Thesis advisor: Theresa Betancourt / Families in low- and middle-income countries (LMICs) face significant mental health and psychosocial care gaps. In recent years, researchers and practitioners have addressed these gaps by task-sharing evidence-based mental health and psychosocial support (MHPSS) interventions to nonspecialist community providers. Task-shared interventions have demonstrated effectiveness (improvements in MHPSS outcomes of intervention participants); however, quality of delivery (fidelity and competence) has rarely been examined as a factor associated with effectiveness despite implementation science models suggesting a causal link between quality of delivery and effectiveness. In this study, I apply a mixed methods approach to examine the quality of delivery by nonspecialists who are facilitating an evidence-based, early childhood development and family violence prevention program, known as Sugira Muryango, in Rwanda. Currently, Sugira Muryango is being expanded and implemented through the Promoting Lasting Anthropometric Change and Young Children’s Development (PLAY) Collaborative, which scales up Sugira Muryango to 10,000 households living in extreme poverty in Ngoma, Nyanza, and Rubavu districts. The program has strong ties to the Rwanda National Government and their social protection and policy goals. I find that nonspecialist age is significantly associated with higher initial fidelity and competence scores and smaller improvements in fidelity and competence over time. In addition, nonspecialists in Nyanza district were more likely to have higher initial fidelity and competence scores but also see smaller changes over time. Fidelity and competence were found to significantly co-vary. Multi-level growth models revealed that fidelity was not a significant predictor of changes in any child discipline outcomes or of any responsive caregiving outcomes. However, competence significantly predicted changes in some responsive caregiving practices, specifically acceptance and learning materials, and it predicted decreases in physical punishment. In semi-structured interviews, the nonspecialists provided examples of using skills such as rapport-building, empathy, and active listening to deliver Sugira Muryango effectively. Nonspecialists also provided examples of barriers to quality of delivery, including compensation and technology issues. Overall, this dissertation contributes empirical evidence to what we understand theoretically and moves towards development of best practices for monitoring and supervising nonspecialists in task-shared MHPSS interventions. / Thesis (PhD) — Boston College, 2024. / Submitted to: Boston College. Graduate School of Social Work. / Discipline: Social Work.