The analysis of social reasoning is at the core of understanding how to manage social networks. Since interpersonal relations are composed of multiple factors with different nature (i.e., structural and social factors), we explore their influence on the strategizing processes in social networks. The research is based on the consideration of social networks in terms of network games. Therefore, we formalize interpersonal relations using the methods of structural and social analysis combined with game theoretic approach. Specifically, we formalize social power of an agent and use it to quantify payoffs. Based on reasoning over values of power we show how individuals reach stability and Nash equilibrium with their peers in network games.

This dissertation investigated online aggression in emerging adults to understand the contextual power of anonymity and social modeling. Emerging adults are characterized as undergoing a period of identity exploration, instability, self-focus, transition, and possibility (Arnett, 2004). Given the importance of identity development at this stage of the lifespan, this research explored religiosity/spirituality and political ideology; two pivotal belief systems that are introspectively evaluated and molded in emerging adults as they separate their identities from their world views (Barry & Nelson, 2004). Furthermore, this dissertation sought to apply religiosity/spirituality and political ideology to the previously established link of anonymity and social modeling and their joined impact on online aggression (Zimmerman & Ybarra, 2016). Behavioral temptation to aggress and participant responses following interaction on a mock blog was recorded and analyzed in situations of anonymity and positive or neutral social models. Aggressive social modeling influenced blog posts and behavioral temptation to aggress. Religiosity/spirituality and political attitudes moderated aggression in blog posts.

Behavioral temptation to aggress and participant blog responses following a group word unscrambling game were examined in situations of anonymity and positive or negative social modeling. Anonymous participants were more aggressive than non-anonymous participants. Also, social modeling seemed to moderate the effect of anonymity on behavioral temptation to aggress as well as verbal aggression via blog posts. Specifically, anonymous participants responded more aggressively when they viewed aggressive models following failure in a team word unscrambling game. These findings suggest that although anonymity may increase the likelihood that individuals will aggress, social modeling may influence aggressive outcomes.