Eyewitness Testimony

Eyewitness testimony refers to the report of an event provided by someone who observed this event and often involves the recollection of details surrounding a witnessed crime as well as identification of participating offenders. Frequently, eyewitness testimony is the most compelling source of evidence in criminal trials and may be the only evidentiary support that links a suspect to the crime committed. Research shows that statements presented by eyewitnesses are influenced by a number of different factors that extend beyond the mere recollection of the event. This is of particular relevance given that inaccurate eyewitness testimony is the leading cause of false convictions in the United States. This article introduces the role of eyewitness testimony in criminal cases, discusses variables that can influence the accuracy of eyewitness testimony, and reviews both current and proposed methods for mitigating potential biases that arise.

Role and Influence in Criminal Cases

Eyewitness testimony remains an important source of evidence in many criminal cases, even with the availability of more technologically advanced sources such as forensic DNA testing. This is in large part due to the wider availability of eyewitness testimony over sources of DNA when gathering evidence in criminal cases. Still, the technological advances in forensic DNA testing have helped to inform the role and limitations of eyewitness testimony, notably in the exoneration of persons convicted of crimes whose innocence is ultimately demonstrated through DNA analysis, a technology that was not widely available in the recent past. The Innocence Project website has documented 344 such cases, with inaccurate eyewitness testimony influencing over 70% of these wrongful convictions. Indeed, errors in eyewitness testimony remain the number one cause of wrongful convictions, with an estimated one third of eyewitnesses making an inaccurate identification.

Inherent in establishing eyewitness testimony as the leading cause of wrongful convictions is the court’s reliance on inaccurate evidence in rendering a verdict, which is influenced by a number of factors. First, judges may have limited understanding of the factors that influence eyewitness testimony and frequently also misjudge the knowledge that jurors may have on the subject. This in turn limits the utilization of an eyewitness expert and prevents the orientation of jurors to the factors that influence the accuracy of eyewitness testimony. This gap in knowledge surrounding eyewitness testimony may prevent the court from making an accurate assessment of a witness’s statement. Indeed, eyewitness confidence, a poor predictor of eyewitness accuracy, is often emphasized over other factors.

Factors Influencing Eyewitness Testimony

There are multiple factors that influence the accuracy of an eyewitness statement. First, one should consider the complex process of memory formation. With very few exceptions, individuals are unable to register and recall an event with the precision and accuracy of a recording device. In fact, memory formation and recall consist of multiple stages recognized to be dynamic, malleable, and susceptible to influence. Perception, the observation of an event, can be biased based on the particular aspects to which an individual pays attention. Encoding, the transfer of the observed event into a memory, can be subject to interpretation as well as biases that are intrinsic (e.g., knowledge base, prior experiences) or extrinsic (e.g., the presence of a weapon or a disguise) to the individual. Storage refers to the preservation of the encoded memory and is susceptible to both modification and deterioration over time. Retrieval represents the recreation of a memory, which is vulnerable to unconscious modifications in order to fill in gaps that may be present in the encoded memory.

Other factors that influence the accuracy of eyewitness testimony can be divided into two broad categories. System variables refer to factors that affect eyewitness testimony and are controlled by the criminal justice system. Conversely, estimator variables are factors that are inherent within the witnessed event. The criminal justice system can estimate the significance in relation to eyewitness testimony, but estimator variables are beyond its direct control. Taken collectively, estimator and system variables may contribute to an erroneous identification of a suspect, with an effect that is likely additive.

Estimator Variables

Characteristics of the witness prior to observing the event in question are an important consideration when evaluating the accuracy of eyewitness testimony. Among these characteristics are an eyewitness’s age, with children less than 10 years of age and individuals more than 70 years more likely to make erroneous identifications than the general population. Additionally, individuals tend to have less accurate facial recognition with suspects who are outside of their own ethnic group, a phenomenon described as the own group bias or cross-race effect. In the same vein, an individual tends to recognize faces with familiar physical characteristics, which can erroneously influence an eyewitness’s identification if that individual mistakes the context of this person’s familiarity and wrongfully associates it with the memory of the perpetrator.

Other estimator variables that should be considered include characteristics of the event in question as well as the mental state of the witness during said event. For example, the distance between the witness and perpetrator is inversely correlated to an accurate identification. Similarly, lower levels of illumination and shorter exposure to the facial features of the perpetrator are also inversely correlated. Additionally, if perpetrators wore disguises or changed their appearance by altering their head or facial hair, a witness may struggle with an accurate identification.

If drugs or alcohol clouded the sensorium of a witness at the time of the event, the ability to make an accurate identification may be impaired. These effects have been demonstrated specifically for alcohol and marijuana intoxication. A final consideration that can influence eyewitness accuracy is the perceived stress of the event in question. When the perceived threat is minor, the associated stress may enhance one’s memory of the event and alleged perpetrator. Conversely, situations that are perceived as highly threatening activate the fight-or-flight response in a witness, which may impair recognition and recall. This is particularly relevant when the perpetrator wields a weapon and potentially impairs a witness’s memory, a phenomenon described as the weapon focus effect.

System Variables

An inherent concept of system variables is that they occur following the alleged event. They include both eyewitness interviews and suspect identification by lineup conducted by law enforcement officers, both of which have the potential to introduce bias if not conducted optimally. For example, interviews conducted during an investigation frequently rely on close-ended questions, which not only may limit the amount of information gleaned from an eyewitness but also have the potential to unknowingly introduce bias given the suggestive nature of some close-ended questions. Additionally, the content provided by an eyewitness during an interview can be influenced by exposure to additional information regarding the event received from other individuals, the news media, or even law enforcement. As memory formation and maintenance are malleable, such postevent information may unknowingly influence what the witness recalls.

The process whereby a witness identifies an alleged suspect is typically a lineup conducted by law enforcement, which has the potential to introduce bias in multiple ways. For instance, the number of individuals participating in a lineup affects the fairness of suspect identification. An insufficient number of participants increases the likelihood that a suspect will be selected by random chance rather than by accurate identification. Similarly, the composition of the lineup may also sway the eyewitness. For this reason, a suspect should not be obviously distinct from other participants in terms of appearance, attire, and physical positioning in relation to the eyewitness.

The verbal interactions between an eyewitness and law enforcement immediately before and after the lineup can also constitute a source of bias. Specifically, if a witness is not advised that the alleged perpetrator may not be present in the current lineup, then that witness has a significantly higher chance of making a false identification. The relative judgment process refers to the tendency of a witness to identify the person who most closely resembles the perpetrator when the actual perpetrator is not present within the lineup. The risk of making an erroneous identification based on similarities in appearance can be minimized by a pre-lineup admonition as described earlier. An additional consideration involves law enforcement’s response when a witness identifies the suspect within the lineup. If officers confirm that the suspect was correctly identified, the witness’s confidence that the suspect is the actual perpetrator may be bolstered. The post-identification feedback effect refers to the augmentation of a witness’s memory due to a newfound confidence that the perpetrator was correctly identified. This may have a significant impact, given that confidence is often regarded by the court as a predictor of accuracy in eyewitness testimony.

Methods to Mitigate Inaccurate Testimony

A standard measure of evaluation exists that functions to ensure eyewitness testimony is sufficiently reliable for inclusion into evidence. Modifications to this standard have been proposed that seek to more stringently evaluate for accuracy, recognizing that any degree of suggestibility is a form of bias that can negatively impact the veracity of eyewitness testimony.

Current Standard

Although states may vary slightly in this regard, the national standard for assessing reliability of eyewitness testimony is based on the 1977 decision of the Supreme Court of the United States in Manson v. Brathwaite (1977). Referred to as the Manson test, this standard involves consideration of two aspects of eyewitness testimony. First is an evaluation of whether the procedure surrounding the identification of the alleged witness was unnecessarily suggestive. Close attention is paid to (1) the presentation of the alleged suspect during the lineup, (2) whether a pre-lineup admonition was administered, and (3) the presence of any confirmatory statements made by law enforcement once the suspect was identified. If there is no evidence of suggestibility, then the testimony is admitted into evidence.

On the other hand, if there is evidence of suggestibility, then five additional criteria are used to assess the reliability of the testimony. These include (1) the extent to which the witness was able to observe the alleged perpetrator, (2) the attention paid by the witness during the event, (3) the confidence of the witness at time of identification, (4) the degree to which the description provided by the witness matches the appearance of the suspect, and (5) the time between the event and the identification.

Considering such additional notions seems reasonable when suggestibility is evident; however, it is important to note that the first three criteria (observation, attention, and confidence) are subjective, relying purely on witness self-report, with confidence being particularly problematic in this regard. Furthermore, the accuracy of witness descriptions has been shown to lack significant correlation to reliability. Finally, although the time that lapsed between the event and the identification is correlated to the strength of a memory, the greatest degree of memory attenuation occurs within minutes of the event. Accordingly, there may not be a meaningful difference in memory strength when the event extends beyond even 24 hr.

Proposed Modifications

A proposed standard for assessing the accuracy of eyewitness testimony is the four-step I-I-Eye method, which has been demonstrated to increase significantly the sensitivity of judges and jurors alike to these important issues. Inherent in this method is considering eyewitness testimony as a type of trace evidence, like DNA and fingerprints. With trace evidence, judges determine if appropriate, scientifically based protocols were followed prior to the material being admitted into evidence. The following four steps comprise a proposed protocol for judges to employ in making this determination.

First, one should determine if the interview was optimally conducted by law enforcement by (1) maximizing the amount of information obtained from an eyewitness, (2) not providing postevent information that may bias the recollection of the event, and (3) avoiding confirmatory statements that may increase an eyewitness’s confidence. Second, one should assess the fairness of the identification procedures employed. Third, one should identify and consider any estimator variables that may affect subsequent testimony. Finally, conclusions should be based on the following four questions:

1. Did law enforcement obtain the maximum amount of information from the witness?
2. Prior to taking a statement on the confidence of the eyewitness, was any feedback given that may have increased confidence?
3. Is there a low, medium, or high probability that the testimony is accurate?
4. Is there a low, medium, or high probability that the identification is accurate?

A final implication of this proposed standard is its emphasis on how suggestibility may decisively influence eyewitness testimony. Specifically, if bias has occurred during either the interview step or the identification step, it cannot be undone, even if those steps are subsequently repeated. In such instances, with few exceptions, the testimony would be deemed unreliable and would not be admitted into evidence.

References:

1. Larson, R. P., Goldfarb, D., & Goodman, G. S. (2015). Introduction to this issue: Children’s eyewitness memory and testimony in context. Behavioral Sciences & the Law, 33, 367–371.
2. Manson v. Brathwaite, 432 U.S. 98 (1977).
3. Ross, S. J., Tredoux, C. G., & Malpass, R. S. (2014). Evaluating eyewitness testimony of adults. In I. B. Weiner & R. K. Otto (Eds.), The handbook of forensic psychology (4th ed., pp. 513–559). Hoboken, NJ: Wiley.
4. Safer, M. A., Murphy, R. P., Wise, R. A., Bussey, L., Millett, C., & Holfeld, B. (2016). Educating jurors about eyewitness testimony in criminal cases with circumstantial and forensic evidence. International Journal of Law and Psychiatry, 47, 86–92.
5. Smalarz, L., & Wells, G. L. (2011). Eyewitness identification evidence: Scientific advances and the new burden on trial judges. Court Review, 48, 14–21.
6. Wise, R. A., & Safer, M. A. (2012). A method for analyzing the accuracy of eyewitness testimony in criminal cases. Court Review, 48, 22–34.