Reliability of Eyewitness Testimony


Eyewitness testimony has been relied upon for decades by the police and the courts. It is used to determine details about crimes and guilt or innocence of suspected perpetrators. Unfortunately, it has been show that eyewitness testimony is not as accurate as it is portrayed to be, and this inaccuracy seems to be hidden from the public. This subject has been a major topic of study in the field of Psychology – Cattell, Binet, Ceci, Bruck and Stern have all contributed to research in this area. Many studies and theories   have shown that the human memory frequently fails and that memories are often distorted. The extent of the role that eyewitness testimony should play in our legal system is questionable.

Human memory consists of five stages, all of which play a large part in eyewitness testimony. In sequential order, these five stages are: 1) the perception/attention stage, in which one simply perceives a stimulus; 2) the encoding stage, in which one pays attention to details in the environment; 3) short-term memory, which has a limit of approximately two minutes; 4) long-term memory, where stored information can be accessed/retrieved as needed, and; 5) retrieval stage, in which an individual actively remembers information from their long-term memory. Memory can be affected in any of these five stages and, as a result, might not be passed on to the subsequent stage or be passed on inaccurately. For example, many details from short-term memory will not enter one’s long-term memory store and will be forgotten within a couple of minutes. Human memory can potentially make many mistakes and be incomplete, yet our continuous use of and dependence on eyewitness testimony in our legal system suggests otherwise.

It has been shown that one’s own pre-existing schemas and comprehension processes influence how one perceives reality. This can affect accuracy when distinguishing between what actually happened in a particular event versus what we think happened. Leading questions are often used by the police when discussing what happened with eyewitnesses. Ashcraft stated that leading questions “tend to suggest to the subject what answer is appropriate.” Studies have shown that such questions influence the subject’s memory of what happened and how they respond. The most popular study that illustrated this was performed by Elizabeth Loftus. A group of students were shown a video of a car crash. They were then asked how fast they believed the car was going when it either smashed into, collided with, bumped into, or contacted the other car. When stronger verbs were used (eg., smashed), the subjects estimated the car was moving at a higher speed. The way the question was asked led the subjects to biased answers. Leading questions often result in memory impairments, which Ashcraft defines as “genuine changes or alterations in memory of an experienced event as a function of some later event.” In the example of the Loftus study, many of the subjects who heard the word smashed also had a false memory of broken glass at the scene. Those who were told the other descriptive words did not recall broken glass. It is known that children are especially susceptible to suggestive questioning.

It is clear that the human memory is not fully accurate and this has been proven to be especially true in situations that require technical accuracy. Individuals often incorrectly claim to remember false information – this is called the misinformation effect. Not only is the information incorrect, but people are often very confident in remembering such events. If another person gives the subject false information which is believed, this is called misinformation acceptance. Ashcraft defines this as when people “accept additional information as having been part of an earlier experience without actually remembering that information.” People also have ‘false memories’ which are simply memories of something that did not happen. Another downfall of human memory that causes issues within the legal system is called source misattribution, which is a confusion in memory where one cannot remember the source of information.

Another flaw in memory that affects eyewitness testimony is the hindsight bias. Weitnen & McCann define this as “the tendency to mould our interpretation of the past to fit how events actually turned out.” People believe that they ‘knew it all along’ when explaining events that would actually be difficult to foresee. Many people are overly confident when identifying a perpetrator in a lineup, even when the perpetrator is not present. Lineup identification is one of the most common methods of identifying a culprit. Unfortunately, many of these lineups are considered unfair. A fair lineup is one in which the suspect does not stand out from the foils/distractors (people who are known to be innocent). There are three main types of bias when forming a lineup: the foil bias is when the suspect is the only member in the lineup who matches the description of the culprit, the clothing bias is when the suspect is the only lineup member wearing similar clothing to that worn by the culprit, and the instruction bias, which involves the police failing to inform the witness that the culprit may not be present. Photo arrays are more commonly used than live lineups and multiple identifications increase the likelihood that the selection is correct. Voice identification is also used – accuracy is higher with longer voice samples and whispering significantly decreases accuracy.

Confidence tends to raise even higher when subjects are told they chose the right suspect. Also, it has been shown that jurors tend to believe confident eyewitnesses moreso than witnesses that appear less confident. In reality, there is only a low correlation between confidence and accuracy of recall. It is possible that such confidence is fueled by the failure to seek disconfirming evidence – the individual cannot think of any reason why they would be wrong, do they assume they must be right. Studies also show that witnesses remember faces of their own race with greater accuracy than those of other races, which is known as the cross-race effect. One is less likely to remember details about the culprit and environment if there is a weapon present. A weapon is considered unusual and threatening. As a result, an individual will focus more on the weapon than on other factors.

Police interrogation is often a procedure aimed at obtaining the answers the police want rather than what may be true and complete. They use many techniques to lead eyewitnesses to say only what is wanted. For example, police officers often interrupt eyewitnesses when they are re-telling their experiences. The police often limit the information provided by the witness by distracting them (eg., asking questions) and preventing them from speaking. Police tend to ask short, specific and sometimes irrelevant questions. They may also ask questions in a random order that is inconsistent with the information witnesses are providing at the time. Police often evoke emotional arousal in eyewitnesses, which has been proven to affect the accuracy of testimony – as emotional arousal increases, attentional capacity decreases. Eyewitnesses are usually asked to testify long after the incident took place (often six months later or more). This will have an obvious affect on one’s memory. The graph below illustrates the ‘forgetting curve,’ which is the decrease in memory capacity as time elapses.


Pozzulo, Bennell and Forth outlined four recommendations to increase the accuracy of lineup or photo array identification. The first states that the person who conducts the lineup or photo array should not know which person is the suspect – this reduces the potential for bias. Secondly, eyewitnesses should be told explicitly that the criminal may not be present in the lineup and, therefore, witnesses should not feel that they must make an identification. The third recommendation states that the suspect should not stand out in the lineup as being different from the foils/distractors based on the eyewitness’ previous description of the criminal or other factors that would draw extra attention to the suspect. Finally, a clear statement should be taken from the witness at the time of the identification and prior to feedback on his or her confidence on the selection made. It is possible that abiding by these recommendations would decrease the number of wrongful convictions made.

It has been well-demonstrated by researchers that fundamental errors are being made when using eyewitness testimony in a legal context. It seems that we are guilty of overestimating identification accuracy, not fully understanding the influence of situational factors and ignoring system variables that may alter accuracy. The public needs to be more informed about how our legal system works and it’s limitations. Eyewitness testimony is commonly used and relied upon in determining guilt or innocence, despite it’s risk of inaccuracy. It is time for this to be re-visited and modified in order to prevent innocent people being locked away and potentially dangerous criminals roaming free.

2 thoughts on “Reliability of Eyewitness Testimony

  1. Tom Regehr

    These are important ideas, they effect – or should effect – everything we do. How do we decide what we are going to do daily? React to anything? The answer of ‘I no not know’, seems so ideal, so often.

Leave a Reply to Tom Regehr Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s