Peer Victimization, Risky Decision Making, and Approach/Avoidance in Mid-Adolescence

• In Event: 1-038 - Neural Underpinnings of Childhood and Adolescent Adverse Social Experiences With Peers

Integrative Statement

Advances in developmental neuroscience have provided insights into the link between adolescents’ susceptibility to peer influence and the maturational imbalance between cognitive-control and social-reward networks that is displayed as risky decision making (RDM) (Steinberg, 2007; Van Leijenhorst et al., 2010). Yet the bulk of this work reports effects of peer presence and risk proclivity (Gardner & Steinberg, 2005). The current study shifts attention to the potential role of adverse social experiences on risk orientation. We hypothesized that peer victimization would be associated with higher risk seeking under conditions of uncertainty, particularly for youth who are temperamentally oriented toward greater risk (greater behavioral approach and lower behavioral avoidance).
Our sample included 140 adolescents from understudied Appalachian communities. Moving away from more commonly employed behavioral risk measures, the current study hypothesis was examined at the neural level (e.g., using fMRI insula activity and risky decision making in an uncertainty task, in a reliable composite measure of RDM) (Paulus et al., 2003), as well as with self-reported attributes (e.g., approach/avoidance questionnaire indicators). Brain imaging and uncertainty task data were collected at 14yrs, and questionnaires at 15 yrs.
In order to obtain a more integrated understanding of social experiences and adolescent RDM, variable-centered and person-centered analyses (Laursen & Hoff, 2006) were employed. The variable-centered analysis utilized correlation analysis to estimate the association between victimization and RDM. As anticipated, greater victimization was associated with higher RDM, r (139) = .15, one-tailed p < .05. Given the modest size of the association, we computed an absolute difference score between victimization and RDM z-scores to explore potential heterogeneity in the sample in the strength of this theorized association. There was wide variation in the sample, with a very strong association found for those with small differences in their victimization and RDM levels relative to others [r (69) = .71, p < .001], whereas no association was observed [r = -.05, p = .65] for those with large differences between victimization and RDM scores.
Given this striking heterogeneity, we employed a person-centered analysis to identify potential sub-groups, based on gender and approach/avoid tendency. Using regression (centered predictors), we statistically predicted approach and avoidance (separate equations) from the victimization/RDM difference score, gender (0 = male, 1 = female), and their two-way interaction. For approach, the full equation was significant, F (3, 136) = 4.24, p < .01, R2 = .09, and the gender-by-difference score interaction was significant, β = -.22, p < .05. Post-hoc probing revealed: for males, β = .54, p < .001; for females, β = .03, p = .77. For males only, those who were lowest in approach tendency were most likely to show a strong association between victimization and RDM. For these males, being risk-averse in decision making may lower risk of peer victimization, and vice versa.

Authors

• Mengjiao Li, University of Massachusetts, Amherst
  Presenting Author

• Shereen El Mallah, University of Virginia
  Non-Presenting Author

• Kirby Deater-Deckard, University Massachusetts
  Non-Presenting Author

• Brooks King-Casas, Department of Psychology, Virginia Tech
  Non-Presenting Author

• Jungmeen Kim-Spoon, Virginia Tech
  Non-Presenting Author