Repetition makes a statement feel true, including one you know is false

The effect, and why it is unusually well supported

Show people a set of statements, wait, then show them a larger set containing some of the originals, and ask for truth ratings. Repeated statements are rated as more likely to be true than new ones. The pattern has been reproduced over decades, across a wide range of materials, and it survived the period in which a great many neighbouring findings did not. It is one of the sturdier results in this area.

That durability is worth stating plainly, because so much of this site is about claims that turned out to be fragile. This one has been examined by many independent groups, with pre-specified analyses, and with samples large enough to detect what they were looking for. The core finding stands.

The mechanism is processing ease, not persuasion

The usual account is that a repeated statement is easier to process the second time, and that the mind reads this ease as familiarity, and familiarity as a weak signal of truth. This is not stupid. Things you have encountered before genuinely are more likely to be true, on average, because false claims tend to be corrected out of circulation. The shortcut has a reasonable design and a predictable failure mode.

The failure mode is that ease of processing is produced by exposure regardless of the source, the context or the accuracy of the claim. A statement encountered three times in a critical article is easier to process afterwards than one never encountered, and the ease does not carry a label saying where it came from.

It is not the same as thinking a thing is common

This is worth separating from the more familiar idea that easily recalled events are judged frequent. That concerns estimates of how often something happens, and it runs on how readily examples come to mind. What is described here concerns the believed truth of a proposition, which is a different judgement with a different input.

The two can point in opposite directions. A vivid single event may be highly available while nobody believes it is typical, and a dull statement repeated in the background may acquire credibility without any accompanying image at all. Keeping the distinction makes it easier to see which one is operating in a given case.

Knowing better provides less protection than expected

The uncomfortable part of this literature is that prior knowledge does not fully shield people. Studies have found the effect for statements that participants could correctly identify as false when asked directly, meaning the relevant knowledge was present and simply not consulted during the truth judgement. Knowledge has to be retrieved to be used, and a fluent statement does not prompt retrieval.

The boundaries of this are still being worked out, and the honest summary is that the effect shrinks with implausibility rather than disappearing. Highly implausible claims resist repetition better than plausible ones; where exactly the resistance becomes reliable is a live question, and studies disagree on how far it extends.

Why the current information environment is the worst case

Nothing about this requires anybody to intend it. A claim circulating in a network is repeated by supporters, by critics quoting it, by summaries of the argument and by people describing the controversy. All of that is exposure. Correction formats that restate the false claim in order to deny it are performing the same repetition as the original, which is why the design of corrections has become a research question in its own right.

The practical implication is unwelcome: the volume of exposure to a claim is close to independent of its truth, and volume is one of the inputs to how true it feels. Attention is not a filter.

What actually helps, which is not much

Warning people that some statements are false, before they see them, reduces the effect somewhat in experimental settings. So does anything that pushes the judgement from feeling towards checking — asking on what basis you believe something, or where you first encountered it, or whether you could name a source. The gains are real and partial.

The stronger defence is structural rather than psychological. Notice that a sense of familiarity is not evidence, treat “I have heard this before” as information about your own exposure history rather than about the world, and be especially wary when a claim feels obvious and you cannot say where it came from. That combination — high confidence, no traceable source — is the characteristic signature of something that was simply repeated.