Eyewitness Misidentification

As exhaustively indicated by hundreds of peer-reviewed social science publications and by the recent spate of exonerations based on DNA evidence, the memory and testimony of eyewitnesses is far from infallible. An eyewitness’ false identification of an innocent suspect as the perpetrator of a crime, or “eyewitness misidentification,” is the single greatest cause of wrongful conviction in the U.S. According to the Innocence Project at the Cardozo School of Law, eyewitness misidentification has played a significant role in over 75% of the more than 230 exonerations that have occurred to date based on DNA evidence. Despite the scientific consensus that eyewitness testimony is unreliable and the Supreme Court’s recognition that “the annals of criminal law are rife with instances of mistaken identification,” eyewitness identification testimony remains among the most convincing evidence presented to jurors (Manson v. Brathwaite, 1977). As former Supreme Court Justice William J. Brennan stated, there is “nothing more convincing [to a jury] than a live human being who takes the stand, points a finger at the defendant, and says ‘That’s the one!’” (Watkins v. Souders, 1982). Law enforcement officials also often mistakenly overestimate the reliability of eyewitness evidence, experiencing “tunnel vision” in which they prematurely devote all their efforts to suspects identified by eyewitnesses, regardless of the accuracy of the identifications.
Patterns in three decades of social science research indicate a variety of causes of eyewitness misidentification. These include system variables, which can be substantially reduced by reforming traditional identification procedures, and estimator variables—situational factors that cannot be directly addressed by systemic reform. By adopting the reforms recommended by the United States Department of Justice and following the lead of states such as North Carolina and Wisconsin, the Commonwealth of Kentucky can considerably reduce the number of innocent people it wrongfully imprisons. In addition, these reforms increase the likelihood that law enforcement officials will not waste time and resources following erroneous leads while actual perpetrators escape.

What causes eyewitness misidentification?
System Variables are factors related to the administration of live and photographic lineups and other identification procedures by law enforcement officials. Systemic weaknesses in traditional identification procedures that cause misidentification of perpetrators include the following: suggestive instructions to witnesses that cue them to pick the suspect or that fail to mention that the perpetrator may not be in the lineup; use of a lineup administrator who knows the identity of the suspect; selection of fillers who look significantly unlike the suspect and cause him/her to stand out; the use of “simultaneous” rather than “sequential” lineups; and post-identification comments that increase the witnesses’ potentially false confidence in his/her identification.

  1. ​Misleading or Lacking Pre-Identification Instructions: In traditional identification procedures, eyewitnesses are presented with one or more live or photograph line-ups that include a suspect and non-suspect “fillers.” If not provided with instructions to the contrary, the witness may assume that the perpetrator is in the lineup, that the witness must identify one of the lineup members as the perpetrator, or that the investigation depends on their immediately making a successful identification. Therefore, witnesses often make “relative judgments,” choosing the individual who looks most like the perpetrator when compared to the other lineup members rather than the person who actually committed the crime. This problem is exacerbated when the witness strives to make an identification at all costs.
  2. Non-“Blind” Lineup Administrator: If the administrator of an identification procedure knows the identity of the suspect, (s)he may unintentionally cue the witness, including through facial expressions and gestures, as to the “correct” answer and undermine the ability of the witness to make an independent identification from memory.
  3. Poor Filler Selection: If fillers are not chosen who look somewhat like the suspect, the suspect may stand out. For instance, if the suspect is bald, has an unusual hairstyle, has tattoos, is of a different race, or has other distinguishing features that cause the suspect to look significantly unlike the fillers, the witness may be unconsciously led to pick the suspect, especially if the suspect’s unique features match the witnesses’ description of the perpetrator and the features of the other lineup members do not.
  4. Reinforcing Post-Identification Comments: After an identification procedure, law enforcement officials sometimes make statements that inflate the witness’ confidence in his/her identification. For instance, an administrator may tell the witness that (s)he has done a good job or has identified the same person as another witness. The witness may later experience and express much greater confidence in his/her identification than the witness felt prior to hearing such comments, and may falsely attribute this greater confidence to a restoration of memory rather than to reinforcing comments or cues.
  5. Simultaneous Lineups: Members of a lineup or photo array are traditionally presented to a witness all at once rather than one at a time. This exacerbates the tendency described above for witnesses to identify the individual who looks most like the perpetrator relative to the other lineup members rather than comparing each lineup member to their independent memories of the perpetrator’s appearance.

Estimator Variables are factors related to the witness him/herself or to the circumstances surrounding the witnesses’ observation of the perpetrator, including gaps in memory, witness stress/trauma, the relative difficulty of cross-racial identification, and the degree of attention paid by the witness to the appearance of the perpetrator during the incident.

  1. Memory Gaps and Unconscious Transference: Eyewitness memory is not a transcript or video-like record. Memory is an imprecise, interpretive reconstruction of events subject to contamination in both its initial encoding and subsequent retrieval. Accurate memory of an event decreases over time. During memory reconstruction, individuals often unconsciously fill in memory gaps with information from other memories; general knowledge and assumptions; and details picked up from other witnesses, newspapers, or other sources. For instance, a victim who did not fully view an attacker’s face or can only fuzzily recall its details may incorporate a familiar face from his/her surroundings, a face that (s)he viewed around the time of the attack, or a face viewed in multiple lineups into his/her memory of the attack. This inadvertent process of filling in memory gaps is referred to as “unconscious transference.” Witnesses may later experience “source confusion,” attributing information gained from other sources to direct recall of an event.
  2. Witness Stress/Trauma: High stress levels drastically reduce the accuracy of eyewitness memory. Individuals tend to recall information best under conditions of moderate stress/arousal. This is evidenced in “Accuracy of Eyewitness Memory for Persons Encountered During Exposure to Highly Intense Stress” (Charles Morgan, et al., 2004). The experiment detailed in the article found that military survival school students interrogated face-to-face for 40 minutes in a low-stress scenario could correctly identify their interrogators in a simultaneous live or photo lineup conducted the following day 62-76% of the time, while those exposed to a high-stress scenario could do so only 30-34% of the time. The weapon-focus effect is related to this—the presence of a weapon draws the witnesses’ attention away from the perpetrator’s appearance to the weapon itself. 
  3. Cross-Racial Identification:  Eyewitness identifications of individuals of races other than the witness are, on average, less accurate than identifications of individuals of the same race, regardless of the races studied, the exposure of the eyewitness to individuals of the other race, or the degree to which the eyewitness does or does not express racist views. This may contribute to the disproportionate number of the wrongfully convicted who are minorities compared to the prison population as a whole. In “Judging Innocence,” Brandon Garrett writes that 73% of individuals wrongfully convicted of rape and exonerated between 1989 and 2007 were Black or Hispanic, while 37% of all inmates serving sentences for rape in 2003 were minorities.

Other estimator variables, as identified by the Texas Innocence Network, include:

  • Age of witness – the identifications of children and the elderly tend to be less accurate than those of adults 
  • Target salience – if a large number of people were present at the scene of the crime, the difficulty of identifying the perpetrator(s) increases 
  • Exposure duration –all things being equal, the longer the witness is exposed to the perpetrator, the more accurate the identification  
  • Cross-gender bias – identifications of males by females or females by males tend to be less accurate than identifications of males by males or females by females 
  • Changes in experiential context – when a witness views an individual in a different location, it tends to be more difficult to identify him/her 
  • Post-event information –a witness’ memory can be distorted by information obtained after an event 
  • Crime scene variables – the lighting and/or layout of the crime scene can affect the witness’ ability to perceive, and therefore recall, the identity of the perpetrator 
  • Distance between the witness and perpetrator: all things being equal, the greater the distance, the more difficult the identification 

What can be done to reduce eyewitness misidentification?
The following reforms are recommended for immediate adoption by all law enforcement jurisdictions by the national Innocence Project. Reforms also recommended in “Eyewitness Evidence: A Guide for Law Enforcement are marked with an asterisk (*). This report was independently written by a working group composed of 50% law enforcement officials and 50% lawyers and researchers and organized by the U.S. Department of Justice’s research agency, the National Institute of Justice (NIJ). 

  1. ​Specific Instructions: The lineup administrator should instruct the witness that the perpetrator “may or may not be present” in the lineup, that the investigation will continue regardless of whether the witness identifies the perpetrator during a lineup, and that the witness should not seek guidance from the administrator. 
  2. Double-Blind Administration: The lineup administrator should not know the identity of the suspect.  (The National Institute of Justice guide states that “Blind procedures are not included in the Guide but are identified as a direction for future exploration and field testing” and recognizes that “investigators’ unintentional cues […] may negatively impact the reliability of eyewitness evidence” and that psychology researches have found that “such influences could be avoided if “blind” identification procedures were employed.”)
  3. Careful Lineup Composition: Lineup administrators should select fillers “who generally fit the witness’ description of the perpetrator” (NIJ Guide). However, “when the description of the perpetrator differs significantly from the appearance of the suspect, fillers should resemble the suspect in significant features,” especially age, race, and such unique markings as scars and tattoos. The suspect should not stand out. Administrators should not include the same suspect in multiple lineups viewed by the same witness. When showing a new suspect to the same witness, administrators should avoid reusing the same fillers. Additionally, they should not place a suspect in the same position in multiple lineups. *
  4. Confidence Statements: The lineup administrator should ask the witness immediately after the identification procedure to provide a statement, in his/her own words, detailing his/her level of confidence in the identification. This statement must be recorded in writing and signed and dated by the witness. The lineup administrator should not make statements or ask questions that could influence the witnesses’ level of confidence prior to the transcription of this statement.*
  5. Recording of Identification Procedures: All identification procedures should be videotaped whenever possible to increase legitimacy and decrease the potential for misconduct. When videotaping is not possible, identifications should be photographed in a manner that “represents the lineup clearly and fairly.” The lineup and its results should also be documented in writing. *
  6. Optional: Sequential Presentation of Lineup Members: Presenting eyewitnesses with photographs or live lineup members one-at-a-time (sequentially) rather than simultaneously decreases the incidence of false identification, as indicated by a pilot project conducted in Hennepin County, Minnesota in 2006. However, the national Innocence Project considers this reform optional due to the recent controversy surrounding a similar pilot performed in 2005 by the Chicago Police Department and the “Illinois Report” that published its findings. (The National Institute of Justice guide remarks, “Scientific research indicates that identification procedures such as lineups and photo arrays produce more reliable evidence when the individual lineup members or photographs are shown to the witness sequentially—one at a time—rather than simultaneously.”)


Learn More About KIP

Click the links below to learn more about the Kentucky Innocence Project