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

balanced in terms of perceived validity, objective validity, and topic. Inspecting either counter- balance separately did not affect the main findings. Experiment 1 had two main sections. To capture the beliefs and attitudes of our sample, participants first answered an 11-item collection of 7-point Likert scales measuring belief and attitudes towards the science of climate change [46]. These scales comprised of climate- adapted versions of the Belief in Science Scale (e.g., “We can only rationally believe in what is scientifically provable about climate change”) and the Attitudes Towards the Scientific Method Scale (e.g., “The use of the scientific method to form opinions, make decisions, and better understand climate change is Worthless/Useful”). Next, participants completed the classic illusory truth effect paradigm with three phases: encoding, delay, and test. Firstly, in the encoding phase, participants were told that they would see a series of trivia statements. Participants were exposed to one out of two counterbalanced subsets of claims (4 scientist, 4 skeptic, 8 weather). Each claim was shown for 8 seconds, indi- vidually in large black font on a white background. Participants then completed a delay task of 15 mins, where they read and answered questions about a passage on the topic of bread, and completed some visual rotation exercises. In the final test phase, participants rated the truth of all 32 claims (8 scientist, 8 skeptic, 16 weather claims), half of which they had seen before.

Truth ratings were made on 6-point Likert scales (1 = Definitely True to 6 = Definitely False), which we recoded for analysis so that larger numerals reflect greater perceived truth. All claims were presented in random order within the encoding and test phases, and we used two counterbalance groups (as described in the Materials section above) to ensure that each claim appeared equally often as repeated or non-repeated across participants. Thus, a claim may act as a repeated claim for one participant but appear as a non-repeated claim for another. Experiment 2 Experiment 2 is a replication of Experiment 1 with several minor changes. We i) doubled the number of participants to increase precision of measurement, ii) refined our set of target claims for objective validity, and iii) included an additional measurement of climate change attitudes which allowed greater discernment within climate science endorsers and skeptics (the Six Americas Super Short Survey [44]). In addition, iv) we asked participants to classify each claim as scientist or skeptic-aligning, thus providing a subjective measure of claim type that bears on whether participants perceived the claim as pro- or counter-attitudinal for themselves. Results for both experiments are reported together in the results section below. Participants. In Experiment 2 we doubled the number of HITs posted (N = 200) to increase precision of measurement. We received 197 completed responses (51% identified as male, 48% identified as female, 1% other; aged 23–75; 97% American) in the recruitment

period from 03/06/2021 to 09/06/2021. To control for data quality, we excluded (preregistered at https://aspredicted.org/blind.php?x=vn5y53) 26 participants who looked up the answers and a further 51 participants who failed the bot-check question, leaving 120 responses for anal- ysis (52% identified as male, 47% identified as female, 1% identified as other/non-disclosed; aged 23–75; 97% American). Design, materials, and procedure. Experiment 2 had the same design as Experiment 1. In Experiment 2, we further refined our target claims and piloted them again using the same category options as in Experiment 1. Drawing on the same selection criteria, we obtained a set of 8 target claims (4 climate scientist claims, 4 skeptic claims). Again, we created two PLOS ONE Illusory truth effect for climate-skeptical claims PLOS ONE | https://doi.org/10.1371/journal.pone.0307294 August 7, 2024 5 / 19