Research on adolescents has shown that friends are more likely to conflict than strangers, but are best friends more likely to conflict than strangers and friends? Peer conflict is part of adolescent interactions. Therefore, ordinary friends who spend a long time together are more likely to have conflicts than strangers with whom they have little interaction. However, adolescents' definition of friendship includes mutual liking and expectations of behavioral norms. Therefore, friends with higher tie strength would theoretically have more reciprocal and consistent behaviors and less likelihood of conflict. To test this research question, the present study uses data from “Changing Climates of Conflict: A Social Network Experiment in 56 Schools, New Jersey, 2012-2013.” It includes 56 distinct social networks among 5th to 8th-grade students with two waves of information about friendship networks, best friend networks, and conflict networks. Therefore, the temporal exponential random graph model (TERGM) alongside meta-analysis techniques is used to see adolescent conflict dynamics. The major contribution of this study is to identify the influence of tie strength on adolescent conflict.

While numerous interventions have been implemented to curb adolescent smoking, most existing research primarily focuses on their direct effects on targeted individuals. However, little is known about the extent and mechanisms of spillover effects that extend beyond the targeted individuals to their friends, friends' friends, and even friends' friends' friends. Understanding these extended spillover effects is essential for optimizing public health strategies, as indirect benefits can amplify the overall impact of interventions. This study employs simulations based on Stochastic Agent-Oriented Models (SAOMs) to provide a novel perspective on how behavioral interventions diffuse through complex social networks.

Using longitudinal data from two high schools in the Add Health study, this research simulates a smoking intervention that targets between 10% and 100% of heavy smokers. Three selection strategies are examined: random selection, in-degree centrality, and eigenvector centrality. The findings indicate that increasing the proportion of targeted heavy smokers enhances both direct and spillover effects. However, these benefits diminish beyond a 40%-50% threshold, signaling network saturation. Centrality-based selection strategies yield stronger effects at lower targeting proportions but lose their advantage over random selection as the intervention expands.

This study advances the understanding of network interventions by demonstrating how social structures influence the spread of behavioral change. The SAOM-based simulations allow for a detailed analysis of the cascading effects of interventions across multiple degrees of social separation. These findings provide practical guidance for designing more effective public health strategies that strategically leverage social networks to maximize intervention reach and impact.

This study examines the formation and evolution of elementary school social networks with a focus on gender differences, exploring the dynamic mechanisms through which these networks develop. We collected data from 96 fifth-grade students across four classes from a school in Gyeonggi Province, Republic of Korea over a two-year period, with 12 data collection waves. We employ the Stochastic Actor-Oriented Model (SAOM) to analyze the processes driving student interactions within both friendship and help networks. Specifically, the research addresses three primary questions: (1) How do gender differences influence the density, reciprocity, and overall structure of friendship networks among elementary students? (2) In what ways do help networks demonstrate centralization around key individuals, and how does this pattern vary by gender? (3) What variations in endogenous, selection, and influence effects can be observed between male and female students during network evolution?

Results are anticipated to show that friendship networks exhibit high density and strong reciprocity, reflecting robust and stable interpersonal relationships among students. In contrast, help networks are anticipated to display a centralized structure concentrated around certain individuals, indicating an uneven distribution of social support. Gender-specific patterns are expected, with female students forming reciprocal relationships based on emotional connection and cooperation, while male students are more likely to develop activity-oriented and asymmetric interactions.

The findings of this study are expected to provide valuable insights for the development of gender-sensitive educational interventions and support strategies in elementary school settings, ultimately contributing to a more tailored approach in fostering social development among students.

Beginning university represents a period of great transition, often marked by significant changes in students' social networks and available support systems. This study employed a longitudinal ego-network approach to identify patterns in relationship change and loneliness among 1st year university students in the UK. A sample of 90 students reported on their social networks during the summer before university and about halfway through their first year. Using descriptive analyses and statistical tests, we examined overall network churn and differences between pre-university and university networks, including changes in alter attributes, dyadic connections, and structural features of egonets. We then conducted a series of linear regressions to test the effect of network change on students’ loneliness. Our findings provide novel insights into how shifts in social networks shape students’ experiences of loneliness, highlighting the need for socially inclusive induction events and support initiatives to foster connection and belonging during this critical transition.