Computers in Human Behavior 71 (2017) 189e195 Contents lists available at ScienceDirect Computers in Human Behavior journal homepage: www.elsevier.com/locate/comphumbeh Parental mediation, cyberbullying, and cybertrolling: The role of gender Michelle F. Wright a, b, * a b Department of Psychology, Pennsylvania State University, USA Faculty of Social Studies, Masaryk University, Czech Republic a r t i c l e i n f o a b s t r a c t Article history: Received 20 October 2016 Received in revised form 26 January 2017 Accepted 28 January 2017 Available online 3 February 2017 Researchers are concerned with identifying the risk and protective factors associated with adolescents' involvement in cyberharassment. One such factor is parental mediation of children's electronic technology use. Little attention has been given to how different parental mediation strategies (i.e., instructive, restrictive, co-viewing) relate to adolescents' cyberharassment (i.e., cyberbullying, cybertrolling) perpetration and victimization, and how gender might impact these associations. To this end, the present study examined the moderating effect of gender on the longitudinal association between parental mediation strategies and adolescents' cyberharassment involvement. This focus was examined using a sample of 568 (53% female) adolescents from the 8th grade in the United States. The association between restrictive mediation and cybervictimization was more positive for girls when compared to boys, while the relationship between instructive mediation and cybervictimization was more negative for girls than for boys. The findings underscore the importance of considering parents in adolescents' cyberharassment involvement. © 2017 Elsevier Ltd. All rights reserved. Keywords: Cybertrolling Cyberbullying Cybervictimization Adolescent Parental mediation 1. Introduction Cyberharassment is a risk factor associated with adolescents' electronic technology use. Cyberharassment involves an assortment of behavioral risks, including cyberbullying and cybertrolling (Bhem-Morawitz & Schipper, 2015). To mitigate the risk factors associated with adolescents' electronic technology use, many parents have increased their supervision of their children's online lives. Parental mediation reduces adolescents experience of cybervictimization (Lwin, Stanaland, & Miyazaki, 2008; Mesch, 2009; Wright, 2015). Less attention has been given to how parental mediation relates to adolescents' cyberbullying perpetration, although a few studies have found negative associations (Chng, Liau, Khoo, & Li, 2014). It is also unclear how various parental mediation strategies (i.e., restrictive, co-viewing, instructive) might reduce adolescents' risk of cyberbullying involvement. Little attention has been given to predictors of cybertrolling, and parental mediation might be one such predictor. Girls typically report more parental mediation (Livingstone & Helsper, 2008). * Department of Psychology, Child Study Center, Pennsylvania State University, University Park, PA 16801, USA. E-mail address: mfw5215@psu.edu. http://dx.doi.org/10.1016/j.chb.2017.01.059 0747-5632/© 2017 Elsevier Ltd. All rights reserved. Thus, it is also important to understand how gender might impact the relationships between parental mediation and cyberbullying and cybertrolling. To this end, the present study examined the moderating effect of gender on these associations, utilizing a longitudinal design and different types of parental mediation strategies. 2. Literature review 2.1. Cyberharassment Cyberharassment is defined as “computer-mediated obscene comments, sexual harassment, and generally harassing behaviors aimed at debasing and/or driving out a virtual world user” (BhemMorawitz & Schipper, 2015 p. 2). Such behaviors can include flaming, nasty comments in chatrooms, offensive emails, or harassing blog posts. Furthermore, cyberharassment is a broad term for an array of negative and hostile online behaviors, including cyberbullying and cybertrolling. Typically conceptualized as an extension of face-to-face bullying, cyberbullying involves malicious behaviors carried out through a variety of information and communication technologies, such as instant messaging tools, social networking sites, gaming 190 M.F. Wright / Computers in Human Behavior 71 (2017) 189e195 consoles, text messages and cell phones, email, and websites (Smith et al., 2008). These malicious behaviors are often tormenting threatening, and/or perceived as harassing by a victim or a group of victims (Dooley, Pyzalski, & Cross, 2009; Grigg, 2010). Other examples of cyberbullying include sharing secrets about the victim, spreading nasty rumors, and threatening to harm someone (Ybarra & Mitchell, 2004). Similar to traditional forms of face-to-face bullying, cyberbullying often involves repetition of the behavior or behaviors and sometimes an imbalance of power between the victim and the perpetrator. The concept of repetitiveness is complex in the cyber context because it is possible to share harassing messages or videos one time or multiple times and with one person or multiple people (Nocentini et al., 2010; Smith, Del Barrio, & Tokunaga, 2013; Vandebosch & van Cleemput, 2008). In addition, recipients of such messages or videos can also choose to re-share such content. Cyberbullying has received increased attention because of its linkages to depression, anxiety, loneliness, suicidal ideation and attempts, poor academic performance, and alcohol and drug use (Bauman, Toomey, & Walker, 2013; Campbell, Spears, Slee, Butler, & Kift, 2012; Huang & Chou, 2010; Wright, 2016). Gender differences in cyberbullying perpetration and cybervictimization are mixed. Some researchers (e.g., Boulton, Lloyd, Down, & Marx, 2012; Li, 2007; Ybarra, Diener-West, & Leaf, 2007) have found that boys were more often cyberbullies, while girls were more likely than boys to report cybervictimization (e.g., Adams, 2010; Hinduja & Patchin, 2007; Kowalski & Limber, 2007). However, opposite patterns have been found in other research. In particular, girls in some samples have self-reported more cyberbullying perpetration when compared to boys (e.g., Cao & Lin, 2015; Dehue, Bolman, & Vollink, 2012; Pornari & Wood, 2010). Other researchers (e.g., Akbulut, Sahin, & Eristi, 2010; Del Rey et al., 2016; Sjurso, Fandream, & Roland, 2016) have found that boys were more often cybervictims than girls. No gender differences in cyberbullying involvement have been found in other research as well (e.g., Beran & Li, 2005; Didden et al., 2009; Marcum, Higgins, Freiburger, & Ricketts, 2012; Wright & Li, 2013). Gender has continued to be an inconsistent predictor of cyberbullying involvement. Cybertrolling is defined as “the practice of behaving in a deceptive, destructive, or disruptive manner in a social setting on the internet with no apparent instrumental purpose” (Buckels, Trapnell, & Paulhus, 2014). Few studies have focused on examining cybertrolling. In one study, Shachaf and Hara (2010) identified boredom, attention seeking, revenge, pleasure, and a desire to cause damage to the community as underlying motivations for engaging in cybertrolling. In addition, aggression, deception, disruption, and success have also been identified as characteristics of cybertrolling (Hardarker, 2010). Cybertrolling is distinguishable from cyberbullying or other forms of cyberharassment due to the pointlessness and disruptiveness of such behavior (Buckels et al., 2014). The intention of cybertrolls is not typically well-known, unlike cyberbullying in which the perpetrator is intending to harm the victim or victims (Lenhardt, 2013). Little attention has been given to the predictors of cybertrolling. Buckels et al. (2014) conducted one of the few studies to examine predictors of cybertrolling among adults. Their focus was on the association between cybertrolling and the Dark Triad of Personality (i.e., sadism, psychopathy, Machiavellianism, narcissism). Findings revealed that cybertrolling was related positively to sadism, psychopathy, and Machiavellianism. Furthermore, sadism was the one personality trait that was consistently associated with cybertrolling. Consequently, Buckels et al. suggest that cybertrolling might be an example of everyday sadism. It is unclear how gender might impact adolescents’ involvement in cybertrolling as little attention has been given to this topic. Additional research should focus on the potential risk and protective factors associated with cybertrolling as well as whether such factors might differ from cyberbullying and how gender could influence such differences. 2.2. Parental mediation Investigating parental mediation in relation to cyberbullying and cybertrolling provides the opportunity to investigate both risk and protective factors associated with these behaviors. Defined as using an assortment of prevention strategies to manage children's relationship with electronic media, parental mediation involves setting rules concerning their children's use of electronic media (Livingstone & Helsper, 2008). Such strategies might involve discussing appropriate use and setting limits on their children's electronic media use (Dehue et al., 2012). For parental mediation to be effective, parents must maintain an open dialogue with their children concerning appropriate content to view and how to use online tools. Some parents set rules regarding negative online behaviors, like cyberbullying. Three parental mediation strategies have also been identified in the literature, including restrictive, coviewing, and instructive (Arrizabalaga-Crespo, Aierbe-Barandiaran, & Medrano-Samaniego, 2010). Restrictive mediation is defined as parents' use of strategies employed to prevent children's access to certain online content. On the other hand, co-viewing mediation involves parents accessing online content with their children, although they might not discuss the content that they encounter. Parents who use instructive mediation maintain an active and continuing dialogue with their children regarding online content and the risks associated with electronic technology use. Attention has been given to parental mediation strategies as some strategies are effective for reducing adolescents' cyberbullying involvement. In one study, Mesch (2009) found that parents' use of monitoring strategies concerned their children's access to appropriate and inappropriate websites. Specifically, parents discussed which websites their children were allowed to visit. When such strategies were implemented, their children experienced lower levels of cybervictimization. Other research has corroborated the negative relationship between cybervictimization and high levels of parental mediation (Chng et al., 2014; Wright, 2015). In their study, Chng et al. (2014) examined active and restrictive parental mediation, and the role of these strategies in cyberbullying. Both parental mediation strategies reduced adolescents' risk of cyberbullying perpetration. However, these patterns were found for boys only and not girls, although girls reported higher levels of parental mediation when compared to boys. Such a finding suggests the importance of also considering gender when examining parental mediation strategies and how such strategies reduce adolescents' risks. They did not investigate how these various parental mediation strategies might mitigate adolescents' risk of cybervictimization. It is unknown how parental mediation might influence cybertrolling. Like cyberbullying, parental mediation might reduce adolescents’ risk of cybertrolling. Given that some parents communicate with their children about appropriate online behavior, it is likely that cybertrolling might be another behavior which is discussed with their children. 2.3. The present study Follow-up research should be conducted to better understand the longitudinal relationships between parental mediation strategies (i.e., restrictive, co-viewing, instructive) and cyberbullying and cybertrolling perpetration and victimization, and how gender might impact these associations. To this end, the present study had the following aims: M.F. Wright / Computers in Human Behavior 71 (2017) 189e195 (1) to examine the relationships between parental mediation strategies and cyberbullying perpetration and victimization. (2) to investigate the relationships between parental mediation strategies and cybertrolling perpetration and victimization. (3) to examine the moderating role of gender in these associations. The pattern of these relationships were examined over one year among adolescents. Data were first collected in the Spring of 7th grade and the next set of data were collected in the Spring of 8th grade. 3. Methods 3.1. Participants The participants for this study were 568 (52% female; M age ¼ 13.48 years) 8th graders from the Midwestern area of the United States. Their schools were located within predominantly middle-class neighborhoods. Participants primarily self-identified as White (73%), Latino/a (20%), Black/African American (5%), Asian (1%), and biracial (1%). Around 20%e31% of students at the schools qualified for free or reduced cost lunch. No income data was gathered. 3.2. Procedures Schools were recruited from a list of over 150 public middle schools located in the suburbs of a large Midwestern city. Of this list, 10 middle schools were randomly selected. The school principals at these 10 middle schools were sent an email. The email described the study's purpose, what adolescents would be expected to do, and how long adolescents would participate in the study. Six school principals expressed their intention to have their students participate in the study. The principal investigator setup a meeting with school principals and 7th and 8th grade teachers to discuss the project, what adolescents would do if they were to participate, and the length of data collection. Classroom announcements were made to made to the 7th grade classes, which explained how important adolescents' participation is, what they would do, the confidentiality and privacy of their answers, and how long they would participate in the study. A total of 713 parental permission slips were sent home with adolescents. The parental permission slip explained the nature of the study. There were 578 parental permission slips returned with permission, 13 that were returned without permission, and 2 that were unreturned. During the Spring of 7th grade, adolescents participated in the first wave of data collection (Time 1; T1). Some adolescents (n ¼ 11) were absent on the first day of data collection. Of these, 10 were able to complete the questionnaires on a make-up day. Before completing the questionnaires, adolescents provided their assent and they were allowed to ask questions in private, if they desired. They completed questionnaires on their demographic background (i.e., age, gender, ethnicity), self-reported cyberharassment (i.e., cyberbullying perpetration, cybervictimization, cybertrolling perpetration, cybertrolling victimization), and parental mediation of their technology use at the first wave. A total of 577 adolescents participated during the first wave of data collection. One year later, during the Spring of 8th grade, adolescents participated in the second wave of data collection (Time 2; T2). A letter was sent home to parents reminding them on their child's participation in the study. Parents wishing to no longer have their child participate in the study were asked to return the letter back to their child's school with their child's name written on it. There were no letters returned to the school. For the second wave of data 191 collection, ten adolescents did not participate because they had moved away. A total of 568 adolescents participated at the second wave. They completed the self-reported cyberharassment questionnaire. The study was approved by the principal investigator's university's IRB. Permission was also granted from three of the school districts, while the other three schools had principal permission. APA ethical standards were followed throughout this study. 3.3. Measures 3.3.1. Parental mediation strategies Adolescents were asked how often they perceived their parents as being involving in their technology use (Arrizabalaga-Crespo et al., 2010). There were 8 items included for this questionnaire, which were rated on a scale of 1 (completely disagree) to 5 (completely agree). The questionnaire included three subscales: restrictive mediation, co-viewing mediation, and instructive mediation. Four items were included for restrictive mediation, including “My parents impose a time limit on the amount of time that I surf the internet”. Three items were included on the coviewing mediation subscale. A sample item from this subscale included “My parents surf the internet with me”. There were two items included for the instructive mediation subscale (e.g., “My parents show me how to use the internet and warn me about the risks). This questionnaire was administered in the 7th grade only. Cronbach's alphas were 0.88 for restrictive, 0.83 for co-viewing, and 0.80 for instructive. 3.3.2. Self-reported cyberharassment This questionnaire assessed how often adolescents perpetrated and experienced cyberbullying and cybertrolling, yielding four subscales: cyberbullying perpetration, cybervictimization, cybertrolling perpetration, and cybertrolling victimization. Sixteen items were included on this questionnaire (Wright & Li, 2013). There were 12 items for cyberbullying, 6 for cyberbullying perpetration (e.g., “I called someone insulting names online or through text messages”) and 6 for cybervictimization (e.g., “Someone called me insulting names online or through text messages”). Four items were included for cybertrolling, 2 for cybertrolling perpetration (i.e., purposefully upsetting people online with the intent to trigger an emotional response, purposefully starting arguments online with the intent to trigger an emotional response) and 2 for cybertrolling victimization (i.e., were upset by something someone purposefully wrote online to get an emotional response, were upset by a purposeful argument online intended to trigger an emotional response). This questionnaire was administered in the 7th and 8th grades. All items were rated on a scale of 1 (Never) to 5 (All of the time). Cronbach's alphas were 0.88 for T1 cyberbullying perpetration, 0.87 for T2 cyberbullying perpetration, 0.92 for T1 cybervictimization, 0.92 for T2 cybervictimization, 0.78 for T1 cybertrolling perpetration, 0.77 for T2 cybertrolling perpetration, and 0.74 for T1 and T2 cybertrolling victimization. 3.4. Analytic plan Before conducting the analyses, analysis of variances (ANOVA) were conducted to examine differences among the six schools on the variables of restrictive mediation, co-viewing mediation, instructive mediation, T1 and T2 cyberbullying perpetration, T1 and T2 cybervictimization, T1 and T2 cybertrolling perpetration, and T1 and T2 cybertrolling victimization. No differences were found among any of the schools concerning each of these variables. Readers interested in more information about these analyses are urged to contact the author. Confirmatory factor analysis was 192 M.F. Wright / Computers in Human Behavior 71 (2017) 189e195 conducted to examine the measurement model using Mplus 7.3. The model had adequate fit, c2 ¼ 688.04, df ¼ 703, p ¼ n.s., CFI ¼ 0.98, TLI ¼ 0.97, RMSEA ¼ 0.04, SRMR ¼ 0.03. The factor loadings had good magnitudes and were significant (ps < 0.001). The items served as indicators for the latent variables in the structural regression model. For the structural regression model, paths were added from T1 parental mediation strategies (i.e., restrictive, instructive, co-viewing) to T2 cyberharassment (i.e., cyberbullying perpetration, cybervictimization, cybertrolling perpetration, cybertrolling victimization). T1 cyberharassment variables were allowed to predict their respective T2 variable. Gender was added as a moderator between T1 parental mediation strategies and T2 cyberharassment. Interactions were further examined using the Interaction program, if the interaction was significant. This program provides the significance of the unstandardized simple regression slopes and display graphical representations of the simple slopes. 4. Results Descriptive statistics and correlations were conducted among all variables in the study (see Table 1). The mean of restrictive mediation was 2.76, and 2.63 and 2.58 for co-viewing mediation and instructive mediation, respectively. Means were also conducted for cyberbullying and cybertrolling perpetration and victimization. The following means were obtained: 2.03 for T1 cyberbullying perpetration, 2.67 for T1 cyberbullying victimization, 2.18 for T2 cyberbullying perpetration, 2.71 for T2 for cyberbullying victimization, 1.93 for T1 cybertrolling perpetration, 1.86 for T1 cybertrolling victimization, 1.99 for T2 cybertrolling perpetration, and 1.90 for cybertrolling victimization. All variables were correlated with each other, except for a few. T1 restrictive mediation was not associated with T1 co-viewing mediation and T2 cybertrolling victimization. T1 co-viewing mediation was not related to T2 cybertrolling victimization. In addition, T1 instructive mediation was not associated with T1 and T2 cyberbullying perpetration, and T1 and T2 cybertrolling perpetration. T1 cyber aggression perpetration was not related to T2 cybertrolling victimization. T2 cybervictimization was not associated with T2 cybertrolling perpetration and victimization. T1 cybertrolling perpetration was unrelated to T2 cybertrolling victimization. To address the study's aims, a structural regression model was created. The model demonstrated adequate fit, c2 ¼ 991.38, df ¼ 881, p ¼ n.s., CFI ¼ 0.97, TLI ¼ 0.96, RMSEA ¼ 0.04, SRMR ¼ 0.03 (see Fig. 1 for a model depiction and Table 2 for all statistics). T1 restrictive (b ¼ 0.23, p < 0.001) and co-viewing (b ¼ 0.30, p < 0.05) mediation were associated negatively with T2 Fig. 1. Structural regression model for the associations among Time 1 (T1) parental mediational strategies (i.e., restrictive, co-viewing, instructive) and Time 2 (T2) cyberbullying and cybertrolling perpetration and victimization. cyberbullying perpetration. On the other hand, T1 restrictive mediation was related positively with T2 cybertrolling perpetration and cybervictimization (b ¼ 0.30, p < 0.001). Furthermore, T1 coviewing mediation was associated negatively with T2 cybervictimization (b ¼ 0.19, p < 0.05) and T2 cybertrolling perpetration (b ¼ 0.26, p < 0.001). T1 instructive mediation was related negatively to T2 cybervictimization (b ¼ 0.19, p < 0.05) and cybertrolling victimization (b ¼ 0.32, p < 0.001). Gender was examined as a moderator in the relationships between parental mediation strategies and the different forms of cyberharassment (i.e., cyberbullying perpetration, cybervictimization, cybertrolling perpetration, cybertrolling victimization). The moderation effect was significant for T1 restrictive mediation and T2 cybervictimization (b ¼ 0.16, p < 0.05) and T1 instructive mediation and T2 cybervictimization (b ¼ 0.17, p < 0.05). The association between T1 restrictive mediation and T2 cybervictimization was more negative for girls. In contrast, the relationship between T1 instructive mediation and T2 cybervictimization was more positive for girls. Table 1 Correlation among variables in the study. 1. T1 Restrictive 2. T1 Co-viewing 3. T1 Instructive 4. T1 CBP 5. T1 CBV 6. T2 CBP 7. T2 CBV 8. T1 CTP 9. T1 CTV 10. T2 CTP 11. T2 CTV M (SD) 1 2 3 4 5 6 7 8 9 10 11 e 0.03 -0.25** -0.32*** 0.41*** -0.35*** 0.38*** 0.31*** 0.17* 0.24** 0.10 2.76 (0.81) e -0.33*** -0.33*** -0.25** -0.30*** -0.24** -0.30*** -0.16* -0.29** 0.08 2.63 (0.98) e 0.11 -0.28*** 0.08 -0.25** 0.05 -0.33*** 0.04 -0.29*** 2.58 (1.01) e 0.26** 0.39*** 0.42*** 0.29*** 0.17* 0.33*** 0.13 2.03 (0.55) e 0.29*** 0.33*** 0.23** 0.29*** 0.16* 0.25** 2.67 (0.61) e 0.26** 0.29*** 0.16* 0.25** 0.20* 2.18 (0.69) e 0.18* 0.19* 0.13 0.10 2.71 (0.60) e 0.20* 0.36*** 0.15 1.93 (0.61) e 0.17* 0.33*** 1.86 (0.81) e 0.18* 1.99 (0.70) e 1.90 (0.73) Note. T1 ¼ Time 1; T2 ¼ Time 2; CBP ¼ Cyberbullying perpetration; CBV ¼ Cyberbullying victimization; CTP ¼ Cybertrolling perpetration; CTV ¼ Cybertrolling victimization. * p < 0.05. **p < 0.01. ***p < 0.001. M.F. Wright / Computers in Human Behavior 71 (2017) 189e195 193 Table 2 Standardized coefficients for structural regression model. Main Effects T1 Restrictive Mediation T1 Co-Viewing Mediation T1 Instructive Mediation Interactions T1 Restrictive Mediation  Gender T1 Co-Viewing Mediation  Gender T1 Instructive Mediation  Gender Covariates T1 Cyberbullying perpetration T1 Cyberbullying victimization T1 Cybertrolling perpetration T1 Cybertrolling victimization T2 Cyberbullying perpetration T2 Cyberbullying victimization T2 Cybertrolling perpetration T2 Cybertrolling victimization 0.23*** 0.30*** 0.05 0.30*** 0.19* 0.31*** 0.18* 0.26** 0.11 0.03 0.13 0.32*** 0.02 0.03 0.01 0.16* 0.10 0.17* 0.05 0.11 0.10 0.03 0.02 0.05 0.43*** e e e e 0.48*** e e e e 0.44*** e e e e 0.32*** Note. T1 ¼ Time 1; T2 ¼ Time 2; CBP ¼ Cyberbullying perpetration; CV ¼ Cybervictimization; CTP ¼ Cybertrolling perpetration; CTV ¼ Cybertrolling victimization. * p < 0.05. **p < 0.01. ***p < 0.001. 5. Discussion This study investigated the associations between parental mediation strategies (i.e., instructive, co-viewing, restrictive) and cyberharassment (i.e., cyberbullying and cybertrolling perpetration and victimization). The moderating effect of gender in these relationships was also examined. Consistent with the previous literature, some parental mediation strategies were related positively to cyberharassment, while some strategies were associated negatively with these behaviors (Chng et al., 2014; Wright, 2015). Parental mediation might function as a form of social support (Livingstone,  €rzig, & Olafsson, Haddon, Go 2011). It provides adolescents with the opportunity to discuss various negative situations that they encounter through electronic technologies. Parents are able to discuss strategies that have the potential to mitigate or reduce their children's exposure to cyberharassment (Wright, 2015). Furthermore, parents who take an active role in their children's digital lives might also provide guidance on appropriate online behaviors. Such guidance might help to reduce adolescents' perpetration of cyberharassment. Actively taking a role in adolescents' digital lives might further encourage adolescents to seek out social support and guidance from their parents regarding the potential exposure and experience of various online risks (Nikken & de Haan, 2015; Talves & Kalmus, 2015). There were a few distinctive findings from the present study. For instance, T1 restrictive mediation was related negatively to T2 cyberbullying perpetration. On the other hand, this type of mediation was associated positively with T2 cybervictimization and cybertrolling victimization. Restrictive mediation concerns parents' use of strategies that prevent their children's access to online content (Arrizabalaga-Crespo et al., 2010). Such prevention of adolescents' online content seems to reduce their perpetration of cyberbullying one year later. It might be likely that parents are sources of information regarding what content adolescents should avoid online, and this knowledge could lead to future conversations about online activities (Chng et al., 2014; Nikken & de Haan, 2015). Restricting adolescents' online activities might reduce their opportunities for perpetrating negative behaviors. In contrast, restrictive mediation might not prevent cybervictimization and cybertrolling victimization. Although it might be that restrictive mediation strategies prevent adolescents from accessing various online content, this strategy might not involve sharing methods to mitigate the potential for victimization by cyberharassment. Restrictive mediation might be similar to the overprotective parenting style. The overprotective parenting style involves parenting behaviors that do not encourage children to develop autonomy, problem-solving skills, and social skills (Clarke, Cooper, & Creswell, 2013; Lereya, Samara, & Wolke, 2013). Overprotective parenting also involves fear. Such fear leaves children vulnerable to risks, particularly cyberharassment. The exposure to risky online behaviors is probably inevitable and parents who use restrictive mediation strategies might not allow their children to develop strategies for dealing with these situations (Smahel & Wright, 2014). Restrictive mediation might prevent adolescents from accessing content, which reduces their risk of perpetrating cyberbullying, while these strategies do not involve sharing methods for avoiding or dealing with these behaviors. Mesch’s (2009) study provided some evidence that parents enforce strict rules regarding their children's access to online content, but that they rarely discuss strategies for dealing with the exposure to negative online interactions. Follow-up research should be conducted on restrictive mediation and the differential associations of these strategies to cyberharassment perpetration and victimization. T1 co-viewing mediation was related negatively to all forms of T2 cyberharassment, except cybertrolling victimization. Coviewing mediation involves parents using electronic technologies with their children (Arrizabalaga-Crespo et al., 2010). Parents who use this strategy are not likely to discuss content with their children, though they might discuss ways to deal with such content. Such minimal levels of intervention might reduce adolescents’ risk of experiencing cybervictimization. Furthermore, co-viewing mediation involves high levels of monitoring and supervision, which might prevent adolescents from having the opportunity to engage in cyberhassment. T1 instructive mediation was negatively associated with T2 cyberharassment victimization. However, no significant associations were found between this form of mediation and T2 cyberharassment perpetration. Instructive mediation involves parents and their children being actively involved in communication about electronic technologies (Arrizabalaga-Crespo et al., 2010). This form of mediation involves parents who discuss the risk of using electronic technologies and various strategies for avoiding these risks (Mesch, 2009; Nikken & de Haan, 2015; Talves & Kalmus, 2015). Consequently, adolescents might avoid situations that could potentially be problematic and if they do encounter such situations they are able to mitigate the effects and potentially reduce their exposure in the future. These parents might also discuss ways to encourage appropriate online interactions and the consequences of engaging in negative online behaviors. Due to such discussions, adolescents might begin to think about their online actions before impulsively responding or initiating cyberharassment. Furthermore, parents who stress the consequences of negative online behaviors might also have adolescents who think about their actions before performing them and are concerned with the consequences 194 M.F. Wright / Computers in Human Behavior 71 (2017) 189e195 of their actions. We are better able to understand the relationships among the variables in this study by examining the moderating effects of gender. In particular, there were gender differences in the associations between T1 parental mediation and T2 cybervictimization. The association between T1 restrictive mediation and T2 cybervictimization was more positive for girls when compared to boys. On the other hand, the relationship between T1 instructive mediation and T2 cybervictimization was more negative for girls in comparison to boys. A potential reason for these findings might be that girls typically report more restrictive and instructive parental mediation when compared to boys (Chng et al., 2014; Mesch, 2009; Wright, 2015). Parents might believe that online interactions are potentially riskier for girls than for boys, due to their belief that girls might be more vulnerable to exploitation online. Such beliefs might increase the likelihood that parents implement mediational strategies with their daughters more often than with their sons. Moderation effects were not found for parental mediation, and cyberbullying perpetration, cybertrolling perpetration, and cybertrolling victimization. It is unclear why gender did not impact the association between parental mediation and cyberbullying perpetration and cybertrolling involvement. More research attention should be given to the role of parental mediation in cyberbullying perpetration. Most of the research focuses on how parental mediation can reduce adolescents' exposure to cyberbullying victimization. Restrictive, instructive, and co-viewing mediation strategies might involve more opportunities for parents to discuss strategies to mitigate or reduce adolescents' vulnerability to online risks, including cyberbullying. These strategies might not have a role in reducing adolescents' perpetration of cyberbullying. Given that cybertrolling often occurs for no apparent reason, it could be that neither girls or boys are likely to engage in these behaviors, despite their parents’ use of mediational strategies. Additional research should focus on cybertrolling and the predictors of these variables. This research should also focus on strategies to prevent or reduce cybertrolling. 5.1. Limitations and future directions The present study relied on self-reports to assess cyberharassment and parental mediation. Self-reports are subject to biases, and therefore follow-up research should include multiple informants to assess cyber behaviors and parental mediation strategies. The study involved a two-wave longitudinal design. Although general patterns were found concerning the associations among the variables, additional research should be conducted to better understand the relationships between these variables and how parental mediation might impact cyberharassment over time. Furthermore, parental mediation might change as children become adolescents. Therefore, longitudinal studies should be designed to investigate changes in parental mediation over time and how mediation might relate to cyberharassment. The quality of the parent-adolescent relationship and parenting styles might impact parental mediation and adolescents' cyberharassment involvement. Thus, additional research should be conducted to investigate how parent-adolescent relationships and parenting styles might influence the association between parental mediation and cyberharassment. Cultural background might also have a role in parents’ use of technology mediational strategies, and these strategies might also vary based on the gender of the child. Follow-up research should be conducted to better understand the role of culture in technology mediation. 6. Conclusions This study examined the moderation of gender in the associations between parental mediation (i.e., instructive, restrictive, co-viewing) and cyberharassment. The findings from the study suggested that the impact of parental mediation on adolescents' cyberharassment depends on the type of behavior. Furthermore, gender also had a role in these relationships. Taken together, the results from this study indicate the importance of raising awareness of cyberharassment and how parental mediation might impact adolescents' online behaviors. Parents have a supportive role in adolescents' electronic technology use and such a role might prevent negative cyber behaviors. Furthermore, parents should be involved in their children's electronic technology use as such involvement reduces cyberbullying and cybertrolling perpetration and victimization. The findings from the present study underscore the importance of conducting more research on parenting youths in the digital age. References Adams, C. (2010). Cyberbullying: How to make it stop. Instructor, 120(2), 44e49. Akbulut, Y., Sahin, T. L., & Eristi, B. (2010). Cyberbullying victimization among Turkish online social utility members. Educational Technology & Society, 13(4), 192e201. Arrizabalaga-Crespo, C., Aierbe-Barandiaran, A., & Medrano-Samaniego, C. (2010). Internet uses and parental mediation in adolescents with ADHD. Revista Latina de Comunicacion, 65, 561e571. Bauman, S., Toomey, R. B., & Walker, J. L. (2013). Associations among bullying, cyberbullying, and suicide in high school students. Journal of Adolescence, 36(2), 341e350. http://dx.doi.org/10.1016/j.adolescence.2012.12.001. Beran, T., & Li, Q. (2005). Cyber-harassment: A new method for an old behavior. Journal of Educational Computing Research, 32(3), 265e277. Bhem-Morawitz, E., & Schipper, S. (2015). Sexing the avatar: Gender, sexualization, and cyber-harassment in a virtual world. Journal of Media Psychology: Theories, Methods, and Applications. http://dx.doi.org/10.1027/1864-1105/a000152. Boulton, M., Lloyd, J., Down, J., & Marx, H. (2012). Predicting undergraduates' selfreported engaged in traditional and cyberbullying from attitudes. CyberPsychology, Behavior, and Social Networking, 15(3), 141e147. http://dx.doi.org/ 10.1089/cyber.2011.0369. Buckels, E. E., Trapnell, P. D., & Paulhus, D. L. (2014). Trolls just want to have fun. Personality and Individual Differences, 67, 97e102. http://dx.doi.org/10.1016/ j.paid.2014.01.016. Campbell, M., Spears, B., Slee, P. H., Butler, D., & Kift, S. (2012). Victims' perceptions of traditional and cyberbullying, and the psychosocial correlates of their victimisation. Emotional and Behavioural Difficulties, 17(3e4), 389e401. http:// dx.doi.org/10.1080/13632752.2012.704316. Cao, B., & Lin, W. (2015). How do victims reach to cyberbullying on social networking sites? The influence of previous cyberbullying victimization experiences. Computers in Human Behavior, 52(C), 458e465. http://dx.doi.org/ 10.1016/j.chb.2015.06.009. Chng, G. S., Liau, A., Khoo, A., & Li, D. (2014). Parental mediation and cyberbullying e a longitudinal study. Annual Review of CyberTherapy and Telemedicine, 12, 98e102. Clarke, K., Cooper, P., & Creswell, C. (2013). The parental overprotection scale: Associations with child and parental anxiety. Journal of Affective Disorders, 151(2), 618e624. http://dx.doi.org/10.1016/j.jad.2013.07.007. Dehue, F., Bolman, C., Vollink, T., & Pouwelse, M. (2012). Cyberbullying and traditional bullying in relation to adolescents' perceptions of parenting. Journal of CyberTherapy & Rehabilitation, 5(1), 25e34. Del Rey, R., Lazuras, L., Casas, J., Barkoukis, V., Ortega-Ruiz, R., & Haralambos, T. (2016). Does empathy predict (cyber) bullying perpetration, and how do age, gender and nationality affect this relationship? Learning and Individual Differences, 45, 275e281. http://dx.doi.org/10.1016/j.lindif.2015.11.021. Didden, R., Scholte, R. H. J., Korzilius, H., de Moore, J. M. H., Vermeulen, A., O'Reilly, M., et al. (2009). Cyberbullying among students with intellectual and developmental disability in special education settings. Developmental Neurorehabilitation, 12(3), 146e151. http://dx.doi.org/10.1080/17518420902971356. Dooley, J. J., Pyzalski, J., & Cross, D. (2009). Cyberbullying versus face-to-face bullying: A theoretical and conceptual review. Journal of Psychology, 217, 182e188. http://dx.doi.org/10.1027/0044-3409.217.4.182. Grigg, D. W. (2010). Cyber-aggression: Definition and concept of cyberbullying. Australian Journal of Guidance and Counselling, 20(2), 143e156. http://dx.doi.org/ 10.1375/ajgc.20.2.143. Hardaker, C. (2010). Trolling in asynchronous computer-mediated communication: From user discussions to academic definitions. Journal of Politeness Research, 6, 215e242. http://dx.doi.org/10.1515/Jplr.2010.011. Hinduja, S., & Patchin, J. W. (2007). Offline consequences of online victimization. Journal of School Violence, 6(3), 89e112. http://dx.doi.org/10.1300/J202v06n03_ 06. Huang, Y., & Chou, C. (2010). An analysis of multiple factors of cyberbullying among junior high school students in Taiwan. Computers in Human Behavior, 26(6), M.F. Wright / Computers in Human Behavior 71 (2017) 189e195 1581e1590. http://dx.doi.org/10.1016/j.chb.2010.06.005. Kowalski, R. M., & Limber, S. P. (2007). Electronic bullying among middle school students. Journal of Adolescent Health, 41(6), 22e30. http://dx.doi.org/10.1016/ j.jadohealth.2007.08.017. Lenhardt, A. (2013). Teens, smartphones & texting. Washington, DC: Pew Research Center’s Internet & American Life Project. Retrieved from http://pewinternet. org/Reports/2012/Teens-and-smartphones.aspx. Lereya, S. T., Samara, M., & Wolke, D. (2013). Parenting behavior and the risk of becoming a victim and a bully/victim: A meta-analysis study. Child Abuse and Neglect, 37(12), 1091e1103. http://dx.doi.org/10.1016/j.chiabu.2013.03.001. Li, Q. (2007). Bullying in the new playground: Research into cyberbullying and cybervictimization. Australian Journal of Educational Technology, 23(4), 435e454.  €rzig, A., & Olafsson, Livingstone, S., Haddon, L., Go K. (2011). Risks and safety on the internet: The perspective of European children. EU Kids Online. London: LSE. Livingstone, S., & Helsper, E. J. (2008). Parental mediation of children's internet use. Journal of Broadcasting & Electronic Media, 52(4), 581e599. http://dx.doi.org/ 10.1080/08838150802437396. Lwin, M. O., Stanaland, A., & Miyazaki, A. D. (2008). Protecting children's privacy online: How parental mediation strategies affect website safeguard effectiveness. Journal of Retailing, 84(2), 2015e2217. http://dx.doi.org/10.1016/j.jretai.2008.04.004. Marcum, C. D., Higgins, G. E., Freiburger, T. L., & Ricketts, M. L. (2012). Battle of the sexes: An examination of male and female cyber bullying. International Journal of Cyber Criminology, 6(1), 904e911. Mesch, G. S. (2009). Parental mediation, online activities, and cyberbullying. CyberPsychology & Behavior, 12(4), 387e393. http://dx.doi.org/10.1089/cpb.2009.0068. Nikken, P., & de Haan, J. (2015). Guiding young children's internet use at home: Problems that parents experience in their parental mediation and the need for parenting support. CyberPsychology: Journal of Psychosocial Research on Cyberspace, 9. article 3. Nocentini, A., Calmaestra, J., Schultze-Krumbholz, A., Scheithauer, H., Ortega, R., & Menesini, E. (2010). Cyberbullying: Labels, behaviours and definition in three European countries. Australian Journal of Guidance and Counselling, 20(2), 129e142. http://dx.doi.org/10.1375/AJGC.20.2.129. Pornari, C. D., & Wood, J. (2010). Peer and cyber aggression in secondary school students: The role of moral disengagement, hostile attribution bias, and outcome expectancies. Aggressive Behavior, 36(2), 81e94. http://dx.doi.org/ 10.1002/ab.20336. Shachaf, P., & Hara, N. (2010). Beyond vandalism: Wikipedia trolls. Journal of Information Science, 36(3), 357e370. http://dx.doi.org/10.1177/016555150365390. 195 Sjurso, I. R., Fandream, H., & Roland, E. (2016). Emotional problems in traditional and cyber victimization. Journal of School Violence, 15(1), 114e131. http:// dx.doi.org/10.1080/15388220.2014.996718. Smahel, D., & Wright, M. F. (2014). Meaning of online problematic situations for children. Results of qualitative cross-cultural investigation in nine European countries. London: London School of Economics and Political Science. EU Kids Online. Smith, P. K., Del Barrio, C., & Tokunaga, R. S. (2013). Definitions of bullying and cyberbullying: How useful are the terms? In S. Bauman, D. Cross, & J. Walker (Eds.), Principles of cyberbullying research: Definitions, measures, methodology (pp. 26e40). New York, NY: Routledge. 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 Clinical Psychology and Psychiatry, 49(4), 376e385. http://dx.doi.org/10.1111/ j.1469-7610.2007.01846.x. Talves, K., & Kalmus, V. (2015). Gendered mediation of children's internet use: A keyhole for looking into changing socialization practices. CyberPsychology: Journal of Psychosocial Research on Cyberspace, 9. article 4. Vandebosch, H., & van Cleemput, K. (2008). Defining cyberbullying: A qualitative research into the perceptions of youngsters. CyberPsychology & Behavior, 11(4), 499e503. http://dx.doi.org/10.1089/cpb.2007.0042. Wright, M. F. (2015). Cyber victimization and adjustment difficulties: The mediation of Chinese and American adolescents' digital technology usage. CyberPsychology: Journal of Psychosocial Research in Cyberspace, 9. article 1. Retrieved from http://cyberpsychology.eu/view.php?cisloclanku¼2015051102&article¼1. Wright, M. F. (2016). Cyber victimization and substance use among adolescents: The moderation of perceived social support. Journal of Social Work Practice in the Addictions, 16(1e2), 93e112. http://dx.doi.org/10.1080/1533256X.2016.1143371. Wright, M. F., & Li, Y. (2013). The association between cyber victimization and subsequent cyber aggression: The moderating effect of peer rejection. Journal of Youth & Adolescence, 42(5), 662e674. http://dx.doi.org/10.1007/s10964-0129903-3. Ybarra, M. L., Diener-West, M., & Leaf, P. (2007). Examining the overlap in internet harassment and school bullying: Implications for school intervention. Journal of Adolescent Health, 41(6), 42e50. http://dx.doi.org/10.1016/ j.jadohealth.2007.09.004. Ybarra, M. L., & Mitchell, K. J. (2004). Online aggressor/targets, aggressors, and targets: A comparison of associated youth characteristics. Journal of Child Psychology and Psychiatry, 45(7), 1308e1316. http://dx.doi.org/10.1111/j.14697610.2004.00328.x. Health Communication ISSN: 1041-0236 (Print) 1532-7027 (Online) Journal homepage: http://www.tandfonline.com/loi/hhth20 Social Representation of Cyberbullying and Adolescent Suicide: A Mixed-Method Analysis of News Stories Rachel Young, Roma Subramanian, Stephanie Miles, Amanda Hinnant & Julie L. Andsager To cite this article: Rachel Young, Roma Subramanian, Stephanie Miles, Amanda Hinnant & Julie L. Andsager (2017) Social Representation of Cyberbullying and Adolescent Suicide: A Mixed-Method Analysis of News Stories, Health Communication, 32:9, 1082-1092, DOI: 10.1080/10410236.2016.1214214 To link to this article: https://doi.org/10.1080/10410236.2016.1214214 Published online: 26 Aug 2016. Submit your article to this journal Article views: 2120 View Crossmark data Citing articles: 6 View citing articles Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=hhth20 HEALTH COMMUNICATION 2017, VOL. 32, NO. 9, 1082–1092 http://dx.doi.org/10.1080/10410236.2016.1214214 Social Representation of Cyberbullying and Adolescent Suicide: A Mixed-Method Analysis of News Stories Rachel Younga, Roma Subramanianb, Stephanie Milesa, Amanda Hinnantb, and Julie L. Andsagerc a School of Journalism and Mass Communication, University of Iowa; bSchool of Journalism, University of Missouri; cSchool of Journalism and Electronic Media, University of Tennessee ABSTRACT Cyberbullying has provoked public concern after well-publicized suicides of adolescents. This mixed-methods study investigates the social representation of these suicides. A content analysis of 184 U.S. newspaper articles on death by suicide associated with cyberbullying or aggression found that few articles adhered to guidelines suggested by the World Health Organization and the American Foundation for Suicide Prevention to protect against suicidal behavioral contagion. Few articles made reference to suicide or bullying prevention resources, and most suggested that the suicide had a single cause. Thematic analysis of a subset of articles found that individual deaths by suicide were used as cautionary tales to prompt attention to cyberbullying. This research suggests that newspaper coverage of these events veers from evidence-based guidelines and that more work is needed to determine how best to engage with journalists about the potential consequences of cyberbullying and suicide coverage. In public discourse, cyberbullying and suicide are causally linked. In his eulogy for Missouri State Auditor Tom Schweich, whose death by suicide was associated with a rumor campaign about his religious affiliation, former U.S. Senator John C. Danforth (2015) described Schweich’s political opponents as bullies, saying “We read stories about cyberbullying, and hear about young girls who kill themselves because of it” (para. 31). In a blog post about Twitter “trolls” who posted violent and harassing statements about his daughter, former Major League pitcher Curt Schilling justified making the names of these Twitter users public by saying “Young women, and men, are KILLING themselves after being cyberbullied” (Schilling, 2015, para. 17). These examples and others suggest that the cyberbullying/suicide link propels public concern and discussion and may also be used to justify punitive actions against those who bully online. Cyberbullying is defined generally as bullying that occurs online, or more specifically, as an “aggressive, intentional act carried out by a group or individual, using electronic forms of contact, repeatedly and over time against a victim who cannot easily defend him or herself” (Smith et al., 2008, p. 376). The mediated nature of cyberbullying, which occurs using communication tools and platforms, distinguishes it from face-toface bullying (Ramirez, Eastin, Chakroff, & Cicchirillo, 2008). Cyberbullying is a significant public health concern because of its prevalence and its long-term negative health effects in both victims and bullies (Herz, Donato, & Wright, 2013; Karch, Logan, McDaniel, Floyd, & Vagi, 2013; Tokunaga, 2010). As with the parent term bullying, cyberbullying is used primarily to describe aggressive interactions among adolescents. Rates CONTACT Rachel Young, MPH, PhD rachel-young@uiowa.edu Adler Journalism Building, Iowa City, IA 52242. © 2017 Taylor & Francis of cyberbullying perpetration and victimization among adolescents vary substantially among studies, based on how the concept is defined and measured. A recent meta-analysis found that from 5.3% to 31.5% of adolescents reported bullying others online, while 2.2% to 56.2% reported having been bullied (Modecki, Minchin, Harbaugh, Guerra, & Runions, 2014). Since the term cyberbullying emerged less than a decade ago, the phenomenon has sparked news media coverage of adolescent suicides preceded by cyberbullying. Suicide is the third most common cause of death among 10- to 14-year-olds and the second leading cause of death among 15- to 24-year-olds (Centers for Disease Control and Prevention [CDC], 2011). Suicidal ideation is also common among adolescents. In 2013, 13.6% of adolescent respondents in the Youth Risk Behavior Surveillance Survey reported having made a plan about how they would attempt suicide, and 8.0% had attempted suicide in the year prior to the survey (CDC, 2014). From 2009 to 2013, the time span covered by this study, the rate of adolescents seriously considering suicide increased from 13.8% to 17.0% (CDC, 2014). A national epidemiological survey of nonlethal suicidal behavior in U.S. adolescents (e.g., suicidal ideation and attempts) found that most had preexisting mental disorders (Nock et al., 2013). News media coverage of suicide has been linked to increased suicide rates among adolescents and adults, a phenomenon called suicide contagion (Phillips, 1974; Romer, Jamieson, & Jamieson, 2006). Both the volume of coverage and specific elements of coverage have been linked to increased suicide rates. Problematic elements include providing identifying details about victims as well as explicit details about the methods of suicide, Assistant Professor, School of Journalism and Mass Communication, University of Iowa, 100 HEALTH COMMUNICATION describing suicide as an epidemic or as increasing, focusing attention on a single suicide, and suggesting that suicide has a single cause (Romer et al., 2006). Although no published content analyses have yet examined news stories about cyberbullying, news media coverage of bullying and cyberbullying has also been criticized as alarmist and oversimplified (Tenore, 2013). Media guidelines published at StopBullying.gov advise against simplifying the issue by blaming or criminalizing those who bully, overstating the problem, or implying that bullying is the sole cause of suicide. This study explores the social representation of suicides associated with cyberbullying in news articles, using content analysis to investigate the representation of these incidents on a larger scale. This study’s use of mixed methods allows us to look at representations in depth, as well as at how frequently aspects of those representations, shown in past research to be particularly salient in suicide coverage, occur in these news stories. Although bullying and cyberbullying often co-occur and it is thus impossible to draw a clear line between coverage of the two, this article focuses on cyberbullying because of the heightened public concern about it as a relatively new phenomenon and to discover how technologies that enable cyberbullying are portrayed in news media coverage. Literature Review Social Representation in News Media Social representations refer to shared cognitive schema or images that are developed and altered over time (Schmitz, Filippone, & Edelman, 2003). As compared to other explications of schema, the emphasis is on “shared” and on the social processes through which schema are set and reset (Sibley, Liu, & Kirkwood, 2006). Social representations function as “reference points for the selection, categorization and organization of social information and experiences” (Schmitz et al., 2003, p. 384), and they “give us a way of making sense of and so constituting socially significant phenomena” (Howarth, 2006, p. 69). Social representations are established and transformed through two processes: anchoring and objectification (Sammut & Howarth, 2014). Anchoring refers to the process through which social representations are developed by associating novel or unfamiliar ideas with familiar categories or images. By “identifying characteristics of the phenomenon that are familiar,” the “unfamiliar becomes embedded in the social classifications available in everyday communication” (Schmitz et al., 2003, p. 385). Objectification refers to the process through which social representations are set or reified. Moscovici (1984), who originally defined the term “social representations,” described objectifications as a means of turning “something abstract into something almost concrete” (p. 29). The concept of social representation has particular resonance in analysis of news media coverage and relates to how messages both “reflect and refract social attitudes” (Liebler, Schwartz, & Harper, 2009, p. 665). Bauer and Gaskell (1999) classified media as a type of formal communication through which social representations are developed and distributed. 1083 Analysis of news media coverage is commonly used as a method of discerning and describing social representations (e.g., Joffe & Haarhoff, 2002; Len-Ríos, Bhandari, & Medvedeva, 2014; Schmitz et al., 2003). Social representations may develop as a way of making sense of new threats. Joffe and Haarhoff (2002) identified social representations of Ebola. They found that British broadsheet newspapers emphasized structural features leading to Ebola’s spread in Africa, while tabloids depicted the disease as fundamentally African and in opposition to Western practices. In interviews with British adults, the salience of these representations differed by the type of newspaper participants read (Joffe & Haarhoff, 2002), a finding that echoes the often-described “frame setting” function of the news media (Scheufele, 1999). Social representations of phenomena in news media may also serve as prototypes (Schmitz et al., 2003). Gibbons and Gerrard (1997) explain how images of prototypical individuals influence behavior through social comparison, particularly among adolescents. News coverage of suicides that implicates cyberbullying as a causal factor may be especially salient to adolescents, who are positioned as both the victims and perpetrators of cyberbullying. News Media Coverage and Behavioral Contagion Although news about suicide does not in itself cause suicide, it can imply permission or invitation to those who are already at risk, a phenomenon called suicide contagion (Colt, 2006). This has sometimes been called a “copycat effect” or the “Werther effect,” after the protagonist in Goethe’s novel The Sorrows of Young Werther (1774), which is believed to have triggered imitation suicides after its release (Phillips, 1974). A meta-analysis of 42 studies found that this copycat effect was 4.03 times stronger for real than for fictional stories of suicide, and stronger for print media than for television (Stack, 2003). In addition, the behavioral contagion effect of suicide coverage in the mass media is strongest among adolescents (Gould, Jamieson, & Romer, 2003; Phillips & Carstensen, 1986). Suicide contagion operates through either of two mechanisms, by providing news audiences information about effective methods of killing themselves or by lowering vulnerable individuals’ inhibition to commit the act (Jamieson, Jamieson, & Romer, 2003). Research on whether news reports of suicide serve as a catalyst for suicide is somewhat inconsistent (Stack, 2005). A study of suicides in urban areas following news coverage of one suicide found evidence that the media may influence some individuals to end their lives, though for others it may provide some protection from doing so (Romer et al., 2006). The varied content of media suicide coverage may be one factor that determines suicide contagion effects. In Austria, news stories about individuals who had considered but had not attempted suicide were associated with a decrease in suicide rates, while stories that mentioned suicides were increasing were associated with an increase in suicide rates (Niederkrotenthaler et al., 2010). When suicide contagion does occur, the individuals most susceptible to its effects are likely to be those most demographically similar to the victim, though others may be affected as well (Gould et al., 2003). Thus, for young people, reporting 1084 R. YOUNG ET AL. deaths by suicide can be especially problematic as younger individuals are more likely to experience disinhibitory effects from negative media messages, according to social learning theory (Bandura, 2001). In a study of suicide coverage, 94% of articles contained identifiable suicidal individuals (Tatum, Canetto, & Slater, 2010). In other words, these articles provided enough social and personal details about the individual that readers might identify closely with them as suicide models. From a social learning perspective, these articles may be modeling suicidal behavior that could influence vulnerable readers who might attempt to imitate these behaviors. Preliminary research has suggested that Internet searches for keywords related to method increase dramatically after news reporting in which suicide methods are included (Chang, Page, & Gunnell, 2011). Journalists have been asked to exercise particular caution when reporting incidents of suicide because these articles may be more likely to include story elements found to be harmful, such as detailed descriptions of the victim, speculation on events preceding the suicide, or the method of suicide (Pirkis, Blood, Skehan, & Dare, 2010). Governmental, news media, public health, and advocacy organizations, including the World Health Organization (WHO, 2008) and the American Foundation for Suicide Prevention (AFSP) (reportingonsuicide.org), have released research-based guidelines that instruct reporters how to minimize suicide contagion effects. Guidelines warn against providing too much detail on the method by which an individual committed suicide and avoiding the tendency to pin the rationale for the act on a single event. Analyses of news media reports on suicide are mixed on whether recommendations have influenced how deaths by suicide are reported in news media (Tatum et al., 2010). A large-scale study of U.S. newspapers found that more than 60% of articles on suicide included the word in the headline, and more than half of the articles described the method (Gould, Midle, Insel, & Kleinman, 2007). Similarly, Tatum et al. (2010) found 77% of U.S. newspaper stories mentioned the suicide method, and preventative and risk factor information was relatively scarce. A majority of articles (69%) contained no information on suicide prevention, and only 1% of articles included suicide risk factors or warnings. There is some evidence from other countries that guidelines do improve news coverage. For example, news media coverage of suicide in the Australian media was evaluated to determine whether the distribution of the suicide reporting guideline document “Reporting Suicide and Mental Illness” had improved coverage of this issue from 2000 to 2007 (Pirkis et al., 2009). The study found that there was a significant improvement in the quality of reporting with regard to several variables such as decrease in the use of inappropriate language, decrease in noting in detail the method of suicide, and increase in providing information about help services. An evaluation of the effect of suicide reporting guidelines in Austria found that the guidelines not only improved the quality of reporting but also were associated with a decrease in the number of suicides (Niederkrotenthaler & Sonneck, 2007). A study to evaluate changes in suicide reportage in Hong Kong newspapers before and after the distribution of the WHO media guidelines—“Preventing Suicide: A Resource for Media Professionals”—as well as the launch of an awareness campaign found that there was an improvement in certain aspects of suicide coverage, particularly in tabloid newspapers; for example, fewer photographs were used following the recommendations (Fu & Yip, 2008). Inconsistent adherence to coverage guidelines is perhaps not surprising, given that the guidelines contradict many established news values of truth-telling, negativity, unexpectedness, and controversy (Friedman, Dunwoody, & Rogers, 1999; Shoemaker & Vos, 2009). Interview research with journalists found that although those who cover suicide are amenable to describing their work as being in the public good, they believe accurate depictions of suicide could deter individuals from taking the same action (Collings & Kemp, 2010; Subramanian, 2014), an interpretation that runs counter to research on suicide contagion. Indeed, journalists have expressed skepticism about the copycat suicide phenomenon, explaining that it is too simplistic to assume that media coverage can trigger suicides (Collings & Kemp, 2010; Subramanian, 2014). Also, in some stories, providing the details of a suicide is perceived as essential as it can serve to expose or hold accountable institutions, such as hospitals, that are perceived as having failed to protect high-risk individuals (Subramanian, 2014). In addition, guidelines that suggest avoiding identifying details to prevent copycat suicides may contradict the desires of friends and family, who discuss their desire for sympathetic stories that accurately represent the individual (Chapple, Ziebland, Simkin, & Hawton, 2013). A detailed analysis of a sample of stories describing individual deaths by suicide associated with cyberbullying allows us to examine whether these deaths coalesce in a coherent social representation and how this representation functions within the broader narrative of cyberbullying. A mixture of quantitative and qualitative analysis of the same news media content has the benefit of clarifying social representations and examining their distribution in a broader sample (Liebler et al., 2009; Schmitz et al., 2003). This work was guided by the following questions: ● To what extent do newspaper articles describing suicides include elements considered to be harmful based on guidelines for suicide coverage? ● What themes emerge in social representations of suicides associated with cyberbullying? Methods Sample The goal of the study was to investigate news coverage of suicides associated with cyberbullying or aggression online. The sample was collected by searching for the names of adolescents who died by suicide in which cyberbullying was publically identified as a factor. A Web and database search identified individuals using the terms “cyberbullying” and “suicide.” Incidents of individuals who committed suicide outside the U.S. were excluded. Also excluded were incidents when cyberbullying was preceded by an HEALTH COMMUNICATION alleged sexual assault because coverage of sexual assault is associated with additional problematic elements (Franiuk, Seefelt, & Vandello, 2008). The individuals used to comprise the final sample included four girls and two boys, ages 12 to 16 years, who died from 2009 to 2013. All incidents received some national and regional media coverage, though the amount of coverage varied. The Lexis-Nexis database was used to locate U.S. newspaper articles using the name of each individual in quotation marks, from the date of death until 1 year later, or until March 12, 2013, if the incident had occurred less than a year previously. Including newspapers that ranged in size and geographic location allowed for the examination of commonalities among articles from smaller daily newspapers close to where the suicides occurred, as well as national newspapers that covered incidents episodically or as exemplars in stories about cyberbullying. The initial search yielded 818 articles. Articles were included in the initial sample if the primary topic was determined to be suicide or if an individual who had died by suicide was used as an exemplar. An exemplar is an example, usually describing a specific case or incident, used to illustrate an issue (Zillmann & Brosius, 2000). Articles were also excluded if they were duplicates (n = 634), briefs (100,000) included the word “suicide” in the headline (small = 8.9%, medium = 11.3%, large = 34.9%; χ2 = 16.48; p < .001; Cramer’s V = 0.30). Mental illness was described as contributing to suicide in 7.5% of articles, while 92.5% did not list mental illness as a contributing factor. Again, there was a difference based on circulation category. A greater proportion of stories in larger newspapers mentioned mental illness as a contributing factor (small = 0%, medium = 1.9%, large = 15.1%; p = .001, χ2 = 13.07; Cramer’s V = 0.27). Few articles (12.8%) provided any mobilizing information related to bullying, mental health, or suicide prevention. The most common information given, either Web links or phone numbers, was from organizations dedicated to preventing bullying (8.0%), followed by suicide prevention or mental health organizations (3.2%), or other organizations (3.2%). Only 0.5% of articles described steps that could be taken to help those considering suicide, such as calling a hotline, seeking counseling, telling adults, or calling the police. Victim Actions Coders also assessed what victim actions were mentioned as having preceded cyberbullying: 8.0% mentioned sending HEALTH COMMUNICATION sexually explicit text messages or images; 15.0% mentioned a romantic relationship that led to conflict with peers; and 10.7% sexual orientation, specifically the action of making sexual orientation public. Technological Affordances The technological affordance mentioned most frequently was anonymity (11.8%), followed by forwardability (9.6%), and permanence of content online (5.9%). Textual Analysis In answer to the second research question, about the social representation of suicides associated with cyberbullying, four main themes emerged: (a) attribution of blame for suicide as a causal chain, from victim action to bullying online and offline to suicide; (b) attribution of blame for bullying shifting among targets, but settling primarily on schools; (c) the sense that cyberbullying is bullying amplified; and (d) publicized suicides function as cautionary tales or alarms that prompt action. Attribution of Blame for Suicide In the articles reviewed, cyberbullying was positioned as a much more salient predictor of suicide than other potential contributors, like mental illness or family troubles. Commonly used phrases such as “bullied to death” or “died after being bullied online” positioned suicide as having been triggered primarily by bullying. Stories also included digitally mediated remarks from bullies that referenced suicide, such as “drink bleach and die” (Sullivan & Hernandez, 2013). Articles about the most recent death in this sample described this suicide as another example of a familiar narrative rather than an extreme outcome, stating that the girl “became one of the youngest members of a growing list of children and teenagers apparently driven to suicide” (Alvarez, 2013a, p. 1A). This suggests that the social representation of cyberbullying as an action that triggers suicide when it occurs among adolescents is becoming more concrete, as described by the social representation process of objectification. The narrative is described here as a causal chain ending in suicide because many articles also described a triggering event that instigated the bullying. These events could be classified as disrupting peer social structures or challenging social—particularly gender—norms. For two girls whose suicides garnered the most news coverage in the sample, the bullying began, online and offline, in a conflict over a romantic relationship. For one girl who was harassed after dating several older boys, her younger age and social status were contrasted with the social power of the older, “popular” students who were her tormentors, primarily other girls. Another girl had sent a naked picture (or “sext”) to a boy who was dating another girl, and her online transgression was punished with harassment at school. Sources in the articles indicated that she perceived bullying as a natural punishment: “At the same time, friends say, [she] knew that the biggest mistakes made were her own” (Meacham, 2009, p. 1A). Both boys included in the analysis were targeted after making their sexual orientation public or because classmates perceived them as gay. In 1087 both cases, articles position “coming out” or “declaring his sexual orientation” as when the boys set themselves apart as targets. With one boy in particular, not only his sexual orientation but also his decision to announce, discuss, and defend it publically were seen as the factors that triggered online bullying. Attribution of Blame for Bullying Stories addressing deaths by suicide associated with cyberbullying also attribute blame for bullying. Because the aggressors were also minors, blame for stopping cyberbullying primarily accrued to individuals and institutions charged with overseeing their behavior, such as schools, counselors, or administrators. Schools were blamed for failing to intervene in this chain by first acknowledging and then stopping bullying, or, to a lesser extent, for failing to notice the suffering of bullying victims and not offering effective support. In several stories, contention centered on what schools knew about the bullying and when they knew it, and with whom their knowledge should be shared, such as parents and law enforcement. Some blame also accrued to peers, defined in criminal parlance as “witnesses,” who saw bullying but did not step in. For parents, the digital world where cyberbullying occurs is seen as foreign. Bereaved parents talk about how they “had no idea” about their child’s online torments and interactions, a result of the “digital disconnect” between parent and child. A parent who tried to delete apps from her daughter’s phone found that her daughter had migrated her online presence to new apps as a way to evade detection. Cyberbullying as Bullying Amplified The concept of bullying is the main anchor used to describe cyberbullying in these stories. Cyberbullying was seen as an amplification of bullying and was often described as an increasing trend that “has affected every school in every community” or as more frequent than offline bullying, though research does not always support that claim (Donelan, 2010, para. 10). In contrast, a researcher stated cyberbullying is not an “epidemic” but is instead a novel manifestation of “the ageold issue of relentless torment” (Vander Velde, 2013, p. 1A). The social representation of cyberbullying reflected tension between this phenomenon as familiar but repackaged and a concern that translation of aggressive behavior to the Internet has magnified its reach and damage. Stories alluded to the centrality of social media and cell phones in adolescent social life. The centrality is implied in conjecture that some individuals who eventually died by suicide failed to discuss cyberbullying with parents because they feared that they would lose access to technology and be cut off from a social lifeline, in addition to the fear that reporting to parents would make the problem worse. Some stories positioned offline bullying as seeming less threatening because it is more familiar or time-delimited. An attorney described how an 11-year-old preferred physical to online aggression: “At least in a physical fight, there’s a start and an end, but when the taunts and humiliation follow a child into their home, it’s ‘torture’ and it doesn’t stop” (Donelan, 2010, para. 2). The Internet and particularly social media are seen as fertile ground for bullying based on several 1088 R. YOUNG ET AL. features. Salient affordances are the permanence of content on the Internet, the ability for anonymous bullying, and the forwardability of content or its potential to spread quickly, like a nude picture meant for a boyfriend that is forwarded to peers. Both the portability and ubiquity of the Internet create an environment for amplifying bullying. Though there is a clear research-based definition of bullying, as intentional, sustained aggression from a stronger bully to weaker victim, it is unclear how this operationalization translates to the Internet, where power is not physical strength and one mean comment can be repeated and spread. This confusion is seen in articles negotiating concerns about cyberbullying. In one article, a sheriff describes it as “girl talk that goes further than girl talk” (Allen, 2013, para. 19). Cautionary Tales Another way in which stories sought to make sense of suicides is by describing a cautionary tale that might spark discussion of a problem’s severity and spurs a search for solutions, in the form of antibullying forums at schools and communities, as well as new legislation. In several states, a well-publicized suicide sparked antibullying legislation, sometimes named for the individual who had died by suicide. These individual incidents may be described as “watersheds” and “wake-up calls” that “put the issue in the spotlight” and signal to the public that “we have to do something.” Individual deaths were used as an opportunity to bring attention, raise funds, or redefine an issue. Prosecution of bullies whose actions are seen as having caused suicide is not merely punishment for those individuals but also deterrent to potential bullies. Said a sheriff, “We’ve raised awareness, and we’ve helped kids” by charging junior high students with felonies, even though the charges were dropped (Alvarez, 2013b, p. 14A). News stories are also cautionary tales because they reflect contradiction and confusion about how best to tackle the problem of bullying online. While in some cases schools were sued for failing to act, appropriate action often seemed undefined except by its outcome, of having stopped the bullying or the suicide. Legislation often codifies schools’ responsibilities for dealing with bullying on- and offline through prevention and punishment. The onus was on schools to determine what incidents classified as criminal and which could be handled internally. Bullies and victims were also consigned to clear categories, despite evidence that adolescents may switch between these roles online (e.g., Roberto, Eden, Savage, Ramos-Salazar, & Deiss, 2014a). Discussion Social representations are not static or singular, but are characterized by a “dialectic of cooperation and conflict” (Howarth, 2006, p. 71). A rise in online aggression among adolescents, or cyberbullying, has coincided with increasing public concern about the health effects of bullying and ambivalence about rampant technology use among youth (Tenore, 2013). This study used the social representations framework to examine how media coverage of suicides related to cyberbullying reflects social conflict about how cyberbullying should be defined and addressed. Then the study analyzes how these stories perpetuating social representations reproduce words and descriptions that suicide-reporting guidelines warn against using. In other words, this research explains how society journalistically tries to make sense of a new trend, cyberbullying, and its involvement in suicide deaths among youth. The scaffolding of these stories includes story elements that have been associated with the potential to normalize suicide as a response to cyberbullying. The textual analysis revealed that stories about suicides associated with cyberbullying exemplify anxieties about online bullying and serve as cautionary tales that focus attention on the issue and force action. The social representation was developed in news stories through an iterative process of media attention that coalesces around a particular representation, the process called objectification. Although it seems clear cyberbullying played a role in triggering the suicides included in this analysis, this research found evidence that stories may oversimplify suicides preceded by cyberbullying. While cyberbullying and bullying have been associated with increased suicidal ideation in surveys, teasing apart causality is complicated (Hinduja & Patchin, 2010). Other research suggests that bullying should be considered in conjunction with other risk factors for suicide, such as depression or other stressful life events (Saleh, Feldman, Grudzinskas, Ravven, & Cody, 2014). A content analysis of a larger sample of stories found that the focus on these individual exemplars of cyberbullying effects was also linked with several other story elements considered potentially harmful in studies of suicide contagion, such as the use of the word “suicide” in the headline, detailed descriptions of suicide methods, and reference to other acts of suicide associated with cyberbullying. According to the 2015 version of the Associated Press Stylebook, the Associated Press (AP) “generally . . . does not cover suicides or suicide attempts, unless the person involved is a well-known figure or the circumstances are particularly unusual or publicly disruptive” (American Copy Editor Society, 2015). While it is aligned with previous suicide coverage guidelines, this new wording also captures some ways in which research-based guidelines for journalists may be at odds with journalistic norms and news values. In health coverage, for example, critiques of coverage of research controversies or fledgling results ignored the fact that novelty “is at the core of what journalism is as a social practice and a form of knowledge production” (Hallin & Briggs, 2015, p. 93). The suicides included in this analysis were not of well-known figures, but the concept of unusual or disruptive circumstances leaves the door open for coverage of any suicides seen as socially resonant. The social representation of cyberbullying that emerged in news stories suggests a society troubled by what’s seen as the increasing trend of aggression online that disrupts traditional notions of bullying. Joffe and Haarhoff (2002) argued that “the typicality of a theme, even in a nonrepresentative sample, provides an indication of the degree to which it is shared” but that “typicality is reported but not reified” (p. 959). While the articles in the sample varied by length, topic, and newspaper size, and the individual incidents varied by gender of victim and context, both thematic and content analysis revealed clear consistencies in how these suicides were covered. Cyberbullying was HEALTH COMMUNICATION emphasized whether the suicide was the center of a story or an exemplar because it was what made these events newsworthy, and because emphasizing the destructive power of cyberbullying could prompt attention and action. Other aspects of victims’ lives that might serve as risk factors for suicide (mental illness, family issues) were less visible, although larger newspapers did address mental illness more frequently, a finding that could be attributed to increased specialization among reporters or awareness of guidelines at bigger papers, though the fact that larger newspapers were also more likely to include “suicide” in the headline may contradict this interpretation. Although it seems clear that cyberbullying was a triggering factor in the suicides included in this analysis, reification of this linkage is potentially problematic from the perspective of suicide contagion (Saleh et al., 2014). The behavioral contagion effects for suicide coverage are strongest among adolescents (Gould et al., 2003), who are also most likely to identify with the victims described in these stories due to their proximity in age and potentially other perceived similarities (Bandura, 2001). In describing the events that precede suicide, journalists are dependent on sources, primarily family members, who are grieving and seeking their own explanations for events (Chapple et al., 2013). It is common for suicide survivors to attempt to find external causes for the suicide (Joiner, 2010). Evidence of cyberbullying can serve as an anchor for understanding tragic acts. Social representations are sometimes described as commonsense theories (Sammut & Howarth, 2014; Schmitz et al., 2003), or as socially shared sense-making. This aspect of social representations highlights how these representations also address cause and effect, and understanding the “why” of processes such as cyberbullying fits with the understanding that humans attempt to make sense of events by attributing causality (Malle, 2011). The causal chain in stories matters because when concepts are linked in media messages, those concepts are then more likely to be linked in memory, and conceptual linkages stored in memory grow stronger with repeated retrieval (Coleman & Thorson, 2002). When blame for cyberbullying is attributed even indirectly to victim actions, such as difference or deviance from what is socially acceptable (Thornberg, 2015), both bullies and bystanders may position the victim as responsible for the situation preceding his or her death (Gini, Albiero, Benelli, & Altoè, 2008). Victim behavior is then the instigator in a causal chain that links deviation from what’s normal or moral to bullying and then to suicide. This aligns with past research that suggests suicide is often framed from an agentic perspective—something over which the victim has control—even when other causal factors, such as mental illness, are blamed (Boudry, 2008). In this sample, coverage of females, bullied for their explicit sexuality, and males, for their sexual orientation, suggests troubling trends toward unquestioned and implicit victim blaming (Press, 2011), even in coverage of cyberbullying that in other ways seems extremely sympathetic toward victims. Other themes in the main social representation in these stories include the laying of blame on schools and the implications that someone should have known adolescents were being victimized and acted to prevent it. The assumption that 1089 knowledge of bullying compels preventative action is complicated by the lack of consensus about how to address cyberbullying and the wide range of potential preventive or punitive actions, from school-based mediation to suspension and from juvenile courts to felonies, and a lack of guidance for schools on how best to manage bullying online. This relates to another theme this study unearthed, which is that publicized suicides function as exemplars but also as alarms that prompt action. Media guidelines for bullying and suicide coverage both suggest that journalists provide prevention resources, such as the contact number for a suicide-prevention helpline. Coverage that lacks information about prevention “implies that bullying has no solution, a misrepresentation of the current state of the research” (stopbullying.gov). Critiques of suicide coverage also suggest that failing to describe the complicated context for suicidal behavior “may result in too narrow a focus of preventive action” (Herz et al., 2013, p. S3). Many articles in the sample included an incident of suicide as an exemplar in a discussion of legislative solutions to cyberbullying. All 50 states have enacted legislation related to bullying, and 23 states thus far have legislation that specifically addresses cyberbullying (Cyberbullying Research Center, 2016). Future research should investigate the relationship between bullying coverage and criminalization of bullying, as well as attribution of responsibility for bullying solutions. In addition, future research should investigate whether focusing attention on legislative solutions diminishes support for alternative or complementary approaches to addressing cyberbullying. Recent research identified communication skills deficiencies as a significant predictor of cyberbullying behavior (Roberto et al., 2014a), suggesting that more work should be done to develop communication-based interventions that address the root causes of cyberbullying. Even limited cyberbullying prevention interventions can influence adolescents’ intentions to retaliate in cyberbullying encounters and intention to report cyberbullying (Roberto, Eden, Savage, Ramos-Salazar, & Deiss, 2014b). Health communication has an important role to play in future studies of bullying because of its emphasis on mediated communication and theory-based interventions (Ramirez et al., 2008). Finally, the theme that cyberbullying is an unsettling amplification of a familiar phenomenon taps into social fears based in technological determinism. The idea that social media are out of everyone’s control and are not, in fact, socially shaped tools allows for a distancing of responsibility. The content analysis determined that most stories described specific platforms as enabling cyberbullying, and that social media were mentioned most frequently. Future research ought to investigate how adolescents perceive the venues upon which cyberbullying occurs. A minority of stories also mentioned the features of digital, and especially social, media that are associated with cyberbullying and that differentiate bullying online from offline. Future research should investigate the valence of the conversation about technology and cyberbullying, especially since automatic monitoring of technology to detect online aggression is one method recently proposed for preventing the behavior (Van Royen, Poels, Daelemans, & Vandebosch, 2015). 1090 R. YOUNG ET AL. This study has limitations. News media coverage of suicides likely varies by socioeconomic status, sexual orientation, race, geography, and gender, so not all suicides would receive regional or national attention. It is important to stress that a significant limitation of the study is the lack of comparison of news coverage of adolescent suicides in which cyberbullying was not mentioned as a contributing factor. How might attribution of responsibility, for example, differ when cyberbullying is not included? Another important limitation is that this study only describes articles in which both cyberbullying and suicide are mentioned, which means conclusions cannot be extended to all newspaper coverage of cyberbullying. Future research could investigate cyberbullying coverage more generally to see whether suicide is mentioned as an outcome of bullying and to explore whether coverage of incidents of cyberbullying or suicide varies by demographic characteristics. In addition, this sample included a broad range of newspapers but was not designed primarily to identify differences in coverage based on newspaper size or other factors that may influence journalistic norms and resources. Conclusion The social representation of suicide associated with cyberbullying reflects anxiety about an old phenomenon translated to a new medium. Social representation analysis positions journalism as a means of formal communication through which society processes this new phenomenon, while content analysis points to the persistence of elements of that social representation in a broader range of articles. The fact that few articles adhere to suicide coverage guidelines intended to reduce suicide contagion suggests there is still room for discussion about how best to engage journalists on these issues. It also illustrates the challenges inherent in suicide coverage, especially for topics as emotionally fraught as cyberbullying. Our content analysis aligns with past research that newspaper articles do not always adhere to suicide coverage guidelines (e.g., Jamieson et al., 2003; WHO, 2008) and that cover...
Purchase answer to see full attachment