magazine issue 11 2015 / Issue 28

Seeing and Believing: Common Courtroom Myths in Eyewitness Memory

• written by Nina Tupper, Melanie Sauerland, Lorraine Hope & Harald Merckelbach
• edited by Emir Efendic

In the lab, psychologists demonstrate how seemingly-harmless decisions in constructing a lineup can easily bias someone to choose the suspect - whether that suspect is guilty or not. Lineup fairness can be assessed by providing non-witnesses (people who are ignorant to the identity of the perpetrator) with a description of the perpetrator and then asking them to pick out the lineup-member that best matches that description (Doob & Kirshenbaum, 1973). If the lineup is fair, responses should be evenly spread among the lineup members presented. This is what the Eyewitness Identification Research Laboratory at the University of Texas (El Paso) did with the lineup of a perpetrator described as an African-American, male teenager with long, braided hair (n.d.). Although the lineup was constructed with six African-American teens, 95% of non-witnesses managed to pick out the police suspect - the only person with braided hair. In this case, a positive identification only demonstrates that the eyewitness is able to identify the most reasonable option. Unfortunately, juries and judges often only see the final result of a police identification lineup. Without the relevant background information or documentation of the actual lineup, there is no real means of judging the reliability of a positive suspect identification. 

In cases like Osagiede, it is clear that influences proven to impact the ability of eyewitnesses to make accurate identifications were not considered. Even more alarming is when such influences are not revealed to or considered by judges and juries making life-altering decisions to convict or acquit the accused. Perhaps “everyone knows” that eyewitnesses can be inaccurate, but it can easily be forgotten, ignored, or lost in translation when applied to investigatory and legal decision-making.

Myth #3: Consistency Is the Hallmark of the Reliable Witness 

In the legal arena, consistency is considered a vital means of determining witness credibility. Police officers intuitively distrust accounts that change from one telling to the next (Krix, Sauerland, Lorei, & Rispens, 2015) and lawyers use inconsistencies to discredit a witness on the stand (Fisher, Brewer, & Mitchell, 2009). In fact, juries in the U.S. are instructed to use statement consistency as a critical factor in determining the credibility of an eyewitness. However, inconsistencies appear any time we rely on memory, and are not necessarily evidence of a defective or deceptive witness. 

In the spring of 2011, six-year-old Vicky¹ witnessed the murder of her mother (Brackmann, Otgaar, Sauerland, & Jelicic, 2015). When police, responding to neighbors’ reports of domestic disturbances next door, arrived at the house, Vicky answered the door and immediately said her father had done it. She was interviewed officially the same day, and she repeated her allegation that her father was the murderer. Two months later, she was questioned a second time. She again blamed her father, but this time produced new details about how the crime was committed, including the murder weapon used (a knife). Since the father denied the crime, Vicky became a key witness against him. 

Imagine now a typical courtroom tactic in which the defense cross-examined Vicky on the stand. Given that statement inconsistency is used as a means of determining a credible witness, lawyers would likely pursue these new details to suggest that Vicky was unreliable, had been given external information, or had made up a story to explain the traumatic death of her mother. However, when we report about an event several times, as witnesses are often required to do, it is not unusual for different information to appear across accounts.   

When inconsistencies do arise, they can be details that are a) initially reported and later changed (contradictory), b) reported in an earlier interview, but not a later one (forgotten or non-reported), or c) initially not reported, but reported later (reminiscent). Despite general suspicion that all inconsistent details reflect unreliability, reminiscent and forgotten details themselves tend to be as accurate, or only slightly less accurate, than consistent ones (Krix et al., 2015; Oeberst, 2012). Furthermore, inconsistencies across statements show little to no relationship with the accuracy of the overall account (Smeets, Candel, & Merckelbach, 2004).