PROJECT ABSTRACT The accumulation of epidemiologic evidence on the roles that policies, systems and environments (PSE) play in influencing health behaviors such as eating and physical activity has led to increased interest in applying the socio-ecological framework and community-engaged approaches to the development of “multi- level interventions” targeting at-risk communities. Evaluation of such innovative interventions, often to address social determinants of health and reduce health disparities, is often challenged as it may be impractical to use the `gold-standard' cluster randomized trial design. Further, the social and behavioral science theories that provide the foundation for the development of interventions may be overlooked when efforts are made to scale- up an intervention. Systems science methods, which have been used to investigate complex causal mechanisms, have been proposed as an alternative for evaluating such interventions. However, their application, which typically involves computational modeling, requires researchers from disparate disciplines to integrate their knowledge and skills at a level that can be difficult to attain. The proposed summer research education program aims to create a learning environment that will support the training of researchers from diverse disciplines to collaborate effectively with each other and with community stakeholders, with the goal of accelerating the translation of research into practice. Specifically, over three summers, it will (i) provide 75 researchers (pre-docs, postdocs, investigators) interested in addressing social determinants of health, with the basic knowledge and skills necessary for applying systems science and data science methods, while also (ii) providing more advanced knowledge and skills to apply either of these methods as well as foundational knowledge of social and behavioral science theories and approaches, to researchers in quantitative methods-focused disciplines (such as math, computer science, biostatistics). In addition, this program will also (iii) provide a three-week practicum to 30 of the 75 trainees who are interested in gaining hands-on experience in working with large datasets. The proposed program will consist of two components: (i) a 3-week curriculum consisting of virtual didactic sessions with learning activities; and (ii) a 3-week practicum for a subset of trainees who will work closely with each other to build and refine agent-based models and interpret results for policy-makers with the engagement of stakeholders. The recruitment and selection plan will be developed with an equity lens, and with an Advisory Committee providing guidance and oversight. To create the learning environment for this training program, we will select applicants to create a cohort that is also diverse in academic background and lived experiences.