Recent research has revealed the complex origins of political identification and the possible effects of this identification on social and political behavior. This article reports the results of a structural equation analysis of national survey data that attempts to replicate the finding that an individual's negativity bias predicts conservative ideology. The analysis employs the Motivational Activation Measure (MAM) as an index of an individual's positivity offset and negativity bias. In addition, information-seeking behavior is assessed in relation to traditional and interactive media sources of political information. Results show that although MAM does not consistently predict political identification, it can be used to predict extremeness of political views. Specifically, high negativity bias was associated with extreme conservatism, whereas low negativity bias was associated with extreme liberalism. In addition, political identification was found to moderate the relationship between motivational traits and information-seeking behavior.