By Susan Gaertner, Ramsey County Attorney, Saint Paul, Minnesota; and John Harrington, Chief of Police, Saint Paul, Minnesota

n article in the October 2008 issue of this magazine raises one of the most critical issues in the law enforcement community today—the problem of faulty eyewitness identification—and urges police chiefs to lead the discussion of reform.¹ Highly publicized DNA exoneration cases, most of which involved mistaken eyewitness identifications, have focused public scrutiny on law enforcement procedures and challenged law enforcement practitioners both to reexamine long-standing practices and to implement change that will reduce the likelihood of such misidentifications in the future. No one in the law enforcement profession has any interest in convicting the innocent.

The Mecklenburg article makes eight general recommendations² but is short on practical specifics. It also sidesteps the two essentials in lineup reform that are supported by more than a quarter century of scientific study—that lineups be “blind” (that is, conducted by someone who does not know who the suspect is) and sequential.³ The article argues that both points require more field study and that there are practical impediments to the implementation of these reforms in a law enforcement setting.

The present article, intended as a pragmatic contribution to the reform discussion, describes the successful implementation of a blind sequential lineup protocol in Ramsey County, Minnesota. Its purpose is not to debate the science but to address the practical application of this change in a law enforcement setting and to provide specifics for any other agency interested in doing the same.

Persuaded by its own review of the science (for reasons stated below), Ramsey County decided to move ahead in 2005 with a year-long pilot project to test whether implementation of blind sequential lineups was feasible and practical for law enforcement practitioners. The pilot was successful and, in April 2006, the new lineup protocol was adopted as countywide policy by all law enforcement agencies in the jurisdiction. The Illinois Field Study (described in the Mecklenburg article) notwithstanding, Ramsey County continues to be satisfied that the bulk of scientific evidence—and pragmatic considerations—support this change. After almost four years of experience with this protocol, the county has concluded that it is feasible, practical, and superior to past eyewitness identification procedures.

Lineup reform need not be delayed for years awaiting the outcome of additional field study when the basic reforms make sense objectively and when experience demonstrates that they can be successfully applied in the field.⁴ Implementing this change in protocol has not only proved workable for investigators in Ramsey County but has also improved the uniformity and the quality of eyewitness identification investigations. It has also yielded bonus practical benefits in court.

The county’s experience is adaptable to other law enforcement jurisdictions; hopefully, outlining it here will encourage other jurisdictions to replicate this effort—and perhaps provide some shortcuts for doing so.

Background: The Sutherlin Exoneration

Ramsey County is a jurisdiction of 493,215 residents.⁵ It has 10 separate law enforcement agencies, including the Saint Paul Police Department (the largest agency, currently with 599 sworn officers), the Ramsey County Sheriff’s Department, seven smaller suburban departments, and a campus police department. The Ramsey County Attorney’s Office prosecutes all felony crimes committed in the jurisdiction.

As a result of national discussion of the DNA exoneration cases, the county attorney in March 2001 decided to initiate a systematic postconviction review of cases of all persons convicted in Ramsey County prior to 1995 (when current DNA technology became widely available) to determine if DNA testing could shed new light on the validity of their convictions.⁶ Of the 116 cases identified, three, all from the 1980s, were found in which DNA evidence could be outcome-determinative. In only one was the evidence still available. Testing of that evidence resulted in the exoneration of David Sutherlin for a 1985 rape. The primary evidence against Sutherlin was mistaken eyewitness identification.

In the Sutherlin case, the investigator suspected Sutherlin was the rapist based on the victim’s physical description of her assailant, which matched Sutherlin’s, and the fact that Sutherlin lived close to the location of the crime. His photo was included in a simultaneous photo display, which was shown to the victim by the investigator. The victim said he looked exactly like her assailant. At trial, she could only say, however, that he resembled her attacker. Other evidence presented included that enzymes found on the victim’s clothing were consistent with those of Sutherlin as well as 22 percent of the population and that Sutherlin’s car had a noisy muffler similar to one heard leaving the scene. The defendant was convicted and sentenced to serve time consecutive to an unrelated double murder.⁷

The 2002 DNA testing of the evidence in the 1985 rape case that exonerated Sutherlin also positively identified the real assailant: it turned out he lived in the same neighborhood as Sutherlin and matched the physical description provided by the victim.⁸ Unlike Sutherlin, however, his photo was not on file with the Saint Paul Police Department at the time.⁹

The indisputable evidence of a miscarriage of justice in the Sutherlin case and the failure to retain evidence that would either have confirmed or refuted the validity of two other convictions prompted two reforms in Ramsey County: the development of a countywide evidence retention policy and a reexamination of eyewitness identification procedures.¹⁰

Review of Practice and Science

The traditional photo identification procedure in Ramsey County, as it has been across the United States, was the simultaneous photo array. The typical photo array contained six photos. It was the usual practice for an investigator with knowledge of the case to show the photo lineup to witnesses.

In the wake of the DNA exonerations, the U.S. Department of Justice’s National Institute of Justice (NIJ), under then U.S. attorney general Janet Reno, formed a study group to examine eyewitness identification procedures.¹¹ The panel reviewed what was then 20 years of scientific research on memory and interview techniques, including scientific evidence supporting blind sequential lineup administration. ¹² Its landmark 1999 report began the national law enforcement discussion of lineup reform and made numerous reform recommendations (many of which parallel those in the Mecklenburg article).¹³ The panel acknowledged the scientific evidence at that time showing eyewitness identification to be more reliable when photos were shown sequentially and blind procedures were used.¹⁴ However, operating on consensus, the report presented both sequential and simultaneous as acceptable alternative procedures and stopped short of including blind administration in its final recommendations, instead stating that it “may be impractical for some jurisdictions to implement.”¹⁵

The first meta-analysis assessing all laboratory studies comparing the simultaneous and sequential lineup formats came in 2001.¹⁶ It showed that identification errors are significantly more likely (up to three times higher) in a simultaneous format; conversely, the sequential procedure produces fewer misidentifications. The sequential format also produces fewer correct identifications overall. However, the reduction in correct identifications using this method is offset by the much larger reduction of incorrect identifications.¹⁷

The DNA exonerations, the NIJ report, scientific data, and subsequent open discussion of the science involved in eyewitness identification caused law enforcement jurisdictions throughout the United States to reassess eyewitness identification procedures and whether those procedures contributed to misidentification. New Jersey was the first state to implement statewide blind sequential lineup reform, followed more recently by North Carolina.¹⁸ Ramsey County’s review of the NIJ report and the underlying science convinced the jurisdiction that it should attempt to implement the reform, or at least test its feasibility in a law enforcement setting.

Rationale for Change

On reflection, the startling scientific evidence that the traditional law enforcement method of offering photo displays might have inadvertently increased the likelihood of misidentification makes sense objectively for two reasons:

• The risk of administrator influence on lineup results can be essentially eliminated by having someone other than the investigator, who knows the identity of the suspect, conduct the photo lineup, whether the lineup is simultaneous or sequential. (Ramsey County has the utmost confidence in the integrity and professionalism of its investigators but recognizes that even the most well-intentioned investigator who knows which person in the lineup is the suspect cannot eliminate all possibility of unintentionally cueing the witness or responding to the witness’s identification or nonidentification of that person.)


• Simultaneous photo display tends to encourage relative judgment; that is, a witness comparing photos side by side can more easily pick the person who looks most like the perpetrator—regardless of whether that person is, in fact, the perpetrator.

The scientific evidence made clear that both components were necessary and that each accomplished a different objective. If a nonblind administrator is used, switching from simultaneous to sequential photo displays alone may even increase misidentification, since there is a danger of inadvertently highlighting an individual suspect photo. On the other hand, while recognizing the necessity of blind administration, the county understood the practical realities of conducting photo lineups under the time pressures and the personnel constraints in a criminal investigation and that, especially in a small department, the investigator with knowledge of the case (a nonblind administrator) might be the only one available. The county therefore set up a practical alternative that still maintained the blind component: If no independent administrator is available, the investigator must use a method that is the functional equivalent (FE)—that is, a method that assures that the investigator does not know and cannot see the order in which the photos are presented and that the witness knows this (and therefore will not be looking to the administrator for cues). This is sometimes called a “blinded” method.

Blind Sequential Lineup Pilot Project

In the spring of 2005, after reviewing the NIJ report and the scientific literature, the county attorney approached the Saint Paul chief of police and Ramsey County Sheriff Bob Fletcher about conducting a pilot project to test the feasibility of employing blind sequential lineups in the field.¹⁹ They agreed. The county attorney developed a new lineup form to facilitate uniformity of application of this protocol change and conducted training of the pilot investigators in April 2005.²⁰ The training explained the rationale behind both the blind and the sequential components of the protocol.²¹ Although using an independent administrator was the preferred method, the training also included the alternative FE (blinded) method. The pilot began May 1, 2005, and lasted one year.

At the outset, some law enforcement personnel were skeptical. Would having an independent administrator conducting the lineups cause unacceptable delays? Would this procedure result in overtime or other administrative complications or costs? Would lineup results be adversely affected because the administrator was not the investigator who had developed a rapport with the witness? Would there be a decline in positive identifications?

After one year, the results were clear: investigators who used this method found it not only workable but no more difficult to apply than the traditional method. There were no associated administrative difficulties or additional overtime costs. However, there was an unexpected benefit: most investigators involved in the pilot came to prefer the new method and felt more confident in the eyewitness identifications that resulted. Notwithstanding laboratory study results,²² no one felt that appreciably fewer identifications had been made. The fear that independent administrators’ lack of rapport with witnesses would adversely affect identifications turned out to be a nonissue. To Ramsey County’s surprise, most investigators ended up preferring to use and managed to find an independent administrator; as a result, there was less need, in practice, to use the FE (blinded) option than was anticipated.

Taking the time to have a pilot project and being open to the suggestions of the investigators involved were crucial to winning support for the change. The blind sequential lineup procedure was adopted by all Ramsey County law enforcement agencies in April 2006. Countywide training of all investigators was conducted that month, and the protocol has been used since that time.

Ramsey County Blind Sequential Photo Identification Protocol

A one-page (two-sided) Sequential Photo Display Form was developed specifically for use by all Ramsey County law enforcement officers.²³ It contains all instructions needed in the field to ensure that the basic elements of the new protocol are uniformly applied. On one side are all the instructions for the administrator (whether independent or an FE), a section to be filled out by the administrator during the photo display, and a section to be filled out by the administrator after the photo display. On the other side are the instructions to be read to the witness before the photo display (including alternative language, depending on whether an independent or an FE administrator is used) and a section to be completed by the witness after the photo display. The form is designed to be selfexplanatory, enabling an officer from the street with no knowledge of the case to be called in, if necessary, to be the independent administrator.

In formulating the photo display (typically, six photos), administrators are instructed to use descriptors given by the witness to find fillers. They are instructed to select at least five fillers and, before finalizing the array, to look at them together to make sure the suspect does not stand out.

An independent administrator is to be used whenever possible. Agencies deal with the rapport issue directly. If an investigator has already interviewed and cultivated a relationship with a witness, the investigator explains that department procedures require another officer to show the photos but that the investigator will be available to talk to the witness as soon as the procedure is done. Whenever an independent administrator conducts the procedure, the investigator with knowledge of the suspect’s identity must remain out of sight of the witness while the lineup is being conducted.

When no independent administrator is available, the FE method may be used, ensuring first that the investigator in the case with knowledge of the identity of the suspect does not know and cannot see where the suspect is in the order of photos displayed and, second, that the witness knows this. Using the alternate FE language on the Sequential Photo Display Form instructions to the witness, the investigator tells the witness, “I do not know the order of the photos,” and instructs the witness not to let the investigator see which photo is being viewed.

Lineup Methods

There are several methods by which the photos may be displayed sequentially. All have in common the fact that the witness sees only one photo at a time and must continue to view all the photos even if a person is identified before all are shown.²⁴

The high-tech version of this method would be to use a laptop computer to display and maintain a record of the photos shown. However, no adequate software program was available to Ramsey County in 2005 that adequately addressed technical issues.²⁵

The low-tech version can be as simple as numbering the photos and placing each in a separate folder (or, for the FE alternative, having someone else do so for the investigator). The witness can then be instructed to look at the folders one at a time (for the FE alternative, this must be done in such a manner that the investigator cannot see the photos). The remaining folders should be kept out of sight so that the witness does not know how many there will be.

The “box with six doors” method, developed and used in the state of New Jersey since 2001, has been used widely by Ramsey County investigators and is the preferred method for many departments in this jurisdiction. The standard six-person simultaneous photo display sheet is simply inserted in a box built specifically for this purpose. Only one door at a time is opened by the witness. The box is equally adaptable to the FE method since someone else can prepare the six-person lineup sheet and insert it into the box.

Regardless of the method used, a record must be kept of the photos used and the order shown.

Whether an independent administrator or an FE is used, the sequential photo display procedure is the same and includes the following:

• Reading the administrator instructions on the form and completing the case and witness identification information at the top of the form


• Selecting the applicable independent administrator or FE language on the form; reading and checking off the six instructions²⁶ to the witness and having the witness initial in the spot indicated that she understands the instructions (an alternate simplified instruction form has been developed for use with children, mentally impaired adults, and persons with limited knowledge of English)²⁷


• Displaying the photos to the witness one at a time, with only one photo visible at a time, being careful not to give the witness any feedback during or after the photo display


• Recording any comments the witness makes while examining the photos and any other relevant observations (such as the witness’s physical reactions)


• If the witness makes an identification, asking the witness how certain he is of the identification and recording the answer on the form


• Not showing the lineup more than once unless the witness requests it; if requested, the entire lineup must be shown, even if the witness requests only one or a few


• Recording the number of times the lineup was shown and what the witness said each time


• At the conclusion of the photo display, having the witness complete the witness section of the form and, if an identification is made, having the witness sign the selected photo


• Not giving the witness feedback on the accuracy of any selection, as this can artificially inflate witness confidence in the selection and skew further investigation

The witness section of the form is the record, in the witness’s own handwriting, of whether any identification was made and, if so, how certain the witness is of the identification. This portion of the form is to be signed and dated by the witness.

It is critical for administrators to record the number of times a witness is shown the sequence because the scientific evidence strongly supports the conclusion that there is a sharp drop-off in the accuracy of any identification made after two showings.²⁸

The completed Sequential Photo Display Form covers all these points and is sufficient to document the procedure; however, some departments require that a general supplementary report be written as well.

Follow-up Interview by Investigator

After an independent administrator has conducted a lineup, the investigator assigned to the case may follow up with a supplemental interview, as needed. This is the time to develop as many additional facts as possible about the details of any identification made, including exploring what it was about the photo (or photos) selected that made the witness pick it, any follow-up on the witness’s level of certainty or confidence in the selection, and, in the case of multiple perpetrators, what each individual did.

It is critical that investigators not provide any information that could artificially inflate the level of confidence witnesses express in their selections (for example, witnesses must not be told that they did a good job or picked the same person other witnesses picked). If witnesses ask whether they picked the right person, that question should not be answered until investigators have already asked all follow-up questions, if at all. If an investigation is ongoing—and especially if other potential eyewitnesses are yet to be questioned—witnesses may simply be told that the investigation needs to be concluded before that question can be answered. If the person a witness picked is ultimately charged with a crime, the witness will eventually find this out. However, being careful not to reinforce the witness’s selection, and documenting this, should eliminate any defense claim that the investigator artificially inflated the witness’s level of confidence in his or her selection. If witnesses are concerned for their safety and want to know if the suspect is in custody, that question, of course, may be answered.

Other Implementations of Blind Sequential Lineup Protocols

Ramsey County is not alone in the adoption of this new eyewitness identification protocol. The neighboring jurisdiction, Hennepin County (which contains the city of Minneapolis and suburbs), after a similar pilot in 2004, successfully implemented blind sequential lineups countywide in 2005. Like Ramsey County, officers in Hennepin County may use an alternate “blinded” procedure if an independent administrator is unavailable.²⁹ In addition to the statewide adoption of this procedure in New Jersey and North Carolina, the Wisconsin Attorney General’s Office has formally recommended that this change be adopted statewide,³⁰ and more than 200 Wisconsin law enforcement agencies, including those in Milwaukee and Madison, have done so to date. Others using this procedure include Denver, Colorado; Suffolk County, Massachusetts (including the Boston Police Department and suburbs); the Northampton, Massachusetts, Police Department; the Westfield, Massachusetts, Police Department; the Chaska, Minnesota, Police Department; the Virginia Beach, Virginia, Police Department; and Santa Clara County, California.³¹ In January 2009, the Dallas, Texas, Police Department announced it was withdrawing from the NIJ-sponsored field study mentioned in the Mecklenburg article and was adopting the blind sequential protocol.³² Seven other Texas jurisdictions had already adopted it. These examples indicate that the protocol can be adapted for large and small agencies in a variety of settings, urban, suburban, small-town, and rural.³³

Assessment of the Ramsey County Experience

After almost four years of experience with the blind sequential eyewitness identification protocol, Ramsey County investigators and prosecutors report no problems with the change. Notwithstanding laboratory findings that this method may produce a slight drop in correct identifications (as the price for significantly fewer faulty identifications), investigators do not feel that they have lost identifications, and they have increased confidence in the identifications that are made. Use of the form has contributed to uniformity in application of the protocol. Overall, the additional attention paid to this issue has also heightened the county’s awareness that no case should be based on eyewitness identification alone and that further investigation is always needed.

Keeping track of the number of times the photos are displayed before an identification is made provides an additional tool for evaluating the reliability of the identification, since it provides a mechanism for assessing whether the witness is “comparison shopping.” This alone is an improvement over the simultaneous display procedure, in which comparing photos to find the person most like the perpetrator (regardless of whether it is, in fact, the perpetrator) was always an inherent, but usually unmeasurable, risk.

The change in protocol has not resulted in overtime, delay, or other administrative difficulties. The unexpected bonus of adopting this reform is that, in the courtroom, a major avenue of defense attack on both the admissibility and the reliability of eyewitness identification procedures has been eliminated. Most importantly, however, Ramsey County feels it has taken a significant step in reducing the likelihood of false eyewitness identification.

Conclusion

Ramsey County’s change in protocol is not the only model for blind sequential identification reform, nor is it perfect. The county is prepared to make adjustments in the future as scientific knowledge in this area evolves. However, it is confident that this change represents an improvement over prior practice. The Ramsey County experience unequivocally demonstrates that this model can work in a law enforcement setting.

The authors wish to acknowledge the assistance of retired assistant Ramsey County attorney Jeanne Schleh in the writing of this article. Ms. Schleh also designed and implemented the blind sequential training method described herein.■

Notes:

¹See Sheri Mecklenburg et al., “Eyewitness Identification: What Chiefs Need to Know
Now,” The Police Chief 75, no. 10 (October 2008): 68–81 (hereafter referred to as the Mecklenburg article).
²Ibid., 74–76. These recommendations include standards and training on appropriate interviewing, selecting fillers, instructions to witnesses, memorializing witness verbal and nonverbal reactions in report writing, and raising officer awareness of the potential for influencing the outcome.
³This procedure is sometimes also referred to as “double blind” to include the fact that the witness likewise does not know the location of the suspect in the sequence (or even if the suspect is included at all). Since this should be a given in any lineup, this article will simply use the term blind to cover both.
⁴Action on this front should not be postponed pending the current NIJ-sponsored field study to which the Mecklenburg article alludes because that study appears to contain the same design flaw that existed in the Illinois study: it is designed to test whether witnesses correctly identify the person the police suspect. However, the crux of the matter in the DNA exoneration cases was that the police did not identify the correct suspect. Unlike laboratory studies, in which scientists can control ground truth (that is, who the real perpetrator is), a field study based on real crime cannot control for a known perpetrator.
⁵U.S. Bureau of the Census, Population Estimates Program, “Minnesota—County,” 2006 Population Estimates, http://factfinder.census.gov/servlet/GCTTable?-ds_name=PEP_2006_EST&-mt_name=PEP_2006_EST_GCTT1R_ST2S&-geo_id=04000US27&-format=ST-2&-tree_id=806&-context=gct (accessed February 27, 2009).
⁶Only cases of persons convicted before 1995 who were still imprisoned were reviewed.
⁷While the rape investigation was under way, Sutherlin was arrested on a double murder and was convicted and sentenced for those offenses first, with a consecutive sentence imposed for his ultimate rape conviction. He is still serving time on the murders, but his 2002 exoneration of the rape means he will never serve any prison time for the sex offense. He will also be eligible for earlier release on the murders and, upon release, will not be subject to sex offender registration requirements that would otherwise apply.
⁸He could not be prosecuted due to expiration of the statute of limitations in effect at the time of the crime.
⁹Within a year of the 1985 rape, he did develop an arrest record. By 2002, when the DNA match was made, he had a lengthy record of convictions, although none was sex-related.
¹⁰This policy, first implemented countywide in 2004, is available on the county attorney’s Web site at http://www.co.ramsey.mn.us/attorney/SPDNA.asp.
¹¹The interdisciplinary panel of 34 persons included representatives of the criminal justice community (law enforcement agencies, prosecutors, and the defense bar) and social science researchers.
¹²A good summary of scientific recommendations from that time is found in Gary L. Wells et al., “Eyewitness Identification Procedures: Recommendations for Lineups and Photo Spreads,” Law and Human Behavior 22, no. 6 (December 1998): 603–647.
¹³See U.S. Department of Justice, Office of Justice Programs, National Institute of Justice, Eyewitness Evidence: A Guide for Law Enforcement, NCJ 178240, October 1999, http://www.ncjrs.org/pdffiles1/nij/178240.pdf (accessed February 26, 2009).
¹⁴Ibid., 9.
¹⁵Ibid.
¹⁶A meta-analysis is a review of existing studies providing a cumulative quantitative summary of outcomes across the studies.
¹⁷Nancy Steblay et al., “Eyewitness Accuracy Rates in Sequential and Simultaneous Lineup Presentation: A Meta-analytic Comparison,” Law and Human Behavior 25, no. 5 (October 2001): 459–473. Specifically, it shows an aggregate error rate of 27 percent (incorrect identification of a specific filler photo, the designated innocent suspect) for the simultaneous method compared to a 9 percent error rate using the sequential method. An updated meta-analysis, currently under review, again shows significantly greater risk of mistaken identification for simultaneous lineups and somewhat small correct identification rates for sequential lineups. Steblay presented the preliminary data for this more recent meta-analysis at the John Jay College of Criminal Justice as “2001+6: An Updated Meta-analysis of Eyewitness Lineup Performance under Sequential versus Simultaneous Formats,” (paper presented at conference titled “Off the Witness Stand: Using Psychology in the Practice of Justice,” New York City, March 1–3, 2007).
¹⁸See New Jersey’s Attorney General Guidelines for Preparing and Conducting Photo and Live Lineup Identification Procedures, April 2001, http://www.njdcj.org/agguide/photoid.pdf (accessed February 26, 2009); the North Carolina Actual Innocence Commission Recommendations for Eyewitness Identification, October 2003, http://www.ncids.org/New%20Legal%20Resources/Eyewitness%20ID.pdf (accessed February 26, 2009); and that state’s subsequent enactment in 2006 of sections 15A-284.50 to 15A-284.53 of the North Carolina General Statute, adopting its recommendations.
¹⁹The terms lineup and photo display are used interchangeably here. Ramsey County rarely conducts in-person lineups.
²⁰Participating in the training were the Saint Paul Police Department’s homicide and sex crime investigators and the Ramsey County Sheriff’s Department’s general investigators, as well as investigators from three smaller suburban departments.
²¹Ramsey County’s scientific consultant for the training and development of its Sequential Photo Display Form is Nancy Steblay, professor of psychology at Augsburg College in Minneapolis. She also provided oversight and review of scientific references in this article. Dr. Steblay may be contacted at steblay@augsburg.edu.
²²See note 17.
²³The form may be downloaded at http://www.co.ramsey.mn.us/attorney/SPDNA.asp.
²⁴Continuing the photo display when the witness identifies a photo before all photos are shown avoids later legal criticism that the lineup was improperly truncated. In the circumstance that a witness makes an early “filler” pick without having seen the suspect’s photo, it also allows an answer to the question of what would have happened had the witness seen the suspect’s photo (in some cases, a witness might revoke the filler pick). Conversely, the witness may pick the suspect but then discredit this pick when a later photo is shown. Regardless of what happens when the full display is completed, documenting all of these reactions is important and useful information about the witness’s memory, the quality of lineup fillers, and the strength of the witness’s identification—and ultimately whether the police have the right suspect.
²⁵An elegant software program has recently been developed by SunGard OSSI for this purpose. It randomly sorts photos, maintains a record of which version witnesses saw, guides witnesses through a series of questions on each photo by blind administration, preserves an audio recording of all answers (which can be downloaded to a discloseable CD), and generates a written report of results. (See Jerry Farris, “Remote Lineup Application,” The Police Chief 75, no. 8 (August 2008): 96–98, for a description of its use for blind sequential lineups by the Winston-Salem, North Carolina, Police Department.) A variation of this program is currently used by the Tucson, Arizona, Police Department in a field study comparing simultaneous and sequential lineups. It uses blind administration in either format (which the computer randomly assigns). The study, for which Dr. Steblay is the scientific consultant, will run approximately through October 2009.
²⁶These instructions advise witnesses (1) that the person who committed the crime may or may not be included in the photos shown; (2) that the administrator does not know whether the person being investigated is included (or, for an FE, that the administrator does not know the order of the photos); (3) that even if the witness identifies someone during the procedure, all photos will continue to be shown; (4) that a photo may be an old one and that some things, like hair styles, can be changed, and skin colors may look slightly different in photos; (5) that witnesses should not feel compelled to make an identification, since it is just as important to clear innocent persons as it is to identify the guilty and that the investigation will continue whether or not an identification is made; and (6) that witnesses will see only one photo at a time, that the photos are not in any particular order, and that they may take as much time as needed to look at each one. Witnesses are also cautioned to avoid discussing this procedure or the results with any other potential witnesses in the case.
²⁷This form may also be downloaded at http://www.co.ramsey.mn.us/attorney/SPDNA.asp.
²⁸This effect has been found in both field and laboratory studies. For field tests, see Amy Klobuchar et al., “Improving Eyewitness Identifications: Hennepin County’s Blind Sequential Lineup Pilot Project,” Cardozo Public Law, Policy, and Ethics Journal 4, no. 2 (April 2006): 381–414; for laboratory tests, see Nancy Steblay, Double-Blind Sequential Police Lineup Procedures: Toward an Integrated Laboratory and Field Practice Perspective, final report to the National Institute of Justice, grant no. 2004-IJ-CX-0044, http://web.augsburg.edu/~steblay/March2007_Final_NIJ_report.pdf (accessed February 26, 2009).
²⁹The inference in the Mecklenburg article that the blind component is optional in Hennepin County is incorrect, as is the statement that the Hennepin County sequential protocol called for witnesses to make a choice after seeing all the photos. The county’s protocol requires an alternative blinded procedure if an independent administrator is not available. For the complete report on the field study, see Klobuchar et al., “Improving Eyewitness Identifications.”
³⁰These very comprehensive guidelines may be viewed at http://www.thejusticeproject.org/reports/model-policy-and-procedure/.
³¹Northampton Police Captain Ken Patenaude was one of the law enforcement members on the original NIJ panel and an early proponent of this change. He remains active in police training on this issue. His protocol materials may be viewed at http://www.innocenceproject.org/docs/Northampton_eyewitness.pdf (accessed February 26, 2009).
³²Jennifer Emily, “Dallas Police Drop Study, Plan Photo-Lineup Changes,” Dallas Morning News, January 16, 2009, http://www.dallasnews.com/sharedcontent/dws/news/localnews/stories/011609dnmetsequentialblind.4311ff6.html (accessed February 27, 2009).
³³This list does not purport to be comprehensive.

Top