Scott D. Gronlund - US grants

The advent of DNA testing has demonstrated that innocent people are in prison, and faulty eyewitness identification has been shown as the primary contributor to the conviction of the innocent. However, a set of procedural safeguards now exist that can improve the reliability of eyewitness identification. Among the most promising is the use of sequential lineups. A sequential lineup (i.e., viewing suspects one at a time) is thought to be superior to a simultaneous lineup (i.e., viewing all suspects at the same time) because it leads a witness to compare each successive person in the lineup to the witness' memory of the perpetrator. In contrast, a simultaneous lineup leads a witness to choose the person in the lineup who looks most like the perpetrator, which is problematic if the police have the wrong man and the actual perpetrator is still at-large. Without an theoretical understanding of the mechanisms underlying the sequential lineup advantage, however, the nationwide adoption of sequential lineups as a procedural safeguard will be hampered. The goal of this research is to develop a quantitatively specified model of the mechanisms underlying the sequential lineup advantage.

The principal investigator hypothesizes that the sequential lineup advantage is due to: 1) enhanced encoding of distinctive information and 2) use of a recall process to access that information. This project comprises experiments that test these hypotheses. In one line of experiments, the encoding conditions will be modified to make the originally distinctive stimulus (i.e., height) no longer distinctive. If the model is right, this should make the sequential lineup advantage disappear. To evaluate the contribution of recall, attention is divided during lineup testing. Dividing attention nullifies the use of a recall process, and the sequential lineup advantage should disappear. Procedural safeguards are available that can enhance the accuracy of eyewitness identification, but without full understanding of what these safeguards do and why, their influence will be limited. This grant will redress this limitation by developing and testing quantitatively specified explanations of the mechanisms underlying one such procedural safeguard, sequential lineups. Finally, an improved understanding of how a sequential lineup protects the innocent may point to other improvements to eyewitness identification procedures that will further enhance the reliability of eyewitness identification.

Recent advances in DNA testing have uncovered an alarming observation: innocent people are in prison and the guilty people that committed those crimes are at large. In the majority of the false conviction cases, faulty eyewitness identification is to blame. One standard police practice, showing witnesses mugshots of potential offenders (or people who may resemble them), accounts for some of these errors. This research identifies cases where the ability of an eyewitness to identify a perpetrator might be harmed or enhanced by the process of searching though a series of mugshots following the crime. Research has shown that witnesses who view mugshots may be less able to correctly identify the perpetrator of the crime. Moreover, witnesses can mistake individuals they view in a mugshot search for the perpetrator, even when subsequently presented with the actual perpetrator. From a memory and decision-making perspective, techniques are needed that avoid biasing a witness? memory, while at the same time aiding the police search for the culprit of the crime. This research includes an empirical exploration of the factors that produce lineup identification errors due to a prior exposure to a mugshot search and develops a formally-specified theoretical explanation for how these effects operate. In addition, new techniques designed to avoid these errors will be tested. This will allow us to develop better police practices for aiding a culprit search without harming the ability of the eyewitness to make an identification.