Dynamic Pathways Between Externalizing and Internalizing Behavior in Adolescent Friendship Networks

• In Event: 3-100 - Peer Network Dynamics and Child Development: Appraising the Evidence and Charting New Directions

Integrative Statement

During adolescence, peer relationships are highly salient and peer dynamics are crucial for shaping the trajectories of externalizing and internalizing behaviors (Prinstein & Giletta, 2016). Ample research from a network perspective (Veenstra et al., 2013) has shown that adolescents befriend others who are similar to them on depressive symptoms (Schaefer et al., 2011) and risk-taking and antisocial behaviors (meta-analysis of 24 studies; Galupe et al., 2018) and that over time youth become similar to their friends on these outcomes (Giletta et al., 2011; Galupe et al., 2018). Yet, limited attention has focused on unpacking co-evolving trajectories of internalizing and externalizing behaviors as embedded in changing networks, and this is an important gap because internalizing and externalizing behavior is comorbid for some youth (Reinke et al., 2012). We fill this gap by examining how new friendships are created and existing ones are maintained as a function of main and interactive associations between externalizing and internalizing behaviors as well as direct and indirect (friends’ externalizing behavior is associated with changes in focal adolescent internalizing behavior; Prinstein & Gilleta, 2016) forms of peer influence on these outcomes.
We used an ethnically diverse sample of 6-8th graders attending a public middle school in the southwestern US (N=951; 50% boys; Mage = 12.10; 13.7% European American, 18.2% African American, 54.3% Latino/a, 8.5% Native American, and 5.3% other race/ethnicity) who completed two assessments across a year. Students nominated up to 10 friends. Students reported on depressive symptoms (DS; Kovacs, 1992; Cronbach alphas .85 and .88) and risk-taking behaviors (RB; Eccles & Barber, 1990; Cronbach alphas .88 and .91). We used a stochastic actor-based modeling approach implemented in RSiena (Snijders et al., 2010) to accomplish our goals.
Results from RSiena models revealed that there were significant interactive peer selection effects between DS and RB. After accounting for the mean centering, we calculated selection estimates for youth with low and high level of DS by summing together the scaled interaction and main effects estimates and calculated standard errors for the selection estimates to test if these estimates are significantly different from zero (computed according to the variance sum law for correlated random variables; Haynie et al., 2014 for more details). We found that for nondepressed youth, RB was positively associated with maintenance of outgoing ties (est.=2.95, p<.05), whereas for depressed youth, RB is negatively associated with maintenance of outgoing ties (est.=-10.04, p<.05). For nondepressed youth, RB was negatively associated with creation of new outgoing ties (est.=-3.14, p<.05), and for depressed youth, RB was positively associated with creation of new outgoing ties (est=10.17, p<.05). Adolescents increased their levels of RB to become similar to their friends (direct influence) and being friends with others who had higher levels of RB was associated with an increase in DS (indirect influence).
This study advances developmental research by illuminating the complex links between internalizing and externalizing behavior as embedded in changing networks and elaborating on theories of peer influence. Theoretical and applied implications of this work will be discussed.

Authors

• Olga Kornienko, George Mason University
  Presenting Author

• Davila Marissa, George Mason University
  Non-Presenting Author

• Carlos E. Santos, University of California Los Angeles
  Non-Presenting Author