National Research Council Urges New Best Practices for Eyewitness Identification

The National Research Council has released a new report on best practices for law enforcement agencies and courts to improve the accuracy and reliability of eyewitness identifications in criminal cases.

Rape, robbery, and other assaults by strangers are the cases in which eyewitness identification is most often important.

Research has increasingly uncovered the limits of eyewitness identifications—limits imposed both by human nature and by flaws in law enforcement agencies’ procedures. For example, unintentional cues from law enforcement officers during a line-up can lead to a misidentification. Also, conditions such as lighting of the scene and stress of the situation create gaps in a witness’s perception of an event. These gaps are can be filled by the individual’s prior experiences, leading to a mistaken identification.

Studies have shown that a witness is more likely to make a mistaken identification when identifying a person of a different race. Also, according to the report, “people’s memories are continuously evolving;” one’s memory of an event can change without one’s realizing it.

The report calls for several reforms in the way law enforcement agencies conduct eyewitness identifications. First, officers should receive training in minimizing the risk of contaminating identifications. Next, the report calls for double-blind lineups; if the lineup administrator does not know the suspect’s identity, then the administrator cannot give any cues—even unconscious ones—to the witness.

The report also says officers should document a witness’s confidence in an identification immediately. The witness’s immediate assessment of his or her confidence in the identification is more accurate than later assessments. In other words, one’s memory of how confident one was, like one’s memory of the crime, often changes over time.

For reforms in the court system, the report calls on judges to conduct pretrial inquiries into the identification process to ensure its fairness and reliability. The court should also give juries detailed information of any prior identifications; that would help juries gauge a witness’s credibility.

Finally, the report calls for permitting expert witnesses to give juries detailed information about how eyewitness identifications work and the hazards of human perception. In Washington state and elsewhere, the courts have been reluctant to allow expert testimony on these subjects. The courts have often said that these are matters of common experience, so expert testimony is not needed.