Eyewitness Testimony: What Happens When You Do Not See What You Thought You Saw

By: Carla Vestal

Earlier this year, there was a dress that had everyone on the internet up in arms, and no, it was not something a reality television person was wearing. The most talked about dress this year was from the online retailer Roman Originals and the question asked was, “What color is the dress?” Some people looked at the image and saw gold and white. Others looked at the dress and saw black and blue. When a picture of the dress was placed in Photoshop, the picture-altering program labeled it as both, white and gold and black and blue! Personally, the picture entertained me for a few hours because sometimes I saw blue and black and other times the alternate. I would look at the picture seeing white and gold, close my eyes, count to three, reopen my eyes and see black and blue.

Other than this image being a cool optical illusion online, I began to question, “How much should you trust your own eyesight?” Or, more importantly, “How much weight should be placed on eyewitness testimony in criminal trials?” Black’s Law Dictionary defines eyewitness testimony as “a statement given under oath by a witness.” Simple enough. But, what if our eyes, like in viewing the now most famous dress of this decade, aren’t as trustworthy as we think they are?

How long have our eyes been playing tricks on us?

Critics have questioned eyewitness testimony as early as the 1900s, and major criminal defense advocates, such as the Innocence Project, claim that “eyewitness misidentification is the greatest contributing factor to wrongful convictions, playing a role in about 75% of convictions overturned through DNA testing nationwide.” The raw numbers indicate some other troubling data. The most current data on wrongful convictions cite that, of the 318 cases overturned, 72% were overturned due to bogus eyewitness testimony. That is an astounding 229 innocent people who have been sentenced to prison because of what later turned out to be erroneous eyewitnesses. And that is not all that happened because of this faulty testimony. While the innocent were serving sentences for crimes they did not commit, the real perpetrators, who have been identified in 90 of these cases, were out committing more crimes: 63 rapes, 17 murders, and 18 other violent crimes.

Do people of different races all look alike?

Our perceptions of people who look differently than we do is called the “cross-race effect.” The cross-race effect asserts that people of the same race are better able to recognize faces of the same race than the faces of people who are of a different race. In a study by Behrman and Davey in 2001, 271 police cases were analyzed and the results showed that people of the same race correctly identified the suspect 65% of the time, whereas, only 45% of the witnesses were able to correctly identify the suspect if the suspect were of another race.[1] The theory behind this disparity is, regardless of race, people share an ingroup advantage with members of the same race. This ingroup advantage allows people of the same race to process facial features of their own group in a more holistic manner than when looking at a person of a different race. When observing a person of a different race, an individual recognizes characteristics that are attributable to that race as a whole. The consequence of this cross-race effect causes disparity when people witness a person of another race committing a crime, because it is more likely that the witness will misidentify the suspect due to the decreased ability of different races to identify the facial characteristics of another race’s face.

The brain does not remember what the eyes see

Vision is composed of three distinct, yet highly interdependent phases: sensation, attention, and perception. The first step of sensation is detecting light and basic shapes. Sensations are ephemeral and a very small amount of what is sensed is perceived. Attention is the process used by the visual system to send selected information that has been sensed for further processing. Perception is when the attended sensations are linked to “environmental cause, made coherent, and categorized through the assignment of meaning, utility, value, and emotional valence.” Memories and emotions from prior experiences also interweave within these new images, adding value and meaning to the individual.[2]

When analyzing how vision works, it is equally important to consider how vision works in relation to distance. People who identified as having “good eyesight” could not see the eyelashes on a person from ten feet away. When 200 feet away, the eyes themselves blur, and at 500 feet, no recognizable facial features can be ascertained. Other factors that have been found to affect the witness include the amount of stress involved, weather conditions, noise levels, and if a weapon had been used.

The memory itself is not stored in one part of the brain, but is recalled through the activation of several parts of the brain that work together. For example, the emotion that a memory brings is stored in one area of the brain, and the content of the memory is stored in another place. Hormones help engrave memories in our brains and then those same hormones categorize new experiences based on old ones. Adding to the matrix of how vision and memory work, “there are between 200 and 400 billion neurons in the brain and each neuron has about 10,000 connections.” Nobel Prize winning neuroscientist Gerald Edelman, Ph.D., sums memory up as the ability to repeat a performance – with mistakes.

How mistakes are added into a memory and how false memories are made

Memory is malleable, so malleable in fact, that studies have shown in studies that memories can be recalled by 25% of people who have never had the experience by mere suggestion and non-verbal clues given by others. Studies show that witnesses will pick the wrong suspect out of a line-up 25% of the time, and when the suspect is absent from the line-up the witnesses will pick an innocent person more than 33% of the time. Witnesses pick-up on nonverbal communication from the officers conducting the line-up when the officer knows the suspect is present, regardless of whether it is the correct person or not. Thus, this solidifies the witness’s choice and the witness becomes more confident in the identification. As litigators know, a confident eyewitness on the stand, regardless of whether the witness is correct or not, has a great impact on a jury.

The court and science come together

Over the past 30 years, there have been hundreds of sociological studies performed to test the reliability of eyewitness testimony and to determine what factors, if any, influence the brain’s recall system and the witness’s self-affirmance of what she or he may have experienced. The New Jersey Supreme Court in 2011 made a landmark decision where it reviewed a special master record of 300 exhibits and 200 scientific studies in the case of State of New Jersey v. Henderson. The court decided that when studying eyewitness testimony,2 groups should be identified: “estimator variables” and “system variables.”

Estimator variables are things that cannot be controlled by the criminal justice system. Examples of estimator variables are the amount of light between the alleged and the distance between the witness and alleged perpetrator. Things that the criminal justice system can and should control for are system variables. System variables include procedures such as how law enforcement retrieves and stores witness testimony, how it conducts line-ups, photo arrays (six-packs), and instructions given to witnesses when participating in identification procedures. State of New Jersey v. Henderson, 27 A.3d 872 (N.J. 2011).

What can be done to improve the system used?

From the scientific evidence that supports the fallibility of eyewitness testimony and the decision rendered in Henderson, there are several methods that can be used to ensure that innocent people are not identified through eyewitnesses. For example, blind administration of the officer conducting line-up, where the officer conducting the procedure does not know who the alleged suspect is, would greatly reduce the non-verbal communication signals the witness may interpret as positive feedback. Also, when using fillers in line-up composition, the fillers should resemble the description on the person given by the witness, and the same fillers should not be used in multiple line-ups. This would reduce the likelihood that seeing the same person multiple times would implant a false recognition in the eyewitness’s memory. Other improvements would allow for the identification process to be recorded; eyewitnesses would memorialize their own confidence level in the identification made; and the conducting officer would tell the witness that the suspect may or may not be in the line-up, and the investigation will continue regardless of the outcome. During the investigation process, implementation of these procedures would lessen misidentification. The lack of these processes being used gives the defense an opportunity to attack the eyewitness’s identification during a trial. If the procedures were implemented nationwide they would protect innocent people from overzealous police and prosecutors, thus helping to ensure that a suspect will have a fair trial.

But what color is the dress?

The designers have confirmed the dress is blue and black.

To learn more about eyewitness identification, vision processing and how the brain stores memories visit:





[1] B. W. Behrman & S. L. Davey, Eyewitness Identification in Actual Criminal Cases: An Archival Analysis, 25 Law & Hum. Behav. 475 (2001).

[2] Comm. on Scientific Approaches to Understanding and Maximizing the Validity and Reliability of Eyewitness Identification in Law Enforcement and the Court et al., Identifying the Culprit: Accessing Eyewitness Identification 46-47 (2014).

Comments are closed.

%d bloggers like this: