Repetition increases belief in climate-skeptical claims, even for climate science endorsers

The observed resistance to counter-attitudinal arguments increases with attitude strength— that is, the extremity of one’s own position and the conviction with which it is held [38, 39]. If people’s truth assessment—in the present studies, their assessment of the perceived truth of climate-related claims—is largely guided by motivated reasoning, repeated exposure to counter-attitudinal claims that align with climate skeptics and climate deniers should not increase truth perceptions among people who endorse climate science. The present research In light of these different perspectives, what might we expect when people who endorse climate science are repeatedly exposed to climate-skeptic claims? A motivational account suggests that mere repetition will do very little to shape their perceptions of truth. People will presumably attend to the content of the claims and reject them as inaccurate when they conflict with their own beliefs and attitudes. In contrast, a metacognitive account suggests that people also attend to how easily a claim can be processed and use this metacognitive experience as information in judging its likely truth. If so, climate science endorsers may (i) evaluate claims that are consis- tent with their beliefs as truer than claims that are not but may (ii) nevertheless evaluate repeated counter-attitudinal claims as truer than non-repeated ones. We tested the influence of repetition on the perceived truth of pro- and counter-attitudinal claims in two experiments. In both studies, we showed people who endorsed climate science a

series of climate related claims that were either aligned with climate science (pro-attitudinal) or with climate skepticism (counter-attitudinal). After a 15-minute delay, participants evaluated the truth of claims they had seen previously and claims they had not seen before. Of interest was (i) whether a single presentation during the exposure phase increases the acceptance of a given claim as true above baseline even when (ii) the claim contradicts participants’ pro-climate sci- ence attitudes. We found that this was the case. A single repetition increased perceived validity of all claim types, including claims that contradicted the participants’ pro-climate science atti- tudes. This held even for claims that participants themselves later classified as counter-attitudi- nal and even for the strongest climate science endorsers, those classified as “Alarmed” on the Six Americas Super Short Survey on climate change beliefs scale (SASSY) [40]. Methods Power analysis and participant sample The between-items effect size for the illusory truth effect is estimated as d = 0.49 95% CI [0.43, 0.57] [2,19]. Based on the smaller estimated effect in a meta-analysis [2], and assuming α = .05, power (1-β) = .95, and a two-tailed analysis, G*Power [40] indicates that 54 participants are required to detect a truth effect in a repeated measures design. Thus, to ensure sufficient power and high precision after allowing for exclusions, we posted 100 HITs in Experiment 1 and 200 HITs in Experiment 2. For both experiments, we recruited US participants on Amazon’s online Mechanical Turk

(MTurk; www.mturk.com/mturk) platform with a � 95% HIT approval ratio and paid them $3.60 USD for completing the 30-minute experiment. When researchers follow best practice recommendations [41], MTurk participants are more attentive and reliable than participants in academic subject pools [42]. We used Qualtrics to randomly assign participants to counter- balances, and we downloaded data only after all available HITs were completed. The current studies were approved by the Human Research Ethics Committee at The Australian National University (protocol 2020/438). All participants provided informed consent before beginning the experiment by ticking a box to indicate that they had read through the participant informa- tion sheet and would like to participate in the experiment. PLOS ONE Illusory truth effect for climate-skeptical claims PLOS ONE | https://doi.org/10.1371/journal.pone.0307294 August 7, 2024 3 / 19