Belief Manipulation Using Informational Treatments in Survey Experiments

• In Event: Information and Attitudes towards Trade

Abstract

Studies in political science increasingly use experimental methods to investigate how exposure to facts affects individual attitudes and behavior. However, the effect of information ought to vary by participant prior knowledge and beliefs. We develop an instrumental-variable approach to account for what amounts to non-compliance, and probe whether exposure to facts changes beliefs, and then attitudes, in a within respondent design. We demonstrate this method with a pre-registered study on how facts about economic interdependence with mainland China affects Taiwanese judgements about the prioritization of economic development and democracy. We find an experimentally induced upward correction causes respondents to assign greater weight on economic development, a China-appeasing position. This finding shows how public knowledge of China's rise affects attitudes toward democracy in its trading partners.

Authors

• Zikai Li, University of Chicago

• Robert Gulotty, University of Chicago