LEWISTON — Eyewitness identifications of criminal defendants are often unreliable and can be influenced by many factors, Bates College Professor Amy Bradfield Douglass said Thursday.

Douglass, who holds a doctorate in psychology and has appeared as an expert witness in state and federal courtrooms throughout New England, was speaking Thursday in Callahan Hall at Lewiston Public Library as part of the Great Falls Forum lecture series.

She has studied how eyewitnesses to crimes make incorrect identifications and has researched reforms aimed at making those identifications more reliable. Her work is supported by the National Science Foundation and the Australian Research Council.

According to a 1993 study by a leading researcher in the field, more than 77,000 people become criminal defendants each year because they have been identified from a lineup or photo spread of possible suspects.

Circumstantial evidence, such as fingerprints, can be presented at trials by prosecutors to help prove a fact. But an eyewitness account, called direct evidence, “is really strong and powerful,” she said.

Just because a person’s fingerprints are at the scene of a crime doesn’t necessarily mean that person committed the crime, Douglass said. But juries find eyewitness testimony compelling.

“If you yourself have served on a jury, you know the experience of having the witness say, ‘I absolutely know that was him.’ That’s a really powerful moment for a jury to think about whether the defendant is guilty,” Douglass said.

Breakthroughs in science and technology have helped to exonerate wrongly convicted defendants, she said. DNA testing has helped free hundreds of prisoners found guilty of older crimes. But DNA is not a solution to wrongful identifications, Douglass said.

A prominent case of wrongful identification featured a rape victim who twice identified the wrong man as her attacker. The woman said she had studied the physical traits of her assailant during the assault in an effort to be able to later identify him to authorities.

“She was like a perfect victim because she was really paying close attention to what he looked like,” Douglass said.

The woman had twice incorrectly picked her assailant from photo spreads.

DNA evidence eventually proved the man she had twice identified as her attacker was a case of mistaken identity after two trials. Even when shown the wrongfully accused man next to the actual assailant, the victim had picked the wrong man because her memory had, over time, supplanted the face of the real attacker with the face of an innocent man, Douglass said.

The woman later reached out to the man she wrongly accused and the two ended up writing a book about the ordeal. Today, both are strong advocates for lineup reform, Douglass said.

But in many crimes, there is no DNA evidence collected at the scene. That’s one reason eyewitness identification is so important, Douglass said. Prosecution of all types of crime should stand or fall on reliable evidence, including physical evidence and witness identifications.

Just as with DNA, blood and other forms of physical evidence, eyewitness identifications should be carefully preserved and protected from contamination, she said.

Understanding how human memory works also is important, Douglass said.

Unlike a video camera, our perceptions often leave out crucial information and may be compromised by optical illusions or by focusing on a single element of what we see, she said.

Lineups and photo spreads must be presented to a witness in a way that isn’t biased, Douglass said. Authorities should tell the witness that the suspect may or may not be pictured in a photo spread or standing in a lineup, she said. And no prompts should be given by the person administering the lineup or photo spread.

If the witness has described the suspect as having certain traits, then all of the people pictured in the lineup or photo spread should possess those traits, Douglass said. And the photos should be shown one at a time.

Law enforcement officers should refrain from reacting to witness identifications, Douglass said, because “sometimes these witnesses are totally confident and totally wrong.”

The person showing the photo spread to the victim shouldn’t know who the suspect is because they may, even without knowing it, act suggestively through subtle verbal language or movements, Douglass said.

Newer recommendations in eyewitness identifications include videotaping all lineup and photo spread procedures, Douglass said. Authorities also are encouraged to establish a reasonable suspicion standard before placing a suspect in a lineup. A third — and more controversial — suggestion is to only allow at trial “pristine confidence” statements by eyewitnesses.

[email protected]

Bates College psychology professor Amy Bradfield Douglass discusses an image that can be seen in two ways during the Great Falls Forum at the Lewiston Public Library on Thursday. 

Bates College psychology professor Amy Bradfield Douglass discusses how eyewitnesses to crimes sometimes make incorrect identifications. She spoke at the Great Falls Forum at the Lewiston Public Library on Thursday. 


Only subscribers are eligible to post comments. Please subscribe or to participate in the conversation. Here’s why.

Use the form below to reset your password. When you've submitted your account email, we will send an email with a reset code.

filed under: