Lecture 8 - Chapter 6: Change Blindness and Eye Witnesses

The Complex Nature of Eyewitness Perception

Seeing is not the same as perceiving. While the human eye can capture vast amounts of visual information, our brains filter, prioritize, and sometimes alter this information based on where our attention is directed. This means that eyewitnesses, including law enforcement officers, are not passive recorders of events. Instead, they actively construct memories based on incomplete and sometimes misleading sensory input.

To understand why eyewitness testimony can be unreliable, we must examine the cognitive processes behind change blindness and inattentional blindness and how they apply to real-world situations, particularly in criminal cases.

1. Change Blindness: Failing to Detect Alterations in the Visual Scene

Definition

Change blindness occurs when a person fails to notice a substantial change in their visual environment. This phenomenon is particularly evident in cases where changes happen gradually or when an observer’s attention is momentarily diverted.

Classic Experiment (Levin & Simons, 1997)

• In this study, participants watched a video where a man answered a phone call after leaving a room.

• However, the man who returned to the scene was played by a different actor.

• Despite the fact that the change involved a central figure, only 33% of participants noticed the switch.

Implications for Eyewitness Testimony

• A witness may confidently testify that they saw a particular individual commit a crime when, in reality, they may have confused that person with someone else who appeared later.

• This is particularly dangerous in cases where multiple individuals resemble one another (e.g., wearing similar clothing, sharing the same race or hairstyle).

• In the case of the police officer arresting the man in the blue hooded sweatshirt, the officer may have unknowingly misidentified the suspect due to change blindness—failing to realize that the original brick-thrower was not the same individual who later emerged from behind the building.

2. Inattentional Blindness: Failing to See What’s Right in Front of You

Definition

Inattentional blindness occurs when a person completely fails to notice an object or event in their field of vision because their attention is focused elsewhere. Unlike change blindness (where the observer does not detect a change), inattentional blindness involves not seeing something that is present at all.

Classic Experiment (Simons & Chabris, 1999 - The Invisible Gorilla Study)

• Participants watched a video where six people passed two basketballs while moving around.

• They were instructed to count the number of times the white-shirted players passed the ball.

• Many failed to notice a person in a gorilla costume walking into the middle of the group, beating their chest, and leaving.

• When asked afterward, many participants were shocked to learn that they had missed such an obvious event.

Implications for Eyewitness Testimony

• If people can overlook a person in a gorilla suit, they can also fail to notice critical details at a crime scene.

• A witness who focuses on one detail (e.g., the suspect's clothing) may fail to notice other important elements (e.g., another suspect, a weapon, an alibi).

• This means that just because a witness did not see something, it does not mean it wasn’t there—raising questions about how much weight we should place on what eyewitnesses claim to have seen.

Real-World Consequences: Eyewitness Errors in the Legal System

Mistaken Identity and Wrongful Convictions

Research shows that eyewitness misidentifications are one of the leading causes of wrongful convictions. Organizations like the Innocence Project have revealed that nearly 70% of wrongful convictions overturned by DNA evidence involved eyewitness errors.

Case Example: The Wrongful Conviction of Ronald Cotton

• In 1984, Jennifer Thompson was attacked and later identified Ronald Cotton as her attacker.

• However, DNA evidence later proved his innocence—he had been mistakenly identified due to faulty eyewitness testimony.

• This case illustrates how eyewitness confidence does not equal accuracy, and how cognitive biases (such as change blindness and inattentional blindness) can lead to devastating errors in justice.

Why Jurors Overestimate Eyewitness Reliability

Even though cognitive psychology research has consistently shown that eyewitness testimony is unreliable, jurors still tend to trust it. This is due to several psychological factors:

1. The Persuasive Power of Personal Testimony

   • Jurors find eyewitness accounts compelling because they feel like firsthand experiences.

   • However, false memories feel just as real as true ones, making it difficult to distinguish between accurate and inaccurate recollections.

2. The Confidence-Accuracy Illusion

   • Studies show that people tend to assume that a confident witness is an accurate witness.

   • However, confidence and accuracy are not strongly correlated—witnesses can be completely confident in their testimony while still being entirely wrong.

3. The "What You See Is All There Is" Bias

   • People assume that if something was important, they would have seen it.

   • Research on inattentional blindness contradicts this assumption—we fail to see things all the time, even in plain sight.

How Courts Are Adapting to Cognitive Science Findings

1. Expert Testimony on Eyewitness Limitations

Some courts now allow expert witnesses (such as cognitive psychologists) to testify about change blindness, inattentional blindness, and memory distortions to help jurors understand the limitations of eyewitnesses.

2. Improving Police Lineup Procedures

• Double-blind lineups: The officer administering the lineup does not know the suspect, preventing subtle bias from influencing the witness.

• Sequential lineups: Witnesses view one person at a time rather than all suspects at once, reducing comparative judgments that can lead to misidentifications.

3. Using Body Cameras and Video Evidence

Since human memory is flawed, objective recordings (such as body cam footage) provide a more reliable alternative to eyewitness accounts.

Theoretical Background

Several theories have been proposed to explain change blindness, including:

• Feature Integration Theory (Treisman & Gelade, 1980): Suggests that attention is required to bind different visual features into a coherent perception. When attention is disrupted, changes in objects may go unnoticed.

• Overwriting Hypothesis: Proposes that new visual input can overwrite previous memories, leading to failures in detecting changes.

• Failure to Compare Hypothesis: Suggests that individuals do not actively compare current and past visual representations, resulting in change blindness.

• Coherence Theory (Rensink, 2002): Argues that visual scenes are perceived as coherent wholes only when attention is directed toward specific elements. Without focused attention, changes are unlikely to be noticed.

Unconscious Transference and Inattentional Blindness

Closely related to change blindness, unconscious transference occurs when witnesses mistakenly identify a familiar but innocent person as the perpetrator of a crime. This error arises because of the misattribution of familiarity to the wrong context. Similarly, inattentional blindness refers to a failure to notice unexpected objects or events in a scene when attention is focused elsewhere.

Research on inattentional blindness has demonstrated that even highly salient stimuli can go unnoticed if individuals are not actively attending to them. This finding has serious implications for eyewitness testimony, as it suggests that witnesses may fail to perceive critical details of a crime, even if they are looking directly at the scene.

Cross-Paradigm Research

Despite the clear connections between change blindness and eyewitness testimony, there has been limited research that explicitly combines these two fields. Only a handful of studies have explored the implications of change blindness for eyewitness accuracy.

Davies and Hine (2007)

This study aimed to merge change blindness and eyewitness testimony research by having 80 participants watch a video of a burglary in which the identity of the burglar changed halfway through. Participants were divided into two conditions:

• Intentional condition: Participants were informed they would be tested on the video later.

• Incidental condition: Participants were told they were watching a safety video.

Findings:

• 39% of participants noticed the change in actors, with most of them (26 of 31) in the intentional condition.

• Participants who detected the change performed significantly better on the lineup identification task.

• Those in the incidental condition had significantly poorer memory accuracy, reflecting real-world crime scenarios where witnesses do not anticipate needing to recall details.

Davis et al. (2008)

This study explored unconscious transference as a form of change blindness, examining whether witnesses falsely identified innocent bystanders as perpetrators due to an illusion of continuity. Participants watched a video of a grocery store theft where an individual was replaced mid-action by another similar-looking person.

Findings:

• 60–68% of participants failed to notice the identity change.

• When distracted by a secondary task, only 20% detected the change, compared to 46% who were not distracted.

• 71–75% of participants made false identifications in a perpetrator-absent lineup.

• Those who failed to notice the change were more likely to misidentify an innocent bystander as the perpetrator.

This study demonstrates that inattentional and change blindness can lead to unconscious transference, where a witness mistakenly attributes a crime to a bystander rather than the actual perpetrator.

Nelson et al. (2009)

This study examined change blindness and unconscious transference using a crime severity manipulation. Over 700 participants watched one of four versions of a video where a student stole either $5 or $500. A change occurred mid-scene where the thief was replaced by an innocent bystander.

Findings:

• Only 4.5% of participants noticed the change in actors.

• Those in the experimental condition (exposed to the identity change) were significantly less likely to correctly identify the perpetrator (36%) compared to the control group (64%).

• Among experimental participants, the innocent bystander was misidentified at the same rate (35%) as the actual perpetrator (36%).

• Crime severity had a minor effect, with participants being slightly more accurate in identifying the perpetrator when the stolen amount was higher ($500 vs. $5).

This study provides further evidence that change blindness can lead to mistaken eyewitness identifications, even when individuals are aware that a crime has occurred.

Implications for Eyewitness Testimony

The research reviewed above highlights the need for caution in relying on eyewitness identifications in criminal cases. Several key takeaways emerge:

• Witnesses may fail to detect changes in perpetrators due to inattentional and change blindness.

• Innocent bystanders are at significant risk of being falsely identified, particularly if they resemble the actual perpetrator.

• Expectation of a memory test (as in the intentional condition) improves recall, but real-world crimes rarely allow for such preparation.

• Distraction (such as focusing on store items in Davis et al.'s study) impairs attention, leading to increased errors in identification.

Future Research Directions

Despite the insights gained from these studies, much remains unknown about the intersection of change blindness and eyewitness testimony. Areas for future research include:

• The influence of witness and perpetrator characteristics (e.g., age, race, disguises) on change blindness.

• The impact of environmental factors, such as stress or poor lighting, on eyewitness accuracy.

• The relationship between change blindness and the misinformation effect, which occurs when post-event information distorts memory.

• The role of inattentional blindness in crime scenarios, particularly in fast-paced or unexpected events.