Perry's earlier point makes me think of a question for any of the political scientists present who know the research: How effective are speeches, ads or talking points at convincing voters of things that are dramatically exaggerated or just plain false?