In its 1972 ruling in Neil v. Biggers, the U.S. Supreme Court outlined five criteria that should be used in evaluating the accuracy of eyewitness identifications: the witness’s certainty, his or her quality of view, the amount of attention paid to the culprit, the agreement between the witness’s description and the suspect, and the amount of time between the crime and the identification attempt. For many reasons, these criteria are suboptimal. Some of them directly contradict empirical research, and others can actually be misleading under certain circum-stances. Preferable methods for evaluating accuracy include assessing the suggestiveness of the identification procedure, including the instructions given to the witness; examining the structure of the lineup or photo spread; and checking whether the person administering the photo spread knew who the suspect was.
The U.S. Supreme Court’s decision in Neil v. Biggers (1972) was the first time that the Court had made explicit recommendations about evaluations of eyewitnesses in criminal cases. These criteria, known to eyewitness researchers as the Biggers criteria, are
the opportunity of the witness to view the criminal at the time of the crime, the witness’ degree of attention, the accuracy of the witness’ prior description of the criminal, the level of certainty demonstrated by the witness at the confrontation, and the length of time between the crime and the confrontation.
Although the Biggers criteria are intuitively appealing, psycholegal researchers generally disapprove of them for several reasons. First, reports on the extent to which these criteria are met, especially confidence, are not reliably related to identification accuracy. Second, most of the reports are subjective, provided by the very person whose memory is in dispute when an identification is challenged. Finally, they attempt to postdict accuracy, a goal that has had limited success throughout the empirical study of eyewitnesses. The background of the criminal case that prompted the criteria and the empirical data related to each criterion are presented below.
In the crime for which Biggers was convicted, the victim was taken from her home and raped along the railroad tracks a short distance away. The attack lasted between 15 and 30 minutes. Several times after the assault, the victim was shown photos in both lineups, where multiple photos are shown at a time, and showups, where only one photo is presented. She did not identify anyone from these photos. Seven months after the assault, Biggers was identified in a police station showup. The showup was conducted because the police claimed that they were unable to locate appropriate fillers for a lineup. After the police escorted the victim past the defendant, she asked them to have him say, “Shut up, or I’ll kill you,” a phrase used by her assailant. She then identified Biggers and indicated that she had “no doubt” about the accuracy of her identification.
The critical issue decided by the Supreme Court was whether the showup was “unnecessarily suggestive” and therefore violated due process. In previous cases, the Court argued that it was possible for identification procedures to be so unnecessarily suggestive as to render the identification unreliable and therefore a violation of due process. However, such a determination must be considered “on the totality of the circumstances.” For example, in Stovall v. Denno (1967), the identification procedure did not violate due process because the victim’s critical medical condition justified the hospital room showup and the victim subsequently identified the defendant at trial. In another case, the totality of the circumstances analysis indicated that due process was violated because the witness was not able to definitively identify the defendant during either a suggestive lineup or a showup. In that case, it was only after the third exposure to the defendant that the witness produced a confident identification.
Using the criteria described above, the Court (with five justices in the majority) concluded that the showup procedure was not unnecessarily suggestive. First, the victim had ample opportunity to view her attacker, both “under adequate artificial light…and under a full moon outdoors.” Second, the witness was “no casual observer,” suggesting that she had paid close attention to the culprit’s face during the assault. Third, her description was “more than ordinarily thorough,” including mention of the assailant’s height, age, skin tone, and voice. Fourth, she displayed a high level of certainty in her identification, saying “I don’t think I could ever forget [that face].” Finally, although 7 months had passed between the crime and the identification of Biggers, the Court reasoned that because the victim never made any other identifications, in spite of many opportunities to do so, “her record for reliability was thus a good one, as she had previously resisted whatever suggestiveness inheres in a showup.” Therefore, according to the totality of the circumstances analysis, reports on the criteria were acceptable, suggesting that the witness’s identification was accurate in spite of the suggestive procedure used to obtain it. The recommendations articulated in Biggers were upheld in Manson v. Brathwaite (1977); the Court has not revisited recommendations for evaluating eyewitnesses since 1977.
At the time these recommendations were issued, the field of psychology and law was in a nascent stage; the flagship journal of the field, Law and Human Behavior, had not yet been established. Therefore, the justices relied on intuition rather than empirical evidence when designing a set of criteria. Although the criteria are intuitively appealing, several significant problems exist.
Certainty, View, and Attention: Subjective Criteria
Three of the five Biggers criteria—certainty, view, and attention—are subjective reports produced by the eyewitness, the very person whose accuracy is at issue in a criminal trial. The most problematic of these is undeniably witnesses’ certainty. Perceivers naturally assume that a confident witness is an accurate one. This belief is well-founded under certain circumstances. When witnessing conditions vary widely, there is a strong relationship between confidence and accuracy: Witnesses who had a poor view are less confident than witnesses who had a good view. However, the integrity of the relationship between confidence and accuracy is easily compromised. For example, information suggesting that a co-witness identified the same person inflates certainty, as does identifying the person believed to be the culprit by the photo-spread administrator. The details of the Court’s certainty recommendation suggest that there may have been some awareness that confidence is malleable. The critical confidence report, from the Court’s perspective, was certainty “at the time of the confrontation,” not confidence at the time of the in-court identification. It is possible that the Court’s specification derived from an awareness that subsequent reports were vulnerable to influence by external variables.
Unfortunately, the Court’s stipulation that certainty at the time of the identification is the relevant report does not ensure that this reported certainty provides useful information about accuracy. Indeed, simple manipulations can dramatically distort witnesses’ memories of how certain they were at the time of their identification. A recent meta-analysis summarized the results of 20experimental tests with more than 2,400 participants. Witnesses who were told that their identification was correct (i.e., “Good, you identified the suspect”) reported recalling greater certainty in their identification than did witnesses who were told nothing about the accuracy of their identification. This inflation is especially troubling because in the original experiments, witnesses made identifications from target-absent photo spreads, meaning that their inflated confidence accompanied an incorrect identification. More troubling, this simple manipulation also distorted reports on the two other subjective criteria. Eyewitnesses who heard that their identification was correct reported better views and paying more attention compared with witnesses who heard nothing about their accuracy. In conclusion, three of the five reports are distorted by simple, legal comments from investigators.
Description and Time: Objective Reports
Description and time are primarily objective reports— evaluations of these criteria do not rely on witnesses’ own reports. Evaluators can examine for themselves the degree of match between a witness’s description and the appearance of the defendant. Similarly, there is a record of the date of the crime and the date the suspect was positively identified. Psychological literature generally supports the time criterion: Accuracy diminishes as the time between the witnessed event and the identification attempt increases. However, the literature on the description criterion is mixed. One study concluded that there is no connection between an eyewitness’s description and identification accuracy. Other studies suggest that witnesses are more likely to identify suspects if the witnesses have given a detailed description. Still other studies complicate the relationship even further by suggesting that the mere process of providing a verbal description of a perpetrator harms identification accuracy (i.e., verbal overshadowing). An empirically validated interviewing style, known as the cognitive interview, increases the quality of witnesses’ descriptions but, as predicted by verbal overshadowing, decreases identification accuracy. Fortunately, identification accuracy rates are preserved if a delay exists between the verbal description and the identification attempt. Even though description and time are primarily objective, they are not immune to influence by external variables such as interviewing style.
System Variables versus Estimator Variables
Most of the Biggers criteria are estimator variables— that is, variables that are not under the control of the justice system. For example, the justice system has no control over what kind of view the eyewitness had of the culprit. Similarly, the justice system has no control over how much or what kind of attention the witness paid to the culprit. In contrast, the time criterion falls into the system variable category—that is, variables that the justice system can control. For example, investigating officers can decide whether to show photos to a witness immediately after a crime is reported or wait until a suspect is located for a live lineup. The ability to make these decisions means that time is a system variable rather than an estimator variable.
The two remaining variables, confidence and description, straddle both categories. In some ways, they are estimator variables because the justice system cannot ensure that crime characteristics lead to high confidence or good descriptions (e.g., by ensuring that the culprit is in view for a long time and has no disguise). However, both criteria have system variable elements. Confidence can be easily manipulated by external factors having nothing to do with identification accuracy. Myriad variables, such as postidentification feedback, co-witness information, and repeated questioning, affect eyewitnesses’ confidence. The quality of a witness’s description is also influenced by external variables such as the style of interviewing. The cognitive interview increases both the amount and the quality of information gathered from eyewitnesses.
The Biggers Criteria Attempt to Postdict Accuracy
The final problem with the Biggers criteria is that they attempt to postdict accuracy—that is, to determine from eyewitnesses’ own reports whether an identification that had already occurred was accurate or inaccurate. Empirical research reveals limited success in postdicting accuracy. This will be especially difficult when the variables intended to postdict accuracy are so vulnerable to distortion. A preferable strategy is to minimize the likelihood of inaccurate identifications at the time of the confrontation. One clear example of such a change is to require investigators to warn eyewitnesses that the culprit might or might not be present in the set of photos. Another is to obtain a report of the witness’s confidence immediately after the identification is made, allowing defense attorneys to challenge inflated confidence reports at trial.
The continuing publicity surrounding DNA exonerations of individuals wrongfully identified should impress on the Court the need to revisit the Biggers criteria. Should the Court undertake such a challenge, some preference for system variable changes would likely be articulated by many psycholegal researchers. In the meantime, defense attorneys and expert witnesses alike should continue to challenge the utility of these criteria, especially confidence, in contributing to meaningful evaluations of eyewitness identification accuracy. At best, the criteria outlined by the Court provide limited information about accuracy. At worst, the criteria provide misleading information, suggesting that triers of fact rely on variables that have tenuous relationships with accuracy.
- Bradfield, A. L., & Wells, G. L. (2000). The perceived validity of eyewitness identification testimony: A test of the five Biggers Law and Human Behavior, 24(5), 581-594.
- Douglass, A. B., & Steblay, N. (2006). Memory distortion in eyewitnesses: A meta-analysis of the post-identification feedback effect. Applied Cognitive Psychology, 20, 859-869.
- Manson v. Brathwaite, 432 U.S. 114 (1977).
- Neil v. Biggers, 409 U.S. 188 (1972).
- Stovall v. Denno, 388 U.S. 293 (1967).
- Wells, G. L. (1978). Applied eyewitness-testimony research: System variables and estimator variables. Journal of Personality and Social Psychology, 36(12), 1546-1557.
- Wells, G. L., & Murray, D. M. (1983). What can psychology say about the Neil v. Biggers criteria for judging eyewitness accuracy? Journal of Applied Psychology, 68(3), 347-362.