This project studies the causal role of peer interactions and the nature of the classroom environment in the formation of individual beliefs, attitudes, and cognitive and non-cognitive skills in childhood, with a view towards improving positive attitudes towards others, improving achievement outcomes, and mitigating inefficient gender gaps in choices and outcomes. To this end, we design and evaluate a series of randomized-controlled trials, based on the idea of creating settings with different types of social interactions among students in the educational environment. The interventions are based on designing a set of novel educational curricula on coding, and using them as a backdrop for inducing certain types of interactions. The context of coding allows us to create a learning environment that can mimic the classroom, with exogenous rules for interactions, performance, and rewards. That is, we create our own educational setting where interactions are induced in pre-determined ways within a learning/performance context. We implement different types of coding curricula that have distinct features. In one, we implement a cooperative learning environment within the coding program, with teamwork. In another, we implement a learning environment within the coding program that involves competitive elements. We compare these with a case where we implement the same coding program but given individually, as well as a pure control group. We aim to measure the effect of the interventions on distinct sets of outcomes, including cognitive and non-cognitive skills, competitive and cooperative behavior, beliefs about oneself and others, attitudes towards outgroups (in terms of gender and refugee status), friendship networks, learning outcomes, and performance at school. In addition, we explore the mechanisms and heterogeneity of treatment effects.
