This study explored various social influences on political beliefs. Specifically, it tested factors to see how they affected a subject's likelihood to model parental and peer voting behavior. Three primary approaches informed this study: social cognitive theory, conformity, and social identity theory. The author collected data from college undergraduates via a two-part survey surrounding the 2008 Presidential election. As interpreted in this study, all four components of Bandura's learning theory proved to be significant predictors for modeling parental voting behavior. Investigation showed that a subject's interest in the election was negatively related to his/her likelihood to model parental voting behavior. As interest decreased, likelihood to model parents increased. Additionally, political experience negatively related to a subject's likelihood to vote for the same candidate his/her parents endorsed. The learning theory approach as conceptualized here was not a good fit for data relative to modeling peer political behavior, however. The presence of unanimous political beliefs among family and friends increased a subject's likelihood to conform to those beliefs. Additionally, ideological conservatism was associated with increased conformity. Lastly, this study reinforced the presence and influence of family reference groups for political behavior. Increased political discussion with parents led to increased conformity with them. Subjects who discussed politics more with parents also viewed them as more powerful political influences. This same relationship existed for talk in the classroom. Peers were not pinpointed as a politically influential reference group.