Policy and Procedure Recommendations for the Collection and Preservation of Eyewitness Identification Evidence

Policy and Procedure Recommendations for the Collection and

Preservation of Eyewitness Identification Evidence

Gary L. Wells

Iowa State University

Margaret Bull Kovera

John Jay College and the Graduate Center, City University of

New York

Amy Bradfield Douglass

Bates College

Neil Brewer

Flinders University

Christian A. Meissner

Iowa State University

John T. Wixted

University of California, San Diego

Objective: The Executive Committee of the American Psychology-Law Society (Division 41 of the

American Psychological Association) appointed a subcommittee to update the influential 1998 scientific

review paper on guidelines for eyewitness identification procedures. Method: This was a collaborative

effort by six senior eyewitness researchers, who all participated in the writing process. Feedback from

members of AP-LS and the legal communities was solicited over an 18-month period. Results: The

results yielded nine recommendations for planning, designing, and conducting eyewitness identification

procedures. Four of the recommendations were from the 1998 article and concerned the selection of

lineup fillers, prelineup instructions to witnesses, the use of double-blind procedures, and collection of

a confidence statement. The additional five recommendations concern the need for law enforcement to

conduct a prelineup interview of the witness, the need for evidence-based suspicion before conducting

an identification procedure, video-recording of the entire procedure, avoiding repeated identification

attempts with the same witness and same suspect, and avoiding the use of showups when possible and

improving how showups are conducted when they are necessary. Conclusions: The reliability and

integrity of eyewitness identification evidence is highly dependent on the procedures used by law

enforcement for collecting and preserving the eyewitness evidence. These nine recommendations can

advance the reliability and integrity of the evidence.

Public Significance Statement

Mistaken eyewitness identification is a primary contributor to criminal convictions of the innocent.

Pristine procedures for collecting and documenting eyewitness identification evidence can help

prevent these mistakes. This scientific review paper makes nine system variable recommendations

concerning eyewitness identification procedures that should be implemented by crime investigators

in eyewitness identification cases.

Keywords: eyewitness identification, lineups, showups, identification procedures, eyewitness

recommendations

X Gary L. Wells, Department of Psychology, Iowa State University;

Margaret Bull Kovera, Department of Psychology, John Jay College

and the Graduate Center, City University of New York; Amy Bradfield

Douglass, Department of Psychology, Bates College; Neil Brewer,

College of Education, Psychology and Social Work, Flinders University;

Christian A. Meissner, Department of Psychology, Iowa State Univer-

sity; John T. Wixted, Department of Psychology, University of California,

San Diego.

This is an official statement of the American Psychology-Law Society,

Division 41 of the American Psychological Association, and does not

represent the position of the American Psychological Association or any of

its other Divisions or subunits.

The authors thank the following individuals for their comments and

suggestions on earlier versions of this article: Alexis Agathodeous, Assis-

tant District Attorney David Angel, Dominick Atkinson, Jen Beaudry,

Shari Berkowitz, Iris Blandón-Gitlin, Chief William Brooks, Brian Cahill,

Steve Charman, Brian Cutler, Jen Dysart, Mitch Eisen, Melanie Fessinger,

Ron Fisher, Nancy Franklin, Valerie Hans, Kate Houston, Jennifer Jones,

Detective Matt Jones, Saul Kassin, Elizabeth Loftus, Jamal Mansour,

Karen Newirth, Steve Penrod, Dan Reisberg, Andrew Smith, Siegfried

Sporer, Nancy Steblay, Jennifer Teitcher, and Adele Quigley-McBride.

Correspondence concerning this article should be addressed to Gary L.

Wells, Department of Psychology, Iowa State University, West 112 Lago-

marcino, Ames, IA 50011. E-mail: [email protected]

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

Law and Human Behavior

ISSN: 0147-7307 http://dx.doi.org/10.1037/lhb0000359

The identification of criminal suspects by eyewitnesses contin-

ues to be a staple form of evidence used by the justice system to

help establish the identity of crime culprits. Over the last few

decades, however, serious concerns have been raised about the

potential unreliability of eyewitness identification evidence.

There have been two primary forces that have helped to shape

this concern about the reliability of eyewitness identifications.

First, psychological scientists began developing programmatic

laboratory-based experimental research starting in the mid to late

1970s that focused on eyewitness identification. Early in this

research, it became apparent that certain variables under control of

the criminal justice system can dramatically inflate the likelihood

of a mistaken identification; these variables were called system

variables (Wells, 1978). An instruction given to eyewitnesses prior

to viewing a lineup, for example, is a system variable because it

influences the reliability of the eyewitness’s identification and is

under the control of the legal system’s policies and procedures for

administering lineups (Malpass & Devine, 1981). By the mid-

1990s, psychological scientists had published hundreds of laboratory-

based experiments in peer-reviewed journals showing that mistaken

identification rates can be very high under certain conditions and had

identified some of the more problematic sets of conditions that can

lead to such errors in real-world circumstances (e.g., Cutler & Penrod,

1995).

The second major force propelling a strong awareness of the

potential fallibilities of eyewitness identification was the applica-

tion of forensic DNA testing to claims of innocence. Although

forensic DNA testing was conceived primarily as a tool to incrim-

inate the guilty, the exonerating powers of forensic DNA testing

quickly became evident. Starting in the 1990s, DNA was used to

test claims of innocence in selected postconviction cases and a

cascade of exonerations of innocent people began to unfold. In a

report commissioned by the U.S. Department of Justice examining

the first 30 exonerations, it was quite clear that the vast majority of

these cases involved mistaken eyewitness identifications (Connors,

Lundregan, Miller, & McEwan, 1996). Although legal scholars

had described a number of cases of innocent people being con-

victed in the United States based on mistaken identification before

the development of DNA testing (e.g., Borchard, 1932; Brandon &

Davies, 1973; Frank & Frank, 1957; Huff, Rattner, & Sagarin,

1986), the DNA exoneration cases quickly outnumbered previ-

ously known cases of convictions of the innocent. Moreover, the

pre-DNA cases tended to be less definitive about whether the

person was, in fact, innocent, often just indicating “probable in-

nocence” or “legal innocence” rather than the more definitive

“actual innocence” characterization that accompanied DNA-based

exonerations (Garrett, 2011; Scheck, Neufeld, & Dwyer, 2000;

West & Meterko, 2017).

The development of a scientific literature on eyewitness identi-

fication and the use of forensic DNA testing to uncover mistaken

eyewitness identifications were a powerful combination. In 1996,

the Executive Committee of the American Psychology-Law Soci-

ety (AP-LS, Division 41 of the American Psychological Associa-

tion), the primary scholarly organization for eyewitness identifi-

cation researchers, appointed a committee to draft a scientific

review paper on recommendations for how to collect eyewitness

identification evidence. The result of that review, vetted by the

membership of AP-LS and subjected to peer review, was published

in Law and Human Behavior (Wells et al., 1998). That article,

commonly referred to as the “white article on lineups,” is the

forerunner to the current article. In effect, the current article is an

update of the original scientific review paper on lineups and, like

the original scientific review paper, also represents the official

position of the AP-LS on these issues. The 1998 scientific review

paper was the first set of science-based recommendations regard-

ing how to conduct lineups that was endorsed by a scientific

society. In addition, the 1998 scientific review paper played an

important role in subsequent developments. For example, it was

the model on which the U.S. Department of Justice made its

recommendations in its 1999 guide for law enforcement on col-

lecting and preserving eyewitness evidence. The NIJ Guide was

mailed to all of the more than 17,000 law enforcement agencies in

the U.S. (Technical Working Group for Eyewitness Evidence,

1999). Moreover, the 1998 article was a model on which the New

Jersey Department of Justice created the first statewide guidelines

on eyewitness identification that law enforcement were required to

follow. It was also the first peer-reviewed publication to include a

description of the unfolding DNA exoneration cases (40 at that

time) and what these DNA cases might tell us about eyewitness

identification evidence. The 1998 scientific review paper has been

used as supporting written material in countless seminars and

workshops of lawyers, police, prosecutors, and judges, as well as

in police training. Finally, the 1998 scientific review paper has

been relied on by state and federal courts across the U.S. (e.g., New

Jersey v. Henderson, 2011; State v. Lawson, 2012; Young v.

Conway, 2013).

The Need for an Updated Scientific Review Paper

on Lineups

Today, our understanding of eyewitness identification has ma-

tured well beyond where it was when the previous scientific

review paper was published. Clearly, experimental laboratory

studies have grown immensely in number and breadth over the last

20 years. However, the general methodology of laboratory eye-

witness identification studies has largely remained the same. Peo-

ple are exposed to a simulated crime, sometimes live, sometimes a

video, and the researchers know the true identity of the culprit.

These participant-witnesses are then shown an identification pro-

cedure, typically a photo lineup, in which the culprit’s photo is

embedded among filler photos or the culprit’s photo is absent from

the array and replaced with the photo of an innocent person. Using

this basic paradigm, researchers then systematically manipulate

variables, such as the view the witness had of the culprit, the

similarity of the fillers to the culprit, the instructions given to the

eyewitness prior to viewing, suggestive behaviors of the lineup

administrator, and so on, to see how those variables affect the

responses of the eyewitnesses. This experimental laboratory meth-

odology has a number of strengths that arise from the fact that

“ground truth” is known (i.e., the researchers know which person

is the actual “culprit”) and from the use of random assignment to

conditions that permit inferences about the cause of any effects

observed in responses.

One type of data that was largely unavailable at the time of the

1998 review comes from published field studies of police lineups.

By field study we mean either an archival or prospective exami-

nation of the results of lineups conducted by police investigators in

actual cases. To be included as a field study in our analysis, the

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

WELLS ET AL.

study had to report how often the eyewitnesses identified the

suspect, identified a filler, or made no identification and the study

had to be published in a peer-reviewed journal. Only one field

study had been published in a peer-reviewed journal at the time of

the 1998 review but now there are 11 peer-reviewed published

field studies. In these field studies of lineup outcomes, the lineup

contains “fillers” who are known to be innocent of the crime (such

as people who were in prison at the time of the crime). If the

eyewitness identifies the suspect, it might or might not be an

accurate identification because ground truth is not known with

certainty in actual cases. But when an eyewitness identifies a filler,

it is clearly a mistaken identification because these fillers are

known to be innocent. A mistaken identification of a filler will not

result in charges against that filler, but a filler identification is a

mistaken identification nevertheless. These field studies have clear

limits, in large part because of the difficulty of establishing ground

truth for identifications of suspects (i.e., not knowing whether the

suspect is innocent or guilty, see Horry, Halford, Brewer, Milne, &

Bull, 2014). However, these field study data help counter some of

the criticisms of laboratory-based experiments—in particular, the

criticism that participants in laboratory experiments make a lot of

mistaken identifications only because the witnessed events are not

real and the consequences of mistaken identification are trivial

(Mecklenburg, Bailey, & Larson, 2008). The argument is that real

witnesses to serious crimes would not be so careless. Hence,

perhaps laboratory studies vastly overestimate the eyewitness mis-

identification problem. The results of peer-reviewed field studies,

however, show otherwise.

Based on these field studies, we can now estimate how often

actual eyewitnesses in serious crime cases mistakenly identify a

filler from a lineup. These 11 peer-reviewed published studies

collected data from a total of 6,734 lineups. These field studies are

from highly varied jurisdictions (e.g., California, Arizona, Texas,

London, England) and a summary of these data is shown in Table

1. For current purposes, two statistics of note from Table 1 speak

to the question of whether actual witnesses to serious crimes are

too cautious to make mistaken identifications at rates like those

observed in lab experiments. First, nearly one of every four wit-

nesses (23.7%) who was shown a lineup selected an innocent filler.

Second, among those who made an identification (35.5% made no

identification), over one third (36.8%) identified a known-innocent

filler. A summary of 94 laboratory eyewitness identification stud-

ies showed that filler identification rates averaged 21.2% when the

culprit was present and 34.6% when the culprit was absent (Clark,

Howell, & Davey, 2008). Averaging these two filler identification

rates from lab studies yields 27.9% filler identifications in the

average laboratory study versus 23.7% filler identifications found

in actual cases. These field study data, which were not available at

the time of the 1998 scientific review paper, suggest that experi-

mental laboratory studies are not producing highly inflated rates of

mistaken identification compared with what happens with actual

eyewitnesses to serious crimes. Of course, lab experiments and

actual eyewitness identification cases differ in many ways. For

example, actual eyewitness cases involve longer retention inter-

vals, more violence, and more guns than do lab studies (see Flowe,

Carline, & Karoglu, 2018). Nevertheless, lab studies are particu-

larly valuable for isolating cause-effect relations among variables,

which is a feature that tends to elude field studies for a variety of

reasons (see Horry et al., 2014).

Table 1

Summary Statistics on 11 Published Field Studies of Eyewitness Identification

Authors

Number of

possible IDs

ID of

suspect

IDs of

filler No ID Suspect % Filler % No ID%

% Making

an ID

Suspect ID

rate among

all IDs

Filler ID

rate among

all IDs

Behrman and Davey (2001) 58 29 14 15 50.0% 24.1% 25.9% 74.1% 67.4% 32.6%

Behrman and Richards (2005) 461 238 68 155 51.6% 14.8% 33.6% 66.4% 77.8% 22.2%

Horry, Halford, Brewer, Milne, and Bull (2014) 833 382 149 302 45.9% 17.9% 36.3% 63.7% 71.9% 28.1%

Horry, Memon, Wright, and Milne (2012) 1,039 406 273 360 39.1% 26.3% 34.6% 65.4% 59.8% 40.2%

Klobuchar, Steblay, and Caligiuri (2006) 178 63 20 95 35.4% 11.2% 53.4% 46.6% 75.9% 24.1%

Memon, Havard, Clifford, Gabbert, and Watt (2011) 1,044 456 437 151 43.7% 41.9% 14.5% 85.5% 51.1% 48.9%

Valentine, Pickering, and Darling (2003) 584 237 121 226 40.6% 20.7% 38.7% 61.3% 66.2% 33.8%

Wells, Steblay, and Dysart (2015) 494 132 75 287 26.7% 15.2% 58.1% 41.9% 63.8% 36.2%

Wixted, Mickes, Dunn, Clark, and Wells (2016) 348 114 104 130 32.8% 29.9% 37.4% 62.6% 52.3% 47.7%

Wright and McDaid (1996) 1,561 611 310 640 39.1% 19.9% 41.0% 59.0% 66.3% 33.7%

Wright and Skagerberg (2007) 134 78 28 28 58.2% 20.9% 20.9% 79.1% 73.6% 26.4%

Overall sum 6,734 2,746 1,599 2,389

Weighted means 40.8% 23.7% 35.5% 64.5% 63.2% 36.8%

Note. Some studies reported data that included identifications by witnesses who knew the culprit (prior familiarity) and those data are excluded from Table 1.

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

EYEWITNESS IDENTIFICATION RECOMMENDATIONS

In addition to a large growth in the number of peer-reviewed

laboratory-based experiments, as well as the addition of field

studies of lineups in actual cases, DNA exoneration cases involv-

ing mistaken eyewitness identification have continued to accumu-

late since the 1998 review paper. As of this writing, DNA has

established that at least 365 people in the U.S. were convicted and

imprisoned (some on death row) for crimes they did not commit

(Innocence Project, 2019). More than 70% of these DNA exoner-

ation cases involved mistaken eyewitness identification. Although

the number of DNA exonerations represent only a small fraction of

convictions, it is essential to note that these exonerees were the

lucky few for whom DNA-rich trace evidence for the crime

existed, was collected and preserved properly, and was tested.

Contrary to public perceptions based on TV programs, culprits

leave behind DNA-rich trace evidence in only a small fraction of

cases. The largest category of convictions based on eyewitness

identification evidence is robberies, and culprits of robbery almost

never leave behind DNA-rich trace evidence that could exculpate

a mistakenly identified person. For cases in which DNA did exist

(primarily restricted to sexual assault cases), no one anticipated

that forensic DNA testing would later develop and so for many

cases prior to the advent of forensic DNA testing, the trace

evidence was not collected, was not preserved properly (allowing

it to deteriorate), was lost, or was destroyed by the time forensic

DNA testing came along. Even today, individuals who claim

innocence may not be permitted by statute or discretionary deci-

sion (by a court or a prosecutor) to access and test available DNA

evidence. Hence, DNA testing could only reverse a very small

fraction of possible mistaken identifications. Indeed, eyewitness

identification evidence is still heavily relied upon today because

DNA and other forms of definitive evidence remain extremely

rare. Improving the reliability of eyewitness identification evi-

dence therefore remains an important priority in preventing mis-

carriages of justice.

Focus of the Current Article

The current article, like the 1998 scientific review paper, is not

a broad review of all issues in eyewitness identification. It is,

instead, a focused examination of system variables in eyewitness

identification: factors that relate to the reliability of eyewitness

identifications over which the justice system has (or can have)

control. Hence, despite the fact that there are many powerful

variables that affect the reliability of eyewitness identification

evidence, which are called estimator variables (e.g., same vs.

cross-race identifications, stress during the witnessing of a crime,

quality of view), such variables are not under control of the justice

system and are therefore not the focus of this scientific review

paper.

In some ways, the definition of system variables that is used

today has broadened from its original definition. Originally, sys-

tem variables in eyewitness identification referred to variables that

influence the accuracy of eyewitness identifications over which the

justice system has control (Wells, 1978). Over time, however, the

definition of system variables has broadened to include factors

under the control of the justice system that relate to (as opposed to

influence) the accuracy of eyewitness identifications. Eyewitness

confidence, for example, does not itself influence the accuracy of

an eyewitness identification; however, eyewitness confidence is

related to the accuracy of eyewitness identification and it is easily

contaminated by events that are under the control of the criminal

justice system (such as feedback from the lineup administrator).

Accordingly, securing a confidence statement at the time of iden-

tification using a double-blind lineup administrator is a system

variable (see Wilford & Wells, 2013, for a more extended discus-

sion of this broader view of system variables).

In the current article, other examples of this broader definition

of system variables will become apparent. For example, one of the

recommendations in the current scientific review paper is video-

recording the entire identification procedure. Obviously, video-

recording the identification procedure is not meant to increase the

accuracy of eyewitness identifications. Instead, the purpose of the

video-recording recommendation is to secure a record that might

help to assess the quality of the identification and the procedure

(Sporer, 1992, 1993). That is, a video can shed light on the likely

accuracy of eyewitness identifications via creating a record of

behavioral cues (such as decision time, spontaneous comments,

and confidence cues) that are diagnostic of accuracy (Kaminski &

Sporer, 2017). In addition, video-recordings of the identification

procedure can help document that police followed recommended

procedures and provide other potential benefits as discussed later

in this article. Using the broader definition, video-recording qual-

ifies as a system variable because it is under the control of the

justice system to either video-record or not record, and the record-

ing is relevant to the eyewitness accuracy problem.

A central issue in the development of recommendations on

policies and procedures in eyewitness identification is how to

decide which recommendations are the most important and what

criteria should be used to decide whether to include a recommen-

dation. For most of our recommendations, there is a solid and

specific body of scientific evidence to support the recommendation

and we review that scientific evidence. In some cases, however,

the recommendation is based primarily on reasonably well-

established understandings of human memory and social influence

and our general understanding of problems that we have observed

in actual cases. Consider, for example, our recommendation that

the entire identification procedure be video-recorded. We believe

that video-recording can have many benefits, including but not

restricted to: moderating potential suggestive behaviors by the

lineup administrator, establishing proof as to exactly what instruc-

tions were given to the eyewitness, recording information about

how long it took the eyewitness to make an identification, and

establishing both verbal and nonverbal records of the confidence

expressed by the eyewitness. We believe that the arguments fa-

voring this recommendation are compelling and elements of the

recommendation are grounded in the science, such as the science

showing that the verbal and nonverbal behaviors of the witness

during the lineup are diagnostic of accurate versus mistaken iden-

tifications (e.g., Kaminski & Sporer, 2017). Hence, video-

recording could be justified on that scientific ground alone.

However, there has been little research on the issue of whether

video-recording serves a prophylactic function, how the videos

should be used, or on whom the camera should be focused. Hence,

not every element of the benefits that we propose for video-

recording of the identification procedure have been fully studied.

Nevertheless, we believe that it would be irresponsible to not

include a video-recording recommendation even if a subset of the

benefits (e.g., its prophylactic function) has not yet been thor-

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

WELLS ET AL.

oughly demonstrated. For each of the recommendations that we

offer, we include justifications for the recommendation as well as

a discussion of any concerns or caveats.

Live Versus Photo Lineups

Throughout this article, we use the term lineup to refer to both

live lineups and photo lineups. We do not include separate sections

on these two common procedures because we know of no evidence

to indicate that the principles governing photo lineups and live

lineups are different (see Fitzgerald, Price, & Valentine, 2018). In

others words, every recommendation that we make about lineups

applies equally to live and photographic displays.

Photo lineups are far more common than live lineups in most

U.S. jurisdictions (Police Executive Research Forum, 2013). Photo

lineups are sometimes called photo-spreads, photo-arrays, or

photo-montages. Another common name for photo lineups in

many U.S. law enforcement circles is “six packs,” which refers to

the most frequent size and arrangement of a photo lineup in which

the lineup contains six photos arranged in two rows of three. Live

lineups, sometimes called corporeal or physical lineups, are rela-

tively rare (in research and in practice) compared with photo

lineups, likely due to the greater difficulty, time, and expense of

constructing and conducting live lineups relative to photographic

lineups. But some jurisdictions, such as New York City, com-

monly conduct both live and photo lineups.

Even though the principles governing live versus photo lineups

are thought to be the same, questions have been raised as to

whether performance overall might be better for live lineups than

for photo lineups. The live-superiority hypothesis predicts that the

three-dimensional nature of a live showing (vs. two-dimensional

photos) of the lineup members, along with having visual informa-

tion about the full-bodies (not just faces) of the lineup members,

would clearly render live lineups superior to photo lineups. How-

ever, a review of the scientific evidence comparing live with photo

lineups (as well as video lineups) showed no support for the

live-superiority hypothesis (Fitzgerald et al., 2018). In addition,

there are a number of practical difficulties involved in organizing

and administering live lineups, including the greater difficulty of

finding appropriate fillers for live lineups, and the need to carefully

orchestrate the timing and roles of various people at the lineup

event. Some problematic elements are also difficult to control in

live lineups, such as the appearance of nervousness on the part of

the suspect, a nervousness not likely to be shared by fillers.

Because of the apparent absence of any significant advantage in

accuracy along with the greater practical difficulty of live lineups,

“live lineups are rarely the best option in practice” (Fitzgerald et

al., 2018, p. 307).

We take no position on live versus photo because preference for

one over the other is likely to depend on the circumstances of the

particular case. For example, if the witness described something

distinctive about the body of the culprit or the culprit’s gait, then

a live lineup might be preferred. But the difficulty of constructing

such a lineup with live fillers who match the witness’s description

of the culprit can be very challenging. Also, it is not uncommon for

a suspect to be at-large, which precludes the use of a live lineup.

In other cases, the stress induced when a victim-witness is asked to

view their assailant live for purposes of identification might cause

difficulties that could be avoided with a photo lineup. Some law

enforcement agencies have used live lineups because there were

reasons to believe that the witness might be able to identify the

culprit’s voice by having lineup members speak which, of course,

is not possible while doing a photo lineup. Of course, doing a

photo lineup does not preclude a later voice lineup with that

suspect using only recorded voices and no visual information. In

fact, some have argued that voice identification should be con-

ducted separately from the visual lineup because the diagnosticity

of the information obtained is greater if the witness can identify the

voice and the face independently of each other (Pryke, Lindsay,

Dysart, & DuPuis, 2004). Hence, conducting photo lineups does

not preclude the identification of voices using a separate proce-

dure.

Lineups as Distinguished From Showups

The 1998 scientific review paper dealt only with lineups, but

there is another common identification procedure, called showups,

that is included in this new scientific review paper. The basic

distinction between lineups and showups is that lineups embed the

suspect among known-innocent fillers whereas showups do not use

fillers and instead simply present the suspect alone. There is no

debate among eyewitness scientists about the fact that lineups

produce better outcomes than do showups, whether the outcomes

are measured in terms of diagnosticity ratios or measured using

signal-detection based methods. There is some debate about the

process by which the outcomes for lineups are superior to the

outcomes for showups. For example, it has been suggested that

the use of good lineup fillers can help the witness decide which

facial features are relevant for making an identification decision

(e.g., Wetmore et al., 2015; Wixted & Mickes, 2014). Others,

however, note that lineups appear to be superior to showups only

because a large share of mistaken identifications are siphoned off

to fillers when lineups are used whereas all mistaken identifica-

tions land on the innocent suspect for showups because showups

have no fillers (e.g., see Smith, Wells, Lindsay, & Penrod, 2017;

Wells, Smith, & Smalarz, 2015). These two accounts of how

lineups manage to produce better outcomes than showups are very

different, but no eyewitness scientist contests the general observa-

tion that lineups with good fillers result in better applied outcomes

than do showups.

If there is no question about the superiority of lineups over

showups, why do we have recommendations about how showups

should be conducted? Why not simply state that showups should

never be conducted and that lineups should be conducted instead?

There have been calls by some eyewitness scientists for the elim-

ination of showups based on the clear evidence that showups are

inferior to lineups (see Levi & Lindsay, 2001). But there are legal

and policy reasons to permit showups under certain conditions

even though a lineup would be more diagnostic. We review those

reasons in the section that details the showups recommendation.

As a final note, it should be apparent that there should never be

such a thing as a photographic showup. After all, the justification

for a showup is that the individual has been detained on the street

and there is a very limited time frame for conducting an identifi-

cation procedure. If investigators are merely in possession of a

photo of a suspect, there is no reasonable excuse for not taking the

time to embed the photo among filler photos and conduct a proper

photo lineup.

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

EYEWITNESS IDENTIFICATION RECOMMENDATIONS

Overview of Recommendations

The 1998 scientific review paper had only four recommenda-

tions. The current scientific review paper endorses these same four

recommendations but expands the number of recommendations to

nine. The original four recommendations in the 1998 scientific

review paper included: (a) the identification procedure should be

administered using a double-blind procedure (i.e., the lineup

should be administered by someone who does not know which

person is the suspect and which persons are fillers); (b) prelineup

instructions to the witness should emphasize that the culprit might

or might not be in the lineup and that the lineup administrator does

not know which person is suspected of being the culprit; (c) there

should be only one suspect per lineup and the suspect should not

stand out from the fillers based on the witness’s description of the

culprit or other factors that would draw attention to the suspect;

and (d) a confidence statement should be secured from the witness

at the time of identification and prior to any opportunity to get

feedback about the identification decision. Although we include

these four recommendations in our new set of nine, we have

modified them in certain ways. For example, the double-blind

recommendation now includes other means for accomplishing the

goal of preventing influence from the lineup administrator that do

not necessarily require a neutral administrator. The prelineup in-

structions include reworked language that is intended to make the

instructions more effective. Finally, the securing of a confidence

statement includes the recommendation of recording a confidence

statement for both affirmative identification decisions and rejec-

tion decisions rather than only if the eyewitness makes an affir-

mative identification decision.

The original four recommendations in the 1998 scientific review

paper were restricted almost exclusively to matters that occurred

only during the lineup itself. Some of the five new recommenda-

tions, however, cover broader territory. For example, new recom-

mendations concern matters that occur before the commencement

of an identification procedure, including consideration of when it

might be unwise to conduct an identification procedure, the prob-

lem of repeated identification procedures with the same witness

and suspect, and the importance of conducting a proper interview

of the witness prior to conducting the identification procedure. In

addition, we make a recommendation concerning the appropriate

use of showups.

The following is a brief description of each of the nine

recommendations. The numeric order of the recommendations

corresponds roughly to the temporal order in which police/

administrators would likely encounter the matters covered by

the recommendation (except for the last recommendation,

which concerns showups).

1. Prelineup Interview Recommendation. Before conduct-

ing an identification procedure and as soon as practicable

after the commission of the crime, an officer should

interview witnesses to document their descriptions of the

culprit, obtain their self-report of viewing conditions and

attention during the crime, document any claims of prior

familiarity with the culprit, instruct witnesses to not

discuss the event with other cowitnesses, and warn the

witnesses against attempting to identify the culprit on

their own. The entire interview should be video-recorded.

2. Evidence-Based Suspicion Recommendation. There

should be evidence-based grounds to suspect that an

individual is guilty of the specific crime being investi-

gated before including that individual in an identification

procedure and that evidence should be documented in

writing prior to the lineup.

3. Double-Blind (or Equivalent) Recommendation. Line-

ups should be conducted using a double-blind procedure

(i.e., neither the administrator nor the witness should

know who the suspect is in the lineup) or an equally

effective method of preventing the lineup administrator

from inadvertently influencing the witness.

4. Lineup Fillers Recommendation. There should be only

one suspect per lineup and the lineup should contain at

least five appropriate fillers who do not make the suspect

stand out in the lineup based upon physical appearances

or other contextual factors such as clothing or back-

ground.

5. Prelineup Instructions Recommendation. When inviting

an eyewitness to attend a lineup procedure (photo lineup

or live lineup), police should not inform the eyewitness

of any information that the witness has not already pro-

vided and certainly should not suggest that the suspect

who will be in the lineup has been arrested or that the

culprit will be present in the identification procedure. The

eyewitness should be instructed that (a) the lineup ad-

ministrator does not know which person is the suspect

and which persons are fillers; (b) the culprit might not be

in the lineup at all, so the correct answer might be “not

present” or “none of these”; (c) if they feel unable to

make a decision they have the option of responding

“don’t know”; (d) after making a decision they will be

asked to state how confident they are in that decision; and

(e) the investigation will continue even if no identifica-

tion is made.

6. Immediate Confidence Statement Recommendation. A

confidence statement should be taken from witnesses as

soon as an identification decision (either positive or neg-

ative) is made.

7. Video-Recording Recommendation. The entire identifi-

cation procedure, including prelineup instructions and

witness confidence statement, should be video-recorded.

8. Avoid Repeated Identifications Recommendation. Re-

peating an identification procedure with the same suspect

and same eyewitness should be avoided regardless of

whether the eyewitness identified the suspect in the initial

identification procedure.

9. Showups Recommendation. Showups should be avoided

whenever it is possible to conduct a lineup (e.g., if

probable cause exists to arrest the person then a showup

should not be conducted). Cases in which it is necessary

to conduct a showup should use the procedural safe-

guards that are recommended for lineups, including the

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

WELLS ET AL.

elimination of suggestive cues, a warning that the de-

tained person might not be the culprit, video-recording

the procedure, and securing a confidence statement.

In the detailed treatments of the nine recommendations that

follow, we placed a strong emphasis on communicating an under-

lying principle for each recommendation. The underlying principle

for a recommendation is important because there can be times in

which circumstances might require deviation from the literal spe-

cifics of the recommendation, but the deviation would still be able

to conform to the underlying principle. For example, our recom-

mendation for how to select fillers to use in a lineup is relatively

specific but, in the end, it is more important that the underlying

principle be achieved, namely that the fillers should be chosen in

a way that would not make an innocent suspect stand out in the

lineup. Hence, although we provide specific recommendations, it

is more important to use procedures that reflect the principles

behind the recommendations than to follow the specific recom-

mendations.

Recommendations

Each of the nine recommendations begins with a statement of

the recommendation. We then describe the rationale for the rec-

ommendation, including relevant data and the reasoning behind the

recommendation. In addition, most of the recommendations have

nuances or caveats, and some have practical concerns that are

discussed.

Recommendation 1: Prelineup Interview

Before conducting an identification procedure and as soon as

practicable after the commission of the crime, an officer should

interview eyewitnesses to document their descriptions of the cul-

prit, obtain their self-report of viewing conditions and attention

during the crime, document any claims of prior familiarity with the

culprit, instruct witnesses to not discuss the event with other

cowitnesses, and warn the witnesses against attempting to identify

the culprit on their own. The entire interview should be video-

recorded.

In many cases there might be 911 (emergency call) recordings

or initial witness statements by first responders that can prove

useful to an investigation; however, this recommendation concerns

a more extensive interview that would be conducted by an inves-

tigative officer. Recommendation 1 relates to the conduct of this

interview with a witness or victim, during which time an investi-

gator collects a statement relating to the person’s memory for the

event and the culprit(s). Collection of a detailed description of the

culprit is a critical form of evidence that can facilitate investiga-

tors’ attempts to locate a suspect (Brown, Lloyd-Jones, & Robin-

son, 2008; Kebbell & Milne, 1998). There is now substantial

research on the most effective procedures for interviewing a wit-

ness or victim following an event (see Dando, Geiselman, Ma-

cLeod, & Griffiths, 2015; Fisher, Schreiber Compo, Rivard, &

Hirn, 2014), as well as the harmful effects of suggestive or mis-

leading interviewing procedures that should be avoided (see Brain-

erd & Reyna, 2005; Loftus, 2017; Newman & Garry, 2013).

Specific interviewing procedures have also been developed for the

collection of person descriptions (see Demarchi & Py, 2009;

Gabbert & Brown, 2015; Meissner, Sporer, & Schooler, 2007;

Satin & Fisher, 2019; Sporer, 1996). Although the current recom-

mendations focus largely on the interviewing of adult witnesses,

many of the same principles of memory apply to the interviewing

of child witnesses (who are particularly susceptible to suggestion,

see Ceci & Bruck, 1995). A robust literature is available for

interested readers documenting the challenges of interviewing

child witnesses (see Kask & Bull, 2009), including the develop-

ment of effective, evidence-based protocols for interviewing chil-

dren (see LaRooy et al., 2015; Sternberg, Lamb, Esplin, Orbach, &

Hershkowitz, 2002).

The Contents and Accuracy of Person Descriptions

Obtaining an accurate and complete description of the culprit is

important to furthering an investigation and ultimately can facili-

tate identification of the culprit. Archival studies suggest that

witnesses tend to provide between seven and nine descriptors of a

culprit on average, frequently including information about per-

ceived height, weight, gender, ethnicity, and age (Fahsing, Ask, &

Granhag, 2004; Granhag, Ask, Rebelius, Öhman, & MacGiolla,

2013; Sporer, 1992, 1996; van Koppen & Lochun, 1997; Yuille &

Cutshall, 1986). Descriptions of the culprit’s clothing, stature, and

facial features are generally less frequent. When specific facial

descriptors are provided, the majority refer to upper regions of the

face, in particular the hair, eyes, and nose. Although estimates of

height, weight, and age can be biased by the witness’s own

characteristics (e.g., individuals who are less than average height

tend to underestimate height; see Flin & Shepherd, 1986), wit-

nesses otherwise appear to provide an accurate, general impression

of the culprit. Such descriptions, however, are often lacking in

specific details (Douglass, Brewer, Semmler, Bustamante, &

Hiley, 2013; Fahsing et al., 2004) that might prove useful for the

construction and assessment of identification arrays (Corey, Mal-

pass, & McQuiston, 1999), and it is therefore important that

investigators use evidence-based procedures to enhance the quality

of witnesses’ accounts.

System Variables That Influence the Quality of

Witness Accounts

Much like other memory phenomena, a host of factors can

influence the accuracy and completeness of a witness’s memory

for the event and culprit (see Granhag, Ask, & MacGiolla, 2013;

Meissner et al., 2007). Consistent with the general eyewitness

literature, a distinction can be drawn between system and estimator

variables (Wells, 1978). With respect to the former, the manner in

which a witness is interviewed by an investigator can undermine

the accuracy of a witness’s statement. In particular, witnesses

appear quite susceptible to the misinformation effect (see Berkow-

itz & Loftus, 2018; Loftus, 2017; Newman & Garry, 2013)in

which leading or suggestive questioning from an investigator can

distort memory reports, and witnesses can be induced to self-

generate errors in their descriptions when forced or encouraged to

provide a “complete” description of the event or culprit (Ackil &

Zaragoza, 1998; Meissner, Brigham, & Kelley, 2001). In this

respect, the use of facial feature checklists is not recommended, as

they can subtly encourage “complete” responses that produce less

accurate person descriptions (Wogalter, 1991, 1996). Finally, ex-

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

EYEWITNESS IDENTIFICATION RECOMMENDATIONS

posure to media coverage of an incident before an interview can

also lead witnesses to recall incorrect details that were suggested

or inferred (Crombag, Wagenaar, & van Koppen, 1996), and

contact with other witnesses can similarly introduce systematic

errors in memory (Gabbert, Memon, & Allan, 2003; Gabbert,

Memon, Allan, & Wright, 2004; Eisen, Gabbert, Ying, & Wil-

liams, 2017; Loftus & Greene, 1980). We recommend that inves-

tigators avoid suggestive or leading interviewing practices and that

they instruct witnesses not to discuss their accounts with or in front

of one another. Whereas it is possible that discussions between

witnesses could lead to an increase in reported details (e.g., Vre-

develdt, Hildebrandt, & van Koppen, 2016), the dangers associated

with contamination suggest that witnesses should be interviewed

individually to preserve the independence of each statement. In-

vestigators should also document whether a witness has spoken

previously with other witnesses or has been exposed to media

reports related to the incident.

Documenting Factors That Can Influence the Quality

of Witness Accounts

It is also important that investigators note the conditions

under which the witness may have viewed or interacted with the

culprit, as certain factors can influence the likely quality of a

witness’s recollection (for a review see Meissner et al., 2007).

Documenting such factors can aid both investigators and fact

finders in assessing the likely reliability of a witness’s memory.

With respect to naturally occurring estimator variables, factors

at the time of encoding such as low illumination (Wagenaar &

van der Schrier, 1996), greater distance from the culprit (Loftus

& Harley, 2005), and limited time of exposure can lead to

poorer quality person descriptions (Sporer, 1992; van Koppen

& Lochun, 1997; Yarmey, 1986; Yarmey, Jacob, & Porter,

2002). The presence of a weapon can draw attention away from

the culprit’s appearance (Fahsing et al., 2004; Fawcett, Russell,

Peace, & Christie, 2011; Kocab & Sporer, 2016; Pickel, 1998,

1999). The consumption of alcohol or drugs by a witness can

similarly reduce the amount of information provided (Flowe,

Takarangi, Humphries, & Wright, 2016; Read, Yuille, & Toll-

estrup, 1992; Schreiber Compo et al., 2017; Yuille & Toll-

estrup, 1990). Extensive delays between encoding and the time

of interviewing can diminish the amount of detail provided by

a witness (Ellis, Shepherd, & Davies, 1980; Meissner, 2002;

Tuckey & Brewer, 2003; van Koppen & Lochun, 1997). In

contrast with the previous factors, prior familiarity with the

culprit (i.e., an individual known to the witness) generally

increases the accuracy of a witness’s description and identifi-

cation (Vallano, Steele, Slapinski, Briggs, & Pozzulo, 2019).

Given the influence of these estimator variables on both the

quality of person descriptions and subsequent attempts to iden-

tify the culprit from a lineup, it is recommended that investi-

gators clearly document the presence of such factors in their

report.

Evidence-Based Approaches for Interviewing

Witnesses and Victims

Acquiring a complete, yet accurate, statement from the witness

is critical to furthering an investigation. Considerable research

has documented the most effective methods for interviewing a

witness or victim (Dando et al., 2015; Fisher et al., 2014). In

general, it is common for investigators to invite an open-ended

response from the witness, followed by specific probes associated

with key details such as the culprit’s physical characteristics (e.g.,

height, build, age, race, sex, etc.), clothing, or any distinguishing

characteristics (Brown et al., 2008; Launay & Py, 2015; Wise,

Safer, & Maro, 2011). The use of open-ended, nonsuggestive

questioning tactics (Clarke, Milne, & Bull, 2011; Walsh & Bull,

2010) is recommended for eliciting a complete narrative from the

witness. Although the use of specific probes can increase the

number of details provided, such details may come at the expense

of lower accuracy of responding (Sauerland, Krix, van Kan, Glunz,

& Sak, 2014). As such, caution should be used in moving to

closed-ended or two-alternative questions and the use of sugges-

tive/leading prompts should be avoided altogether.

Evidence-based interviewing protocols have been developed

that both avoid the pitfalls of leading and suggestive questioning

and enhance witness reporting by facilitating the retrieval of in-

formation from memory. One of the most notable and empirically

validated protocols is the Cognitive Interview (Fisher & Geisel-

man, 1992). A robust literature has demonstrated the effectiveness

of the Cognitive Interview for eliciting both detailed event narra-

tives and person descriptions from cooperative witnesses (Memon,

Meissner, & Fraser, 2010). Several instructional and mnemonic

aspects of the Cognitive Interview appear to be particularly useful,

including: (a) encouraging witnesses to “report all” of the infor-

mation they can recall but not to guess about anything they are

unsure of (e.g., Clifford & George, 1996); and (b) using context

reinstatement by asking witnesses to close their eyes and think

back to the event context (e.g., Smith-Spark, Bartimus, & Wilcock,

2017; Vredeveldt, Baddeley, & Hitch, 2012, 2014). Lab-based

research indicates that the Cognitive Interview (compared with a

standard police interview) increases the number of descriptors of

the culprit and increases the chances that using the description can

lead to finding the culprit’s photo among a larger set of photos

(Satin & Fisher, 2019). The increase in descriptors for the Cogni-

tive Interview is typically quite large for correct details; and

although a small increase in incorrect details has been noted across

studies, the accuracy rate for the Cognitive Interview does not

differ from that of a standard interview (see Memon et al., 2010).

We encourage investigators to seek training in and adopt the

Cognitive Interview protocol when interviewing witnesses and

victims.

Another specific protocol for eliciting person descriptions,

termed the Person Description Interview, significantly increases

the quantity of person descriptors provided by witnesses (Demar-

chi & Py, 2009; Demarchi, Py, Groud-Than, Parain, & Brunel,

2013). The Person Description Interview incorporates two key

instructions to the witness with respect to describing a person of

interest: (a) to provide general information about the person before

moving to specific featural aspects of the face, and (b) when

describing the face to begin with the lower regions of the face

(chin and lips) and to move up to the top regions (eyes and hair).

Consistent with the Person Description Interview instructions,

encouraging witnesses to provide more general, coarse-grained

information during an interview can enhance the quantity of in-

formation absent a cost to accuracy (Brewer, Vagadia, Hope, &

Gabbert, 2018).

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

WELLS ET AL.

Finally, although witness descriptions are frequently collected

via an oral interview conducted by an investigator, at times a

witness may be asked to directly provide a written statement.

There is mixed evidence with respect to how the format for

eliciting a witness’s recall might influence the quantity and quality

of information provided, with some studies suggesting that oral

interviews produce more information from witnesses than does

written statements (Kraus, Zeier, Wagner, Paelecke, & Hewig,

2017; Sauerland & Sporer, 2011) and others finding no difference

as a function of modality (McPhee, Paterson, & Kemp, 2014;

Sauerland et al., 2014). A recently developed protocol, referred to

as the Self-Administered Interview (see Hope, Gabbert, & Fisher,

2011), allows cooperative witnesses to self-generate high-quality

descriptions of their experience. The Self-Administered Interview

prompts witnesses to recall details of the event, including a person

description of the culprit (e.g., hair, complexion, build, distin-

guishing features). To facilitate recall, the Self-Administered In-

terview incorporates the two key elements of the Cognitive Inter-

view as previously described—a “report everything” instruction

and a context reinstatement prompt. The Self-Administered Inter-

view has been shown to significantly increase the quantity of

person descriptors when compared with a standard free recall

prompt, at a level comparable with that of the Cognitive Interview

(Gabbert, Hope, & Fisher, 2009; Hope, Gabbert, Fisher, & Ja-

mieson, 2014). Such an interview protocol is particularly useful

when an incident involves many possible witnesses or victims and

when such conditions could lead to significant delays in eliciting a

statement from witnesses or victims.

Finally, we recommend that all interviews with witnesses should

be video-recorded. Such an objective record of the interview will

allow both investigators and fact finders the opportunity to review

the information provided the witness and evaluate its evidential

value. Importantly, studies suggest that investigators fail to accu-

rately record or recall key details of statements provided in inter-

views (Kassin, Kukucka, Lawson, & DeCarlo, 2017; Lamb, Or-

bach, Sternberg, Hershkowitz, & Horowitz, 2000); thus, recording

the interview with a witness provides an objective record of the

information elicited, absent omissions or errors that may be intro-

duced via the investigators’ recollection of the interview. Lastly,

the interview should be video-recorded from a perspective that

captures both the investigator and the witness, as studies suggest

that this perspective can enhance fact finders’ evaluations of the

evidence (Lassiter, 2010; Lassiter, Ware, Ratcliff, & Irvin, 2009;

Ratcliff, Lassiter, Schmidt, & Snyder, 2006).

Instructions to Witnesses Following an Interview

Interviewers should instruct witnesses to not discuss the wit-

nessed event or what they have told investigators with other

potential witnesses in the case. As noted earlier, when eyewit-

nesses talk with each other about their memories, they can influ-

ence one another such that their subsequent individual memory

reports can become contaminated with what others have recalled

(for reviews, see Gabbert & Hope, 2013; Wright, Memon, Skager-

berg, & Gabbert, 2009). Encountering a piece of misinformation

from a cowitness about a facial feature can lead witnesses to later

misidentify someone from a lineup who has that feature (e.g.,

Eisen et al., 2017; Zajac & Henderson, 2009). Moreover, hearing

that a cowitness had made an identification from a lineup can

increase the chances that a witness would also make an identifi-

cation; hearing that a cowitness identification was made more

confidently, as opposed to less confidently, can also increase the

confidence that a witness expresses in the accuracy of their own

identification (Levett, 2013). It is recommended (a) that witnesses

should be cautioned to avoid discussing the case with others and

(b) that investigators should refrain from sharing any information

that other witnesses had previously provided. As with the lineup

identification procedure (see Recommendation 7), we recommend

that the entire prelineup interview be video-recorded.

Finally, interviewers should instruct witnesses not to conduct

their own investigation of the crime. Increasingly, we are seeing

that the first identification that witnesses make is the result of a

self-directed search on the Internet, including social media sites.

These searches by witnesses lack many of the protections of a

well-conducted lineup; all faces viewed are possible suspects,

there are no instructions reminding the witnesses that the culprit

may not be among the faces they viewed, some faces may stand

out more than others, witnesses might engage in these searches

alongside cowitnesses, and there is no recording of confidence

immediately after the identification. Once this initial identification

of a suspect is made through an Internet search, it is not possible

to conduct an uncontaminated identification procedure using better

methods (see Recommendation 8).

Recommendation 2: Evidence-Based Suspicion

There should be evidence-based grounds to suspect that an

individual is guilty of the specific crime being investigated before

including that individual in an identification procedure and that

evidence should be documented in writing prior to the lineup.

Conducting lineups in the absence of evidence-based reasons for

suspicion is a risk factor for mistaken identification. In the par-

lance of eyewitness science, making an individual the focus of a

lineup in the absence of evidence that the individual is likely to be

the culprit (e.g., having only a hunch) contributes to a low base rate

for culprit-present lineups (i.e., a high base rate for culprit-absent

lineups). In the case of lineups, base rate refers to the rate for

which the suspect in the lineup is guilty versus innocent. A proper

lineup contains only one suspect, who might or might not be the

culprit (see Recommendation 4). It follows from this structure of

lineups that a mistaken identification of an innocent suspect cannot

happen with a culprit-present lineup and, of course, an identifica-

tion of the culprit cannot happen with a culprit-absent lineup

(Wells & Turtle, 1986). Therefore, low base rates for culprit-

present lineups (high base rates for culprit-absent lineups) create

fertile ground for mistaken identifications of innocent suspects and

reduce the chances of identifying the culprit. Moreover, culprit-

absent lineups inflate the rate at which eyewitnesses identify

known-innocent fillers (Smith, Wilford, Quigley-McBride, &

Wells, 2019), thereby tainting that witness’s credibility for any

later lineup that might include the culprit.

The evidence-based suspicion recommendation derives from the

observation that there are no laws or other mechanisms in place to

prevent jurisdictions from making investigative decisions that re-

sult in extremely low base rates for culprit-present lineups (i.e., a

high rate of culprit-absent lineups; Wells, 2006). In fact, the only

study of actual lineups to estimate the base rate for culprit-present

lineups in any jurisdiction (in this case the Houston, Texas Police

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

EYEWITNESS IDENTIFICATION RECOMMENDATIONS

Department) yielded an estimate of a mere 35% (Wixted, Mickes,

Dunn, Clark, & Wells, 2016). If this estimate is correct for Hous-

ton, then the suspect was innocent in 65% of their lineups. Another

field study found that 40% of the lineups (in Northern California

jurisdictions) had no prelineup evidence at all indicating that the

suspect was the culprit; and for an additional 30% of the lineups

there was minimal evidence (Behrman & Richards, 2005). More-

over, a national survey of U.S. law enforcement agencies reported

that more than one third of the agencies stated that they needed no

evidence at all or needed only a mere hunch that a person might be

the culprit before placing that person in a lineup (Wise et al.,

2011). Of course, the base rate is likely to vary from one jurisdic-

tion to the next depending on the practices and policies in place

(Wells, 1993). Nevertheless, at the time of this writing we know of

no jurisdiction in the U.S. whose policies or written procedures

require, urge, or even mention that there should be some form of

concrete evidence against a person before conducting an identifi-

cation procedure focused on that person.

It is unclear why so many crime investigators do not seem to be

concerned about the problem with having little or no evidence

before placing someone in a lineup. Perhaps this lack of concern

stems from an assumption that an eyewitness would not pick an

innocent individual and would only pick someone if they remem-

bered the person committing the crime.

The Importance of Base Rates

There are many studies in the basic judgment and decision-

making literature showing that people struggle to grasp the strong

impact that prior probabilities and base rates have on test outcomes

(e.g., Kahneman & Tversky, 1973). To illustrate, it is perhaps

instructive to draw a close analogy between eyewitness identifi-

cation testing and medical diagnostic testing. In medical diagnostic

testing, it is common for medical organizations to issue guidelines

about when to perform diagnostic tests versus forgo such tests.

Consider, for example, the prostate-specific antigen (PSA) test for

prostate cancer (see Vollmer, 2006). Although the PSA test is just

as accurate for men under 30 as it is for men over 50, almost every

positive PSA test result on men under 30 is a false alarm whereas

only a small fraction of positive PSA test results on men over 50

are false alarms. Because the base rate (or prior probability) that an

under-30 male will have prostate cancer is nearly zero, almost

every positive result is a false alarm.

This same principle applies to eyewitness identification proce-

dures. In the case of lineups, the base rate is the rate at which the

suspect in the lineup is guilty versus innocent. More formal treat-

ments of the (Bayesian) mathematics behind this problem are

available (see Wells, Yang, & Smalarz, 2015; Wixted & Wells,

2017), but a simple version of the problem is presented here.

Assume that the chances that an innocent suspect will be identified

from a culprit-absent lineup is 6% and the chances that a guilty

suspect will be identified from a culprit-present lineup is 60%.

Assume as well that the long-term base rate for culprit-present

lineups is 50% (and the culprit-absent lineup base rate is therefore

50%). Suppose now that 1,000 lineups were conducted (500

culprit-present and 500 culprit-absent). We would expect 300

identifications of guilty suspects (60% of 500) and 30 identifica-

tions of innocent suspects (6% of 500). In this example, 330

suspects are identified and 9.1% of these suspects (30/330 ⫽ .091)

are innocent.

Now suppose that, instead of a 50% base rate, the base rate were

lowered to 30% (300 culprit-present lineups and 700 culprit-absent

lineups). Now, the 1,000 lineups would be expected to yield 180

identifications of guilty suspects (60% of 300) and 42 identifica-

tions of innocent suspects (6% of 700). The result is that 222

suspects are identified (180 ⫹ 42) and 18.9% of these (42/222 ⫽

.189) are innocent. In this 30% base-rate example, the percentage

of identified suspects who are innocent more than doubles com-

pared to when the base rate is 50%. Clearly, things get better if the

base rate for the suspect being guilty is increased to 70%. At a base

rate of 70%, the 1,000 lineups would yield 420 identifications of

guilty suspects (60% of 700) and only 18 identifications of inno-

cent suspect (6% of 300). In this example, 438 suspects are

identified (420 ⫹ 18) and only 4.1% are innocent.

Notice in the earlier examples that the eyewitnesses themselves

are performing just as well when the base rate is 30% as they are

when the base rate is 70% (just as the PSA test performs as well

when used on men of age 30 as it does on men of age 60). The

difference is that the 30% base rate allows for many more false

alarms than does the 70% base rate. Every time a culprit-absent

lineup is conducted, there exists some probabilistic jeopardy for an

innocent suspect. Therefore, minimizing the chances of presenting

witnesses with culprit-absent lineups is one way to reduce the

problem of wrongful convictions.

Even when the witness does not identify the innocent suspect in

a culprit-absent lineup, they often identify a known-innocent

lineup filler (Clark & Wells, 2008; Wells & Lindsay, 1980; Wells

& Olson, 2002; Wells et al., 2015). Most filler identifications are

made with low confidence, clearly signaling their error-prone

nature (Wixted & Wells, 2017); however, some are made with

higher confidence. Suppose, for example, an eyewitness is shown

a culprit-absent lineup and identifies a known-innocent filler.

Later, police receive information about who the actual culprit is. It

is too late to undo the fact that showing the eyewitness a culprit-

absent lineup led the witness to identify a known-innocent filler as

the culprit. This prior identification of a known-innocent filler

makes the prosecution of any newly identified person in a later

lineup substantially more difficult for prosecutors. In this sense,

culprit-absent lineups not only create risk for innocent suspects but

also elevate rates of filler identifications that, in turn, undermine

eyewitnesses’ credibility on any later identification opportunities

that could involve a culprit-present lineup (Wells, Steblay, &

Dysart, 2012).

What Is Evidence-Based Suspicion?

By evidence-based suspicion, we mean that there is articulable

evidence that leads to a reasonable inference that a particular

person, to the exclusion of most other people, likely committed the

crime in question. As with other standards used in the legal system

(such as reasonable suspicion or probable cause), there is no

precise probability associated with the concept of evidence-based

suspicion. However, a mere hunch is not evidence-based suspi-

cion. Moreover, merely fitting a general description that the wit-

ness gave of the culprit (e.g., young male, mid-20s, dark hair,

normal build) is not evidence-based suspicion as it could be

applied to large numbers of people. Nor can this notion of

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

WELLS ET AL.

evidence-based suspicion be based on backward reasoning in

which a pick of that person in the lineup is used retroactively to

justify the placement of the person in the lineup. In other words,

evidence-based suspicion is something that must be established

prior to the lineup.

Not all articulable evidence qualifies as good evidence that

connects a specific suspect to a specific crime. However, there are

many possible examples of what could qualify as evidence-based

suspicion for purposes of justifying the placement of an individual

in a lineup identification procedure. Examples include:

• A unique fit to a specific description that was given by the

eyewitness (e.g., blue teardrop tattoo under left eye; moon

shaped scar on chin);

• Self-incriminating statements;

• Being in possession of materials linked to the crime along

with a fit to the general physical description given by the

eyewitness;

• Known to be in the area of the crime around the time of the

crime along with a fit to the general physical description

given by the eyewitness;

• Physical evidence at the crime scene linked to the person

along with a fit to the general physical description given

by the eyewitness;

• A unique pattern to the crime that is known to be associ-

ated with a particular offender along with a fit to the

general physical description given by the eyewitness.

These examples of evidence-based suspicion are certainly not

exhaustive. But they capture the idea that the evidence should be

articulable, not based on mere hunch, and lead to a reasonable

inference that there is individuating evidence that makes this

person, to the exclusion of most other people, a reasonable candi-

date to be the one who committed the crime in question.

In contrast, articulable evidence that fails to link the suspect to

the specific crime for which an identification is sought does not

qualify as evidence-based suspicion. Some examples include:

• A search of police records reveals that the suspect was

convicted of a similar crime in the same jurisdiction, has

been released from prison, and is now living in the neigh-

borhood where the crime was committed.

• There are multiple witnesses to a crime and the identifi-

cation of the suspect made by the first witness is used to

establish evidence-based suspicion for the remaining wit-

nesses. If there was no articulable evidence for the first

witness then the first lineup should never have been con-

ducted.

• The suspect resembles a composite sketch or rendering of

the culprit made with the assistance of the witness. This

type of evidence does not clear the threshold of reason-

able, articulable suspicion linking the suspect to the crime

under investigation because composites do not reliably

represent a recognizable representation of the culprit (Ko-

vera, Penrod, Pappas, & Thill, 1997).

• A suspect who was apprehended in the vicinity of one

crime happens to match the description of the culprit not

only for that crime but also for several other similar crimes

recently committed elsewhere in the community. If a wit-

ness to the crime committed in the vicinity of the suspect’s

apprehension does not identify the suspect, then there is no

evidence to support placing that suspect in lineups shown

to witnesses to the other similar crimes committed else-

where in the community. The nonidentification fails to

establish any link between the suspect and the other sim-

ilar crimes.

Again, this list of articulable evidence that does not meet criteria

for evidence-based suspicion is not exhaustive. However, these

examples illustrate that the evidence supporting the placement of a

suspect in an identification procedure must be evaluated for

whether it actually provides a nexus between the suspect and the

crime witnessed.

Final Comments on Evidence-Based Suspicion

The medical field’s understanding of the impact of base rates on

medical diagnostic test outcomes is far ahead of the legal system’s

understanding of the impact of base rates in eyewitness lineup test

outcomes (Wells et al., 2015). Furthermore, the concern about base

rates in eyewitness identification might be even more important

than are base rate concerns in medical testing because diagnostic

medical tests can be repeated to confirm reliability of the result, or

a different type of test can be performed to look for convergence

of results. An eyewitness identification test, in contrast, cannot be

repeated with that same witness and same suspect without being

contaminated (see Recommendation 8); it is important to ensure

that the chances of presenting eyewitnesses with a culprit-absent

lineup are not unduly high.

Although our discussion of the importance of having evidence-

based suspicion has been centered on the chances that an innocent

suspect will be identified, there is an additional reason to be

concerned about presenting eyewitnesses with a culprit-absent

lineup. Even if the eyewitness does not identify an innocent

suspect, culprit-absent lineups strongly increase the chances that

the eyewitness will identify a filler (Wells, 1984; Wells et al.,

2015). When an eyewitness identifies a filler, it “burns” the cred-

ibility of that eyewitness for purposes of any later identification

(e.g., Wells et al., 2012). Suppose, for example, an innocent

suspect was placed in a lineup, the eyewitness picks a filler, and

investigators later discover evidence-based suspicion against a

new suspect. Can they simply bring the eyewitness back and show

the eyewitness a new lineup with the new suspect? The empirical

data indicate that eyewitnesses who identify a filler from a culprit-

absent lineup are highly error prone on any later lineup, even if that

later lineup includes the culprit (Smalarz, Kornell, Vaughn, &

Palmer, 2019). Moreover, research indicates that giving discon-

firming feedback to witnesses who identify a filler reduces per-

formance on subsequent identification tests (Palmer, Brewer, &

Weber, 2010).

An evidence-based suspicion standard could be implemented

easily by requiring detectives to present their proposal for con-

ducting a lineup to a supervisor of detectives. The supervisor of

detectives could then question the detective about why this lineup

is being conducted with this particular person as its focus (Wells,

2006). The detective should be able to point to some concrete

evidence that could lead to a reasonable inference that this person

should be suspected of being the culprit in question; if not, a

supervisor of detectives could suggest instead that the detective

investigate further so as to have more confidence that the subject

of the lineup is the culprit.

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

EYEWITNESS IDENTIFICATION RECOMMENDATIONS

Although a lineup should be conducted only after establishment

of evidence-based suspicion, eyewitness memory can fade with the

passage of time. Hence, a lineup should be conducted as soon as

possible after establishing evidence-based suspicion.

Recommendation 3: Double-Blind (or Equivalent)

Lineups should be conducted using a double-blind procedure

(i.e., neither the administrator nor the witness should know who

the suspect is in the lineup) or an equally effective method of

preventing the lineup administrator from inadvertently influencing

the witness.

A lineup administration is a social interaction between a witness

and an administrator. Like in other social situations, interpersonal

expectancies operate in the context of a lineup administration.

Why is the social interaction aspect of lineup administration con-

cerning? When someone has an expectation about how another is

likely or ought to behave, this expectation can cause the person

with the expectation to behave differently toward the target of the

expectation. This change in the expectation-holder’s behavior elic-

its the very behavior that was expected from the target (Harris &

Rosenthal, 1985; Rosenthal, 2002; Snyder & Swann, 1978). The

social interaction that takes place during the administration of a

lineup is not immune from this interpersonal expectancy phenom-

enon. There is no presumption that the influence of the lineup

administrator is intentional or even that the lineup administrator or

witness is aware of the influence.

Lineups-As-Experiments

A lineup is a test of the hypothesis that the person whom the

police suspect is in fact the culprit of the crime. The lineup

administrator is fundamentally an experimenter who is conducting

a procedure to test this hypothesis (Wells & Luus, 1990). Because

people tend to test hypotheses in a way that will confirm their

expectations (e.g., Klayman & Ha, 1987; Skov & Sherman, 1986),

a lineup administrator, like any other experimenter, should follow

protocols that will prevent them and their expectations from influ-

encing the results of their tests. Double-blind testing, in which the

lineup administrator does not know which person is the suspect

and which are merely fillers (i.e., a blind administrator), is the best

way of ensuring that any information that administrators have

about which lineup member is the suspect will not influence the

witnesses’ behavior, including any identification decision they

might make or their confidence in that decision (see also Recom-

mendation 6 for a discussion of how double-blind administration

eliminates the opportunity for postidentification feedback that

could influence witness confidence). Double-blind testing can also

prevent administrators’ expectations from influencing their reports

of witnesses’ behaviors during the procedure. In contrast, single-

blind lineup administration, in which the administrator knows

which lineup member is the suspect and which are fillers (i.e.,

nonblind administrator), allows for the possibility that the admin-

istrator will communicate the identity of the suspect to the witness

through intentional or unintentional verbal or nonverbal behaviors.

In a single-blind lineup procedure, the eyewitness does not know

which person is suspected of being the culprit and which ones are

fillers, but the lineup administrator knows. In a double-blind lineup

procedure, neither the eyewitness nor the lineup administrator

knows which person is suspected of being the culprit and which

are fillers.

The double-blind recommendation is primarily focused on keep-

ing knowledge about the suspect from the administrator of a lineup

so that this knowledge cannot influence the administrator’s behav-

ior while conducting the identification procedure. However, the

purpose of the recommendation is to keep anyone who knows

which lineup member is the suspect from influencing the witness.

Thus, there should be no officers (e.g., the lead detective) in the

room where the identification procedure is conducted who know

which lineup member is the suspect, even if they are not the officer

administering the procedure. Moreover, if there are multiple wit-

nesses, a different blind administrator should conduct the lineup

with each witness because conducting a procedure with one wit-

ness may provide an administrator with clues about which lineup

member is the suspect, which might then influence how that

administrator interacts with the next witness while administering

the identification procedure (Douglass, Smith, & Fraser-Thill,

2005).

It is equally important to keep information about which person

is the suspect and which ones are fillers from the witness. It might

seem odd to explicitly warn against letting a witness know who the

suspect is before they make an identification decision. In practice,

however, we find it not uncommon for circumstances surrounding

the identification procedure to alert the witness to which lineup

member is the suspect. For example, the witness may be tipped off

to who the suspect is after being presented with multiple photo

arrays that share only one lineup member (the suspect) in common,

which is one of the many reasons for our recommendation to avoid

repeated lineup procedures (Recommendation 8).

The recommendation for double-blind administration of lineups

was included among the original four recommendations made in

the previous scientific review paper (Wells et al., 1998). At the

time that article was written, however, there were no studies that

directly tested whether a lineup administrator’s knowledge of

which lineup member was the suspect influenced witness identi-

fications. Without studies directly testing the effects of double-

blind administration of lineups, the recommendation was made

based on generalizations from basic studies on experimenter ex-

pectancy effects (Harris & Rosenthal, 1985) and early research

suggesting that positive feedback to witnesses after they choose

the suspect increases their confidence in that choice (Wells &

Bradfield, 1998). Since the previous scientific review paper was

written, researchers have conducted a number of studies demon-

strating that the single-blind administration of lineups increases the

likelihood that witnesses will identify the suspect (for a review, see

Kovera & Evelo, 2017), irrespective of whether the suspect is the

culprit (Charman & Quiroz, 2016; Greathouse & Kovera, 2009)or

an innocent suspect (Charman & Quiroz, 2016; Greathouse &

Kovera, 2009; Zimmerman, Chorn, Rhead, Evelo, & Kovera,

2017).

Paradigms for Examining Lineup Administrator

Influence

Scholars have developed several paradigms to examine the

effects of administrator influence on witness decisions. In one

paradigm, which has been termed the steering paradigm (Ko-

vera & Evelo, 2017), the lineup administrator is a confederate

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

WELLS ET AL.

of the experimenter who intentionally engages in behaviors that

steer that witness toward the suspect (e.g., Rhead, Rodriguez,

Korobeynikov, Yip, & Kovera, 2015) or encourages the witness to

make an identification (Clark, Brower, Rosenthal, Hicks, & Mo-

reland, 2013). However, the influence of administrators on wit-

nesses need not be intentional. In the cue-disruption paradigm, all

administrators know who the suspect is but half of them are

prevented from sending cues (whether intentionally or unintention-

ally) to the witness during the administration of the lineup. In one

study, for example, the contact between the administrator and the

witness was limited by having the administrators stand behind the

witnesses while they viewed a photo-array. Witnesses were less

likely to identify the suspect when the administrator stood behind

them than when the administrator sat in front of or beside the

witness (Haw & Fisher, 2004). In the double-blind paradigm,

participants are randomly assigned to be either witnesses or lineup

administrators; half of the lineup administrators are told who the

suspect is and the other half are not (Phillips, McAuliff, Kovera, &

Cutler, 1999). These participant administrators then present the

photo-array to the participant witnesses. Across all paradigms,

when administrators know who the suspect is and are not pre-

vented from sending cues to the witness, witnesses are more likely

to choose the suspect from the lineup, whether the suspect is the

culprit or not (Kovera & Evelo, 2017).

These studies tell us that changes in the behavior of adminis-

trators during the administration of the lineup are responsible for

this increase in witness picks of the suspect. In the steering

paradigm studies, the administrators’ behaviors were intentionally

manipulated to steer the witness toward the suspect and away from

fillers (e.g., Rhead et al., 2015). In double-blind paradigm studies,

observers reported that nonblind administrators placed more pres-

sure on witnesses to choose someone from the lineup than did

blind administrators (Greathouse & Kovera, 2009) and that pres-

sure was directed toward choosing the suspect rather than a filler

(Zimmerman et al., 2017). Nonblind administrators were more

likely to directly ask witnesses about the suspect than were blind

administrators (Zimmerman et al., 2017). Nonblind administrator

influence can be nonverbal as well; nonblind administrators are

also more likely than blind administrators to smile when a witness

is looking at the suspect rather than a filler (Charman & Quiroz,

2016; Zimmerman et al., 2017).

These differences in behaviors between blind and nonblind

administrators affect which photo witnesses choose from lineups,

not whether they make a choice at all. Witnesses are equally likely

to choose someone from a lineup, irrespective of whether the

lineup administrator knows who the suspect is (Greathouse &

Kovera, 2009; Kovera & Evelo, 2017). The increase in witness

identifications of the suspect from single-blind lineup administra-

tions appears to be the result of witnesses who would have iden-

tified a filler (and do so under blind administration) identifying the

suspect instead due to influence from the nonblind administrator

(Kovera & Evelo, 2017). This pattern of findings, replicated in a

number of studies (Charman & Quiroz, 2016; Greathouse & Ko-

vera, 2009; Kovera & Evelo, 2017), is known as the filler-to-

suspect shift and provides compelling evidence that single-blind

lineup administration allows administrators to transmit informa-

tion about who the suspect is to witnesses, even if unintentionally.

Double-Blind Administration Helps Prevent

Postidentification Feedback

In addition to affecting witnesses’ identification decisions,

single-blind lineup administration allows administrators to provide

feedback to witnesses about their decisions. Nonblind administra-

tors react to witness identifications in ways that send information

to witnesses about whether their choice was “correct” (i.e., an

identification of the suspect; Charman & Quiroz, 2016; Garrioch &

Brimacombe, 2001). Two decades of research supports the con-

clusion that providing feedback to witnesses that they identified

the suspect increases their confidence in the accuracy of their

decision, especially among eyewitnesses who have made a mis-

taken identification (Steblay, Wells, & Douglass, 2014; see Rec-

ommendation 6 for a more complete discussion of this research).

This confirming feedback effect attenuates the relationship be-

tween confidence and accuracy (Bradfield, Wells, & Olson, 2002),

rendering witnesses’ reports of their confidence useless for judging

their accuracy (Wixted & Wells, 2017). In addition to preventing

administrators from providing feedback that will influence wit-

nesses’ reports of their confidence, double-blind administration

will also prevent other unwelcome effects of feedback such as the

contamination of witnesses’ memory for the conditions under

which they witnessed the crime (Steblay et al., 2014), the impair-

ment of witness memory for the culprit (Smalarz & Wells, 2014a),

and lessening the ability of jurors to differentiate between accurate

and inaccurate witnesses (Smalarz & Wells, 2014b).

Double-Blind Administration Helps Ensure Full and

Accurate Reports

Knowing who the suspect is may also influence what informa-

tion administrators record about witnesses’ behavior during the

identification procedure. Even though we recommend video-

recording the lineup administration (see Recommendation 7), a

2013 survey indicated that most lineups in the U.S. (⬎75%) are

not video-recorded (Police Executive Research Forum, 2013). In

cases that fail to video-record, the only contemporaneous record of

what happened during the procedure is information memorialized

by the administrator. Lineup administrators often fail to make a

record of the verbatim statement of the eyewitness but instead will

make a note of the gist of what the eyewitness said. It is also

possible that knowing who the suspect is may change how admin-

istrators assess and record witnesses’ choices from lineups. If so,

when witnesses make tentative identifications (e.g., “I don’t know.

I think it may be Number 4, but I’m not certain.”), administrators

who know that the witness is talking about a suspect may record a

positive identification of the suspect whereas administrators who

know that the witness is talking about a filler may record the very

same behavior as a nonidentification or rejection (Rodriguez &

Berry, 2014). Moreover, lineup administrators’ interpretations of

ambiguous eyewitness statements and administrators’ perceptions

of the witness are biased by whether the lineup administrator is

blind or not blind (Charman, Matuku, & Mook, 2019).

Although there are limited empirical data that directly bear on

the effects of administrators’ knowledge of who the suspect is on

their reports, evidence continues to mount that forensic examiners’

expectancies influence their evaluations (for a review, see Kassin,

Dror, & Kukucka, 2013). In addition, there are data from both

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

EYEWITNESS IDENTIFICATION RECOMMENDATIONS

laboratory and field studies suggesting that administrators who are

nonblind record witness choices differently than do administrators

who are blind. Indeed, for 5 years in Queens County, NY, the

District Attorneys’ Office recorded choices made by witnesses

from live single-blind lineups. Supposedly, choices were only

recorded when the administrators judged that it had been made

with a high degree of confidence and was not tentative (as reported

in Mecklenburg, 2006). But in these Queen’s county cases, it was

nonblind administrators who made decisions as to whether to

report an affirmative response as an identification or whether to

dismiss it as something else. In other words, the administrators in

the Queens County cases knew that these were filler picks. This

procedure resulted in a very low rate of reported filler identifica-

tions (between 0.56% and 5.62%). Field studies for which the

lineups were conducted double-blind, however, report filler iden-

tification rates that are much higher 11%–15% (see Klobuchar,

Steblay, & Caligiuri, 2006; Wells, Steblay, & Dysart, 2015). This

difference in filler identification rates could represent a differ-

ence in reporting that derives from whether the lineup was

double-blind.

In other field data from the Evanston Police Department in

Illinois, reports from double-blind lineups were more likely to

involve verbatim reports of witness statements than were reports

from single-blind lineups (83% vs. 39%, Steblay, 2011). The

interpretation of the Evanston data is problematic because only

double-blind administrators were instructed to record what words

the witnesses used to make their identification whereas single-

blind administrators were not.

Controlled experiments have tested how administrators make

records of the behavior of eyewitnesses as a function of whether

the administrator of a lineup was blind and whether a confederate-

witness chose the suspect or a filler (Rodriguez & Berry, 2014,

2019). Although double-blind administrators were just as likely to

report that witnesses had made a positive identification when the

witness identified a filler as when the witness identified the sus-

pect, single-blind administrators were more likely to report incor-

rectly that witnesses who identified a filler had not made an

identification. Moreover, when nonblind administrators recorded

the confidence reported by witnesses, independent coders who

were blind to condition judged the confidence levels of those who

identified suspects to be higher than those who identified fillers

even though the confederate-witness expressed the same level of

confidence in both types of identifications (Rodriguez & Berry,

2019).

These effects of single-blind lineup administration on witnesses’

identification decisions, their confidence, and administrators’ re-

porting behavior support the use of double-blind procedures when

collecting eyewitness identification evidence. The U.S. Depart-

ment of Justice (2017) and the National Research Council (2014)

have made similar recommendations. Nevertheless, many jurisdic-

tions have yet to put this procedure, or a similar alternative, into

practice (Kovera & Evelo, 2017).

Practical Issues in Implementing

Double-Blind Lineups

Some resistance to double-blind lineup administration has come

from a limited-resources argument that certain police departments

are so small that every officer in the department knows the identity

of the suspect. At the state level, this concern has not been found

in actual practice. In 2002, New Jersey became the first state to

mandate double-blind administration, and state officials have re-

ported no problems implementing this policy. In cases of very

small police departments, for example, cooperative agreements

were created to loan officers to nearby departments for the purpose

of conducting double-blind identification procedures. Other states,

from Florida to California, have similarly reported no problems

with conducting double-blind lineups.

Hence, we recommend double-blind lineup administration and

believe that actual practice has proven it to be viable for all

jurisdictions. Nevertheless, in theory any procedure that prevents

the possibility of a nonblind lineup administrator influencing the

eyewitness could be used. With photo lineups for example, it is

possible to use a laptop computer with software that delivers

prelineup instructions, randomizes and presents the photo lineup,

records any identification decision from mouse clicks, and collects

a confidence statement from the eyewitness. With such software,

the eyewitness can self-administer the photo lineup without any-

one else present in the room, thereby guaranteeing that there could

not have been lineup administrator influence over the eyewit-

nesses’ identification or confidence statements. We recommend

that video-recording be used with the laptop procedure just as it is

with a double-blind administrator (see Recommendation 7).

A low-tech alternative to the self-administered laptop procedure

for photo lineups is the self-administered envelope method. With

the envelope method, a photo lineup is prepared with clearly

numbered photos and the page should clearly state the options (to

identify one of the photos, indicate “not there,” or indicate do not

know in a way that is parallel to the instructions in Recommen-

dation 5). The page should also include a confidence question (see

Recommendation 6). A photo lineup with these items should be

placed in a large envelope and sealed. After giving complete

instructions to the eyewitness (see Recommendation 5), the lineup

administrator should tell the eyewitness that the photos are inside

of the envelope. Of course, when using the self-administered

envelope method as an alternative to the double-blind method, the

instruction that the lineup administrator does not know which

lineup member is the suspect cannot be used. The witness should

be instructed to make an identification decision by either circling

the photo of the person they believe to be the culprit or circling the

“none” or “do not know” option below the photos. The witness

should be instructed to place the photo-lineup and responses back

in the envelope before opening the door to tell the lineup admin-

istrator that she or he has finished. The witness should not be

handed the envelope until the lineup administrator is prepared to

leave the room. The outside of the envelope should again tell the

eyewitness to open the envelope and to view the photos only after

the officer has left the room and to replace the photos and the

answers to questions back in the envelope before opening the door

to let the officer know that they are finished. Only when the

witness has confirmed that the identification decision and confi-

dence statement have been completed and placed back in the

envelope should the officer reenter the room and examine the

results.

This envelope method could be adapted for sequential presen-

tation of the photos, with photos placed individually in smaller,

numbered envelopes and instructions to look at each photo in

numerical order and record an identification decision and a con-

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

WELLS ET AL.

fidence judgment before replacing the photo in its envelope and

proceeding to the next envelope. Backloading of the lineup could

be achieved by placing additional envelopes with blank photo

pages in the later numbered envelopes, with an instruction in the

first envelope used for backloading that the lineup procedure is

complete and witnesses should return all materials to the large

envelope and let the administrator know that they are done. As

with all identification procedures, the self-administered envelope

procedure should be video-recorded to ensure that the witness

follows the instructions given.

In addition to concerns about limited resources making it diffi-

cult to implement double-blind procedures, some have objected to

their adoption because of the loss of correct identifications asso-

ciated with double-blind administration (Clark, 2012a, 2012b).

Although it is true that double-blind procedures can reduce both

correct and mistaken identifications, they do so by eliminating the

opportunity for administrators to cue the witness to which lineup

member is the suspect. Given that the legal system requires that an

eyewitness identification be based on the independent memory of

the witness (Perry v. New Hampshire, 2012), the loss of an

identification obtained through administrator influence should not

be a concern. Indeed, some scholars have termed these correct

identifications obtained through suggestive procedures to be “ille-

gitimate hits” (i.e., correct identifications that are not based in the

witness’s memory but instead a product of the cues received from

an administrator; Wells et al., 2012). Thus, double-blind proce-

dures serve to protect suspects’ rights to due process.

Recommendation 4: Lineup Fillers

There should be only one suspect per lineup, and the lineup

should contain at least five appropriate fillers who do not make the

suspect stand out in the lineup based upon physical appearances

or other contextual factors such as clothing or background.

Recommendation 4 concerns what might be considered the most

widely known problem that can afflict lineups, namely a lineup

that is constructed in a way that makes it obvious which member

is the suspect. This idea is the source of scores of cartoons and

jokes about the perceived failings of criminal justice, such as one

that depicts a person embedded in a lineup composed of a dog, cat,

refrigerator, and a microwave oven. Nevertheless, jocular treat-

ments of biased lineups hide a serious problem and this problem

has much greater complexity than meets the eye.

The problem of biased lineups is one of the oldest in the

scientific study of eyewitness identification. In fact, the first pub-

lished experiment on eyewitness identification that manipulated

the presence versus absence of the culprit in the lineup (a now

routine feature of eyewitness identification experiments) was an

experiment in which the researchers also manipulated the lineup

fillers to be either similar or dissimilar to the suspect (Lindsay &

Wells, 1980). Not surprisingly, in culprit-absent lineup conditions

the use of high-similarity lineup fillers strongly reduced mistaken

identifications of the innocent suspect compared with the use of

low-similarity fillers. In culprit-present conditions, the use of these

same high-similarity fillers had only a minor impact on accurate

identifications of the culprit relative to the use of low-similarity

fillers. This pattern of results, showing that using low-similarity

fillers increases the chances of mistaken identification of an inno-

cent suspect, has been repeatedly replicated (Fitzgerald, Price,

Oriet, & Charman, 2013).

Despite the relative ease of replicating the basic finding that

low-similarity fillers increase the risk of mistaken identification of

an innocent suspect, there is not total agreement among eyewitness

scientists regarding the best strategy for choosing fillers to serve in

a lineup. There are two primary strategies for selecting fillers

(Luus & Wells, 1991). One strategy uses the verbal description of

the culprit that the eyewitness provided (e.g., “White male, mid-

20s in age, about 5 feet 10 inches tall, short dark hair, no facial

hair, medium build”). This method of selecting fillers is called the

match-to-description strategy. The alternative strategy, called the

resemble-suspect strategy, involves selecting fillers who physi-

cally resemble the suspect. The resemble-suspect strategy can be

problematic because it has no criterion or “stopping point” for

determining how similar the fillers should be, at times resulting in

lineup fillers who are too similar and leading to a different problem

than a biased lineup (Luus & Wells, 1991). In effect, extremely

high similarity creates a lineup of near-clones, thereby making it

too difficult to identify the culprit from a culprit-present lineup.

The match-to-description strategy, in contrast, has a natural stop-

ping point (the description) and does not risk creating such high

levels of similarity between fillers and the suspect that would

interfere with obtaining accurate identifications of the culprit

(Luus & Wells, 1991). An experiment comparing the two strate-

gies to a biased (low similarity) lineup showed the two strategies

to be equally effective in reducing innocent suspect identifications;

however, the resemble-suspect strategy produced a reduction in

accurate identifications of the culprit, whereas the match-to-

description strategy did not (Wells, Rydell, & Seelau, 1993). These

findings have been replicated (Juslin, Olsson, & Winman, 1996).

Other studies have shown either no detrimental effect on culprit

identifications from using the resemble-suspect strategy and no

evidence that it made the innocent suspect stand out (Tunnicliff &

Clark, 2000), or no advantage one way or the other for match-to-

description versus suspect resemblance strategies (Darling, Valen-

tine, & Memon, 2008). But a large-scale study comparing

description-matched fillers to suspect-matched fillers showed clear

evidence favoring description-matched fillers (Carlson et al.,

2019).

In a meta-analysis of the data on filler similarity, lineups clas-

sified as high similarity produced a reduction in culprit identifica-

tions relative to low similarity lineups but not relative to moderate

similarity lineups (Fitzgerald et al., 2013). As noted in the meta-

analysis, the categorization of low, medium, and high similarity

lineups reflected relative rather than absolute levels of similarity

(the concept of similarity could not be defined in absolute terms,

with clear criteria for each level of sameness). In a more recent

study, fillers who resembled the suspect were selected from either

an extremely large database of faces (which produced very high

similarity fillers) or a more modest sized database (which pro-

duced more moderate levels of similarity; Bergold & Heaton,

2018). Compared with the more modest size database of faces,

using the large database of faces for selecting fillers resulted in a

reduction in accurate identifications of the culprit by producing too

much similarity between the fillers and the suspect. Overall, the

data suggest that using the resemble-suspect strategy could pro-

duce too much similarity between the suspect and the fillers such

that it interferes with identifications of the culprit, especially when

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

EYEWITNESS IDENTIFICATION RECOMMENDATIONS

fillers are selected from large databases. This problematic condi-

tion could become more prevalent as the ability to rapidly search

large databases of faces becomes increasingly more common.

A Blended Approach for Choosing Fillers

The net result of these complex problems is that the science has

not yet been able to specify what the optimal level of similarity of

fillers to the suspect ought to be and thus, at this time, there is no

single strategy or formula for selecting fillers to be used in a

lineup. Nevertheless, there are generally accepted principles re-

garding how fillers should be selected for a lineup, and they tend

to involve a blend of the match-to-description and resemble-

suspect strategies. First, there is general agreement among experts

for a minimal requirement that fillers should fit the description that

the eyewitness gave of the culprit. A failure to match the witness’s

description might introduce a serious bias even though lineup

members may look very similar. For example, a witness might

describe the culprit as “male, 40s, shaven head, striking blue eyes,

very solid build around the neck and shoulders.” A subsequent

lineup could have an array of people who are very similar in

appearance; however, if only one of the lineup members had blue

eyes, that individual becomes distinct within the array in that his

face will match the memory of the perpetrator more so than will

the fillers.

An exception to this general match-to-description principle

should be made when the suspect does not fit the witness’s

description. For example, a person might become a suspect for

reasons other than his or her appearance. Furthermore, if the

description of the culprit mentioned a moustache, but the suspect

does not have a moustache, then the fillers also should not have a

moustache. In other words, if there is a discrepancy on some

physical feature between the eyewitness’s description of the culprit

and the appearance of the suspect, the fillers should match the

suspect’s appearance (rather than the witness’s description of the

culprit) on that feature.

Another situation in which match-to-description is inadequate is

when the description is vague, general, or sparse. For example, the

description “young White male” is inadequate. In this case, the

fillers should match the suspect on basic “default” characteristics

such as facial hair, hairstyle, and general body build (Lindsay,

Martin, & Webber, 1994). Hence, the match-to-description method

for selecting fillers should be used only if the description is

complete.

Sometimes a suspect has a unique feature such as a tattoo or a

scar, making it very difficult to find a filler who matches on that

unique feature. There are two general approaches to dealing with

this issue, made easier by advances in technology. One approach is

to duplicate this feature on the fillers, which could be done

electronically in the case of photo lineups. The other approach is

to cover the unique feature on the suspect and then place that same

cover on each of the fillers at the same location on their bodies.

Both the “duplicate” and “cover” approaches appear to be equally

effective (Colloff, Wade, & Strange, 2016). However, these alter-

ations must be done in a way that does not make the suspect stand

out in a mock witness test (see next section).

Another complication in selecting fillers can occur when a

person becomes a suspect based on resemblance to a facial com-

posite (e.g., forensic sketch or computer-generated face) or a

surveillance image (although see discussion of the problems asso-

ciated with using culprit-match to a composite in Recommendation

2: The Evidence-Based Suspicion Recommendation). If the person

became a suspect based on resemblance to a composite or to a

surveillance image whereas the fillers were chosen based merely

on their match to the witness’s verbal description, then there is a

risk that the suspect will stand out. Hence, in such cases, fillers for

a lineup need to be chosen based on their similarity to that same

composite or surveillance image rather than chosen based on the

verbal description given by the eyewitness (see Wixted & Wells,

2017).

“Mock Witness” Testing

There is broad agreement that the lineup that is created should

be able to pass a “mock witness” test. A mock witness test is one

in which a large number of people are individually given the

description that the witness had given of the culprit, then shown

the lineup and asked which person they think is the suspect. The

ideal outcome from such a test would be if the suspect were

selected by these mock witnesses only 1/Nth of the time, where N

is the nominal number of lineup members. So, for a six-person

lineup, a good outcome would be if the suspect were picked one

sixth of the time. Consider again the witness description “male,

40s, shaven head, striking blue eyes, very solid build around neck

and shoulders” and the suspect is the only one with blue eyes. This

lineup is likely to result in most mock witnesses choosing the

suspect rather than spreading their choices across the lineup mem-

bers. We are not suggesting that police should be required to

conduct a mock witness test on each lineup they create. Instead, we

believe that a conscientious and objective detective would have a

good sense of whether the lineup was fair without conducting a

mock witness test with a large number of people. However, we do

recommend that a nonblind officer building the lineup ask at least

one or two other people (blind as to which person is the suspect)

to review the witness description and evaluate the lineup with

respect to whether it would pass a mock witness test. We also

recommend that every lineup report include a written record of

how the fillers were selected for the lineup.

Mock witness tests have been around since the 1970s (Doob &

Kirshenbaum, 1973) and a number of different statistics have been

developed to estimate lineup bias from mock witness results (e.g.,

Malpass, 1981; Tredoux, 1998; Wells, Leippe, & Ostrom, 1979).

Mock witness measures tend to predict choices of eyewitnesses

from culprit-absent lineups (Tredoux, Parker, & Nunez, 2007),

though mock witness tests do have limits (see Wells & Bradfield,

1999). One of these limits is that a mock witness test is insensitive

to whether the level of similarity between fillers and the suspect is

too high, so high that it would likely harm rates of accurate

identifications if the suspect is the actual culprit. For example, a

lineup of clones would produce a good result from a mock witness

test (one sixth of choices are of the suspect), even though it would

not be a good lineup in that most witnesses would not be able to

distinguish between the culprit and the fillers. In addition, because

mock witness tests use the description of the culprit provided by

the eyewitness, a mock witness test can appear perfectly fair if the

description is sparse whereas that same lineup can appear quite

unfair if the description is detailed (Mansour, Beaudry, Kalmet,

Bertrand, & Lindsay, 2017).

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

WELLS ET AL.

Background, Clothing, and Other Contextual Factors

The physical characteristics of the fillers are not the only factors

that can make the suspect stand out in a lineup. In photo lineups,

for example, the background of the photos, the size or brightness

of the images, and the source of the photo could make the suspect’s

photo stand out from the others. Backgrounds are relatively easy to

fix on photos with modern editing software. Similarly, clothing

can easily be eliminated from headshots with editing software.

Sometimes, a suspect’s photo is from a different source (e.g.,

employment ID, social media, Department of Motor Vehicles

driver’s license) than a police department’s usual source for photo

lineups (e.g., mugshots), and this can make the suspect’s photo

stand out. This type of discrepancy will require either careful

electronic editing or perhaps going to the same source (e.g.,

Department of Motor Vehicles drivers’ licenses) to locate appro-

priate fillers.

It is not the case that every aspect of background, clothing, and

other features must be exactly the same. The critical issue is

whether the suspect stands-out in the context of the other fillers.

So, for example, if the background of every photo was different

from every other photo in the lineup, then perhaps the suspect’s

photo would not be any more distinct than other lineup members.

Articles of clothing (e.g., hats, shirts, etc.), however, could be of

special concern. Specifically, if there are reasons to believe that the

clothing worn by the suspect is similar to that worn by the culprit,

then (a) every filler needs to be clothed that same way, (b) the

suspect needs to have his or her clothes changed to blend in with

the fillers, or (c) each lineup member’s clothing must be obscured

from the view of the witness. For cases in which the identification

of clothing is thought to have potential probative value, a separate

identification procedure involving only the papers of clothing (a

clothing lineup) could be conducted independently of the identi-

fication of the suspect (see Lindsay et al., 1994; Lindsay, Wall-

bridge, & Drennan, 1987).

The Single-Suspect and Minimum of Five

Fillers Requirement

A central feature of this recommendation is that the lineup

should have only one suspect. There are several reasons why there

should be only one suspected person in the lineup with the remain-

der having the status of being known-innocent fillers. Suppose, for

example, that a lineup were composed entirely of suspects and no

fillers. All-suspect lineups have been likened to a multiple-choice

test in which there is no wrong answer (Wells & Turtle, 1986;

Wixted & Wells, 2017). The value of having known-innocent

fillers is that unreliable eyewitnesses are likely to err on a filler

rather than on an innocent suspect (assuming that the lineup is

composed of good fillers).

Consider again Table 1 of the current article in which we

displayed the outcomes of lineups in actual cases. These were all

single suspect lineups in which the suspect was embedded among

fillers. Notice that these eyewitnesses to serious crimes identified

fillers approximately 37% of the time they made an identification.

If everyone in those lineups had been a suspect, all 37% of these

would be mistaken identifications of innocent suspects who would

then be subject to arrest and possible prosecution. These field data

reinforce the dangers of having multiple suspects in a lineup, an

issue that was first documented over 30 years ago through statis-

tical proofs using data from eyewitness identification experiments

(Wells & Turtle, 1986).

The recommendation that there should be at least five known-

innocent fillers (thereby creating a six-person lineup) for a single

suspect is somewhat arbitrary. At a theoretical level, we can say

that an innocent suspect is better protected from mistaken identi-

fication with a six-person lineup than a five-person lineup, which

is better protection than a four-person lineup, and so on, as long as

other things are equal (e.g., how good the fillers are). However,

there are diminishing returns (in terms of restricting mistaken

identifications of innocent suspects). After all, in terms of protect-

ing an innocent suspect, an increase from two fillers to three fillers

is greater than an increase from five to six, which in turn will have

more impact than an increase from seven to eight fillers.

Some jurisdictions in the U.S. use more than six. Some juris-

dictions in Australia use lineups that vary from eight to 10 mem-

bers, and England and Wales also use more than six members. Of

course, as noted in the previous discussion on selecting fillers for

lineups, it is not the nominal size of the lineup that matters so much

as the number of lineup members who fit the description of the

culprit. A lineup of 12 people would be less effective than a lineup

of six people if the 11 fillers in the 12-person lineup did not fit the

description of the suspect, whereas the five fillers in the six-person

lineup did fit the description. Of course, a 12-person lineup in

which all fillers were a good match to the suspect would provide

more protection for the innocent suspect than would a six-person

lineup in which all the fillers were good fillers. Although one

eyewitness identification researcher has argued strongly for large

increases in photo lineup sizes to as high as 120 (Levi, 2011), at

this point we are not convinced that large increases in lineup size

are warranted in practice. Among other things, all but one of the

lineup members must be a priori cleared as possible suspects so

that they can have the definitive status of known-innocent fillers.

Establishing the innocence of a large number of fillers is not an

issue in a lab experiment, but it would be in actual practice. As

lineup size increases, it is also increasingly difficult to locate fillers

who properly fit the description of the culprit. Adding extremely

poor fillers to a lineup can enhance eyewitnesses’ confidence in a

mistaken identification (Charman, Wells, & Joy, 2011). In addi-

tion, there are concerns about potential loss of correct identifica-

tions (of the culprit) if a lineup becomes too large.

A Theoretical Note About Lineup Fillers

There is currently debate in the eyewitness identification liter-

ature about how good lineup fillers manage to improve overall

lineup performance (i.e., how do they reduce mistaken identifica-

tions of innocent suspects more than they interfere with identifi-

cations of the culprit?). Some have suggested that the use of good

fillers helps witnesses decide which facial features are diagnostic

(e.g., Colloff et al., 2016; Wixted & Mickes, 2014), whereas others

have argued that good fillers simply siphon false positive identi-

fications away from the innocent suspect more than they siphon

from the culprit (e.g., Smith et al., 2017; Smith, Wells, Smalarz, &

Lampinen, 2018; Wells, Smith, & Smalarz, 2015). It is possible

that both of these processes are involved and these might not be the

only two possibilities. Regardless of which processes underlie the

contribution that good fillers make to improved witness accuracy,

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

EYEWITNESS IDENTIFICATION RECOMMENDATIONS

the answer to this theoretical question ultimately could help to

clarify the best strategies for choosing fillers for lineups, and

should be considered in future research.

With photo lineups, care should be taken not only with respect

to the choice of filler photos but also the photo of the suspect.

Often, there are multiple possible photos of the suspect to choose

from. When possible, the suspect’s photo should be clear and look

as much as possible like the suspect appeared at the time of the

crime. If the crime was recent, for example, older photos of the

suspect should not be used if a more recent photo is available.

Recommendation 5: Prelineup Instructions

When inviting an eyewitness to attend a lineup procedure (photo

lineup or live lineup), police should not inform the eyewitness of

any information that the witness has not already provided and

certainly should not suggest that the suspect who will be in the

lineup has been arrested or that the culprit will be present in the

identification procedure. The eyewitness should be instructed that

(a) the lineup administrator does not know which person is the

suspect and which persons are fillers; (b) the culprit might not be

in the lineup at all, so the correct answer might be “not present”

or “none of these”; (c) if they feel unable to make a decision they

have the option of responding “don’t know”; (d) after making a

decision they will be asked to state how confident they are in that

decision; and (e) the investigation will continue even if no identi-

fication is made.

This recommendation addresses the concern that, if witnesses

approach the identification test with the mistaken belief that the

culprit must be present in the lineup, they may be predisposed

toward making a positive identification. Intuition would suggest

that many witnesses are likely to presume—based on the invitation

to view a lineup—that the police must have a strong suspect and,

thus, their task is to determine which lineup member the suspect is.

There are various strands of evidence suggesting that witnesses

make this assumption. For example, in one study, 90% of a large

sample of witnesses indicated immediately after making their

identification decision that they had expected the culprit to be

present in the lineup and believed their task was to identify him or

her (Memon, Gabbert, & Hope, 2004). In another study, witnesses

either viewed a lineup containing the culprit or a lineup with the

culprit removed but not replaced by another filler (Wells, 1993). In

the former condition, 54% of witnesses picked the culprit, 25%

picked one of the fillers, and 21% made no choice. When the

culprit was removed but not replaced, it might be expected that

around 75% of witnesses (i.e., 54% ⫹ 21%) would make no

choice. Instead, only 32% of witnesses made no choice, with 68%

distributing their choices across the various fillers. This finding

suggests that witnesses are predisposed toward making a positive

identification, though not necessarily with high confidence, pro-

vided some lineup member appears to be a reasonable match to

their memory.

The most compelling evidence of the usefulness of this recom-

mended instruction comes from studies comparing the identifica-

tion performance obtained when witnesses are instructed that the

culprit might or might not be present in the lineup to identification

performance obtained under conditions where no such warning is

provided. The former condition has typically been referred to as an

unbiased instructions condition, the latter as a biased instructions

condition. The way in which these two conditions have been

enacted has varied. For example, unbiased instructions have often

simply involved providing the culprit might or might not be

present warning, although sometimes this has been accompanied

by an instruction that there is no need to pick anyone or even an

instruction emphasizing that the consequences of a wrong decision

may be dire. Biased instructions may also take many forms, such

as failing to forewarn the witness that the culprit might not be in

the lineup or by strongly implying or even stating that the culprit

is in the lineup. For example, asking the witness to select which

lineup member is the culprit strongly implies that the witness is

expected to make a positive identification decision.

Several features of the empirical findings on the effects of

biased versus unbiased instructions warrant mention. First, the

findings of three major reviews using meta-analytic procedures

demonstrate that witnesses were more likely to make a positive

identification decision when the lineup instructions were biased

(i.e., no warning regarding possible absence of the culprit) than

when they were unbiased (Clark, 2005; Steblay, 1997, 2013).

Second, although the magnitude of this effect varied across the

reviewed studies, presumably depending on the conditions at

memory encoding and the identification test, the increased likeli-

hood of choosing was reflected in increased positive identifica-

tions from both culprit-absent and culprit-present lineups. Thus, it

had both positive and negative effects. Third, subsequent to the

publication of these reviews, findings from two studies with very

large sample sizes have reinforced the impact of biased versus

unbiased instructions with both adult and child witnesses (Brewer

& Wells, 2006; Keast, Brewer, & Wells, 2007). Biased instructions

contributed to higher rates of both mistaken identifications and

correct identifications, suggesting that a failure to warn witnesses

that the culprit may not be present in the lineup contributes to

witnesses being prepared to accept less evidence (i.e., establishing

a more liberal decision criterion) for making a positive identifica-

tion decision. This lower threshold for choosing could be benefi-

cial if the base rate of guilty suspects were high, but it would be

detrimental if the base rate of guilty suspects were low, as may

often be true of real police lineups (e.g., Wixted et al., 2016).

The precise impact of witnesses lowering their decision criterion

will also depend on factors such as the characteristics of the

various lineup members (Brewer, Weber, & Semmler, 2005). For

example, if the lineup is biased against the suspect by virtue of the

suspect being the only plausible lineup member, a lower decision

threshold would increase the likelihood of (a) a correct identifica-

tion if the suspect is the culprit and (b) a mistaken identification of

an innocent suspect if the suspect is innocent. In contrast, if the

suspect is presented in an array of highly plausible fillers, instruc-

tions that lead witnesses to set a lower decision threshold may lead

to responses being spread more evenly across all lineup members.

It is important to note one qualification of the pattern of findings

typically found when contrasting the impact of unbiased versus

biased instructions. The effect of unbiased instructions may be

negated if the witness receives an explicit suggestion prior to

viewing the lineup that the culprit may be present in the lineup. In

a study in which witnesses were presented with culprit-absent

lineups only, witnesses received the suggestion surely you are

going to be able to pick the person out from the lineup prior to

receiving the instruction that the culprit may or may not be present

in the lineup (Quinlivan et al., 2012). Witnesses who received that

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

WELLS ET AL.

suggestion, followed by unbiased instructions, were three times

more likely to identify the suspect (who, in this study, was inno-

cent) than were witnesses who received unbiased instructions

without any prior suggestive comment. Witnesses in the former

condition were also more confident in the accuracy of their erro-

neous identification decision than those in the latter condition.

The research findings on instructing witnesses prior to their

viewing a lineup have clear implications. First, when inviting an

eyewitness to attend a lineup procedure, police should not suggest

that a suspect has been arrested or that the culprit will be present

in the identification procedure. Second, in our experience some

witnesses seem to be under the misconception that the investiga-

tion hinges on their identification decision. Consequently, wit-

nesses should also be told that the investigation will continue even

if no identification is made.

Third, it should be made quite clear to the witness that the

culprit may or may not be in the lineup and that they do not have

to select any of the lineup members. In other words, responses

such as not present or none of these are quite appropriate. A

reminder that the witness does not have to choose anyone from the

lineup is important. A large percentage of witnesses are under the

impression that the culprit is present and their task is to identify

him (Memon et al., 2004). Fourth, to ensure that the witness does

not lose sight of the fact that such response options are appropriate,

there should be an explicit not present response option accompa-

nying the lineup members from which the eyewitness can choose.

In the case of a photo lineup, this option may be located below the

array of lineup faces. In the case of a live lineup, a response sheet

that shows the possible response options can be used: lineup

member numbers (i.e., 1, 2,...,6),notpresent, and do not know.

Finally, lineup administration procedures should accommodate

the possibility that the witness may look at the lineup and be

unwilling to pick someone or to respond not present because, for

example, they cannot decide between two or more lineup members

or they are uncertain about whether the culprit is in the lineup. For

that witness (i.e., one who really has no idea about what to do), an

appropriate response may be to say “don’t know” rather than not

present. Both adult and child witnesses use options such as do not

know or not sure when they are made explicitly available, with

frequency of use varying considerably depending on the encoding

stimuli and lineup materials (Brewer, Keast, & Sauer, 2010; Weber

& Perfect, 2012; Zajac & Karageorge, 2009). Moreover, there is

some evidence indicating that positive identifications of a suspect

are more diagnostic of suspect guilt when they are made in the

presence of a do not know option compared with when no such

option existed (Weber & Perfect, 2012).

The availability of an option to respond do not know is likely to

reduce the likelihood of low confidence positive identifications,

which research shows are often inaccurate (Brewer & Wells, 2006;

Wixted & Wells, 2017). Although there are strong grounds for

always questioning the reliability of low confidence identifica-

tions, there may be a tendency on the part of police or prosecutors

to argue, for example, that the initial low confidence identification

was made from a photo that was not a good likeness to the suspect,

thereby purportedly providing a reasonable explanation for the

witness’s low confidence. Further, although studies have shown

that mock-jurors (appropriately) downgrade the credibility of the

witness and the culpability of the defendant when cross-

examination highlights a disparity between a witness’s expressed

confidence at initial identification (low confidence) versus an

in-court identification (high confidence; Bradfield & McQuiston,

2004; Jones, Williams, & Brewer, 2008), this effect does not

always occur (e.g., Douglass & Jones, 2013). For example, if a

witness expressed high confidence in their identification of the

suspect during a trial despite a low confidence initial identification,

jurors tended to excuse the confidence inflation if the witness

appeared to have experienced an epiphany about their initial low

confidence identification (e.g., “I wasn’t very confident at the time

of the identification because I was scared back then”). As long as

these types of excuses for initial low-confidence identifications are

permitted, we argue that it is crucial that there is an explicit do not

know response option, which can be located alongside the not

present option.

We note that some jurisdictions have used what has been re-

ferred to as an “appearance change” instruction. This instruction

was among a set of guidelines developed by a U.S. Department of

Justice (DoJ) working group on the collection of eyewitness evi-

dence. Specifically, the DoJ guidelines recommended that, prior to

being shown a lineup, eyewitnesses should be told that “individ-

uals depicted in lineup photos may not appear exactly as they did

on the date of the incident because features such as head and facial

hair are subject to change” (Technical Working Group for Eye-

witness Evidence, 1999, p. 32). We have not included this instruc-

tion in the current set of recommendations because subsequent

research has shown that the appearance change instruction in-

creased false identifications but did not increase culprit identifica-

tions (Charman & Wells, 2007; Molinaro, Arndorfer, & Charman,

2013).

As a final note on instructions, we recommend that in addition

to the witness having the instructions in writing, the lineup admin-

istrator should read the instructions aloud to witnesses, pausing

after each point to make sure that the witness understands each

point.

Recommendation 6: Obtain an Immediate

Confidence Statement

A confidence statement should be taken from witnesses as soon

as an identification decision (either positive or negative) is made.

For double-blind lineups, “immediate” means that the confi-

dence statement should be secured with only the blind adminis-

trator in the room and before the case detective or any other

nonblind individuals are allowed into the room. Note that a con-

fidence statement should be recorded if a witness positively iden-

tifies someone or if a witness indicates the culprit is not present,

says they do not know, or indicates that they are not sure.Asan

example, the confidence statement could be collected as a numeric

response (i.e., on a scale from 0% confident to 100% confident).

Alternatively, confidence could be collected using a verbal scale

(e.g., “positive,” “probably,” “maybe”). If neither scale is used and

witnesses simply use their own words, a verbatim record of their

verbal statements (or preferably a video-recording, see Recom-

mendation 7) should be made, not a summary or paraphrase

generated by the lineup administrator. If the witness’s response is

“don’t know,” a confidence statement should be recorded if the

witness spontaneously provides one. Otherwise, no confidence

statement should be solicited for “don’t know” responses. How-

ever, it could be useful to let the eyewitness state a basis for the

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

EYEWITNESS IDENTIFICATION RECOMMENDATIONS

“don’t know” response (e.g., did not get a good view, none are

familiar). The prelineup instructions should have already commu-

nicated to the eyewitness that a confidence statement will be

requested (see Recommendation 5). These prelineup instructions

help to prevent witnesses from drawing erroneous conclusions that

their confidence is only being assessed because the lineup admin-

istrator thinks their decision is incorrect.

This recommendation is based on the fact that eyewitness con-

fidence is a useful cue to the accuracy of a witness’s decision when

instructions do not imply the presence of a culprit (Quinlivan et al.,

2012), double-blind administration is used, and fair lineups are

presented (Wixted & Wells, 2017). Specifically, confidence pre-

dicts accuracy among witnesses who choose from a photospread

when immediate confidence reports are obtained (e.g., Brewer &

Wells, 2006; Palmer, Brewer, Weber, & Nagesh, 2013; Sauer,

Brewer, Zweck, & Weber, 2010). Note that forcing witnesses to

withhold their confidence reports for as little as 5 min has been

found to undermine the predictive value of confidence, unless

witnesses are required to spend that time reflecting on the witness-

ing conditions and identification procedure, or producing reasons

why they might have made an incorrect decision (Brewer, Keast,

& Rishworth, 2002).

The corpus of data suggesting that confidence is a useful cue to

eyewitness accuracy stands in stark contrast to DNA exonerations

in which innocent people were mistakenly identified in court by

highly confident eyewitnesses, most of whom were demonstrably

less confident at the time of the initial identification (e.g., Garrett,

2011). If confidence is a useful cue to accuracy, how could these

mistaken witnesses have been so confident? The answer lies in the

fact that intervening postdecision events can dramatically shift

witnesses’ reports of their confidence from low at the time of

identification to high at the time of trial. As a result, these shifts

render delayed reports nearly useless as cues to accuracy. How-

ever, if confidence reports are taken immediately after an identi-

fication decision, the integrity of confidence as a cue to accuracy

is enhanced considerably.

We recommend that confidence statements be collected on a

graded scale using words (e.g., “positive,” “probably,” “maybe”)

or numbers (e.g., from 0% confident to 100% confident). The key

element of this recommendation is that an immediate record of a

witness’s confidence is collected. Immediate confidence estimates

are the only way to ensure that postidentification variables do not

contaminate subsequent confidence reports.

We focus on immediate confidence reports because seemingly

innocuous postdecision events can contaminate witnesses’ confi-

dence reports, undermining what could have been forensically

meaningful information from an eyewitness. One of the most

heavily researched of these events involves postidentification feed-

back offered by a lineup administrator in the form of a simple

comment confirming the witness’s decision. Such feedback can

dramatically inflate confidence reports. In the original test of the

postidentification feedback effect (Wells & Bradfield, 1998), re-

searchers provided inaccurate witnesses with such a comment:

“Good, you identified the suspect.” That simple statement resulted

in 50% of inaccurate witnesses reporting that their confidence was

a 6 or 7 on a 7-point scale (compared with only 15% of witnesses

in the control condition). Importantly, the inflated confidence

report created by this feedback is a retrospective judgment because

witnesses indicate how confident they were at the time of their

identification, before they knew their decision was correct. Any

resulting confidence inflation obviously obscures a true picture of

the witness’s experience at the time of the identification decision.

The effect of postidentification feedback is robust and reliable (see

meta-analysis by Steblay et al., 2014). It also features prominently

in some judicial decisions as courts grapple with how to ensure

that eyewitness identification testimony truly reflects the witness’s

experience of making an identification decision, rather than the

influence of extramemorial variables (e.g., New Jersey v. Hender-

son, 2011; Oregon v. Lawson, 2012).

Beyond affecting witnesses’ self-reports, distorted confidence

judgments complicate assessments of witness identification deci-

sions. Indeed, several experiments show us that evaluators rate

inaccurate witnesses who have received confirming feedback as

more credible than those who received disconfirming feedback or

no feedback, even when evaluators are instructed to ignore the

feedback and even when an explicit confidence statement is not

available (Douglass, Neuschatz, Imrich, & Wilkinson, 2010). Be-

cause postidentification feedback inflates the confidence of inac-

curate witnesses more than the confidence of accurate witnesses, it

also impairs fact finders abilities to distinguish accurate from

inaccurate witnesses (Bradfield et al., 2002; Smalarz & Wells,

2014b).

Recommending an immediate confidence report is an important

companion to the recommendation that the entire lineup procedure

be video-recorded (see Recommendation 7). Not only can video-

recording demonstrate whether postidentification feedback was

given, but it can also preserve witness nonverbal cues that may

signal accuracy (Matuku, Douglass, & Charman, 2018). In the

absence of a video-recorded identification procedure, triers of fact

who learn that a witness’s confidence has inflated over time are

sometimes unwilling to impugn the witness’s credibility (Bradfield

& McQuiston, 2004), especially if a compelling explanation ac-

companies the inflation (e.g., “I had an epiphany!”; Jones et al.,

2008). However, if evaluators see the identification procedure in

which a witness’s initial confidence is lackluster, their assessments

of a highly confident trial witness are less positive, which is an

important shift when the highly confident witness has identified

the wrong person (Douglass & Jones, 2013).

Beyond distorting evaluators’ ability to assess witnesses, con-

taminated witness reports may also unduly shape preliminary

investigations by (a) triggering biased evaluations of subsequent

pieces of evidence and/or (b) biasing the integration of evidence

against the identified suspect. For example, a witness who is

highly confident in a mistaken identification may trigger investi-

gators to view the suspect’s alibi as weaker than it would have

been otherwise or may suggest to investigators that they suspend

pursuit of additional suspects (for a discussion of these effects see

Charman, Douglass, & Mook, 2019). Recording a witness’s im-

mediate confidence eliminates the potential for subsequent infla-

tions to go unnoticed.

If lineup administrators follow the recommendation to conduct

double-blind procedures (Recommendation 3), they will be unable

to provide postidentification feedback because they will not know

which person is the suspect. Therefore, they will be unable to

confirm (or disconfirm) any decision made by the witness. How-

ever, even vague positive comments (“You have been a good

witness”) can be interpreted by witnesses as confirming feedback

(Dysart, Lawson, & Rainey, 2012). Therefore, consistent with the

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

WELLS ET AL.

recommendation on prelineup instructions (Recommendation 5),

witnesses should be explicitly told that the lineup administrator

does not know which person is the suspect and which people are

fillers.

Even if administrators correctly implement double-blind proce-

dures and prelineup instructions, it is still imperative to collect

immediate confidence reports because witnesses may encounter

postidentification feedback in multiple other ways. For example,

the mere fact that a case proceeds to trial is a form of confirmation

that a prior identification was correct (see also Berkowitz &

Frenda, 2018; Wade, Nash, & Lindsay, 2018 for other types of real

world contamination). Another potential source of contamination

comes from witnesses who conduct their own investigations via

social media searches, often accompanied by cowitnesses who

may be able to “confirm” that an Instagram or Facebook photo

represents the culprit (Douglass & Smalarz, 2019).

Recommendation 7: Video-Recording

The entire identification procedure, including prelineup instruc-

tions and witness confidence statement, should be video-recorded.

To preserve a faithful record of the conditions under which

witnesses make their identifications, we recommend that the entire

interaction between the police and the witness be video-record-

ed—from the time when the witness is given the very first pre-

lineup instructions through to the completion of the procedure

when the witness has provided identification and confidence state-

ments. The video must capture all features of the administration,

including the interactions among the lineup administrator, the

witness, and the lineup members (whether they are presented live

or by photo). Under some circumstances, fully capturing the lineup

administration may necessitate cameras recording from multiple

angles. Although the authors of the original scientific review paper

(Wells et al., 1998) recognized some of the benefits of video-

recording a lineup administration, they did not include video-

recording as one of their recommendations at that time. In part,

video-recording the identification procedure was left out of the

original set of recommendations because of concerns about the

costs associated with making the record, including costs for equip-

ment and materials. Since that time, the cost of video-recording

interactions has decreased considerably and most adults have

cellular phones capable of rendering high-quality video-records.

Furthermore, increasing numbers of police now have access to

body cameras that can be positioned to make video-recordings of

identification procedures. Thus, we believe it is time to make the

video-recording of lineup administrations standard practice as it is

in some jurisdictions and in other countries (e.g., Australia).

In his response to the initial scientific review paper, Kassin

(1998) noted two reasons why video-recording identification pro-

cedures should be considered best practice. First, police reports of

what happened during an identification procedure may be incom-

plete or even inaccurate given that they are based on officers’

recollections of what happened during the procedure; recollections

that are subject to the typical foibles of human memory. Second, it

is possible that video-recording the identification procedure could

encourage administrators to carefully adhere to best practices and

deter them from engaging in any suggestive practices (Kassin,

1998). In the remainder of this section, we expand upon each of

these justifications for video-recording identification procedures,

citing relevant research when it is available.

Clearly, video-recording identification procedures has the ben-

efit of providing a more precise and accurate accounting of what

happened during the procedure. Although it might be a rare oc-

currence, video-recording the procedure will make it more difficult

for police officers to intentionally fabricate their reports of what

occurred during the lineup administration. There is evidence of

some police officers purposefully misrepresenting case-related

events (Orfield, 1987, 1992; Slobogin, 1996), but even in the case

of a conscientious, well-meaning officer, there are benefits of a

more accurate reporting of the procedure through video-recording.

For example, the memories of police officers are subject to the

same cognitive errors as are those of others, including interference

(e.g., Kane & Engle, 2000; Nairne, 2002) and memory intrusions

from mental scripts about what usually occurs (e.g., Greenberg,

Westcott, & Bailey, 1998; Kleider, Pezdek, Goldinger, & Kirk,

2008). One can easily imagine that memories from other lineups

conducted or mental scripts for what should have happened could

interfere with what a police officer remembers, and consequently

reports, about an identification procedure. Indeed, there have been

cases in which a police officer testified to reading mandated

instructions verbatim to the witness, whereas the video-recording

subsequently revealed improvisation that introduced suggestive-

ness into the procedure. Whether or not the officer was attempting

to intentionally mislead or was genuinely mistaken, the actual

procedure could be easily reviewed if it is video-recorded, and the

fallibility of memory is no longer an issue.

In contrast, witnesses’ memories for a procedure are not likely

to be influenced by interference or mental scripts regarding what

usually happens given that most witnesses have never participated

in a lineup before. However, their lack of expertise with the

procedure will likely result in reports that are less complete (e.g.,

Chase & Simon, 1973; Gobet & Simon, 1996). Moreover, their

reports of the procedure may be affected by the decay of their

memory trace over time (Deffenbacher, Bornstein, McGorty, &

Penrod, 2008) or postevent information (Loftus, 2004). The video-

record would also provide clear information about how long it took

the witness to make an identification. Witnesses’ estimates of time

are often inaccurate (Yarmey, 2000), but time-to-identification is

an important postdictor of witness accuracy: Identifications made

more quickly are more likely to be accurate than are those made

more slowly (Sporer, 1992). Although research has not provided a

definitive cutoff that allows us to discriminate between accurate

and inaccurate identifications (Brewer, Caon, Todd, & Weber,

2006; Weber, Brewer, Wells, Semmler, & Keast, 2004), having an

accurate record of a witness’s time-to-identification can provide

useful information to fact finders. In sum, despite knowing no

evidence directly examining the accuracy of police officers’ and

witnesses’ memories of identification procedures, we are confident

that the basic cognitive research on memory errors generalizes to

this context.

Although there is no direct evidence about the accuracy of

police reports of identifications, police reports of witness inter-

views and suspect interrogation procedures omit important details

about the procedures used (Kassin et al., 2017; Lamb et al., 2000).

Moreover, testimony from witnesses about what happened during

the procedure is even more likely to be subject to error given that

police reports are likely to be written shortly after the event,

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

EYEWITNESS IDENTIFICATION RECOMMENDATIONS

whereas testimony may be given months or years later. Video-

recording the identification procedure allows police command

staff to confirm that proper procedures are being followed. In

addition, video-recording the procedure eliminates the need to rely

on police officers’ or witnesses’ memories for the event at trial. In

many cases, detectives might not have to testify during suppression

hearings if there is a video available, thereby saving staff resources

and money by not having to pay officers to attend court.

When deciding a motion to suppress the identification, the judge

could review the video and evaluate the suggestiveness of the

procedure herself rather than relying on attorneys’ characteriza-

tions of the procedure based on their readings of police reports and

witness testimony. Defense attorneys might review the video-

recording when deciding whether to offer a plea or encourage a

client to accept one, whereas prosecutors might review the video-

recording when evaluating the strength of a case and how to

proceed. Moreover, in the United States a suspect does not have

the right to have an attorney present at a preindictment identifica-

tion procedure and never has the right to have an attorney present

at a photo lineup procedure (United States v. Ash, 1973). In the

absence of an attorney to view the identification procedure, the

video-recording could help defense attorneys to better represent

their clients following identification. In addition, eyewitness ex-

perts could be asked to review the video-recording and either

testify or prepare a written report about the identification proce-

dure. Finally, the video-recording could be introduced as evidence

at trial so that fact finders can judge for themselves whether the

procedure was suggestive, whether the witnesses engaged in any

behaviors that either enhanced or diminished their credibility, or

whether the witness’s confidence was inflated (e.g., Douglass &

Jones, 2013).

Although the data on this topic are limited, and there are no

known data on how judges or attorneys might evaluate video-

records of identification procedures, there are a handful of studies

that explore the effects of watching the video-recording of an

eyewitness identification procedure on mock juror judgments.

Most of these studies have tested whether viewing a video-

recording of the identification procedure helps jurors evaluate the

accuracy of witness identifications (Beaudry et al., 2015; Reardon

& Fisher, 2011). In one study, watching the video-recording

helped participants distinguish between accurate and inaccurate

witnesses (Reardon & Fisher, 2011). In another study (Beaudry et

al., 2015), confirmatory feedback interfered with the ability of

participant-judges to differentiate between accurate and inaccurate

witnesses, with participants judging the witnesses who received

confirmatory feedback to be more accurate irrespective of their

actual accuracy. Participants’ judgments of accuracy were unaf-

fected by whether the lineup was conducted using single- or

double-blind procedures, irrespective of whether the participants

had viewed the video of the procedure. However, these videos

were relatively short (M ⫽ 1 m, 37 s) and it is not clear from the

report of the study how much influence the administrators exerted

in the single-blind conditions. In a study that manipulated whether

the videotape contained evidence of administrator influence,

watching a video-recording of a single-blind versus a double-blind

procedure did influence jurors’ verdicts (Modjadidi & Kovera,

2018). Specifically, watching the video-recording increased par-

ticipants’ ratings of procedural suggestiveness and decreased their

guilty verdicts when the video-recording depicted witness steering

that can occur in a single-blind administration as opposed to a

double-blind administration. Thus, the limited evidence available

suggests that watching the video-recording of a lineup administra-

tion will help jurors (and perhaps judges and attorneys) evaluate

whether a lineup procedure was suggestive.

Finally, video-recording the identification procedure could en-

courage administrators to ensure that their conduct conforms to

best practices and deter them from engaging in any suggestive

procedures. Although we know of no studies that directly test

whether video-recording can change the behaviors of administra-

tors of identification procedures, the issue has been examined in

the context of interrogation practices. In a mock crime and inter-

rogation paradigm, researchers manipulated the actual guilt-

innocence of suspects who were then subjected to interrogations

conducted by police officers who were either informed or unin-

formed regarding the researcher’s recording of the interrogation

session (Kassin, Kukucka, Lawson, & DeCarlo, 2014). Police

officers who knew that they were being recorded were signifi-

cantly less likely to use minimization tactics and somewhat less

likely (although not significantly so) to use maximization tactics;

both tactics are known to increase rates of false confessions

(Kassin et al., 2010). In addition, participant-suspects (who were

made aware of the video-recording manipulation) reported that the

police officers who did not know they were being recorded tried

harder to make them confess than did police officers who knew

about the recording. Thus, there is evidence, albeit from outside

the eyewitness arena, that video-recording police procedure can

deter undesirable behaviors.

There are few data on the question of whether video-recording

eyewitness identification procedures might change the behavior of

eyewitnesses. But there are good data indicating that even crime

suspects are not inhibited by video-recording during interrogations

and that people quickly forget that they are being recorded (Kassin

et al., 2019).

In sum, both logic and the available empirical evidence support

the recommendation to video-record all identification procedures.

If the procedure is to be recorded, it should be recorded in a way

that captures all relevant information about the procedure, includ-

ing the verbal and nonverbal behaviors of the witness, the admin-

istrator, and the lineup members. Research from the interrogation

literature makes it clear that camera angle matters, in that people

attribute causality to the person who is the focus of the video-

recording (Lassiter, 2010; Lassiter et al., 2009; Ratcliff et al.,

2006). In the case of interrogations, a camera focused on the

suspect causes evaluators to be more likely to view a confession as

voluntary and more likely to judge them to be guilty than if the

camera was focused on the interrogator or equally on the suspect

and the interrogator. It is reasonable to believe that a camera

focused on the witness may similarly cause evaluators to overlook

suggestive behaviors on the part of administrator, or features and

behaviors of suspects that make them stand out from the other

lineup members. However, a video-recording of an identification

procedure that includes information about the witness, the admin-

istrator, and lineup members can provide a complete record of the

procedure that documents suggestive practices when they are

present and protects the police from unjustified and time-

consuming claims of bias when the procedure was free from

suggestion.

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

WELLS ET AL.

Recommendation 8: Avoid Repeated Identification

Procedures With the Same Witness and Suspect

Repeating an identification procedure with the same suspect

and same eyewitness should be avoided regardless of whether the

eyewitness identified the suspect in the initial identification pro-

cedure.

This recommendation holds no matter how compelling the ar-

gument in favor of a second identification might seem (e.g., the

original photo of the suspect was not as good as it could have been;

the witness was nervous during the first identification test and is

calmer now; the initial identification was made from a social

media profile, but it would be more desirable to have an identifi-

cation made using proper police procedures). The importance of

focusing on the first identification test cannot be emphasized

strongly enough.

There are certain kinds of forensic evidence for which repeated

testing is not only acceptable but also desirable. A crime scene

fingerprint comparison with a suspect’s prints, for example, might

be subjected to repeated comparisons to confirm a conclusion.

Likewise, given a sufficient amount biological material from a

crime scene, forensic examiners might test only a small portion of

the sample for DNA, allowing for subsequent testing by a different

analyst or different means of evaluation. However, eyewitness

identification evidence has a unique characteristic that makes it

unsuitable for what might be called “repeated testing.” Whether

the eyewitness is asked to make an identification with a showup or

a lineup, there is only one uncontaminated opportunity for a given

eyewitness to make an identification of a particular suspect. Any

subsequent identification test with that same eyewitness and that

same suspect is contaminated by the eyewitness’s experience on

the initial test.

For purposes of our recommendation, repeated identification

tests refer to a situation in which an eyewitness is given a subse-

quent identification test (or more) with the same suspect that

appeared in an earlier identification test. We are not referring to a

situation in which an eyewitness is given an identification test with

one suspect and then, after rejecting that suspect, is given an

identification test with a different suspect and different fillers. Nor

are we referring to a situation in which there were multiple culprits

and the eyewitness is given a separate identification test for each

culprit.

There are many ways in which the use of repeated identification

procedures surface in criminal cases. For example, eyewitnesses

might view a mug book that contains the suspect prior to viewing

a lineup that includes that suspect, or an eyewitness might first

view a showup and then later be shown a lineup with that same

suspect. A repeated identification can also occur when the eyewit-

ness views a lineup and makes a tentative pick (or no pick) and

then, at a later time, sees a second lineup with that same suspect

and new fillers. Sometimes a witness is first shown a photo lineup

and then later is shown a live lineup using that same suspect. Other

times the first lineup uses one photo of the suspect, and the second

lineup contains a different photo of the same suspect. Perhaps the

most common repeated identification procedure of all is when the

witness makes an out-of-court identification (from a showup or a

lineup) and then is asked to repeat that identification in court (i.e.,

an in-court or “dock” identification) at pretrial hearings or at trial

or at both the pretrial hearing and at trial.

At a theoretical level, there are at least three processes by which

an initial identification test that includes a given suspect can

contaminate a later identification test if the later test includes that

same suspect (Deffenbacher, Bornstein, & Penrod, 2006; Steblay

& Dysart, 2016). One such process is memory-source error (or

“source monitoring error”; see Johnson, Hashtroudi, & Lindsay,

1993). Memory-source error is implicated when the eyewitness

perceives the suspect in the second identification procedure to be

familiar and misattributes the familiarity to the original witnessed

event rather than to the fact that the previous identification proce-

dure included that face. In this type of effect, even though the

eyewitness did not identify the suspect in the first identification

procedure, that person’s face was made familiar by its appearance

in the first procedure. This source misattribution effect, which

involves a dissociation between familiarity and an awareness of

the source of that familiarity, was first demonstrated over 40 years

ago (Brown, Deffenbacher, & Sturgill, 1977). It is closely related

to other phenomena such as familiarity without awareness (Man-

dler, 1980), the false fame effect (Jacoby, Woloshyn, & Kelley,

1989), and the “bystander effect,” in which a bystander to the

crime is identified as the culprit due to a misattribution of famil-

iarity (Ross, Ceci, Dunning, & Toglia, 1994). Simply being ex-

posed to an innocent suspect in a mug book, showup, or a lineup

increases the chances of that person being identified in a later

lineup even if the witness did not choose the person in the first

identification procedure (e.g., Brown et al., 1977; Dysart, Lindsay,

Hammond, & Dupuis, 2001; Haw, Dickinson, & Meissner, 2007;

Hinz & Pezdek, 2001; Lawson & Dysart, 2014; Steblay, Tix, &

Benson, 2013).

A second process by which the first identification procedure can

contaminate the later identification procedure is when the eyewit-

ness identifies the suspect in the initial identification procedure

and is later given another identification procedure with that same

suspect and a different set of fillers. In this case, the initial

identification, even if mistaken, causes the witness to simply

repeat the same identification in the second identification proce-

dure. A meta-analysis of this commitment effect has provided

strong evidence that a mistaken identification in an initial identi-

fication procedure tends to be repeated in a second identification

procedure if that lineup contains the mistakenly identified person

(Deffenbacher et al., 2006). More recent studies have provided

additional support for this effect (e.g., Godfrey & Clark, 2010;

Goodsell, Neuschatz, & Gronlund, 2009; Hinz & Pezdek, 2001;

Lawson & Dysart, 2014; Pezdek & Blandon-Gitlin, 2005; Steblay

et al., 2013; Valentine, Davis, Memon, & Roberts, 2012). Re-

searchers have argued that commitment, which is the powerful

tendency to stick with an earlier decision that was freely made, is

the psychological mechanism underlying this effect. But commit-

ment is not the only process by which an initial mistaken identi-

fication results in a repeat mistaken identification of the same

person in a subsequent identification test. For example, there is

evidence that the act of identifying an innocent person in an initial

identification procedure changes the eyewitness’s memory away

from the culprit and toward the person identified, a process that is

intensified if the witness is given confirming feedback following

the initial mistaken identification (Smalarz & Wells, 2014a).

Although not yet specifically tested in controlled experiments,

there is a third possible process by which repeating the same

suspect in a second identification procedure can contaminate the

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

EYEWITNESS IDENTIFICATION RECOMMENDATIONS

second identification decision. Specifically, it is possible for the

eyewitness who makes no identification in the first procedure (e.g.,

a photo lineup) to later consciously and explicitly recognize that

there is only one person in the second procedure (e.g., a live

lineup) who was also in the first procedure. In effect, this “gives

away” the hypothesis of police investigators by communicating to

the eyewitness which person is the suspect (i.e., the person in

common between the two procedures). In other words, this type of

repeated procedure violates a fundamental characteristic of a good

lineup, namely that there not be aspects of the procedure that leak

information about which person is the suspect in the lineup (Wells

& Luus, 1990).

Any of these processes might occur in a given situation that

involves repeating the suspect in an identification procedure. For

this reason, we recommend that repeated identification procedures

be avoided. Of course, it could be argued that if the witness

identifies the suspect in the first identification procedure then there

is no harm in repeating the identification procedure. After all, the

identification has already been made and the second identification

procedure is merely a confirmatory process. However, there is

good evidence that repeated testing of eyewitnesses leads to arti-

ficially elevated levels of eyewitness confidence (Shaw, Garven, &

Wood, 1997; Shaw & McClure, 1996). Repeatedly asking a ques-

tion appears to lead to increased ease or fluency of answering the

question, which leads witnesses to develop a stronger sense of

confidence that their answer is correct, even when their answer is

incorrect (Shaw, 1996).

Finally, it is important to note that when witnesses make iden-

tifications outside of a police procedure, any additional identifica-

tion procedure conducted by the police is a second identification

attempt. Witnesses may spontaneously identify someone as the

culprit as they walk about their daily lives. Sometimes witnesses to

crimes launch their own investigations. They may hear the culprit

referred to by a nickname or street name and then search social

media for someone using that name. Or they may search the social

media connections of someone they knew was present during a

crime looking for the culprit. Whether an identification is made

through a self-directed search of social media or spontaneously

spotting a suspect on the street, this identification is the first

identification, and it will contaminate any subsequent identifica-

tion procedure the police might present to the witness.

Recommendation 9: Showups

Showups should be avoided whenever it is possible to conduct a

lineup (e.g., if probable cause exists to arrest the person then a

showup should not be conducted.). Cases in which it is necessary

to conduct a showup should use the procedural safeguards that are

recommended for lineups, including the elimination of suggestive

cues, a warning that the detained person might not be the culprit,

video-recording the procedure, and securing a confidence state-

ment.

The term showup refers to an eyewitness identification proce-

dure in which a single individual is presented to the eyewitness and

the eyewitness is asked whether this is the person who committed

the crime in question. The primary defining feature that separates

a showup from a lineup is the use of fillers: A showup has no

fillers, whereas a lineup does. Showups have been heavily criti-

cized as being extremely suggestive, a criticism that dates back

more than 100 years (Gross, 1911). In 1967, the U.S. Supreme

Court, in reference to a showup identification, said that “It is hard

to imagine a situation more clearly conveying the suggestion to the

witness that the one presented is believed to be guilty by the

police” (United States v. Wade, 1967, p. 226). Despite the strong

language seemingly condemning showups, the U.S. Supreme

Court has consistently supported the admissibility of showups

(e.g., see Manson v. Braithwaite, 1977; Neil v. Biggers, 1972)

provided that the witness shows evidence of reliability in other

ways. The Court listed five factors to consider that might indicate

that a suggestive procedure could nevertheless be reliable, namely

that (a) the witness had a good opportunity to view the culprit, (b)

the witness paid good attention while witnessing, (c) the witness

gave a good description of the culprit, (d) there was a short delay

between the witnessed event and the identification, and (e) the

witness made the identification with high confidence. These cri-

teria have been strongly criticized by eyewitness scientists on a

variety of counts, especially when the procedure was suggestive

(Wells & Quinlivan, 2009).

In practice, showups tend to be conducted under specific con-

ditions. In particular, showups tend to be restricted to situations in

which an individual who fits the description of the culprit of a

crime is detained in the general vicinity of the crime shortly after

the crime has occurred. As discussed in the introduction to this

article, there is a rationale for sometimes permitting showups

under these conditions (fit description, proximity to crime, and

soon after the crime occurred). The rationale is that there is not

enough time to construct and conduct a lineup procedure because

the detained person can only be detained for a relatively short

period of time unless there are grounds for arrest. Fitting a de-

scription of the culprit is not, in and of itself, grounds for arrest.

So, unless there is probable cause for arresting the detained person,

the choice is not between conducting a lineup and conducting a

showup; rather, the choice is between conducting a showup or

setting the detained person free, thereby potentially creating a

public safety issue. Moreover, because showups sometimes result

in eyewitnesses rejecting the detained person as being the culprit

(see Gonzalez, Ellsworth, & Pembroke, 1993), showups can result

in innocent people being quickly excluded as possible suspects,

thereby allowing a continuation of the search for the true offender.

About 15% of DNA exonerations involving eyewitness identi-

fication involved showups (West & Meterko, 2017). Experiments

comparing lineups with showups reveal that lineups are clearly

superior to showups in terms of the lineup procedure’s ability to

distinguish between innocent and guilty suspects (e.g., Clark,

2012a; Dekle, Beal, Elliot, & Huneycutt, 1996; Gronlund et al.,

2012; Mickes, 2015; Steblay, Dysart, Fulero, & Lindsay, 2003;

Wetmore et al., 2015; Yarmey, Yarmey, & Yarmey, 1996; see

meta-analysis by Neuschatz et al., 2016). Interestingly, however,

witnesses are no more likely to make an affirmative identification

with a showup than with a lineup (Gonzalez et al., 1993; Wells,

2001). In fact, witnesses actually make more affirmative identifi-

cation responses to lineups than to showups, perhaps because there

are more people from whom to choose. However, inaccurate

identifications from showups always fall on the innocent suspect,

whereas in lineups such inaccurate choices tend to spread across

known-innocent fillers (Smith et al., 2017; Wells, 2001). In several

ways, the fillers used in a lineup act as a safeguard, protecting the

innocent suspect from mistaken identification—a form of protec-

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

WELLS ET AL.

tion that does not exist for showups (see Wixted & Mickes, 2014,

for a different account of why lineups work better than showups).

Our recommendation is that showups should be avoided (and a

lineup conducted instead), if at all possible. For example, a showup

can be avoided and a lineup used instead if there is probable cause

for arresting a detained person instead of doing a showup. If, for

example, the detained person is in possession of stolen goods, is in

possession of an unauthorized firearm or other weapon, or has an

outstanding warrant, then the person should be arrested. Indeed,

without evidence linking the detained person to the crime (other

than being in the vicinity of where it was committed, which—as

noted previously—is not evidence-based suspicion), officers are

using the most suggestive identification procedure (a showup)

when base rates of guilt are the lowest. A lineup for the arrested

individual can then be arranged rather than a showup. Another

situation in which showups can sometimes be avoided occurs

when there are multiple witnesses. In cases of multiple witnesses,

a showup can be conducted with one of the witnesses and, if an

identification is made, the identification becomes grounds for

arrest. The remaining witnesses can then be preserved for more

reliable lineup procedures, which should be conducted only after

evidence is developed connecting the suspect to the specific wit-

nessed crime.

Reducing the Suggestiveness of Showups

Although showups are clearly less reliable than lineups, there

are some ways to reduce the suggestiveness of showups. First,

many of the features of a good lineup procedure can be incorpo-

rated into showup procedures. For example, recent research finds

that a preshowup instruction about additional opportunities can be

effective in reducing mistaken identifications with little or no

reduction in accurate identifications (Eisen, Smith, Olaguez, &

Skerritt-Perta, 2017; Smith, Wells, Lindsay, & Myerson, 2018).

This additional-opportunities instruction simply tells witnesses

prior to viewing the detained person that if they do not think the

person is the culprit, they might have additional opportunities to

view someone else. The theory behind the additional-opportunities

instruction is that witnesses set a low criterion for making an

identification with showups because they assume that this is their

only opportunity to identify the culprit.

As with lineups, eyewitnesses who participate in a showup

procedure should be asked to report how confident they are in their

identification or rejection decision. In addition, as with lineups,

showups should be video-recorded, a recommendation that is

relatively easy to implement now that dashboard-mounted and

body-worn cameras are increasingly commonplace (see Recom-

mendation 7). The video of the showup should start before the

witness is instructed and continue through the witness’s confi-

dence statement, and should include the officer, witness, and

suspect in view.

Because showups are conducted in the field during which

search-and-detain operations are actively unfolding, care should be

taken to ensure that witnesses do not overhear police radio con-

versations that could prejudice a showup procedure. In an exper-

iment using a high-realism paradigm for studying showup identi-

fications, overhearing the suggestion that the sheriff had caught the

guy significantly increased false identifications from showups but

did not affect accurate culprit identifications (Eisen, Skerrit-Perta,

Jones, Owen, & Cedre, 2017).

Clothing is often a central feature of an eyewitness’s description

that police use for finding a person who is then detained for a

showup. The concern is that the eyewitness might identify the

person based primarily on the clothing (see Lindsay et al., 1987).

In cases in which the person was detained based on a clothing

description, consideration should be given to covering the person’s

clothing with a blanket for the showup identification test.

For legal reasons, eyewitnesses are usually brought to the loca-

tion of the detained person for a showup rather than the detained

person being brought to the eyewitness, because transporting the

detained person is usually considered an arrest. Consistent with our

recommendations, if the detained person can be arrested, then a

lineup should be conducted rather than a showup. Taking the

eyewitness to the detained individual rather than the reverse also

makes it easier in multiple-witness cases to ensure that the wit-

nesses are not contaminated by observing an identification deci-

sion made by another witness. As noted previously, if one witness

makes an identification, then that would normally be considered

probable cause for arrest and the remaining witnesses can later be

shown a lineup instead of a showup. Placing the detained person in

handcuffs or in the back seat of a patrol car for the showup can

suggest to the witness that the person has been arrested. This

indication that the person has been arrested, in turn, can suggest to

the eyewitness that there is evidence against the detained person

beyond simply matching the description of the suspect. Hence,

unless there is reason to believe that the detained person is a flight

risk, these types of restraints should be avoided during a showup.

Severe limits on safeguards with showups. Obviously, there

are some critical safeguards that can be used with lineups that

cannot be used with showups. By definition, showups do not

include fillers, which is a key safeguard. But showups also cannot

be double-blind. In fact, showups cannot ever be single blind.

After all, both the showup administrator and the eyewitness know

which person is the suspect, namely the person being presented to

the eyewitness.

In-court identifications as showups. The current article con-

cerns policies and procedures for collecting and preserving eye-

witness identification evidence rather than how courts handle

eyewitness evidence. Nevertheless, there is a common courtroom

procedure known as an in-court identification (or a dock identifi-

cation) in which an eyewitness on the stand is asked if she or he

can identify the culprit in the courtroom. The defendant typically

is sitting at the defense counsel table rather than embedded among

known-innocent fillers who fit the description of the culprit.

Hence, an in-court identification is closely analogous to a showup.

In terms of suggestiveness, the in-court identification is arguably

even more suggestive than a typical showup because it clear to the

witness that the defendant has already been indicted. Moreover, it

is usually the case that the eyewitness has already identified the

defendant in a precourt procedure, which means that the in-court

identification is a repeated identification that goes directly against

Recommendation 8 (e.g., avoid repeated identifications). And, if

the eyewitness has not already identified the defendant in a proper

precourt procedure, the showup nature of an in-court identification

should not be considered an acceptable alternative to a properly

constructed and conducted lineup. The low probative value of an

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

EYEWITNESS IDENTIFICATION RECOMMENDATIONS

in-court identification raises serious concerns that its prejudicial

value exceeds it probative value.

Final Observations

The current set of recommendations for the collection and

preservation of eyewitness identification evidence is being pro-

posed under a much different set of conditions than existed when

the 1998 scientific review paper recommendations were published.

Prior to 1998, there was no official set of recommendations from

a scientific body on eyewitness identification nor were there any

guidelines from the legal system that were grounded in science.

Since the 1998 scientific review paper, however, numerous

agencies, governing bodies, and organizations have proposed

guidelines for collecting and preserving eyewitness identification

evidence. These include the U.S. Department of Justice Guide

(Technical Working Group on Eyewitness Evidence, 1999; rein-

forced recently by a directive from the United States Department

of Justice, 2017), which followed closely on the 1998 scientific

review paper as discussed in the introduction to the current article.

Furthermore, a large set of U.S. state-wide reform guidelines have

been enacted that range from the recommendations of state justice

departments (e.g., Wisconsin), to executive orders from their At-

torneys General (e.g., New Jersey), to court mandates (e.g., Ore-

gon), to laws passed by state legislatures (e.g., North Carolina). In

fact, in the United States there are now 21 states that require

reforms through one or more of these mechanisms, namely Cali-

fornia, Colorado, Connecticut, Florida, Georgia, Kansas, Illinois,

Louisiana, Maryland, New Hampshire, Nebraska, Nevada, New

Jersey, New York, North Carolina, Ohio, Oregon, Texas, Vermont,

West Virginia, and Wisconsin. In other states, such as Montana,

Hawaii, Minnesota, Pennsylvania, Iowa, Maine, Michigan, Wash-

ington, Idaho, and Rhode Island, police departments in major

metropolitan areas have revamped their procedures. In effect, the

reform of eyewitness identification procedures has been a chain-

reaction going back to the 1998 scientific review paper, the De-

partment of Justice guide in 1999, and the first state to adopt

reforms (New Jersey in 2002). Each jurisdiction has modeled its

reforms around those that preceded it and, hence, include the core

four recommendations that were presented in the original 1998

scientific review paper (i.e., prelineup instructions, how to select

fillers for the lineup, the need to collect a measure of witness

confidence at the time of the identification, and the use of a

double-blind procedure).

An important document relating to eyewitness identification

procedures that was generated in recent years deserves special

notice. The International Association of Chiefs of Police (IACP)

produced a position article in 2016, in which they made specific

recommendations regarding how lineups and showups should be

conducted (International Association of Chiefs of Police, 2016).

The IACP policy document goes beyond what most state reforms

have done and goes beyond the recommendations in the 1998

scientific review paper. Specifically, the IACP recommendations

not only include the four core recommendations, but also, like the

current article, recommend avoiding multiple identifications pro-

cedures using the same witness and suspect, recommend video-

recording the identification procedure, and make recommendations

about showups that are similar to the recommendations that we

offer in the current article. There are two recommendations in the

current scientific review paper, however, that are not in the IACP

policy, namely Recommendation 1 (Prelineup interview of the

eyewitness) and Recommendation 2 (Evidence-based suspicion).

We note that the IACP policy document, like the policy docu-

ments that have been adopted in various states and localities, is

almost exclusively a prescriptive document. Short, prescriptive

documents can make sense for policy documents. The IACP pol-

icy, for example, is a mere four pages covering lineups and

showups. Similarly, eyewitness identification policies and proce-

dures documents adopted in various states and agencies are largely

devoid of sustained rationale and explanations. In other words,

these policy documents specify certain procedures, but they do not

explain the rationale behind the procedures or review relevant

science pertaining to the prescribed procedures. Hence, the current

scientific review paper, like the 1998 scientific review paper,

serves an important function that other policy documents do not,

namely the function of documenting relevant science, where avail-

able, and an articulating rationale around each of the recommen-

dations.

Earlier in this article we briefly alluded to an observation that

deserves to be reemphasized. Although our nine recommendations

are relatively specific and detailed, the most important aspects of

our recommendations are the underlying principles associated with

those recommendations. The double-blind administration of line-

ups, for example, is based on the principle that procedures must be

in place to prevent the lineup administrator’s knowledge of which

person is the suspect to influence the eyewitness. Our specific

preference for double-blind administration, however, is not the

only way to achieve this prevention. With photo lineups, for

instance, a properly programmed computer could administer the

lineup without any administrator in the room at all. The underlying

principle is important for another reason as well, namely that the

specific recommendation might be technically followed and yet

the principle itself is violated at some other point or level. Con-

sider, for example, the invitation to view a lineup recommenda-

tion: “When inviting an eyewitness to attend a lineup procedure,

police should not suggest that a suspect has been arrested or that

the culprit will be present in the identification procedure.” The

underlying principle concerns the need to avoid communications

to the eyewitness that could undermine the later prelineup instruc-

tions emphasizing that the culprit might not be in the lineup.

Although the invitation to view a lineup might be the most likely

time at which this principle is violated, it is not the only possible

time. In this sense, the underlying principle is more important than

the specific wording of the recommendation.

Finally, we think it is important for law enforcement to under-

stand why certain procedures are recommended for constructing

and conducting eyewitness identification procedures rather than

simply being instructed on how to conduct those procedures. The

current article provides valuable background on how research in

areas such as perception, memory, decision making, and social

influence can inform recommendations on lineup construction and

presentation. Thus, in line with the National Research Council

(2014) report on eyewitness evidence that highlights the impor-

tance of training for law enforcement, we believe that this article

could serve as a useful resource for developing and implementing

such training.

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

WELLS ET AL.

References

Ackil, J. K., & Zaragoza, M. S. (1998). Memorial consequences of forced

confabulation: Age differences in susceptibility to false memories. De-

velopmental Psychology, 34, 1358–1372. http://dx.doi.org/10.1037/

0012-1649.34.6.1358

Beaudry, J. L., Lindsay, R. C. L., Leach, A.-M., Mansour, J. K., Bertrand,

M. I., & Kalmet, N. (2015). The effect of evidence type, identification

accuracy, line-up presentation, and line-up administration on observers’

perceptions of eyewitnesses. Legal and Criminological Psychology, 20,

343–364. http://dx.doi.org/10.1111/lcrp.12030

Behrman, B. W., & Davey, S. L. (2001). Eyewitness identification in actual

criminal cases: An archival analysis. Law and Human Behavior, 25,

475–491. http://dx.doi.org/10.1023/A:1012840831846

Behrman, B. W., & Richards, R. E. (2005). Suspect/foil identification in

actual crimes and in the laboratory: A reality monitoring analysis. Law

and Human Behavior, 29, 279–301. http://dx.doi.org/10.1007/s10979-

005-3617-y

Bergold, A. N., & Heaton, P. (2018). Does filler database size influence

identification accuracy? Law and Human Behavior, 42, 227–243. http://

dx.doi.org/10.1037/lhb0000289

Berkowitz, S. R., & Frenda, S. J. (2018). Rethinking the confident eyewitness:

A reply to Wixted, Mickes, & Fisher. Perspectives on Psychological Sci-

ence, 13, 336–338. http://dx.doi.org/10.1177/1745691617751883

Berkowitz, S. R., & Loftus, E. F. (2018). Misinformation in the courtroom.

In H. Otgaar & M. L. Howe (Eds.), Finding the truth in the courtroom:

Dealing with deception, lies, and memories (pp. 11–30). New York, NY:

Oxford University Press.

Borchard, E. (1932). Convicting the innocent: Errors of criminal justice.

New Haven, CT: Yale University Press.

Bradfield, A., & McQuiston, D. E. (2004). When does evidence of eye-

witness confidence inflation affect judgments in a criminal trial? Law

and Human Behavior, 28, 369–387. http://dx.doi.org/10.1023/B:LAHU

.0000039331.54147.ff

Bradfield, A. L., Wells, G. L., & Olson, E. A. (2002). The damaging effect

of confirming feedback on the relation between eyewitness certainty and

identification accuracy. Journal of Applied Psychology, 87, 112–120.

http://dx.doi.org/10.1037/0021-9010.87.1.112

Brainerd, C. J., & Reyna, V. F. (2005). The science of false memory. New

York, NY: Oxford University Press. http://dx.doi.org/10.1093/acprof:

oso/9780195154054.001.0001

Brandon, R., & Davies, C. (1973). Wrongful imprisonment: Mistaken

convictions and their consequences. London, UK: George Allen and

Unwin.

Brewer, N., Caon, A., Todd, C., & Weber, N. (2006). Eyewitness identi-

fication accuracy and response latency. Law and Human Behavior, 30,

31–50. http://dx.doi.org/10.1007/s10979-006-9002-7

Brewer, N., Keast, A., & Rishworth, A. (2002). The confidence-accuracy

relationship in eyewitness identification: The effects of reflection and

disconfirmation on correlation and calibration. Journal of Experimental

Psychology: Applied, 8, 44–56. http://dx.doi.org/10.1037/1076-898X.8

.1.44

Brewer, N., Keast, A., & Sauer, J. D. (2010). Children’s eyewitness

identification performance: Effects of a Not Sure response option and

accuracy motivation. Legal and Criminological Psychology, 15, 261–

277. http://dx.doi.org/10.1348/135532509X474822

Brewer, N., Vagadia, A. N., Hope, L., & Gabbert, F. (2018). Interviewing

witnesses: Eliciting coarse-grain information. Law and Human Behav-

ior, 42, 458–471. http://dx.doi.org/10.1037/lhb0000294

Brewer, N., Weber, N., & Semmler, C. (2005). Eyewitness identification.

In N. Brewer & K. D. Williams (Eds.), Psychology and law: An

empirical perspective (pp. 177–221). New York, NY: Guilford Press.

Brewer, N., & Wells, G. L. (2006). The confidence-accuracy relationship

in eyewitness identification: Effects of lineup instructions, foil similar-

ity, and target-absent base rates. Journal of Experimental Psychology:

Applied,

12, 11–30. http://dx.doi.org/10.1037/1076-898X.12.1.11

Brown,

C., Lloyd-Jones, T. J., & Robinson, M. (2008). Eliciting person

descriptions from eyewitnesses: A survey of police perceptions of eye-

witness performance and reported use of interview techniques. The

European Journal of Cognitive Psychology, 20, 529–560. http://dx.doi

.org/10.1080/09541440701728474

Brown, E., Deffenbacher, K., & Sturgill, W. (1977). Memory for faces and

the circumstances of the encounter. Journal of Applied Psychology, 62,

311–318. http://dx.doi.org/10.1037/0021-9010.62.3.311

Carlson, C. A., Jones, A. R., Whittington, J. E., Lockamyeir, R. F., Carlson,

M. A., & Wooten, A. R. (2019). Lineup fairness: Propitious heteroge-

neity and the diagnostic feature-detection hypothesis. Cognitive Re-

search: Principles and Implications, 4, 20–26. http://dx.doi.org/10

.1186/s41235-019-0172-5

Ceci, S. J., & Bruck, M. (1995). Jeopardy in the courtroom: A scientific

analysis of children’s testimony. Washington, DC: American Psycho-

logical Association. http://dx.doi.org/10.1037/10180-000

Charman, S. D., Douglass, A. B., & Mook, A. (2019). Cognitive bias in

forensic decision making. In N. Brewer & A. B. Douglass (Eds.),

Psychological science and the law (pp. 30–53). New York, NY: Guil-

ford Press.

Charman, S. D., Matuku, K., & Mook, A. (2019). Non-blind lineup

administration biases administrator’s interpretation of ambiguous wit-

ness statements and their perceptions of the witness. Applied Cognitive

Psychology, 33, 1260–1270. http://dx.doi.org/10.1002/acp.3579

Charman, S. D., & Quiroz, V. (2016). Blind sequential lineup administra-

tion reduces both false identifications and confidence in those false

identifications. Law and Human Behavior, 40, 477–487. http://dx.doi

.org/10.1037/lhb0000197

Charman, S. D., & Wells, G. L. (2007). Eyewitness lineups: Is the

appearance-change instruction a good idea? Law and Human Behavior,

31, 3–22. http://dx.doi.org/10.1007/s10979-006-9006-3

Charman, S. D., Wells, G. L., & Joy, S. W. (2011). The dud effect: Adding

highly dissimilar fillers increases confidence in lineup identifications.

Law and Human Behavior, 35, 479–500. http://dx.doi.org/10.1007/

s10979-010-9261-1

Chase, W. G., & Simon, H. A. (1973). Perceptions in chess. Cognitive

Psychology, 4, 55–81. http://dx.doi.org/10.1016/0010-0285(73)90004-2

Clark, S. E. (2005). A re-examination of the effects of biased lineup

instructions in eyewitness identification. Law and Human Behavior, 29,

395–424. http://dx.doi.org/10.1007/s10979-005-5690-7

Clark, S. E. (2012a). Costs and benefits of eyewitness identification

reform: Psychological science and public policy. Perspectives on

Psychological Science, 7, 238–259. http://dx.doi.org/10.1177/

1745691612439584

Clark, S. E. (2012b). Eyewitness identification reform: Data, theory, and

due process. Perspectives on Psychological Science, 7, 279–283. http://

dx.doi.org/10.1177/1745691612444136

Clark, S. E., Brower, G. L., Rosenthal, R., Hicks, J. M., & Moreland, M. B.

(2013). Lineup administrator influences on eyewitness identification and

eyewitness confidence. Journal of Applied Research in Memory &

Cognition, 2, 158–165. http://dx.doi.org/10.1016/j.jarmac.2013.06.003

Clark, S. E., Howell, R. T., & Davey, S. L. (2008). Regularities in

eyewitness identification. Law and Human Behavior, 32, 187–218.

http://dx.doi.org/10.1007/s10979-006-9082-4

Clark, S. E., & Wells, G. L. (2008). On the diagnosticity of multiple-

witness identifications. Law and Human Behavior, 32, 406–422. http://

dx.doi.org/10.1007/s10979-007-9115-7

Clarke, C., Milne, R., & Bull, R. (2011). Interviewing suspects of crime:

The impact of PEACE training, supervision and the presence of a legal

advisor. Journal of Investigative Psychology and Offender Profiling, 8,

149–162. http://dx.doi.org/10.1002/jip.144

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

EYEWITNESS IDENTIFICATION RECOMMENDATIONS

Clifford, B. R., & George, R. (1996). A field evaluation of training in

three methods of witness/victim investigative interviewing. Psychol-

ogy, Crime & Law, 2, 231–248. http://dx.doi.org/10.1080/

10683169608409780

Colloff, M. F., Wade, K. A., & Strange, D. (2016). Unfair lineups make

eyewitnesses more likely to confuse innocent and guilty suspects. Psy-

chological Science, 27, 1227–1239. http://dx.doi.org/10.1177/

0956797616655789

Connors, E., Lundregan, T., Miller, N., & McEwan, T. (1996). Convicted

by juries, exonerated by science: Case studies in the use of DNA

evidence to establish innocence after trial. Alexandria, VA: National

Institute of Justice.

Corey, D., Malpass, R. S., & McQuiston, D. E. (1999). Parallelism in

eyewitness and mock witness identifications. Applied Cognitive Psy-

chology, 13, S41–S58.

Crombag, H. F., Wagenaar, W. A., & Van Koppen, P. J. (1996). Crashing

memories and the problem of ‘source monitoring’. Applied Cognitive

Psychology, 10, 95–104.

Cutler, B. L., & Penrod, S. D. (1995). Mistaken identification: The eye-

witness, psychology, and the law. New York, NY: Cambridge University

Press.

Dando, C. J., Geiselman, R. E., MacLeod, N., & Griffiths, A. (2015).

Interviewing adult witnesses and victims. In G. Oxburgh, T. Myklebust,

T. Grant, & R. Milne (Eds.), Communication in investigative and legal

contexts: Integrated approaches from forensic psychology, linguistics

and law enforcement (pp. 79–106). Hoboken, NJ: Wiley-Blackwell.

http://dx.doi.org/10.1002/9781118769133.ch5

Darling, S., Valentine, T., & Memon, A. (2008). Selection of lineup foils

in operational contexts. Applied Cognitive Psychology, 22, 159–168.

http://dx.doi.org/10.1002/acp.1366

Deffenbacher, K. A., Bornstein, B. H., McGorty, E. K., & Penrod, S. D.

(2008). Forgetting the once-seen face: Estimating the strength of an

eyewitness’s memory representation. Journal of Experimental Psychol-

ogy: Applied, 14, 139–150. http://dx.doi.org/10.1037/1076-898X.14.2

.139

Deffenbacher, K. A., Bornstein, B. H., & Penrod, S. D. (2006). Mugshot

exposure effects: Retroactive interference, source confusion, and uncon-

scious transference. Law and Human Behavior, 30, 287–307. http://dx

.doi.org/10.1007/s10979-006-9008-1

Dekle, D. J., Beal, C. R., Elliot, R., & Huneycutt, D. (1996). Children as

witnesses: A comparison of lineup versus showup identification meth-

ods. Applied Cognitive Psychology, 10, 1–12. http://dx.doi.org/10.1002/

(SICI)1099-0720(199602)10:1⬍1::AID-ACP354⬎3.0.CO;2-Y

Demarchi, S., & Py, J. (2009). A method to enhance person description: A

field study. In R. Bull, T. Valentine, & T. Williamson (Eds.), Handbook

of psychology of investigative interviewing: Current developments and

future directions (pp. 241–256). Chichester, UK: Wiley. http://dx.doi

.org/10.1002/9780470747599.ch14

Demarchi, S., Py, J., Groud-Than, S., Parain, T., & Brunel, M. (2013).

Describing a face without overshadowing effect: Another benefice of the

Person Description Interview. Psychologie Française, 58, 123–133.

http://dx.doi.org/10.1016/j.psfr.2013.01.002

Doob, A. N., & Kirshenbaum, H. M. (1973). Bias in police lineups—

Partial remembering. Journal of Police Science and Administration, 1,

287–293. http://dx.doi.org/10.1007/BF01039807

Douglass, A. B., Brewer, N., Semmler, C., Bustamante, L., & Hiley, A.

(2013). The dynamic interaction between eyewitnesses and interviewers:

The impact of differences in perspective on memory reports and inter-

viewer behavior. Law and Human Behavior, 37, 290–301. http://dx.doi

.org/10.1037/lhb0000034

Douglass, A. B., & Jones, E. E. (2013). Confidence inflation in eyewit-

nesses: Seeing is not believing. Legal and Criminological Psychology,

18, 152–167. http://dx.doi.org/10.1111/j.2044-8333.2011.02031.x

Douglass, A. B., Neuschatz, J. S., Imrich, J., & Wilkinson, M. (2010). Does

post-identification feedback affect evaluations of eyewitness testimony

and identification procedures? Law and Human Behavior, 34, 282–294.

http://dx.doi.org/10.1007/s10979-009-9189-5

Douglass,

B., & Smalarz, L. (2019). Post-identification feedback: State

of the science. In M. Miller & B. Bornstein (Eds.), Advances in psy-

chology and law (pp. 101–136). New York, NY: Springer. http://dx.doi

.org/10.1007/978-3-030-11042-0_4

Douglass, A. B., Smith, C., & Fraser-Thill, R. (2005). A problem with

double-blind photospread procedures: Photospread administrators use

one eyewitness’s confidence to influence the identification of another

eyewitness. Law and Human Behavior, 29, 543–562. http://dx.doi.org/

10.1007/s10979-005-6830-9

Dysart, J. E., Lawson, V. Z., & Rainey, A. (2012). Blind lineup adminis-

tration as a prophylactic against the postidentification feedback effect.

Law and Human Behavior, 36, 312–319. http://dx.doi.org/10.1037/

h0093921

Dysart, J. E., Lindsay, R. C. L., Hammond, R., & Dupuis, P. (2001). Mug

shot exposure prior to lineup identification: Interference, transference,

and commitment effects. Journal of Applied Psychology, 86, 1280–

1284. http://dx.doi.org/10.1037/0021-9010.86.6.1280

Eisen, M. L., Gabbert, F., Ying, R., & Williams, J. (2017). “I think he had

a tattoo on his neck”: How co-witness discussions about a perpetrator’s

description can affect eyewitness identification decisions. Journal of

Applied Research in Memory & Cognition, 6, 274–282. http://dx.doi

.org/10.1016/j.jarmac.2017.01.009

Eisen, M., Skerrit-Perta, A., Jones, J. M., Owen, J., & Cedre, G. C. (2017).

Pre-admonition suggestion in live showups: When witnesses learn that

the cops caught “the” guy. Applied Cognitive Psychology, 31, 520–529.

http://dx.doi.org/10.1002/acp.3349

Eisen, M., Smith, A. M., Olaguez, A. P., & Skerritt-Perta, A. S. (2017). An

examination of showups conducted by law enforcement using a field-

simulation paradigm. Psychology, Public Policy, and Law, 23, 1–22.

http://dx.doi.org/10.1037/law0000115

Ellis, H. D., Shepherd, J. W., & Davies, G. M. (1980). The deterioration of

verbal descriptions of faces over different delay intervals. Journal of

Police Science and Administration, 8, 101–106.

Fahsing, I. A., Ask, K., & Granhag, P. A. (2004). The man behind the

mask: Accuracy and predictors of eyewitness offender descriptions.

Journal of Applied Psychology, 89, 722–729. http://dx.doi.org/10.1037/

0021-9010.89.4.722

Fawcett, J. M., Russell, E. J., Peace, K. A., & Christie, J. (2011). Of guns

and geese: A meta-analytic review of the ‘weapon focus’ literature.

Psychology, Crime & Law, 19, 35–66. http://dx.doi.org/10.1080/

1068316X.2011.599325

Fisher, R. P., & Geiselman, R. E. (1992). Memory-enhancing techniques

for investigative interviewing: The cognitive interview. Springfield, IL:

Charles C Thomas.

Fisher, R. P., Schreiber Compo, N., Rivard, J., & Hirn, D. (2014). Inter-

viewing witnesses. In T. J. Perfect & D. S. Lindsay’s (Eds.), The SAGE

handbook of applied memory (pp. 559–578). Los Angeles, CA: SAGE

Publications. http://dx.doi.org/10.4135/9781446294703.n31

Fitzgerald, R. J., Price, H. L., Oriet, C., & Charman, S. D. (2013). The

effect of suspect-filler similarity on eyewitness identification decisions:

A meta-analysis. Psychology, Public Policy, and Law, 19, 151–164.

http://dx.doi.org/10.1037/a0030618

Fitzgerald, R. J., Price, H. L., & Valentine, T. (2018). Eyewitness identi-

fication: Live, photo, and video lineups. Psychology, Public Policy, and

Law, 24, 307–325. http://dx.doi.org/10.1037/law0000164

Flin, R. H., & Shepherd, J. W. (1986). Tall stories: Eyewitnesses’ ability to

estimate height and weight characteristics. Human Learning: Journal of

Practical Research & Applications, 5, 29–38.

Flowe, H. D., Carline, A., & Karoglu, N. (2018). Testing the reflection

assumption: A comparison of eyewitness ecology in the laboratory and

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

WELLS ET AL.

criminal cases. The International Journal of Evidence & Proof, 22,

239–261. http://dx.doi.org/10.1177/1365712718782996

Flowe, H. D., Takarangi, M. K. T., Humphries, J. E., & Wright, D. S.

(2016). Alcohol and remembering a hypothetical sexual assault: Can

people who were under the influence of alcohol during the event provide

accurate testimony? Memory, 24, 1042–1061. http://dx.doi.org/10.1080/

09658211.2015.1064536

Frank, J., & Frank, B. (1957). Not guilty. London, UK: Gallanez.

Gabbert, F., & Brown, C. (2015). Interviewing for face identification. In T.

Valentine & J. P. Davis (Eds.), Forensic facial identification: Theory

and practice of identification from eyewitnesses, composites and CCTV

(pp. 15–41). Chichester, UK: Wiley.

Gabbert, F., & Hope, L. (2013). Suggestibility and memory conformity. In

A. M. Ridley, F. Gabbert, & D. J. La Rooy (Eds.), Suggestibility in legal

contexts: Psychological research and forensic implications (pp. 63–84).

London, UK: Wiley-Blackwell.

Gabbert, F., Hope, L., & Fisher, R. P. (2009). Protecting eyewitness

evidence: Examining the efficacy of a self-administered interview tool.

Law and Human Behavior, 33, 298–307. http://dx.doi.org/10.1007/

s10979-008-9146-8

Gabbert, F., Memon, A., & Allan, K. (2003). Memory conformity: Can

eyewitnesses influence each other’s memories for an event? Applied

Cognitive Psychology, 17, 533–543. http://dx.doi.org/10.1002/acp.885

Gabbert, F., Memon, A., Allan, K., & Wright, D. B. (2004). Say it to my

face: Examining the effects of socially encountered misinformation.

Legal and Criminological Psychology, 9, 215–227. http://dx.doi.org/10

.1348/1355325041719428

Garrett, B. (2011). Convicting the innocent: Where criminal prosecutions

go wrong. Cambridge, MA: Harvard University Press. http://dx.doi.org/

10.4159/harvard.9780674060982

Garrioch, L., & Brimacombe, C. A. E. (2001). Lineup administrators’

expectations: Their impact on eyewitness confidence. Law and Human

Behavior, 25, 299–315. http://dx.doi.org/10.1023/A:1010750028643

Gobet, F., & Simon, H. A. (1996). Templates in chess memory: A mech-

anism for recalling several boards. Cognitive Psychology, 31, 1–40.

http://dx.doi.org/10.1006/cogp.1996.0011

Godfrey, R. D., & Clark, S. E. (2010). Repeated eyewitness identification

procedures: Memory, decision making, and probative value. Law and

Human Behavior, 34, 241–258. http://dx.doi.org/10.1007/s10979-009-

9187-7

Gonzalez, R., Ellsworth, P. C., & Pembroke, M. (1993). Misidentifications

and failures to identify in lineups and showups. Journal of Personality

and Social Psychology, 64, 525–537. http://dx.doi.org/10.1037/0022-

3514.64.4.525

Goodsell, C. A., Neuschatz, J. S., & Gronlund, S. D. (2009). Effects of

mugshot commitment on lineup performance in young and older adults.

Applied Cognitive Psychology, 23, 788–803. http://dx.doi.org/10.1002/

acp.1512

Granhag, P. A., Ask, K., & MacGiolla, E. (2013). Eyewitness recall: An

overview of estimator-based research. In T. J. Perfect & D. S. Lindsay

(Eds.), SAGE handbook of applied memory (pp. 35–55). Los Angeles,

CA: SAGE Publications.

Granhag, P. A., Ask, K., Rebelius, A., Öhman, L., & MacGiolla, E. (2013).

‘I saw the man who killed Anna Lindh!’ An archival study of witnesses’

offender descriptions. Psychology, Crime & Law, 19, 921–931. http://

dx.doi.org/10.1080/1068316X.2012.719620

Greathouse, S. M., & Kovera, M. B. (2009). Instruction bias and lineup

presentation moderate the effects of administrator knowledge on eye-

witness identification. Law and Human Behavior, 33, 70–82. http://dx

.doi.org/10.1007/s10979-008-9136-x

Greenberg, M. S., Westcott, D. R., & Bailey, S. E. (1998). When believing

is seeing: The effect of scripts on eyewitness memory. Law and Human

Behavior, 22, 685–694. http://dx.doi.org/10.1023/A:1025758807624

Gronlund, S. D., Carlson, C. A., Neuschatz, J. S., Goodsell, C. A., Wet-

more, S., Wooten, A., & Graham, M. (2012). Showups versus lineups:

An evaluation using ROC analysis. Journal of Applied Research in

Memory & Cognition, 1, 221–228. http://dx.doi.org/10.1016/j.jarmac

.2012.09.003

Gross,

(1911). Criminal psychology: A manual for judges, practitioners,

and students. Boston, MA: Little, Brown.

Harris, M. J., & Rosenthal, R. (1985). Mediation of interpersonal expec-

tancy effects: 31 meta-analyses. Psychological Bulletin, 97, 363–386.

http://dx.doi.org/10.1037/0033-2909.97.3.363

Haw, R. M., Dickinson, J. J., & Meissner, C. A. (2007). The phenome-

nology of carryover effects between show-up and line-up identification.

Memory, 15, 117–127. http://dx.doi.org/10.1080/09658210601171672

Haw, R. M., & Fisher, R. P. (2004). Effects of administrator-witness

contact on eyewitness identification accuracy. Journal of Applied Psy-

chology, 89, 1106–1112. http://dx.doi.org/10.1037/0021-9010.89.6

.1106

Hinz, T., & Pezdek, K. (2001). The effect of exposure to multiple lineups

on face identification accuracy. Law and Human Behavior, 25, 185–198.

http://dx.doi.org/10.1023/A:1005697431830

Hope, L., Gabbert, F., & Fisher, R. P. (2011). From laboratory to the street:

Capturing witness memory using the Self-Administered Interview. Le-

gal and Criminological Psychology, 16, 211–226. http://dx.doi.org/10

.1111/j.2044-8333.2011.02015.x

Hope, L., Gabbert, F., Fisher, R. P., & Jamieson, K. (2014). Protecting and

enhancing eyewitness memory: The impact of an initial recall attempt on

performance in an investigative interview. Applied Cognitive Psychol-

ogy, 28, 304–313. http://dx.doi.org/10.1002/acp.2984

Horry, R., Halford, P., Brewer, N., Milne, R., & Bull, R. (2014). Archival

analyses of eyewitness identification test outcomes: What can they tell

us about eyewitness memory? Law and Human Behavior, 38, 94–108.

http://dx.doi.org/10.1037/h0093959

Horry, R., Memon, A., Wright, D. B., & Milne, R. (2012). Predictors of

eyewitness identification decisions from video lineups in England: A

field study. Law and Human Behavior, 36, 257–265. http://dx.doi.org/

10.1037/h0093959

Huff, R., Rattner, A., & Sagarin, E. (1986). Guilty until proven innocent.

Crime and Delinquency, 32, 518–544. http://dx.doi.org/10.1177/

0011128786032004007

Innocence Project (2019). Retrieved from https://www.innocenceproject

.org/all-cases/

International Association of Chiefs of Police. (2016). Model policy on

eyewitness identification. Retrieved from https://www.theiacp.org/

resources/policy-center-resource/eyewitness-identification

Jacoby, L. L., Woloshyn, V., & Kelley, C. (1989). Becoming famous

without being recognized: Unconscious influences of memory produced

by dividing attention. Journal of Experimental Psychology: General,

118, 115–125. http://dx.doi.org/10.1037/0096-3445.118.2.115

Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitor-

ing. Psychological Bulletin, 114, 3–28. http://dx.doi.org/10.1037/0033-

2909.114.1.3

Jones, E. E., Williams, K. D., & Brewer, N. (2008). “I had a confidence

epiphany!”: Obstacles to combating post-identification confidence infla-

tion. Law and Human Behavior, 32, 164–176. http://dx.doi.org/10.1007/

s10979-007-9101-0

Juslin, P., Olsson, N., & Winman, A. (1996). Calibration and diagnosticity

of confidence in eyewitness identification: Comments on what can be

inferred from the low confidence-accuracy correlation. Journal of Ex-

perimental Psychology: Learning, Memory, and Cognition, 22, 1304–

1316. http://dx.doi.org/10.1037/0278-7393.22.5.1304

Kahneman, D., & Tversky, A. (1973). On the psychology of prediction.

Psychological Review, 80, 237–251. http://dx.doi.org/10.1037/h0034747

Kaminski, K. S., & Sporer, S. L. (2017). Discriminating between correct

and incorrect eyewitness identifications: The use of appropriate cues.

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

EYEWITNESS IDENTIFICATION RECOMMENDATIONS

Journal of Experimental Psychology: Applied, 23, 59–70. http://dx.doi

.org/10.1037/xap0000110

Kane, M. J., & Engle, R. W. (2000). Working-memory capacity, proactive

interference, and divided attention: Limits on long-term memory re-

trieval. Journal of Experimental Psychology: Learning, Memory, and

Cognition, 26, 336–358. http://dx.doi.org/10.1037/0278-7393.26.2.336

Kask, K., & Bull, R. (2009). From person descriptions to interviewing

methods: What can be done to improve child witness testimonies?

Trames, 13, 95–108. http://dx.doi.org/10.3176/tr.2009.2.01

Kassin, S. M. (1998). Eyewitness identification procedures: The fifth rule.

Law and Human Behavior, 22, 649–653. http://dx.doi.org/10.1023/A:

1025702722645

Kassin, S. M., Drizin, S. A., Grisso, T., Gudjonsson, G. H., Leo, R. A., &

Redlich, A. D. (2010). Police-induced confessions: Risk factors and

recommendations. Law and Human Behavior, 34, 3–38. http://dx.doi

.org/10.1007/s10979-009-9188-6

Kassin, S. M., Dror, I. E., & Kukucka, J. (2013). The forensic confirmation

bias: Problems, perspectives, and proposed solutions. Journal of Applied

Research in Memory & Cognition, 2, 42–52. http://dx.doi.org/10.1016/

j.jarmac.2013.01.001

Kassin, S. M., Kukucka, J., Lawson, V. Z., & DeCarlo, J. (2014). Does

video recording alter the behavior of police during interrogation? A

mock crime-and-investigation study. Law and Human Behavior, 38,

73–83. http://dx.doi.org/10.1037/lhb0000047

Kassin, S. M., Kukucka, J., Lawson, V. Z., & DeCarlo, J. (2017). Police

reports of mock suspect interrogations: A test of accuracy and percep-

tion. Law and Human Behavior, 41, 230–243. http://dx.doi.org/10.1037/

lhb0000225

Kassin, S. M., Russano, M. B., Amrom, A. D., Hellgren, J., Kukucka, J.,

& Lawson, V. Z. (2019). Does video recording inhibit crime suspects?

Evidence from a fully randomized field experiment. Law and Human

Behavior, 43, 45–55. http://dx.doi.org/10.1037/lhb0000319

Keast, A., Brewer, N., & Wells, G. L. (2007). Children’s metacognitive

judgments in an eyewitness identification task. Journal of Experimental

Child Psychology, 97, 286–314. http://dx.doi.org/10.1016/j.jecp.2007

.01.007

Kebbell, M. R., & Milne, R. (1998). Police officers’ perceptions of

eyewitness performance in forensic investigations. The Journal of

Social Psychology, 138, 323–330. http://dx.doi.org/10.1080/

00224549809600384

Klayman, J., & Ha, Y. (1987). Confirmation, disconfirmation, and infor-

mation in hypothesis testing. Psychological Review, 94, 211–228. http://

dx.doi.org/10.1037/0033-295X.94.2.211

Kleider, H. M., Pezdek, K., Goldinger, S. D., & Kirk, A. (2008). Schema-

driven source misattribution errors: Remembering the expected from a

witnessed event. Applied Cognitive Psychology, 22, 1–20. http://dx.doi

.org/10.1002/acp.1361

Klobuchar, A., Steblay, N. M., & Caligiuri, H. L. (2006). Improving

eyewitness identifications: Hennepin County’s blind sequential lineup

pilot project. Cardozo Public Law, Policy, and Ethics Journal, 2, 381–

414.

Kocab, K., & Sporer, S. L. (2016). The weapon focus effect for person

identifications and descriptions: A meta-analysis. In M. Miller & B. H.

Bornstein (Eds.), Advances in psychology and law (Vol. 1, pp. 71–117).

New York, NY: Springer. http://dx.doi.org/10.1007/978-3-319-29406-

3_3

Kovera, M. B., & Evelo, A. J. (2017). The case for double-blind lineup

administration. Psychology, Public Policy, and Law, 23, 421–437.

http://dx.doi.org/10.1037/law0000139

Kovera, M. B., Penrod, S. D., Pappas, C., & Thill, D. L. (1997). Identifi-

cation of computer-generated facial composites. Journal of Applied

Psychology, 82, 235–246. http://dx.doi.org/10.1037/0021-9010.82.2.235

Kraus, U., Zeier, F., Wagner, W., Paelecke, M., & Hewig, J. S. (2017).

Comparing the quality of memory reports in different initial eyewitness

questioning approaches.

Cogent Psychology, 4, 143–163. http://dx.doi

.org/10.1080/23311908.2017.1403063

Lamb,

M. E., Orbach, Y., Sternberg, K. J., Hershkowitz, I., & Horowitz, D.

(2000). Accuracy of investigators’ verbatim notes of their forensic

interviews with alleged child abuse victims. Law and Human Behavior,

24, 699–708. http://dx.doi.org/10.1023/A:1005556404636

LaRooy, D., Brubacher, S. P., Aromäki-Stratos, A., Cyr, M., Hershkowitz,

I., Korkman, J., & Stewart, H. (2015). The NICHD protocol: A review

of an internationally-used evidence-based tool for training child forensic

interviewers. Journal of Criminological Research. Policy & Practice, 1,

76–89.

Lassiter, G. D. (2010). Psychological science and sound public policy:

Video recording of custodial interrogations. American Psychologist, 65,

768–779. http://dx.doi.org/10.1037/0003-066X.65.8.768

Lassiter, G. D., Ware, L. J., Ratcliff, J. J., & Irvin, C. R. (2009). Evidence

of the camera perspective bias in authentic videotaped interrogations:

Implications for emerging reform in the criminal justice system. Legal

and Criminological Psychology, 14, 157–170. http://dx.doi.org/10.1348/

135532508X284293

Launay, C., & Py, J. (2015). Methods and aims of investigative interview-

ing of adult witnesses: An analysis of professional practices. Pratiques

Psychologiques, 21, 55–70. http://dx.doi.org/10.1016/j.prps.2014.11.001

Lawson, V. Z., & Dysart, J. E. (2014). The showup identification proce-

dure: An exploration of systematic biases. Legal and Criminological

Psychology, 19, 54–68. http://dx.doi.org/10.1111/j.2044-8333.2012

.02057.x

Levett, L. M. (2013). Co-witness information influences whether a witness

is likely to choose from a lineup. Legal and Criminological Psychology,

18, 168–180. http://dx.doi.org/10.1111/j.2044-8333.2011.02033.x

Levi, A. M. (2011). Much better than the sequential lineup: A 120-person

lineup. Psychology, Crime & Law, 18, 631–640. http://dx.doi.org/10

.1080/1068316X.2010.526120

Levi, A. M., & Lindsay, R. C. L. (2001). Lineup and photo spread

procedures: Issues concerning policy recommendations. Psychology,

Public Policy, and Law, 7, 776–790. http://dx.doi.org/10.1037/1076-

8971.7.4.776

Lindsay, R. C., Martin, R., & Webber, L. (1994). Default values in

eyewitness descriptions. Law and Human Behavior, 18, 527–541. http://

dx.doi.org/10.1007/BF01499172

Lindsay, R. C. L., Nosworthy, G. J., Martin, R., & Martynuck, C. (1994).

Using mug shots to find suspects. Journal of Applied Psychology, 79,

121–130. http://dx.doi.org/10.1037/0021-9010.79.1.121

Lindsay, R. C. L., Wallbridge, H., & Drennan, D. (1987). Do clothes make

the man? An exploration of the effect of lineup attire on eyewitness

identification accuracy. Canadian Journal of Behavioural Science, 19,

463–478. http://dx.doi.org/10.1037/h0079998

Lindsay, R. C. L., & Wells, G. L. (1980). What price justice? Exploring the

relationship of lineup fairness to identification accuracy. Law and Hu-

man Behavior, 4, 303–313. http://dx.doi.org/10.1007/BF01040622

Loftus, E. F. (2004). Memories of things unseen. Current Directions in

Psychological Science, 13, 145–147. http://dx.doi.org/10.1111/j.0963-

7214.2004.00294.x

Loftus, E. F. (2017). Eavesdropping on memory. Annual Review of Psy-

chology, 68, 1–18. http://dx.doi.org/10.1146/annurev-psych-010416-

044138

Loftus, E. F., & Greene, E. (1980). Warning: Even memory for faces may

be contagious. Law and Human Behavior, 4, 323–334. http://dx.doi.org/

10.1007/BF01040624

Loftus, G. R., & Harley, E. M. (2005). Why is it easier to identify someone

close than far away? Psychonomic

Bulletin & Review, 12, 43–65.

http://dx.doi.org/10.3758/BF03196348

Luus, C. A. E., & Wells, G. L. (1991). Eyewitness identification and the

selection of distracters for lineups. Law and Human Behavior, 15,

43–57. http://dx.doi.org/10.1007/BF01044829

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

WELLS ET AL.

Malpass, R. S. (1981). Effective size and defendant bias in eyewitness

identification lineups. Law and Human Behavior, 5, 299–309. http://dx

.doi.org/10.1007/BF01044945

Malpass, R. S., & Devine, P. G. (1981). Eyewitness identification: Lineup

instructions and the absence of the offender. Journal of Applied Psy-

chology, 66, 482–489. http://dx.doi.org/10.1037/0021-9010.66.4.482

Mandler, G. (1980). Recognizing: The judgment of previous occurrence.

Psychological Review, 87, 252–271. http://dx.doi.org/10.1037/0033-

295X.87.3.252

Manson v. Braithwaite, 432 U.S. 98 (1977).

Mansour, J. K., Beaudry, J. L., Kalmet, N., Bertrand, M. I., & Lindsay,

R. C. L. (2017). Evaluating lineup fairness: Variations across methods

and measures. Law and Human Behavior, 41, 103–115. http://dx.doi

.org/10.1037/lhb0000203

Matuku, K., Douglass, A. B., & Charman, S. (2018, March). A cautionary

note about videotaped identification procedures. Paper presented at the

annual meeting of the American Psychology-Law Society, Memphis,

TN.

McPhee, I., Paterson, H. M., & Kemp, R. I. (2014). The power of the

spoken word: Can spoken recall enhance eyewitness evidence? Psychi-

atry, Psychology and Law, 21, 551–566. http://dx.doi.org/10.1080/

13218719.2013.848001

Mecklenburg, S. H. (2006). Report to the legislature of the State of Illinois:

The Illinois pilot program on sequential double-blind identifica-

tion procedures. Retrieved from https://www.cga.ct.gov/jud/tfs/

20130901_Eyewitness%20Identification%20Task%20Force/20110921/

Illinois%20ID%20Pilot%20Program.pdf

Mecklenburg, S. H., Bailey, P. J., & Larson, M. R. (2008). The Illinois

Field Study: A significant contribution to understanding real world

eyewitness identification issues. Law and Human Behavior, 32, 22–27.

http://dx.doi.org/10.1007/s10979-007-9108-6

Meissner, C. A. (2002). Applied aspects of the instructional bias effect in

verbal overshadowing. Applied Cognitive Psychology, 16, 911–928.

http://dx.doi.org/10.1002/acp.918

Meissner, C. A., Brigham, J. C., & Kelley, C. M. (2001). The influence of

retrieval processes in verbal overshadowing. Memory & Cognition, 29,

176–186. http://dx.doi.org/10.3758/BF03195751

Meissner, C. A., Sporer, S. L., & Schooler, J. W. (2007). Person descrip-

tions as eyewitness evidence. In R. C. L. Lindsay, D. F. Ross, J. D. Read,

& M. P. Toglia (Eds.), Handbook of eyewitness psychology: Vol. II.

Memory for people (pp. 1–34). Mahwah, NJ: Erlbaum.

Memon, A., Gabbert, F., & Hope, L. (2004). The ageing eyewitness. In J.

Adler (Ed.), Forensic psychology: Debates, concepts and practice (pp.

96–112). Devon, UK: Willan.

Memon, A., Havard, C., Clifford, B., Gabbert, F., & Watt, M. (2011). A

field evaluation of the VIPER system: A new technique for eliciting

eyewitness evidence. Psychology, Crime & Law, 17, 711–729. http://dx

.doi.org/10.1080/10683160903524333

Memon, A., Meissner, C. A., & Fraser, J. (2010). The Cognitive Interview:

A meta-analytic review and study space analysis of the past 25 years.

Psychology, Public Policy, and Law, 16, 340–372. http://dx.doi.org/10

.1037/a0020518

Mickes, L. (2015). Receiver operating analysis and confidence-accuracy

characteristic analysis in investigations of system variables and estima-

tor variables that affect eyewitness memory. Journal of Applied Re-

search in Memory & Cognition, 4, 93–102. http://dx.doi.org/10.1016/j

.jarmac.2015.01.003

Modjadidi, K., & Kovera, M. B. (2018). Viewing videotaped identification

procedure increases juror sensitivity to single-blind photo-array admin-

istration. Law and Human Behavior, 42, 244–257. http://dx.doi.org/10

.1037/lhb0000288

Molinaro, P. F., Arndorfer, A., & Charman, S. D. (2013). Appearance-

change instruction effects on eyewitness lineup identification accuracy

are not moderated by amount of appearance change. Law and Human

Behavior, 37, 432–440. http://dx.doi.org/10.1037/lhb0000049

Nairne, J. S. (2002). Remembering over the short-term: The case against

the standard model.

Annual Review of Psychology, 53, 53–

81. http://dx

.doi.org/10.1146/annurev.psych.53.100901.135131

National

Research Council. (2014). Identifying the culprit: Assessing eye-

witness identification. Washington, DC: The National Academies Press.

Retrieved from https://www.nap.edu/catalog/18891/identifyingthe-

culprit-assessing-eyewitness-identification

Neil v. Biggers, 409 U.S. 188 (1972).

Neuschatz, J. S., Wetmore, S. A., Key, K. N., Cash, D. K., Gronlund, S. D.,

& Goodsell, C. A. (2016). A comprehensive evaluation of showups. In

B. Bornstein & M. K. Miller (Eds.), Advances in psychology and law

(pp. 43–69). Cham, Switzerland: Springer International Publishing.

http://dx.doi.org/10.1007/978-3-319-29406-3_2

New Jersey v. Henderson, 27 A. 3d 872 (2011).

Newman, E. J., & Garry, M. (2013). False memory. In T. Perfect & D. S.

Lindsay (Eds.), SAGE handbook of applied memory (pp. 110–126). Los

Angeles, CA: SAGE Publications.

Oregon v. Lawson, 352 Or. 724 (2012).

Orfield, M. W., Jr. (1987). The exclusionary rule and deterrence: An

empirical study of Chicago narcotics officers. Chicano Law Review, 54,

1016–1069. http://dx.doi.org/10.2307/1599834

Orfield, M. W., Jr. (1992). Deterrence, perjury, and the heater factor: An

exclusionary rule in the Chicago criminal courts. University of Colorado

Law Review, 63, 75–161.

Palmer, M. A., Brewer, N., & Weber, N. (2010). Postidentification feed-

back affects subsequent eyewitness identification performance. Journal

of Experimental Psychology: Applied, 16, 387–398. http://dx.doi.org/10

.1037/a0021034

Palmer, M. A., Brewer, N., Weber, N., & Nagesh, A. (2013). The

confidence-accuracy relationship for eyewitness identification decisions:

Effects of exposure duration, retention interval, and divided attention.

Journal of Experimental Psychology: Applied, 19, 55–71. http://dx.doi

.org/10.1037/a0031602

Perry v. New Hampshire, 565 U.S. 228 (2012).

Pezdek, K., & Blandon-Gitlin, I. (2005). When is an intervening lineup

most likely to affect eyewitness identification accuracy? Legal and

Criminological Psychology, 10, 247–263. http://dx.doi.org/10.1348/

135532505X49846

Phillips, M. R., McAuliff, B. D., Kovera, M. B., & Cutler, B. L. (1999).

Double-blind photoarray administration as a safeguard against investi-

gator bias. Journal of Applied Psychology, 84, 940–951. http://dx.doi

.org/10.1037/0021-9010.84.6.940

Pickel, K. L. (1998). Unusualness and threat as possible causes of “weapon

focus”. Memory, 6, 277–295. http://dx.doi.org/10.1080/741942361

Pickel, K. L. (1999). The influence of context on the “weapon focus”

effect. Law and Human Behavior, 23, 299–311. http://dx.doi.org/10

.1023/A:1022356431375

Police Executive Research Forum. (2013). A national survey of eyewitness

identification procedures in law enforcement agencies. Retrieved from

http://www.policeforum.org/assets/docs/Free_Online_Documents/

Eyewitness_Identification/a national survey of eyewitness identification

procedures in law enforcement agencies 2013.pdf

Pryke, S., Lindsay, R. C. L., Dysart, J. E., & Dupuis, P. (2004). Multiple

independent identification decisions: A method of calibrating eyewitness

identifications. Journal of Applied Psychology, 89, 73–84. http://dx.doi

.org/10.1037/0021-9010.89.1.73

Quinlivan, D. S., Neuschatz, J. S., Cutler, B. L., Wells, G. L., McLung, J.,

& Harker, D. L. (2012). Do pre-admonition suggestions moderate the

effect of unbiased lineup instructions? Legal and Criminological Psy-

chology, 17, 165–176. http://dx.doi.org/10.1348/135532510X533554

Ratcliff, J. J., Lassiter, G. D., Schmidt, H. C., & Snyder, C. J. (2006).

Camera perspective bias in videotaped confessions: Experimental evi-

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

EYEWITNESS IDENTIFICATION RECOMMENDATIONS

dence of its perceptual basis. Journal of Experimental Psychology:

Applied, 12, 197–206. http://dx.doi.org/10.1037/1076-898X.12.4.197

Read, J. D., Yuille, J. C., & Tollestrup, P. (1992). Recollections of a

robbery. Law and Human Behavior, 16, 425–446. http://dx.doi.org/10

.1007/BF02352268

Reardon, M. C., & Fisher, R. P. (2011). Effect of viewing the interview and

identification process on juror perceptions of eyewitness accuracy. Ap-

plied Cognitive Psychology, 25, 68–77. http://dx.doi.org/10.1002/acp

.1643

Rhead, L. M., Rodriguez, D. N., Korobeynikov, V., Yip, J. H., & Kovera,

M. B. (2015). The effects of lineup administrator influence and mortality

salience on witness identification accuracy. Journal of Forensic Psy-

chology Practice, 15, 248–274. http://dx.doi.org/10.1080/15228932

.2015.1041362

Rodriguez, D. N., & Berry, M. A. (2014). The effect of line-up adminis-

trator blindness on the recording of eyewitness identification decisions.

Legal and Criminological Psychology, 19, 69–79. http://dx.doi.org/10

.1111/j.2044-8333.2012.02058.x

Rodriguez, D. N., & Berry, M. A. (2019). Administrator blindness affects

the recording of eyewitness lineup outcomes. Law and Human Behavior,

44, 71–87. http://dx.doi.org/10.1037/lhb0000352

Rosenthal, R. (2002). Covert communication in classrooms, clinics, court-

rooms, and cubicles. American Psychologist, 57, 839–849. http://dx.doi

.org/10.1037/0003-066X.57.11.839

Ross, D. F., Ceci, S. J., Dunning, D., & Toglia, M. P. (1994). Unconscious

transference and mistaken identity: When a witness misidentifies a

familiar but innocent person. Journal of Applied Psychology, 79, 918–

930. http://dx.doi.org/10.1037/0021-9010.79.6.918

Satin, G. E., & Fisher, R. P. (2019). Investigative utility of the Cognitive

Interview: Describing and finding perpetrators. Law and Human Behav-

ior, 43, 491–506. http://dx.doi.org/10.1037/lhb0000326

Sauer, J., Brewer, N., Zweck, T., & Weber, N. (2010). The effect of

retention interval on the confidence-accuracy relationship for eyewitness

identification. Law and Human Behavior, 34, 337–347. http://dx.doi.org/

10.1007/s10979-009-9192-x

Sauerland, M., Krix, A. C., van Kan, N., Glunz, S., & Sak, A. (2014).

Speaking is silver, writing is golden? The role of cognitive and social

factors in written versus spoken witness accounts. Memory & Cognition,

42, 978–992. http://dx.doi.org/10.3758/s13421-014-0401-6

Sauerland, M., & Sporer, S. L. (2011). Written vs. spoken eyewitness

accounts: Does modality of testing matter? Behavioral Sciences & the

Law, 29, 846–857. http://dx.doi.org/10.1002/bsl.1013

Scheck, B., Neufeld, P., & Dwyer, J. (2000). Actual innocence. New York,

NY: Random House.

Schreiber Compo, N., Carol, R. N., Evans, J. R., Pimentel, P., Holness, H.,

Nichols-Lopez, K.,...Furton, K. G. (2017). Witness memory and

alcohol: The effects of state-dependent recall. Law and Human Behav-

ior, 41, 202–215. http://dx.doi.org/10.1037/lhb0000224

Shaw, J. S., III. (1996). Increases in eyewitness confidence resulting from

post-event questioning. Journal of Experimental Psychology: Applied, 2,

126–146. http://dx.doi.org/10.1037/1076-898X.2.2.126

Shaw, J. S., III, Garven, S., & Wood, J. M. (1997). Co-witness information

can have immediate effects on eyewitness memory reports. Law and

Human Behavior, 21, 503–523. http://dx.doi.org/10.1023/A:

1024875723399

Shaw, J. S., III, & McClure, K. A. (1996). Repeated post-event questioning

can lead to elevated levels of eyewitness confidence. Law and Human

Behavior, 20,

629–653. http://dx.doi.org/10.1007/BF01499235

Skov, R.

B., & Sherman, S. J. (1986). Information-gathering processes:

Diagnosticity, hypothesis-confirmatory strategies, and perceived hy-

pothesis confirmation. Journal of Experimental Social Psychology, 22,

93–121. http://dx.doi.org/10.1016/0022-1031(86)90031-4

Slobogin, C. (1996). Testifying: Police perjury and what to do about it.

University of Colorado Law Review, 67, 1037–1060.

Smalarz, L., Kornell, N., Vaughn, K. E., & Palmer, M. A. (2019). Identi-

fication performance from multiple lineups: Should eyewitnesses who

pick fillers be burned? Journal of Applied Research in Memory &

Cognition, 8, 221–232. http://dx.doi.org/10.1016/j.jarmac.2019.03.001

Smalarz, L., & Wells, G. L. (2014a). Confirming feedback following a

mistaken identification impairs memory for the culprit. Law and Human

Behavior, 38, 283–292. http://dx.doi.org/10.1037/lhb0000078

Smalarz, L., & Wells, G. L. (2014b). Post-identification feedback to

eyewitnesses impairs evaluators’ abilities to discriminate between accu-

rate and mistaken testimony. Law and Human Behavior, 38, 194–202.

http://dx.doi.org/10.1037/lhb0000067

Smith, A. M., Wells, G. L., Lindsay, R. C. L., & Myerson, T. (2018).

Eyewitness identification performance on showups improves with an

additional-opportunities instruction: Evidence for present-absent criteria

discrepancy. Law and Human Behavior, 42, 215–226. http://dx.doi.org/

10.1037/lhb0000284

Smith, A. M., Wells, G. L., Lindsay, R. C. L., & Penrod, S. D. (2017). Fair

lineups are better than biased lineups and showups, but not because they

increase underlying discriminability. Law and Human Behavior, 41,

127–145. http://dx.doi.org/10.1037/lhb0000219

Smith, A. M., Wells, G. L., Smalarz, L., & Lampinen, J. M. (2018).

Increasing the similarity of lineup fillers to the suspect improves applied

value of lineups without improving memory performance. Psychological

Science, 29, 1548–1551. http://dx.doi.org/10.1177/0956797617698528

Smith, A. M., Wilford, M. M., Quigley-McBride, A., & Wells, G. L.

(2019). Mistaken eyewitness identification rates increase when either

witnessing or testing conditions get worse. Law and Human Behavior,

43, 358–368. http://dx.doi.org/10.1037/lhb0000334

Smith-Spark, J. H., Bartimus, J., & Wilcock, R. (2017). Mental time travel

ability and the Mental Reinstatement of Context for crime witnesses.

Consciousness and Cognition, 48, 1–10. http://dx.doi.org/10.1016/j

.concog.2016.10.002

Snyder, M., & Swann, W. B. (1978). Hypothesis-testing processes in social

interaction. Journal of Personality and Social Psychology, 36, 1202–

1212. http://dx.doi.org/10.1037/0022-3514.36.11.1202

Sporer, S. L. (1992). Post-dicting eyewitness accuracy: Confidence,

decision-times and person descriptions of choosers and non-choosers.

European Journal of Social Psychology, 22, 157–180. http://dx.doi.org/

10.1002/ejsp.2420220205

Sporer, S. L. (1993). Eyewitness identification accuracy, confidence and

decision-times in simultaneous and sequential lineups. Journal of Ap-

plied Psychology, 78, 22–33. http://dx.doi.org/10.1037/0021-9010.78

.1.22

Sporer, S. L. (1996). Psychological aspects of person descriptions. In S. L.

Sporer, R. S. Malpass, & G. Koehnken (Eds.), Psychological issues in

eyewitness identification (pp. 53–86). Hillsdale, NJ: Erlbaum, Inc.

State v. Lawson, 352 Or. 724, 291 P. 3d 673 (2012).

Steblay, N. M. (1997). Social influence in eyewitness recall: A meta-

analytic review of lineup instruction effects. Law and Human Behavior,

21, 283–297. http://dx.doi.org/10.1023/A:1024890732059

Steblay, N. K. (2011). What we know now: The Evanston Illinois field

lineups. Law and Human Behavior, 35, 1–12. http://dx.doi.org/10.1007/

s10979-009-9207-7

Steblay, N. K. (2013). Lineup Instructions. In B. Cutler (Ed.), Reform of

eyewitness identification procedures (pp. 65–86). Washington, DC:

APA Press. http://dx.doi.org/10.1037/14094-004

Steblay, N. K., & Dysart, J. E. (2016). Repeated eyewitness identification

procedures with the same suspect. Journal of Applied Research in

Memory & Cognition, 5, 284–289.

http://dx.doi.org/10.1016/j.jarmac

.2016.06.010

Steblay,

N.,

Dysart, J., Fulero, S., & Lindsay, R. C. L. (2003). Eyewitness

accuracy rates in police showup and lineup presentations: A meta-

analytic comparison. Law and Human Behavior, 27, 523–540. http://dx

.doi.org/10.1023/A:1025438223608

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

WELLS ET AL.

Steblay, N. K., Tix, R. W., & Benson, S. L. (2013). Double exposure: The

effects of repeated identification lineups on eyewitness accuracy. Ap-

plied Cognitive Psychology, 27, 644–654. http://dx.doi.org/10.1002/acp

.2944

Steblay, N., Wells, G. L., & Douglass, A. B. (2014). The eyewitness post

identification feedback effect 15 years later: Theoretical and policy

implications. Psychology, Public Policy, and Law, 20, 1–18. http://dx

.doi.org/10.1037/law0000001

Sternberg, K. J., Lamb, M. E., Esplin, P. W., Orbach, Y., & Hershkowitz,

I. (2002). Using a structure interview protocol to improve the quality of

investigative interviews. In M. Eisen (Ed.), Memory and suggestibility in

the forensic interview (pp. 409–436). Mahwah, NJ: Erlbaum.

Technical Working Group for Eyewitness Evidence. (1999). Eyewitness

evidence: A guide for law enforcement. Washington, DC: U.S. Depart-

ment of Justice, Office of Justice Programs.

Tredoux, C. (1998). Statistical inference on measures of lineup fairness.

Law and Human Behavior, 22, 217–237. http://dx.doi.org/10.1023/A:

1025746220886

Tredoux, C., Parker, J. F., & Nunez, D. (2007). Predicting eyewitness

identification accuracy with mock witness measures of lineup fair-

ness: Quality of encoding interacts with line up format. South African

Journal of Psychology, 37, 207–222. http://dx.doi.org/10.1177/

008124630703700201

Tuckey, M. R., & Brewer, N. (2003). The influence of schemas, stimulus

ambiguity, and interview schedule on eyewitness memory over time.

Journal of Experimental Psychology: Applied, 9, 101–118. http://dx.doi

.org/10.1037/1076-898X.9.2.101

Tunnicliff, J. L., & Clark, S. E. (2000). Selecting foils for identification

lineups: Matching suspects or descriptions? Law and Human Behavior,

24, 231–258. http://dx.doi.org/10.1023/A:1005463020252

United States Department of Justice. (2017, January). Eyewitness identifi-

cation: Procedures for conducting photo arrays. Memorandum for heads

of department law enforcement components and all department prosecutors

from Sally Yates, Deputy Attorney General, Washington, DC.

United States v. Ash, 413 U.S. 300 (1973).

United States vs. Wade, 388 U.S. 218, 234 (1967).

Valentine, T., Davis, J. P., Memon, A., & Roberts, A. (2012). Live

show-ups and their influence on a subsequent video lineup. Applied

Cognitive Psychology, 26, 1–23. http://dx.doi.org/10.1002/acp.1796

Valentine, T., Pickering, A., & Darling, S. (2003). Characteristics of

eyewitness identification that predict the outcome of real lineups. Ap-

plied Cognitive Psychology, 17, 969–993. http://dx.doi.org/10.1002/acp

.939

Vallano, J., Steele, L. E., Slapinski, K. A., Briggs, A. P., & Pozzulo, J. D.

(2019). Familiar eyewitness identifications: The current state of affairs.

Psychology, Public Policy, and Law, 25, 128–146. http://dx.doi.org/10

.1037/law0000204

Van Koppen, P. J., & Lochun, S. K. (1997). Portraying perpetrators: The

validity of offender descriptions by witnesses. Law and Human Behav-

ior, 21, 661–685. http://dx.doi.org/10.1023/A:1024812831576

Vollmer, R. T. (2006). Predictive probability of serum prostate-specific

antigen for prostate cancer: An approach using Bayes rule. American

Journal of Clinical Pathology, 125, 336–342. http://dx.doi.org/10.1309/

R5H6VUQ32KGJW448

Vredeveldt, A., Baddeley, A. D., & Hitch, G. J. (2012). The effects of

eye-closure and “ear-closure” on recall of visual and auditory aspects of

a criminal event. Europe’s Journal of Psychology, 8, 284–299. http://

dx.doi.org/10.5964/ejop.v8i2.472

Vredeveldt, A., Baddeley, A. D., & Hitch, G. J. (2014). The effectiveness

of eye-closure in repeated interviews. Legal and Criminological Psy-

chology, 19, 282–295. http://dx.doi.org/10.1111/lcrp.12013

Vredeveldt, A., Hildebrandt, A., & van Koppen, P. J. (2016). Acknowl-

edge, repeat, rephrase, elaborate: Witnesses can help each other remem-

ber more. Memory, 24, 669–682. http://dx.doi.org/10.1080/09658211

.2015.1042884

Wade, K. A., Nash, R. A., & Lindsay, D. S. (2018). Reasons to doubt the

reliability

eyewitness memory: Commentary on Wixted, Mickes, and

Fisher (2018). Perspectives on Psychological Science, 13, 339–342.

http://dx.doi.org/10.1177/1745691618758261

Wagenaar, W. A., & van der Schrier, J. H. (1996). Face recognition as a

function of distance and illumination: A practical tool for use in the

courtroom. Psychology, Crime & Law, 2, 321–332. http://dx.doi.org/10

.1080/10683169608409787

Walsh, D., & Bull, R. (2010). What really is effective in interviews with

suspects? A study comparing interviewing skills against interviewing

outcomes. Legal and Criminological Psychology, 15, 305–321. http://

dx.doi.org/10.1348/135532509X463356

Weber, N., Brewer, N., Wells, G. L., Semmler, C., & Keast, A. (2004).

Eyewitness identification accuracy and response latency: The unruly

10–12-second rule. Journal of Experimental Psychology: Applied, 10,

139–147. http://dx.doi.org/10.1037/1076-898X.10.3.139

Weber, N., & Perfect, T. J. (2012). Improving eyewitness identification

accuracy by screening out those who say they don’t know. Law and

Human Behavior, 36, 28–36. http://dx.doi.org/10.1037/h0093976

Wells, G. L. (1978). Applied eyewitness testimony research: System vari-

ables and estimator variables. Journal of Personality and Social Psy-

chology, 36, 1546–1557. http://dx.doi.org/10.1037/0022-3514.36.12

.1546

Wells, G. L. (1984). The psychology of lineup identifications. Journal of

Applied Social Psychology, 14, 89–103. http://dx.doi.org/10.1111/j

.1559-1816.1984.tb02223.x

Wells, G. L. (1993). What do we know about eyewitness identification?

American Psychologist, 48, 553–571. http://dx.doi.org/10.1037/0003-

066X.48.5.553

Wells, G. L. (2001). Eyewitness lineups: Data, theory, and policy. Psy-

chology, Public Policy, and Law, 7, 791–801. http://dx.doi.org/10.1037/

1076-8971.7.4.791

Wells, G. L. (2006). Eyewitness identification: Systemic reforms. Wiscon-

sin Law Review, 2006, 615–643.

Wells, G. L., & Bradfield, A. L. (1998). “Good, you identified the suspect”:

Feedback to eyewitnesses distorts their reports of the witnessing expe-

rience. Journal of Applied Psychology, 83, 360–376. http://dx.doi.org/

10.1037/0021-9010.83.3.360

Wells, G. L., & Bradfield, A. L. (1999). Measuring the goodness of

lineups: Parameter estimation, question effects, and limits to the mock

witness paradigm. Applied Cognitive Psychology, 13, 27–40.

Wells, G. L., Leippe, M. R., & Ostrom, T. M. (1979). Guidelines for

empirically assessing the fairness of a lineup. Law and Human Behavior,

3, 285–293. http://dx.doi.org/10.1007/BF01039807

Wells, G. L., & Lindsay, R. C. L. (1980). On estimating the diagnosticity

of eyewitness nonidentifications. Psychological Bulletin, 88, 776–784.

http://dx.doi.org/10.1037/0033-2909.88.3.776

Wells, G. L., & Luus, E. (1990). Police lineups as experiments: Social

methodology as a framework for properly-conducted lineups. Personal-

ity and Social Psychology Bulletin, 16, 106–117. http://dx.doi.org/10

.1177/0146167290161008

Wells, G. L., & Olson, E. A. (2002). Eyewitness identification: Information

gain from incriminating and exonerating behaviors. Journal of Experi-

mental Psychology: Applied, 8, 155–167. http://dx.doi.org/10.1037/

1076-898X.8.3.155

Wells, G. L., & Quinlivan, D. S. (2009). Suggestive eyewitness identifi-

cation procedures and the Supreme Court’s reliability test in light of

eyewitness science: 30 years later. Law and Human Behavior, 33, 1–24.

http://dx.doi.org/10.1007/s10979-008-9130-3

Wells, G. L., Rydell, S. M., & Seelau, E. P. (1993). On the selection of

distractors for eyewitness lineups. Journal of Applied Psychology, 78,

835–844. http://dx.doi.org/10.1037/0021-9010.78.5.835

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

EYEWITNESS IDENTIFICATION RECOMMENDATIONS

Wells, G. L., Small, M., Penrod, S. J., Malpass, R. S., Fulero, S. M., &

Brimacombe, C. A. E. (1998). Eyewitness identification procedures:

Recommendations for lineups and photospreads. Law and Human Be-

havior, 22, 603–647. http://dx.doi.org/10.1023/A:1025750605807

Wells, G. L., Smith, A. M., & Smalarz, L. (2015). ROC analysis of lineups

obscures information that is critical for both theoretical understanding

and applied purposes. Journal of Applied Research in Memory & Cog-

nition, 4, 324–328. http://dx.doi.org/10.1016/j.jarmac.2015.08.010

Wells, G. L., Steblay, N. K., & Dysart, J. E. (2012). Eyewitness identifi-

cation reforms: Are suggestiveness-induced hits and guesses true hits?

Perspectives on Psychological Science, 7, 264–271. http://dx.doi.org/10

.1177/1745691612443368

Wells, G. L., Steblay, N. K., & Dysart, J. E. (2015). Double-blind photo

lineups using actual eyewitnesses: An experimental test of a sequential

versus simultaneous lineup procedure. Law and Human Behavior, 39,

1–14. http://dx.doi.org/10.1037/lhb0000096

Wells, G. L., & Turtle, J. W. (1986). Eyewitness identification: The

importance of lineup models. Psychological Bulletin, 99, 320–329.

http://dx.doi.org/10.1037/0033-2909.99.3.320

Wells, G. L., Yang, Y., & Smalarz, L. (2015). Eyewitness identification:

Bayesian information gain, base-rate effect equivalency curves, and

reasonable suspicion. Law and Human Behavior, 39, 99–122. http://dx

.doi.org/10.1037/lhb0000125

West, E., & Meterko, V. (2017). Innocence Project: DNA exonerations,

1989–2014: Review of data and findings from the first 25 years. Albany

Law Review, 79, 717–795.

Wetmore, S. A., Neuschatz, J. S., Gronlund, S. D., Wooten, A., Goodsell,

C. A., & Carlson, C. A. (2015). Effect of retention interval on showup

and lineup performance. Journal of Applied Research in Memory &

Cognition, 4, 8–14. http://dx.doi.org/10.1016/j.jarmac.2014.07.003

Wilford, M. M., & Wells, G. L. (2013). Eyewitness system variables:

Revisiting the system-variable concept and the transfer of system vari-

ables to the legal system. In B. L. Cutler (Ed.), Reform of eyewitness

identification procedures (pp. 23–43). Washington, DC: American Psy-

chological Association. http://dx.doi.org/10.1037/14094-002

Wise, R. A., Safer, M. A., & Maro, C. M. (2011). What U.S. law

enforcement officers know and believe about eyewitness factors, eye-

witness interviews, and identification procedures. Applied Cognitive

Psychology, 25, 488–500. http://dx.doi.org/10.1002/acp.1717

Wixted, J. T., & Mickes, L. (2014). A signal-detection-based diagnostic-

feature-detection model of eyewitness identification. Psychological Re-

view, 121, 262–276. http://dx.doi.org/10.1037/a0035940

Wixted, J. T., Mickes, L., Dunn, J. C., Clark, S. E., & Wells, W. (2016).

Estimating the reliability of eyewitness identifications from police line-

ups. Proceedings of the National Academies of Science, 113, 304–309.

http://dx.doi.org/10.1073/pnas.1516814112

Wixted, J. T., & Wells, G. L. (2017). The relationship between eyewitness

confidence and identification accuracy: A new synthesis. Psychological

Science in the Public Interest, 18, 10–65. http://dx.doi.org/10.1177/

1529100616686966

Wogalter, M. S. (1991). Effects of post-exposure description and im-

aging on subsequent face recognition performance. Proceedings of

the Human Factors Society, 35, 575–579. http://dx.doi.org/10.1518/

107118191786754680

Wogalter, M. S. (1996). Describing faces from memory: Accuracy and

effects on subsequent recognition performance. Proceedings of the Hu-

man Factors Society, 40, 536–540. http://dx.doi.org/10.1177/

154193129604001103

Wright, D. B., & McDaid, A. T. (1996). Comparing system and estimator

variables using data from real lineups. Applied Cognitive Psychology,

10, 75–84. http://dx.doi.org/10.1002/(SICI)1099-0720(199602)10:

1⬍75::AID-ACP364⬎3.0.CO;2-E

Wright, D. B., Memon, A., Skagerberg, E. M., & Gabbert, F. (2009). When

eyewitnesses talk. Current

Directions in Psychological Science, 18,

174–178. http://dx.doi.org/10.1111/j.1467-8721.2009.01631.x

Wright, D. B., & Skagerberg, E. M. (2007). Postidentification feedback

affects real eyewitnesses. Psychological Science, 18, 172–178. http://dx

.doi.org/10.1111/j.1467-9280.2007.01868.x

Yarmey, A. D. (1986). Verbal, visual, and voice identification of a rape

suspect under different levels of illumination. Journal of Applied Psy-

chology, 71, 363–370. http://dx.doi.org/10.1037/0021-9010.71.3.363

Yarmey, A. D. (2000). Retrospective duration estimations for variant and

invariant events in field situations. Applied Cognitive Psychology, 14,

45–57. http://dx.doi.org/10.1002/(SICI)1099-0720(200001)14:1⬍45::

AID-ACP623⬎3.0.CO;2-U

Yarmey, A. D., Jacob, J., & Porter, A. (2002). Person recall in field

settings. Journal of Applied Social Psychology, 32, 2354–2367. http://

dx.doi.org/10.1111/j.1559-1816.2002.tb01866.x

Yarmey, A. D., Yarmey, M. J., & Yarmey, A. L. (1996). Accuracy of

eyewitness identifications in showups and lineups. Law and Human

Behavior, 20, 459–477. http://dx.doi.org/10.1007/BF01498981

Young v. Conway 715 F. 3d 79 (2d Cir. 2013).

Yuille, J. C., & Cutshall, J. L. (1986). A case study of eyewitness memory

of a crime. Journal of Applied Psychology, 71, 291–301. http://dx.doi

.org/10.1037/0021-9010.71.2.291

Yuille, J. C., & Tollestrup, P. A. (1990). Some effects of alcohol on

eyewitness memory. Journal of Applied Psychology, 75, 268–273.

http://dx.doi.org/10.1037/0021-9010.75.3.268

Zajac, R., & Henderson, N. (2009). Don’t it make my brown eyes blue:

Co-witness misinformation about a target’s appearance can impair

target-absent line-up performance. Memory, 17, 266–278. http://dx.doi

.org/10.1080/09658210802623950

Zajac, R., & Karageorge, A. (2009). The wildcard: A simple technique for

improving children’s target-absent lineup performance. Applied Cogni-

tive Psychology, 23, 358–368. http://dx.doi.org/10.1002/acp.1511

Zimmerman, D. M., Chorn, J. A., Rhead, L. M., Evelo, A. J., & Kovera,

M. B. (2017). Memory strength and lineup presentation moderate effects

of administrator influence on mistaken identifications. Journal of Ex-

perimental Psychology: Applied, 23, 460–473. http://dx.doi.org/10

.1037/xap0000147

Received June 4, 2019

Revision received November 19, 2019

Accepted November 20, 2019 䡲

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Content may be shared at no cost, but any requests to reuse this content in part or whole must go through the American Psychological Association.

WELLS ET AL.