Report details ways to improve eyewitness testimony's value

The testimony of eyewitnesses plays a key role in a majority of criminal cases.

“There is no more powerful evidence in a criminal trial than a witness pointing at someone and saying, ‘That’s the guy,’” said Arizona State University criminology and criminal justice professor Henry F. Fradella.

The problem, Fradella said, is eyewitnesses oftentimes are flat wrong.

To strengthen the value of eyewitness accounts, the National Academy of Sciences in October issued a report titled “Identifying the Culprit: Assessing Eyewitness Identification,” which provides a list of recommendations designed to set a national standard of best practices for police and courts in handling eyewitness accounts.

The report, a distillation of decades of independent research, makes recommendations such as training law enforcement in issues with eyewitness memory, developing standardized witness instructions and videotaping the witness-identification process, among others.

There’s also a recommendation for police to implement double-blind photo arrays and lineup procedures, where the individual administering the lineup doesn’t know the identity of the suspect as a way to eliminate the possibility of influencing a witness.

“If the state is going to take away your liberty, they better get it right,” said Amshula Jayaram, state policy advocate with the New York-based Innocence Project.

The group has helped to exonerate hundreds of wrongfully convicted people.

Inaccurate eyewitness identification played a role in nearly three-quarters of the wrongful conviction cases the project has worked on, Jayaram said.

The memory hole

Criminologists and psychologists have long known the inherent problems with eyewitness identifications.

A key issue has to do with memory, and how people naturally acquire, retain and retrieve events they have witnessed.

“Memory is flawed,” Fradella said. “Our memory does not function like a videotape.”

Factors such as the immediate surroundings of an event, short opportunities for observation, poor lighting and sensory overload can all affect a person’s memory.

Added to that, the mind will often naturally fill in the gaps of a memory with other nonrelated events or with memories of things seen and heard at different times in a phenomenon known as confabulation.

Upon retrieving memories, a person’s mind similarly can transfer recollections of unrelated events or of things that may not have occurred at all.

In this instance, researchers have found people’s memories can be influenced by suggestion of outside forces such as the questioning from law enforcement or from images a person was exposed to in the media.

Cross-racial identification has also been shown to be a problem in many instances.

For many eyewitnesses, accurately identifying people of other races proves difficult.

Researchers have found people can more readily discern subtleties of expressions, emotions and facial features of members of their own racial group than those of other groups.

The so-called “weapons focus” also affects a person’s recollections.

This effect holds that people confronted with a weapon-wielding person naturally tend to focus their attention on the weapon and not the individual.

“These are the sorts of things the justice system can’t do anything about,” Fradella said.

What the system can do

While policy and procedure changes can’t alter the subtleties of the mind, they can help to minimize the risk of misidentification.

The Pima County Sheriff’s Department has for many years sought to have uniformity in creating photo lineups.

The photos are drawn from a database of mugshots taken at the jail where all arrestees are placed in front of a neutral background and covered with a sheet from the neck down.

In assembling a six-pack photo lineup, a computer system selects photos based on similar physical features as the suspect.

Back to homepage

Subscriber Login

Keep reading with a digital access subscription.

Subscribe now

Police can also manually reject images that in some way stand out from the others in a lineup.

The department hasn’t implemented double-blind lineups as the NAS report recommends, but witnesses are dealt with in a low-pressure manner, said David Theel, commander of PCSD’s Violent Crimes and Targeted Offenders division.

“We tell our witnesses they are not obligated to make an identification,” Theel said.

The department also will not mingle nonmugshot photos — such as those from driver’s licenses, for example — with mugshots in order to preserve uniformity of appearance.

The Tucson Police Department does use a double-blind procedure in photo arrays.

Veteran Tucson defense attorney Brick Storts said avoiding even such a minor difference as the color of backgrounds in photo arrays has helped to eliminate many misidentifications.

“I think they try to do a fairly decent job of it,” Storts said.

That’s in particular contrast to how photo lineups were approached in decades past, he said. Storts remembers detectives simply handing witnesses stacks of photographs to sort through.

Detectives questioning witnesses also use uniform language in describing lineups to witnesses.

They tell witnesses the suspect may not be in the lineup, to study each image independently and to avoid guessing.

The department does not routinely videotape or make audio recordings of eyewitness interactions.

The department would consider adopting some of the recommendations made in the NAS report, Theel said, but the changes would have to made in cooperation with the Pima County Attorney’s Office, which prosecutes the cases police present.

Courtroom changes

The report also recommends changes to address some of the issues of misidentification in the courtroom setting.

“Unfortunately, eyewitness identification is devastating from the standpoint of what happens at trial,” Storts said.

Storts, whose legal career began in 1962, said he would like to see the courts use a more explicit jury instruction on eyewitness testimony, as is recommended in the report.

The current instruction used in most criminal cases in Arizona courts tells jurors to give all testimony the weight they think it deserves.

An instruction from the U.S. Court of Appeals for the Ninth Circuit, however, gives explicit direction to jurors, listing five factors to take into account when considering eyewitness testimony.

They include: the opportunity a witness had to observe a suspect, whether the identification was influenced later on, inconsistent identifications, if the witness had previously seen the suspect, and the totality of the circumstances surrounding the identification.

“How do we get the jury to understand that eyewitness testimony is really unreliable evidence,” Fradella said. “Some of these things that prevent wrongful conviction really are empirically sound.”

One way, he said, is for the courts to allow more expert testimony at trial explaining to jurors the problems of eyewitness accounts.

The NAS report addresses the issue with recommendations for judges on allowing expert witness testimony on eyewitness identification.

In the name of justice

Since eyewitness accounts play a key role in so many criminal prosecutions, getting it right is seen as central to the quest for justice.

“Everybody benefits when you get it right,” Jayaram said. “When you get it wrong, the only person who benefits is the real perpetrator.”

The costs of wrongful conviction go beyond the individual miscarriage of justice and spill over into the community, she said, noting every wrongful conviction leaves a criminal free to commit more crimes.

Fradella said the societal costs all adversely affect the community’s bottom line, too, by diverting money from other areas to pay for prosecution and incarceration.

“We live in an age where our resources are scarce,” he said. “These are resources that could have been used for education.”