Misinformation Effect.

01Overview

The misinformation effect is the finding that information encountered after an event can alter — and in many cases replace — a person's memory of the original event. Memory is not a recording; it is a reconstruction. And reconstructions can be contaminated by what happens after the fact.

For designers, this is a critical finding for user research methodology. The way you ask questions in an interview, the order in which you show designs, and the framing you use to introduce a test can all contaminate participants' memories of their actual experience — and produce data that reflects your framing, not their truth.

02Detailed explanation

Elizabeth Loftus's foundational 1974 research showed that asking participants "How fast were the cars going when they smashed into each other?" produced higher speed estimates than "How fast were the cars going when they hit each other?" — for the same footage. A week later, participants who heard "smashed" were also more likely to report seeing broken glass in footage that contained none.

• The effect is strongest when: the original memory is weak or ambiguous; the post-event information is plausible; the source of the post-event information is authoritative; and there is a time delay between the event and the recall.
• Leading questions are the primary vehicle for the misinformation effect in user research: "Did you find the checkout confusing?" is not a neutral question — it suggests that confusion was the expected response and biases recall toward that interpretation.
• Showing a competitor's design before asking about your own can contaminate memory of your design with the comparative frame — users may recall your product as more or less confusing than they actually found it, relative to the comparison.

03Why it exists

Memory is not stored like a file — it is reconstructed each time it is accessed, using available information to fill gaps. Post-event information is "available" during reconstruction and gets incorporated into the rebuilt memory. The original trace and the new information merge in ways that are largely invisible to the person remembering.

04Effects on users

• Users who read a product's marketing copy before a usability test will recall the product experience through the lens of those claims — positively or negatively, depending on whether the experience matched.
• After a redesign, users' memories of the old design are immediately contaminated by the new one — making retrospective comparisons unreliable. Users genuinely believe they preferred or disliked the old design based on a memory that no longer accurately represents it.
• In diary studies, prompts that contain embedded assumptions ("what was frustrating about this step?") contaminate recall — even if the user didn't experience frustration at the time.

05Effects on designers & teams

• Interview design: the most dangerous research questions are not wrong questions — they are subtly leading questions. "Did you feel confident navigating?" contains an assumption about the relevance of confidence that contaminates the response.
• Usability test facilitation: a facilitator who reacts — visually or verbally — to user struggle can contaminate the user's memory of the experience. Participants may recall being more or less confused than they were, calibrated to the facilitator's evident expectations.
• Research synthesis: reading previous research reports before conducting new interviews introduces misinformation into the analyst's own recall of what participants said — prior findings contaminate fresh interpretation.
• Stakeholder walkthroughs: narrating a design during a demonstration — "this is where users can do X" — can contaminate observers' memory of what they saw with what they were told, rather than what they actually perceived.

6Introspective view

Look inward. Post-event information — a leading debrief question — silently rewrites what participants remember.

From an introspective perspective, ask how Misinformation Effect may already be shaping your research, critique, planning, and interpretation — not only what users encounter in the finished interface.

07Practical takeaways

• Use open, non-leading questions: "What did you notice?" rather than "Did you notice the confirmation message?" Open questions allow participants to reconstruct their memory without post-event contamination from the question itself.
• Separate think-aloud from retrospective recall: contemporaneous verbalisations during a task are less contaminated than post-task recall. Ask users to narrate as they go, not to summarise afterwards.
• Record observations first, questions second: note what you observed before asking participants what they remember. This prevents your questions from retroactively shaping your interpretation of the behaviour you saw.
• Be careful with comparative testing: showing multiple designs in sequence contaminates memory of each design with the comparison. Counterbalance order; use between-subjects designs for high-stakes comparisons.
• Minimise the gap between event and recall: the longer the delay between the experience and the interview, the more time for post-event information to contaminate the memory. Interview as close to the experience as possible.

08Design examples

09Ethical risks

The misinformation effect can be exploited intentionally in research design — asking leading questions to produce a desired finding, framing research to contaminate memories in favourable ways, or presenting stakeholders with research that reflects the questions asked rather than the users' actual experience.

The obligation of rigorous user research is to produce data that reflects user experience — not to validate pre-existing decisions. Questions that contaminate memory are not just methodologically weak; they are a form of research dishonesty, producing pseudo-evidence that misleads product teams and harms users.

Good research questions don't just avoid leading — they actively work to hear what users actually experienced, not what the researcher hoped they experienced. That requires vigilance, humility, and genuine commitment to the participant's truth.

10Related biases