When Beliefs Collide with Evidence: Testing Motivated and Classical Reasoning Accounts of Evidence Interpretation

his study investigates how people reason when presented with evidence, emphasizing the role of cognitive effort. Across three experiments (NStudy1a = 505, NStudy1b = 113, NStudy2 = 767), participants completed a data interpretation task varying in difficulty, while effort was measured via reaction times (Studies 1a and 2) or cardiovascular response (Study 1b). While accuracy was higher when correct responses were concordant with participants’ beliefs and among more numerate individuals, there was no evidence that numerical ability magnified concordance effects. We also do not find evidence that investing cognitive effort depended on concordance. Instead, cognitive effort was higher among individuals with greater numeracy. Providing instructions on how to solve the data interpretation task (in Study 2) increased the overall accuracy, but there was no evidence that it altered concordance or numeracy effects. However, only under instructions we found robust, positive relationship between effort and accuracy. Overall, the results suggest that effort and ability jointly promote making accurate rather than belief-concordant conclusions, providing evidence for classical rather than motivated reasoning account of evidence interpretation.