Many groups require cooperation in order to efficiently complete tasks in a manner that benefits all group members. The antecedents of cooperative and competitive behavior have been well studied using a particular class of problems called social dilemmas. Cultural variables, such as collectivism, are often thought to influence cooperative behavior in groups, but experimental evidence has seen mixed results. The current study attempts to add to our understanding of the effects of cultural variables on cooperative and competitive behavior in groups by advancing two major ideas: (1) that the Input-Process-Output (I-PO) framework-a theoretical framework of group functioning which proposes that group members' individual characteristics, dispositions, etc. influence the processes of groups when interacting which, in turn, impact the outcomes the group produces-can be used as a conceptualization for understanding the impact of cultural variables on potential group outcomes, provided that a distinction is made between potentially meaningful but taskunrelated distal inputs such as collectivism and task-related proximal inputs, and (2) that group process can be indexed using variance components calculated from the Social Relations Model (SRM}-a statistical tool used to analyze dyadic data. Using two social dilemmas as experimental media, participants were placed in groups of four and asked to report what they expected each of their group members to do during the social dilemmas and how much they trusted each of their fellow group members. Results demonstrate that collectivism increases the tendency to expect similar behavior from fellow group members and to trust fellow group members at similar levels when given little diagnostic information. In turn, more competitive behavior is demonstrated in groups that have members who all expect similar behavior from each group member, but show variability regarding what the behavior will be. The study demonstrates (1) a significant relationship between collectivism and expectations of other group members' behavior, (2) expectations will synthesize into meaningful variance components as calculated using SRM, (3) SRM variance components serve as useful indicators of group process and, (4) SRM variance components can be used to predict cooperative and competitive behavior in social dilemma situations. This research demonstrates the value of using SRM variances as indices of process and underscores the theoretical utility of the I-P-O framework as an explanatory tool of group behavior.