Abstract Internalizing symptoms (i.e., anxiety, depressive symptoms, suicidality) are alarmingly common among adolescents. In 2019, nearly 37% of high school students reported feeling anxious, sad, or hopeless, representing an increase from 2017 (i.e., 31%), and nearly 19% seriously considered suicide1, an increase from 2017 (17%). The coronavirus pandemic has exacerbated this problem, with research finding that social isolation resulting from the pandemic was linked to higher levels of stress, fear, loneliness, anxiety, depression, and suicide ideation among adolescents2-3. Importantly, ethnic/racial disparities have been documented in internalizing symptoms, with more negative outcomes for Latinas and Black males39-45. Current universal school-based approaches to prevention have reported uneven or limited effects, or no effects at all6-12. Questions have also been raised regarding cost and accessibility49-59. Despite the uneven track record of universal school-based (Tier 1) approaches, which has led some to recommend an emphasis on targeted (Tier 2) approaches48, they possess several advantages. First, given the demographic heterogeneity in risk factors, as well as disparate access to high-quality, culturally-sensitive health care23, schools remain attractive as they can ensure equitable access to primary prevention services. Second, universal programs avoid the difficulty of identifying adolescents at risk12, and third, they minimize the risk of stigmatizing adolescents who seek out or are referred to services20-22. In this project, submitted to the Transformative Research Award initiative (RFA- RM-21-017), we will use cooperative learning (CL) as a universal (Tier 1) school-based prevention program to target malleable peer-based risk factors and subsequently evaluate how change in these mechanisms can reduce adolescent internalizing symptoms. CL targets various forms of maladaptive peer relations that create stress and comprise a significant risk factor for internalizing symptoms in adolescence; CL also promotes peer protective factors that can reduce the likelihood of internalizing symptoms (e.g., peer acceptance)27-34. CL has also been found to promote more cross-race interaction and interracial attraction, greater cross-ethnic academic support, and more frequent cross-ethnic friendships95-97. To support teachers in implementing CL, we will use a Web-based software platform that provides an accessible, high-fidelity mechanism to deliver CL lessons. We will not only test for intervention effects, but will also evaluate the relative strength of specific, theoretically-derived change mechanisms, and uncover key issues/barriers related to implementation that will allow us to dynamically adjust our implementation approach to achieve maximum impact during subsequent scale-up. Our project is designed to establish feasibility with teachers and schools and counter potential threats to implementation fidelity while also creating a roadmap to enhanced scalability and sustained implementation. Significant results from this project could inspire broad uptake of our approach in educational contexts, potentially addressing a significant public health need during a time of crisis due to the COVID-19 pandemic.

Narrative Adolescence is a developmental period of significant risk for anxiety, depressive symptoms, and suicidality, and we propose to target key peer-based risk and protective factors using Cooperative Learning (CL). CL is a small-group instructional approach that can enhance peer relations and reduce peer-related risks, as well as promote academic engagement and achievement and reduce racial disparities. CL will be delivered with the aid of technology that automates the design and delivery of CL lessons, promoting rapid implementation, scalability, high fidelity, accessibility, and sustainability.