The Misinformation Effect in Eyewitness Memory
Author: Ukrainian Psychological Hub · Published: October 11, 2026 · Editorial Policy
The misinformation effect in eyewitness memory is the influence of misleading information encountered after an event on what a person later reports remembering about that event. A witness may see one traffic sign, hear a different sign mentioned in a question, and subsequently report the introduced sign as though it appeared in the original scene. In the justice system, this phenomenon matters because testimony can develop through interviews, conversations, identification procedures, exposure to other accounts, and later review of evidence. The sequence through which information was acquired is therefore an essential part of evaluating an eyewitness report.
Experimental research provides compelling evidence that inaccurate post-event suggestions can reduce accuracy and introduce erroneous details. The classic Loftus, Miller, and Burns experiments established a controlled laboratory paradigm, and a 2026 contemporary replication confirmed a measurable effect under newer testing conditions. The strength of this evidence concerns effects on later memory reports across experimental groups. Whether an individual witness's answer reflects source confusion, genuine change in recollection, a response strategy, or several processes together requires additional analysis.
The forensic implication is practical: statements have histories, and corroboration has to be independent. A detail repeated consistently after suggestive questioning is not equivalent to a detail recorded before the suggestion. The National Research Council's scientific consensus report places contamination risk within the wider science of eyewitness perception, memory, identification and investigative procedure.
Definition: What Is the Misinformation Effect?
The misinformation effect refers to the way misleading information received after an event can influence subsequent reporting of that event. The defining temporal sequence is original exposure, post-event misinformation, and a later memory test or report. This is an experimentally studied psychological phenomenon, not a clinical disorder, a diagnosis, an indicator of dishonesty, or a legal finding. An eyewitness can remain sincere while giving an inaccurate answer.
In the typical experiment, participants first view a video, slideshow or staged incident. Researchers then present information that is consistent with the scene, neutral, or misleading. Finally, participants are tested about details for which the original scene is known. Randomly assigned conditions let researchers compare error rates and estimate how introducing misinformation affected responses. The 2024 systematic review by Brassil, O'Mahony, and Greene describes this core design and its distinction from other false-memory paradigms.
For example, a witness sees someone carrying a dark object but cannot identify it. A later investigator asks about the 'black phone,' presenting a plausible but unverified description. If the witness repeats 'black phone' in a subsequent account, several interpretations remain possible: the description may have been observed but initially omitted, borrowed from the question, inferred from contextual expectations, or strategically selected as a best guess. Documentation must establish what was said and when before attributing a mechanism.
Misinformation need not be intentionally deceptive. Ordinary conversations, inaccurate questions, mistaken captions, suggestive images, police feedback, and repeated online stories can introduce details after an event. A witness who is trying to cooperate may treat an investigator's wording as informative. The effect is therefore relevant to the preservation of evidence even when everyone involved is acting in good faith.
Landmark Experiments: What Did Loftus Actually Show?
Loftus and Palmer (1974): The Words Used in a Question Matter
In Reconstruction of Automobile Destruction, Elizabeth Loftus and John Palmer showed participants traffic-accident films. Different groups heard different verbs in a speed-estimation question, including versions implying more or less forceful impact. The question using 'smashed' produced higher average speed estimates than questions with milder verbs. In a later test, that group was more likely to report broken glass, even though the film did not contain any.
The study showed that post-event language can affect later answers under controlled conditions. It remains a foundational demonstration of suggestion, but its numerical results should not be repurposed as a universal error percentage for actual witnesses. Laboratory participants saw specified films, answered particular questions, and completed tests selected by the investigators. Generalization requires attention to the event, wording, delay, and response task.
Loftus, Miller, and Burns (1978): A Visual Detail Changed After Verbal Misinformation
The researchers' later Semantic Integration of Verbal Information Into a Visual Memory involved 1,242 participants across five experiments and a pilot study. Participants saw slides of an accident and subsequently received accurate, irrelevant, or misleading descriptions. The misled groups showed poorer performance on later recognition tasks. This work helped define the three-stage experimental paradigm still used to study eyewitness misinformation.
These studies are informative because researchers know the originally presented detail. In legal cases, the ground truth may be incomplete or contested. A laboratory effect identifies a causal impact of misinformation on reporting within its design; a real investigation additionally needs the earliest account, the precise later exposure, and independent records of the disputed event.
Wiechert et al. (2026): Replication and the Debate About Mechanisms
A contemporary replication and extension used an accident video and later co-witness statements about a traffic sign. Misled participants gave the correct answer less often. When the researchers altered the response format and introduced confidence-based betting to reduce some strategic response tendencies, the misinformation effect remained but became smaller. The paper appeared online in September 2025 and in a July 2026 journal issue.
This result supports the robustness of the phenomenon while challenging overly simple explanations. Some apparent errors can depend on how people interpret a test or choose an answer. Other errors persist when those influences are reduced. The primary evidence supports multiple contributing mechanisms rather than any claim that each incorrect answer proves permanent erasure or replacement of an original memory trace.
How the Misinformation Effect Works
Source Monitoring: Was It Seen or Learned Later?
Source monitoring is the process of deciding where knowledge originated: personal observation, another person's statement, inference, imagination, or a document. A witness may remember that someone wore a red jacket but fail to identify that the color first appeared in a witness discussion rather than direct perception. The source-monitoring framework of Johnson, Hashtroudi, and Lindsay provides an influential explanation of how remembered information can be attributed to the wrong source.
Familiarity, detail and vividness can encourage an impression that information came from firsthand observation. Yet a frequently repeated detail can become familiar without becoming accurate. Source monitoring depends partly on retaining contextual cues about the time, place and manner of acquisition. The 2024 systematic review identifies source monitoring as a central theoretical account, while describing several approaches to the phenomenon.
Reconstruction and Integration of Event Information
People reconstruct past events using retained sensory material, contextual knowledge, expectations and information acquired later. A suggested detail that fits the general event may become part of a coherent account. The original 1978 researchers used the language of semantic integration to describe how verbal descriptions influence later recognition of visual material. Modern evidence also supports retrieval and decision processes that can produce similar observable results.
Memory reconsolidation is sometimes used in popular accounts to explain all such errors. However, a change in a behavioral response does not by itself demonstrate a particular neural storage mechanism. Cognitive models explain performance patterns; claims about specific brain processes require their own evidence. In a forensic opinion, the level of explanation needs to match the measurements actually available.
Retrieval Competition and Test Demands
A witness may retain access to the original detail while finding the misleading version easier to retrieve. Asking a forced-choice question can encourage a response even when a person is uncertain, whereas allowing uncertainty and asking about sources may produce different results. A memory report is an observable behavior influenced by retrieval, judgment and the response options provided.
The 2026 replication is relevant here because reducing strategic responding changed the magnitude of the effect. It demonstrates why researchers distinguish memory distortion from performance changes created partly by task demands. Forensic experts should describe a reporting error as an error, while explaining the remaining uncertainty about the person's internal representation.
Social Influence and Co-Witness Memory Conformity
Two eyewitnesses may see an event from different angles, then discuss it and begin using the same description. Their later agreement can arise from both independent observation and the exchange of information. Confidence or apparent expertise in one speaker may affect what the other accepts. Memory conformity can involve genuine source confusion, social agreement, or both.
The 2023 research by O'Donnell, Chan, Foster, and Garry studied repeated misinformation and variations in its source. Their experimental and meta-analytic results supported the importance of repetition but did not establish an independent effect of source variability under the conditions examined. Thus, the number of people repeating a statement and the number of times a statement is repeated should be documented separately.
Repetition, Delay, Attention, and Other Moderators
Repeated claims can become easier to retrieve and may displace less accessible event details in a later response. Repetition is particularly important where a witness encounters the same mistaken description through interviews, discussions, reports and online coverage. O'Donnell and colleagues found that repeated misinformation can influence reporting, while showing why repetition and variation in sources must be experimentally disentangled.
Retention interval matters because witnesses may lose contextual detail over time and encounter additional information. Delay is not a diagnostic sign that misinformation occurred; forgetting can happen without misinformation. When evaluating statements taken weeks or months apart, the analyst should consider the entire exposure history, not just elapsed time.
The quality of original encoding also matters. A clearly observed central action and a briefly seen peripheral object may have different chances of being recalled accurately. Darkness, obstruction, divided attention and complex scene dynamics can limit the original detail available. These factors concern initial perception and are distinct from subsequent misinformation, though they may interact with it.
Individual differences in cognitive functioning have been investigated, but the evidence does not supply a reliable individual-level 'misinformation susceptibility score' for use as proof in a case. The 2024 systematic review examined performance-based cognitive abilities and found a varied literature. One intelligence test, age category or personality label should not be used to certify that a witness either resisted or absorbed a suggestion.
Stress and trauma can influence several stages of eyewitness memory. Their effects need to be assessed through research appropriate to the particular mechanism; stress-related encoding changes are not themselves evidence of later misinformation. Clinical diagnosis, cognitive capacity, eyewitness vulnerability, and legal credibility are distinct questions. No diagnosis establishes that testimony is true or false.
Misinformation Effect Versus Other Memory Errors
The misinformation effect is a specific post-event influence. An initial perceptual error occurs when a witness misidentifies something during the original event. Ordinary forgetting occurs when information becomes unavailable or difficult to retrieve, with or without misleading input. An inference error occurs when a witness fills in an uncertain detail using expectations. Misinformation may compound these vulnerabilities, but they should remain analytically separate.
'False memory' is a broader expression that includes errors produced in other experimental tasks and autobiographical contexts. The Brassil et al. review explicitly distinguishes the eyewitness misinformation paradigm from other false-memory methods because effects in one task do not automatically establish identical mechanisms in another. The present article concerns the impact of post-event information on eyewitness reporting, leaving comprehensive treatment of general false-memory research to its own canonical topic.
Leading questions are one possible delivery mechanism. A question can imply that an unverified object existed, as in 'Where was the knife?' rather than 'What, if anything, did the person hold?' A leading question does not automatically cause memory distortion, and a misleading statement can be delivered without a formal question. The forensic priority is to preserve the actual wording and examine subsequent changes.
Post-identification feedback is another related phenomenon. Confirmation after a lineup choice may affect later confidence or memories of the viewing conditions. This issue belongs to the specialized science of identification procedures and initial confidence measurement. The American Psychology-Law Society recommendations explain safeguards intended to minimize these system-level influences.
Memory conformity describes the convergence of accounts following social exchange. It overlaps with misinformation research when co-witness statements contain inaccurate details. A shared narrative may be valuable corroboration only if its independence can be established. Similar stories after extensive discussion do not carry the same evidentiary meaning as independently recorded first reports.
Can a Witness Be Highly Confident and Still Be Wrong?
Yes. Subjective confidence measures the witness's certainty, whereas accuracy concerns correspondence with the event. A person can confidently recall information originally introduced after the event. Conversely, uncertainty does not make an account automatically inaccurate. The question is what information was available to the witness when confidence was expressed and whether feedback or rehearsal occurred.
There is a crucial nuance in identification research. The Wixted and Wells 2017 scientific synthesis found that an initial confidence statement collected under properly controlled conditions may contain useful information about identification accuracy. This does not make later courtroom confidence a direct gauge of every narrative detail, and confidence–accuracy relationships vary with the procedure and task.
Therefore, later certainty should be interpreted with the earliest statement and documentation of subsequent influences. An increase in confidence after police feedback, repeated questioning, social discussion or evidence review deserves contextual examination. It does not prove contamination by itself, nor does it independently verify the original observation.
How Strong Is the Evidence?
The basic phenomenon is well supported by decades of controlled experiments. The 1974 and 1978 landmark studies, newer replications, systematic reviews and analyses of corrective warnings supply converging evidence that exposure to inaccurate post-event details can shift memory reporting. Elizabeth Loftus's historical scientific review of the idea places the research across decades of experimental development.
The precise mechanism is more contested. Report changes can reflect source misattribution, changes in accessibility, interference, expectations, response selection or integrated reconstruction. A finding that people reported misleading details is not automatically evidence that their original memory was erased or that they experienced a vivid false recollection. Tests designed to separate explanations can produce different effect sizes.
Laboratory designs provide knowledge of the original event and control over misinformation. Real investigations involve complex events, personal stakes, unequal observation, incomplete records, media exposure and conversations that researchers cannot fully manipulate. The National Academies' eyewitness report explicitly emphasizes the limits of translating experimental results into field practice and judicial procedure.
The percentages reported in one experiment are not a calibrated probability that a particular witness is wrong. Experimental rates depend on stimulus material, test design, delay, measurement, scoring, and participants. No accepted universal equation converts a misinformation exposure into an individual error probability. A case requires its own evidence, timeline, documented statements and corroboration.
What Investigators Can Do to Protect Eyewitness Evidence
Obtain and Preserve an Early Independent Account
An investigator should seek a prompt, neutral account under appropriate circumstances, letting the witness initially describe the event in their own words. Open invitations reduce the necessity of proposing unverified facts. Precise follow-up questions can then clarify location, sequence, sensory information and uncertainty, while documenting whether each answer reflects direct observation or later information.
The first statement should be preserved in its actual wording where possible. Recording the full interaction helps later reviewers distinguish the witness's words from the interviewer's assumptions. A short investigator summary may omit the questions that shaped a response. Preservation supports transparent evaluation without assuming that the first account is perfect.
Identify and Document Post-Event Information Sources
Investigators can ask neutrally whether the witness saw media coverage, spoke with other witnesses, read messages, reviewed recordings or heard official descriptions after the event. The record should preserve dates, channels and content where possible. Documenting those exposures is more informative than treating every later change as proof of deception or every repeated detail as independent evidence.
Separating witnesses before independent initial accounts, where practical and lawful, can reduce cross-contamination. Perfect isolation is often impossible, especially after highly publicized events. The aim is to preserve the chain of information and identify what each witness knew before communicating with others.
Protect Identification Procedures Against Feedback
The National Research Council recommends training, standardized witness instructions, blind lineup administration, confidence documentation and recording. The 2020 American Psychology-Law Society review details nine policy recommendations, including a prelineup interview, evidence-based suspicion, video recording and limiting repeated identification attempts with the same suspect.
These measures address identification evidence specifically; they should not be confused with a complete protocol for all investigative interviews. Their broader lesson is that investigator behavior and procedure can change the quality of evidence. Recording initial confidence before feedback helps preserve a more informative measure than retrospective courtroom certainty.
Permit Uncertainty and Avoid Supplying Answers
Questions should allow 'I don't know,' corrections and distinctions between what was seen and what was learned later. Pressure to choose between options can turn uncertainty into a specific but weakly grounded response. Interviewers should refrain from introducing contested details as established facts and should avoid inadvertently rewarding one version.
A witness who changes an earlier statement should be invited to explain the change without accusation or coaching. The original and revised accounts, question wording, timing and exposure history should all remain available. The goal is to collect the best available account and preserve its provenance, not to engineer consistency.
Can Warnings and Corrections Reduce the Effect?
Yes, under studied conditions. Blank and Launay's meta-analysis of 25 post-warning studies found that warnings can substantially reduce the average misinformation effect, although results differ with warning design and other experimental factors. The evidence supports source-focused correction and attention to the possibility that later information was mistaken.
A warning does not guarantee exact restoration of a prior memory. Witnesses can continue to confuse sources, especially when the original detail is weak and the suggested version is familiar. Repeated efforts to challenge an answer can also change later reporting. Corrections should therefore be conducted within applicable professional procedures rather than as improvised attempts to force an original memory to return.
The research distinguishes preventing contamination, detecting an exposure after it occurred, and correcting an inaccurate belief or memory report. These are related goals with different evidence and practical constraints. A court may evaluate a revised account, but the rules for taking, disclosing and admitting that account are jurisdiction-specific.
Eyewitness Misinformation and Courtroom Decisions
Psychology addresses how post-event information affects perception, memory reporting and confidence. Law governs whether testimony is admissible, what material must be disclosed, how experts may testify, how evidence is weighed, and who makes the ultimate factual findings. These legal standards differ by jurisdiction. The existence of a scientifically supported misinformation effect is neither a verdict about one witness nor a universal rule requiring exclusion of testimony.
A properly limited expert opinion can explain relevant research, outline contamination opportunities documented in a case, describe methodological uncertainties, and identify which inferences cannot be made. Experts should not present a laboratory percentage as a personal accuracy probability. They also cannot directly inspect the witness's original memory or decide credibility as a matter reserved to the fact-finder.
Case records useful for forensic analysis include the first statement, precise interview questions, recording dates, co-witness interactions, media and evidence exposure, lineup materials, contemporaneous confidence statements and later revisions. These records support an assessment of whether information was independently witnessed or introduced afterward. They may also disclose alternate explanations for differences between statements.
The National Academies' findings and recommendations stress careful presentation and preservation of eyewitness identification evidence while acknowledging uncertainties in applied research. Their recommendations are professional scientific guidance rather than binding law in every American state or other legal system.
Worked Example: Why an Exposure Timeline Matters
In a hypothetical case, a witness initially tells police that a person carried 'a small dark object' while leaving a store. The next day, an interviewer asks whether the witness noticed a black handgun. A news report later describes a handgun. During a subsequent interview, the witness says with confidence that a black gun was visible.
This history identifies opportunities for contamination, because the more specific detail appeared in post-event materials before the later answer. It does not establish whether the original object was a gun. The witness might have seen it clearly but described it vaguely, might have inferred it from the question, or might have incorporated the later account. Video, physical evidence, early notes and other independent records would affect the evaluation.
Suppose a second witness repeats the handgun description after speaking with the first witness. Their agreement now has an uncertain degree of independence. An expert should trace who told whom, what was known at each stage, and whether independent observation can be separated from later discussion. Agreement is strongest as corroboration when independent origins are supported by evidence.
What the Misinformation Effect Does Not Prove
Misinformation does not automatically prove that the witness lied. A person may be sincerely mistaken or uncertain about where a detail originated. It also does not prove that all eyewitness evidence is unreliable: accounts vary in encoding conditions, timing, interviewing procedures and corroboration.
Nor does vividness guarantee authenticity. A highly detailed recollection may incorporate borrowed information, and an uncertain recollection may still be accurate. The most useful question is what can be established about the content and history of a report, not how cinematic or emotionally convincing it sounds.
Finally, the effect does not provide a clinical screening score, a method for predicting guilt, a legal threshold for admissibility, or a definitive diagnosis of an individual witness. Research supports a class of vulnerabilities that courts and investigators should consider alongside case-specific evidence.
Frequently Asked Questions
What is the misinformation effect in eyewitness memory in simple terms?
It occurs when information received after an event changes what an eyewitness later reports. For example, a mistaken detail in an interview question may appear in a later description even when the detail was absent from the event.
Who discovered the misinformation effect?
Elizabeth F. Loftus and colleagues developed influential experimental demonstrations in the 1970s. The 1974 Loftus and Palmer experiment and 1978 Loftus, Miller, and Burns series helped establish the modern research paradigm.
Is the misinformation effect the same as a false memory?
It is one studied pathway to inaccurate memory reporting. False memory is a broader label covering different experimental phenomena. Researchers cannot assume that all false-memory tasks have identical mechanisms or implications.
How do leading questions affect eyewitness testimony?
They can introduce presuppositions or misleading details that shape later answers. The effect depends on timing, wording, what was actually observed and how the response is measured. Asking about an object does not prove that the witness saw it.
Can eyewitnesses affect one another's memories?
Yes. Discussion and repeated claims can influence later reports. In a case, it matters whether the witnesses provided independent accounts before any communication and whether their later convergence can be explained partly by shared information.
Why can confidence be high even when an account is inaccurate?
Familiarity, repeated retrieval and post-event feedback can affect certainty. Initial confidence in well-controlled identification procedures can still be informative under defined conditions. Later certainty and accuracy remain separate questions.
Is this a clinical disorder or a diagnosis?
No. The misinformation effect is an experimental finding about memory reports. It has no standalone DSM or ICD diagnosis and does not independently identify deception, a psychiatric disorder or criminal responsibility.
Does a warning erase the effect?
Warnings can reduce the average effect under laboratory conditions, as supported by Blank and Launay. They do not guarantee that an individual memory returns to its original form.
Can science assign an accuracy percentage to a particular eyewitness?
There is no universal validated conversion from laboratory misinformation rates to the probability that one witness is accurate. A case requires attention to observation, exposure, interview history, independent evidence and the specific question being evaluated.
What should investigators do if a witness changes a statement?
Preserve both accounts, document the exact questions and the timing of changes, identify sources of later information, and investigate corroboration. Applicable interviewing, evidence-disclosure and court rules guide the legal handling of those records.
Does this research make all eyewitness testimony unreliable?
No. Eyewitness information can be significant evidence. The research shows why independent early statements, neutral procedures, records of questions, and appropriate corroboration matter to its interpretation.
For broader context, see Eyewitness Testimony: How Memory Works in Legal Cases.
Related Articles
References
Blank, H., & Launay, C. (2014). How to protect eyewitness memory against the misinformation effect: A meta-analysis of post-warning studies. Journal of Applied Research in Memory and Cognition, 3(2), 77–88. DOI
Brassil, M., O'Mahony, C., & Greene, C. M. (2024). Do cognitive abilities reduce eyewitness susceptibility to the misinformation effect? A systematic review. Psychonomic Bulletin & Review, 31, 2410–2436. DOI
Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring. Psychological Bulletin, 114(1), 3–28. DOI
Loftus, E. F. (2026). The history of an idea: The misinformation effect. Legal and Criminological Psychology, 31(2), 595–609. DOI
Loftus, E. F., & Palmer, J. C. (1974). Reconstruction of automobile destruction: An example of the interaction between language and memory. Journal of Verbal Learning and Verbal Behavior, 13(5), 585–589. DOI
Loftus, E. F., Miller, D. G., & Burns, H. J. (1978). Semantic integration of verbal information into a visual memory. Journal of Experimental Psychology: Human Learning and Memory, 4(1), 19–31. PubMed
National Research Council. (2014). Identifying the Culprit: Assessing Eyewitness Identification. National Academies Press. DOI
O'Donnell, R., Chan, J. C. K., Foster, J. L., & Garry, M. (2023). Experimental and meta-analytic evidence that source variability of misinformation does not increase eyewitness suggestibility independently of repetition of misinformation. Frontiers in Psychology, 14, 1201674. DOI
Wells, G. L., Kovera, M. B., Douglass, A. B., Brewer, N., Meissner, C. A., & Wixted, J. T. (2020). Policy and procedure recommendations for the collection and preservation of eyewitness identification evidence. Law and Human Behavior, 44(1), 3–36. DOI
Wiechert, S., Verschueren, A., Ben-Shakhar, G., Pertzov, Y., Loftus, E. F., & Verschuere, B. (2026). The misinformation effect: A contemporary replication and extension of Loftus et al. (1978) to investigate its underlying mechanisms. Journal of Experimental Psychology: Learning, Memory, and Cognition, 52(7), 1225–1241. DOI
Wixted, J. T., & Wells, G. L. (2017). The relationship between eyewitness confidence and identification accuracy: A new synthesis. Psychological Science in the Public Interest, 18(1), 10–65. DOI