ENCOURAGING BYSTANDER INTERVENTIONS THROUGH PERSONAL RESPONSIBILITY, EMPATHY, AND SOCIAL MEDIA TESTDRIVE TRAINING

• Sponsoring Institution: National Institute of Food and Agriculture
• Project Status: COMPLETE
• Funding Source: HATCH
• Reporting Frequency: Annual
• Accession No.: 1024234
• Project Start Date: Oct 22, 2020
• Project End Date: Oct 1, 2023
• Program Code: [(N/A)]- (N/A)

Recipient Organization
CORNELL UNIVERSITY
(N/A)
ITHACA,NY 14853

Performing Department
Communication

Non Technical Summary
Cyberbullying and harassmentarewidespread online, with young people being the most vulnerable group of social media users. According to the 2018 Pew Research Center report, 59% of young people in the US have experienced online harassment, which includes offensive name-calling, physical threats, and spreading of false rumors online.One key group that can mitigate negative effects of online harassment is bystanders, people who witness offensive behaviorbut are neither perpetrators nor direct targets. They represent by far the largest group of social media users involved with online harassment, with over 85% of young people having witnessed cyberbullying, according to the 2017 Pew Research report. While bystanders can reduce adverse effects of cyberbullying and prevent subsequent cyberbullying incidents,only few intervene. In fact, only three in ten Americans have taken steps after witnessing online abuse directed at someone else (Pew Research report, 2017). Building on our previous studies (DiFranzo et al., 2018; Taylor et al., 2019), the goal of the proposed research is to motivate bystanders to interveneby increasing their sense of personal responsibility and empathy towards online victims. The proposed research will explore effects of transparency and empathy nudges on bystander interventions in a simulated social media environment. The findingswill be used to enhance and deliver an online training via Social Media TestDrive,an interactive and realistic social media simulation that uses an experiential learning method to help young people learn and practice digital citizenship skills.The Social Media TestDrive training will be disseminated in New York Statethrough our partnerships with 4-H clubs and Cornell Cooperative Extension educators, andnationally through our partnership with Common Sense.

Animal Health Component

(N/A)

Research Effort Categories

Basic

100%

Applied

(N/A)

Developmental

(N/A)

Classification

Knowledge Area (KA)	Subject of Investigation (SOI)	Field of Science (FOS)	Percent
903	6010	3030	100%

Knowledge Area
903 - Communication, Education, and Information Delivery;

Subject Of Investigation
6010 - Individuals;

Field Of Science
3030 - Information and communication;

Keywords

bystander intervention

cyberbullying

positive youth development

social media

well-being

Goals / Objectives
The overall goal of this research is to find solutions to cyberbullying prevention and mitigation through encouraging bystander interventions. Our first bystander intervention is rooted in empathy induction by using a targeted deployment of empathy nudges that encourage perspective-taking and empathic concern. Our second objective is to examine bystanders' personal responsibility when they encounter cyberbullying content flagged by the online platform, and how information related to sources of content moderation affects bystanders' subsequent actions. Both projects will be implemented using a realistic social media simulation that allows for a high level of realism and control. The results will inform the development andenhancement of an online educational program - Social Media TestDrive - aimed at training young people to learn and practice online empathy and responsibility skills in social media interactions.

Project Methods
We will conduct two studies using a simulated social media platform used in our previous studies (DiFranzo et al., 2018; Taylor et al., 2019). This platform creates a realistic yet controlled social media experience. It mirrors basic functionality of other social media, such as a curated activity feed, status updates, one-click interactions, and threaded comments after each post. A custom-developed simulation run on this platform will provide an experimental control and create identical social media experience for all participants, except for variations accounted by an experimental design and participants' own activities on the site.During a two-day participation on the site, all participants will witness instances of cyberbullying, and we will examine the effects of empathy and personal responsibility interventions on their actions on the site. For empathy, we will build on our previous study in which we manipulated the effect of empathy nudges designed to promote perspective-taking and empathy. It was implemented through a feeling prompt that asked the question above the comment box in an activity feed. For example, "How is Jane Doe feeling?" (the name of the victim/poster). Our previous study has shown that the empathy nudge intervention failed to scale up when presented with every post in an activity feed. To reduce potential empathy fatigue and empathy burn-out, this study will manipulate a selective exposure to feeling prompts by embedding them only within a selected number of posts (the type and number of posts will be experimentally manipulated). If successful, this study will have important theoretical and practical implications for understanding bystanders' solutions to cyberbullying through encouraging empathy and perspective-taking online. The second study will examine a combination of design interventions with computational approaches to regulate cyberbullying and online harassment, in an effort to increase bystanders' sense of personal responsibility. Whereas social media platforms often automatically flag offensive content, they are often inaccurate because of the lack of contextual information and rely on users to confirm or disconfirm problematic content. We will examine how transparency around algorithmic content regulation will affect users' perceptions and reactions to moderation. By using the same experimental social media platform as in Study 1, we will have participants witness cyberbullying content, which will be either ambiguous or unambiguous harassment, that will be flagged by the system. The flagging message will display different source of moderation (system vs. other users vs. unidentified) to test the effect of attribution source on bystanders' perceptions of social norms and taking responsibility for subsequent cyberbullying content they encounter on the site. The results of Studies 1 and 2 will inform the development and delivery of TestDrive modules aimed atencouragingbystander interventions through empathy and personal responsibility on social media.

Progress 10/22/20 to 09/30/21

Outputs
Target Audience: Target audiences served by the project during this reporting period include: a) youth who are taught prosocial and responsible conduct online as a result of experiential learning opportunities through the Social Media TestDrive platform; b) parents whose concern about children's online safety and use of social media are addressed through our ongoing research and dissemination of the Social Media TestDrive platform; c) teachers, extension educators, and schools, which have benefitted from the digital literacy education and dissemination efforts undertaken by my team. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?Both of the research studies described above have been conducted by multi-member teams consisting of professionals (faculty, programmer, administrative support) and both graduate and undergraduate students. These efforts have resulted in important training and professional development opportunities for graduate students whoparticipated in different phases of these projects, from conception and study design to data analysis, result presentation and reporting, and publication. Undergraduate students have had an opportunity to be trained in responsible, replicable, and ethical ways of conducting behavioral studies online, data collection procedures and data coding, as well innovative ways for social media experimentation through realistic and safe simulations of online behavior. How have the results been disseminated to communities of interest?The results of both studies have been disseminated to professional communities at annual conventions of International Communication Association (2021 Convention for the Bystander Intervention Study and 2022 Convention for the Youth Social Media Literacy Scale). Additionally, the bystander intervention study was published in the Journal of Computer-Mediated Communication, and the digital literacy scale study is currently being prepared for publication at the Communication Methods and Measures Journal. Some of the items from the newly developed digital literacy scale are being prepared for public releaseas an assessment tool for the Social Media TestDrive modules, which have been actively used by educators and families across the US. The information about the assessment tool will be included in the nextSocial Media TestDrive newsletter, which is sent quarterly to over 200 educators. Finally, we will disseminate it onthe Social Media TestDrive site accessed by 12,000-15,000 learners monthly. What do you plan to do during the next reporting period to accomplish the goals?The next reporting period will continue our efforts with both bystander intervention and digital literacy projects, as well as ways how our research efforts into understanding bystander behaviors can be translated into digital literacy interventions to promote youth's ethical and responsible behaviors online.

Impacts
What was accomplished under these goals? During this reporting period, we conducted two studies. The goal of the first study was to understand how to promote bystander intervention in social media, or how to encourage people to act when they encounter online harassment and cyberbullying directed at other users. Many platforms rely on both users and automated systems to detect and flag offensive posts, but it is not understood how seeing the source of moderation (who flagged an offensive post) affects users' subsequent behavior. Using an innovative experimental design through our custom-developed social media simulation, this study examined people's reactions to the flagging of offensive posts by other users, an automated system, or an unspecified process. In addition to determininghow visibility around source of flagging affects the users' ensuing behavior, we also examined potential mechanisms that can encourage bystanders' interventions, such as perceived social norms and assuming accountability and responsibility for one's online actions. The results provide new insights into the challenge of communicating transparency around content moderation of offensive posts in a way that will empower bystanders to have agency and become active upstanders on social media platforms. The goal of the second studywas to develop a reliable, valid, and objective social media literacy scale that is specifically designed for young adolescent populations. The scale development was done through four consecutive steps: 1) an in-depth review of the available literature to identify key core competencies and specific domains of social media literacy, 2) creation of a large item pool that assesses these core competencies within six different domains of online behavior, 3) expert review and cognitive pretesting with youth, and 4) empirical validation of the final 90-item pool using item response theory based on a sample of 306 youth participants. This newly developed and validated scale willprovide an objective knowledge test to measurethe effectiveness of various youth social media literacy interventions and programs.

Publications

• Type: Journal Articles Status: Published Year Published: 2021 Citation: Bhandari, A., Ozanne, M., Bazarova, N. N., & DiFranzo, D. (2021). Do you care who flagged this post? Effects of moderator visibility on bystander behavior. Journal of Computer-Mediated Communication, 26, 284-300.
• Type: Conference Papers and Presentations Status: Accepted Year Published: 2022 Citation: Purington, A., Masur, P., Bazarova, N. N., Zou, E., & Whitlock, J. (2022). The Youth Social Media Literacy Scale: Scale Development and Validation Using Item Response Theory. Accepted for the 2022 International Communication Association Convention.
• Type: Journal Articles Status: Under Review Year Published: 2022 Citation: Ozanne, M., Bhandari, A., DiFranzo, D., & Bazarova, N. N. (Revise & Resubmit).Perception of Bias and Trust in Moderation Systems on Social Media Platforms. Big Data & Society.
• Type: Journal Articles Status: Other Year Published: 2022 Citation: Purington, A., Masur, P., Bazarova, N. N., Zou, E., & Whitlock, J. (In preparation for submission). The Youth Social Media Literacy Scale: Scale Development and Validation Using Item Response Theory. Communication Methods and Measures.