.99 1.05  -0.67 0.35  .47* -.65 Note. All t-test results were statistically insignificant at the .05 significance level. Range = 0 - 4. O = outing; B = bully; A = audience. n = 25. df = 24. * p < .05. ** p < .001.    Figure 4. Changes in the Intensity of Perceived Anger, Fear, Sadness, and Disgust about the Outing Incident by Comment Valence   -1-0.8-0.6-0.4-0.200.2AngerFearSadnessDisgustChange in Intensity of EmotionEmotion about the Outing IncidentPositive CommentsNegative Comments 51 Figure 5. Changes in the Intensity of Perceived Anger, Fear, Sadness, and Disgust toward the Bully by Comment Valence   Figure 6. Changes in the Intensity of Perceived Anger, Fear, Sadness, and Disgust toward the Audience by Comment Valence   -1-0.8-0.6-0.4-0.200.2AngerFearSadnessDisgustChange in Intensity of EmotionEmotion toward the BullyPositive CommentsNegative Comments-1-0.8-0.6-0.4-0.200.2AngerFearSadnessDisgustChange in Intensity of EmotionEmotion toward the AudiencePositive CommentsNegative Comments 52 Because anger, fear, sadness, and disgust showed similar tendencies in response to the valence of comments by three different targets, these four items were combined into an overall composite measure of negative emotion to further analyze the overall tendency. A two-way mixed ANOVA was computed to test the effects of comment valence and exposure to audience comments (pre- and post-test) on the negative emotions for each of three different targets (about the incident, toward the bully, and toward the audience).  Composite negative emotion about the outing incident. Results indicated that there was no statistically significant interaction between the comment valence and exposure to audience comments on negative emotion about the incident, F(1, 47) = .38, p = .54, partial 2 = .008. The main effect of exposure to audience comments showed no statistically significant difference in mean negative emotion about the incident at the different time points, F(1, 47) = 2.43, p = .13, partial 2 = .05. The main effect of comment valence showed that there was no statistically significant difference in mean negative emotion about the incident between the two groups who read negative or positive comments, F(1, 47) = .13, p = .71, partial 2 = .003. Figure 7 shows the pattern. Composite negative emotion toward the bully. There was no statistically significant interaction between the comment valence and exposure to audience comments on negative emotion toward the bully, F(1, 47) = .02, p = .88, partial 2 = .001. The main effect of exposure to audience comments showed no statistically significant difference in mean negative emotion toward the bully at the different time points, F(1, 47) = .01, p = .93, partial 2 < .001. The main effect of comment valence showed that there was no statistically significant difference in mean negative emotion toward the bully between the two groups who read negative or positive comments, F(1, 47) = .93, p = .34, partial 2 = .02. Figure 8 shows the pattern.   53 Figure 7. Estimated Marginal Means of Perceived Negative Emotion about the Incident by Comment Valence and Exposure to Audience Comments    Figure 8. Estimated Marginal Means of Perceived Negative Emotion toward the Bully by Comment Valence and Exposure to Audience Comments    54 Composite negative emotion toward the audience. There was a statistically significant interaction between the comment valence and exposure to audience comments on negative emotion toward the audience, F(1, 47) = 8.70, p = .005, partial 2 = .16. There was a statistically significant difference in negative emotion toward the audience between the negative and positive comment conditions at posttest, F(1, 47) = 11.35, p = .002, partial 2 = .20, but not at pretest, F(1, 47) = .39, p = .53, partial 2 = .008. Negative emotion toward the audience was significantly greater in the negative comment condition (M = 2.71, SD = .88), compared to the positive comment condition (M = 1.72, SD = 1.16). There was a statistically significant effect of exposure to audience comments on negative emotion toward the audience in the positive comment condition, F(1, 23) = 10.79, p = .003, partial 2 = .32, but not in the negative comment condition, F(1, 24) = .27, p = .61, partial 2 = .01. In the positive comment condition, negative emotion toward the audience was reduced at posttest (M = 1.72, SD = 1.16) compared to pretest (M = 2.42, SD = 1.24), a mean difference of .70, 95% CI [.26, 1.14]. Figure 9 shows the pattern. Figure 9. Estimated Marginal Means of Perceived Negative Emotion toward the Audience by Comment Valence and Exposure to Audience Comments    55 The Spotlight Effect in Outing H2: The spotlight effect in outing. The second hypothesis predicted that victims of outing overestimate the negativity of audience judgments such that their predicted audience judgments are more negative than the actual audience judgments. An independent sample t-test was computed to compare the predicted audience judgments of the victims and the actual audience judgments about the person shown in the video. Results indicated that participants perspective reported lower ratings of judgments (i.e., more negative judgments) (M = 2.12, SD = 1.14) than those M = 3.44, SD = .87), t(99) = -6.58, p < .001, d = 1.30. As predicted, vaudience judgments were more negative than the  about the victim. The data were consistent with the hypothesis, thus hypothesis two was supported. H3: The moderation effect of comment valence on the spotlight effect. The third hypothesis predicted that the spotlight effect is greater in the negative comment condition compared to the positive comment condition. A two-way ANOVA was computed with perspective and comment valence on judgment. Results indicated that there was a statistically significant interaction between perspective and comment valence for the judgment score, F(1, 97) = 5.72, p = .02, partial 2 = .06. Therefore, an analysis of simple main effects for comment valence was performed. All pairwise comparisons were run for each simple main effect with reported 95% confidence intervals and p-values Bonferroni-adjusted at the p < .025 level within each simple main effect. There was a statistically significant difference in the predicted and actual judgment scores in the negative comment condition, F(1, 97) = 28.61, p < .001, 2 = .23. In the negative comment condition, predicted judgment (M = 1.75, SD = 0.99) was more negative than the actual  56 judgment (M = 3.41, SD = 0.78), a mean difference of 1.66, 95% CI [1.04, 2.27]. There was no statistically significant difference between the predicted and the actual judgment scores in the positive comment condition, F(1, 97) = 3.22, p = .08, partial 2 = .03.  There was a statistically significant effect for comment valence on predicted judgment, F(1, 97) = 31.75, p < .001, parti2 = .25. The predicted judgment score was lower in the negative comment condition (M = 1.75, SD = 0.99), compared to the positive comment condition (M = 3.56, SD = 1.56), a mean difference of 1.81, 95% CI [1.18, 2.45]. There was a statistically significant difference in the actual judgment scores between the positive and negative comment conditions, F(1, 97) = 5.57, p 2 = .05. The actual judgment score was lower in the negative comment condition (M = 3.41, SD = 0.78), compared to the positive comment condition (M = 4.15, SD = 1.09), a mean difference of 0.74, 95% CI [0.12, 1.36].  As predicted, the spotlight effect was greater in the negative comment condition compared to the positive comment condition. Thus, hypothesis three was supported. Table 5 presents the means and standard deviations for judgment ratings by perspective and comment valence and Figure 10 illustrates the pattern. Table 5  Means and Standard Deviations of Pre- and Post-Stimulus Judgment Ratings by Perspective and Comment Valence Perspective Comment Valence Pre-Stimulus  Post-Stimulus  M SD  M SD n Victim Positive 2.26 1.04  3.56 1.56 24  Negative 1.99 1.23  1.75 0.99 25  Total 2.12 1.14  2.64 1.58 49 Audience Positive 3.59 0.87  4.15 1.09 24  Negative 3.32 0.87  3.41 0.78 28  Total 3.44 0.87  3.75 1.00 52     57 Figure 10. Estimated Marginal Means of Judgment Ratings by Perspective and Comment Valence  H4: The effects of comment valence on predicted audience judgments. The fourth hypothesis predicted that the valence of comments about a victim of outing affects predicted audience judgments such that (a) negative comments increase the negativity of the predicted audience judgments while (b) positive comments increase the positivity of the predicted audience judgments. A two-way mixed ANOVA was computed to test the effects of comment valence and exposure to audience comments (pre- and post-judgment. Results indicated that there was a statistically significant interaction between comment valence and exposure to audience comments on predicted judgments, F(1, 47) = 13.17, p < .001, partial 2 = .22. There was a statistically significant difference in predicted judgment between the negative and positive comment conditions at posttest, F(1, 47) = 23.80, p < .001, partial 2 = .34, but not at pretest, F(1, 47) = .72, p = .40, partial 2 = .02. Predicted judgment ratings were  58 significantly higher in the positive comment condition (M = 3.56, SD = 1.56), compared to the negative comment condition (M = 1.75, SD = .99). There was a statistically significant effect of exposure to audience comments on predicted judgment in the positive comment condition, F(1, 23) = 3.85, p = .001, partial 2 = .38, but not in the negative comment condition, F(1, 24) = .93, p = .34, partial 2 = .04. In the positive comment condition, predicted judgment ratings were statistically significantly higher at posttest (M = 3.57, SD = 1.56) compared to pretest (M = 2.26, SD = 1.04), a mean difference of 1.30, 95% CI [.58, 2.02].  As predicted, positive comments increased the positivity of the predicted audience judgments; however, negative comments did not increase the negativity of the predicted audience judgments. Thus, hypothesis four was partially supported. The means and standard deviations for pre- and post-stimulus predicted judgment ratings are presented in Table 5. Figure 11 shows the pattern. Figure 11. Estimated Marginal Means of Predicted Audience Judgment Ratings by Comment Valence and Exposure to Audience Comments    59 H5: Relationship between predicted audience judgments and negative emotions. The fifth hypothesis predicted that negative audience judgments perceived by victims of outing are positively correlated with their perceived negative emotions. The perceived audience judgments were measured by semantic differential bipolar items ranging from 1 to 7 with higher scores indicating more positivity. To test hypothesis five, the predicted audience judgment ratings were reverse-coded such that higher ratings indicated more negativity. Pearson bivariate correlation tests were conducted to assess the correlations between predicted negative audience judgments and perceived negative emotions about the outing incident, toward the bully, and toward the audience at posttest. An examination of the correlation matrix revealed a small to medium positive relationship between the negativity of predicted audience judgments and perceived negative emotions about the outing incident. Predicted negative audience judgment was positively associated with anger (r = .33, p = .02), fear (r = .19, p = .20), sadness (r = .13, p = .38), and disgust (r = .29, p = .04) about the outing incident. Predicted negative audience judgment was also positively associated with anger (r = .29, p = .046), fear (r = .14, p = .33), sadness (r = .20, p = .16), and disgust (r = .32, p = .03) toward the bully. Predicted negative audience judgment was most strongly positively associated with anger (r = .69, p < .001), fear (r = .35, p = .01), sadness (r = .52, p < .001), and disgust (r = .64, p < .001) toward the audience. The data were consistent with the hypothesis, thus hypothesis five was supported. Table 6 displays the results of the correlation analysis.    60 Table 6  Pearson Product-Moment Correlations for Predicted Negative Audience Judgment and Perceived Negative Emotions about the Outing Incident, toward the Bully, and toward the Audience  Variable  1 2 3 4 5 6 7 8 9 10 11 12 1. Judgmenta              2. Anger (O)  .33***            3. Fear (O) .19*** .36***           4. Sadness (O) .13*** .47*** .62***          5. Disgust (O) .29*** .77*** .46*** .58***         6. Anger (B) .29*** .84*** .19*** .32*** .55***        7. Fear (B) .14***            .26*** .87*** .61*** .39*** .18***       8. Sadness (B) .20*** .25*** .66*** .76*** .41*** .18*** .78***      9. Disgust (B) .32*** .63*** .44*** .50*** .85*** .58*** .46*** .39***     10. Anger (A) .69*** .46*** .31*** .35*** .43*** .43*** .38*** .39*** .54***    11. Fear (A) .35*** .19*** .75*** .42*** .33*** .08*** .75*** .54*** .41*** .50***   12. Sadness (A) .52*** .34*** .60*** .66*** .44*** .24*** .66*** .74*** .49*** .69*** .77***  13. Disgust (A) .64*** .33*** .39*** .37*** .52*** .23*** .46*** .47*** .58*** .85*** .63*** .75*** Note. a The predicted audience judgment ratings were reverse-coded such that higher ratings indicate more negativity.  O = outing; B = bully; A = audience. n = 49. * p < .05. ** p < .01. *** p < .001.    61 The Influence of Online Comments and Empathy on Audience Perception  H6: The effects of comment valence on audience judgments. Hypothesis six predicted that the valence of comments about a victim of outing affects audience judgment such that (a) negative comments increase the negativity of the audience judgment while (b) positive comments increase the positivity of the audience judgment. A two-way mixed ANOVA was computed to test the effects of comment valence and exposure to audience comments (pre- and post-test) on audience judgments on the victim.  Results show that there was no statistically significant interaction between the comment valence and exposure to audience comments on audience judgments on the victim, F(1, 50) = 2.41, p = .13, partial 2 = .05. The main effect of exposure to audience comments showed a statistically significant difference in mean judgment about the victim at the different time points, F(1, 50) = 4.67, p = .04, partial 2 = .09. There was an increase in judgment scores from 3.46 at pretest to 3.78 at posttest, a mean difference of 0.32, 95% CI [0.02, 0.06]. The main effect of comment valence showed that there was a statistically significant difference in mean judgment about the victim between the negative and positive comment conditions, F(1, 50) = 6.37, p = .02, partial 2 = .11. The marginal means for audience judgment score were 3.36 (SE = 0.14) for the negative comment condition and 3.87 (SE = 0.15) for the positive comment condition, a statistically significant mean difference of .51, 95% CI [0.10, 0.91].  To further analyze changes in audience judgment ratings before and after reading comments, a paired-sample t-test was computed for each of the positive and negative comment conditions. In the positive comment condition, there was a statistically significant increase in judgment scores from 3.59 (SD = .87) at pretest to 4.15 (SD = 1.09) at posttest, a mean difference of 0.56, 95% CI [0.003, 1.11], t(23) = 2.08, p = .049, d = .42. No statistically significant change  62 in judgment scores between pre- and post-tests was found in the negative comment condition, t (27) = .58, p = .56.  Positive comments increased positive judgments as expected; however, negative comments did not affect the audience judgments. Therefore, hypothesis six was partially supported. The means and standard deviations for pre- and post-stimulus actual judgment ratings are presented in Table 5. Figure 12 shows that pattern. Figure 12. Estimated Marginal Means of Actual Audience Judgment Ratings by Comment Valence and Exposure to Audience Comments    H7: Relationship between empathy and audience judgments. Hypothesis seven predicted that empathy is positively associated with positive audience judgments about the victim of outing. Pearson bivariate correlation tests were conducted to assess the correlations between audience judgments and two dimensions of empathy (cognitive and affective empathy) at posttest. An examination of the correlation matrix revealed a medium negative relationship between positive audience judgments and empathy. Positive audience judgment was positively  63 associated with cognitive empathy (r = .33, p = .02) and affective empathy (r = .41, p = .002). The data were consistent with the hypothesis, thus hypothesis seven was supported. Table 7 depicts the results of the correlation analysis.  Table 7  Pearson Product-Moment Correlations and Descriptive Statistics for Actual Audience Judgment and Empathy  Variable  1 2 M (SD) 1. Audience Judgment   3.75 (1.00) 2. Cognitive Empathy .33*  2.54 (1.16) 3. Affective Empathy .41** .88*** 2.57 (1.19)    Note. n = 52. * p < .05. ** p < .01. *** p < .001.    64 CHAPTER 5 DISCUSSION -based audience cyberbullying focusing on perceptions and emotions associated with outing. The main purpose of this study was to test predictions based on the spotlight effect and appraisal theories of emotion. This study identified emotions perceived toward different targets in response to an outing incident and examined how audience comments, particularly the valence of comments, empathy on audience perception also was explored. The following sections provide the summary and interpretation of the findings in this study; discuss the limitations and implications; and make suggestions for future research. Summary and Interpretation         emotional reactions and psychological responses to outing were investigated regarding three different targets, specifically perceived negative emotions about the incident, and toward the emotional and psychological responses toward the bully and the audience. Analyses of the frequencies of each emotion by target revealed that different types of emotions were identified toward different targets in the outing incident. Pdepressed, humiliated, and embarrassed when they thought of the audience compared to when they thought of the bully. No statistically significant difference was found in the other emotions, indicating that there was considerable overlap in the other types of emotions perceived toward the bully and the audience. Responses to the open-ended question identified that four emotional  65 responses of paranoid, concerned, worried, and violated were felt in common across the three different targets. Additionally, there were several differences regarding the specific targets. Hurt was an emotion perceived toward the bully and about the incident; suicidal was reported toward the audience and about the incident; and vengeful was directed toward the bully and the audience. Moreover, each target elicited a unique set of emotions in response to the open-ended question. The incident evoked responses of exploited, nervous, and betrayed; the bully elicited responses of enraged and murderous; and the audience elicited responses of confused and invaded. It is notable that participants reported more intense emotions toward the bully (e.g., enraged and murderous) than toward the audience (e.g., confused and invaded). Furthermore, the t-tests examining the differences in the intensity of each emotion revealed that victims of outing experienced stronger emotions of depression and embarrassment toward the audience than the bully, and stronger emotions of anger, disgust, and frustration toward the bully than the audience. These differences make sense given the definitional distinctions between sadness and embarrassment compared to anger. Sadness, a major component of depression, occurs in response to an irrevocable loss, such as the loss of public face (Smith & Lazarus, 1993) while embarrassment has feels that he/she has lost control in a very public social medium. In contrast, anger occurs when someone else is responsible for a bad situation, and especially when this is seen as an unjust violation (Smith & Lazarus, 1993). As such, appraisal theories explained about the link between accountability and anger when it comes to a negative event, and though the results from the McNemar test indicated that participants in the victim condition reported that they would feel anger toward the audience as  66 well as toward the bully, the t-test results further clarified that victims experienced more intense anger toward the bully than the audience. It is notable that anger was perceived toward the audience even if the audience is not held directly accountable for the incident. An interview from a victim of public humiliation provides further insight about the accountability of an audience in an outing incident. Jennifer Lawrence, a victim of Hollywood photo hacking in 2014, called the incident a sex crime and said that anybody who looked at the nude pictures perpetuates a sexual offense (Kashner, 2014). She considered the audience who were eager to view the photos as aggressors because the audience did not have her permission to look at her naked body. Similarly, victims of outing may hold an audience partially accountable for the incident, even though the audience did not have intention to harm the victim, because the audience takes part in -ended question in this study the audience. Based on the knowledge gained from parsing out the targets (e.g., the bully vs. the audience) associated with each emotion in response to an outing incident, one contribution of this research is to provide evidence for the audience effect. The imThis study further theories (Lazarus, 1993; Lazarus & Folkman, 1984). The first hypothesis predicted that the such that positive audience comments are reappraised more positively than negative audience comments. As predicted, participants in the positive comment condition reported higher ratings on emotion-focused coping potential, problem-focused coping potential, and future expectancy than those in the negative comment condition. These results justify the claim that the valence of comments  67 affects emotional responses due to reappraisal as reappraisals lead to changes in emotional responses, according to appraisal theories (Lazarus, 1993; Lazarus & Folkman, 1984). Additionally, appraisal theories posit that social support facilitates reappraisal of the source of stress and thus modifies the initial appraisals (Lazarus & Folkman, 1984). It is suspected that positive audience comments led to enhanced coping because positive comments in this study were interpreted as socially supportive. In fact, the assessments of audience comments by the participants in the victim condition revealed that positive comments were perceived as MP = 5.83, SDP = 1.55, n = 24; MN = 1.60, SDN = 1.19, n = 25, t(47) = 10.75, p < .001, d = 3.07).  The second research question asked how the valence of audience comments affects the intensity of the negative emotions (anger, sadness, fear, and disgust) perceived (a) about the incident, (b) toward the bully, and (c) toward the audience. Changes in each negative discrete emotion toward different targets before and after exposure to audience comments were analyzed. The only statistically significant difference was found in perceived anger, fear, sadness, and disgust toward the audience in the positive comment condition. Positive comments reduced the intensity of anger, fear, sadness, and disgust perceived toward the audience. Negative comments did not affect these emotions toward the audience. Neither positive nor negative comments affected these four types of emotions about the outing incident and toward the bully. Furthermore, these four discrete emotions were combined to generate a perceived negative emotion composite measure to explore an overall tendency. Perceived negative emotions about the incident, toward the bully, and toward the audience were examined separately with regard to the effects of comment valence and exposure to audience comments. A statistically significant interaction effect of the comment valence and exposure to audience  68 comments was found in the perceived negative emotion toward the audience. Further analyses revealed that the interaction effect was mainly due to the effect of positive comments on the perceived negative emotion. Positive comments significantly reduced the intensity of the perceived negative emotion toward the audience; however, negative comments did not elicit more intense perceived negative emotion toward the audience. It is suspected that negative comments were not perceived as harsh enough to increase the intensity of perceived negative emotion toward the audience even if participants reported that the negative comments led to negative reappraisals.  Also, the results indicate that the comment valence had no effect on the perceived negative emotion toward the bully and about the incident. It is likely that the online comments in response to the post were portrayed as an audience, thus the online comments enhanced the salience of the audience in the minds of the participants and led to changes in emotions directly related to this target, and not related to the bully or the incident. Exposure to the positive or negative audience responses did not influence perceived negative emotion toward the bully and about the incident likely because no matter what others say about the incident, the bully is still accountable for the bad situation and the quality of the incident, and the fact that it occurred, do not change. Given these stable attributions about the bully and the incident, it follows that the emotions related to these targets also would be stable. The spotlight effect in outing. This study examined the spotlight effect by comparing predicted judgments actual judgments about a victim of outing. The second hypothesis predicted that victims of outing overestimate the negativity of audience judgments such that their predicted audience judgments are more negative than the actual audience judgments. In this study, p 69 ratings of audience judgments of the victim compared to how the audience participants actually judged the victim (with lower ratings representing more negative judgments). This result is consistent with previous research findings on the spotlight effect that have found predicted judgments tend to be more negative than actual judgments in response to an embarrassing event (Savitsky et al., 2001). The spotlight effect helps us understand how victims and an audience of outing make perceptions regarding the incident. Combined with the findings above regarding emotions associated with the audience, this research extends our understanding of the spotlight effect. Specifically, the spotlight effect occurs in an event where an audience is present, but their immediate reactions to the event are not explicitly expressed. Given the limited cues from the audience, the person in an embarrassing situation makes perceptions and predictions of what the audience would think of him/her. Given the negative emotions associated with an audience revealed by this research, including depression, humiliation, and embarrassment, it follows that these self-focused emotions would lead to attributional errors, namely overestimations of negative audience judgments. The third hypothesis predicted that the spotlight effect is greater in the negative comment condition compared to the positive comment condition. It was predicted that the valence of comments moderates the relationship between predicted and actual judgments such that predicted judgment is much more negative than the actual judgment in the negative comment condition while the difference between the predicted and actual judgment is smaller in the positive comment condition. To test this hypothesis, participants were presented with a set of positive or negative comments toward the victim. As expected, the perceptual difference between the victim and the audience was greater in the negative comment condition while the difference was minimal in the positive comment condition. With negative comments, the predicted  70 judgment was more negative and the actual judgment did not significantly change. On the other hand, positive responses from the audience increased the positivity of predicted judgment greatly while they did not change the actual judgment much. As a result, the difference between the predicted and the actual judgments was slim. This result suggests that the spotlight effect is more pronounced in the negative comment condition than in the positive comment condition. Thus, the comment valence moderated the spotlight effect. Victims of outing overestimated the negativity while the predicted and the actual perceptions did not differ in the positive comment condition. This finding could represent an important avenue for communication interventions regarding cyberbullying. In particular, it may be helpful for online audiences to circumvent the negative outcomes of cyberbullying by posting positive, supportive comments about victims. The fourth hypothesis predicted that the valence of comments about a victim of outing affects the  predicted audience judgments such that negative comments increase the negativity of the predicted audience judgments while positive comments increase the positivity of the predicted audience judgments. This hypothesis was partially supported. As predicted, positive comments increased the positivity of the predicted audience judgments. Negative comments slightly increased the negativity of the predicted judgments; however, the change was not statistically significant. These results are consistent with predictions made by the spotlight effect and appraisal theories. Predicted audience judgments were negatively overestimated when the audience feedback was not available as a cue, according to the spotlight effect literature (Gilovich et al, 2000; Savitsky et al., 2001). Moreover, as appraisal theories posit (Lazarus, 1991, 1993), the concepts of goal congruence and future expectations can account for how the  71 comments were appraised. For example, positive comments may have modified  negatively overestimated predictions because these comments wergoals and desires, and their future expectations hoping to cope with the incident. In contrast, negative comments may have confirmed their negative predictions because these comments were against their goals and desires as appraisal theories predicted. It is suspected that the statistically insignificant change in the negative comments condition indicates that the negative comments used in this study closely matched predicted audience responses of the respondents. Negative comments with harsher remarks might have resulted in a statistically significant change.   The fifth hypothesis predicted that negative audience judgments perceived by victims of outing are positively correlated with their perceived negative emotions. Overall, predicted negative judgments were positively correlated with perceived negative emotions (anger, fear, sadness, and disgust) about the outing incident, toward the bully, and toward the audience. By target, the predicted negative judgments were more strongly positively correlated with perceived negative emotions toward the audience, compared to perceived negative emotions about the incident and toward the bully. This strength of association is likely attributable to the fact that the comments were from an online audience, thus leading to a stronger association with the audience, rather than the incident or bully. Regarding the audience as a target, anger was the strongest perceived negative emotion positively associated with predicted negative judgments, followed by disgust, sadness, and fear. Regarding the bully as a target, disgust was the strongest perceived negative emotion positively associated with predicted negative judgments, followed by anger, sadness, and fear. Regarding the incident as a target, anger was the strongest perceived negative emotion positively associated with predicted negative judgments, followed by disgust, fear, and sadness. Overall, anger and disgust were more strongly perceived negative emotions  72 positively associated with perceived negative judgments compared to sadness and fear. These results indicate that predicted audience judgments can be a predictor of negative emotions associated with cyberbullying, and thus suggest potential for an effective communication campaign to facilitate coping. As previously mentioned, positive, supportive comments can be posted online to aid coping and emotional reappraisal perceptions. One benefit of posting positive comments online is that they can be viewed anytime the victim needs support.  Peer influence on audience judgments. The sixth hypothesis predicted that the valence of comments about a victim of outing affects the audience judgments such that negative comments increase the negativity of the audience judgments while positive comments increase the positivity of the audience judgments. This hypothesis was partially supported. As predicted and consistent with previous research (Ballantine et al., 2015; Edwards et al., 2007; Walther et al., 2008, 2010), positive comments increased positive audience judgment ratings. However, negative comments did not increase the negativity of the audience judgments, a result that contradicted predictions. This contradicting result might be due to the quality of the online comments used in this study. Participants in the audience perspective with the negative comment condition perceived the comments as less realistic. They might have thought the negative audience feedback less realistic because it was not socially desirable. Leaving a negative comment toward the victim in an outing incident might have been perceived as an act of joining in bullying, thus it might have been perceived against the social norm. Also, it should be noted that although the negative comments made little change, which is statistically insignificant, in audience judgments, the change occurred in the opposite direction. Surprisingly, the negative comments slightly increased positive audience judgment ratings. It might be that those who read  73 the negative comments were against the negative remarks and became protective of the victim. This assumption was not directly tested but could provide a meaningful avenue for future research. All in all, these results can be used to inform victims of outing that an audience judges victim of outing will make more favorable impressions of the victim; however, the audience hus they will not think less of the victim. The role of empathy on audience judgments. The seventh hypothesis predicted that empathy is positively correlated with the positivity of audience judgments about the victim of outing. Parallel to the previous research findings that more empathic people tend to be more supportive of cyberbullied victims than less empathic people (Shultz et al. 2014; Van Cleemput et al., 2014), this study found that both cognitive and affective empathy are positively correlated with favorable audience judgments of the victim in an outing incident. Affective empathy was slightly more strongly correlated with positive audience judgments than cognitive empathy (r = .41 vs. r = .33, respectively). These results indicate that those with higher levels of empathy tend to judge a victim of outing less harshly than those with lower levels of empathy. Thus, empathy is a key factor in making favorable impressions of a victim of outing and other types of cyberbullying. Epley, Savitsky, and Gilovich (2002) argued that the spotlight effect occurs because people overlook that other people can empathize with them thus make more charitable judgments on them. Their definition of empathy only included cognitive empathy. This current study suggests that affective empathy is also an important component of empathy that leads people to make favorable judgments.   74 Limitations There are several limitations in this study. First, this study used a hypothetical scenario with fabricated comments and that it examined perceptual responses to the stimuli rather than this study perceived the hypothetical scenario to be realistic according to their ratings of the event. Despite these perceptions, the negative comments about the victim of outing were perceived to be less realistic, compared to the positive comments. Higher realism ratings for the positive comments suggest that these comments may have been viewed as more socially desirable responses to the scenario. However, these assumptions were not directly verified. The results need to be interpreted with caution considering this limitation. Another limitation is that this study only assessed whether there was an impact for comment valence and did not parse out particular aspects of the content of audience comments or their unique influence. The audience comments created for this study were multi-dimensional in terms of content, including accountability for the incident (e.g., blaming the bully or the victim), different types of social support (e.g., boosting the victim self-esteem or offering general support/kindness), and different targets of the message (e.g., appealing to the victim or to the audience). This multi-dimensionality of the audience comments also was reflected in the induction check measures, which included multiple items beyond the general positivity or negativity of audience comments (e.g., discouraging-encouraging, unsupportive-supportive, unkind-kind, not helpful-helpful, discomforting-comforting, and unfavorable-favorable). Third, repeated measures were used for a short period of time in this study. The average completion time for the survey questionnaire was about nine minutes. Processing a stimulus message and the accompanying audience responses within less than ten minutes can be  75 problematic for a longitudinal study; however, this sequential repeated measure design should be considered as an experimental treatment to separate comment effects from participreactions to the incident. This design has a strength in that it can separate comment effects from the initial perception about the incident. Presenting the posting with comments at the same time could have been more realistic than the sequential approach; however, it is also realistic that one finds a posting without comments and then views posted comments later. Fourth, another issue centers on how emotional responses were measured. In this study, some emotional responses were measured with multiple items (anger, fear, sadness, disgust, and happiness) while other emotional responses were measured with single items (depressed, anxious, stressed, humiliated, embarrassed, frustrated, and ashamed) drawing from the previous research. The inconsistency in emotional response measures is mainly due to the lack of comprehensive emotion measures for cyberbullying and the exploratory approach for examining different emotions associated with different targets (RQ1) in this study. Also, this study used a pretest-posttest design to show the real change in emotional responses; however, 23 emotion items were repeatedly measured about three different targets at two time points and this might have been associated with a higher dropout rate in the victim condition compared to the audience condition (18% vs. 6%). Fifth, there is a possibility of Type I error for findings that were significant at the .05 level in this study due to the number of significance tests computed. However, given the large effect sizes for several of  findings, this is less likely to be a problem.   Lastly, it is notable that it took about 10 weeks to collect data on MTurk for this study. This is an unusually long time for online data collection, especially considering that the target sample size of this study was relatively small. This could be due to the sensitive, sexual nature of  76 the outing scenario used for this research. It may have been associated with a self-selection bias, specifically, certain people may have opted out of participating in the study because of the topic. Implications This study extends the current cyberbullying literature by examining the audience effect in broad-based audience cyberbullying. Much of the cyberbullying research has defined cyberbullying acts broadly and included multiple types of cyberbullying when they study cyberbullying behavior and its relation to other variables such as empathy and emotion. By using a broad and inclusive definition of cyberbullying, studies have failed to specify the effects of distinctive types of cyberbullying on their emotional consequences. Taking this limitation into consideration, the current study focused on one specific type of cyberbullying that involves a broad-based audience in order to effectively address the influence of an audience on the victim perspectives at the same time. Studying how victims and an audience of outing respond to the same stimuli provided a more complete view of interactive features of cyberbullying. Also, the experimental design employed in this study led to a better understanding of comment valence by time. A significant implication of this study is that it is the first study to test the spotlight effect in cyberbullying. Among the several types of cyberbullying, the spotlight effect is expected to be quite pronounced in outing as it involves a large audience. The current experimental survey research provides empirical evidence that the spotlight effect persists in cyberbullying. The harsh judgments of others in various contexts (Epley et al., 2000; Savitsky et al., 2001).  77 Furthermore, this study suggests that positive audience comments and audience empathy are important factors that can remedy the spotlight effect, representing an important opportunity for communication researchers to generate effective campaigns to facilitate the coping of cyberbullied victims. This study adds to the emotion literature by studying discrete emotions associated with outing. Most cyberbullying research used a single item measure for each emotion associated with cyberbullying such as anger or sadness (e.g., Ortega et al., 2012; Patchin & Hinduja, 2006). Some researchers simply asked study participants if cyberbullying was emotionally distressing (Straude-Müller et al., 2012) or if it was perceived to have negative impact (Slonje & Smith, 2008; Smith et al., 2008). To the best of my knowledge, no study has attempted to differentiate emotional responses toward different targets in cyberbullying. Considering these limitations in previous research, the current study provides an example of how emotions can be studied in the context of cyberbullying. This study used discrete emotion measures because the discrete emotion model has better predictive power than a single factor item (Dillard & Peck, 2001; Dillard & Seo, 2013). The multifaceted emotional responses were observed toward different targets in this studyresearch has uncovered the specific emotions that need to be managed in order to facilitate or not they post supportive comments) in cyberbullying may determine the magnitude of  toward the audience. Lastly, this study has real world implications as it provided the knowledge with which those who have been victims of cyberbullying can reduce their emotional distress. This study revealed that the harshness of predicted audience judgments on a victim of outing was  78 exaggerated and that people did not judge the victim as harshly as the victim expected. This fact needs to be communicated to cyberbullied victims, repeatedly, in order to aid their coping and help them positively reappraise the event. Furthermore, positive comments led to more favorable audience judgments on the victim of outing while negative comments did not make the audience judgments more negative. Also, empathic people tended to make favorable impressions of the victim. Taken as a whole, the findings from this research can be applied in the real world to help victims of broad-based audience cyberbullying reduce their emotional distress. Future Research Although the present study offered contributions to the literature concerning broad-based audience cyberbullying, the results of this study leave more to be investigated and answered. This study calls attention to several topics in need of further investigation. First, follow-up study is needed to confirm the generalizability of the research findings in this study. Because this study focused solely on outing, it raises some questions about its interpretation in comparison with other types of cyberbullying. Future research needs to study if the spotlight effect consistently occurs in outing and in other types of cyberbullying. Also, to further test the spotlight effect in an online setting, future research may examine the online comment effect on the spotlight effect in a positive event. Previous research showed that the spotlight effect persists both in positive and negative incidents (Gilovich et al., 2000). Since the current study focused only on a negative event, it calls into question how the online comments influence the spotlight effect in a positive event. Second, studying how different emotions toward different targets are linked to other important variables, such as suicidal ideation, will provide valuable information. For example, whether sadness perceived toward the audience is more strongly associated with depression than  79 sadness perceived toward the bully; whether anger perceived toward the bully and toward the audience elicits different psychological consequences can be examined.  Third, this study found that positive comments helped the victims of outing reduce their emotions. Future research is needed on message features . What specific message features can be used to facilitate positive reappraisals, correct misperception, and reduce stress is a question for further examination. For example, how mixed valence comments affect can be examined. Mixed valence messages containing both negative and positive comments in different proportions may have  Furthermore, future work should distinguish the separate dimensions of audience comments and examine the effects of each unique dimension. For example, audience comments focusing on emotional or esteem support for the victim, accountability of the incident (e.g., blaming the bully vs. blaming the victim), or target audiences (e.g., offering help to the victim vs. encouraging other audience members to support the victim) may have  perceptions. Also, it is necessary to develop a measurement scale that captures each unique dimension. Lastly, future research needs to study which factors turn online observers into actors and lead people to leave positive comments instead of bypassing. This is an important area of study because research has shown that bystanders who are inactive are more likely to also become a victim or an aggressor themselves, whereas those who intervene are less likely to become a victim or an aggressor (Ferreira, Simao, Ferreira, Souza, & Francisco, 2016). A clear distinction among the bully, the victim, and the audience might not be possible. A bully at one point can become a victim and an audience can become a bully or a victim, and vice versa. The current  80 study considered empathy as a key factor in audience behavior and discussed the role of an audience as helpers, or supportive bystanders. More studies need to provide answers to important questions related to this topic, such as what motivates the audience to become a neutral bystander, a judge, or a social support provider in cyberbullying and whether the audience chooses their role based on their nature or situational factors. Furthermore, communication intervention efforts should focus on moving a passive yet empathic bystander to an active sympathetic upstander who speaks up against cyberbullying by posting positive, supportive comments. An anti-cyberbullying campaign can be designed to emphasize the role of an audience to be an upstander instead of being a bystander it is not okay to bully, but it is also important to speak up against bullying.  Conclusion This study investigated perceptions and emotions associated with broad-based audience cyberbullying. This study found that the audience and the bully evoked different negative emotional responses from victims of outing. Victims of outing tended to overestimate the ormation revealed online. Positive audience comments remedied this perceptual discrepancy between the victim and the audience and led to emotions of the victim and facilitated more positive perceived judgment of the victims while negative comments had little effect. Empathy toward the victim played an important role in . This study emphasized the important role of an audience in cyberbullying and discussed how the mere existence of an audience can significantly impact the  and that positive audience comments can  81 help the victims cope with the negative consequences of outing. This study will be a step toward a richer and more inclusive understanding of cyberbullying involving a broad-based audience.    82             APPENDICES    83 APPENDIX A Hypothetical Scenario for Victim Condition  Below is a description of a hypothetical scenario. Please try to imagine the following situation occurring to you. Even if the scenario describes behavior that you believe could not happen to you, we are asking you to imagine your reactions and perceptions to the event if it did happen to you:  Figure 13. Hypothetical Scenario for Victim Condition     84 APPENDIX B  Hypothetical Scenario for Audience Condition  Below is a description of a hypothetical scenario. Please try to imagine the following situation occurring to you. Even if the scenario describes behavior that you believe could not happen to you, we are asking you to imagine your reactions and perceptions to the event if it did happen to you:  Figure 14. Hypothetical Scenario for Audience Condition      85 APPENDIX C  Stimulus Materials for Positive Comments Condition  Now imagine that you were informed of comments people left under the video on YouTube. Please read the following comments carefully and answer the following questions.  Figure 15. Stimulus Materials for Positive Comments Condition      86 APPENDIX D  Stimulus Materials for Negative Comments Condition  Now imagine that you were informed of comments people left under the video on YouTube. Please read the following comments carefully and answer the following questions.  Figure 16. Stimulus Materials for Negative Comments Condition      87 APPENDIX E  Measurement Items  A. Assessments of Hypothetical Scenario  A1. Realism 1. A situation like this could happen to someone. 2. A scenario like this could potentially occur to anyone.   3. This incident could realistically happen.   4. It is possible for an event like this to happen to someone. 5. It is probable that something like this could happen to someone. 6. It is conceivable that an event like this could take place.  A2. Sensitivity of Information 1. The video contains private information. 2. The video contains sensitive information. 3. The video contains risky information. 4. The video contains intimate information. 5. The video contains personal information. 6. The video contains confidential information.  A3. Intention of Disclosure 1. The video was shared with your consent (the consent of the person in the video). 2. You (the person in the video) voluntarily shared the video. 3. You (the person in the video) had intention to share the video. 4. You (the person in the video) were aware that the video would be seen by other people. 5. You (the person in the video) wanted other people to see the video.   B. Assessments of Online Comments  B1. Valence 1. Negative - Positive 2. Discouraging - Encouraging 3. Unsupportive - Supportive 4. Unkind - Kind 5. Not helpful - Helpful 6. Discomforting - Comforting 7. Unfavorable  Favorable  B2. Realism 1. Unrealistic - Realistic 2. Impossible - Possible 3. Unlikely - Likely 4. Implausible - Plausible 5. Untruthful  Truthful 6. Unbelievable - Believable       88 C. Judgments  C1. Predicted Audience Judgment ink that people who watched this video would form ____________      C2. Actual Audience Judgment ____________   1. Negative - Positive 2. Unpleasant - Pleasant 3. Bad - Good 4. Undesirable - Desirable  5. Worthless - Worthy 6. Shameful - Admirable 7. Unattractive - Attractive 8. Disagreeable - Agreeable 9. Unfavorable - Favorable  D. Reappraisal  D1. Emotion-Focused Coping Potential 1. believe I could deal with my emotions in this situation. 2. improve the way I think about this situation. 3. improve the way I see this situation. 4. view this situation more positively. 5. change my perspective on this situation. 6. help me to feel better.  D2. Problem-Focused Coping Potential 1. I could do what is needed to manage this situation. 2. I could take appropriate action in this situation. 3. I could do what needs to be done in this situation. 4. I would be capable of acting appropriately in this situation. 5. I am capable of dealing with this situation.   D3. Future Expectancy 1. This situation will improve. 2. This situation will get better. 3. This situation will turn around.     89 E. Emotion Items  1. Irritated 2. Angry 3. Annoyed 4. Aggravated 5. Fearful 6. Afraid 7. Scared 8. Sad 9. Dreary 10. Dismal 11. Sickened 12. Disgusted 13. Revolted 14. Happy 15. Content 16. Cheerful 17. Depressed 18. Stressed 19. Anxious 20. Frustrated 21. Ashamed 22. Humiliated 23. Embarrassed  F. Empathy  F1. Affective Empathy (Empathic Concern) 1. I have tender, concerned feelings for the person. 2. I feel very sorry for the person. 3. I feel kind of protective toward the person. 4. I feel very much pity for the person.  F2. Cognitive Empathy (Perspective Taking) 1. perspective. 2.  3. I try to imagine how I     90               REFERENCES               91 REFERENCES  Altman, I., & Taylor, D. (1973). Social penetration: The development of interpersonal relationships. New York: Holt. Ang, R. P., & Goh, D. H. (2010). Cyberbullying among adolescents: The role of affective and cognitive empathy, and gender. Child Psychiatry & Human Development, 41, 387397. Ballantine, P. W., Lin, Y., & Veer, E. (2015). The influence of user comments on perceptions of Facebook relationship status updates. Computers in Human Behavior, 49, 5055. Brown, M.A., & Stopa, L. (2007). The spotlight effect and the illusion of transparency in social anxiety. Journal of Anxiety Disorders, 21, 804819. Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazonof inexpensive, yet high-quality, data? Perspectives on Psychological Science, 6, 35.  Burkard, A. W., & Knox, S. (2004). Effect of therapist color-blindness on empathy and attributions in cross-cultural counseling. Journal of Counseling Psychology, 51, 387397. Davis, M. H. (1980). A multidimensional approach to individual differences in empathy. JSAS Catalog of Selected Documents in Psychology, 10, 85.  Davis, M. H. (1994). Empathy: A social psychological approach. Madison, WI: Brown and Benchmark. Dillard, J. P., & Peck, E. (2001). Persuasion and the structure of affect: Dual systems and discrete emotions as complementary models. Human Communication Research, 27, 3868. Dillard, J. P., & Seo, K. (2013). Affect and persuasion. In J. Dillard & L. Shen (Eds.), The SAGE handbook of persuasion: Developments in theory and practice (2nd ed., pp. 150166). Thousand Oaks, CA: Sage. Duke, A. (2014, October 12). 5 things to know about the celebrity nude photo hacking scandal. CNN. Retrieved November 12, 2014, from http://edition.cnn.com/2014/09/02/showbiz/hacked-nude-photos-five-things Edwards, C., Edwards, A., Qing, Q., & Wahl, S. (2007). The influence of computer-mediated word-of-mouth communication on student perceptions of instructors and attitudes toward learning course content. Communication Education, 53, 255277. Epley, N., Savitsky, K., & Gilovich, T. (2002). Empathy neglect: reconciling the spotlight effect and the correspondence bias. Journal of Personality and Social Psychology, 83, 300312. Erdfelder, E., Faul, F., & Buchner, A. (1996). GPOWER: A general power analysis program. Behavior Research Methods, Instruments, & Computers, 28, 111.  92 Folkman, S., Lazarus, R.S., Dunkel-Schetter, C., DeLongis, A., & Gruen, R.J. (1986). Dynamics of a stressful encounter: Cognitive appraisal, coping, and encounter outcomes. Journal of Personality and Social Psychology, 50, 9921003. Gilovich, T., Kruger, J., & Medvec, V. H. (2002). The spotlight effect revisited: Overestimating the manifest variability of our actions and appearance. Journal of Experimental Social Psychology, 38, 9399.  Gilovich, T., Medvec, V. H., & Savitsky, K. (2000). The spotlight effect in social judgment: An Journal of Personality and Social Psychology, 78, 211222. Gilovich, T., & Savitsky, K. (1999). The spotlight effect and the illusion of transparency: Current Directions in Psychological Science, 8, 165168. Gini, G., Albiero, P., Benelli, B., & Altoe, Gbullying and defending behavior? Aggressive Behavior, 33, 467476. Griffin, D. W., & Ross, L. (1991). Subjective construal, social inference, and human misunderstanding. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 24, pp. 319359). San Diego, CA: Academic Press. Guo, S. (2016). A meta-analysis of the predictors of cyberbullying perpetration and victimization. Psychology in the schools, 53, 432453. Haikal, M. & Hong, R.Y. (2010). The effects of social evaluation and looming threat on self-attentional biases and social anxiety. Journal of Anxiety Disorders, 24, 345352. Harmon, A. (2004, August 26). Internet gives teenage bullies weapons to wound from afar. New York Times. Retrieved February 2, 2011, from http://query.nytimes.com/gst/fullpage.html?res=9E00E6D8133EF935A1575BC0A9629C8B63&pagewanted=all Hawker, D., & Boulton, M. (2000). Twenty years' research on peer victimization and psychosocial maladjustment: A meta-analytic review of cross-sectional studies. Journal of Child Psychology and Psychiatry, 41, 441455.  Hay, C., Meldrum, R., & Mann, K. (2010). Traditional bullying, cyber bullying, and deviance: A general strain theory approach. Journal of Contemporary Criminal Justice, 26, 130147. Heirman, W., & Walrave, M. (2008). Assessing concerns and issues about the mediation of technology in cyberbullying. Cyberpsychology: Journal of Psychosocial Research on Cyberspace, 2(2), article 1. Retrieved October 18, 2010, from http://cyberpsychology.eu/view.php?cisloclanku=2008111401&article=1 Hinduja, S. & Patchin, J. W. (2010). Bullying, cyberbullying, and suicide. Archives of Suicide Research, 14, 206221.  93 Hinduja S, & Patchin J. W. (2016). State cyberbullying laws: A brief review of state cyberbullying laws and policies. Cyberbullying Research Center. Retrieved May 11, 2016, from http://www.cyberbullying.us/Bullying-and-Cyberbullying-Laws.pdf Hoadley, C. M., Xu, H., Lee, J. J., & Rosson, M. B. (2010). Privacy as information access and illusory control: The case of the Facebook News Feed privacy outcry. Electronic Commerce Research and Applications, 9, 5060. Holmstrom, A., & Kim, S. (2015). The mediating role of cognitive reattribution and reappraisal in the esteem support process. Communication Research, 42, 6086. Holt, M. K., & Espelage, D. L. (2007). Perceived social support among bullies, victims, and bully-victims. Journal of Youth and Adolescence, 36, 984994. Hu, W. (2010, October 1). Legal debate swirled over charges in a student's suicide. New York Times. Retrieved October 17, 2010, from http://www.nytimes.com/2010/10/02/nyregion/02suicide.html?_r=0 Hunter, S.C., & Borg, M. (2006). The influence of emotional reaction on help seeking by victims of school bullying. Educational Psychology, 26, 813826. Jolliffe, D., & Farrington, D. P. (2004). Empathy and offending: A systematic review and meta-analysis. Aggressive and Violent Behavior, 9, 441476. Jourard, S.M. (1964). The transparent self: Self-disclosure and wellbeing. New York: Van Nostrand Reinhold. Kashner, S. (2014, November). Both huntress and prey. Vanity Fair, 183, 130137. Kowalski, R. M., Giumetti, G. W., Schroeder, A. N., & Lattanner, M. R. (2014). Bullying in the digital age: A critical review and meta-analysis of cyberbullying research among youth. Psychological Bulletin, 140, 10731137. Kowalski, R. M., Limber, S. E., & Agatston, P. W. (2012). Cyberbullying: Bullying in the digital age (2nd ed.). Malden, MA: Wiley-Blackwell. Lazarus, R. S. (1991). Emotion and adaptation. New York: Oxford University Press. Lazarus, R. S. (1993). Coping theory and research: Past, Present, and future. Psychosomatic Medicine, 55, 234247. Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer. Martin, R. C., & Dahlen, E. R. (2005). Cognitive emotion regulation in the prediction of depression, anxiety, stress, and anger. Personality and Individual Differences, 39, 12491260. Matsunaga, M. (2011). Underlying circuits of social support for bullied victims: An appraisal-based perspective on supportive communication and postbullying adjustment. Human Communication Research, 37, 174206.  94 McKenna, K. Y. A., & Bargh, J. A. (2000). Plan 9 from cyberspace: The implications of the Internet for personality and social psychology. Personality and Social Psychology Review, 4, 5775. Miller, P. A., & Eisenberg, N. (1988). The relation of empathy to aggressive and externalizing/antisocial behavior. Psychological Bulletin, 103, 324344. Moreno, M. A., Parks, M. R., Zimmerman, F. J., Brito, T. E., & Christakis, D. A. (2009). Display of health risk behaviors on MySpace by adolescents. Archives of Pediatric and Adolescent Medicine, 163, 2734. Morgan, E. M., Snelson, C., & Elison-Bowers, P. (2010). Image and video disclosure of substance use on social media websites. Computers and Human Behavior, 26, 14051411. Newton, P. (2013, April 10). Canadian teen commits suicide after alleged rape, bullying. CNN. Retrieved September 20, 2013, from http://edition.cnn.com/2013/04/10/justice/canada-teen-suicide/index.html Olweus, D. (2003). A profile of bullying at school. Educational Leadership, 60, 1217. Ortega, R., Elipe, P., Mora-Merchán, J. A., Genta, M. L., Brighi, A., Guarini, A., Smith, P. K., Thompson, F. and Tippett, N. (2012). The emotional impact of bullying and cyberbullying on victims: A European cross-national study. Aggressive. Behavior, 38, 342356. Pabian, S., & Vandebosch, H. (2016). An investigation of short-term longitudinal associations between social anxiety and victimization and perpetration of traditional bullying and cyberbullying. Journal of Youth and Adolescence, 45, 328339. Parrott, W.G., & Harré, R. (1996). Embarrassment and the threat to character. In R. Harré & W.G. Parrott (Eds.) The emotions: Social, cultural and biological dimensions (pp. 3956). Thousand Oaks, CA: Sage Publications. Patchin, J. W., & Hinduja, S. (2006). Bullies move beyond the schoolyard: A preliminary look at cyberbullying. Youth Violence and Juvenile Justice, 4, 148169. Patchin, J. W., & Hinduja, S. (2010). Cyberbullying and self-esteem. Journal of School Health, 80, 614621. Patchin, J. W., & Hinduja, S. (2013). Cyberbullying facts. Retrieved on May 1, 2016, from http://cyberbullying.org/facts Patchin, J. W., & Hinduja, S. (2015). Summary of our cyberbullying research (2004-2015). Retrieved on May 1, 2016, from http://cyberbullying.org/summary-of-our-cyberbullying-research Rainie, L., Brenner, J., & Purcell, K. (2012). Photos and videos as social currency online. Washington, DC: Pew Internet & American Life Project.  95 Regan, D., & Totten, J. (1975). Empathy and attribution: Turning observers into actors. Journal of Personality and Social Psychology, 32, 850856. Savitsky, K., Epley, N., & Gilovich, T. (2001). Do others judge us as harshly as we think? Overestimating the impact of our failures, shortcomings, and mishaps. Journal of Personality and Social Psychology, 81, 4456. Shen, L., & Dillard, J. P. (2007). The influence of BIS/BAS and message framing on the processing of persuasive health messages. Communication Research, 34, 433467. Shultz, E., Heilman, R., & Hart, K. J. (2014). cyberbullying: An exploration of bystander behavior and motivation. Cyberpsychology: Journal of Psychosocial Research on Cyberspace, 8(4), article 3. Slonje, R. & Smith, P. K. (2008). Cyberbullying: Another main type of bullying? Scandinavian Journal of Psychology, 49, 147154.  Slonje, R., Smith, P. K., & Frisén, A. (2013). The nature of cyberbullying, and strategies for prevention. Computers in Human Behavior, 29, 2632. Smith, C. A., & Lazarus, R. S. (1993). Appraisal components, core relational themes, and the emotions. Cognition and Emotion, 7, 233269. Smith, P. K., Mahdavi, J., Carvalho, M., Fisher, S., Russell, S., & Tippett, N. (2008). Cyberbullying: its nature and impact in secondary school pupils. Journal of Child Psychology and Psychiatry, 49, 376385. Steffgen, G., , A., Pfetsch, J., & Melzer, A. (2011). Are cyberbullies less empathic? Cyberpsychology, Behavior, and Social Networking, 14, 643648. Straude-Müller, F., Hansen, B., & Voss, M. (2012). How stressful is online victimization? European Journal of Developmental Psychology, 9, 260274. Suler, J. (2004). The online disinhibition effect. CyberPsychology & Behavior, 7, 321326. Tokunaga, R. S. (2010). Following you home from school: A critical review and synthesis of research on cyberbullying victimization. Computers in human behavior, 26, 277287. Tufekci Z. (2008). Can you see me now? Audience and disclosure management in online social network sites. Bulletin of Science, Technology & Society, 18, 2036. Van Cleemput, K., Vandebosch, H., & Pabian, S. (2014). Personal characteristics and contextual  cyberbullying. Aggressive behavior, 40, 383396. van Geel, M., Vedder, P., & Tanilon, J. (2014). Relationship between peer victimization, cyberbullying, and suicide in children and adolescents: a meta-analysis. JAMA Pediatrics, 168, 435442.  96 Walsh, M. (2012, April 12). Three California teens arrested for rape that allegedly drove teen girl to suicide, parents demand justice. New York Daily News. Retrieved May 17, 2012, from http://www.nydailynews.com/news/crime/teen-commits-suicide-rape-endless-bullying-family-article-1.1314786#ixzz2YhjOsd6Q Walther, J. B., DeAndrea, D., Kim, J., & Anthony, J. C. (2010). The influence of online comments on perceptions of antimarijuana public service announcements on YouTube. Human Communication Research, 36, 469492. Wang, J., Nansel, T. R., & Iannotti, R. J. (2011). Cyber and traditional bullying: Differential association with depression. Journal of Adolescent Health, 48, 415417. Watson, D., & Friend, R. (1969). Measurement of social-evaluative anxiety. Journal of Consulting and Clinical Psychology, 33, 448457. What is Cyberbullying. (n.d.). Retrieved March 05, 2016, from http://www.stopbullying.gov/cyberbullying/what-is-it/index.html  Willard, N. E. (2007). cyberbullying and cyberthreats: Responding to the challenge of online social aggression, threats, and distress. Champaign, IL: Research Press. Wright, M. F. (2016). Cyber victimization on college campuses-longitudinal associations with suicidal ideation, depression, and anxiety. Criminal Justice Review, 41, 190203.  Ybarra, M. L. (2004). Linkages between depressive symptomatology and Internet harassment among young regular Internet users. CyberPsychology & Behavior, 7, 247257. Ybarra, M. L., Mitchell, K. J., Wolak, J., & Finkelhor, D. (2006). Examining characteristics and associated distress related to Internet harassment: Findings from the second youth Internet safety survey. Pediatrics, 118, 11691177.