Do individuals selectively apply their scientific reasoning ability when communicating about scientific evidence on polarized topics?

Classical research on motivated reasoning finds that prior beliefs can influence how people use their reasoning ability in a directional manner, making them more likely to arrive at belief-consistent conclusions.Recent research has tested the related hypothesis of selective application of reasoning ability: Do those with greater reasoning ability selectively apply their greater ability to form interpretations of evidence that are more biased by their prior beliefs, compared to those with less ability?This research has found mixed results.We report the results of two preregistered experiments using a paradigm designed to directly test for the selective application of reasoning ability.We ask whether individuals with greater ability to evaluate scientific evidence quality selectively apply that ability when communicating about flawed evidence, depending upon whether the evidence is consistent or inconsistent with their beliefs.Across both experiments, we fail to find evidence for selective application of reasoning ability: higher ability participants' communications are generally biased by their prior beliefs, but no more biased than those of low-ability participants.Overall, high-ability participants generate communications that are generally more accurate and transparent than those of low-ability participants. Public Significance StatementPrior research on motivated reasoning finds that people's prior beliefs can influence how they use their reasoning ability, making them more likely to arrive at belief-consistent conclusions.We conducted two experiments to test whether those with greater reasoning ability, when confronted with evidence that is consistent or inconsistent with their prior beliefs, selectively apply their greater ability to form interpretations that are more biased than lower ability participants'.We find that, when asked to communicate about flawed research evidence, higher ability participants' communications are biased by their prior beliefs, but no more biased than those of lower ability participants.Instead, compared to those with less ability, they are more likely to create accurate and transparent communications about scientific findings on polarized topics.