The political landscape in the United States has become increasingly polarized in recent years, with political discourse often becoming heated and people relying mainly on information congenial to their ideology. This paper examines the propensity of individuals (college students and faculty in our sample) to engage in motivated reasoning — the tendency to engage in self-serving beliefs — when evaluating the accuracy of electoral polls for the 2024 Presidential Election.

Using a between subjects randomized experiment, I assess the levels of motivated reasoning when participants are presented with polling information which either favors their preferred candidate (good news) or the opponent (bad news). I hypothesize that participants are more likely to incorporate good news than bad news in their beliefs (motivated reasoning). I also expect that the degree of motivated reasoning is positively correlated with both cognitive reasoning skills and the strength of political ideology, as supported by previous research. 

Specifically, to measure cognitive reasoning skills, participants complete a Cognitive Reflection Test (CRT). Strength of political beliefs will be assessed through questions on political affiliation and ideology. Motivated reasoning is measured by the degree of belief updating. Specifically, participants first estimate their preferred candidate’s ‘true’ popularity after viewing an electoral poll showing a tie. Then, they are presented with a second poll — randomly assigned to show either good or bad news — and asked to revise their estimate. Lower levels of belief updating in response to bad news, compared to good news, would indicate a reluctance to accept unfavorable information, thus showing evidence of motivated reasoning. Therefore, my methodology allows for a nuanced analysis of the relationship between motivated reasoning, cognitive reasoning skills, and the strength of political ideology.