Generalization of rejection and acceptance in social networks

Social environments present opportunities for connection and resources, but they also involve the risk of rejection. How do people learn which individuals will reject or accept them upon entering a novel environment? Here, we propose a route to such learning: people use knowledge of relationships in social networks to infer who will be likely to accept or reject them. Previous research shows that people generalize trust from one individual to that individual’s friends, yet it remains unclear whether rejection and acceptance experiences generalize in similar ways in social network contexts. We designed a novel experimental paradigm in which participants experienced rejection and acceptance within an artificial group, learned about network connections among group members, and decided which members to approach in a new task. Study 1 found that participants generalized rejection by avoiding individuals socially closer to a rejector and approaching those closer to an accepter, forming a gradient of avoidance and approach based on network distance. Study 2 further demonstrated stronger generalization when networks reflected friendship as opposed to randomly assigned ties, suggesting partner choices depend on explicit inferences about meaningful relationships rather than associative learning alone. Finally, in a longitudinal survey of student groups, Study 3 extended these findings to real-world social networks, revealing similar patterns of generalization in college student organizations. Together, our findings inform the cognitive processes that help humans successfully navigate social environments by adaptively forming new connections.