Why Don't You Take Another Look at Number Three?: Investigator Knowledge And Its Effects on Eyewitness Confidence and Identification Decisions

• Author: Russano/Dickinson/Greathouse/Kovera
• Position: Assistant Professor of Criminal Justice
• Pages: 354-380

Page 355

Melissa B. Russano received her B.A. from the University of Virginia and her M.S. and her Ph.D. in psychology from Florida International University. She is an Assistant Professor of Criminal Justice at Roger Williams University. Jason J. Dickinson received his B.A. and his M.A. from Central Michigan University and his Ph.D. in psychology from Florida International University. He is an Assistant Professor of Psychology at Montclair State University. Sarah M. Greathouse received her B.A. from Western Illinois University and her M.S. from Florida International University. She is a doctoral student in psychology at John Jay College, City University of New York. Margaret Bull Kovera received her B.A. from Northwestern University and her Ph.D. from the University of Minnesota. She is a Professor of Psychology at John Jay College of Criminal Justice, City University of New York. Questions and comments should be directed to the first author at mrussano@rwu.edu.

I Introduction

Between 1977 and 1979, at least twenty-five women in Coral Gables, Florida reported being sexually assaulted by a man who became known as the "Bird Road rapist." The rapist would typically locate his victims by flashing his car's headlights, prompting the women to pull over, and he would then force the victims into his car at gunpoint. A number of victims described the assailant as a tall, heavyset (approximately 6'0" to 6'2", 200 pounds) Latin male who spoke good English with an accent.¹ After one of the victims thought she saw her assailant at a gas station and copied down a license plate number, the owner of the car, Luis Diaz, became a police suspect. Although the victim later identified Diaz from a photoarray, Diaz did not match the witness'

This material is based upon work supported by the National Science Foundation under Grant No. 9986240. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.Page 356 description (e.g., he was 5'3", 134 pounds, and spoke little to no English), and no charges were filed.²

As the attacks continued, pressure from the victims and media to capture the perpetrator mounted.³Another victim was asked to view a photo lineup with Diaz's picture in it. Notably, while she first said that she did not recognize any of the men as her assailant, she "eventually said Diaz resembled her attacker" and "under pressure, she ultimately identified him as her attacker."⁴In the end, a total of eight victims identified Diaz from photoarrays or live lineups.⁵Although there was no corroborating physical evidence, in 1980 Diaz was found guilty of seven counts of rape or attempted rape, and he was sentenced to multiple life sentences in prison.⁶Diaz always maintained his innocence, and the first break in his case came in 1993 when two victims recanted their identifications. In 2003, with the help of the Innocence Project, Diaz filed for post-conviction DNA testing. Two semen samples attributed to the Bird Road rapist were tested; neither matched Luis Diaz's DNA. After serving twenty-six years in prison, on August 3, 2005, Diaz was released from prison as a free man.⁷

The Luis Diaz case illustrates an increasingly recognized phenomenon-that mistaken eyewitness identification is a leading cause of wrongful convictions in the United States.⁸In attempting to identify problematic identification procedures that may adversely affect the reliability of eyewitness identifications, researchers have expressed concern over the role that a lineup administrator's knowledge of a suspect's identity may have on identification decisions and witness confidence.⁹Spe-Page 357cifically, the issue of interest is whether an investigator who is aware of a suspect's identity might intentionally or unintentionally influence a witness to identify the suspect or influence a witness' confidence in his or her decision. Of particular concern is the possibility that the investigator's knowledge might precipitate a dangerous chain of events, in which an investigator's knowledge of a suspect's identity is communicated to a witness, leading to a false identification of an innocent suspect by a highly confident witness, which in turn results in a wrongful conviction by a convinced jury.

A number of leading researchers in the area have called for the implementation of "double-blind" procedures, in which both the lineup administrator and the witness are unaware of the suspect's identity prior to and during lineup administration, to safeguard against these possible influences.¹⁰Would the outcome of Luis Diaz's case been different if the lineup investigators had been "blind" to the suspect's identity? Would eight victims still have identified Luis Diaz as the man who had raped them? Obviously, the answers to these questions will remain forever unknown; however, one cannot help but wonder exactly what kind of "pressure" the second witness was under to identify Diaz as her attacker after saying he resembled the assailant. Is it possible that she, and the other witnesses, made their identifications after investigators somehow communicated their knowledge of Diaz's identity? What form could that influence have taken? Does lineup administrator knowledge really affect witness decision-making? The purpose of this article is to address these issues by exploring the current knowledge base in the area of investigator influence, including a review of the empirical research on the effect of investigator knowledge on both eyewitness confidence and identification accuracy. Policy implications will also be discussed.

II Background

Wells and Luus suggested that identification tasks can be compared in a number of ways to scientific experiments, and that it would be useful to apply many of the same procedures and principles researchers use to conduct methodologically strong experiments to identificationPage 358tasks.¹¹The logic of the lineup-as-experiment analogy is that police begin with a hypothesis that a particular suspect is the perpetrator, and to test that hypothesis, they conduct an experiment (i.e., the identification procedure).¹²The investigator conducting the lineup is analogous to the experimenter, and the witness is analogous to a research participant. Police construct a lineup by selecting fillers or foils (i.e., they create the study materials) and conduct the identification task (i.e., they execute the design and procedure of the study).¹³They record the witness' decision (i.e., the data) and based on the witness' decision, they reevaluate the evidence against the suspect (i.e., they decide whether the data supports the research hypothesis).¹⁴

One of the principles of methodologically sound research is that an experimenter should be kept blind to the hypothesis of the study (or at least to the research condition) to prevent experimenter-expectancy or experimenter bias effects.¹⁵In other words, if an experimenter knows the hypothesis of the study, he or she may cue or encourage the witness to respond in a particular way, especially in a way that confirms the experimenter's hypothesis. For example, experimenters may inadvertently leak the hypothesis of the study or engage in confirmation bias behavior. In this manner, the experimenter is seeking out and eliciting information consistent with the hypothesis and ignoring inconsistent information.¹⁶

Researchers have expressed concerns that experimenter-expectancy effects may also occur during identification tasks.¹⁷Specifically, police administrators who know the identity of the suspect may communicate that knowledge to the witness through verbal or non-verbal cues.¹⁸It is important to note that this process may be intentional or unintentional;Page 359 it is possible for neither the administrator nor the witness to be aware that the influence is occurring. For instance, the administrator may suggest to the witness that the witness reconsider a specific photograph or through eye contact or body language signal to the witness that they are focusing on the suspect.¹⁹After a witness makes an identification, the police officer might directly tell the witness whether he or she identified the suspect, which in turn can inflate the witness' confidence.²⁰They may also engage in behavior driven by confirmation bias by asking witnesses questions that direct the witness' focus to the suspect and away from the other photos (e.g., do you think he had bushy eyebrows?).²¹Therefore, one implication of the lineup-as-experiment analogy is that police investigators conducting an identification task should be kept blind to the suspect's identity (i.e., therefore preventing hypothesis-leaking and confirmation bias), and a number of leading researchers in the field have recommended that police departments adopt this as a good-practice procedure.²²

Partly due to increasing recognition of the growing number of wrongful conviction cases involving faulty eyewitness identifications,²³in the late 1990s then Attorney General Janet Reno commissioned a taskforce of researchers, police officers, prosecutors, and defense attorneys to develop national guidelines for the collection and preservation of eyewitness evidence.²⁴The resulting guidelines, issued by the National Institute of Justice (NIJ), were released in 1999 and are supported by years of scientific research on eyewitness memory and identification.²⁵Some of the major guidelines with respect to identification procedures addressed how fillers/foils (i.e., known innocents) should be selected,²⁶the instructions that should be given to the witness,²⁷the introduction of the sequential...