Eyewitness Identification Resource Guide:
A Primer for Reform
“A major factor contributing to the high incidence of miscarriages of justice from mistaken identification has been the degree of suggestion inherent in the manner in which the prosecution presents the suspect to witnesses for pretrial identification. . . . There is grave potential for prejudice, intentional or not, in the pretrial lineup, which may not be capable of reconstruction at trial.”
—Supreme Court Justice William Brennan, writing for the majority in United States v. Wade (1967)
Table of Contents.
I. Blind Administration:
Recommended Methods for Implementation…………………..9-10
II. Providing Instructions to Witnesses:
III. Composition of Photo Array and Live Lineup Members:
IV. Confidence Statements:
V. Recording the Identification Procedure:
VI. Sequential Presentation (Optional):
Administrator: The person conducting the show-up, photo array, or live lineup.
Blind: When the administrator does not know which photo array or live lineup member being viewed by the eyewitness is the suspect.
Confidence Statement: A statement provided by the eyewitness in his own words upon identifying a photo array or show-up or lineup member, describing the eyewitness’s level of certainty regarding his selection.
Eyewitness: A person who observes another person at or near the scene of an offense.
Filler: Either a person or a photograph of a person who is not suspected of an offense and is included in an identification procedure.
Identification procedure: A photo array, live lineup or show-up.
Instructions: Directives provided by the administrator to the eyewitness before the eyewitness identification procedure.
Live lineup: An identification procedure in which a group of persons, including the suspected perpetrator of an offense and other persons not suspected of the offense, is displayed to an eyewitness for the purpose of determining whether the eyewitness identifies the suspect as the perpetrator.
Photo array: An identification procedure in which an array of photographs, including a photograph of the suspected perpetrator of an offense and additional photographs of other persons not suspected of the offense, is displayed to an eyewitness for the purpose of determining whether the eyewitness identifies the suspect as the perpetrator.
Sequential: An identification procedure in which the eyewitness views only one member of the live lineup or photo array at a time and is shown all persons or photographs even if an identification has been made before all the persons or photographs have been viewed.
Show-up: A show-up is an eyewitness identification procedure in which an eyewitness is presented with a single suspect in order to determine if the eyewitness identifies this individual as the perpetrator of the crime.
Simultaneous: An identification procedure format in which the eyewitness views all of the members of a lineup in a live lineup or shown all of the photographs in a photo array at the same time.
Misidentifications have contributed to the wrongful conviction of more than 75% of the 220 people (as of June, 2008) exonerated through post-conviction DNA testing in the United States. The single largest cause of wrongful conviction, misidentifications plague our criminal justice system from the investigative stage through final disposition.
The Innocence Project closely tracks all eyewitness identification studies, in an effort to help jurisdictions understand how such misidentifications happen – and as importantly, what can prevent them from happening. This Resource Guide synthesizes the most important information for those considering how to improve the accuracy of eyewitness identifications.
Mindful of the different political landscapes of each state, this Guide attempts only to provide its readers with uncontested scientific research, grounded in years of examination and endorsed by a wide range of justice and law enforcement organizations, such as the National Institute of Justice and the International Association of Chiefs of Police.
The average citizen might assume that eyewitness identifications are typically accurate, and that DNA exonerations are anomalous flukes. Surprisingly, however, studies show that the rate at which eyewitnesses select non-suspect from photo and live lineup members during the course of an identification procedure hovers around 20%.1 Of the 220 post-conviction exonerations proven through DNA testing to date, over 75% included at least one misidentification. These facts demonstrate that misidentifications are not irregularities, but rather common occurrences in the course of criminal investigations.
Misidentifications not only harm the innocent, but damage the ability of law enforcement to properly investigate crime. Inaccurate eyewitness identifications can “burn witnesses” and otherwise confound investigations from the earliest stages. Time is a critical factor in a successful criminal investigation; when a misidentification distracts police from the real perpetrator, the potential for justice diminishes, while the threat to public safety increaeses.
Despite solid and growing proof of the inaccuracy of traditional eyewitness identification procedures – and the availability of simple measures to reform them - traditional eyewitness identifications remain among the most commonly used. This raises serious concerns, as eyewitness identifications are regarded by juries as some of the most compelling evidence against criminal defendants.
The good news, however, is that we can eliminate much of this error. Over the past 25 years, a large body of peer-reviewed, scientific research and practice shows that simple and easily implemented systemic changes in administering eyewitness identification procedures can greatly improve the accuracy of those identifications. States, counties and municipalities across the nation are now rapidly adopting and implementing eyewitness identification reforms into their daily practice.
The Innocence Project endorses a range of reforms, based upon the results of nearly three decades of scientific research as well as surveys of jurisdictions, both large and small, which have implemented these best practices.
This resource guide will serve as a primer, addressing the five primary reform recommendations recognized by police, prosecutorial, and judicial experience, the Innocence Project, national justice organizations, and academic research:
Blind Administration of the Eyewitness Identification Procedure
Provision of Instructions to the Eyewitness
Proper Selection of Photographs/Live Participants (“Fillers”) to be Used in the Identification Procedure
Immediate Acquisition of a “Confidence Statement” From the Eyewitness
Recordation of the Identification Procedure
The experiences of the jurisdictions that have implemented these reforms is that when they are introduced with the proper education and training for the officers that administer them, they are met with little resistance.
The Innocence Project also supports the sequential (as opposed to simultaneous) presentation of lineup members, as there is a large body of research demonstrating that this significantly reduces the number of incorrect identifications. The sequential presentation has also been shown to reduce significantly the number of correct identifications, albeit at a substantially lower rate. We recognize, however, that the choice to present lineup members sequentially is a policy decision. The consideration of this policy decision will be better informed by a series of field studies that are already underway in a number of jurisdictions across the country. We hope this research will further clarify the benefits of employing sequential lineups in actual practice and resolve any lingering questions. We currently offer it as an optional reform opportunity. In the meantime, there is no reason to stall reform efforts related to the other five areas addressed in this Resource Guide.
Blind Administration of the Identification Procedure.
“Blind administration” of an eyewitness identification procedure refers to:
When the person administering the eyewitness identification procedure does not know which photo array or live lineup member is the person suspected by the police of having committed the crime under investigation.
When the person administering the identification procedure knows which individual is the suspect, a procedure is employed that prevents the administrator from knowing which lineup member is being viewed by the witness at any given time. [As a practical matter, this method of blind administration can only happen in the context of a photo array.]
Blind Administration of the Identification Procedure.
Scientific research has consistently shown that test subjects are influenced by the expectations of those who perform the tests. The idea that test administrators’ expectations are communicated either openly or indirectly to test subjects, who then modify their behavior in response, has been corroborated by over forty years of general social science research.2 A prominent meta-analysis conducted at Harvard University, which combined the findings of 345 previous studies, concluded that in the absence of a blind administrator, individuals typically tailor their responses to meet the expectations of the administrator.3
Integral to any scientific inquiry, blind administrators are employed in a range of research, from clinical trials in medical and pharmaceutical research to social psychology experiments. Blind administrators are important for eyewitness identification procedures, as eyewitness memories are easily manipulated by outside influences. The most likely source of such influence is the traditional (non-blind) identification procedure administrator, as that person knows who police believe is the suspect and may inadvertently transmit this knowledge to the eyewitness attempting to identify the real perpetrator. Specifically, a non-blind administrator may provide subtle influences that:
Lead the eyewitness towards a suspect;
Steer the eyewitness away from non-suspect (filler) members; and/or
Provide post-identification feedback to the eyewitness, which influences the eyewitness’s confidence in his selection, thereby affecting his testimony at trial.
The eyewitnesses themselves may seek clues from an identification procedure administrator. A recent experiment that sought to examine the decision-making processes of eyewitness test subjects concluded that “witnesses were more likely to make decisions consistent with lineup administrator expectations when the level of contact between the administrator and the witness was high than when it was low.”4 Eyewitness identification administrators’ power of influence is similarly documented in a number of other studies.5 In fact, the damage of a non-blind administrator does not necessarily contain itself to one eyewitness; one study indicates that information gleaned from one eyewitness by the administrator during an identification procedure also affects the identification choice of the second eyewitness if the same administrator conducts both identification procedures.6
Gary Wells, a leading authority on eyewitness identification, provides another example of how non-blind administration can negatively impact an investigation. If, for instance, an eyewitness viewing a photo array or live lineup states that, “number three is…well it looks like number three,” the identification procedure’s administrator has discretion as to whether a positive identification was actually made. Understandably, this discretion, according to Wells, tends to favor a selection of the police suspect when the administrator is not blind – even if the police suspect is not the actual perpetrator of the crime.7
Advocating for the use of a blind administrator does not call into question the integrity of law enforcement; rather it acknowledges a commonly-held precept in the scientific community and applies it to the eyewitness procedure. Employing a blind administrator in the area of eyewitness identification was first introduced in 1988,8 and has been shown to minimize both intentional and inadvertent suggestion.
A staple of science, the use of a blind administrator in a photo array or live line-up procedure will:
prevent the administrator from inadvertently leading the eyewitness towards the suspect;
prevent the administrator from inadvertently leading the eyewitness away from a ‘filler’; and
prevent post-identification feedback from occurring, thereby affecting the eyewitness’s level of confidence in his selection.
Blind Administration of the Eyewitness Identification Procedure - Methods.
It is standard practice for the case detective or investigator assigned to a particular investigation to administer an identification procedure associated with that case. This method evolved without regard for the effect it might have on eyewitness selection, because at the time there was no research upon which to determine its possible impact. Years of subsequent psychological studies, however, have demonstrated the significant value of a blind administrator.
A large body of peer-reviewed, scientific research now proves that both photo arrays and live lineups should be administered by a “blind” administrator. This eliminates any opportunity for administrator influence, however unintentional that may be.
Understandably, small police departments with limited officer power, or those that conduct numerous identifications off-site rather than at department locations, may believe that ‘blind administration’ is infeasible. The “Folder System”9 was devised to address these concerns yet enable blind administration. Even with the Folder System, however, it is still best, whenever possible, that the officer administrating the procedure not be one involved in the investigation. Should the investigating officer of a particular case be the only law enforcement personnel available to conduct a photo array, the Folder System should be employed. The following are the recommended instructions for the Folder System:
Use one suspect photograph that resembles the description of the perpetrator provided by the witness, five filler photographs that match that description and do not cause the suspect photograph to unduly stand out, and ten folders in total. [Four of the folders will not contain any photos and will serve as ‘dummy folders.’]
Affix one filler photo to Folder #1 and number the folder.
The individual administering the photo array should place the suspect photograph and the remaining four filler photographs into Folders #2-6 and shuffle the photographs so that he is unaware of which folder the suspect is in, and then number the remaining folders, including Folders #7-10, which will remain empty. [This prevents the witness from knowing when he has seen the last photo].
The administrator should provide instructions to the witness. The witness should be informed that the perpetrator may or may not be contained in the photos he is about to see and that the administrator does not know which folder contains the suspect.
Without looking at the photo in the folder, the administrator is to hand each folder to the witness individually. Each time the witness has viewed a folder, the witness should return the photo to the administrator. The order of the photos should be preserved, in a facedown position, in order to document in Step 6.
The administrator should then document and record the results of the procedure. This should include: the date, time and location of the photo array procedure; the name of the administrator; the names of all of the individuals present during the procedure; the number of photos shown; copies of the photographs themselves; the order in which the folders were presented; the sources of all of the photos that were used; a statement of confidence in the witness’s own words as to the certainty of his identification, taken immediately upon reaction to viewing; and any additional information the administrator deems pertinent to the procedure.
The Provision of Instructions to the Eyewitness.
“Instructions” are a series of statements issued by the identification procedure administrator to the eyewitness that deter the eyewitness from feeling compelled to make a selection, seek clues from the administrator about whom to (not) pick, or whether or not a selection was correct, and otherwise help minimize the likelihood of a misidentification.
Providing Instructions to Witnesses -
Understandably, the victim or observer of a crime may fear that his failure to identify a suspect will end the criminal investigation. Alternately, the eyewitness may think the procedure is simply a technicality to confirm what an investigation has already proven. What most do not realize is that eyewitness identifications have proven to be one of the most compelling, yet error-prone, forms of evidence used to secure a conviction. Instructions are meant to prevent the eyewitness from contributing unknowingly to a wrongful conviction.
Providing instructions to eyewitnesses helps them understand the role of the identification procedure in the investigative process. It is also helps decrease the pressure witnesses may feel to make a selection, which has been shown to contribute to misidentifications. Instructions can take many forms, but perhaps paramount in protecting the innocent is the instruction that the perpetrator “may or may not be present” in the photo array, live line-up or show-up. This “unbiased” instruction can be compared to a “biased” directive, in which the witness is told that the perpetrator is present in the photo array, live lineup, or show-up.
In identification procedures where the perpetrator is present, a landmark study found that witnesses choose at a high rate, regardless of the instructions provided to the witnesses.
In identification procedures where the perpetrator is absent (and therefore any identification would be incorrect), the type of instructions provided – “biased” or “unbiased” - have a significant effect on eyewitness’s accuracy. When “unbiased” instructions are given, the number of (incorrect) identifications decrease. When “biased” instructions were given, the number of (incorrect) identifications increase. In other words, when the perpetrator is not in the photo array or live lineup, and thus when the innocent suspect is more vulnerable, unbiased instructions lead to fewer false identifications, whereas biased instructions lead to an increased rate of false identifications.10
These findings are supported by a meta-analysis of results from twenty-two individual studies on the effects of instructions, which found that when biased instructions were provided to witnesses, a higher level of choosing occurred.11
This point is important since any resistance to mandating the use of unbiased instructions is likely rooted in a fear that it would cause more cautious witnesses to become wary of affirmatively identifying a correct choice. Yet research indicates that correct identifications remain constant regardless of the type of instruction offered, and the only effect of implementing an unbiased, or cautionary, set of instructions is a decrease in guessing, which can only help the innocent.12
It has been argued that the presence of counsel during the course of an identification procedure will act as an appropriate safeguard against suggestiveness. Research, however, indicates otherwise. It has been shown that attorneys have great difficulty detecting suggestive bias while instructions are provided to witnesses, and are therefore powerless to correct it.13 Further, in cases where attorneys correctly identify instances of identification procedure suggestiveness, research suggests that attempts to suppress the results of the suggestive procedure in court are only as effective as the judicial sensitivity to factors that underlie potential identification procedure bias.14
Because neither attorneys nor courts can be counted on to “trump” a suggestive eyewitness identification procedure, it is imperative that precautions are taken to protect the purity of the identification procedure at the time at which it is administered.
SPECIAL NOTE: The importance of providing instructions is particularly important in the case of a show-up procedure where law enforcement believe the exigent circumstances necessitate the one-on-one identification procedure format. In State v. Ledbetter15, the Connecticut Supreme Court ruled that when, at a showup, a positive identification is made in the absence of an instruction to the eyewitness that “the perpetrator may not be present,” the jury shall be warned that this failure increased the probability of a misidentification. This judicial concern is supported by a body of research that reveals that show-ups put the innocent at greater peril of misidentification than properly administered photo or live lineups.16
Providing Instructions to Eyewitnesses-
Substance of Recommended Instructions.
According to the National Institute of Justice Technical Working Group on Eyewitness Identification, comprised of police, prosecutors, judges, defense lawyers, and some of the nation’s leading eyewitness identification researchers, the following instructions should be given to the eyewitness by the administrator of the identification procedure17:
1. Instruct each eyewitness without other persons present.
2. Describe the photo array to the eyewitness only as a “collection of photographs.”
3. Instruct the eyewitness that the person who committed the crime may or may not be present in the identification procedure.
4. Consider suggesting to the eyewitness to think back to the event and his/her frame of mind at the time.
5. Instruct the eyewitness to select a photo array or live lineup member if he/she can and to state how he/she knows the person if he/she can.
6. Assure the eyewitness that regardless of whether he/she makes an identification, the police will continue to investigate the case.
7. Instruct the eyewitness that the procedure requires the investigator to ask the eyewitness to state, in his/her own words, how certain he/she is of any identification.
In addition to the recommendations promulgated by the National Institute of Justice, other jurisdictions have disseminated model instructions guidelines aimed at reducing the incidence of misidentifications. These include:
Advising the eyewitness that the administrator does not know who the suspect is.18
Directing the eyewitness not to discuss the identification procedure or its results with other eyewitnesses involved in the case and discouraging contact with the media.19
Composition of Photo Array and
Live Lineup Members.
“Fillers” should be selected based on their resemblance to the description provided by the witness, yet should not stand out unduly from the suspect.
Composition of Photo Array and Live Lineup Members – Scientific Support.
In the United States, photo arrays containing six photographs (including one suspect and five non-suspect, or filler, photographs) are generally viewed as the minimally acceptable lineup size. Yet, in order for this array to be adequate and just, the five filler photographs must be reasonable alternatives for the description provided by the eyewitness.20
In fact, when the innocent person is the only person to fit the description provided by the eyewitness, the confidence level of the eyewitness in his selection of the innocent person is greater than when other photo array or lineup members also fit the eyewitness’s description. Therefore, when photo array or live lineup members are selected that match the eyewitness’s description, high rates of accurate identifications can be maintained while reducing false identifications characterized by an inflated sense of confidence.21
Research suggests that impartial photo array and live lineup composition requires not only that filler members look like the eyewitness’s description, but also that the fillers:
do not make the suspect unduly stand out;
look different enough to be discernible from each other; and
possess features that reduce the likelihood of guessing or the use of deductive reasoning.
In other words, photo arrays and live lineups should be constructed in such a way that they are truly testing the memory of the eyewitness, rather than setting the stage for a guessing game in which one person stands out over the others.
Composition of Photo Array and Live Lineup Members – Recommendations.
While appropriate construction requires skill, available research has identified a number of suggestions that limit the subjectivity seemingly inherent in the photo array or live lineup construction decision-making process. 22
The Innocence Project, based upon a review of peer-reviewed research, recommends the following steps be taken when composing a photo array or live lineup:
For the suspect:
In the case of a photo array, the photograph of the suspected perpetrator should be contemporary, resemble his or her appearance at the time of the offense, and not unduly stand out when compared with the fillers.
If there are multiple eyewitnesses, the suspect should be placed in a different position in the live lineup and/or photo array for each eyewitness.
During an identification procedure, no writings or information concerning any previous arrest, indictment or conviction of the suspected perpetrator should be visible or made known to the eyewitness.
The suspected perpetrator should be the only suspected perpetrator included in the identification procedure.
For the fillers:
At least five fillers should be included in a photo array, in addition to the suspected perpetrator.
At least four fillers should be included in a live lineup, in addition to the suspected perpetrator.
Fillers should be selected that resemble the description of the perpetrator provided by the eyewitness and should resemble the eyewitness’s description in significant features (i.e. face, weight, and build), including any unique or unusual features (i.e. scar, tattoo, etc).
If the eyewitness previously viewed a photo array or live lineup in connection with the identification of another person suspected of involvement in the offense, the fillers in the photo array or live lineup in which the suspected perpetrator participates should be different from the fillers used in any prior identification procedures.
Confidence Statements- Explanation.
A confidence statement is a declaration provided by the eyewitness immediately upon identification and before any feedback is provided, in which he articulates in his own words the level of confidence he has in the identification he has made.
Confidence Statements –
The identification of a defendant made in open court by a victim or eyewitness is so persuasive that it has, at times, served as the only piece of evidence needed to secure a conviction. Research has consistently shown that the eyewitness’s degree of confidence in his identification at trial is the single largest factor affecting whether observers believe that the identification is accurate.23 In other words, the more confidence the eyewitness exudes, the more likely a juror will believe that the identification he made is accurate.
Yet research has also shown that a witness’s confidence in his identification is malleable, and susceptible to influences and suggestion, which can be unintended and unrecognized.24 Typically, these changes to witness memory occur after the administrator provides some form of feedback, either confirming or disconfirming, to the eyewitness after the identification has been made.
In the case of an accurate identification, confirming feedback will serve to bolster the eyewitness’s confidence, which in turn will likely sway jurors toward a justifiable conviction. However, when confirming feedback is provided to an eyewitness who has incorrectly identified an innocent person, the feedback can be dangerous. A study that examined the effects of feedback found that post-identification feedback produced “strong effects” on the witnesses’ reports of a range of factors, from overall certainty to clarity of memory.25
For many reasons, however, eyewitnesses cannot be expected to completely refrain from discussing their identifications before they appear at trial. To remedy the potential harm to an innocent person misidentified by an eyewitness, whose confidence might be bolstered by confirming feedback, it is recommended that a “confidence statement,” in the witness’s own words, be taken at the time that the identification is made.
A recent meta-analysis of post-identification feedback also emphasized the need to secure a confidence statement at the time that an identification is made.26 A separate study also demonstrated that the relationship between confidence and accuracy can be enhanced by asking for a confidence statement before any feedback about the identification is provided.27
Substance of Recommendations.
In order to diminish the possible harm wrought by confirming feedback, it is recommended that the administrator seek and document a clear statement from the eyewitness, at the time of the identification and in the eyewitness’s own words, that indicates the eyewitness’s confidence in the selection he has made during the course of an eyewitness identification procedure.
Further, if at the time of the procedure, the eyewitness identifies a person as the perpetrator, it is critical that the eyewitness not be provided any information concerning the selection he has made before the administrator obtains the eyewitness’s confidence statement about the selection.
Electronic Recording of the Identification Procedure- Explanation.
In order to provide courts with access to the most reliable evidence of any possible eyewitness identification, identification procedures must be electronically recorded.
Creating an electronic record of the identification procedure is important because the precise communications, both verbal and non-verbal, made by both the eyewitness and administrator are critically important to investigator and juror assessments of the accuracy of any identification. There are various steps in the identification procedure where an administrator could, with absolutely no prejudicial intent, provide subtle verbal or non-verbal cues to the eyewitness that would affect the eyewitness’s identification of a person as a suspect, or the eyewitness’s confidence in any identification made. The eyewitness’s specific statement of certainty in any identification, made upon an identification, is also critical. Without a record of what transpired during an identification procedure, such communications may be missed, or misinterpreted.
Creating an electronic record of the identification procedure aids law enforcement. A criminal investigation requires that the investigating officer amass a large amount of complex information and track a number of details. Additional information that might not have seemed pertinent during an identification procedure may prove fruitful later in the investigation, but this is only possible if a record is kept. In addition, a fair and just identification procedure that is preserved for the record can withstand any attacks raised at trial.
Video records of well-executed identification procedures can also be used to train other law enforcement officers. A video record can, over time, help demonstrate the proper technique for filler selection, blind administration, the provision of instructions, and other best practices.
Finally, the knowledge that these types of procedures are being recorded boosts public confidence in the criminal justice process. Simply put, creating an electronic record of eyewitness identification procedures provides everyone with the best evidence of what specifically transpired during those identification procedures.
Electronic Recording of the Identification Procedure- Recommendations.
Ideally, the identification procedure will be electronically recorded, using video equipment. If a video record is impracticable, the officer conducting the identification procedure should document, in writing, the reason a video recording could not be made, and audio record the identification procedure. If neither a video nor audio record is possible, the officer conducting the identification procedure should document, in writing, the reasons why and a written record of the identification procedure should be made, and then closely document, in writing, all aspects of the identification procedure.
Regardless of the type of record undertaken, the entire identification procedure for each eyewitness, beginning when the eyewitness enters the identification room (or other venue for viewing identification procedures) and ending when the eyewitness has completed the eyewitness identification procedure and left the procedure room or location, should be preserved for the record. Specifically, this should include:
1. All identification and non-identification results obtained during the identification procedure, including the eyewitness’s confidence statement. (If in writing, these should be signed by the eyewitness);
2. The names of all persons present at the identification procedure;
3. The date and time of the identification procedure;
4. Any eyewitness identification(s) of (a) filler(s);
5. In a photo array or live lineup, identification information and the sources of all photographs or persons used;
6. In a photo array, the photographs themselves*; and
7. In a live lineup, a photo or other visual recording of the lineup that includes all persons who participated in the lineup*.
* In the course of the identification procedure, if the lineup members (either photos or live members) are presented to the eyewitness in a sequential format, the order in which the lineup members are shown to the eyewitness should be preserved for the record.
The Sequential Presentation - Explanation.
In a sequential presentation of an identification procedure, live lineup or photo array members are presented to the eyewitness one-by-one (sequentially), rather than all at once (simultaneously). This presentation format reduces the eyewitness’s use of relative judgment by requiring the eyewitness to compare each photo or live lineup member to his own memory of the perpetrator rather than choosing from among the photos or lineup members the individual who most resembles his memory of the perpetrator.
An important note about sequential identification procedures is that they can actually decrease the likelihood of a correct identification if they are not accompanied by a “blind administrator.” Therefore, “sequentially” presenting photo arrays or live lineups is a recommended reform only when teamed with a blind administrator.
The Sequential Presentation – Scientific Support.
When combined with a blind administrator, the sequential presentation of photographs or live lineup members has been shown to significantly increase the overall accuracy of eyewitness identifications. In order to reduce the prevalence of false identifications, academic research has pointed to the importance of a sequential presentation. Presenting photographs or lineup members sequentially, as opposed to simultaneously, deters the eyewitness from making a “relative judgment,” i.e. selecting from among the photographs or lineup members the person who most resembles her memory of the perpetrator. When photo array or live lineup members are presented sequentially, the eyewitness is more likely to assess the resemblance of each person against her memory of the perpetrator, and is less likely to simply make a relative judgment across all members of the identification procedure.28
There is a strong body of scientific support favoring the sequential over simultaneous presentation of photo array and lineup members. Research indicates that while presenting members sequentially versus simultaneously results in a loss of some correct identifications,29 there is a concomitant – and significantly greater - reduction in incorrect identifications. Thus, sequential presentations greatly increase the overall accuracy of eyewitness identifications.
In further support of sequential presentation of photo arrays and live lineups, a recent meta-analysis that combined the scientific results of thirty individual tests indicated that when crimes and police procedures are simulated realistically, the difference between correct identification rates for sequential and simultaneous identification procedures are small, if nonexistent.30 The decrease in incorrect identifications was found to remain significant, however. In other words, use of the sequential presentation, especially when it mirrors real world conditions, significantly reduces false identifications.31 Research has also shown that correct identifications are also diminished—albeit at a considerably lower rate Therefore, we recognize that this is a policy determination and offer the reform as a discretionary option.
Further supporting the superiority of the sequential presentation in protecting the innocent was a recent field study conducted in Hennepin County, Minnesota which sought to test the efficacy of sequential presentation in Minneapolis and three suburban Minnesota communities. The field study found a significantly lower rate of non-suspect (filler) identifications than the rate from the comparative data set (which combined both field and laboratory results from previous experiments). This demonstrated that the sequential presentation, in the context of real world conditions, produced protections to the innocent not provided by the traditional, simultaneous method.32
These results differ absolutely from those offered by a well-publicized, but severely flawed, evaluation that was recently conducted in three Illinois jurisdictions, including Chicago.33 The “Illinois Report,” which is an “evaluation” of the sequential identification procedure that was written not by a scientist, but by the General Counsel to the Chicago Superintendant of Police34, suggests that a sequential presentation of identification procedures does not work in practice. Because the Report was not published in a scientific publication, however, it was not peer-reviewed by social scientists. Nonetheless, a host of social science researchers and practitioners have questioned the report’s assertions, assumptions, and methodology, indicating that the Report’s flaws undercut the reliability and validity of results necessary to raise legitimate questions about sequential presentation.35
Most notably, a recent peer-reviewed publication disseminated the findings of a blue ribbon panel, composed of leading social science researchers and including a Nobel Laureate, which concluded that the methodology employed by the Illinois Report evaluators rendered its conclusions unreliable. The blue ribbon panel noted that “The confound (failing to properly isolate the variables) has devastating consequences for assessing the real world implications of this particular study” and that, therefore, “(t)he results to not inform everyday practice in a useful manner.”36 The Illinois Report is noted in this resource guide only because it has been used by opponents of the reform as the basis for rejecting sequential presentation.
It is worth noting further that following (and despite) publication of the Illinois Report, the California Commission on the Fair Administration of Justice, comprised of a range of stakeholders from all corners of the criminal justice community, reviewed best practices relating to eyewitness identification and recommended the sequential presentation format for use in the entire state. The Innocence Project understands the confusion caused by the Illinois Report and, to help address concerns about the benefits of sequential lineup presentations, supports field studies of sequential lineup procedures which employ the use of laptop computers and blind administrators. We hope this research will further clarify the benefits of employing sequential lineups in actual practice and resolve any lingering questions about sequential presentations.
The Innocence Project continues to provide all interested parties with substantial information on the scientific research about sequential presentation and the experiences of jurisdictions that are effectively using the practice, including studies, protocol, law enforcement references, and other information.
A Discretionary Option:
The Sequential Presentation - Recommendations.
Sequential presentation of photo array and live lineup members —when combined with blind administration— is an important reform to increase the accuracy of eyewitness identification procedures.
The sequential presentation should always be accompanied by a blind administrator.
Lineup members, both photographs and live individuals, should be shown to eyewitnesses one at a time.
The police suspect should never be the first person presented in a sequential identification procedure.
Photo array and live lineup members should be shown to the eyewitness in random order.
Even if an eyewitness makes an identification before all of the photos or live lineup members are shown, all of the photos and lineup members should be shown to the eyewitness.
Eyewitnesses should not be aware of how many individual photographs or live lineup members will be shown to him or her during the course of the identification procedure.
If, in your particular jurisdiction, the political reality is such that use of the sequential presentation will make adoption of eyewitness identification reforms difficult or impossible, the other five non-controversial reforms recommended in this guide should be put into practice without delay.
Misidentifications have contributed to over 75% of the nation’s 220 wrongful convictions proven by DNA evidence. These eyewitness identification reforms, thoroughly supported by a large body of peer-reviewed, scientific research, must be implemented in order to eliminate the risk of misidentification posed by traditional eyewitness identification procedures that cause irreparable damage to criminal investigations and prosecutions. Indeed, there are a range of jurisdictions that have already implemented these reforms and reported favorably upon them.
The Innocence Project is available to explain any of these reforms in greater detail.
We have and are willing to provide interested parties with the scientific literature on the matter, recommendations from leading national organizations, and sample protocols from practicing jurisdictions. If you would like any of this information, or have any other questions about eyewitness identification procedures, please contact Policy Analyst Rebecca Brown at RBrown@innocenceproject.org or (212) 364-5360 or Staff Attorney and Eyewitness Fellow Ezekiel Edwards at ZEdwards@innocenceproject.org or (212) 364-5349.