Every year in North America at least 75,000 people are identified from police lineups and subsequently prosecuted. There are hundreds of documented cases in which mistaken eyewitness identification has led to false imprisonment. Although it is impossible to know how often eyewitnesses make mistakes, it is known that mistakes are made. For example, of approximately 8,000 sexual assault cases in which DNA was tested by the U.S. Federal Bureau of Investigation, the suspect was exonerated approximately 25% of the time. In most of those cases, eyewitness identification was the primary way in which suspects were identified. Furthermore, of 140 cases in which people have been falsely imprisoned and subsequently exonerated, more than 80% involved mistaken eyewitness identification. Not surprisingly, improving the accuracy of eyewitness evidence has been a focus of legal psychologists for many years.
History of Research on Eyewitness Testimony Accuracy
For many years, psychologists have been trying actively to understand how and why eyewitnesses make mistakes. Hugo Munsterburg published his groundbreaking volume on the topic, titled On the Witness Stand, in 1906. This book, based on basic memory research, detailed how eyewitnesses were prone to a number of errors and did not have the perfect memories the legal system often assumed. Although Munsterburg’s book stimulated a great deal of discussion after it was published, it was largely dismissed by the legal system because it did not describe research that had been done specifically on eyewitnesses to crimes.
Contemporary Research on Eyewitness Testimony Accuracy
Hundreds of more recent studies have explored the accuracy of eyewitness identifications. The basic distinction these researchers have made is between system variables, factors the legal system can control, and estimator variables, factors the legal system cannot control. With system variables, one can construct the situation so that errors may be avoided. For example, a biased lineup instruction from the police (e.g., “The guy is in the lineup, all you have to do is pick him out”) will lead to more eyewitness errors than will an unbiased instruction (e.g., “The criminal may or may not be present in the lineup”).
An example of an estimator variable is the race of the criminal relative to the witness. Research has shown a consistent decrease in eyewitness accuracy if the witness and the suspect are of different races. Because this decrement is greater when the majority group member identifies a minority, relative to when a minority attempts to identify a majority group member, some researchers have suggested familiarity with the other group may be the cause of the increased error. Other researchers have suggested that different racial groups focus on different facial features. However, a definitive cause has thus far eluded researchers. Importantly, there is little the legal system can do to alleviate this problem. Thus, many researchers have argued it may be best to focus on system variables, as these are the factors that the legal system will be able to change.
Guidelines for Investigators
Although research in the field has been around for 35 years, it is only recently that practitioners have begun to embrace the findings and change their procedures to avoid errors. In 1999, the U.S. Department of Justice published a landmark set of recommendations on the treatment of eyewitness evidence, titled Eyewitness Evidence: A Guide for Law Enforcement. Then attorney general of the United States, Janet Reno called together the relevant stakeholder groups and asked them to come up with an empirically validated guide for law enforcement that the whole group could support. The final group consisted of 33 members: 17 police officers, 4 district attorneys, 6 public defenders, and 6 eyewitness researchers. The Guide, jointly developed by the legal community and researchers, may represent the most comprehensive and potentially influential work of its type ever developed.
The purpose was to provide step-by-step procedures for law enforcement in dealing with eyewitnesses. These procedures were designed to minimize the risk of contaminating eyewitness evidence and reduce the likelihood of mistaken identifications. One set of procedures, based on the empirically validated “cognitive interview” technique, was designed to improve how police interviewed witnesses. Specifically, law enforcement officials were encouraged to establish rapport with the witnesses; encourage witnesses to give complete answers; ask only open-ended, nonleading questions; and caution people against guessing about what they had witnessed. As for the lineup process used by police, the Guide recommends the use of single suspect lineups (i.e., one suspect with all others being known innocent fillers); ensuring the use of unbiased lineup instructions; and avoiding post-identification suggestions and feedback to witnesses.
In addition, researchers in the group wanted the Guide to recommend the sequential lineup procedure (where photos are shown one at a time) as preferable to show-up (one photo) and simultaneous (where the photos are all shown at the same time) lineups. Sequential lineups have dramatically reduced error rates compared to other types of lineups. In addition, the researchers wanted to recommend double-blind procedures such that neither the officer conducting the lineup nor the witness is aware of the identity of the suspect, a procedure that further reduces false identifications. However, other members of the group were reluctant to include these two elements in the Guide as they might not be feasible for some police departments. Interestingly, many law enforcement agencies that have adopted the Guide report little difficulty implementing these procedures.
Thus, it is clear that the psychological research on eyewitness identification evidence has had a great deal to contribute to the legal system. Indeed, this area is an excellent example of how researchers and practitioners can work together to ensure a better outcome for society. However, though the eyewitness literature has provided a great deal of insight, there remain many questions that need to be answered.
- Wells, G. L. (1993). What do we know about eyewitness identification? American Psychologist, 48, 553-571.
- Wells, G. L., Malpass, R. S., Linsay, R. C., Fisher, R. P., Turtle, J. W., & Fulero, S. M. (2000). From the lab to the police station: A successful application of eyewitness research. American Psychologist, 55, 581-598.