Poster #172 - Longitudinal Effects of Intergroup Attitudes and Peer Norms on Ethnic Segregation of Friendships in Middle School

• In Event: 1-169 - Poster Session 04
  In Poster Session: PS 04 Section - Race, Ethnicity, Culture, Context

Integrative Statement

Adolescent friendship networks tend to be segregated by race, even in diverse schools (Moody, 2001). What leads youth to prefer a same-ethnic friend? And what factors might encourage reaching out to peers of a different racial group? Prior research suggests attitudes (Wolfer & Hewstone, 2018) and peer norms supporting intergroup contact (Tropp et al., 2016) affect adolescent’s interest in outgroups. Yet few studies have examined whether attitudes and norms influence actual friendship patterns longitudinally. Using latent curve modeling with structured residuals (LCM-SR), we simultaneously model developmental trajectories and within-person cross-lagged associations to examine how attitudes and peer norms are related to friendship segregation during middle school, and to test directionality of effects.

The sample included 4,215 students (52% girls) from 26 urban middle schools (41% Latino, 26% White, 18% Asian, and 15% Black). Schools systematically varied in ethnic diversity. Friendship segregation was assessed using an invariant log-odds measure, which accounts for network size and opportunities for same- and cross-ethnic ties. Attitudes were measured using items assessing how adolescents felt toward ethnic ingroups and outgroups (e.g., trust, liking). Average ratings for outgroups were subtracted from ingroups (α = .93). Higher scores indicate greater ingroup preference. The School Interracial Climate Scale assessed peer norms supporting associations with outgroups (α = .71). All models controlled for gender, parent education, ethnicity, and contact opportunities in school measured by Simpson’s Diversity Index and proportion same-ethnic peers.

All models fit the data well. Unconditional models revealed that friendships tend to be segregated at the start of middle school (B = 1.02) and become more segregated over time (B = .10, p < .01). Adolescents also showed ingroup preference (B = .49), which declined over time (B = -.08, p < .001). Peer norms were relatively positive (B = 4.23), and improved over time (B = .04, p < .001). Conditional models (Figure 1) revealed between-person associations: adolescents with greater ingroup preference had more segregated friendships at the start of middle school (B = .38, p < .001), and adolescents who showed less steep decreases in ingroup bias showed greater increases in segregation over time (B = .38, p < .001). Within-person associations revealed time-dependent effects such that greater than usual ingroup preference at any point predicted greater than usual segregation at the following time point. Paths from segregation to subsequent ingroup preference were non-significant indicating attitude-driven effects. Findings were similar for peer norms: more positive than usual perceptions of peer norms predicted less subsequent segregation (Figure 2).

Although individual differences influence stable, systematic developmental growth (adolescents with greater ingroup preference or who perceive less supportive norms show greater segregation), time-specific deviations were found. Specifically, within-person dynamic effects indicate trajectories are malleable during middle school, with changing attitudes and norms affecting subsequent segregation. These findings suggest prejudice reduction interventions that target norm perceptions and personal attitudes, rather than behavior, may be especially effective at any point in middle school to promote greater racial integration in multiethnic schools (Paluck & Shepherd, 2012). Future analyses will examine ethnic differences.

Authors

• Kara Kogachi, University of California, Los Angeles
  Presenting Author

• Sandra Graham, University of California, Los Angeles
  Non-Presenting Author