UNDERSTANDING NEWS MEDIA COVERAGE ON BULLYING AND CYBERBULLYING

• Published in: Journal of adolescent health 2019-02, Vol.64 (2S), p.S94
• Main Authors: Gower, Aubrey D, Vaillancourt, Tracy, Brittain, Heather, Pletta, Karen, Moreno, Megan A
• Format: Article
• Language: English
• Subjects: Anger; Anxiety; Bullying; Content analysis; Cyberbullying; Dictionaries; Fear & phobias; Health information; Journalists; Mass media; Media coverage; News media; Prevention programs; Public health; Words
• ISSN: 1054-139X; 1879-1972

Purpose: Bullying and cyberbullying are common phenomena with negative health, academic, and social consequences. One way in which youth and families are exposed to bullying is through news media, which often highlights tragic cyberbullying cases and can reach a national audience. Previous work suggests that many people rely on news media at least as much as on healthcare providers for health information. Thus, it is important to understand the content of bullying-focused news media. It remains unclear whether the current state of news media coverage is fear-based or public health-oriented, and whether this differs by topics of bullying or cyberbullying. Methods: This content analysis study used natural language processing on a sample of published news media articles. We used Proquest to identify newspapers that varied geographically by region and balanced national/regional outlets. We searched within newspapers using terms "bullying" and "cyberbullying" over the past five years. Potentially eligible articles were downloaded from both searches, and then evaluated to apply inclusion criteria ensuring the article focused on bullying or cyberbullying and on adolescents. For selected articles, we applied natural language processing using Linguistic Inquiry and Word Count (LIWC). From existing LIWC dictionaries, we identified 2 a priori topics of interest: 1) Affective Processes, including words about anxiety and anger; 2) Drives, including words related to risk. We then used a modified Delphi process to develop and pilot test a new LIWC dictionary to characterize articles as Fear-based or Public Health-oriented. We used Mann-Whitney U tests to compare LIWC output for bullying and cyberbullying news stories and chi-square tests for proportions of articles that were Fear-based and Public Health-oriented. Results: A total of 464 articles met inclusion criteria: 323 bullying-focused articles and 140 cyberbullying-focused articles. For affective processes, we found that 0.22% of words in bullying articles focused on anxiety, compared to 0.34% of words in cyberbullying articles (p=.0007). Further, 0.61% of words in bullying articles focused on anger, compared to 0.85% of words in cyberbullying articles (p=.0003). For drives, 0.7% of words in bullying articles implied risk, compared to 0.83% of words in cyberbullying articles (p=.008). Among bullying articles, 19.5% were categorized as Fear-Based and 49.8% were categorized as Public Health-Oriented; whereas cyberbull