What factors drive the thoughts that come to respondents' minds at the particular moment that a survey question is asked about their policy preferences? We have a partial answer to this question from the long tradition of literature stemming from The American Voter -- those with strong ideological positions respond to policy questions with views that mostly concur with their ideology -- but we know very little about why moderates respond the way they do to policy questions. This dissertation develops a theoretical model of how preferences are formed and analyzes how mechanisms of preference formation vary based on strength of ideological views. The Cultural Preference Model put forth posits that while ideologues rely on political heuristics to respond to survey questions, moderates default to a more primitive source of their worldviews: culture. Further, the model states that ideology is simply cultural worldviews that have been forced to fit the specific setup of political institutions. The project uses an anthropological operationalization of cultural worldview, Douglas and Wildavsky's Cultural Theory, to analyze the relationships between culture, ideology, and policy preferences for all individuals, and then for moderates and ideologues separately. Results indicate that the model is supported: moderates rely more heavily on their cultural views when forming and reporting their policy preferences, whereas ideologues rely primarily on political heuristics.