Pamela J. Wisniewski

To work toward developing trauma-informed research practices in HCI for youth, we engaged with youth (ages 13-21) who were asked to participate in a study [54] in which they donated their Instagram data for the purpose of studying potentially traumatic online risks they encountered in their private messages. Youth were asked to flag conversations that made them feel uncomfortable or unsafe. As a follow-up, youth participants were asked to participate in a retrospective interview regarding their experience participating in the study. As part of the larger research project, we also had undergraduate research assistants (RAs) annotate youth data for risky interactions. For this paper, we interviewed youth participants and RAs regarding their experiences reflecting on the risk-flagged data. In doing so, we conducted a meta-level research study on the potentially traumatic or uncomfortable experiences of participants and the researcher assistants who annotated their data. Our engagement with youth participants and RAs was driven by the following high-level research questions:

We worked with seven teens (aged 15-17) in a Youth Advisory Board program (YAB) for over a year to involve them in online safety research from reviewing online safety research protocols to co-designing online safety interventions that cater to their needs by teaching them essential UX design tools and techniques. Teens created storyboards, user personas, mind maps, and high-fidelity prototypes for their ideas regarding online safety and privacy features. Our case study outlines the overview, methodology we used, and lessons learned from the long journey with teens in the YAB program for online safety research. We provide heuristic guidelines for the research community that aim to build similar research programs for teens, including aligning long-term program goals with teens’ needs, ensuring equal participation from diverse teens, building trust, maintaining maximum engagement, and communicating the outcomes and impact of their contributions.

We conducted an ethnographically-informed study with 28 participants (9 autistic Young Adults or" YAs" in need of substantial daily support, 6 parents, 13 support staff) to understand how autistic YAs self-regulate and receive mediation on social media. We found that autistic YAs relied on blanket boundary rules and struggled with impulse control; therefore, they coped by asking their support network to help them deal with negative social experiences. Their support networks responded by providing informal advice, inthe-moment instruction, and formal education, but often resorted to monitoring and restrictive mediation when more proactive approaches were ineffective. Overall, we saw boundary tensions arise between Autistic YAs and their support networks as they struggled to find the right balance between providing oversight versus promoting autonomy. This work contributes to the critical disability literature by revealing the benefits and tensions of allyship in the context of helping young autistic adults navigate social media.

Background: Suicide safety planning is an evidence-based approach used to help individuals identify strategies to keep themselves safe during a mental health crisis. This study systematically reviewed the literature focused on mobile health (mHealth) suicide safety planning apps.Objective: This study aims to evaluate the extent to which apps integrated components of the safety planning intervention (SPI), and if so, how these safety planning components were integrated into the design-based features of the apps.Methods: Following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, we systematically analyzed 14 peer-reviewed studies specific to mHealth apps for suicide safety planning. We conducted an analysis of the literature to evaluate how the apps incorporated SPI components and examined similarities and differences among the apps by conducting a comparative analysis of app features. An independent review of SPI components and app features was conducted by downloading the available apps.Results: Most of the mHealth apps (5/7, 71%) integrated SPI components and provided customizable features that expanded upon traditional paper-based safety planning processes. App design features were categorized into 5 themes, including interactive features, individualized user experiences, interface design, guidance and training, and privacy and sharing. All apps included access to community supports and revisable safety plans. Fewer mHealth apps (3/7, 43%) included interactive features, such as associating coping strategies with specific stressors. Most studies (10/14, 71 …

Authors’ addresses: Ashwaq Alsoubai, ashwaq. alsoubai@ vanderbilt. edu, atalsoubai@ kau. edu. sa, Vanderbilt University, Nashville, Tennessee, USA; Afsaneh Razi, afsaneh. razi@ drexel. edu, Drexel University, Philadelphia, Pennsylvania, USA; Zainab Agha, zainab. agha@ vanderbilt. edu, Vanderbilt University, Nashville, Tennessee, USA; Shiza Ali, shiza@ bu. edu, Boston University, Boston, Massachusetts, USA; Gianluca Stringhini, gian@ bu. edu, Boston University, Boston, Massachusetts, USA; Munmun De Chodhury, munmund@ gatech. edu, Georgia Institute of Technology, 30318, Atlanta, Georgia, USA; Pamela J. Wisniewski, pamela. wisniewski@ vanderbilt. edu, Vanderbilt University, Nashville, Tennessee, USA.

Emotion detection is a rapidly advancing method of quantifying human experience. Past literature shows emotional data are highly sensitive and personal. This has made gathering of emotional data for research difficult and subjective. The data that have been collected have shown that habituation effects (previous exposure to a concept) can result in more lenient ethical evaluation of actions. This inadvertent leniency may lead to design and acceptance of virtual agents that do not perform or behave in a manner commensurate with our societal standards. To build effective virtual agents, emotional data must be used in a way that is ethical and inoffensive to humans. Agents designed to interact with humans in virtually any capacity should strive to better understand them and to use that understanding within prescribed bounds. We must understand the impact that agents using such data will have on people. To …

We collected and analyzed Instagram direct messages (DMs) from 173 youth aged 13-21 (including 86 LGBTQ+ youth). We examined youth's risk-flagged social media trace data with their self-reported mental health outcomes to examine how the differing online experiences of LGBTQ+ youth compare with their heterosexual counterparts. We found that LGBTQ+ youth experienced significantly more high-risk online interactions compared to heterosexual youth. LGBTQ+ youth reported overall poorer mental health, with online harassment specifically amplifying Self-Harm and Injury. LGBTQ+ youth's mental well-being linked positively to sexual messages, unlike heterosexual youth. Qualitatively, we found that most of the risk-flagged messages of LGBTQ+ youth were sexually motivated; however, a silver lining was that they sought support for their sexual identity from peers on the platform. The study highlights the importance of tailored online safety and inclusive design for LGBTQ+ youth, with implications for CHI community advancements in fostering a supportive online environments.

Feasible and developmentally appropriate sociotechnical approaches for protecting youth from online risks have become a paramount concern among human-computer interaction research communities. Therefore, we conducted 38 interviews with entrepreneurs, IT professionals, clinicians, educators, and researchers who currently work in the space of youth online safety to understand the different sociotechnical approaches they proposed to keep youth safe online, while overcoming key challenges associated with these approaches. We identified three approaches taken among these stakeholders, which included 1) leveraging artificial intelligence (AI)/machine learning to detect risks, 2) building security/safety tools, and 3) developing new forms of parental control software. The trade-offs between privacy and protection, as well as other tensions among different stakeholders (e.g., tensions toward the big-tech …

Authors’ addresses: Karla Badillo-Urquiola, kbadillou@ nd. edu, University of Notre Dame, South Bend, Indiana, USA; Zainab Agha, zainab. agha@ vanderbilt. edu, Vanderbilt University, Nashville, TN, USA; Danielle Abaquita, denielle. abaquita@ gmail. com, University of Central Florida, Orlando, FL, USA; Scott B. Harpin, scott. harpin@ cuanschutz. edu, University of Colorado, Aurora, Colorado, USA; Pamela J. Wisniewski, pamela. wisniewski@ vanderbilt. edu, Vanderbilt University, Nashville, TN, USA.

Recent increases in self-harm and suicide rates among youth have coincided with prevalent social media use; therefore, making these sensitive topics of critical importance to the HCI research community. We analyzed 1,224 direct message conversations (DMs) from 151 young Instagram users (ages 13-21), who engaged in private conversations using self-harm and suicide-related language. We found that youth discussed their personal experiences, including imminent thoughts of suicide and/or self-harm, as well as their past attempts and recovery. They gossiped about others, including complaining about triggering content and coercive threats of selfharm and suicide but also tried to intervene when a friend was in danger. Most of the conversations involved suicide or self-harm language that did not indicate the intent to harm but instead used hyperbolical language or humor. Our results shed light on youth perceptions, norms, and experiences of self-harm and suicide to inform future efforts towards risk detection and prevention. Content Warning: This paper discusses the sensitive topics of self-harm and suicide. Reader discretion is advised.

This study aimed to measure validity and reliability of the Persian version of the Cyber-Bullying/Victimization Experience Questionnaire (CBVEQ) among Iranian adolescents. This cross-sectional validation study was conducted on Iranian adolescents between January 2022 and September 2022. Online questionnaires were completed by a total of 1439 Iranian high school students, with an age range of 14–18 years (M = 15.93 years, SD = 1.43). Participants completed Cyber-Bullying/Victimization Experience Questionnaire (CBVEQ), Online Disinhibition, Sensation Seeking, Empathy, and School Connectedness questionnaires. The validity and reliability of the CBVEQ were assessed through measures including Cronbach's alpha, construct validity via Confirmatory Factor Analysis (CFA), and evaluations of divergent and convergent validity. The CFA indicated that a two-factor model provides a good fit for the …

RQ1: What stakeholder (Streamers and Viewers) values should be embedded in an interpersonal telepresence system?

Accurate real-time risk identification is vital to protecting social media users from online harm, which has driven research towards advancements in machine learning (ML). While strides have been made regarding the computational facets of algorithms for “realtime” risk detection, such research has not yet evaluated these advancements through a human-centered lens. To this end, we conducted a systematic literature review of 53 peer-reviewed articles on real-time risk detection on social media. Real-time detection was mainly operationalized as “early” detection after-the-fact based on pre-defined chunks of data and evaluated based on standard performance metrics, such as timeliness. We identified several human-centered opportunities for advancing current algorithms, such as integrating human insight in feature selection, algorithms’ improvement considering human behavior, and utilizing human evaluations. This work serves as a critical call-to-action for the HCI and ML communities to work together to protect social media users before, during, and after exposure to risks.

Computer-Based Open-Ended Learning Environments (OELEs) are designed to challenge learners to become proficient problem-solvers and develop the ability to independently solve complex problems. However, the traditional focus of OELE research has been on demonstrating overall learning gains, potentially overlooking students who struggle in these environments. To address this gap, we take a social justice-based approach by studying 99 sixth-grade students who participated in a week-long classroom study. We first assessed learning outcomes across all then identified 20 students who failed to do well. We qualitatively analyzed video recordings of their interactions with the OELE to understand why they struggled and to determine if interface issues inhibited their learning. Five themes emerged:(1) challenges in knowledge acquisition;(2) challenges in scaffolding learning;(3) disregarding system guidance,(4) not leveraging supporting tools;(5) and getting discouraged by incorrect answers. Based on our findings, we make design recommendations for OELEs to better support underserved learners, recognizing that failure is an important catalyst for motivating improvements in child-centered design.

In this position paper, we discuss the potential for leveraging LLMs as interactive research tools to facilitate collaboration between human coders and AI to effectively annotate online risk data at scale. Collaborative human-AI labeling is a promising approach to annotating large-scale and complex data for various tasks. Yet, tools and methods to support effective human-AI collaboration for data annotation are under-studied. This gap is pertinent because co-labeling tasks need to support a two-way interactive discussion that can add nuance and context, particularly in the context of online risk, which is highly subjective and contextualized. Therefore, we provide some of the early benefits and challenges of using LLMs-based tools for risk annotation and suggest future directions for the HCI research community to leverage LLMs as research tools to facilitate human-AI collaboration in contextualized online data annotation. Our research interests align very well with the purposes of the LLMs as Research Tools workshop to identify ongoing applications and challenges of using LLMs to work with data in HCI research. We anticipate learning valuable insights from organizers and participants into how LLMs can help reshape the HCI community's methods for working with data.

With the prevalence of risks encountered by youth online, strength-based approaches such as nudges have been recommended as potential solutions to guide teens toward safer decisions. However, most nudging interventions to date have not been designed to cater to teens’ unique needs and online safety concerns. To address this gap, this study provided a comprehensive view of adolescents’ feedback on online safety nudges to inform the design of more effective online safety interventions. We conducted 12 semi-structured interviews and 3 focus group sessions with 21 teens (13–17 years old) via Zoom to get their feedback on three types of nudge designs from two opposing perspectives (ie, risk victim and perpetrator) and for two different online risks (ie, Information Breaches and Cyberbullying). Based on the teens’ responses, they expressed a desire that nudges need to move beyond solely warning the user to providing a clear and effective action to take in response to the risk. They also identified key differences that affect the perception of nudges in effectively addressing an online risk, they include age, risk medium, risk awareness, and perceived risk severity. Finally, the teens identified several challenges with nudges such as them being easy to ignore, disruptive, untimely, and possibly escalating the risk. To address these, teens recommended clearer and contextualized warnings, risk prevention, and nudge personalization as solutions to ensure effective nudging. Overall, we recommend online safety nudges be designed for victim guidance while providing autonomy to control their experiences, and to ensure accountability and …

Advisor-advisee relationships between PhD students and faculty are vital to research, but advising dynamics can be challenging for both student and advisor. Though advising can involve egregious problems such as sexual harassment, we focus on what might be less serious but more common issues such as exploitation, unprofessional behavior, mishandling of credit, and inadequate communication. While problems can be caused by advisor or advisee, the power imbalance exacerbates problems for PhD students. In any case, open discussion about PhD advising is rare. In this panel, we hope to start a much-needed conversation about PhD advising to raise awareness within the SIGCHI community about common advising problems; and to begin brainstorming solutions that faculty, administrators, and PhD students can implement.

Online dating platforms have gained widespread popularity as a means for individuals to seek potential romantic relationships. While recommender systems have been designed to improve the user experience in dating platforms by providing personalized recommendations, increasing concerns about fairness have encouraged the development of fairness-aware recommender systems from various perspectives (e.g., gender and race). However, sexual orientation, which plays a significant role in finding a satisfying relationship, is under-investigated. To fill this crucial gap, we propose a novel metric, Opposite Gender Interaction Ratio (OGIR), as a way to investigate potential unfairness for users with varying preferences towards the opposite gender. We empirically analyze a real online dating dataset and observe existing recommender algorithms could suffer from group unfairness according to OGIR. We further investigate the potential causes for such gaps in recommendation quality, which lead to the challenges of group quantity imbalance and group calibration imbalance. Ultimately, we propose a fair recommender system based on re-weighting and re-ranking strategies to respectively mitigate these associated imbalance challenges. Experimental results demonstrate both strategies improve fairness while their combination achieves the best performance towards maintaining model utility while improving fairness.

Traditional online safety technologies often overly restrict teens and invade their privacy, while parents often lack knowledge regarding their digital privacy. As such, prior researchers have called for more collaborative approaches on adolescent online safety and networked privacy. In this paper, we propose family-centered approaches to foster parent-teen collaboration in ensuring their mobile privacy and online safety while respecting individual privacy, to enhance open discussion and teens' self-regulation. However, challenges such as power imbalances and conflicts with family values arise when implementing such approaches, making parent-teen collaboration difficult. Therefore, attending the family-centered design workshop will provide an invaluable opportunity for us to discuss these challenges and identify best research practices for the future of collaborative online safety and privacy within families.

As part of a Youth Advisory Board of teens (YAB), a longitudinal and interactive program to engage with teens for adolescent online safety research, we used an Asynchronous Remote Community (ARC) method with seven teens to explore their social media usage and perspectives on privacy on social media. There was a spectrum of privacy levels in our teen participants’ preferred social media platforms and preferences varied depending on their user goals such as content viewing and socializing. They recognized privacy risks they could encounter on social media, hence, actively used privacy features afforded by platforms to stay safe while meeting their goals. In addition, our teen participants designed solutions that can aid users to exercise more granular control over determining what information on their accounts is to be shared with which groups of users. Our findings highlight the need to ensure researchers …