It’s a scene that has become a staple in modern pop culture.
A witness to a crime walks into a room, looks at a lineup of stone-faced suspects from behind a two-way mirror and dramatically points to the person they saw commit a crime.
From there, it usually only takes a few dramatic speeches from lawyers before that suspect is on his or her way to prison.
As it turns out, the reality of eyewitness identifications is much more complicated and much less reliable than it is presented in movies and television shows.
“Lawyers and psychologists have all examined the subject and determined eyewitnesses are unreliable,” said Stephen Saltzburg, professor of law at the Wallace and Beverly Woodbury University at George Washington University.
The Innocence Project, a non-profit organization focused on overturning wrongful convictions through the use of DNA evidence, found 72 percent of the 317 wrongful convictions it has overturned involved mistaken eyewitness identifications.
So, the focus of experts, lawyers and advocates is on ensuring the processes police departments use to identify suspects promote the most accurate identifications possible.
The Case for Eyewitnesses
Not all eyewitness are the same though, former Washington homicide prosecutor Tad DiBiase said.
DiBiase insisted witnesses who know the suspect are much more reliable than witnesses identifying a stranger or someone with whom they have only passing familiarity. Fortunately, he believes these are the most common type of identification.
“People are going to believe an identification more if they know the two people knew each other,” DiBiase said. “That’s just common sense.”
The key for prosecutors, DiBiase said, is to not rely on just one eyewitness identification in a court case, especially if that eyewitness did not know the suspect before the crime.
“As a prosecutor, you’re always trying to corroborate what may be your weak evidence,” DiBiase said.
In some cases, this means trying to explain to the jury any mistakes in an identification. DiBiase said especially in Washington, some witnesses will make an improper identification due to fear or social pressure.
Prosecutors also need to know if the witness used drugs or alcohol before seeing the crime take place so they can determine if the witness is still reliable.
Police use a few different techniques to have eyewitnesses identify suspects, but generally the witness is asked to pick the person they saw commit the crime from a selection of other people, sometimes called foils.
This can either be done through a live lineup, which involves an eyewitness picking a suspect from a group of people standing in a room, or through a photo lineup, which is the same concept, but uses photos of suspects and foils rather than people who are physically present.
While this system is relatively standard, some experts worry about its ability to produce accurate eyewitness identifications.
One problem, said Dr. Jason Doll, chair of the Department of Forensic and Legal Psychology at Marymount University, is witnesses can look to the person administering the test for cues that tell the witness if he or she is selecting the right suspect. These cues can be subconscious, like a slight smile or movement from the administrator, and can affect the eyewitness’ choice, even if neither party is aware of the cue.
To fix this, groups like the Innocence Project insist the person administering the lineup or photo identification should not know who the suspect is, or if the suspect is even present.
Doll also said the administrator should let the eyewitness know this fact so the witness is not pressured into selecting somebody.
“They think they have to make a choice,” Doll said, adding that double blind tests let the witness have the option to say the suspect is not in the lineup or photo array.
The Innocence Project also advocates for the use of sequential lineups rather than the more traditional simultaneous method. Other experts agree, but Doll admitted this is perhaps the most controversial method for improving the accuracy of eyewitness identifications.
In sequential lineups, the eyewitness looks at one person or picture at a time and simply tells the administrator if the image represents the suspect. The administrator continues to show the witness one picture at a time until the witness makes an identification.
This doesn’t necessary increase the likelihood of a witness identifying a suspect, but it might reduce the chances of the witness choosing someone who did not commit the crime, Doll said.
After the identification is made, the administrator should ask the witness how confident he or she is in the choice, Doll said. This statement should be in the witness’ own words, not placed on a scale of one to 10 or expressed as a percentage, he added.
There is demand from certain scholars to allow the use of polygraphs to help determine if an eyewitness’ identification is accurate, but Saltzburg is convinced polygraphs are not the answer to the troubles of eyewitness identifications.
The problem, he said, is the polygraph cannot test if the identification is truthful, just that the eyewitness believes it to be so.
“There is no evidence that certainty correlates with accuracy,” Saltzburg said.
The Metropolitan Police Department did not respond to a request concerning its methods of obtaining eyewitness identifications. However, charging documents from the department often refer to the use of photo arrays in identifications.
“Memory is malleable”
There are some problems with eyewitness identifications over which police departments have little or no control. These problems are associated with the human mind, and how people store and recall memories.
“Most people are not very good observers,” Saltzburg said.
One issue that is intertwined with eyewitness identifications in criminal cases is the ways in which stress can impact human memory.
Michael Dougherty, Director of the Decision, Attention and Memory lab at the University of Maryland, said stressful situations can divide a person’s attention, and compared trying to remember details from such situations to a student studying for a test with a television on.
It might work for a short time, but as soon as the television draws that student’s attention away from her books, her ability to recall memories is decreased, Dougherty said.
Research shows this is especially true with crimes involving weapons.
People tend to focus on unique or out of place items, Doll said, and victims of crimes involving weapons often have difficulty remembering the face of the person who committed the crime because they were staring at the weapon.
Stress doesn’t affect everybody in the same way, however, and some studies have shown some stress is useful when recalling memories, Dougherty said. He described the relationship between stress and the accurate recollection of events as an inverted ‘U’ shape.
Eyewitnesses can also recognize a person in the lineup or photo array, but cannot determine where they have seen the person before.
Known as a source memory error, this means witnesses can mistakenly identify someone they simply saw walking on the street in the days leading up to the identification as the person they saw commit a crime, Dougherty said.
“The problem with humans — with brains in general — is we start incorporating other things into our memory over time,” Doll said.
Witnesses can also add details of a crime they have heard from other witnesses or from reports on the news to their memories, even if they did not see these details when they witnessed the crime.
“Memory is malleable,” Doll said.
Studies have also shown people can have a hard time distinguishing between people of a different race then their own.
In practice, this means that if a black witness sees a robbery committed by a white person, the witness might have difficulty identifying the correct suspect, Doll said.
Other circumstances, such as the time of day an eyewitness sees a crime occur or the lighting conditions in the area can have a negative impact on the witness’ ability to recall the identity of the criminal accurately.
“We can’t control the circumstances under which crimes are committed,” Saltzburg said.
The Justice System Still Needs Witnesses
All of these flaws don’t make eyewitness identification any less valuable to the justice system.
Saltzburg said many jurors come into the courtroom believing eyewitnesses to be infallible, in part because people think their memories would be reliable in a similar situation.
As a result, some jurisdictions allow judges to give jury instructions specifically regarding the accuracy of eyewitness testimony, Saltzburg said. Others also allow expert witnesses to take the stand to testify on the problems with memory and identifications to let the jury know the perils of trusting eyewitnesses outright.
“Largely, it comes down to better educating the juries when it comes to eyewitness testimonies,” said Paul Cates, communications director for the Innocence Project.
Not everybody agrees with this, however. DiBiase, the former prosecutor, compared allowing an expert to tell a jury eyewitness can be unreliable to allowing a lawyer to introduce statistics that say members of a certain race are more likely to commit crimes as evidence that a defendant of that race is guilty.
“You cannot always draw facts from the general and apply them to the specific,” DiBiase said.
Defense lawyers can also play a role in limiting the damage a mistaken identification can have on a criminal trial’s result.
Saltzburg recommended defense lawyers always ask an eyewitness on the stand to give his or her description of the suspect they saw commit the crime. If this description is very general, or if the defense can prove the witness was using drugs or alcohol when the crime occurred, jurors might give less weight to the identification.
In the end, eyewitness identifications will likely continue to be a pillar of the criminal justice system.
“We need people who see crimes and victims of crimes to tell us what happened,” Saltzburg said.
Through better jury education and improved police practices, experts and advocates hope to ensure eyewitness identifications are more accurate, less prone to mistakes and help the justice system punish criminals and protect the innocent.