By Sheri H. Mecklenburg, Assistant U. S. Attorney, Northern District of Illinois, Chicago, Illinois; Patricia J. Bailey, Assistant District Attorney, Chief of the Special Litigation Bureau, New York County District Attorney’s Office–Manhattan, New York, New York; Mark R. Larson, Chief Deputy, Criminal Division, King County Prosecuting Attorney’s Office, Seattle, Washington
he police properly view their role as detecting what crime has been committed and who, in fact, committed it and who, in fact, did not. The investigative stage may be controlled by legal rules but it is not part of the adversary system and the allegiance of the police is to the truth. They have no mandated role to play for the prosecution or the defense . . . the investigator’s duty is not to present the best case, it is to present the truth.”1 So, it is incumbent upon police to adopt investigative procedures, including eyewitness identification procedures, best designed to get at the truth.
|Sequential, Double-Blind Method|
In September 2010, the IACP National Law Enforcement Policy Center released the Model Policy on Eyewitness Identification, which advocates the use of the sequential, double-blind method for all photo arrays and lineups. This policy, along with the accompanying Concepts & Issues Paper, can be purchased by visiting www.theiacp.org/policycenter or by contacting the Policy Center at firstname.lastname@example.org or 800-843-4227 x 319.
In the October 2008 issue of Police Chief, the authors challenged the law enforcement community to take on the task of improving eyewitness procedures, protocols, and training by including the adoption of standardized instructions, written protocols, and in-depth training.2 Some jurisdictions took that call to heart and have reviewed and revised their identification procedures, policies, and training.
Nevertheless, recent survey research demonstrates that many police departments still have not adopted standard instructions, written protocols, and in-depth training for eyewitness identification procedures.3 The purpose of this article is to update readers on the most recent scientific information regarding the sequential procedure so that police chiefs can objectively evaluate the best way to improve eyewitness identification procedures.
|New York State Experience in Developing Statewide Identification Protocols|
In 2009, the New York State District Attorneys Association formulated a Best Practices Committee. The 25-member committee consists of both elected district attorneys and experienced assistant district attorneys from across the state. The committee members represent the state’s regional diversity so as to ensure a comprehensive understanding of the issues that impact New York State’s criminal justice system.
The Best Practices Committee’s goal is to develop best practices for law enforcement, to review innovative initiatives, and to analyze the causes of wrongful convictions and the various recommendations surrounding that issue. In its first year, the committee began to review a wide variety of topics including videotaping interrogations, forensic evidence, and identification procedures. The first project completed by the committee, in collaboration with police agencies around the state, resulted in the New York State Photo Identification Guidelines.
The committee reviewed the identification procedures used in more than 550 police departments around the state; departments which range in size from the New York (city) Police Department’s 35,000 officers to agencies having 10 full-time officers. During this initial phase it became clear that practices and resources vary widely across the state. For example, a few jurisdictions predominately relied on the use of live lineups, while others almost exclusively used photo identifications. In addition, very few departments had written protocols, and training was sporadic. From this disparity in practices and the fact that identification issues have played a significant role in wrongful convictions, it became clear that there was a need for law enforcement to lead the way to formulate standardized best practices. Protocols that are self-generated and voluntarily implemented are far more likely to be successful than procedures mandated by legislation or executive fiat.
The committee worked to create identification procedures with multiple objectives in mind: accurate identifications of guilty persons, respect for the rights of the accused, protection of witnesses and consideration for the practical realities imposed by location and funding. Many good ideas were found not only through New York’s law enforcement partners, but also by a review of identification procedures in other states. Laboratory research on eyewitness identification was also part of the analysis. Social scientists in the field were consulted. Any procedure used to elicit eyewitness identifications must be fair, reliable, and free of influence—intentional or unintentional.
The committee’s first draft of the guidelines was circulated among all of the state prosecutor offices and various New York police agencies. Agencies across the state gave valuable input. In May 2010, the New York State District Attorneys Association, New York State Police, New York State Sheriff’s Association, the New York State Association of Chiefs of Police, New York City Police Department, and other law enforcement agencies announced the voluntary adoption of the first set of standardized identification procedures for the State of New York.
This unique collaboration amongst law enforcement agencies at all levels of government across the state was the first time in the United States that law enforcement agencies proactively worked on a statewide basis to establish best practices for identification procedures. However, the process did not stop there.
Unlike legislated procedures that do not allow for improvement, voluntary guidelines allow for growth. As a result of suggestions received and concerns raised after the release and implementation of the procedures, the guidelines have already been updated. In addition, everyone involved recognized training as a critical component to the successful implementation of the procedures. The first standardized training on identification procedures in the state was developed and included daylong training sessions throughout the state. In many instances it was the first classroom training some experienced officers have had since the academy. Training videos were created to allow quicker and easier access to instruction on the guidelines.
The Latest Research on the Sequential Lineup
The sequential procedure does not merely consist of showing the witness one photo at a time. Instead, the procedure is a package with the sequential display being the centerpiece.4 The package requires witnesses to make an identification choice of “yes,” “no,” or “I don’t know” for each photo before seeing the next photo. It also requires that the lineup administrator be “blind” to the identity of the suspect—the “double-blind” component—because the sequential presentation increases the opportunity for the administrator to cue or suggest to the witness as compared to the traditional “six-pack” photo array.
The claim of sequential superiority rests on the contention that the sequential method reduces the number of false or filler identifications as compared to the traditional photo arrays. Advocates say the advantage comes from the use of “absolute” over “relative” judgment. The idea being, when faced with a sequential presentation, witnesses compare each photo to their own memory, and make a distinct or absolute decision for each photo. On the other hand, according to this theory, witnesses viewing an array of photos will merely identify a lineup member who most resembles the perpetrator, relative to or compared with the other photos in the array.5 Recent research suggests that reduction in filler identifications in the sequential procedure is not the product of absolute judgment, but rather is due to a criterion shift, or the point at which a witness is willing to identify someone. Thus witnesses will use a stricter standard before making an identification in a sequential lineup—making it less likely for witnesses to make any identification, accurate or false.6
In 2006, McQuiston-Surrett and others, determined that nearly half of the experimental studies used as support for the claims of sequential superiority were unpublished experiments, many of which consisted of undergraduate student projects with unknown methodologies. When McQuiston-Surrett and others, reconstructed the data using only published, peer-reviewed studies, the results were quite different—the sequential method showed no advantage. The implications of these findings caused these researchers to warn that, “it could be difficult to convince legislators, lawyers, judges and juries when fully 40 percent of the knowledge about [sequential lineups] derives from unpublished undergraduate projects and academic colloquia and has not been subject to peer review.”7 They further observed that “outside of the corpus of published studies emanating from the single laboratory, there is no evidence that [sequential lineups] are in overall terms superior to [simultaneous lineups].”8 These findings make it clear that “the research base for [sequential lineups] may not be sufficiently developed to currently advocate for its implementation to the exclusion of other procedures.”9
Then in 2008, Clark and others published a review of the original studies that reveal that the studies showing a sequential advantage were biased to elicit false identifications in the traditional photo arrays.10 Those experiments replaced the mock perpetrator’s photo with that of an innocent suspect chosen specifically because he most looked like the perpetrator. Precisely because the innocent suspect was chosen for his close resemblance to the perpetrator, while the other photos were chosen for their general match to the description, the innocent suspect stood out from all the other photos. As a result, they found that there was a bias built into the experiment for the innocent suspect to stand out among the photos in the simultaneous photo arrays, thus creating a higher rate of false picks in the traditional, simultaneous photo arrays. Subsequent research confirmed this flaw in the initial studies.11 When Clark and others—who do not advocate any one method of presentation—reconstructed the experiments without using biased lineups, the results demonstrated that although the sequential procedure still suppressed the number of accurate identifications, there was no difference in the number of false identifications between the sequential and simultaneous methods.12
The following year, Gronlund and others conducted 24 experiments comparing sequential and simultaneous lineups: 19 experiments showed no difference between the two methods, 3 showed a simultaneous advantage, and 2 showed a sequential advantage.13 Notably, the 2 experiments that showed a sequential advantage did so when compared to traditional lineups, where the suspect stood out. The experimental results led the researchers to conclude the sequential advantage likely emerged from the repeated use of biased photo arrays, particularly since the majority of the experiments were merely replications of the original flawed studies.14
This most recent laboratory research casts doubt on the validity of the claim of a sequential advantage.15 But, even if there is no advantage, when faced with the call to convert to sequential procedures, is there any downside to adopting the sequential procedure? The Police Chief October 2008 article discusses some of the disadvantages to the real-world application of the sequential procedure. Any police chief who wants to make an informed decision on whether to adopt the sequential identification procedure must also consider the potential impact of these disadvantages.
First, it is well-documented that the sequential procedure results in the loss of accurate identifications.16 But, even putting aside the loss of accurate identifications, there are other concerns with the sequential procedure. For example, the sequential method is potentially more suggestive than the simultaneous photo array, since the administrator of the lineup knows exactly which individual photo the witness is viewing at any given moment.17 In comparison, the administrator of a traditional photo array does not know which photo the witness is actually viewing. Thus, the opportunity to cue a witness is built into the very structure of the sequential procedure. It is essential that a blind administrator conduct the sequential procedure. As such, police departments adopting the sequential procedure, but making blind administrators optional due to resource concerns, should limit the use of the sequential lineup to instances where a blind administrator is available.
|The Blind Administrator|
In the October 2008 Police Chief article the authors discuss the movement to adopt a blind administrator for lineups, meaning that an investigator who does not know the identity of the subject conducts the lineup. The eyewitness researchers urge adoption of a blind administrator to prevent “inadvertent influence” by the lineup administrator, but have not identified what kind of inadvertent influence systematically occurs, such that the witness notices, understands and is willing to follow. The research on the use of a blind administrator remains scarce. “[L]ittle research has been conducted on how a lineup administrator’s knowledge might bias identification accuracy.” Greathouse and Kovera, “Instruction Bias and Lineup Presentation Moderate the Effects of Administrator Knowledge on Eyewitness Identification,” Law and Human Behavior, 33, 70-82 (2009). Surprisingly, of all the recommendations made by experts to increase the reliability of eyewitness identification, only the recommendation to use a blind procedure has been made “without any eyewitness identification research to support their position.” Ibid., 71.
The recommendation for a blind administrator is drawn from experimenter expectancy research, meaning that experimenters conducting research can communicate their expectations and thereby influence the responses of the research subjects. Ibid., citing Rosenthal, Experimenter Effects in Behavioral Research (New York: Irvington Publishers, 1976), which found that graduate students who believed that their rats will learn to run a maze quickly or slowly may have produced a behavior change in the rats. However, eyewitness researchers have acknowledged that “[t]he lineup setting is different from the traditional setting in which experimenter expectancy effects have been observed.” Greathouse and Kovera, 72. In the eyewitness context, the witness has an independent memory for the perpetrator, accurate or not, which may limit the ability of the investigator to influence the witness toward or away from a particular photo, particularly if the photo triggers, or fails to trigger, the witness’s memory for the offender. The limited research on blind administrators has failed to demonstrate that witnesses are influenced by administrators in the same manner as subjects of the experimenter expectancy research. See Phillips et al., “Double-blind Photoarray Administration as a Safeguard against Investigator Bias,” Journal of Applied Psychology 84, no.6 (December 1999): 940-951, finding that administrator knowledge influenced the witnesses to choose an innocent suspect only when a sequential lineup was administered and only in the presence of an observer; and Haw and Fisher, “Effects of Administrator-Witness Contact on Eyewitness Identification Accuracy,” Journal of Applied Psychology 89, no. 6 (December 2004): 1106-1112, finding that student administrators were able to influence identifications only where the contact with the witness was high and prolonged—but not in those identification procedures where the contact was low.
Without any science to support it, the use of blind administrators is a pure policy decision, dependent on resources and perception. The use of blind administrators makes the job of prosecutors easier, but the dwindling manpower faced by police chiefs is a very real concern; every jurisdiction that has used blind administrators, to the authors’ knowledge, has had to settle for far less than full compliance. Still, there are other ways to reduce influence during the eyewitness procedure, such as positioning the computer screen so that only the witness can see the displayed photos or having the administrator positioned so that he or she is outside the view of the witness and, most importantly, the adoption of proper witness instructions and administrator protocols. Police chiefs are encouraged to write these methods into their protocols and to implement training on this important issue.
Even advocates caution that sequential presentation can actually increase the number of false identifications in the following five instances: (1) when the witness is 65 or older; (2) when the witness is 12 or younger; (3) when the crime involves multiple offenders; (4) when an offender has changed his appearance; and (5) when the identification is cross-race, meaning that the eyewitnesses are identifying a perpetrator of a different race.18 From a protocol standpoint, agencies adopting the sequential procedure should be evaluating each case and witness to determine which method is “the best” to employ under the circumstances. A blanket sequential policy that fails to account for these issues can result in a significant number of false identifications.
The Police Chief article also raised the legal issues created by the sequential procedure when the suspect is in the first position, since an identification means that the witness has essentially viewed a one-person array or with live lineups, a one-person show-up.19 These concerns remain unresolved. Sequential advocates continue to suggest that this criminal justice dilemma can be avoided by showing the witness the remaining photos in the photo array even though an identification has already been made. Such a procedure has not been studied, and its effects are unknown—that is, does the witness actually evaluate additional photographs after making an identification, or does the witness perceive this as a signal that his choice is “wrong,” or that the police do not trust his judgment. This procedure makes the witness vulnerable to cross-examination because the witness identified the first photo shown and might not have genuinely looked further, or was uncertain of his identification if he actually continued to look for the offender.
Sequential advocates further suggest as a solution to this criminal justice issue that the suspect never be placed in the first position, but this suggestion only shifts the problem to the next position.20 Eliminating position one as a viable option also decreases the size of the lineup, since the procedure does not begin in earnest until the second photo. Similarly, placing the photo in a later position raises concerns—what happens when witnesses gets to photograph number five or six and realize that they are running out of options? To eliminate this problem, sequential advocates suggest fooling witnesses into believing that there are more photos to be seen by holding a stack of “fake” photos at the end that are never intended to be shown. However, conducting procedures which require “fooling” witnesses may have implications for the public trust in law enforcement and adversely affect the cooperative relationships police foster with victims and witnesses.
Perhaps the real question is, if law enforcement has to keep making adjustments to address legal and practical problems with the sequential procedure, can it truly be a superior method, and one that is ready to be implemented as the national standard?
The Latest Field Study: Austin, Texas
In 2012, the American Judicature Society released the report, authored by three sequential advocates, on the first-phase of a field study comparing the sequential identification method with the simultaneous identification method.21 The AJS study originally included four jurisdictions, but three of the four dropped out during the course of the study.22 Charlotte-Mecklenburg, North Carolina, dropped out of the study because in 2008 its state legislature mandated the statewide use of the sequential method. Both Tucson, Arizona, and San Diego, California dropped, out because of difficulties with the computer-driven identification procedures, which were expressly designed for the study. In fact, the computer-driven procedures were “somewhat clumsy and took longer to use than simply printing the photos and administering the lineup in the traditional way.”23 The possibility of computer-driven lineup procedures is still a worthwhile endeavor, which should be explored further, and the AJS study should be commended for a first try.
The AJS study presents a number of issues that make it unpersuasive as a basis for policy decisions on the sequential method. For example, police collected 855 photo lineups: approximately 75 percent coming from Austin (615); 16 percent from Tucson (144), 6 percent from Charlotte-Mecklenburg (53) and 5 percent from San Diego (43). However, of those 855 photo lineups, 358 lineups had to be eliminated because the procedures used did not fit the study protocols, leaving only 497 included in the analysis. Of the 358 eliminated lineups, approximately 59 percent were rejected because a blind administrator was not used. That such a large number of lineups had to be excluded, without explanation as to why so many could not be conducted by a blind administrator, raises questions about the practical difficulty of implementing blind administrators.
The AJS Protocols
Some of the protocols of the AJS study are similar to those that the article in October 2008 Police Chief urged police to adopt, such as documenting the certainty, in the witness’s own words, of any identification, and instructing the witness that the “perpetrator may or may not be in the lineup.” However, several protocols that appear to be unique to the AJS study are not only untested by research, they are also troubling for the criminal justice system. For instance, regardless of presentation method, all witnesses were asked whether any of the photos “looked familiar.” If the witness selected a photo, the computer program asked “whether the person was familiar for reasons related to the crime or unrelated to the crime.” Familiarity is a far lower standard than recognition. For the criminal justice system, a witness’s ability or inability to recognize the suspect as the perpetrator is the critical and relevant issue; a witness’ familiarity with a photo is not.
Under the AJS study protocols, the suspect never appeared in position one of the photo lineup. This protocol essentially concedes the problem of identifying the first photo in a sequential procedure, shifted the problem to the next position while reducing the size of the lineup.
The protocols further required witnesses be told that the lineup contains only one possible suspect, effectively prompting the witness. A witness who is told that there is only one right answer might squelch information about a second photo in the lineup, information that could reflect on the witness’s reliability or add information to the investigation. A witness’s claim to recognize more than one photo—be it fillers or the suspect and a filler—is critical information in assessing the reliability of any eyewitness identification and for Brady disclosure obligations.
The study’s protocol requires witnesses who made an identification to continue to review the remaining photos to see if anyone else “looked familiar.” In the sequential procedure, this was accomplished by having the witness continue to sequentially look through the remaining photos. In the traditional photo array, witnesses were required to return to the array to determine if anyone else “looked familiar.” Such a procedure makes the witness vulnerable to cross-examination about what the witness actually looked for after making an identification. The witness is told initially that he will be returned to the photos even if an identification is made. The effect of these instructions has never been scientifically tested.
A number of the witnesses in the AJS study selected both a filler and the suspect as being familiar. The blind administrators were required to listen to the audio tape of the procedure and attempt to disambiguate whether the witness preferred a filler or the suspect. Police “disambiguation” of a witness’s identification is problematic for the criminal justice system, both at the investigative and trial stages. First, if the witness’s decision is ambiguous, it is just that, ambiguous. The investigation continues from that point with that information. Second, having officers substitute their subjective beliefs as to whom witnesses “preferred,” not only exposes officers to cross-examination, it casts doubt on the reliability of the witness identification.
The authors of the AJS study found that the sequential procedure was superior to the simultaneous procedure in terms of obtaining fewer filler identifications.24 Despite historical evidence that somewhere between 8 and 15 percent of accurate identifications are lost in the sequential procedure, the AJS study found that there was no loss of accurate identifications in the sequential procedure compared to the simultaneous procedure.25 This finding must be analyzed further and duplicated, before this result can be accepted as meaningful.
The study also found that, among the no identification decisions, there were many more “not sure” results in the sequential procedure than in the simultaneous showings, with 19.2 percent of the no identifications being “not sure” with the simultaneous lineups, and 46.5 percent of the no identifications being “not sure” with the sequential lineups. The study’s authors argue that the high number of “not sure” responses is an advantage for the sequential method since it is “[b]etter to get no identification than a filler identification because it keeps the witness unspoiled for a possible new lineup later.” Law enforcement is not interested in “saving” a witness for multiple identification attempts. Police and prosecutors are interested in finding out the truth: in knowing what information the witness has, in knowing who the offender is, and whether a specific witness can make a reliable identification.
Many of the issues with the sequential lineup remain unresolved almost 30 years after its inception. As one of the leading advocates of the procedure admits, “little is known about many issues” surrounding the sequential procedure, but nevertheless he and others urge the adoption of the sequential lineup as a matter of policy.26 Other scientists, however, recognize it is inappropriate to adopt sequential lineups as the only, or even recommended, procedure, when “little is known” about the unresolved issues surrounding this method.27
The results of this study are inconsistent with the laboratory research. The results also differ from those of the Illinois study, the only other real-world field study comparing sequential and simultaneous procedures.28
Random assignment was a main protocol of the Illinois study, and all three participating jurisdictions complied with this mandate.29 The initial report of the AJS study inaccurately alleges that “the Illinois study acknowledged that some unknown number of filler identifications for the simultaneous procedure was not recorded in the results because the nonblind detective decided that the witness was not sure enough in the identification.”30 Given that the results from the Illinois study are largely consistent with the most recent laboratory research, and that both the Illinois data and the most recent laboratory research conflict with the AJS study, the need for an objective and scientific peer review of the AJS raw data is all the more critical before its value can be determined. Until then, the AJS study should not be considered a serious basis for policy decisions.
Until a host of issues surrounding the sequential procedure are resolved and the results of the AJS study are peer reviewed and replicated, law enforcement should not be pushed into adopting the sequential procedure.31 The police, in the quest for the truth, must demand that these issues be resolved before any wholesale change to policy.
Updated research has called into question the validity of the claimed sequential advantage. The urge to do something to reduce mistaken eyewitness identification should not lead to blindly adopting a procedure that has unresolved issues and many disadvantages. Following public policy mandates that do not aid in the quest for the truth can be a costly mistake.♦
|Editor’s Note: This article is distinct from the articles normally published in the Police Chief magazine. Traditionally, the magazine’s articles are pragmatic, giving police executives practical advice on how to handle issues and problems.|
Eyewitness identification is the subject of much debate, and there are many strong and divided opinions on the subject. The purpose of this article is to add to the knowledge and information currently available to help readers reach their own conclusions. This article is based on the opinions of the authors and does not necessarily reflect the views of the Police Chief magazine or the IACP.
This article was written by the authors in their personal roles and represents the authors’ views only. This article does not reflect any position, policy, opinion, or view of their employers, including the United States Attorneys’ Office, the United States Department of Justice, or any other agency or organization.
1James B. Zagel, “Getting to the Truth before It Falls into the Hands of the Lawyers: Pursuing Accuracy in Criminal Cases,” Loyola School of Law Public Interest Law Reporter 11, no. 2 (Summer 2006): 17 et. seq., 34-35. Judge Zagel was the former Director of the Illinois State Police before being appointed to the federal bench.
2Sheri H. Mecklenburg et al., “Eyewitness Identification: What Chiefs Need to Know Now,” The Police Chief 75, no. 10 (October 2008): 68-81.
3Police Executive Research Forum, A National Survey of Eyewitness Identification Processes in Law Enforcement Agencies (March 8, 2013), http://policeforum.org/library/eyewitness-identification/NIJEyewitnessReport.pdf (accessed June 18, 2013).
4R.C.L. Lindsay and Gary Wells, “Improving Eyewitness Identifications from Lineups: Simultaneous versus Sequential Lineup Presentation,” Journal of Applied Psychology 70, no. 3 (August 1985): 556-564. See Jamal K. Lindsay et al., “Beyond Sequential Presentation: Misconceptions and Misrepresentations of Sequential Lineups,” Legal and Criminological Psychology 14, no. 1 (February 2009): 31, in which the creator of the sequential procedure explains that a sequence requires multiple elements.
5Several studies have examined the issue of relative versus absolute judgment as a product of the presentation method and found that the subjects do not uniformly use one strategy over the other just because of the type of lineup method employed. See, e.g., Wendy Kneller et al., “Simultaneous and Sequential Lineups: Decision Processes of Accurate and Inaccurate Eyewitnesses.” Applied Cognitive Psychology 15, no. 6 (November 2001): 659-671, in which participants were just as likely to report the use of an “absolute” strategy whether shown a simultaneous lineup or a sequential lineup.
6Christian A. Meissner et al., “Eyewitness Decisions in Simultaneous and Sequential Lineups: A Dual-Process Signal Detection Theory Analysis,” Memory and Cognition 33, no. 5 (July 2005): 783-792. For an early discussion recognizing the questionable validity of the absolute versus relative judgment theory, see Ebbe Ebbesen and Heather Flowe, “Sequential versus Simultaneous Lineups: What Do We Really Know?” (2001), http://www2.le.ac.uk/departments/psychology/ppl/hf49/SimSeq%20Submit.pdf (accessed June 6, 2013). See also Lindsay et al., “Sequential Lineup Presentations,” 15. Both the relative judgment and conservative criterion processes may be at work in the identification process.
7Dawn McQuiston-Surrett et al., “Sequential versus Simultaneous Lineups: A Review of Methods, Data and Theory,” Psychology, Public Policy and the Law 12, no. 2 (May 2006): 137-169, 139, http://eyewitness.utep.edu/Documents/McQuiston-Surrett&06SequentialVsSimultaneousLineups.pdf (accessed June 6, 2013).
8Ibid. at 141, 155. The single laboratory referred to is that of Professor Rod Lindsay, who along with Gary Wells, gave birth to the sequential procedure concept in 1985, and remain its staunch advocates almost 30 years later. See Lindsay and Wells, “Improving Eyewitness Identifications from Lineups.”
9McQuiston-Surrett et al., “Sequential versus Simultaneous Lineups,” 162. In an attempt to address these concerns, the primary advocates of the sequential procedure conducted their own review in 2011 and still concluded that the research showed a sequential advantage. Nancy K. Steblay et al., “Seventy-two Tests of the Sequential Lineup Superiority Effect: A Meta-Analysis and Policy Discussion,” Psychology, Public Policy, and Law 17, no. 1 (February 2011): 99-139. However, this review included unpublished studies and also omitted published studies, which showed no sequential lineup advantage.
10Steven E. Clark et al., “Regularities in Eyewitness Identification,” Law and Human Behavior 32, no. 3 (June 2008): 187-218.
11Curt Carlson et al., “Lineup Composition, Suspect Position and the Sequential Lineup Advantage,” Journal of Experimental Psychology: Applied 14, no. 2 (June 2008): 118-128.
12For a more in-depth discussion of flaws in the original sequential research, see Roy S. Malpass et al., “Public Policy and Sequential Lineups,” Legal and Criminological Psychology 14 (2009): 1-12, and Roy Malpass et al., “Response to Lindsay, Mansour, Beaudry, Leach and Bertrand’s Sequential Lineup Presentation: Patterns and Policy, ” Legal and Criminological Psychology 14, no. 1 (February 2009): 25-30.
13Scott Gronlund et al., “Robustness of the Sequential Lineup Advantage,” Journal of Experimental Psychology: Applied 15, no. 2 (June 2009): 140–152.
14Ibid. at 141, “it is possible that researchers have been finding similar results [a sequential advantage] because researchers have been conducting similar experiments.”
15An increasing number of scientists now recognize the doubt surrounding the claims of sequential superiority. Malpass, “Response to Lindsay”; and Laura Mickes et al., “Receiver Operating Characteristic Analysis of Eyewitness Memory: Comparing the Diagnostic Accuracy of Simultaneous versus Sequential Lineups,” Journal of Experimental Psychology: Applied 18, no. 4 (December 2012): 361–376.
16The loss is between 8 and 15 percent. McQuiston-Surrett et al., “Sequential versus Simultaneous Lineups,” 138–39. See also Stephen Clark, “Costs and Benefits of Eyewitness Identification Reform: Psychological Science and Public Policy,” Perspectives on Psychological Science 7, no. 3 (May 2012); Steblay et al., “Seventy-two Tests of the Sequential Lineup Superiority Effect”; and Lindsay et al., “Sequential Lineup Presentations,” citing articles concluding that there is a loss of correct identifications accompanying the sequential procedure.
17 Gary L. Wells et al, “Eyewitness Identification Procedure Recommendations for Lineups and Photospreads,” Law and Human Behavior 22, no. 6 (December 1998): 603–647. See also McQuiston-Surrett et al., “Sequential versus Simultaneous Lineups,” 143, citing Mark R. Phillips et al., “Double-Blind Photoarray Administration as a Safeguard Against Investigator Bias,” Journal of Applied Psychology 84, no. 6 (December 1999): 940–951.
18See Lindsay et al., “Sequential Lineup Presentations,” 14, “Despite the initial positive results with sequential lineups, less encouraging results exist.” Ibid.
19McQuiston-Surrett et al., “Sequential versus Simultaneous Lineups,” 149.
20It is naive to believe that witnesses will not figure out that the first position is not a viable position. Police protocols are not secret, nor should they be. They are a matter of public record, widely known among police, prosecutors, defense lawyers, and the public. Thus, as the protocols become known to the public, witnesses—knowing “they never put the offender in the first position”—will merely ignore that first position, making the second position the first viable option.
21Gary Wells et al., A Test of the Simultaneous versus Sequential Lineup Methods: An Initial Report of the AJS National Eyewitness Identification Field Studies (Des Moines, Iowa: American Judicature Society, 2011), http://www.popcenter.org/library/reading/PDFs/lineupmethods.pdf (accessed June 5, 2013).
24Of course, filler identifications are not the type of error the criminal justice system is concerned with in terms of preventing wrongful conviction. An identified filler will not be arrested, prosecuted, or convicted. Invariably, fillers are known innocents.
25See note 19. Even the authors note that this result “raises questions” in light of the laboratory studies. Gary Wells et al., A Test of the Simultaneous versus Sequential Lineup Methods, 15.
26Lindsay et al., “Sequential Lineup Presentations,” 17.
27Malpass, “Response to Lindsay.”
28 Sheri Mecklenburg et al., Illinois Report to the Legislature on Sequential, Double-Blind Procedures (March 2006), “the Illinois Study” discussed in the October 2008 Police Chief article, http://www.chicagopolice.org/IL%20Pilot%20on%20Eyewitness%20ID.pdf (accessed June 7, 2013).
29Mecklenburg et al., Illinois Report, 25-26. As reported there, Evanston alternated the sequential and traditional procedures according to whether the case number was even or odd. Chicago and Joliet assigned crimes occurring in one high-crime geographical area to implement sequential procedures, and collected data on traditional lineups from an adjacent high-crime geographical area. Thus, the investigators in all three jurisdictions had no control over the assignment of sequential or traditional procedures.
30Each participating police department “recorded” each and every line up procedure conducted as part of the study. All of that information was provided to the two eyewitness researchers who were analyzing the data. Those experts, independent of one another, reviewed all the data and determined which lineups met their protocols for inclusion in the analysis. Addendum to the Report to the Legislature of the State of Illinois: The Illinois Pilot Program on Sequential Double-Blind Identification Procedures, 8, note 3. Similarly, the researchers involved in the AJS Study discarded numerous (358) lineups because they believed the lineups did not meet the protocols of their study. It is inaccurate to say that police in the AJS Study failed to record the results of these lineups, just as it is inaccurate to make such a claim about the police participants in the Illinois Study.
31The updated research brings to mind Judge Zagel’s prescience in 2006 that, “it will have to be the work of people new to the field who experiment and confirm whatever results we find. The current experts in the field may be viewed as having too great an interest in proving that they were right.” Zagel, “Getting to the Truth before It Falls into the Hands of the Lawyers,” 34.
Please cite as:
Sheri H. Mecklenburg et al., "Eyewitness Identification: An Update on What Chiefs Need to Know," The Police Chief 80 (July 2013): 60–67.