New Guidelines Strive To Minimize Blind Spots In Eyewitness Testimony, Researcher Says
FAYETTEVILLE, Ark. - A recent report by the National Institute of Justice tackles important issues concerning eyewitness evidence, but needs to go even further to prevent potential miscarriage of justice, a University of Arkansas law professor says.
Eyewitness testimony remains one of the most damning pieces of evidence in a trial, but it, like fingerprints, can be inadvertently contaminated if not handled correctly, said Donald P. Judges, Ben J. Altheimer Professor of Legal Advocacy.
"What we really have are imprints in the brain of the witness," Judges said. Many social and psychological factors contribute to eyewitness accuracy, and the results of social science studies are beginning to change the practice of law enforcement.
The National Institute of Justice report, "Eyewitness Evidence: A Guide for Law Enforcement," (the Guide) outlines practices and procedures law enforcement officials should follow in investigations involving eyewitnesses. A working group of 34 people including law enforcement officials, prosecutors, defense attorneys and social science researchers, was called together by Attorney General Janet Reno to address growing concerns about erroneous convictions.
The new guidelines offer recommendations for changes in the way police officers and detectives interview witnesses and present suspect line-ups, with the intention of helping good law enforcement officials do their job even more effectively. They are a good start, but they don’t go far enough, Judges said.
His article "Two Cheers for the Department of Justice’s Eyewitness Evidence: A Guide for Law Enforcement," has just been published in a special faculty edition of the Arkansas Law Review.
As DNA evidence has become available, some cases involving eyewitnesses have been overturned because the purported perpetrator’s DNA did not match what was found at the crime scene. In these cases, despite eyewitness evidence to the contrary, people were incarcerated for crimes they did not commit, while the real perpetrators remained free.
Studies in the social sciences have shown that the accuracy of the eyewitness interview and suspect identification can be powerfully influenced by the way they are implemented.
Assumptions made by witnesses, coupled with verbal or non-verbal cues or omissions on the part of law enforcement officials, can lead to wrongful accusations.
"There are lots of ways, many of them quite unintentional, to contaminate the witnesses’ memory," Judges said. A chance remark, a sudden blink, or even assumptions made by the witness can all lead to a false identification, and even conviction.
Three main components characterize memory: acquisition, storage and retrieval. Law enforcement officers must concern themselves with how to best retrieve memories.
During an interview, the eyewitness interacts with a law enforcement official, a person perceived by the cooperative witness as trustworthy and in a position of authority. Unconscious social pressures can lead the witness to try to please the officer, even if the information they offer is not entirely accurate.
This social response is heightened by repeating the same questions - witnesses may infer that the first answer was not satisfactory. Response can also be influenced by multiple choice questions - was the car green or blue? The witness may think that the answers "red" or "I don’t know" are not acceptable.
"However, the ultimate goal is not frictionless existence. It’s truth," Judges said.
Eyewitnesses may also hear suggestive information from the media and other witnesses, which can color their testimony.
Social science has determined that the cognitive interview is an effective technique for more accurate memory retrieval in witnesses. This includes establishing rapport, an open-ended narration, probing, review and a closing stage. Such an interview encourages the witness to recall every aspect of the event and can be quite time consuming, Judges said.
Some of the same social factors that govern the witness interview also factor into witness identification of a suspect. When a witness sees a line-up, he or she may infer that the perpetrator must be in it, unless someone specifically states that the guilty party may or may not be present. Also, the person may once again want to please the authority figure by making a positive identification, and may fear that the investigation will end if no identification is made.
Social science has also found that sequential line-ups, where suspects are presented one at a time, yield fewer false positives than simultaneous line-ups, where suspects are shown all at once. This is because people are more prone to make relative judgements, Judges said. Studies have shown that people will identify a suspect in a line-up, but if shown another line-up with the suspect removed they will tend to pick the person who looks most like the first suspect chosen: They do not always say, "The guilty person’s not there anymore."
People can’t compare one suspect to another in a sequential line-up. Instead, they must make an absolute judgement - is this the person or isn’t it?
A chance remark after identification - "That’s the one we caught yesterday," or "You got the right person" - can enhance a witnesses’ confidence but not the accuracy of the identification. Witness confidence should be assessed immediately after the identification, for many witnesses become more certain with the passing of time, and that confidence has little to do with accuracy. Unfortunately, juries place a high value on confident witnesses, despite their demonstrated lack of link to accuracy, Judges said.
The new Justice Department Guide is a landmark event in the use of social science research by law enforcement, Judges said.
"This is a success story for the science of psychology and the legal profession," he said.
In his paper, he addresses several points in which the guidelines could have gone further to prevent erroneous convictions.
With interviews, the Guide instructs 911 emergency call dispatchers to ask open-ended questions followed by more specific questions. "The manner in which facts are elicited from a caller can influence the accuracy of the information obtained," the Guide says.
The Guide also suggests that investigating officers separate multiple witnesses and instruct them to avoid discussing details of the event with other witnesses or the media, and to steer clear of media coverage of the incident.
The Guide tells interviewers to encourage witnesses to report all details of the event and to use drawings, gestures or objects to describe the event. Interviewers should also avoid volunteering information about the event or interrupting the witness, and should tell the witness not to guess. The guide recommends documentation of the interview and encourages follow-up interviews as necessary.
However, the guidelines leave out several points, including the concepts underlying the cognitive interview and obtaining confidence statements at the time of questioning, Judges said.
For eyewitness identification, the Guide makes specific recommendations about instructions to the witness; the investigator should state that the perpetrator may or may not be in the line-up, and that the investigation will continue regardless of whether or not the witness makes an identification. The Guide also recommends that the investigator get a statement of the witnesses’ confidence in the identification, and that the procedure be documented.
The Guide also addresses the issue of line-up bias by recommending that the fillers in the line-up match the suspect’s description, and that any outstanding scars or marks be omitted or added to ensure that the suspect does not stand out in the line-up.
The Guide does not go so far as to recommend some things that social scientists have found yield fewer false accusations, like separate line-ups for multiple witnesses, or "mock witness" procedures to test for false identifications, where witnesses are first shown a line-up without the suspect. Nor does it endorse the double-blind administration of the line-up, which research has shown can help eliminate eyewitness errors, Judges said.
It also fails to recommend sequential line-up identifications over simultaneous ones, which introduces a comparative error, where the witness picks out the person who looks "most like" the one he or she remembers, instead of making a positive identification.
These points remain essential in making accurate identifications, Judges said.
"The vast majority of wrongful convictions are based on mistaken eyewitness testimony," Judges said. "I hope the Justice Department continues to work in this area and to address the issues highlighted in this paper."
# # #
Contacts
Donald P. Judges, Ben J. Altheimer Professor of Legal Advocacy, (479) 575-7571, djudges@comp.uark.eduMelissa Blouin, science and research communications manager, (479) 575-5555, blouin@comp.uark.edu