Experimental Design
We implement a clustered randomized controlled trial (RCT) to evaluate a school-based intervention to promote social cohesion in public elementary schools in Bogotá, Colombia, with a high concentration of migrant students. Schools are the unit of randomization, stratified by locality and absenteeism rates. A total of 121 schools are randomly assigned to treatment (n=60) or control (n=61). Within each school, two to three fifth-grade classrooms (the final grade of primary school) are selected to participate, with an average of 70 students per cluster. The intervention consists of weekly one-hour sessions during the school day, facilitated by trained teachers using materials specifically designed for this project (e.g., illustrated stories, videos, songs, and group activities). The intervention will run for approximately four months. Teachers have already received training on delivering the intervention and will continue to receive ongoing implementation support, including regular monitoring and incentives to maintain high fidelity to the program. Control schools continue their standard curriculum during the trial period but will receive all intervention materials and teacher training after the endline survey. Importantly, current control students will not participate in the intervention, enabling future evaluations of medium-term effects by comparing treated and untreated cohorts within these schools. Data collection includes student surveys, behavioral games, social network mapping, cognitive and social intelligence tests, and teacher surveys. Baseline data is collected before the intervention, with endline data collected six to seven months later. The analysis will estimate intent-to-treat (ITT) effects of assignment to treatment on key outcomes, along with heterogeneity analyses by nationality, gender, and teacher characteristics.