Snap Shot
- The question: Can what a witness hears and sees after an event, from other witnesses, news and social media, investigators, or family, friends, and professionals, change what the witness later remembers and reports?
- What the research shows: Witnesses who discuss an event sometimes adopt each other’s details, particularly when they saw different things; in one study designed that way, 71% adopted at least one item (Gabbert et al., 2003). Friends and partners are especially influential. Discussion is common among real witnesses, news and online material can produce reports of things never seen, and confirming feedback from an investigator inflates a witness’s certainty and recollection of viewing conditions. Correct information from others can also improve accuracy, and challenging a co-witness’s error reduces its influence.
- What it does not show: Contact with other people does not by itself make an account unreliable, and most contamination research concerns peripheral details of filmed events. The research cannot show from an account alone that a particular detail came from an outside source rather than from the event; an expert analyzes the process (questions, repetition, exposure), not the truth of the account.
- The law: Due process screens only identification procedures arranged by law enforcement (Perry v. New Hampshire; Mil. R. Evid. 321); suggestion from co-witnesses, media, or family goes to weight and is tested through cross-examination, instructions, and expert testimony under Rule 702, Houser, and Arizona’s Chapple line. Experts may not opine that a particular witness is mistaken (Lindsey).
- Issues for each side: Prosecutors should separate witnesses, record first accounts, and document each witness’s exposures, without treating agreement among witnesses as independent corroboration; defense counsel should tie a contamination claim to an identified source, a specific detail, and a sequence showing exposure came first.
Between an event and a trial, a witness’s memory is exposed to a steady flow of outside information. Witnesses talk with one another, read news reports and social media posts, answer investigators’ questions, hear how an identification turned out, and discuss the event with family, friends, counselors, and advocates. Any of these contacts can add a detail, change one, or make the witness more confident in one, and the witness may not be aware that a detail came from that contact. A detail absorbed this way may be reported sincerely, as a memory error and not a lie, although conformity can also reflect social pressure rather than changed memory.
These questions reach court in almost every contested case. Several witnesses who agree may be offered as independent corroboration; a witness’s confidence may have grown between the first statement and the trial; or a disputed detail may first appear after a conversation, a news story, or an interview. Researchers have compared witness memory to physical trace evidence such as DNA, which can be contaminated and needs care in collection (Lane & Houston, 2021, pp. 11–13). The same research shows that contact with other people can also strengthen accurate memories.
The sections below take up four outside sources in turn: co-witness discussion and memory conformity, news and social media, investigators, and family, friends, therapists, and advocates. The basic laboratory paradigm is covered in The Misinformation Effect, and the process by which a witness attributes a detail to the wrong origin is covered in Source Monitoring and Misattribution of Memory.
Co-Witness Discussion and Memory Conformity
When two people who saw the same event talk about it, each may later report details that only the other saw. In an experiment in which pairs of participants watched different versions of the same filmed crime, 71% of those who discussed it before recalling it individually reported at least one item they could have learned only from their partner; adults aged 60 to 80 were no more susceptible than adults aged 18 to 30 (Gabbert et al., 2003). A coordinated replication of a related co-witness procedure found distortion after discussion in ten countries (Ito et al., 2019).
Researchers call this memory conformity. A review identified three routes by which it occurs: a witness may report another’s version to avoid disagreement (normative influence), may defer to a partner believed to have the better memory (informational influence), or may genuinely come to remember the other’s detail as his or her own (memory distortion) (Wright et al., 2009). The distinction matters because public agreement need not reflect a changed memory, whereas the third route yields a detail the witness sincerely believes. Relationships also matter: participants were more likely to adopt information from a friend or romantic partner than from a stranger (Hope et al., 2008).
Conformity reaches confidence as well as content. Witnesses told that a co-witness had identified the same person became more confident in their own identification, and those told that a co-witness had chosen someone else became less confident (Luus & Wells, 1994). In three experiments on face recognition, exposure to another person’s mistaken responses reduced the accuracy of participants’ high-confidence judgments (Sousa & Jaeger, 2022). Because high initial confidence is ordinarily one of the more informative signals of accuracy (see Confidence and Accuracy), contact with co-witnesses before a confidence statement is recorded can weaken that signal.
Discussion is common among real witnesses. In a survey of 60 witnesses at a police identification suite, 88% had co-witnesses and 58% of those had discussed the event with at least one (Skagerberg & Wright, 2008). Much of it now happens by text and group chat. In experiments in which participants watched a crime film on a smartphone and then discussed it through an instant-messaging application, they commonly incorporated false details supplied by the other witness, and instructions explaining that the partner had seen a different version reduced the effect (Cadavid & Luna, 2021). In an online study using a video-call application, when one participant introduced an incorrect detail and the other accepted it, the second participant’s later answer about that detail was most often wrong; when the second participant questioned it, the later answer was most often correct (Kękuś et al., 2024).
Discussion is not uniformly harmful. In a field study of pub patrons, correct information from a co-witness statement increased the accuracy of later reports, while incorrect information led some participants to adopt it, regardless of their level of intoxication (Bartlett et al., 2022). In a study modeled on campus sexual assault, a co-witness’s false denial of details during an informal conversation did not measurably change the other participant’s memory of the filmed assault a week later (Bücken et al., 2024). Conversation can also shape memory by omission: a listener can become less able to recall related details that a speaker left unmentioned (Cuc et al., 2007; Lane & Houston, 2021, pp. 133–134). The practical lesson is not that discussion invalidates an account but that what was discussed, with whom, and when can matter.
News, Social Media, and Public Information
Witnesses often encounter news coverage soon after an event, more than once, and from sources they regard as credible. After a cargo plane crashed into an Amsterdam apartment building in 1992, researchers asked people whether they had seen television film of the moment of the crash; no such film existed, yet many survey respondents reported having seen it (Crombag et al., 1996). In a study conducted during the week before Ireland’s 2018 abortion referendum, 3,140 participants read six news stories, two of them fabricated; almost half reported a memory of at least one fabricated event, participants were more likely to remember fabricated scandals that fit their own voting preference, and a later warning reduced but did not eliminate the false memories (Murphy et al., 2019).
These studies show that publicly circulating information can be adopted into memory, especially when it fits what a person already believes; they do not show that a particular witness adopted a particular detail. The practical question is what the coverage, posts, or shared images contained, when the witness saw them, and whether the disputed detail appeared in the account before that exposure, a sequence publication dates and message histories often allow counsel to reconstruct.
Investigators: Feedback and Questioning
Leading and misleading questions are covered in The Misinformation Effect, and questioning of children in Suggestibility of Children’s Memory. Two further investigator-related sources deserve mention here.
The first is feedback after an identification. In the study that named the effect, witnesses who had made a mistaken identification from a culprit-absent lineup and were then told “Good, you identified the suspect” later reported more certainty, a better view, and closer attention than witnesses given no feedback (Wells & Bradfield, 1998). A meta-analysis confirmed the effect on retrospective reports (Douglass & Steblay, 2006), and a later review recommended double-blind procedures and recording confidence before any feedback (Steblay et al., 2014). Because courts consider certainty and view in assessing reliability, feedback can distort the very information used to evaluate an identification. Recent experiments mark limits: feedback reliably inflated retrospective reports but did not consistently change what witnesses did, such as offering a statement (Fukushima et al., 2024). See Confidence and Accuracy.
The second is information shared during an investigation. Showing a witness photographs or telling the witness what another witness said or what the evidence shows can each supply content. Real photographs can lend plausibility to suggested events; in one study, viewing a genuine school class photograph increased the rate at which adults came to report a fabricated childhood event (Lindsay et al., 2004; see Imagination Inflation and False Memory).
Family, Friends, Therapists, and Advocates
Witnesses and complainants talk with the people closest to them, and the research on friends and partners as influential co-witnesses applies to them (Hope et al., 2008). Children are a special case: after children aged 3 to 8 took part in science demonstrations and then heard their parents read a story describing both experienced and nonexperienced events, many described fictitious events in response to open-ended prompts, and accuracy fell further with direct questions, especially for younger children (Poole & Lindsay, 2001). Retelling also changes the teller. People shape a retelling to their audience and purpose, and a retelling slanted toward one interpretation can bias what the teller later remembers (Tversky & Marsh, 2000).
Advocate support is associated with better treatment by police and medical personnel and less distress afterward (Campbell, 2006), and nothing in the contamination research suggests that support itself compromises memory. The contamination question concerns content: whether a specific detail, interpretation, or sequence was supplied to the witness and later appeared in the witness’s account. Therapy raises a distinct issue when memories are recovered after a period of reported amnesia, where research has linked recovered-memory reports to particular therapeutic practices (Patihis & Pendergrast, 2019); that debate, with the evidence on both sides, is covered in Recovered and Repressed Memory. For sexual assault complainants specifically, including advocates, therapists, peer discussion, and repeated interviewing, see the paper “Memory Contamination: Discussion, Interviews, and Post-Event Information” in the Counterintuitive Testimony resource.
Every one of these sources also operates on accused persons and defense witnesses, for example an accused who learns the complainant’s account from investigators, command, friends, or discovery.
Limits of the Evidence
Most of this research uses students or online samples, brief filmed events, low stakes, and peripheral details. Real witnesses may care more about the event, may have seen it longer, and may encounter outside information repeatedly over months; the research does not settle whether those differences make memories more or less resistant. Co-witness studies usually create a direct conflict by showing partners different versions of the event, which may produce more opportunity for conformity than many real events. The field surveys establish that discussion is common, not that it changed any account.
The results are group averages, and in most studies many participants did not adopt the other person’s information. Experiments can show that outside information can change a report; they cannot show that a particular detail in a particular account came from an outside source. Once exposure has occurred, confidence, vividness, and detail do not reliably separate contaminated details from genuine ones (as summarized in Lane & Houston, 2021, p. 81); confidence recorded at a first fair identification is a separate matter (see “Confidence and Accuracy”), so the origin of a disputed detail usually has to be traced through a record of what the witness said, heard, and saw, and when.
Questions to Ask About This Research
When the 71% figure from Gabbert et al. (2003) is cited, counsel should ask what exactly was counted: participants who reported at least one item they could have learned only from a partner who saw a different version in an event designed to produce conflicting details, not the share of each account that was contaminated. When the Amsterdam crash study (Crombag et al., 1996) or the referendum study (Murphy et al., 2019) is cited to suggest that news coverage shaped a witness’s memory, ask what the coverage contained, whether the witness saw it, and whether the disputed detail appeared before the exposure. When feedback research (Wells & Bradfield, 1998; Steblay et al., 2014) is offered, ask whether confidence was recorded before any feedback and whether the feedback here resembled the explicit confirmation used in the studies. When agreement among witnesses is offered as corroboration, ask whether they spoke before giving statements. When discussion is said to have contaminated an account, ask whether anyone challenged the detail (Kękuś et al., 2024) and whether the information exchanged was correct (Bartlett et al., 2022). These questions are explained in “Reading Research Claims: A Guide for Attorneys” on this site.
External Sources of Memory Distortion and the Law
Federal and Military Standards
The constitutional screen for suggestive identification applies only to suggestion arranged by law enforcement. In Perry v. New Hampshire (2012), the Supreme Court held that due process does not require a preliminary judicial inquiry into reliability when law enforcement did not arrange the suggestive circumstances, and relied instead on trial safeguards: counsel, cross-examination, jury instructions on eyewitness fallibility, the rules of evidence, and proof beyond a reasonable doubt. Contamination from co-witnesses, news, social media, or family therefore ordinarily goes to weight. Expert testimony explaining those sources is governed by Federal Rule of Evidence 702, as amended December 1, 2023, under which the proponent must show it is more likely than not that the testimony rests on sufficient facts or data, uses reliable methods, and reflects a reliable application of them; Daubert v. Merrell Dow Pharmaceuticals, Inc. (1993) assigned judges the gatekeeping role, and Kumho Tire Co. v. Carmichael (1999) extended it to all expert testimony, including technical or other specialized knowledge.
In courts-martial, Military Rule of Evidence 321 makes an identification inadmissible on a timely motion or objection if it resulted from an unlawful lineup or other identification process “conducted by the United States or other domestic authorities,” defined to include a process so suggestive as to create a substantial likelihood of misidentification, or if exclusion is required by due process (Mil. R. Evid. 321(b), (c)(1)). On the rule’s terms, suggestion from a co-witness or the media is not a process conducted by authorities; whether an investigator’s disclosures bear on a challenge depends on the circumstances. Military Rule of Evidence 702 tracks the federal rule; under Military Rule of Evidence 1102 the 2023 amendment became part of the military rule in 2025 (Joint Service Committee on Military Justice, 2025; practitioners should confirm current text), and military judges apply the six factors of United States v. Houser (1993). An expert may not offer the functional equivalent of an opinion that a witness is truthful or accurate (United States v. Brooks, 2007).
Arizona
Arizona Rule of Evidence 702 adopted the Daubert standard effective January 1, 2012, and was amended effective January 1, 2024, to conform to the 2023 federal amendment (Ariz. Sup. Ct. Order R-23-0004). State v. Chapple (1983), Arizona’s leading case on memory and eyewitness expert testimony, held that excluding such testimony was an abuse of discretion on facts in which the identifications were critical and contested, while leaving admission otherwise to the trial court’s discretion; State v. Nordstrom (2001) upheld limiting an eyewitness expert to the general factors affecting reliability, without an opinion on the accuracy of a particular witness. Experts may not opine on whether a particular witness is truthful (State v. Lindsey, 1986, 149 Ariz. at 475), and State v. Salazar-Mercado (2014) permits general, educational expert testimony that satisfies Rules 702 and 403. When police arranged a suggestive pretrial identification, a defendant may request a hearing under State v. Dessureault (1969); contamination from other sources is addressed through the safeguards Perry identified.
Special Issues in Legal Practice
The principles described above apply with equal force to every party in a legal proceeding. The science operates the same way regardless of who is asking the questions, and the research summarized here does not by itself favor either side. What differs between the parties is not the science but the vantage point — the particular places where these issues tend to surface, and the particular errors each party is best positioned to guard against. The two summaries below identify those recurring issues. They are offered for orientation, not as litigation strategy; in every case the controlling question is what the evidence in that case will, and will not, support, and the expert’s role is to inform the trier of fact rather than to advocate for either side.
Special Issues for Prosecutors
The research supports several propositions prosecutors can legitimately present. Discussion among witnesses and with family, friends, and advocates is common and does not by itself make an account unreliable; correct information from others can improve accuracy, challenged misinformation is less often adopted, and in at least one study a co-witness’s false denials did not alter memory for the event (Bartlett et al., 2022; Bücken et al., 2024; Kękuś et al., 2024). Support from advocates is associated with better outcomes for complainants and is not evidence of contamination. A detail that appears in a recorded first account given before any identifiable exposure rests on firm ground, and the State is well positioned to create that record by separating witnesses promptly, recording first statements, using open-ended questions, recording confidence before any feedback, and documenting what each witness was shown or told.
The research also counsels caution. Agreement among witnesses who talked with one another is not independent corroboration, and conformity can raise confidence without raising accuracy (Luus & Wells, 1994; Sousa & Jaeger, 2022). Confirming feedback after an identification inflates a witness’s reported certainty and recollection of viewing conditions, so confidence and viewing estimates given after feedback deserve less weight than those recorded before it. When a disputed detail first appears after a conversation, a news story, or a disclosure by an investigator, the factfinder is entitled to consider that sequence, and the same principles apply to an accused’s statements: a nonpublic detail in an admission shows knowledge only if it can be shown that the detail was not supplied.
Special Issues for Defense
The research supports several propositions defense counsel can legitimately present. Witnesses who discuss an event sometimes adopt each other’s details, particularly when they saw different things, and may report them sincerely and sometimes confidently; friends and partners are especially influential (Gabbert et al., 2003; Hope et al., 2008; Wright et al., 2009). Discussion is common among real witnesses, including by text and online (Cadavid & Luna, 2021; Skagerberg & Wright, 2008). Public information can be adopted into memory, especially when it fits a person’s beliefs (Murphy et al., 2019), and confirming feedback distorts witnesses’ reports of certainty and viewing conditions (Wells & Bradfield, 1998). Evidence of what a witness was exposed to, combined with expert testimony explaining these processes, is a legitimate aid to the factfinder, particularly where agreement among witnesses is offered as corroboration.
The research also counsels caution about its limits. Contamination is a risk and a mechanism, not a finding about a particular witness; many participants in these studies did not adopt others’ information, and most studies concern peripheral details of filmed events. A contamination argument is strongest when it identifies a specific source, links it to a specific detail, and shows that the exposure came before the detail first appeared, and weakest when the detail was recorded before any identifiable exposure. Concerns about advocates or therapists are fairly raised only as to specific content, not the fact of support. An expert cannot testify that a particular detail was contaminated or that a witness’s memory is false, and the same principles apply to the accused’s own account.
This paper is for educational purposes only and is not legal advice. The author is a psychologist, not an attorney; summaries of law are general and must be verified against current authority or with a licensed attorney.
References
Bartlett, G., Albery, I. P., Frings, D., & Gawrylowicz, J. (2022). The effects of alcohol and co-witness information on memory reports: A field study. Psychopharmacology, 239(9), 2945–2953. https://doi.org/10.1007/s00213-022-06179-5
Bücken, C. A., Mangiulli, I., & Otgaar, H. (2024). Denials in informal co-witness conversations do not affect memory for witnessed events. Applied Cognitive Psychology, 38(2), Article e4193. https://doi.org/10.1002/acp.4193
Cadavid, S., & Luna, K. (2021). Online co-witness discussions also lead to eyewitness memory distortion: The MORI-v technique. Applied Cognitive Psychology, 35(3), 621–631. https://doi.org/10.1002/acp.3785
Campbell, R. (2006). Rape survivors’ experiences with the legal and medical systems: Do rape victim advocates make a difference? Violence Against Women, 12(1), 30–45. https://doi.org/10.1177/1077801205277539
Crombag, H. F. M., Wagenaar, W. A., & van Koppen, P. J. (1996). Crashing memories and the problem of “source monitoring.” Applied Cognitive Psychology, 10(2), 95–104. https://doi.org/10.1002/(SICI)1099-0720(199604)10:2<95::AID-ACP366>3.0.CO;2-%23
Cuc, A., Koppel, J., & Hirst, W. (2007). Silence is not golden: A case for socially shared retrieval-induced forgetting. Psychological Science, 18(8), 727–733. https://doi.org/10.1111/j.1467-9280.2007.01967.x
Douglass, A. B., & Steblay, N. (2006). Memory distortion in eyewitnesses: A meta-analysis of the post-identification feedback effect. Applied Cognitive Psychology, 20(7), 859–869. https://doi.org/10.1002/acp.1237
Fukushima, Y., Jordan, K., & Garry, M. (2024). To what extent does post-identification feedback translate into witness behaviour? Memory, 32(8), 959–967. https://doi.org/10.1080/09658211.2024.2372384
Gabbert, F., Memon, A., & Allan, K. (2003). Memory conformity: Can eyewitnesses influence each other’s memories for an event? Applied Cognitive Psychology, 17(5), 533–543. https://doi.org/10.1002/acp.885
Hope, L., Ost, J., Gabbert, F., Healey, S., & Lenton, E. (2008). “With a little help from my friends…”: The role of co-witness relationship in susceptibility to misinformation. Acta Psychologica, 127(2), 476–484. https://doi.org/10.1016/j.actpsy.2007.08.010
Ito, H., Barzykowski, K., Grzesik, M., Gülgöz, S., Gürdere, C., Janssen, S. M. J., Khor, J., Rowthorn, H., Wade, K. A., Luna, K., Albuquerque, P. B., Kumar, D., Singh, A. D., Cecconello, W. W., Cadavid, S., Laird, N. C., Baldassari, M. J., Lindsay, D. S., & Mori, K. (2019). Eyewitness memory distortion following co-witness discussion: A replication of Garry, French, Kinzett, and Mori (2008) in ten countries. Journal of Applied Research in Memory and Cognition, 8(1), 68–77. https://doi.org/10.1037/h0101833
Joint Service Committee on Military Justice. (2025). Amendments to the Military Rules of Evidence since 1984 (G. E. Maggs, Ed.; 2025 ed.). https://jsc.defense.gov/Portals/99/MRE%20Amendments%202025.pdf
Kękuś, M., Szpitalak, M., Polczyk, R., & Barzykowski, K. (2024). Online misinformation can distort witnesses’ memories: Analysis of co-witness discussions using an online version of the MORI-v technique. Frontiers in Psychology, 14, Article 1239139. https://doi.org/10.3389/fpsyg.2023.1239139
Lane, S. M., & Houston, K. A. (2021). Understanding eyewitness memory: Theory and applications. New York University Press.
Lindsay, D. S., Hagen, L., Read, J. D., Wade, K. A., & Garry, M. (2004). True photographs and false memories. Psychological Science, 15(3), 149–154. https://doi.org/10.1111/j.0956-7976.2004.01503002.x
Luus, C. A. E., & Wells, G. L. (1994). The malleability of eyewitness confidence: Co-witness and perseverance effects. Journal of Applied Psychology, 79(5), 714–723. https://doi.org/10.1037/0021-9010.79.5.714
Murphy, G., Loftus, E. F., Grady, R. H., Levine, L. J., & Greene, C. M. (2019). False memories for fake news during Ireland’s abortion referendum. Psychological Science, 30(10), 1449–1459. https://doi.org/10.1177/0956797619864887
Patihis, L., & Pendergrast, M. H. (2019). Reports of recovered memories of abuse in therapy in a large age-representative U.S. national sample: Therapy type and decade comparisons. Clinical Psychological Science, 7(1), 3–21. https://doi.org/10.1177/2167702618773315
Poole, D. A., & Lindsay, D. S. (2001). Children’s eyewitness reports after exposure to misinformation from parents. Journal of Experimental Psychology: Applied, 7(1), 27–50. https://doi.org/10.1037/1076-898X.7.1.27
Skagerberg, E. M., & Wright, D. B. (2008). The prevalence of co-witnesses and co-witness discussions in real eyewitnesses. Psychology, Crime & Law, 14(6), 513–521. https://doi.org/10.1080/10683160801948980
Sousa, W. S., & Jaeger, A. (2022). Memory conformity for high-confidence recognition of faces. Memory & Cognition, 50(6), 1147–1156. https://doi.org/10.3758/s13421-022-01325-y
Steblay, N. K., Wells, G. L., & Douglass, A. B. (2014). The eyewitness post identification feedback effect 15 years later: Theoretical and policy implications. Psychology, Public Policy, and Law, 20(1), 1–18. https://doi.org/10.1037/law0000001
Tversky, B., & Marsh, E. J. (2000). Biased retellings of events yield biased memories. Cognitive Psychology, 40(1), 1–38. https://doi.org/10.1006/cogp.1999.0720
Wells, G. L., & Bradfield, A. L. (1998). “Good, you identified the suspect”: Feedback to eyewitnesses distorts their reports of the witnessing experience. Journal of Applied Psychology, 83(3), 360–376. https://doi.org/10.1037/0021-9010.83.3.360
Wright, D. B., Memon, A., Skagerberg, E. M., & Gabbert, F. (2009). When eyewitnesses talk. Current Directions in Psychological Science, 18(3), 174–178. https://doi.org/10.1111/j.1467-8721.2009.01631.x
Legal Authorities
Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993).
Kumho Tire Co. v. Carmichael, 526 U.S. 137 (1999).
Perry v. New Hampshire, 565 U.S. 228 (2012).
State v. Chapple, 135 Ariz. 281, 660 P.2d 1208 (1983).
State v. Dessureault, 104 Ariz. 380, 453 P.2d 951 (1969).
State v. Lindsey, 149 Ariz. 472, 720 P.2d 73 (1986).
State v. Nordstrom, 200 Ariz. 229, 25 P.3d 717 (2001).
State v. Salazar-Mercado, 234 Ariz. 590, 325 P.3d 996 (2014).
United States v. Brooks, 64 M.J. 325 (C.A.A.F. 2007).
United States v. Houser, 36 M.J. 392 (C.M.A. 1993).
Fed. R. Evid. 702.
Ariz. R. Evid. 403.
Ariz. R. Evid. 702.
Ariz. Sup. Ct. Order R-23-0004 (amending Ariz. R. Evid. 702, effective Jan. 1, 2024).
Mil. R. Evid. 321.
Mil. R. Evid. 702.
Mil. R. Evid. 1102.
Related topic papers: The Misinformation Effect · Source Monitoring · Confidence and Accuracy · Remembering Changes Memory · Suggestibility of Children’s Memory · Imagination Inflation
Related topics from Arizona Forensic Services: Counterintuitive Testimony · Child Abuse Allegations