Abstract
Through socialization in friendship networks, individuals' levels of prejudice become more similar to their friends' levels of prejudice over time. However, the potential consequences of socialization for prejudice in a social environment remain unknown, as research has yet to explore the size of the socialization effect. Using longitudinal social network analysis and empirically informed network simulations, the present research investigated the extent to which socialization among low-, medium-, and high-prejudice individuals relates to changes in attitudes toward ethnic and religious outgroups in school friendship networks (
= 2,484 adolescents in 10 German schools). Results from the longitudinal social network model showed that individuals' levels of prejudice became more similar to their friends' levels of prejudice over time, providing evidence of socialization. Results from the empirically informed network simulations revealed that socialization produced at most a 3% change in prejudice over 9 months, reflecting a small effect size akin to the effects of prejudice-reduction interventions. Increases, decreases, and stability in prejudice were observed in the simulations, depending on the initial levels of prejudice among individuals and their friends. Socialization was strongest among friends who held initially opposing attitudes, which led both high- and low-prejudice individuals to become more neutral over time. The findings thus suggest that socialization has a neutralizing effect, rather than a polarizing effect, on prejudice in adolescent friendship networks. This research has methodological implications for the estimation of effect sizes in psychological studies and practical implications for network interventions that aim to utilize socialization to reduce societal prejudice. (PsycInfo Database Record (c) 2026 APA, all rights reserved).