Why Eyewitnesses Can Be Wrong


By Catherine A. SandersonAmherst College

One of the most influential pieces of information police officers can have in trying to solve a crime has traditionally been an eyewitness—a person who has first-hand information about the crime being committed. In some cases, this person was a bystander who saw the crime take place. In other cases, the eyewitness is actually the victim of the crime.

A chemist reading a DNA profile to determine the criminal
Wrong convictions were noticeably reduced thanks to DNA evidence. (Image: James Tourtellotte/Public domain)

Eyewitnesses Are Not Always Right

Eyewitnesses can give police officers information about exactly what happened and describe the suspect. They may later be asked to identify the suspect in a photo or line-up. They may also appear during a trial to describe what they saw. Eyewitnesses who can confidently identify the suspect and confirm that they saw them commit a crime are very convincing and jurors tend to believe them. Unfortunately, eyewitnesses can also be wrong.

One survey of prosecutors across the United States published in 1989 found that an estimated 77,000 people each year had been charged with committing a crime based solely on reports of an eyewitness. Although there had been a few scattered cases over the years of confirmed eyewitness misidentifications, the prevalence of such errors only became widely known in the 1980s, as DNA evidence started to become widely used in criminal investigations.

Memory errors can occur during encoding of a memory, storage of a memory, or retrieval of a memory. All of these errors occur at an unconscious level—witnesses really believe their memories are accurate, even when they aren’t—and confident witnesses are especially convincing, even though there’s often no correlation between confidence and accuracy.

How Perceptions Can Be Wrongly Affected

Encoding a memory is the initial perception of an event. In many cases when an eyewitness observes a crime occurring, the conditions for accurately encoding what is happening just are not present. Crimes may occur at night or in dim light, meaning it’s literally hard to see what’s going on. Crimes often occur quickly, so there isn’t much time to form perceptions. 

A depiction showing the weapon-focus effect on face or other features
Witnesses or victims tend to concentrate on the criminal’s weapon, so they don’t focus on other details. (Image: Jnasco/Public domain)

The person committing the crime may have disguised their appearance in some way. Crimes also occur unexpectedly, meaning we aren’t primed to pay close attention in advance. And watching a crime being committed often creates arousal and anxiety, which can further disrupt our ability to pay close attention to various details.

In particular, the weapon-focus effect describes people’s natural tendency to direct their attention towards a weapon, if one is present. If the person committing a crime is holding a knife or gun, eyewitnesses’ intently focus on that weapon, meaning they are not focusing on aspects of the focusing on aspects of the person’s face or other features that would help with later identification of a suspect.

This article comes directly from content in the video series Introduction to PsychologyWatch it now, on Wondrium.

The Cross-race Bias

Another factor that leads to eyewitness errors is the cross-race identification bias, meaning people’s relative difficulty in identifying a person who is not of their race.

In a 1988 study that was one of the first demonstrations of this bias, convenience-store clerks in Texas were asked to identify one of three customers who had stopped by earlier in the day and made a purchase. One customer was white, one was black, and one was Latino. Clerks were most accurate at identifying people of their own racial or ethnic group.

People who are in the majority population group are more likely to misidentify people than those in minority population groups, presumably at least in part because minority group members have greater exposure to the faces of majority group members than the reverse. In cases of convictions made based on false identifications that were overturned following DNA evidence, 42% involved cross-race errors, typically with a white person misidentifying a black person.

A Real Case of Misidentification

In some cases, this misidentification can occur even when the eyewitnesses are themselves the victim and have relatively extensive contact with the person who committed the crime. In 1984, a woman named Jennifer Thompson was raped in her home in North Carolina. She immediately reported the assault to the police and worked with a sketch artist to create a picture of his face. 

The police then showed her a series of mug shots of men with similar features. She picked out one of these photos as showing the man who assaulted her, a man named Ronald Cotton, whose photo was in the book of mug shots due to a robbery he had committed several years before. But she wasn’t immediately sure it was him; she hesitated for several minutes, and then said, “I think this is the guy. This looks the most like him.” Jennifer Thompson was white; Ronald Cotton was Black. 

A white woman and a black man sitting on chairs on stage.
Getting under the cross-race identification bias effect, Jennifer Thompson confessed that Ronald Cotton was her rapist. (Image: PopTech/Public domain)

Over time, Jennifer’s confidence in her identification grew. By the time of the trial a year later, she confidently pointed to Ronald Cotton as the man who had raped her. He was sentenced to life in prison based on her testimony. Ten years later, DNA technology revealed a match between samples taken from Jennifer’s body and a different man, who also confessed to the assault. Cotton was set free.

What explains the difficulty in identifying people from a different racial group? One explanation is that people generally spend more time with members of their own race, and that leads them to be more attuned to the distinct characteristics of people who are similar to themselves. Conversely, we may tend to see out-group members as “all the same.”

So, one problem with eyewitness identification occurs at the level of what we encode. But even if encoding goes well, and we do pay close attention to the details of the crime when it initially occurs, memories can change while they are in storage, impairing the accuracy of eyewitnesses.

Common Questions about Why Eyewitnesses Can Be Wrong

Q: What is an eyewitness?

An eyewitness is a person with first-hand information about a specific crime. An eyewitness can be a bystander or, in some cases, the victim of the crime itself.

Q: How often do eyewitnesses identify the wrong person?

According to one survey published in 1989, around 77,000 people each year in the United States had been charged with committing a crime based only on eyewitnesses misidentifications. Fortunately, as DNA evidence became widely used, the prevalence of these errors became widely known.

Q: Why is encoding a memory important for witnesses?

The initial perception of an event is called encoding a memory in which a person needs to be in a special condition to form a perception. Many of the witnesses’ reports can be wrong because the crimes often happen quickly or at night, and the witnesses don’t get much time to encode what is happening and to form accurate perceptions.

Keep Reading
Errors in Problem-solving: Jinx, Confirmation Bias, and Instincts
Problem-solving and Mental Shortcuts
Groupthink and the Error of Overconfidence