**Title:** Truth Misattribution **Collaborators:** Ruben Laukkonen, Benjamin Kaveladze, Jason Tangen, Jonathan Schooler. For more information, please contact Ruben Laukkonen: ruben.laukkonen@gmail.com. **Background & Rationale:** Building on the work of Dougal & Schooler (2007) on discovery misattribution, we aim to evaluate whether truth judgments about propositions such as "lithium is the lightest of all metals” (Reber & Unkelbach, 2010), can be confounded by the phenomenology that accompanies solutions to problems, and in particular the Aha! experience. Aha! experiences reliably lead to greater confidence in solutions to problems, often without any deliberate verification of the solution (Danek et al., 2014; Salvi et al., 2016; Danek & Wiley, 2017; Webb, Little, & Cropper; 2016; Hedne et al., 2016). If an Aha! experience—or more subtle phenomenology associated with solving an anagram—occurs at the same time as a propositon is evaluated, then the certainty that accompanies the phenomenology may be misattributed to the proposition, artificially increasing its perceived truth. This line of reasoning is consistent with findings that humans interpret their own phenomenology (Schacter & Singer, 1962; Dutton & Aron, 1974; Schwarz’s, 2010; Schooler, 2001; Latane & Darley, 1970, Nisbett & Wilson, 1997, Caruthers, 2009, Dougal & Schooler, 2007, Whittlesea & Williams, 1998; 2000, Johansson et al., 2004), and they rely on gut feelings and hunches to make truth attributions (Reber & Schwarz, 1999; Schwarz & Newman, 2017; Zajonc, 1968; Weaver, Garcia, Schwarz, & Miller, 2007). This experiment also tests a key hypothesis proposed by Laukkonen, Schooler, & Tangen (forthcoming) in their Eureka Heuristic model of the Aha! experience. If the Aha! experience acts as a heuristic for 'truthiness' about one's own ideas, then this heuristic may lead to false attributions of 'truthiness' if the experience is associated with an unrelated proposition. **Design & Materials:** This experiment is a mixed factorial design with three within participant variables: 2 (proposition: true or false) x 2 (problem: solved or unsolved) x 2 (Aha! experience: yes or no), and one between subjects factor (Anagram: anagram present, anagram absent, and anagram absent with delay). The dependent measure of interest is truth judgments on a 12 point scale ranging from 1 (definitely false), to 12 (definitely true). All participants are presented with the same 26 propositions (13 true and 13 false) and those in the anagram condition are also presented with 26 anagrams derived from the propositions (see materials attachment). In the delay condition no anagram is presented but the missing word in the proposition is presented with a 15 second delay in order to mimic the anagram condition as closely as possible. **Participants:** Based on Dougal and Schooler (2007), 300 participants (100 in each condition) will provide sufficient power (.8) to detect an effect size of .4 as observed in experiment 1 of Dougal and Schooler (2007). Participants are recruited using Amazon Mechanical Turk. **Procedure:** All participants are provided with video instructions, and randomly assigned to either the Anagram, No Anagram, or Delay condition. In the Anagram Condition, there are four phases to a trial: (1) presentation of an incomplete proposition, (2) resolving an anagram that completes the proposition or presentation of the solution after 13 seconds, (3) truth judgment about the completed proposition, (4) self reported Aha! experience. In the No Anagram condition, participants are presented with the same propositions with the following phases: (1) a completed proposition, (2) a truth judgment about the proposition, (3) self reported Aha! experience. In the delay condition participants are presented with (1) the incomplete proposition without the anagram, (2) the completed proposition after a 15 second delay, (3) a truth judgment about the proposition, (4) self reported Aha! experience. Here is one illustrative example from the Anagram Condition. Participants are first presented with the incomplete proposition, for example: "There are more than 100,000 craters on the…". Below the incomplete proposition participants are presented with an anagram that completes the proposition, in this case they are presented with "NOMO" (Moon). When the anagram is resolved participants see the complete proposition as: "There are more than 100,000 craters on the moon." If the anagram is not solved within 20 seconds then the solution, "moon", is presented. Participants then make a truth judgment about the completed proposition, after either solving it themselves or having the solution presented to them. Finally, on a new screen, participants report whether they experienced an Aha! moment (yes or no). In the No Anagram condition participants are simply presented with the completed proposition "There are more than 100,000 craters on the moon". They then make a truth judgment about it, and then report their Aha! experience. In the delay condition participants are presented with "There are more than 100,000 craters on the ...". After 15 seconds they are presented with the completed proposition "There are more than 100,000 craters on the moon." At the end of each condition all participants report their demographic information and three manipulation checks. Participants are asked, "did you find the answers to any of the questions online or elsewhere?". If the answer is Yes, the participant is removed. Participants are asked if they experienced any Aha! moments. If the answer is 'I don't know', or 'No', the participant is removed. Participants are asked, "Is english the language you are most comfortable using?". If the answer is no, the participant is removed. **Data Collection Done to Date:** No data have been collected to date. **Decision Rules:** An anagram is classified as solved if a correct solution is entered within the 20 second time limit. A correct solution is coded if the unscrambled word completes the proposition (see materials), and minor misspellings will be accepted. Any participant who fails to complete the full experiment will be removed. Any participants who fails to solve more than 80% of the anagrams will be removed. Any participants providing more than 80% consistent truth judgments (e.g., always indicating that propositions are false) will be removed. Any participants who provide more than 90% consistent Aha! experiences (e.g., always saying they experienced an Aha! moment) will be removed from relevant analyses. **Planned Analyses:** A mixed analysis of variance with three within participant factors: proposition (true or false), problem (solved or unsolved), Aha! experience (yes or no), and one between subjects variable, anagram (present, absent, or delayed). The dependent measure is average truth judgment ranging from 1 (definitely false), to 12 (definitely true). Appropriate follow up tests will also be conducted. A mixed analysis of variance with two within participants factors, anagram (solved or unsolved), and proposition (true or false), and one between subjects factor, anagram (present, absent, or delayed). The dependent variable is the number of Aha! experiences reported. Pearson's correlations between Aha! experiences and truth judgments for both anagram and no anagram conditions. Any exploratory analyses will be updated at the bottom of this entry. **Hypotheses:** (1) Participants will rate true claims as more likely to be true than false claims overall. (2) Participants will rate propositions as more likely to be true if an Aha! moment was experienced. (3) Participants will rate propositions as more likely to be true if an anagram was solved rather than not solved. (4) Participants in the anagram condition will rate propositions as more likely to be true on average than participants in the no anagram condition. (5) Participants in the anagram condition will report more Aha! moments than in the no anagram condition. We speculate that hypotheses three and four may be conditional on the following: (1) solved anagrams lead to more Aha! experiences than unsolved anagrams, and (2) participants in the anagram condition report more Aha! experiences than participants in the no anagram condition. We included the delay condition to ensure that there was nothing special about the delayed onset of the solution in the anagram condition (e.g., the delayed onset elicits a kind of miniature Aha! moment). Therefore we expect that the delay condition will show the same or similar pattern of results compared to the no-delay no-anagram condition. If there are differences, we should still find that the anagram condition leads to more Aha! moments and stronger truth judgments than the delay condition. **Aha! Instructions Transcript:** After you decide whether the claim is true, you will be asked whether or not you experienced an "Aha!" moment at any point in the trial. Almost everyone has experienced an Aha! moment in the past. Many people report Aha! moments while having a shower, or just before falling asleep. Try to recall an Aha! experience that you've had, and try to remember how it felt. When completing the task, try to pay attention to when Aha! moments occur at any point in the trial. When an Aha! moment occurs, it is as if the solution to the problem suddenly pops into your mind, like a lightbulb turning on. You might experience surprise, you might feel relief, and you might feel a light sense of happiness and ease. You can think of this experience as a miniature ‘Eureka moment’. You might even feel an internal sense of “Aha!,” or you might think to yourself, “of course!,” “that was so obvious”. Not experiencing an Aha! moment might feel like nothing much at all. You might simply think about the problem, and then gradually work out the solution. **Data Collection Done to Date** None **Post Data Collection Analyses** We recently finished data collection and found support for each of our hypotheses, except for the difference between truth judgments in the three between-subject conditions. We reasoned that this may be due to the fact that some anagram trials were too difficult, and therefore very few participants solved them. The anagram condition overall wouldn't be expected to lead to greater truth judgments than the other conditions if the participants were failing to solve the anagrams. Indeed, a failure to solve an anagram could lead to a disfluency effect, decreasing truth judgments. We therefore made the following hypotheses, and will be testing them shortly: 1. If we constrain our analyses to the anagrams that were solved successfully more than 70% of the time, and include only those trials in the no anagram and delay conditions, then we would expect to find that the anagram condition now shows the highest truth ratings. 2. On the other hand, if we constrain the analyses to those trials where anagrams were solved less than 30% of the time, we expect that the difference between the conditions would either not be observed, or would reverse due to the disfluency involved in failing to solve many anagrams. 3. Our argument is that it is specifically the insight experience elicited by the anagram which increases the truth judgment. Therefore, we also predict that within correctly solved anagrams, those accompanied by insight will lead to stronger truth judgments than the solutions without insight. We chose 70% and 30% as the cutoff because there was sufficient anagrams that were solved more than 70% of the time (10 out of 26) and less than 30% of the time (7 out of 26) to provide ample data points for analysis. The results of this analysis are pasted below: **Post Data Collection Results** "We then conducted the same analysis for anagrams that were solved more than 70% of the time, and then again for anagrams that were solved less than 30% of the time. For anagrams that tended to be successfully solved, the main effect of condition was marginal, F(2,265) = 2.8, p = .062. Planned comparisons showed a difference between the Anagram condition and the Delay condition, t(265) = 2.234, p < .026, and a marginal difference between the Anagram condition and the No Anagram condition, t(265) = 1.955, p < .052. The direction of these effects were such that the Anagram condition lead to the highest truth ratings (M = 7.025, SD = 1.5) relative to the Delay (M = 6.51, SD = 1.5) and the the No Anagram condition (M = 6.57, SD = 1.51). On the other hand, for anagrams that tended not to be solved (less than 30% of the time), we found no difference between the conditions, F(2,265) = 1.02, p = .363. Taken together, we found that the mere presence of the anagram is not sufficient to increase truth judgments between-subjects, instead—as in the within-subjects effects—it’s necessary that the participants successfully solve the anagram themselves in order for it to influence truth judgments. In other words, we don’t find evidence that revealing the solution to the participant is sufficient (as in RE experiments), but instead it is important that the participant independently discovers the solution (as found in Dougal & Schooler, 2007)."