May 2020

IZA DP No. 13241: Deliberation Enhances the Confirmation Bias: An Examination of Politics and Religion

Existing research has documented the confirmation bias in the domain of politics, but relatively little research has examined the confirmation bias in religion. I developed a novel task in the religious domain and compare confirmation bias evidence to that observed in the political domain. Using a preregistered data collection and analysis plan, I examined data from n=402 participants prescreened by political and religious beliefs. Participants were administered the online task that examined selective information exposure and perceived strength of arguments that are incongruent to one's own beliefs regarding "gun control" and the "existence of God". Results documented a confirmation bias in both information exposure and perceived argument strength. I also examined the hypothesis that the confirmation bias is stronger in situations where more thought or deliberation is brought to bear on the task. The evidence here depends on the measure of deliberation used, but generally supports this hypothesis. For example, the data showed that individuals who have thought a lot about the topic at hand (gun control and the existence of God) displayed a stronger confirmation bias in perceived argument strength than those having thought less about the issue. This paper contributes by offering new evidence documenting the confirmation bias directly compared across domains. And, the findings regarding how deliberation may worsen the bias are in line with previous research suggesting the confirmation bias may be unlike other decision biases — this bias may thrive when the decision maker is more is more deliberative or thoughtful.