Misleading information can be unknowingly incorporated into an original memory after the fact

What it is

After recalling an experience, exposure to related but inaccurate information — someone else's retelling, a leading question, a report seen afterward — can be folded into the original memory without the person noticing. Afterward they will present the blended content as something they "actually saw." This is the misinformation effect. It differs from ordinary forgetting: subjective confidence in the account is usually high, and the person has no sense that the memory has been tampered with.

Why it happens

Misinformation gets incorporated because of a failure of source monitoring — at recall, the brain must judge whether a given detail came from the actual experience or from something heard, read, or asked about afterward, and the cues this judgment relies on are themselves unreliable, especially when the outside information is close to the original experience in time and content. The two sources end up merged into a single memory trace that can no longer be reliably pulled apart. Question wording is the most common vehicle for misinformation: a leading question that presupposes an object or action that was never actually there prompts the respondent to search their memory for that presupposed detail as if it needed confirming — and that search itself is what produces the integration.

Studying it

The classic paradigm has participants witness an event (typically a short film), then answers questions about details phrased with different presuppositions; the reported details and confidence levels are compared across wording conditions. Follow-up replications further show the effect persists even when participants are explicitly warned that a question may contain misleading information — showing this isn't simple deception but a rewriting at the level of the memory representation itself. Studying this effect requires being able to independently verify the actual event details as a baseline; without that, there's no way to distinguish a misinformation effect from simply never having encoded the detail.

Where it stops holding

Effect size scales with how salient the misled detail is, the interview interval, and how many times the misleading question is repeated — details that are more central or attention-getting are harder to replace. A single mild wording difference typically produces less distortion than repeated suggestion or an explicit false statement from an authority figure. Lab events are typically brief and presented once, which doesn't map cleanly onto the layered impressions built up from repeated real-world product use, so lab effect sizes shouldn't be carried over directly to interface-use settings.

Applying it

• When asking "what did you do at that point" in a usability test or interview, avoid presuppositional wording — don't ask "did you click confirm first," ask "do you remember where you clicked." A presupposed detail gets treated by the respondent as a candidate to search their memory for, and they may "confirm" an action that never happened.
• Don't let a respondent hear another participant's account or the "correct" way to do the task between interview rounds; information picked up during that interval can easily reappear in the next round told as the respondent's own firsthand experience.
• How to check: compare the same respondent's description of a detail before and after switching to neutral wording — a noticeably different answer after the switch is a sign the original wording had already introduced a bias.

Related

• Same group: A6.20.1 Memory is rebuilt at every retrieval, not replayed from a fixed archive · A6.20.2 A retrieved memory absorbs new information, shifting later recall · A6.20.4 A user's account of what they previously did cannot serve as an objective behavioral record
• Nearby: A6.13 Encoding in long-term memory · A7.11 Expectation violation and the explanation gap
• Search terms: misinformation effect · source monitoring · leading question