This proposal aims at exploring the impact of thought on political attitudes. More precisely, it proposes to show that mere thought helps citizens form attitudes that are more in line with their underlying values and interests. For data, I propose a survey-based experiment to examine the impact of thought on support for federal public spending in Brazil. The design involves five treatment groups and a control group, all randomly assigned. Each group will be asked questions measuring both federal spending preferences and political knowledge, but with differences in wording, ordering, and surrounding script to manipulate information and thought. Several decades of empirical research have shown that most citizens know very little about politics. Most recent work in the area has focused on the consequences for vote choices and policy preferences. The emphasis has been on the role played by political information on attitude and preference formation. However, no attention has been given to the effect of thought alone. This proposal intends to show that political information matters, but that mere thought also helps citizens form attitudes that are more closely related to their own values and interests. To my knowledge, no other study has yet examined this question. Previous work in political and social psychology has been criticized for being conducted exclusively in the United States or at best in other industrialized countries. This project breaks new grounds by proposing to test for its hypotheses in Brazil. In sum, this project proposes to study a question left almost unexplored by political scientists, and also promises to contribute to the nascent field of comparative political psychology. Broader Impact The field of political science, and most generally the social sciences, makes extensive use of survey research to learn about individuals. attitudes, preferences, and behaviors. In the past decades, we have learned substantially about various effects of this methodology on the responses we get and on how to improve it. However, with a very few exceptions, no other studies have looked at the effects of thought on survey responses. In that sense, this project might provide new ways to improve how we ask questions during surveys, and, therefore, benefit all the disciplines that make use of this research method
