Memory Errors in the Legal System: The Unreliability of Eyewitness Testimony
Eyewitness testimony is often regarded as a foundational form of evidence and has historically been perceived as one of the strongest, most reliable pieces of evidence one can obtain in a criminal case. [1] However, this is far from the truth. Eyewitness testimony is highly unreliable and plays a role in wrongfully convicting hundreds of innocent people. About sixty-nine percent of exonerations based on DNA evidence found after the trial were from cases that involved eyewitness testimony in the conviction. [2] This is problematic because eyewitness testimony often plays a direct role in the outcomes of criminal cases; about three percent of felony cases use eyewitness identification as the only critical evidence when making a judgment or conviction. This may not seem like a large number, but three percent represents about seventy-seven thousand cases and individuals who are directly influenced by these reports. [1] The unreliability of eye-witness testimony is mainly due to flaws in memory such as weapon focus, source monitoring errors, and the post-identification feedback effect. As such, evidence collection processes need to be amended to reduce wrongful convictions and ensure the justness of our legal system.
A key factor in the unreliability of eyewitness testimony is the weapon focus effect. This refers to the human tendency to focus on the weapon when a crime is occurring as opposed to the criminal. During an important event, people may have poor memory of the perpetrator of the crime due to dedicating most of their attention to the perpetrator’s weapon, and this often results in inaccurate descriptions of the criminal from eyewitnesses. [4] Experiments show that witness accuracy in describing the perpetrator is worse in conditions where a weapon is present in comparison to conditions without a weapon [5] and that witnesses recall less feature information overall about the suspect when a weapon is visible. [5] One possible explanation for this effect is that we dedicate more attention to unexpected objects; another is that we dedicate more attention to objects that may cause us harm due to instincts for self-preservation. [6] Either way, this effect is problematic as a large percentage of crimes in the U.S. involve guns. In 2021, around eighty-one percent of murders in the U.S. involved a firearm [7], and around seventy-eight percent of robberies involved dangerous weapons. [8] In all these cases, eyewitness testimony should not be taken as fact; it should be acknowledged and understood that these accounts are likely flawed. As such, eyewitness testimony should still be used to support other existent evidence but should not be allowed to stand as the only basis for conviction.
Another effect that contributes to the unreliability of eyewitness testimony is source confusion, sometimes known as source monitoring errors. Source confusion occurs when people do not remember the origin of certain memories or beliefs. [9] This is problematic because witnesses may report things about the suspect that they believe they remember, but really arose from other sources. This issue presents itself in eyewitness testimony in three main ways. First, familiarity with a suspect in a lineup from an unrelated source can cause an eyewitness to identify them as the criminal. [10] This occurs due to misattribution of the source of familiarity; people think they recognize someone from the crime but really recognize them from a mundane, unrelated place or previous lineup. For example, one man was wrongfully convicted after being identified as the criminal by an eyewitness in a physical lineup. The witness thought she recognized this man because he was at the crime scene; in reality; she recognized him because she had seen him in a previous photo lineup. This man was convicted and served twenty-seven years in prison before DNA evidence was found that exonerated him. [11] Fully combating this effect is challenging, but an effort should be made to avoid showing witnesses the same person in multiple lineups and to select lineup fillers from areas that aren’t very close to where the witness is from. These measures will reduce the chance that witnesses recognize suspects from unrelated places and misattribute that memory to the crime.
The second main way that source monitoring errors can impair eyewitness testimony is through witness collaboration. In these cases, eyewitnesses can believe they remember something about the suspect from the crime even though they really remember these facts from talking to other witnesses. One example of this is when multiple witnesses in a crime reported remembering a physical characteristic of the criminal that ended up being false. Police later found that only one witness had remembered this attribute and shared it with the others when they were talking after the crime. Later, the other witnesses reported recalling this about the suspect even though they really just recalled hearing it from the other witness in conversation. [12] To combat source confusion due to witness collaboration, witnesses should not be allowed to confer immediately following a crime. This will ensure they maintain independent recollections of events and aren’t influenced by each other’s accounts.
A third example of source monitoring errors in eyewitness testimony is through leading interviews. Leading questions from interviewers containing details of the crime can cause eyewitnesses to later report remembering these details despite reporting they didn’t remember them when they were first questioned. [13] Experiments on this phenomenon show that even slight wording variations in the interview questions can prompt different answers. [13] Due to source monitoring errors, leading questions can create false memories for witnesses who believe they remember these things from the crime instead of their interview. To minimize source confusion from leading questions, interviewers should adopt a more open-ended interview procedure where they ask witnesses to recall the crime with minimal interruption, prodding, questioning, or guidance. This decreases the influence that may arise from interviewer suggestions or leading questions.
The last memory phenomenon that causes unreliability in eyewitness testimony is the post-identification feedback effect. This refers to the tendency of witness reports to get distorted by feedback suggestive of the accuracy of their suspect identifications. [14] For example, if a suspect makes an identification they are initially uncertain of and receives feedback that suggests their choice was correct, they will rate their confidence in their identification higher when asked about it at a later date. [15] The structure of lineups worsen this effect, as they are often presented as though the actual criminal is definitively in the lineup. When faced with this, witnesses often feel more pressure to make a choice and will select a suspect even if they don’t truly recognize any as being the perpetrator. In these cases, witnesses are usually unsure about their choice, but may still report feeling very confident in their selection at a later date due to the post-identification feedback effect. This makes it seem as though the eyewitness is credible and confident in their choice even though they may have been very uncertain at the time of the actual lineup. To reduce the post-identification feedback effect, interviewers should not give any indication of accuracy after the witness chooses a suspect and refrain from giving any response. It should also be standard procedure to tell witnesses that the real criminal might not be in the lineup and to have witnesses rate their confidence in their identification immediately following the lineup. This will reduce pressure to make a selection and combat the post-identification feedback effect. Lastly, a “blind” administrator, someone who doesn’t know the identity of the true suspect, should be used in lineups so their pre-existent biases can’t influence the choice of the witness.
Due to multiple flaws with eyewitness testimony caused by weapon focus, source confusion, and the post-identification feedback effect, the American legal system needs to be amended so eyewitness testimony doesn’t continue to lead to wrongful convictions. These issues raise concerns about the current trial standards for evidence—procedures that are directly determining the lives of thousands of innocent people. We must question our reliance on eyewitness testimony, a type of evidence that is greatly susceptible to distortion, and amend current precedents to uphold the integrity of our legal system.
Edited by Hanrui Huang
Endnotes
[1] Chew, Stephen L, “Myth: Eyewitness Testimony Is the Best Kind of Evidence.” Association for Psychological Science - APS, August 20, 2018, https://www.psychologicalscience.org/uncategorized/myth-eyewitness-testimony-is-the-best-kind-of-evidence.html.
[2] “How Eyewitness Misidentification Can Send Innocent People to Prison,” Innocence Project, April 23, 2023, https://innocenceproject.org/how-eyewitness-misidentification-can-send-innocent-people-to-prison/.
[3] Alvin Goldstein, June Chance, & Gregory Schneller, “Frequency of Eyewitness Identification in Criminal Cases,” University of Missouri, https://link.springer.com/content/pdf/10.3758/BF03329902.pdf.
[4] Kramer, Buckhout, & Eugenio, “Weapon Focus, Arousal, and Eyewitness Memory: Attention Must be Paid,” Office of Justice Programs, https://www.ojp.gov/ncjrs/virtual-library/abstracts/weapon-focus-arousal-and-eyewitness-memory-attention-must-be-paid.
[5] Nancy Steblay, “The Weapon Focus Effect in Eyewitness Memory,” May 26, 2023, https://www.oxfordbibliographies.com/display/document/obo-9780199828340/obo-9780199828340-0313.xml.
[6] Ibid.
[7] Carlson, Pleasant, Weatherford, Carlson, & Bednarz, “The Weapon Focus Effect: Testing an Extension of the Unusualness Hypothesis,” Digital Commons @ Texas A&M University-San Antonio, December 2016, https://digitalcommons.tamusa.edu/psyc_faculty/8/#:~:text=The%20weapon%20focus%20effect%20(WFE,and%20unusual%20objects%20distract%20eyewitnesses.
[8] Gramlich, John, “What the Data Says about Gun Deaths in the U.S,” Pew Research Center, April 26, 2023, https://www.pewresearch.org/short-reads/2023/04/26/what-the-data-says-about-gun-deaths-in-the-u-s/.
[9] “Federal Robbery: Prevalence, Trends, and Factors in Sentencing,” United States Sentencing Commission, December 8, 2022, https://www.ussc.gov/research/research-reports/federal-robbery-prevalence-trends-and-factors-sentencing#:~:text=In%20fiscal%20year%202021%2C%20a,percent%20of%20robberies%20involving%20weapons.
[10] Pilat D. & Sekoul D. “Source Confusion.” The Decision Lab, 2021, https://thedecisionlab.com/biases/source-confusion.
[11] “Don’t I Know You? The Effect of Prior Acquaintance/Familiarity on Witness Identification,” NACDL, April 2012,
https://www.nacdl.org/Article/April2012-DontIKnowYouTheEffectofPriorAc.
[12] Pozzulo, Joanna, “Eyewitness Descriptions and Identifications of Familiar Perpetrators,” OUP Academic, November 21, 2019, https://academic.oup.com/book/32376/chapter/268651175.
[13] Gabbert, Fiona, & Hope, “Suggestibility in the Courtroom: How Memory Can Be Distorted during the Investigative and Legal Process,” OUP Academic, October 19, 2017, https://academic.oup.com/book/2678/chapter/143103519.
[14] “Eyewitness Testimony and Memory Biases,” Noba, https://nobaproject.com/modules/eyewitness-testimony-and-memory-biases#:~:text=Likewise%2C%20eyewitness%20memory%20can%20be,the%20legal%20system%20are%20real.
[15] Steblay, N. K., Wells, G. L., & Douglass, “The Eyewitness Post Identification Feedback Effect 15 Years Later: Theoretical and Policy Implications,” American Psychological Association, https://psycnet.apa.org/record/2014-03984-001.
[15] Gary Wells & Deah Quinlivan, “The Eyewitness Post-Identification Feedback Effect: What is the Function of Flexible Confidence Estimates for Autobiographical Events?”, Applied Cognitive Psychology, 2009, https://onlinelibrary.wiley.com/doi/pdf/10.1002/acp.1616.