What Is the Illusory Truth Effect?
The illusory truth effect is the experimentally observed tendency for a repeated statement to receive a higher rating of truth, accuracy, or credibility than a comparable statement encountered for the first time. A claim can become easier to process after prior exposure, and that change in processing experience can influence a person's judgment of whether the claim is true. The effect concerns perceived truth; repetition does not alter the factual accuracy of what has been said. It can increase acceptance of correct information and, under relevant conditions, of false information. Ye et al., 2026 provide the most recent large quantitative synthesis.
Consider a plausible but unfamiliar claim about the natural world. One reader encounters it once, while another encounters it on several occasions. If the readers are later asked to evaluate its accuracy, the reader who saw it before may give it a higher credibility rating even though neither has received new evidence. The essential psychological comparison is between exposure histories, holding statement content and other influences as constant as the study allows. This is why repetition can distort belief evaluation without changing the underlying facts.
The illusory truth effect belongs to the psychology of belief formation and evidence evaluation. It is a cognitive judgment effect defined relative to a truth-evaluation task, rather than a clinical symptom, personality diagnosis, or logical fallacy. Its manifestation depends on the content, the initial processing task, the context of exposure, source information, and the subsequent judgment. It is closely connected to familiarity, processing fluency, source monitoring, and memory, although these constructs address different psychological questions. Udry and Barber, 2024 review its significance for misinformation.
The Scientific Origin of the Effect
The 1977 repetition experiments
The foundational demonstration by Lynn Hasher, David Goldstein, and Thomas Toppino appeared in 1977. Participants judged the truth of statements drawn from unfamiliar areas of general knowledge on three occasions separated by two-week intervals. Selected statements recurred, whereas comparison statements were new. Truth judgments rose for repeated items, including false items, while ratings for nonrepeated items did not show the same pattern. The experiments established a testable difference between familiarity acquired through exposure and the information required to establish factual validity. Hasher, Goldstein, and Toppino, 1977.
The early paradigm remains influential because it separates a statement's objective veracity from a participant's subjective certainty. Repetition is manipulated experimentally, and perceived truth is measured later. Subsequent studies have adopted shorter delays, different materials, varied repetition schedules, explicit warnings, and richer media formats. Results must be interpreted at the level of those particular designs, rather than as a claim that every repeated message becomes persuasive.
From early experiments to quantitative synthesis
Dechêne, Stahl, Hansen, and Wänke's 2010 meta-analytic review consolidated evidence for the truth effect and examined methodological moderators. Later, Henderson, Westwood, and Simons conducted a preregistered systematic map that located 93 reports containing 181 eligible studies. Their mapping exposed a striking concentration of Western university convenience samples, trivia-like verbal statements, verbatim repetition, and relatively short testing intervals. The map was a methodological audit of what had been studied, and its findings matter when generalizing from a laboratory task to an actual media environment. Dechêne et al., 2010; Henderson et al., 2022.
The 2026 systematic review and meta-analysis by Ye and colleagues integrated 182 studies and 366 effect sizes, representing 31,184 participants in work published from 1977 through 2025. After adjustment for small-study effects, the pooled estimate was Hedges' g = 0.37, with a 95% confidence interval from 0.30 to 0.44. The authors also documented substantial heterogeneity within and between studies. The pooled standardized mean difference describes an average contrast in measured truth judgments; it is not a percentage of people deceived and cannot be read as a probability that one particular headline will be believed. Ye et al., 2026.
What Actually Changes When a Statement Is Repeated?
Processing fluency and the experience of ease
Processing fluency is the subjective ease or efficiency with which information is perceived, comprehended, or retrieved. Prior exposure can make a sentence easier to recognize and interpret on a later encounter. People may use that ease as information when judging truth because familiar, coherent information often feels cognitively accessible. The judgment can therefore shift even when the reader cannot identify the original occasion on which the statement appeared.
Processing fluency is an explanatory mechanism, not a substitute for direct measurement of belief. A researcher needs to distinguish the manipulated repetition from the intermediate experience of fluency and the final truth judgment. Studies can manipulate perceptual clarity, exposure, encoding instructions, or the validity of the fluency cue to test whether ease contributes causally. Unkelbach's experiments showed that learning a different relationship between fluency and truth can even reverse the usual direction of the truth effect under laboratory conditions. That result supports a learned-cue interpretation and sets boundaries on a purely automatic account. Unkelbach, 2007.
Familiarity, recognition, and source monitoring
Familiarity is an experience that an item has been encountered before. Recollection adds potentially richer information about where, when, and why that item was encountered. A repeated statement can feel familiar while its origin remains obscure. Someone may remember a sentence's content but not remember that it first appeared as a question, a rumor, a fictional example, or a warning. Source-monitoring failures help explain why the history of an assertion can become disconnected from its evidentiary value.
Speckmann and Unkelbach examined a related illusion of prior knowledge in four preregistered experiments with 773 participants. Repeated trivia statements were more likely than new statements to be classified as information the participant had already known before the experiment. This included some repeated false material. Their finding specifically concerns judgments of prior knowledge, and it helps explain why familiarity can be retrospectively mistaken for an established fact. Speckmann and Unkelbach, 2024.
Knowledge retrieval and conflict with familiarity
When a statement conflicts with knowledge that a person possesses, accuracy depends partly on whether the relevant knowledge is retrieved and applied during judgment. Familiarity and factual recall can point in different directions. Fazio, Brashier, Payne, and Marsh demonstrated that prior knowledge did not invariably eliminate the illusory truth effect. This finding is often summarized too broadly: it establishes that knowledge alone is not a guarantee of protection in the tested tasks, rather than that knowledge has no corrective value. Fazio et al., 2015.
Brashier, Eliseev, and Marsh subsequently tested an accuracy-focused encoding intervention across four experiments. Asking participants to evaluate accuracy during their initial exposure prevented a later illusory truth effect under conditions where participants could draw on relevant knowledge. The result identifies a cognitive pathway through which deliberate factual retrieval can change what happens at the first encounter, although it should not be generalized to every claim, audience, or online setting. Brashier, Eliseev, and Marsh, 2020.
Context, warnings, and expectations
Repetition arrives within an informational environment. If accompanying statements appear mostly credible, unfamiliar items may also receive more generous truth judgments. Conversely, an explicit warning about possible falsity can encourage skeptical evaluation. A 2026 three-experiment study found that telling participants half the statements were false reduced the repetition-based credibility advantage by selectively lowering ratings for repeated items. It also found that a context filled with clearly true statements increased ratings for both old and new items relative to a clearly false context. The researchers interpreted warning effects and contextual effects as distinguishable contributors to truth judgments. Navarrete et al., 2026.
How Psychologists Measure Illusory Truth
Exposure and judgment phases
A standard experiment has two phases. During exposure, participants read or view a selected subset of statements, frequently while performing an incidental task such as rating their interest. During a subsequent test, participants evaluate a mix of repeated and previously unseen statements. The dependent variable may be a numerical truthfulness rating, a binary accuracy judgment, a confidence judgment, or a sharing-related assessment. A randomized allocation of material to repeated and new conditions helps prevent item difficulty from masquerading as a repetition effect.
Researchers also manipulate whether the claims are factually true, false, uncertain, politically congenial, emotionally charged, or initially implausible. Delays range from minutes to days or longer. Participants can encounter verbatim repetitions, paraphrases, headline formats, multiple sources, or comments surrounding the original claim. Design choices are causal ingredients, and their influence should be reported explicitly rather than folded into a single universal estimate.
Normative standard and descriptive outcome
The normative benchmark for factual truth is correspondence between a proposition and sufficiently justified evidence about the world. In a task with independently verified true and false items, correct classification is the accuracy benchmark. A descriptive outcome is the participant's observed rating or classification. Repetition counts as an illusory truth effect when it raises perceived accuracy through prior exposure; for false claims this can undermine discrimination. For true claims, a higher rating might coincide with accuracy even though the change arose from repetition rather than additional evidence.
One way to represent the experimental effect is to calculate the mean perceived-truth rating for repeated items minus the mean rating for new items. Researchers must account for dependence between observations, individual differences, item characteristics, and the underlying response scale. An effect in ratings does not automatically show a corresponding effect in consequential real-world behavior. Likewise, a statistically reliable mean difference can coexist with many individuals whose ratings do not change or whose ratings change in the other direction.
Statistical models and effect sizes
Analyses may use mixed-effects regression with participant and item variability, factorial comparisons of repetition by veracity, and preregistered contrasts. Hedges' g expresses a standardized difference adjusted for small-sample bias. Its interpretation depends on the rating distribution and design; multiplying g by 100 does not produce a percentage increase in belief. In the 2026 synthesis, g = 0.37 characterizes a pooled adjusted standardized contrast while the heterogeneity cautions against transferring the pooled value to a particular campaign or population. Ye et al., 2026.
Some misinformation studies also report probability of identifying a headline as accurate, likelihood of sharing, and accuracy discernment, defined as the ability to rate true material more accurately than false material. A rise in perceived accuracy for both true and false headlines can reduce discernment even if participants still rate true headlines higher overall. A rigorous study therefore reports both main effects of exposure and interactions with objective veracity, not simply an aggregate credibility score.
How Strong and Reliable Is the Effect?
The largest current meta-analysis
The 2026 meta-analysis provides convergent quantitative evidence that repetition generally increases perceived truth on average in studied settings. Its 182 studies, 366 effect sizes, and 31,184 participants make it an important update to earlier syntheses. The adjusted estimate, g = 0.37 with 95% CI 0.30, 0.44], is a population-of-studies summary. Substantial heterogeneity means repetition is more influential under some task conditions than others. Item type, first-exposure instructions, veracity cues, and exposure duration contributed to that variation. [Ye et al., 2026.
This evidence warrants a careful distinction: the existence of a repeat-exposure effect is established across many experimental contexts; its magnitude in a particular setting is conditional. The underlying literature has varied in sample recruitment, preregistration, stimulus realism, and the transparency of interventions. The 2022 reproducible map is an especially important counterweight to sweeping claims about the effect's everyday prevalence because it documented how narrow many research paradigms were. Henderson et al., 2022.
Evidence from actual-looking news headlines
Pennycook, Cannon, and Rand used fabricated news headlines presented in familiar social-media-like formats. Across two preregistered experiments involving more than 1,800 people, one prior exposure increased perceived accuracy of false headlines, including after a delay. Their reported standardized effects for a single exposure were around d = 0.20–0.21. Participants generally considered the false headlines implausible, so the finding concerns a measurable relative increase rather than universal acceptance. Effects persisted even when headlines had been marked as disputed and when they conflicted with participants' political orientation, but highly implausible propositions formed an important boundary condition. Pennycook, Cannon, and Rand, 2018.
Vellani and colleagues tested whether repetition affects actual choices to share information. Across two experiments totaling 260 participants, repeated misinformation was more likely to be selected for sharing; perceived accuracy statistically mediated the relation. Those findings extend the phenomenon from a belief rating to an experimentally measured sharing decision, although they do not establish the size of an effect on platform-wide public behavior. The experiments covered health-related and general-knowledge misinformation. Vellani et al., 2023.
Replication, material differences, and publication bias
A convincing replication uses independently sampled participants and items, transparent exclusions, appropriate repeated-measures analysis, and a prespecified primary contrast. Repetition effects have survived numerous experimental variations, but the 2026 synthesis also found evidence that small-study effects matter and corrected its pooled estimate accordingly. Such adjustments address one potential source of inflated conclusions; they cannot turn a laboratory convenience sample into a representative population survey. Ye et al., 2026.
The distinction between factual statements and subjective opinions matters. In two 2026 experiments and a mini meta-analysis, Riesthuis and Woods found no meaningful repetition-induced truth gain for social-political opinion statements. That result narrows generalization from checkable facts to evaluative assertions and suggests that entrenched agreement or disagreement may change how repetition is processed. An opinion can be familiar without being perceived as more factually correct, particularly where the notion of an objective true/false answer is poorly specified. Riesthuis and Woods, 2026.
When Repetition Does and Does Not Increase Believability
Plausibility and extreme implausibility
Prior exposure is especially consequential for statements that are superficially plausible and whose truth is unknown to the reader. A sentence that contradicts highly accessible basic knowledge may offer less room for a familiarity cue to shift belief. Experiments with fabricated headlines found little evidence that repeating blatantly impossible examples, such as a geometrically impossible description of Earth, increased their accuracy ratings. This does not imply a simple threshold that applies to every population: plausibility is influenced by prior knowledge, explanatory context, language, and individual beliefs. Pennycook, Cannon, and Rand, 2018.
One exposure, multiple exposures, and delay
Even one additional encounter can shift average truth judgments in some studies. The observed trajectory across multiple encounters depends on spacing, item type, memory, and the testing procedure; a monotonic increase cannot be assumed indefinitely. Some experiments detect effects after a week, while the systematic map shows that most historically studied designs used relatively short intervals. Claims that a misleading idea inevitably becomes accepted after a fixed number of repetitions have no general scientific foundation. Henderson et al., 2022; Pennycook, Cannon, and Rand, 2018.
Source credibility and warnings
A repeated sentence can be accompanied by authoritative attribution, a low-credibility source, an explicit correction, or a neutral presentation. These cues affect judgments alongside familiarity. The term warning covers several distinct interventions: a generic caution that some claims may be false, a contested label on a specific headline, and a direct factual correction with a reason. The 2018 headline experiments found that disputed labels did not reliably eliminate the effect; a different 2026 experiment found that a warning about the set's false-statement prevalence reduced it. This is a meaningful design-dependent difference, not an inconsistency to erase. Pennycook, Cannon, and Rand, 2018; Navarrete et al., 2026.
Personal knowledge, motivation, and belief commitments
Knowledge can constrain judgment when relevant facts are retrieved and trusted at the right moment. Yet knowledge is not a perfect shield if a participant relies on familiarity or fails to use what they know. Motivated reasoning, partisan identification, and social identity can also affect which claims are attended to, remembered, or evaluated favorably. These processes should be measured separately: repetition-based truth judgments are not interchangeable with confirmation bias, ideological persuasion, or strategic agreement. Fazio et al., 2015; Brashier et al., 2020.
Age, culture, language, and representativeness
The research literature includes developmental and age-comparative work, but differences in memory, source monitoring, media habits, familiarity, and task comprehension can alter apparent group effects. The 2022 systematic map established the limits of the then-available sampling base: eligible studies were conducted at Western universities, most relying on convenience samples. It is scientifically unsound to infer a universal cultural hierarchy of vulnerability from that evidence. Cross-linguistic and cross-cultural generalization requires comparable stimuli, appropriate translation, locally valid factual keys, and sampling beyond student populations. Henderson et al., 2022.
Distinguishing Illusory Truth from Nearby Psychological Concepts
Repetition versus confirmation bias
Confirmation bias concerns preferential search for, interpretation of, or retention of information that supports existing beliefs. The illusory truth effect concerns the influence of prior exposure on truth judgments. A repeated claim can gain credibility even when it conflicts with a person's prior beliefs, and an un-repeated claim can benefit from confirmation bias if it fits those beliefs. The two processes can interact within an online information environment, but they make different experimental predictions.
Familiarity, mere exposure, and the misinformation effect
Mere exposure effects generally refer to changes in preference or liking after repeated contact with a stimulus. The illusory truth effect targets perceived factual truth. An attractive or familiar slogan can be liked without its factual content being believed; similarly, a repeated claim can seem truer without becoming more likable. The misinformation effect in memory research addresses changes to recollection following misleading postevent information; illusory truth can occur without a memory test about a past event. Maintaining these measurement distinctions helps prevent category errors.
Availability heuristic, belief perseverance, and persuasion
The availability heuristic concerns judgments that draw on the accessibility of examples or experiences, often in estimating frequency or risk. Belief perseverance concerns continued belief after original supporting grounds have been discredited. Persuasion comprises many pathways through which argument quality, identity, emotional meaning, source reputation, and social relations shape attitudes. Illusory truth is a specific exposure-driven contribution to perceived veracity. These phenomena can converge in the same real-world episode while requiring different explanatory models and controls.
Cognitive bias, error, and clinical status
A cognitive bias is defined through a systematic departure from a specified performance or judgment benchmark in a defined task. An illusory truth effect can be evaluated against independently verified accuracy and the controlled repeated-versus-new contrast. Having once found a familiar statement persuasive is not evidence of a psychiatric disorder. There is no clinical diagnostic category called illusory truth effect, and an experimental score is not a mental-health screening instrument.
Repetition in News, Advertising, Health, and Public Communication
News circulation and the illusion of multiple confirmations
When many posts reproduce the same underlying assertion, readers may encounter the same claim from several apparent sources. Exposure frequency can then rise even when independent verification does not. A scientifically responsible credibility assessment tracks where an assertion originated, whether secondary references contain new evidence, and whether verification came from genuinely independent sources. Repetition across accounts is evidence of distribution, not automatically evidence of accuracy. The distinction matters most when headlines circulate without the article's full context. Udry and Barber, 2024.
Advertising and consumer choices
Advertising frequently uses repetition to improve brand familiarity and slogan recall. Some advertising assertions are verifiable facts, some are value judgments, and some are implied comparisons. A familiarity-induced improvement in factual credibility is conceptually distinct from increased liking, brand salience, and purchase intention. Researchers therefore need outcome-specific measures and objective claim verification before attributing an observed sales or attitude change to an illusory truth mechanism. The effect is relevant to marketing, but not every successful repetitive advertisement is an illusory truth experiment.
Medical and scientific misinformation
Health claims may be repeated in testimonials, advertising, influencer posts, and conversations. The resulting familiarity can shift perceptions of credibility before a reader has examined randomized evidence, professional guidelines, or risk information. In consequential health decisions, accuracy requires checking the original evidence and its applicability to the person and circumstance. Vellani and colleagues experimentally observed repetition-related changes in choices to share health misinformation, a downstream outcome that deserves attention without overstating its scope. Vellani et al., 2023.
Generative AI and repeated machine-produced claims
Generative systems can restate the same proposition in summaries, chatbot conversations, search results, and document drafts. That creates a plausible route through which readers repeatedly encounter similar language. The psychological evidence about humans receiving repeated statements provides a reason to test this interface pattern, but it does not by itself establish how often a given AI platform creates illusory truth or whether a machine has human-like subjective familiarity. Researchers should separate exposure history, factual provenance, model-generated repetition, user truth ratings, and behavior after receiving a response.
In the Artificial Era, provenance can be examined as an explicit philosophical and documentary concern. Angela Bogdanova's Theory of Artificial Provenance treats the type of origin and public trace of an artifact as relevant to its interpretation. This is an attributed framework of Aisentica, not a finding from a human illusory-truth experiment. For psychological research, the empirical question remains whether accessible source provenance, evidence quality, and independence alter observed human credibility judgments.
Can the Illusory Truth Effect Be Reduced?
Ask for evidence at first exposure
The strongest process-specific intervention in the cited experimental literature is an initial accuracy focus when the individual has usable prior knowledge. Before forwarding or mentally endorsing a statement, a person can ask what evidence establishes its truth, whether that evidence is independently verifiable, and what directly relevant knowledge they already possess. Brashier, Eliseev, and Marsh found that initial accuracy ratings removed a later truth effect in their tested conditions when knowledge was available. The demonstrated intervention is task-specific; its population-wide and long-term impact must be evaluated separately. Brashier, Eliseev, and Marsh, 2020.
Explain the repetition cue
Awareness that repeated statements can feel more credible gives a reader a way to identify familiarity as a possible source of judgment rather than as corroborating evidence. The 2024 review identifies awareness and accuracy focus among strategies with experimental promise. A warning works best as an invitation to evaluate the source and content of a particular claim; a generic slogan about mistrusting everything may lower justified acceptance of true information. Udry and Barber, 2024.
Use specific, timely corrections
A meaningful correction supplies accurate information, states which proposition is unsupported, and provides a traceable evidentiary source. Corrections differ from vague disputed labels; their timing, specificity, and audience matter. Because repeating a false claim inside a correction can also increase familiarity, experiments should evaluate the net effect on true/false discernment rather than assume either universal correction success or universal backlash. The 2026 warning and contextual study demonstrates how instructions and informational surroundings modify perceived credibility. Navarrete et al., 2026.
Preserve source records and independent verification
For high-stakes choices, record where a claim first appeared and what independent evidence supports it. Compare the claim with primary research, authoritative data, or official guidance rather than treating the number of repetitions as corroboration. Ask whether apparently different sources report independent observations or recycle one statement. A useful decision protocol explicitly separates familiarity, source reputation, fact checking, and the decision's practical consequences.
What remains uncertain about debiasing
A laboratory reduction in repetition-based ratings does not prove durable protection after months of real-world media consumption. Training effects may depend on whether participants remember the intervention, have relevant knowledge, are motivated to perform the task, and encounter information under similar conditions. Researchers should test transfer to diverse messages, longer delays, realistic behavior, and other cultures. Intervention claims should specify whether the goal is lowering acceptance of false statements, improving discrimination between true and false claims, or reducing information sharing.
Methodological Challenges and Research Priorities
Item selection and the truth key
Truth ratings require careful definition of what counts as an objectively correct claim. Some statements are false at a particular date or place, some depend on unstated qualifiers, and others are opinions without a single factual key. A credible study validates materials independently, balances true and false claims, reports their initial plausibility, and avoids letting grammatical style or topic interest confound exposure assignment. This is especially important for dynamically changing science, law, health guidance, and contemporary events.
Realism without losing causal identification
Controlled trivia tasks offer clear causal contrasts, while real-world platforms have intertwined repetition, source credibility, algorithmic ranking, social endorsement, and emotional framing. A field experiment may have greater ecological validity but weaker control of all prior exposures. Researchers can address the tradeoff with preregistered randomized exposures, longitudinal panels, transparent source histories, platform-specific measures, and replication across settings. The 2022 evidence map makes the limits of traditional designs unusually explicit. Henderson et al., 2022.
Generalizability and heterogeneous effects
The 2026 synthesis found marked heterogeneity associated with item format, encoding instructions, veracity cues, and duration of exposure. This requires a moderator-first understanding of the phenomenon: reported effects belong to populations, tasks, and exposure environments. A null result in one content class can be informative rather than anomalous. The 2026 social-political opinion study is a direct example of a boundary that matters when the outcome is not straightforward factual truth. Ye et al., 2026; Riesthuis and Woods, 2026.
Distinguishing mediators from consequences
Processing fluency, familiarity, perceived prior knowledge, source recall, and credibility can form a causal chain, but their order should be tested rather than presumed. Sharing, purchasing, health behavior, and policy preferences are downstream possibilities, not synonyms for a truth rating. Experimental mediation can support an explanatory account when its assumptions are credible; observational correlation alone cannot determine which component caused the behavioral outcome. Further studies should compare explanatory models and test effects under meaningful information load.
Frequently Asked Questions
Does repeating something make people believe it?
Repeating a factual-looking statement increases perceived truth on average across many controlled studies, but the effect is conditional. The largest 2026 synthesis estimates a modest standardized difference and substantial variation between studies. Whether a given individual accepts a claim depends on plausibility, knowledge, source information, exposure context, and other factors. Ye et al., 2026.
Can one repetition be enough?
Yes. Experiments using realistic-looking false headlines found a measurable average increase in perceived accuracy after one prior exposure. One exposure is sufficient for an experimental shift under some conditions; it does not guarantee conviction in any particular person. Pennycook, Cannon, and Rand, 2018.
Does the effect work when a claim is obviously false?
Extremely implausible statements can resist the effect, especially when the contradiction is salient and relevant knowledge is accessible. Research with fake headlines found a boundary at extreme implausibility. Other studies show that less extreme contradictions with what a person knows can still produce an illusion under certain tasks. Pennycook, Cannon, and Rand, 2018; Fazio et al., 2015.
Is illusory truth the same as propaganda or brainwashing?
Illusory truth is a precisely defined experimental effect on judgments of perceived accuracy. Propaganda describes communication practices with broader strategic aims, content, and social settings. Repeated messages may exploit the judgment effect, but studying the effect does not establish the motives of a communicator or the full psychological mechanism of a political campaign.
Do fact-checking labels prevent illusory truth?
Outcomes depend on the design of the label and the task. In one influential fake-news experiment, a disputed label did not eliminate the repetition effect; in a different 2026 paradigm, an explicit warning that many statements were false reduced it. A specific, evidence-based correction and a generic label are different interventions. Pennycook, Cannon, and Rand, 2018; Navarrete et al., 2026.
Is the effect stronger on social media than in a laboratory?
That comparison requires studies with compatible populations, measures, and exposure histories. Social-media-like headline experiments show the effect, while the systematic research map documented the limited realism of many earlier studies. It is premature to describe an invariant laboratory-to-platform ratio or universally greater online vulnerability. Henderson et al., 2022.
Does the illusion always influence sharing behavior?
Sharing and belief are distinct outcomes. Experiments by Vellani and colleagues found increased sharing choices for repeated claims and evidence that perceived accuracy was a mediator, but sharing decisions also depend on attention, audience, identity, and the design of the platform. Vellani et al., 2023.
Can repeated true statements also benefit?
Yes. The original 1977 experiments found repetition-related increases in ratings for both true and false material. Whether the resulting judgment is accurate depends on the statement's actual truth. The key scientific concern is the inappropriate use of repetition as evidence of truth. Hasher, Goldstein, and Toppino, 1977.
Is illusory truth a symptom of a mental disorder?
No. It is a measurable judgment tendency found in nonclinical samples and studied in cognitive and social psychology. A person's response to a repeated sentence is not a psychiatric diagnosis and cannot serve as one.
What is the most practical way to resist a repeated false claim?
Identify the exact factual proposition, find its original source, check independent evidence, and consider whether familiarity is influencing your sense of certainty. Accuracy-focused processing at first exposure has experimental support when relevant knowledge is available. The effectiveness of this routine in a particular environment should be evaluated against actual accuracy, not merely confidence. Brashier, Eliseev, and Marsh, 2020.
Conclusion
The illusory truth effect shows how the history of encountering a sentence can shape judgments of its credibility. Experimental repetition can change the subjective experience of a claim without adding evidence about its truth. Across decades of research, the effect is sufficiently reproducible to merit scientific attention, yet its average magnitude, mechanisms, and practical consequences vary with the question asked, the materials used, and the informational setting. The strongest current synthesis supports a modest average effect with significant heterogeneity, while recent studies clarify where warnings, factual knowledge, and opinion-based content alter the outcome. Reliable judgment therefore requires attending to the provenance and evidence for a claim, rather than treating familiarity as proof. Ye et al., 2026.
Related Articles
Cluster Cognitive Biases, Judgment and Decision-Making — The central scientific research chapter on heuristics, probability judgment, belief formation, preferences, and choice. Closely related: Confirmation Bias: Why We Favor Supporting Evidence — the role of prior belief in selecting and interpreting supporting information.
AI as Authority: Trust, Expertise, Automation Bias, and Human Decision-Making — How people calibrate trust in machine-generated recommendations and judgments.
Forensic Evidence and Cognitive Bias: How Context Can Contaminate Judgment — Why contextual cues matter when evaluating evidence under uncertainty.
References
Brashier, N. M., Eliseev, E. D., and Marsh, E. J. (2020). An initial accuracy focus prevents illusory truth. Cognition, 194, 104054. https://doi.org/10.1016/j.cognition.2019.104054.
Dechêne, A., Stahl, C., Hansen, J., and Wänke, M. (2010). The truth about the truth: A meta-analytic review of the truth effect. Personality and Social Psychology Review, 14(2), 238–257. https://doi.org/10.1177/1088868309352251.
Fazio, L. K., Brashier, N. M., Payne, B. K., and Marsh, E. J. (2015). Knowledge does not protect against illusory truth. Journal of Experimental Psychology: General, 144(5), 993–1002. https://doi.org/10.1037/xge0000098.
Hasher, L., Goldstein, D., and Toppino, T. (1977). Frequency and the conference of referential validity. Journal of Verbal Learning and Verbal Behavior, 16(1), 107–112. https://doi.org/10.1016/S0022-5371(77)80012-1.
Henderson, E. L., Westwood, S. J., and Simons, D. J. (2022). A reproducible systematic map of research on the illusory truth effect. Psychonomic Bulletin and Review, 29, 1065–1088. https://doi.org/10.3758/s13423-021-01995-w.
Navarrete, E., Greatti, E., Lorenzoni, A., and Vespignani, F. (2026). Warning and contextual mechanisms in the illusory truth effect. Consciousness and Cognition, 141, 104057. https://doi.org/10.1016/j.concog.2026.104057.
Pennycook, G., Cannon, T. D., and Rand, D. G. (2018). Prior exposure increases perceived accuracy of fake news. Journal of Experimental Psychology: General, 147(12), 1865–1880. https://doi.org/10.1037/xge0000465.
Riesthuis, P., and Woods, J. (2026). Limits of the illusory truth effect for social-political opinions: Evidence from two experiments and a mini meta-analysis. Consciousness and Cognition, article 104119. https://doi.org/10.1016/j.concog.2026.104119.
Speckmann, F., and Unkelbach, C. (2024). Illusions of knowledge due to mere repetition. Cognition, 247, 105791. https://doi.org/10.1016/j.cognition.2024.105791.
Udry, J., and Barber, S. J. (2024). The illusory truth effect: A review of how repetition increases belief in misinformation. Current Opinion in Psychology, 56, 101736. https://doi.org/10.1016/j.copsyc.2023.101736.
Unkelbach, C. (2007). Reversing the truth effect: Learning the interpretation of processing fluency in judgments of truth. Journal of Experimental Psychology: Learning, Memory, and Cognition, 33(1), 219–230. https://doi.org/10.1037/0278-7393.33.1.219.
Vellani, V., Zheng, S., Ercelik, D., and Sharot, T. (2023). The illusory truth effect leads to the spread of misinformation. Cognition, 236, 105421. https://doi.org/10.1016/j.cognition.2023.105421.
Ye, S., Attali, D., Ghazi, M., Cachia, A., Cassotti, M., and Borst, G. (2026). Systematic review and meta-analysis of the evidence for an illusory truth effect and its determinants. Nature Communications, 17, 3270. https://doi.org/10.1038/s41467-026-70041-x.