A referential theory of the repetition-induced truth effect

Large language models (LLMs) and the institutionalization of misinformation

Large language models (LLMs), such as ChatGPT, flood the Internet with true and false information, crafted and delivered with techniques that psychological science suggests will encourage people to think that information is true. What's more, as people feed this misinformation back into the Internet, emerging LLMs will adopt it and feed it back in other models. Such a scenario means we could lose access to information that helps us tell what is real from unreal - to do 'reality monitoring.' If that happens, misinformation will be the new foundation we use to plan, to make decisions, and to vote. We will lose trust in our institutions and each other.

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

Does repeated exposure to climate-skeptic claims influence their acceptance as true, even among climate science endorsers? Research with general knowledge claims shows that repeated exposure to a claim increases its perceived truth when it is encountered again. However, motivated cognition research suggests that people primarily endorse what they already believe. Across two experiments, climate science endorsers were more likely to believe claims that were consistent with their prior beliefs, but repeated exposure increased perceptions of truth for climate-science and climate-skeptic claims to a similar extent. Even counter-attitudinal claims benefit from previous exposure, highlighting the insidious effect of repetition.

Are you sure? Examining the potential benefits of truth-checking as a learning activity

Learners may be uncertain about whether encountered information is true. Uncertainty may encourage people to critically assess information's accuracy, serving as a kind of desirable difficulty that benefits learning. Uncertainty may also have negative effects, however, leading people to mistrust true information or to later misremember false information as true. In three experiments, participants read history statements. In one condition, all statements were true, and the participants knew it. In the other two conditions, some statements were true, and others were false. Participants were either told the statements' accuracy or they guessed the statements' accuracy prior to feedback, a manipulation we refer to as truth-checking. All participants were then tested on recalling the true information and on recognising true versus false statements. We observed a significant benefit of truth-checking in one of the three experiments, suggesting that truth-checking may have some potential to enhance learning, perhaps by inducing people to encode to-be-learned information more deeply than they would otherwise. Even so, the benefit may come at a cost-truth-checking took significantly longer than study alone, and it led to a greater likelihood of thinking false information was true, suggesting costs of truth-checking may tend to outweigh benefits.

Simulation-based prior knowledge elicitation for parametric Bayesian models

A central characteristic of Bayesian statistics is the ability to consistently incorporate prior knowledge into various modeling processes. In this paper, we focus on translating domain expert knowledge into corresponding prior distributions over model parameters, a process known as prior elicitation. Expert knowledge can manifest itself in diverse formats, including information about raw data, summary statistics, or model parameters. A major challenge for existing elicitation methods is how to effectively utilize all of these different formats in order to formulate prior distributions that align with the expert's expectations, regardless of the model structure. To address these challenges, we develop a simulation-based elicitation method that can learn the hyperparameters of potentially any parametric prior distribution from a wide spectrum of expert knowledge using stochastic gradient descent. We validate the effectiveness and robustness of our elicitation method in four representative simulation studies covering linear models, generalized linear models, and hierarchical models. Our results support the claim that our method is largely independent of the underlying model structure and adaptable to various elicitation techniques, including quantile-based, moment-based, and histogram-based methods.

Illusions of knowledge due to mere repetition

Repeating information increases people's belief that the repeated information is true. This truth effect has been widely researched and is relevant for topics such as fake news and misinformation. Another effect of repetition, which is also relevant to those topics, has not been extensively studied so far: Do people believe they knew something before it was repeated? We used a standard truth effect paradigm in four pre-registered experiments (total N = 773), including a presentation and judgment phase. However, instead of "true"/"false" judgments, participants indicated whether they knew a given trivia statement before participating in the experiment. Across all experiments, participants judged repeated information as "known" more often than novel information. Participants even judged repeated false information to know it to be false. In addition, participants also generated sources of their knowledge. The inability to distinguish recent information from well-established knowledge in memory adds an explanation for the persistence and strength of repetition effects on truth. The truth effect might be so robust because people believe to know the repeatedly presented information as a matter of fact.

The illusory certainty: Information repetition and impressions of truth enhance subjective confidence in validity judgments independently of the factual truth

People not only judge repeatedly perceived information as more likely being true (the so-called truth effect) they also tend to be more confident after judging the validity of repeated information. These phenomena are assumed to be caused by a higher subjective feeling of ease (i.e., fluency) when processing repeated (vs. new) information. Based on the suggestion that a higher number of coherent mental activations is promoting a fluency experience, we argue that besides repetition an already existing information network, that is (nonspecific) prior knowledge, can enhance fluency. Following this argumentation, information repetition as well as the act of judging incoming information as being true (vs. false) should feed into subjective confidence - independently of the factual truth (when judging under uncertainty). To test this, we reanalyzed two published data sets and conducted a new study. In total, participants (N = 247) gave 29,490 truth judgments and corresponding ratings of subjective confidence while attending two judgement phases (i.e., 10 min and 1 week after the exposure phase in each experiment). Results showed that (a) repetition (in 3 of 3 data sets) and (b) impressions of truth (in 2 of 3 data sets) were systematically related to higher subjective confidence. Moreover, we found (c) a significant positive interaction between repetition and impressions of truth after both intervals in all data sets. Our analyses further underline the moderating effect of time: Influences of repetition significantly decreased with increasing time interval. Notably, the factual truth did not systematically affect any of the above reported effects.

The illusory truth effect: A review of how repetition increases belief in misinformation

Repetition increases belief in information, a phenomenon known as the illusory truth effect. In laboratory experiments, the illusory truth effect has often been examined using general trivia statements as stimuli, but repetition also increases belief in misinformation, such as fake news headlines and conspiracy beliefs. Repetition even increases belief in claims that are implausible or that contradict prior knowledge. Repetition also has broader impacts beyond belief, such as increasing sharing intentions of news headlines and decreasing how unethical an act is perceived to be. Although the illusory truth effect is robust, some interventions reduce its magnitude, including instruction to focus on accuracy and awareness of the illusory truth effect. These strategies may be effective for reducing belief in misinformation.

Does Foreign-Accented Speech Affect Credibility? Evidence from the Illusory-Truth Paradigm

In a pioneering study, Lev-Ari and Keysar (2010) observed that unknown statements are judged less credible when uttered with foreign accent compared to native accent. This finding was interpreted in terms of processing fluency; when intelligibility is reduced, the credibility of the message decreases. Here, we use the illusory truth paradigm to explore how accent affects credibility. In a between-participant design, participants were exposed to unknown statements uttered by native-accented or foreign-accented speakers. After a distractor task, the same statements were presented with new statements, and participants assessed their truthfulness. Truthfulness ratings were higher for repeated statements than for new statements, replicating the illusory truth effect. Contrary to the processing fluency hypothesis, the effect was similar in both the foreign-accented and native-accented speech groups. A new group of participants rated the speakers' voices on various social traits. A negative bias against foreign speakers was observed. However, this negative-bias did not affect truth ratings.The impact of foreign-accented speech on message credibility is discussed in the context of two factors, processing fluency and out-group stereotype activation.

"That's just like, your opinion, man": the illusory truth effect on opinions

With the expanse of technology, people are constantly exposed to an abundance of information. Of vital importance is to understand how people assess the truthfulness of such information. One indicator of perceived truthfulness seems to be whether it is repeated. That is, people tend to perceive repeated information, regardless of its veracity, as more truthful than new information, also known as the illusory truth effect. In the present study, we examined whether such effect is also observed for opinions and whether the manner in which the information is encoded influenced the illusory truth effect. Across three experiments, participants (n = 552) were presented with a list of true information, misinformation, general opinion, and/or social-political opinion statements. First, participants were either instructed to indicate whether the presented statement was a fact or opinion based on its syntax structure (Exp. 1 & 2) or assign each statement to a topic category (Exp. 3). Subsequently, participants rated the truthfulness of various new and repeated statements. Results showed that repeated information, regardless of the type of information, received higher subjective truth ratings when participants simply encoded them by assigning each statement to a topic. However, when general and social-political opinions were encoded as an opinion, we found no evidence of such effect. Moreover, we found a reversed illusory truth effect for general opinion statements when only considering information that was encoded as an opinion. These findings suggest that how information is encoded plays a crucial role in evaluating truth.

An ongoing secondary task can reduce the illusory truth effect

Introduction

People are more likely to believe repeated information-this is known as the Illusory Truth Effect (ITE). Recent research on the ITE has shown that semantic processing of statements plays a key role. In our day to day experience, we are often multi-tasking which can impact our ongoing processing of information around us. In three experiments, we investigate how asking participants to engage in an ongoing secondary task in the ITE paradigm influences the magnitude of the effect of repetition on belief.

Methods

Using an adapted ITE paradigm, we embedded a secondary task into each trial of the encoding and/or test phase (e.g., having participants count the number of vowels in a target word of each trivia claim) and calculated the overall accuracy on the task.

Results

We found that the overall ITE was larger when participants had no ongoing secondary task during the experiment. Further, we predicted and found that higher accuracy on the secondary task was associated with a larger ITE.

Discussion

These findings provide initial evidence that engaging in an ongoing secondary task may reduce the impact of repetition. Our findings suggest that exploring the impact of secondary tasks on the ITE is a fruitful area for further research.

The relevance of syntactic complexity for truth judgments: A registered report

Fluency theories predict higher truth judgments for easily processed statements. We investigated two factors relevant for processing fluency: repetition and syntactic complexity. In three online experiments, we manipulated syntactic complexity by creating simple and complex versions of trivia statements. Experiments 1 and 2 replicated the repetition-based truth effect. However, syntactic complexity did not affect truth judgments although complex statements were processed slower than simple statements. This null effect is surprising given that both studies had high statistical power and varied in the relative salience of syntactic complexity. Experiment 3 provides a preregistered test of the discounting explanation by using improved trivia statements of equal length and by manipulating the salience of complexity in a randomized design. As predicted by fluency theories, simple statements were more likely judged as true than complex ones, while this effect was small and not moderated by the salience of complexity.

The illusory truth effect requires semantic coherence across repetitions

Repeated exposure to information increases its' perceived truth, and this illusory truth effect is often explained by two theoretical frameworks: the fluency account and the referential theory of truth. Whereas the fluency account suggests that prior activation of a single referent within a statement should increase its perceived truth, the referential theory makes no such predictions. The referential theory instead proposes that when a statement is processed, it activates the corresponding memory referents within that statement and strengthens the connection between these referents in the semantic memory network. Because repeated statements will have more coherent corresponding referents than new statements, they are perceived as relatively truer. Experiments 1 and 2 focused on testing the fluency account, with participants exposed to one or two of a statement's referents before evaluating that statement's truth. Experiments 3 and 4 focused on the referential theory by exposing participants to non-critical facts that linked together two of a critical statements' referents before evaluating the truth of the critical statements. We consistently observed a standard illusory truth effect, such that facts that repeated verbatim were rated as truer than new facts. However, perceived truth was not affected by prior exposure to the critical statement's topic (Experiment 1) or by prior exposure to non-critical facts related to the same topic(s) as the critical statement (Experiment 2). There was also no boost in perceived truth following prior exposure to non-critical facts that linked together two of the primary referents of the critical statement but did so in a semantically distinct manner from how those same referents were linked in the critical statement itself (Experiments 3 and 4). However, Experiment 4 demonstrated that perceived truth significantly increased if there was prior exposure to non-critical facts that linked two of the critical statement primary referents in a way that was semantically coherent with how those same referents were linked within the critical statement. Together, these results are consistent with the referential theory, and suggest that semantic consistency across repetitions plays a crucial role in leading to repetition-based illusory truth effects.

Repetition could increase the perceived truth of conspiracy theories

Conspiracy theories can be encountered repeatedly, which raises the issue of the effect of their repeated exposure on beliefs. Earlier studies found that repetition increases truth judgments of factual statements, whether they are uncertain, highly implausible, or fake news, for instance. Would this "truth effect" be observed with conspiracy statements? If so, is the effect size smaller than the typical truth effect, and is it associated with individual differences such as cognitive style and conspiracy mentality? In the present preregistered study, we addressed these three issues. We asked participants to provide binary truth judgments to conspiracy and factual statements already displayed in an exposure phase (an interest judgment task) or that were new (displayed only in the truth judgment task). We measured participants' cognitive style with the three-item Cognitive Reflection Test (CRT), and conspiracy mentality with the Conspiracy Mentality Questionnaire (CMQ). Importantly, we found that repetition increased truth judgments of conspiracy theories, unmoderated by cognitive style and conspiracy mentality. Additionally, we found that the truth effect was smaller with conspiracy theories than with uncertain factual statements, and suggest explanations for this difference. The results suggest that repetition may be a simple way to increase belief in conspiracy theories. Whether repetition increases conspiracy beliefs in natural settings and how it contributes to conspiracism compared to other factors are important questions for future research.

Repeated simulation increases belief in the future occurrence of uncertain events

The feeling that an imagined event will or will not occur in the future - referred to as belief in future occurrence - plays a key role in guiding our decisions and actions. Recent research suggests that this belief may increase with repeated simulation of future events, but the boundary conditions for this effect remain unclear. Considering the key role of autobiographical knowledge in shaping belief in occurrence, we suggest that the effect of repeated simulation only occurs when prior autobiographical knowledge does not clearly support or contradict the occurrence of the imagined event. To test this hypothesis, we investigated the repetition effect for events that were either plausible or implausible due to their coherence or incoherence with autobiographical knowledge (Experiment 1), and for events that initially appeared uncertain because they were not clearly supported or contradicted by autobiographical knowledge (Experiment 2). We found that all types of events became more detailed and took less time to construct after repeated simulation, but belief in their future occurrence increased only for uncertain events; repetition did not influence belief for events already believed or considered implausible. These findings show that the effect of repeated simulation on belief in future occurrence depends on the consistency of imagined events with autobiographical knowledge.

Knowledge-building in surgery: An evolving craft

In surgery, the construction of knowledge is shaped in two scenarios: evidence-based medicine, and experiential learning. The former has become an essential pillar in this process. Based on this fact, clinical practice guidelines have been developed. However, their transient nature, constant renewal, and lack of individualization distance them from daily clinical activity. On the other hand, there is the construction of knowledge through vicarious and experiential learning. Its existence is linked to the beginning of medicine, and its importance is irrefutable. However, it is laden with the "truth effect", surgical dogma and singularism. When integrating the construction of knowledge in these scenarios, critical thinking skills are paramount and necessary to guide the surgeon in the decision-making process within the context of surgical dynamism. This article explores this situation and its impact.

Fake News and the Sleeper Effect in Social Media Posts: the Case of Perception of Safety in the Workplace

Fake news and misinformation on social media platforms are two of the biggest problems of the last few years. Understanding the underlying mechanisms of memory is of fundamental importance to develop specific intervention programs. In this study, 324 white-collar workers viewed Facebook posts focused on coronavirus disease-2019 prevention norms in the workplace. In a within-participants design, we manipulated the message and the source to expose each participant to real news, real news presented by a discounting cue (sleeper effect condition), and fake news. The results show that participants were more susceptible to fake news during a 1-week delayed posttest following a memory recall process. Furthermore, they remembered the message easily, but not the source, which did not differ in the real-news conditions. We discuss the results, mentioning the sleeper effect and fake news theories.

The evasive truth: do mere exposures at the subliminal and supraliminal levels drive the illusory truth effect?

The subjective truth of a statement is boosted by mere exposure to itself or a part of itself. This phenomenon is referred to as the illusory truth effect. We examined whether subliminal pre-exposure to the statement topic would increase its subjective truth. In the exposure phase, participants observed the topic, which was presented supraliminally or subliminally. After the exposure phase, they rated the subjective truth of the statement. If unconscious processing contributed to the illusory truth effect, subliminal exposure to the topic would increase the subjective truth of the statement. On the other hand, if the illusory truth effect required conscious and controlled processing, increases in the subjective truth of a statement would be induced only by supraliminal exposure to the topic. The results showed that the illusory truth effect was not found in either supraliminal or subliminal groups. Our findings provide no reliable evidence that pre-exposure to the statement topic saliently promotes its subjective truth.

Lack of source memory as a potential marker of early assimilation of novel items into current knowledge

In daily life, humans process a plethora of new information that can be either consistent (familiar) or inconsistent (novel) with prior knowledge. Over time, both types of information can integrate into our accumulated knowledge base via distinct pathways. However, the mnemonic processes supporting the integration of information that is inconsistent with prior knowledge remain under-characterized. In the current study, we used functional magnetic resonance imaging (fMRI) to examine the initial assimilation of novel items into the semantic network. Participants saw three repetitions of adjective-noun word pairs that were either consistent or inconsistent with prior knowledge. Twenty-four hours later, they were presented with the same stimuli again while undergoing fMRI scans. Outside the scanner, participants completed a surprise recognition test. We found that when the episodic context associated with initially inconsistent items was irretrievable, the neural signature of these items was indistinguishable from that of consistent items. In contrast, initially inconsistent items with accessible episodic contexts showed neural signatures that differed from those associated with consistent items. We suggest that, at least one day post encoding, items inconsistent with prior knowledge can show early assimilation into the semantic network only when their episodic contexts become inaccessible during retrieval, thus evoking a sense of familiarity.

The Role of Valence Matching in the Truth-by-Repetition Effect

People judge repeated information as truer than new information, a “truth-by-repetition” effect. Because repetition increases processing fluency, which is assumed to elicit positive affect, participants may match their positive experience associated with repeated information with a positive (“true”) rather than negative (“false”) response. We tested this valence-matching hypothesis in a preregistered experiment by manipulating the affective congruency of the response format. Specifically, in the congruent condition, participants had to select a positive (negative) picture to respond “true” (“false”). In the incongruent condition, we reversed these associations. In line with the valence matching hypothesis, the truth-by-repetition effect was larger in the congruent than incongruent condition. However, the effect was small, and Bayesian analyses were inconclusive. In addition, the truth-by-repetition effect was significant in both response format conditions. The results suggest a possible contribution of a valence matching process to the truth-by-repetition effect, but one that does not challenge extant models.

Recalling fake news during real news corrections can impair or enhance memory updating: the role of recollection-based retrieval

Fake news can impair memory leading to societal controversies such as COVID-19 vaccine efficacy. The pernicious influence of fake news is clear when ineffective corrections leave memories outdated. A key theoretical issue is whether people should recall fake news while reading corrections with contradictory details. The familiarity backfire view proposes that recalling fake news increases its familiarity, leading to interference. However, the integrative encoding view proposes that recalling fake news promotes co-activation and binding of contradictory details, leading to facilitation. Two experiments examined if one theory better accounts for memory updating after participants recalled actual fake news details when reading headlines that corrected misinformation. In Phase 1, participants read real and fake news headlines of unclear veracity taken from various internet sources. In Phase 2, participants read real news headlines that reaffirmed real news and corrected fake news from Phase 1. When they detected that Phase 2 real news corrected fake news, they attempted to recall Phase 1 fake news. In Phase 3, participants first recalled real news details. When they remembered that those details were corrections from Phase 2, they attempted to recall fake news from Phase 1. Recalling fake news when noticing corrections in Phase 2 led to better memory for real news in Phase 3 when fake news was recalled again and worse memory for real news in Phase 3 when fake news was not recalled again. Both views explain part of the memory differences associated with recalling fake news during corrections, but only when considering whether people recollected that fake news had been corrected.

A reproducible systematic map of research on the illusory truth effect

People believe information more if they have encountered it before, a finding known as the illusory truth effect. But what is the evidence for the generality and pervasiveness of the illusory truth effect? Our preregistered systematic map describes the existing knowledge base and objectively assesses the quality, completeness and interpretability of the evidence provided by empirical studies in the literature. A systematic search of 16 bibliographic and grey literature databases identified 93 reports with a total of 181 eligible studies. All studies were conducted at Western universities, and most used convenience samples. Most studies used verbatim repetition of trivia statements in a single testing session with a minimal delay between exposure and test. The exposure tasks, filler tasks and truth measures varied substantially across studies, with no standardisation of materials or procedures. Many reports lacked transparency, both in terms of open science practices and reporting of descriptive statistics and exclusions. Systematic mapping resulted in a searchable database of illusory truth effect studies ( https://osf.io/37xma/ ). Key limitations of the current literature include the need for greater diversity of materials as stimuli (e.g., political or health contents), more participants from non-Western countries, studies examining effects of multiple repetitions and longer intersession intervals, and closer examination of the dependency of effects on the choice of exposure task and truth measure. These gaps could be investigated using carefully designed multi-lab studies. With a lack of external replications, preregistrations, data and code, verifying replicability and robustness is only possible for a small number of studies.

Sensitivity to Evidential Dependencies in Judgments Under Uncertainty

According to Bayesian models of judgment, testimony from independent informants has more evidential value than dependent testimony. Three experiments investigated learners' sensitivity to this distinction. Each experiment used a social version of the balls-and-urns task, in which participants judged which of two urns was the most likely source of evidence presented by multiple informants. Informants either provided independent testimony based solely on their own observations or dependent-sequential testimony that considered the testimonies of previous informants. Although participants updated their beliefs with additional evidence, this updating was generally insensitive to evidential dependency (Experiments 1 and 2). A notable exception was when individuals were separated according to their beliefs about the relative value of independent and sequential evidence. Those who viewed independent evidence as having greater value subsequently gave more weight to independent testimony in the balls-and-urns task (Experiment 3), in line with the predictions of a Bayesian model. Our findings suggest that only a minority of individuals conform to Bayesian predictions in the relative weighting of independent and dependent evidence in judgments under uncertainty.

Monetary incentives do not reduce the repetition-induced truth effect

People rate and judge repeated information more true than novel information. This truth-by-repetition effect is of relevance for explaining belief in fake news, conspiracy theories, or misinformation effects. To ascertain whether increased motivation could reduce this effect, we tested the influence of monetary incentives on participants’ truth judgments. We used a standard truth paradigm, consisting of a presentation and judgment phase with factually true and false information, and incentivized every truth judgment. Monetary incentives may influence truth judgments in two ways. First, participants may rely more on relevant knowledge, leading to better discrimination between true and false statements. Second, participants may rely less on repetition, leading to a lower bias to respond “true.” We tested these predictions in a preregistered and high-powered experiment. However, incentives did not influence the percentage of “true” judgments or correct responses in general, despite participants’ longer response times in the incentivized conditions and evidence for knowledge about the statements. Our findings show that even monetary consequences do not protect against the truth-by-repetition effect, further substantiating its robustness and relevance and highlighting its potential hazardous effects when used in purposeful misinformation.

Understanding the dynamics emerging from infodemics: a call to action for interdisciplinary research

Research on infodemics, i.e., the rapid spread of (mis)information related to a hazardous event, such as the COVID-19 pandemic, requires integrating a multiplicity of scientific disciplines. The dynamics emerging from infodemics have the potential to generate complex behavioral patterns. To react appropriately, it is of ultimate importance for the fields of Business and Economics to understand these dynamics. In the short run, they might lead to an adaptation in household spending or to a shift in buying behavior towards online providers. In the long run, changes in investments, consumer behavior, and markets are to be expected. We argue that the dynamics emerge from complex interactions among multiple factors, such as information and misinformation accessible to individuals and the formation and revision of beliefs. (Mis)information accessible to individuals is, amongst others, affected by algorithms specifically designed to provide personalized information, while automated fact-checking algorithms can help reduce the amount of circulating misinformation. The formation and revision of individual (and probably false) beliefs and individual fact-checking and interpretation of information are heavily affected by linguistic patterns inherent to information during pandemics and infodemics and further factors, such as affect, intuition, and motives. We argue that, to get a deep(er) understanding of the dynamics emerging from infodemics, the fields of Business and Economics should integrate the perspectives of Computer Science and Information Systems, (Computational) Linguistics, and Cognitive Science into the wider context of economic systems (e.g., organizations, markets or industries) and propose a way to do so. As research on infodemics is a strongly interdisciplinary field and the integration of the above-mentioned disciplines is a first step towards a holistic approach, we conclude with a call to action which should encourage researchers to collaborate across scientific disciplines and unfold collective creativity, which will substantially advance research on infodemics.

Identifying False Human Papillomavirus (HPV) Vaccine Information and Corresponding Risk Perceptions From Twitter: Advanced Predictive Models

Background

The vaccination uptake rates of the human papillomavirus (HPV) vaccine remain low despite the fact that the effectiveness of HPV vaccines has been established for more than a decade. Vaccine hesitancy is in part due to false information about HPV vaccines on social media. Combating false HPV vaccine information is a reasonable step to addressing vaccine hesitancy.

Objective

Given the substantial harm of false HPV vaccine information, there is an urgent need to identify false social media messages before it goes viral. The goal of the study is to develop a systematic and generalizable approach to identifying false HPV vaccine information on social media.

Methods

This study used machine learning and natural language processing to develop a series of classification models and causality mining methods to identify and examine true and false HPV vaccine–related information on Twitter.

Results

We found that the convolutional neural network model outperformed all other models in identifying tweets containing false HPV vaccine–related information (F score=91.95). We also developed completely unsupervised causality mining models to identify HPV vaccine candidate effects for capturing risk perceptions of HPV vaccines. Furthermore, we found that false information contained mostly loss-framed messages focusing on the potential risk of vaccines covering a variety of topics using more diverse vocabulary, while true information contained both gain- and loss-framed messages focusing on the effectiveness of vaccines covering fewer topics using relatively limited vocabulary.

Conclusions

Our research demonstrated the feasibility and effectiveness of using predictive models to identify false HPV vaccine information and its risk perceptions on social media.

Judging fast and slow: The truth effect does not increase under time-pressure conditions

Due to the information overload in today’s digital age, people may sometimes feel pressured to process and judge information especially fast. In three experiments, we examined whether time pressure increases the repetition-based truth effect — the tendency to judge repeatedly encountered statements more likely as “true” than novel statements. Based on the Heuristic-Systematic Model, a dual-process model in the field of persuasion research, we expected that time pressure would boost the truth effect by increasing reliance on processing fluency as a presumably heuristic cue for truth, and by decreasing knowledge retrieval as a presumably slow and systematic process that determines truth judgments. However, contrary to our expectation, time pressure did not moderate the truth effect. Importantly, this was the case for difficult statements, for which most people lack prior knowledge, as well as for easy statements, for which most people hold relevant knowledge. Overall, the findings clearly speak against the conception of fast, fluency-based truth judgments versus slow, knowledge-based truth judgments. In contrast, the results are compatible with a referential theory of the truth effect that does not distinguish between different types of truth judgments. Instead, it assumes that truth judgments rely on the coherence of localized networks in people’s semantic memory, formed by both repetition and prior knowledge.

The effects of repeating false and misleading information on belief

False and misleading information is readily accessible in people's environments, oftentimes reaching people repeatedly. This repeated exposure can significantly affect people's beliefs about the world, as has been noted by scholars in political science, communication, and cognitive, developmental, and social psychology. In particular, repetition increases belief in false information, even when the misinformation contradicts prior knowledge. We review work across these disciplines, identifying factors that may heighten, diminish, or have no impact on these adverse effects of repetition on belief. Specifically, we organize our discussion around variations in what information is repeated, to whom the information is repeated, how people interact with this repetition, and how people's beliefs are measured. A key cross-disciplinary theme is that the most influential factor is how carefully or critically people process the false information. However, several open questions remain when comparing findings across different fields and approaches. We conclude by noting a need for more interdisciplinary work to help resolve these questions, as well as a need for more work in naturalistic settings so that we can better understand and combat the effects of repeated circulation of false and misleading information in society. This article is categorized under: Psychology > Memory Psychology > Reasoning and Decision Making.

The Trajectory of Truth: A Longitudinal Study of the Illusory Truth Effect

Repeated statements are rated as subjectively truer than comparable new statements, even though repetition alone provides no new, probative information (the illusory truth effect). Contrary to some theoretical predictions, the illusory truth effect seems to be similar in magnitude for repetitions occurring after minutes or weeks. This Registered Report describes a longitudinal investigation of the illusory truth effect (n = 608, n = 567 analysed) in which we systematically manipulated intersession interval (immediately, one day, one week, and one month) in order to test whether the illusory truth effect is immune to time. Both our hypotheses were supported: We observed an illusory truth effect at all four intervals (overall effect: χ 2(1) = 169.91; M repeated = 4.52, M new = 4.14; H1), with the effect diminishing as delay increased (H2). False information repeated over short timescales might have a greater effect on truth judgements than repetitions over longer timescales. Researchers should consider the implications of the choice of intersession interval when designing future illusory truth effect research.

Are foreign-accented speakers that 'incredible'? The impact of the speaker's indexical properties on sentence processing

This study investigated the impact of the speaker's identity generated by the voice on sentence processing. We examined the relation between ERP components associated with the processing of the voice (N100 and P200) from voice onset and those associated with sentence processing (N400 and late positivity) from critical word onset. We presented Dutch native speakers with sentences containing true (and known) information, unknown (but true) information or information violating world knowledge and had them perform a truth evaluation task. Sentences were spoken either in a native or a foreign accent. Truth evaluation judgments were not different for statements spoken by the native-accented and the foreign-accented speakers. Reduced N100 and P200 were observed in response to the foreign speaker's voice compared to the native speaker's. While statements containing unknown information or world knowledge violations generated a larger N400 than true statements in the native condition, they were not significantly different in the foreign condition, suggesting shallower processing of foreign-accented speech. The N100 was a significant predictor for the N400 in that the reduced N100 observed for the foreign speaker compared to the native speaker was related to a smaller N400 effect. These finding suggest that the impression of the speaker that listeners rapidly form from the voice affects semantic processing, which confirms that speaker's identity and language comprehension cannot be dissociated.

The effects of repetition frequency on the illusory truth effect

Repeated information is often perceived as more truthful than new information. This finding is known as the illusory truth effect, and it is typically thought to occur because repetition increases processing fluency. Because fluency and truth are frequently correlated in the real world, people learn to use processing fluency as a marker for truthfulness. Although the illusory truth effect is a robust phenomenon, almost all studies examining it have used three or fewer repetitions. To address this limitation, we conducted two experiments using a larger number of repetitions. In Experiment 1, we showed participants trivia statements up to 9 times and in Experiment 2 statements were shown up to 27 times. Later, participants rated the truthfulness of the previously seen statements and of new statements. In both experiments, we found that perceived truthfulness increased as the number of repetitions increased. However, these truth rating increases were logarithmic in shape. The largest increase in perceived truth came from encountering a statement for the second time, and beyond this were incrementally smaller increases in perceived truth for each additional repetition. These findings add to our theoretical understanding of the illusory truth effect and have applications for advertising, politics, and the propagation of "fake news."

Default Inheritance in Modified Statements: Bias or Inference?

It is a fact that human subjects rate sentences about typical properties such as "Ravens are black" as very likely to be true. In comparison, modified sentences such as "Feathered ravens are black" receive lower ratings, especially if the modifier is atypical for the noun, as in "Jungle ravens are black". This is called the modifier effect. However, the likelihood of the unmodified statement influences the perceived likelihood of the modified statement: the higher the rated likelihood of the unmodified sentence, the higher the rated likelihood of the modified one. That means the modifier effect does not fully block default inheritance of typical properties from nouns to modified nouns. This paper discusses this inheritance effect. In particular, I ask whether it is the direct result of composing concepts from nouns, that is, a bias toward "black" when processing "raven". I report a series of experiments in which I find no evidence for a direct inheritance from composition. This supports the view that default inheritance is rather an inference than a bias.

Is it all about the feeling? Affective and (meta-)cognitive mechanisms underlying the truth effect

People are more likely to judge repeatedly perceived statements as true. A decisive explanation for this so-called truth effect is that the repeated information can be processed more fluently than new information and that this fluency experience renders the information more familiar and trustworthy. Little is known, however, regarding whether and how affective states and dispositional cognitive preferences influence the truth effect. To this end, we conducted two experiments in which we manipulated (a) processing fluency via repetition, (b) the time interval (10 min vs. 1 week) between repetitions, and (c) short-term affective states using the presentation of emotional faces (Experiment 1) or the presence of an irrelevant source for changes in affective states (Experiment 2). Additionally, we assessed the dispositional variables need for cognitive closure (NCC), preference for deliberation (PD) and preference for intuition (PI). Results of Experiment 1 showed that the truth effect was significantly reduced for statements that were followed by a negative prime, although this was the case only for the longer repetition lag. Furthermore, higher NCC and lower PD scores were associated with an increased truth effect. Results of Experiment 2 replicated the moderating role of NCC and further showed that participants, who were provided with an alternative source for changes in their affective states, showed a reduced truth effect. Together, the findings suggest that (a) fluency-related changes in affective states may be (co-)responsible for the truth effect, (b) the truth effect is decreased when the repetition interval is long rather than short, and (c) the truth effect is increased for individuals with a higher need for cognitive closure. Theoretical implications of these findings are discussed.

Perceived truth of statements and simulated social media postings: an experimental investigation of source credibility, repeated exposure, and presentation format

To better understand the spread of fake news in the Internet age, it is important to uncover the variables that influence the perceived truth of information. Although previous research identified several reliable predictors of truth judgments—such as source credibility, repeated information exposure, and presentation format—little is known about their simultaneous effects. In a series of four experiments, we investigated how the abovementioned factors jointly affect the perceived truth of statements (Experiments 1 and 2) and simulated social media postings (Experiments 3 and 4). Experiment 1 explored the role of source credibility (high vs. low vs. no source information) and presentation format (with vs. without a picture). In Experiments 2 and 3, we additionally manipulated repeated exposure (yes vs. no). Finally, Experiment 4 examined the role of source credibility (high vs. low) and type of repetition (congruent vs. incongruent vs. no repetition) in further detail. In sum, we found no effect of presentation format on truth judgments, but strong, additive effects of source credibility and repetition. Truth judgments were higher for information presented by credible sources than non-credible sources and information without sources. Moreover, congruent (i.e., verbatim) repetition increased perceived truth whereas semantically incongruent repetition decreased perceived truth, irrespectively of the source. Our findings show that people do not rely on a single judgment cue when evaluating a statement’s truth but take source credibility and their meta-cognitive feelings into account.

How explicit warnings reduce the truth effect: A multinomial modeling approach

The finding that repeating a statement typically increases its perceived truth has been referred to as the truth effect. Previous research has found that warning participants about the truth effect can successfully reduce, but not eliminate the effect. We used a multinomial modeling approach to investigate how warnings affect the cognitive processes that are assumed to underlie judgments of truth. In a laboratory experiment (N = 167), half of the participants were warned about the truth effect before judging the truth of repeated and new statements. Importantly, whereas half of the presented statements were of relatively unknown validity, participants could likely identify the correct truth status for the other half of the statements by drawing on stored knowledge. Multinomial modeling analyses revealed that warning instructions did not affect the retrieval of knowledge or participants' guessing behavior relative to a control condition. Instead, warned participants exhibited a significantly reduced tendency to rely on experiential information such as processing fluency when judging a repeated statement's truth. However, this was only the case for statements for which participants held relevant knowledge. These results are consistent with the notion that it is possible to discount metacognitive experiences such as processing ease when their informational value is questioned. Specifically, our findings suggest that people are less likely to base their judgments of truth on experiential information and metacognitive experiences induced by repetition if (a) they are warned about the deceptive power of repetition, and (b) other valid cues are available to inform their judgments.

The truth revisited: Bayesian analysis of individual differences in the truth effect

The repetition-induced truth effect refers to a phenomenon where people rate repeated statements as more likely true than novel statements. In this paper, we document qualitative individual differences in the effect. While the overwhelming majority of participants display the usual positive truth effect, a minority are the opposite-they reliably discount the validity of repeated statements, what we refer to as negative truth effect. We examine eight truth-effect data sets where individual-level data are curated. These sets are composed of 1105 individuals performing 38,904 judgments. Through Bayes factor model comparison, we show that reliable negative truth effects occur in five of the eight data sets. The negative truth effect is informative because it seems unreasonable that the mechanisms mediating the positive truth effect are the same that lead to a discounting of repeated statements' validity. Moreover, the presence of qualitative differences motivates a different type of analysis of individual differences based on ordinal (i.e., Which sign does the effect have?) rather than metric measures. To our knowledge, this paper reports the first such reliable qualitative differences in a cognitive task.

When fiction becomes fact: exaggerating host manipulation by parasites

In an era where some find fake news around every corner, the use of sensationalism has inevitably found its way into the scientific literature. This is especially the case for host manipulation by parasites, a phenomenon in which a parasite causes remarkable change in the appearance or behaviour of its host. This concept, which has deservedly garnered popular interest throughout the world in recent years, is nearly 50 years old. In the past two decades, the use of scientific metaphors, including anthropomorphisms and science fiction, to describe host manipulation has become more and more prevalent. It is possible that the repeated use of such catchy, yet misleading words in both the popular media and the scientific literature could unintentionally hamper our understanding of the complexity and extent of host manipulation, ultimately shaping its narrative in part or in full. In this commentary, the impacts of exaggerating host manipulation are brought to light by examining trends in the use of embellishing words. By looking at key examples of exaggerated claims from widely reported host-parasite systems found in the recent scientific literature, it would appear that some of the fiction surrounding host manipulation has since become fact.

Repetition increases both the perceived truth and fakeness of information: An ecological account

People believe repeated statements more compared to new statements - they show a truth by repetition effect. In three pre-registered experiments, we show that repetition may also increase perceptions that statements are used as fake news on social media, irrespective of the factual truth or falsehood of the statements (Experiment 1 & 2), but that repetition reduces perceptions of falsehood when the context of judgment is left unspecified (Experiment 3). On a theoretical level, the findings support an ecological account of repetition effects, as opposed to either a fluency-as-positivity or to an amplification account of these effects. On a practical level, they qualify the influence of repetition on the perception of fake news.

Adding Types, But Not Tokens, Affects Property Induction

The extent to which we generalize a novel property from a sample of familiar instances to novel instances depends on the sample composition. Previous property induction experiments have only used samples consisting of novel types (unique entities). Because real-world evidence samples often contain redundant tokens (repetitions of the same entity), we studied the effects on property induction of adding types and tokens to an observed sample. In Experiments 1-3, we presented participants with a sample of birds or flowers known to have a novel property and probed whether this property generalized to novel items varying in similarity to the initial sample. Increasing the number of novel types (e.g., new birds with the target property) in a sample produced tightening, promoting property generalization to highly similar stimuli but decreasing generalization to less similar stimuli. On the other hand, increasing the number of tokens (e.g., repeated presentations of the same bird with the target property) had little effect on generalization. Experiment 4 showed that repeated tokens are encoded and can benefit recognition, but appear to be given little weight when inferring property generalization. We modified an existing Bayesian model of induction (Navarro, Dry, & Lee, 2012) to account for both the information added by new types and the discounting of information conveyed by tokens.

The Effect of Repetition on Truth Judgments Across Development

According to numerous research studies, when adults hear a statement twice, they are more likely to think it is true compared with when they have heard it only once. Multiple theoretical explanations exist for this illusory-truth effect. However, none of the current theories fully explains how or why people begin to use repetition as a cue for truth. In this preregistered study, we investigated those developmental origins in twenty-four 5-year-olds, twenty-four 10-year-olds, and 32 adults. If the link between repetition and truth is learned implicitly, then even 5-year-olds should show the effect. Alternatively, realizing this connection may require metacognition and intentional reflection, skills acquired later in development. Repetition increased truth judgments for all three age groups, and prior knowledge did not protect participants from the effects of repetition. These results suggest that the illusory-truth effect is a universal effect learned at a young age.

Truthiness, the illusory truth effect, and the role of need for cognition

Ease of processing-cognitive fluency-is a central input in assessments of truth, but little is known about individual differences in susceptibility to fluency-based biases in truth assessment. Focusing on two paradigms-truthiness and the illusory truth effect-we consider the role of Need for Cognition (NFC), an individual difference variable capturing one's preference for elaborative thought. Across five experiments, we replicated basic truthiness and illusory truth effects. We found very little evidence that NFC moderates truthiness. However, we found some evidence that (without an experimental warning), people high on NFC may be more susceptible to the illusory truth effect. This may reflect that elaborative thought increases the fluency with which encoded statements are processed after a delay (thus increasing the illusory truth effect). Future research may fruitfully test whether the influence of NFC and other individual difference measures depends on whether people are making immediate or delayed truth judgments.

Judging Truth

Deceptive claims surround us, embedded in fake news, advertisements, political propaganda, and rumors. How do people know what to believe? Truth judgments reflect inferences drawn from three types of information: base rates, feelings, and consistency with information retrieved from memory. First, people exhibit a bias to accept incoming information, because most claims in our environments are true. Second, people interpret feelings, like ease of processing, as evidence of truth. And third, people can (but do not always) consider whether assertions match facts and source information stored in memory. This three-part framework predicts specific illusions (e.g., truthiness, illusory truth), offers ways to correct stubborn misconceptions, and suggests the importance of converging cues in a post-truth world, where falsehoods travel further and faster than the truth.