Rethinking the link between cognitive sophistication and politically motivated reasoning

Abstract

Partisan disagreement is a salient feature of contemporary American politics. A surprising but robust aspect of this disagreement is that it is often the greatest among individuals who are the most cognitively sophisticated. A popular hypothesis for this phenomenon is that cognitive sophistication magnifies “politically motivated reasoning”—reasoning driven by the motivation to reach conclusions congenial to one’s political group identity. However, in the designs of studies supporting this hypothesis, the effect of political group identity is typically conflated with the effect of specific prior beliefs about the issue under study; and reasoning can be affected by such beliefs in the absence of any political group motivation. The diagnosticity of existing evidence is thus ambiguous. To shed new light on this issue, we conducted three experiments in which we statistically controlled for people’s specific prior beliefs—isolating the direct effect of political group identity—when estimating the association between their cognitive sophistication, political group identity, and reasoning in the paradigmatic design used in the literature. Despite observing a robust direct effect of political group identity (per se) on reasoning, we found no evidence that cognitive sophistication magnifies this effect. In contrast, we found fairly consistent evidence that cognitive sophistication magnifies a direct effect of specific prior beliefs on reasoning. We conclude that there is currently a lack of compelling empirical evidence that cognitive sophistication magnifies politically motivated reasoning as commonly understood.

The Implied Truth Effect: Attaching Warnings to a Subset of Fake News Headlines Increases Perceived Accuracy of Headlines Without Warnings

“…More generally, we found that a headline's veracity had a much bigger impact on accuracy perceptions than the headline's political concordance, which is in line with other recent findings about trust in mainstream versus hyperpartisan or fake news sources (Pennycook and Rand 2019a). This result also resonates with the observation that individuals who are more analytic and deliberative are less likely to believe politically concordant false news (Bronstein et al 2019;Pennycook and Rand 2019b, c), rather than more likely according to the motivated cognition account (Kahan et al 2012, Kahan 2017, and that insofar as analytic thinking is associated with political polarization, it may be because individuals who are more analytic are more likely to defer to their priors when evaluating evidence, rather than more likely to reason with the goal of protecting their identities (Tappin et al 2019). The present results therefore add to evidence suggesting that belief in fake news is not purely (or even largely) a symptom of political partisanship hijacking our ability to reason.…”

“…The logic behind this approach is straightforward: if people are warned that a headline is false, they should be less likely to believe it. Some prior work supports this line of reasoning: explicit warnings have been found to reduce the effects of subsequently corrected misinformation (Ecker et al 2010, Lewandowsky et al 2012, Chan et al 2017 and to combat politicized interpretations of science (Bolsen and Druckman 2015, Cook et al 2017, van der Linden et al 2018. Other work, however, suggests that warnings may be rendered ineffective by politically motivated reasoning, whereby people are biased against believing information that challenges their political ideology (Garrett and Weeks 2013, Flynn et al 2016, Kahan 2017.…”

Thinking clearly about causal inferences of politically motivated reasoning: why paradigmatic study designs often undermine causal inference

“…This brings us to question (2). One approach to improving causal identification of politically motivated reasoning in the paradigmatic designs is to limit the influence of people's relevant prior beliefs and information-for example, by statistical control (31,54) or by features of the design (53,55). These approaches are an improvement, but they are not panaceas: Relatedly, it is not always (or perhaps ever) straightforward to identify a priori which prior beliefs are most relevant to the reasoning task and should thus be accounted for.…”

Biased belief in the Bayesian brain: A deeper look at the evidence

“…That is, the notion that "people are more receptive to evidence that confirms their prior beliefs" (Williams, 2018, p. 142; as cited in Mercier & Sperber, 2017, p. 218). The most straightforward empirical evidence for this notion is that individuals are prone to rate information as stronger or more convincing if it confirms vs. contradicts their prior beliefs (e.g., Koehler, 1993;Lord et al, 1979;Taber & Lodge 2006;Tappin, Pennycook, & Rand, 2018). This effect is extremely robust; indeed, people are "often unable to escape the pull of their prior beliefs, which guide the processing of new information in predictable ways" (Taber & Lodge, 2006, p. 767, our emphasis).…”

“…In other words, it prevents the key inference of motivated reasoning. A direct corollary is that the results from these designs are susceptible to (confounding) explanations based on prior beliefs-because the random assignment of information not only varies the consistency of said information with peoples' preferences, political identities, and so on, but also with their prior beliefs (Tappin, Pennycook, & Rand, 2018); an "empirical catch-22" in motivated reasoning research (Ditto et al, 2018b, p. 13). In this light, then, the observed patterns of information evaluation may reduce to "people are more receptive to evidence that confirms their prior beliefs" (Williams, 2018, p. 142)-and, as pointed out above, confirmation bias in the interpretation of new information does not provide particularly convincing evidence of a violation of Bayesian inference.…”