ArticlePDF Available

Disputed eyewitness identification evidence: Important legal and scientific issues

Authors:
12 Court Review - Summer 1999
Footnotes
1. See, e.g., CUTLER & PENROD, MISTAKEN IDENTIFICATION: THE
EYEWITNESS, PSYCHOLOGY, AND LAW (1995); LOFTUS, EYEWITNESS
TESTIMONY (1979); LOFTUS & DOYLE, EYEWITNESS TESTIMONY: CIVIL
AND CRIMINAL (1987); ROSS, ET AL., ADULT EYEWITNESS TESTIMONY:
CURRENT TRENDS AND DEVELOPMENTS (1994); SPORER, ET AL.,
PSYCHOLOGICAL ISSUES IN EYEWITNESS IDENTIFICATION (1996).
2. HUFF, ET AL., CONVICTED BUT INNOCENT: WRONGFUL CONVICTIONS AND
PUBLIC POLICY (1996); Huff, et al., Guilty until Proved Innocent:
Wrongful Convictions and Public Policy, 32 CRIME & DELINQ. 518
(1986).
3. BORCHARD, CONVICTING THE INNOCENT: SIXTY-FIVE ACTUAL ERRORS OF
CRIMINAL JUSTICE (1932).
4. 388 U.S. 218 (1967)
5. Id. at 228.
6. WALL, EYEWITNESS IDENTIFICATION IN CRIMINAL CASES (1964).
7. Id. at 26.
8. Huff, et al., supra note 2.
Eyewitness testimony, when delivered
in a confident manner by a witness,
may be more convincing to jurors
than any other type of evidence. When a
witness declares, “That’s the man I saw,
right there!” most jurors are persuaded that
the identification is accurate. After all, the
witness was there: why would he or she be
mistaken? This strong belief by jurors in
the accuracy of eyewitnesses has been
demonstrated time after time by research
studies.1Yet there is a major problem con-
cerning eyewitness evidence.
Knowledgeable legal scholars and social
scientists have noted that not only is eye-
witness evidence powerful, it is also more
likely to be erroneous than any other type of
evidence.2
Many legal scholars have been aware of
this weakness at least since 1932, when
Edwin Borchard wrote Convicting the Innocent.3Thereafter, the
problematic nature of eyewitness evidence was explicitly
acknowledged by the U.S. Supreme Court in a 1967 decision,
United States v. Wade.4In Wade, Justice Brennan noted, “The
vagaries of eyewitness identification are well-known; the annals
of criminal law are rife with instances of mistaken identifica-
tion.”5Additionally, the justices cited a well-known legal text by
Patrick Wall,6who had written that many judges and lawyers
agreed with the assertion that “[m]istaken identifications have
been responsible for more miscarriages of justice than any other
factor – more so perhaps, than all other factors combined.”7
Our purpose here is not to suggest that, because of its high
error rate, eyewitness evidence should be excluded at trial.
More often than not, eyewitness testimony is accurate, and it is
frequently the only evidence available. Furthermore, we do not
simply wish to bemoan the weakness of eyewitness evidence.
Because such cautions have already been
given to the legal community, our inten-
tion is not to repeat them here. Rather, our
intent is to provide an up-to-date synopsis
of the results of scientific research on fac-
tors that affect the accuracy of eyewitness
identifications, and to suggest ways in
which knowledge of these research find-
ings may be helpful to judges and jurors. It
is our contention that if judicial decision-
makers are aware of the general unreliabil-
ity of eyewitness evidence, and also are
made aware of scientifically-based knowl-
edge about the specific factors that affect
eyewitness accuracy, then the utilization of
this knowledge might significantly reduce
the number of wrongful convictions that
occur. Furthermore, research results sug-
gest that this can happen without greatly
increasing the chances that guilty defen-
dants will go free.
I. ESTIMATING THE PREVALENCE OF EYEWITNESS
ERRORS
How do we know that eyewitness evidence is so error-prone?
One source of information is the analysis of actual cases of
wrongful convictions. Recently, Huff, Rattner and Sagarin,8
made an exhaustive search for cases in the United States since
1900 in which clear instances of erroneous convictions had
occurred. These were not cases in which substantial doubt
remained, but ones in which indisputable evidence of the per-
son’s innocence came to light after the conviction, by way of new
forensic evidence or confessions by others. They identified 205
such cases and categorized them by the type of error that led
most directly to the conviction (e.g., perjury, forensic errors,
negligence by criminal justice officials, coerced confessions,
Disputed Eyewitness
Identification Evidence:
Important Legal and Scientific Issues
John C. Brigham, Adina W. Wasserman and Christian A. Meissner
Summer 1999 - Court Review 13
9. CONNERS, ET AL., CONVICTED BY JURIES, EXONERATED BY SCIENCE: CASE
STUDIES IN THE USE OF DNA EVIDENCE TO ESTABLISH INNOCENCE AFTER
TRIAL (1996).
10. Huff, et al., supra note 2.
11. See Daniel Givelber, Meaningless Acquittals, Meaningful
Convictions: Do We Reliably Acquit the Innocent?, 49 RUTGERS L.
REV. 1317, 1350 (1997)(noting that 1 of 28 wrongful convictions
found in National Institute of Justice study was result of guilty
plea).
12. U. S. Department of Justice, Bureau of Justice Statistics (1995).
13. WRIGHTSMAN, ET AL., PSYCHOLOGY AND THE LEGAL SYSTEM (4th ed.
1998).
14. Fruzzetti, et al., Memory and Eyewitness Testimony. in ASPECTS OF
MEMORY, VOL. 1: THE PRACTICAL ASPECTS 18 (1992); Ross, et al.,.
ADULT EYEWITNESS TESTIMONY: CURRENT TRENDS AND
DEVELOPMENTS (1994); Sporer, et al., PSYCHOLOGICAL ISSUES IN
EYEWITNESS IDENTIFICATION (1996).
15. Deffenbacher, The Influence of Arousal on Reliability of Testimony,
in EVALUATING WITNESS EVIDENCE 235 (1983).
etc.). In more than half of these cases (fifty-two percent), the pri-
mary causal factor leading to conviction was incorrect eyewit-
ness identification.
The National Institute of Justice, under the auspices of the U.
S. Department of Justice, recently conducted a study of wrong-
ful convictions that had later been identified via DNA evidence.9
They identified twenty-eight such cases in which the convicted
person was later exonerated by DNA tests that showed that the
convicted individual could not have been the perpetrator. All of
these cases involved sexual assault, though many involved other
additional offenses. More importantly, all of these cases
involved eyewitness identifications, and in the vast majority of
them, eighty-six percent, erroneous eyewitness identification
was the primary evidence that had produced the conviction.
The report’s authors pointed out that when DNA evidence is not
admitted, “we force the courts to rely on inferior evidence, such
as eyewitness testimony.” In all twenty-eight cases, DNA evi-
dence was not available at trial and the triers of fact had to rely
on eyewitness testimony, which turned out to be inaccurate.
What is the magnitude of the problem? There can never be
an exact answer, because the prevalence of wrongful convictions
can never be known precisely. Still, some social scientists have
made an educated guess. In the mid-1980s, Huff, Rattner, and
Sagarin10 conducted a mail survey in Ohio of all presiding
judges of common pleas courts, all county prosecutors, all
county public defenders, all county sheriffs, and the police
chiefs of the seven largest cities in Ohio. The overall response
rate was sixty-five percent, and at least sixty percent of those in
each category responded. Survey respondents were asked to
estimate the percentage of convictions that were “wrongful con-
victions” of innocent persons. Based upon the pattern of
responses, the researchers estimated that the average response
was that the overall prevalence of wrongful convictions was
about one-half of one percent, or one in every two hundred con-
victions. Unfortunately, we don’t know whether their respon-
dents were thinking only about contested trials, or were includ-
ing all convictions, most of which result from guilty pleas. To
be sure, an erroneous eyewitness identification could result in a
guilty plea, even if to a lesser charge, just as it could result in a
conviction after trial.11 If the wrongful conviction percentage
estimate is applied to all convictions, then the numbers would
be startling. To illustrate, in the mid-1990s there were 2.8 mil-
lion arrests per year in the U.S. just for the eight most serious
crimes on the FBI index. About seventy percent of those
arrested for serious crimes were convicted, producing almost
two million convictions per year.12 If one-half of one percent of
those convictions were of innocent people, then nearly 10,000
wrongful convictions occurred each year. If more than half of
those wrongful convictions were primarily the result of eyewit-
ness errors, then more than 5,000 innocent people may have
been convicted of felonies each year because of erroneous eye-
witness identification! If, on the other hand, the wrongful con-
viction figures are applied only to contested trials, the numbers
would be considerably smaller (i.e., 500 to 1,000 cases per year),
since only ten to twenty percent of all felony prosecutions go to
trial.13 As noted, however, some innocent defendants may
accept a plea bargain if they fear that they will be convicted at
trial. While we will probably never be able to derive a precise
figure for the number of wrongful convictions of innocent
defendants that occur, this remains an issue of concern to many,
and it is clear that erroneous eyewitness identifications form the
backbone of those wrongful convictions that do occur.
II. SCIENTIFIC RESEARCH ON EYEWITNESS
IDENTIFICATION ACCURACY
A second source of evidence concerning erroneous eyewit-
ness identification consists of the many scientific studies of eye-
witness accuracy that have been undertaken in the past three
decades. Researchers have isolated a host of factors that can
influence the ultimate identification of a suspect. As several
exhaustive reviews are readily available in the scientific litera-
ture,14 our goal here is to describe the most grounded findings,
and to highlight the most recent advances in the field. While
early research began by examining general factors that affect
human perception and memory, recent studies have directly
investigated face recognition and eyewitness identifications in
more applied respects. In general, studies have focused on the
conditions under which a misidentification is likely to occur,
including the initial perception of the event or suspect (the
encoding phase), the period in which the memory trace is stored
in memory (the retention interval), and the subsequent recall or
identification of the suspect (the retrieval phase).
Encoding of the Event/Suspect
During the initial perception of the event or suspect, various
factors may influence an eyewitness’s ability to accurately
encode the stimulus (in this case, the crime event and the crim-
inal’s appearance) in memory. Contrary to popular belief,
human perception does not work like a camera or video
recorder. Rather, what is perceived and stored in memory is
often incomplete or distorted as a result of the individual’s state
of mind or the nature of the event observed. For example, in a
violent crime situation, the victim or eyewitness may be paying
a great deal of attention to the event, but the high level of arousal
that they experience is likely to interfere with their ability to
accurately encode details of the event (including the face of the
perpetrator).15 Furthermore, research on the weapon focus effect
has indicated that when a witness is threatened with a weapon
14 Court Review - Summer 1999
16. Steblay, A Meta-Analytic Review of the Weapon Focus Effect, LAW
AND HUMAN BEH., 16, 413 (1992).
17. Darley & Gross, A Hypothesis-confirming Bias in Labelling Effects,
44 J. PERSONALITY AND SOC. PSYCH. 20 (1983); PETERSON,
WITNESSES: MEMORY OF SOCIAL EVENTS (Unpublished doctoral dis-
sertation, Univ. of California at Los Angeles, 1976).
18. Loftus & Ketcham,The Malleability of Eyewitness Accounts, in
EVALUATING WITNESS EVIDENCE 159 (1983); Shaw, et al., Co-witness
Information Can Have Immediate Effects on Eyewitness Memory
Reports,21 LAW & HUMAN BEH. 503 (1997).
19. Jenkins & Davies, Contamination of Facial Memory through
Exposure to Misleading Composite Pictures, 70 J. APPLIED PSYCH.
164 (1985); Loftus & Green, Warning: Even Memory for Faces
May Be Contagious, 4 LAW & HUMAN BEH. 323 (1980).
20. For a review, see Fruzzetti, et al., Memory and Eyewitness
Testimony, in ASPECTS OF MEMORY, VOL. 1: THE PRACTICAL ASPECTS
18 (1992).
21. Lindsay & Johnson, Misleading Suggestibility and Memory for
Source, 17 MEMORY & COGNITION 349 (1989).
22. MEISSNER. SELF-GENERATED MISINFORMATION: THE INFLUENCE OF
RETRIEVAL PROCESSES IN VERBAL OVERSHADOWING (Unpublished
master’s thesis, Florida State Univ, 1998).
23. See Loftus, Unconscious Transference, 2 LAW & PSYCH. REV. 93
(1976); Read, et al., The Unconscious Transference Effect: Are
Innocent Bystanders Ever Misidentified?, 4 APPLIED COGNITIVE
PSYCH. 3 (1990).
24. Brown, et al., Memory for Faces and the Circumstances of
Encounter, 62 J. APPLIED PSYCH. 311 (1977); Gorenstein &
Ellsworth, Effect of Choosing an Incorrect Photograph on a Later
Identification by an Eyewitness, 65 J. APPLIED PSYCH.616 (1981).
(e.g., a knife or gun), the wit-
ness’s attention is drawn to
focus on the weapon, making it
less likely that the appearance
of the person wielding the
weapon will be accurately
encoded.16 An individual’s
expectations of the event can
also influence the manner in
which details about the event
are recalled. This effect, known
as the confirmation bias, illus-
trates that eyewitnesses tend to report a scenario that is consis-
tent with what they expected to see.17 Other, studies have indi-
cated that, in general, memory is better for faces or events seen
for longer durations under optimal observational conditions
such as good lighting, close distance, low stress, and no disguise.
The Retention Interval
The retention interval can be defined as the time from which
the individual perceives and encodes the information to the time
when he or she is asked to retrieve the information from mem-
ory (e.g., view a lineup). During this interval, a number of fac-
tors can influence a witness’s later recall of the event or suspect
from memory. The major finding has been that witnesses are
highly susceptible to suggestions regarding their memory for the
previously viewed event. Such “post-event suggestions” may
come from overhearing the recall of other witnesses or from
questioning by field officers investigating the crime, and may
involve aspects of the situation or facial characteristics of the
suspect.18 For example, several studies have demonstrated that
if witnesses are given another eyewitness’s description of the
suspect, they will be biased toward selecting a member of the
photo lineup who most closely matches the other witness’s
description, even to the extreme of selecting a lineup member
who has a mustache, when the perpetrator did not have one.19
This post-event misinformation effect has been shown to be quite
powerful across a vast number of studies, stimuli, and situa-
tions. Overall, it appears that individuals tend to “commit” to
the post-event misinformation by accepting it as if it were an
accurate account. Researchers are not certain whether this new
information changes the original memory or instead creates a
new memory that “overlies” the original memory, thus making
the original memory temporarily inaccessible.20 More recent
studies have indicated that the latter explanation may be more
plausible, finding that certain conditions at recall appear to
enhance the accurate retrieval of the original event from mem-
ory.21 But regardless of which process occurs, the result is the
same — the witness’s original memory is no longer accessible.
Retrieving the Memory of the Event/Suspect
Research has shown that many factors can affect the accuracy
with which the memory of a crime or a criminal is retrieved
(e.g., when describing the criminal to police or when viewing a
lineup). For example, the manner in which eyewitnesses are
requested to provide a description of the suspect can have
important implications on their ability to subsequently identify
him/her. Studies from our own lab have indicated that when
people are strongly urged to provide a full and “complete”
description of the suspect, they tend to guess (often inaccu-
rately) about features they are not sure of. These inaccurate
guesses interfere with their later ability to recognize the person’s
face in a lineup, producing a higher level of misidentifications
than for those who were not urged to give a “complete” descrip-
tion.22
At the time of retrieval, factors surrounding identification of
the suspect from a lineup are critically important. A host of
studies have investigated the phenomenon labeled unconscious
transference, wherein different memory images may become
combined or confused with one another. Also termed the
bystander effect, this phenomenon occurs when a witness
misidentifies an individual from a photo lineup as the actual sus-
pect when, in reality, the witness previously saw the individual
either as a bystander at the event or in a completely different
context.23 Additionally, research has shown that selecting an
incorrect person from a showup or lineup strongly increases the
likelihood that the same individual will be selected in future
lineups or in-court identifications, despite the inaccuracy of the
original identification.24 Such errors in memory at the time of
retrieval appear to parallel the misinformation effect discussed
previously, in which witnesses tend to commit to a response and
provide the same response in subsequent attempts at retrieval.
Several theoretical explanations for such retrieval-based phe-
nomena have been put forth. Some researchers believe them to
be the result of source confusion, a common memory error that
occurs when a person knows that a face seems familiar, but
[W]itnesses are
highly susceptible
to suggestions
regarding their
memory for the
previously viewed
event.
Summer 1999 - Court Review 15
25. See Lindsay, Memory Source Monitoring and Eyewitness Testimony,
in ADULT EYEWITNESS TESTIMONY: CURRENT TRENDS AND
DEVELOPMENTS 27 (1994).
26. Ross, et al., Unconscious Transference and Lineup Identification:
Toward a Memory Blending Approach, in ADULT EYEWITNESS
TESTIMONY: CURRENT TRENDS AND DEVELOPMENTS 80 (1994).
27. Brigham, Meissner, & Wasserman, Applied Aspects of the
Construction, Administration and Expert Assessment of Photo
Lineups, APPLIED COGNITIVE PSYCH.(in press).
28. Id.
28a. These research findings were summarized in the first “Scientific
Review Paper” created by a special committee of the American
Pschology-Law Society. See Kassin, Eyewitness Identification
Procedures: The Fifth Rule, 22 LAW & HUMAN BEHAV. 649
(1998); Wells, et al., Eyewitness Identification Procedures:
Recommendations for Lineups and Photospreads, 22 LAW &
HUMAN BEHAV. 603 (1998).
29. Wells & Bradfield, “Good, You Identified the Suspect”: Feedback to
Witnesses Distorts Their Reports of the Witnessing Experience, 83 J.
APPLIED PSYCH. 360 (1998).
incorrectly recalls the source of that familiarity.25 Source infor-
mation is the hardest type of memory information to keep
straight, and virtually everyone has experienced this phenome-
non in everyday life. Usually the result of such source confusion
is often harmless embarrassment; however, when the confusion
involves the suspect in a criminal case, this error is no longer
trivial. Another possibility involves memory blending, a theoret-
ical account which states that the two mental images (e.g., the
criminal and a similar-looking person seen in a showup or
lineup) are mentally combined into a single memory represen-
tation. As a result of this “blending,” the original memory of the
criminal may no longer be accessible.26
When police investigators prepare to present a photo lineup
to the eyewitness, several key factors may influence the likeli-
hood of a misidentification. Elsewhere we have recently
reviewed the scientific research on the various psychological fac-
tors involved in the construction and administration of photo
lineups.27 Here we will provide a brief synopsis of the major
findings.
When constructing a photo lineup, it is critical that the sus-
pect is not a distinctive member of the photo lineup.
Distinctiveness can occur because the suspect has physical char-
acteristics that the other lineup members do not have, or
because he or she more closely fits the description of the perpe-
trator than the other lineup members do. To prevent the possi-
bility that the witness might select the target simply due to his
or her distinctiveness from the other foils, it is important that all
members of the photo lineup fit both a general description and
the visual image of the suspect.
During the administration of the photo lineup, intentional or
unintentional verbal and nonverbal cues given by others (i.e.,
law enforcement personnel, attorneys, other witnesses, or other
lineup members) can significantly bias identification accuracy.
In addition, many witnesses feel strong pressure to make a pos-
itive identification, whether from law enforcement officials, con-
cerned friends or family members, or themselves (e.g., a desire
to be a “good witness” and help the police). Such pressures
increase the chance of an erroneous identification any time that
the criminal is not in the lineup.
The fairness of a lineup can be tested empirically. To do so,
individuals who had never viewed the suspect, or had only read
a description of the suspect, are shown the lineup and asked to
guess who the suspect is. If a lineup is a fair one (constructed
based upon the standard above), the frequency of correct
guesses by these individuals should be no more than what might
be expected by chance (e.g., one in six, or seventeen percent, for
a six-person lineup). By utilizing this technique under both lab-
oratory and case-specific con-
ditions, researchers have devel-
oped several measures of lineup
size (a measure based upon the
premise that a lineup should
have enough suitable members
to ensure that the probability
of a chance identification of an
innocent suspect is low), and
lineup bias (a measure estimat-
ing the degree to which the suspect is distinctive in appearance
in the lineup) that can be used to evaluate the fairness of a given
lineup. We have asserted that the bias measure is most impor-
tant and proposed a reasonable standard for estimating lineup
fairness that involves combining the size and bias estimates to
create an overall lineup fairness index.28
Recent research studies have found that a number of other
variables, often called “system variables,” are influential during
the construction and administration of photo lineups. Factors
that recent research has shown to be important include: (1) hav-
ing the lineup administered by someone who is unaware of
which lineup member is the suspect (also known as “double-
blind testing”); (2) the use of unbiased instructions that explicitly
state that the perpetrator “may or may not be present” in the
lineup, and that the witness may elect to select no one from the
lineup; (3) the use of sequential lineups, in which the eyewitness
views one photo at a time and makes an identification decision
before viewing the next photo, since research indicates that
there are fewer “false alarms,” i.e., erroneous identifications with
this method than with the typical simultaneous lineup, in which
the eyewitness views all photos at the same time; and (4) the
importance of videotaping the entire identification process so
that independent evaluations of the procedures can be made
later, such as in expert testimony.28a
What happens after the eyewitness makes a positive identifi-
cation can also be important. Recent research has shown that if
the eyewitness is told, immediately following the lineup admin-
istration, that he or she correctly identified the suspect, two
results can occur, one obvious and one more subtle. First, not
surprisingly, the eyewitness becomes more confident in the
accuracy of the identification. Second, though, the feedback
also is likely to change his or her memory for the crime itself.
The witness is likely to remember that he or she saw the crimi-
nal longer, and under better viewing conditions, than previously
reported.29
[I]t is critical that
the suspect is
not a distinctive
member of the
photo lineup.
16 Court Review - Summer 1999
30. Marin, et al., The Potential of Children as Eyewitnesses, 3 LAW &
HUMAN BEH., 295 (1979); Pozzulo & Lindsay, Identification
Accuracy of Children versus Adults: A Meta-analysis, 22 LAW &
HUMAN BEH. 549 (1998).
31. Yarmey, The Elderly Witness, in PSYCHOLOGICAL ISSUES IN
EYEWITNESS IDENTIFICATION,259 (1996); Goodman & Reed, Age
Differences in Eyewitness Testimony, 10 LAW & HUMAN BEH. 317
(1986); for a review, see Schooler & Loftus,Multiple Mechanisms
Mediate Individual Differences in Eyewitness Accuracy and
Suggestibility, in MECHANISMS OF EVERYDAY COGNITION, 177 (1993).
32. For a review, see Loftus, et al., Who Remembers What? Gender
Differences in Memory, 26 MICH. QTRLY. REV. 64 (1987).
33. Yuille, Research and Teaching with Police: A Canadian Example, 33
INT. REV. APPLIED PSYCH. 5 (1984).
34. Lavrakas, et al., A Perspective on the Recognition of Other Race
Faces, 20 PERCEPTION & PSYCHOPHYSICS 475 (1976); Malpass,
Training in Face Recognition., in PERCEIVING AND REMEMBERING
FACES (Davies, et al., eds., 1981); Malpass, et al., Verbal and Visual
Training in Face Recognition, 14 PERCEPTION & PSYCHOPHYSICS 330
(1973); Woodhead, et al., On Training People to Recognize Faces,
22 ERGONOMICS 333 (1979).
35. Hosie & Milne, The Effect of Experimental Design on Memory for
Typical and Distinctive Faces, 4 MEMORY 175 (1996).
36. Bothwell, et al., Cross-racial Identifications, 15 PERSONALITY &
SOC. PSYCH. BULL. 19 (1989); Brigham & Barkowitz, Do “They All
Look Alike?” The Effect of Race, Sex, Experience and Attitudes on the
Ability to Recognize Faces, 8 J. APPLIED SOC. PSYCH 306 (1978).
37. Chance & Goldstein, The Other-race Effect and Eyewitness
Identification, in PSYCHOLOGICAL ISSUES IN EYEWITNESS
IDENTIFICATION 153 (1996).
38. Kassin, et al., The “General Acceptance” of Psychological Research
on Eyewitness Testimony: A Survey of the Experts, 44 AMER. PSYCH.
1089 (1989); Yarmey & Jones, Is the Psychology of Eyewitness
Identification a Matter of Common Sense?, in EVALUATING WITNESS
EVIDENCE: RECENT PSYCHOLOGICAL RESEARCH AND NEW
PERSPECTIVES (1983).
39. Brigham, Disputed Eyewitness Identifications: Can Experts Help?, 8
THE CHAMPION 10 (1989); Leippe, The Case for Expert Testimony
about Eyewitness Memory, 1 PSYCH., PUBLIC POLICY, & LAW 909
(1995).
Characteristics of the
Eyewitness and Suspect
Several characteristics of the
eyewitness can influence the
accuracy of his or her mem-
ory and subsequent identifi-
cation of the suspect. For
example, age is important.
Studies have shown that
although children tend to
recall less information when
compared with adults, the
standard proportion of correct information recalled does not
typically differ between the two populations.30 Overall, elderly
adults also tend to perform more poorly than do younger
adults. Of most importance, children and the elderly also
appear to be more susceptible to the effects of suggestive ques-
tioning or post-event misinformation.31 While children may
demonstrate this effect due to their unwillingness to challenge
an adult’s authority, the elderly appear more likely to forget the
source of where they previously learned the (mis)information.
Other demographic variables that have been investigated
include gender and occupation of the eyewitness. It appears
that men and women may differ in the type of information they
recall about an event (e.g., female-oriented items such as cloth-
ing vs. male-oriented items such as a type of car).32 With regard
to occupation, it is a common assumption that law enforcement
officials will be better at identifying faces than will citizens.
However, research has failed to support this assumption, find-
ing that officers perform no better in identifying the faces of
perpetrators than do laypersons. Studies have also shown that
officers are able to provide more detailed accounts of the event,
and that they are less susceptible to the effects of post-event
misinformation when compared to laypersons.33 Empirical
studies attempting to train individuals to remember events and
faces have demonstrated a similar pattern of results; namely,
individuals’ accuracy can be improved for recalling details of an
event, but not for identification of faces.34
With regard to characteristics of the suspect, research has
primarily focused on the perceived typicality of the face. The
presence of unusual attributes that make a face distinctive from
other faces (e.g., Cindy Crawford’s mole, Jay Leno’s chin,
Sylvester Stallone’s droopy eyes) also make it easier to remem-
ber. But such distinctive characteristics also appear to make it
more difficult to construct a fair lineup, due to the difficulty in
finding other individuals with similar distinctive features.
Alternatively, faces that are more typical in appearance are sig-
nificantly more difficult to later recall or to identify from a
photo lineup, and often result in a higher likelihood of false
identification.35 A recent example of this phenomenon involves
the extensive FBI search for Andrew Cunanan, the individual
believed to have murdered fashion designer Gianni Versace in
Miami, Florida. Cunanan had a very typical-looking face that
resulted in thousands of false reported sightings across the
nation.
Interactions between Characteristics of the
Eyewitness and Suspect
Certain characteristics of the eyewitness and the suspect can
also interact to influence identification accuracy. The most
commonly cited example of such an interaction involves the
own-race bias in face recognition.36 This robust phenomenon
reflects the finding that recognition memory tends to be better
for faces of one’s own race than for faces of other races. As
Chance and Goldstein37 noted, “Few psychological findings are
so easy to duplicate.” Furthermore, researchers have endorsed
the importance of the effect in a variety of surveys,38 and expert
witnesses have widely cited its influence in testimony on dis-
puted cross-race identifications.39 Although the mechanisms
responsible for the effect have not been isolated, current
research is examining various aspects of cross-race experience
and its possible influence on the manner in which individuals
attempt to remember same- and other-race faces.
A second example of such an interaction between the eye-
It appears that
men and women
may differ in
the type of
information they
recall about an
event.
Summer 1999 - Court Review 17
40. Jalbert & Getting, Racial and Gender Issues in Facial Recognition,
in PSYCHOLOGY AND LAW: INTERNATIONAL PERSPECTIVES 309 (1992);
Mason, Age and Gender as Factors in Facial Recognition and
Identification, 12 EXP. AGING RESEARCH 151 (1986); Slone,
Brigham, & Meissner, Social and Cognitive Factors Affecting the
Own-race Bias in Whites, BASIC & APPLIED SOC. PSYCH. (in press).
41. Lipton, On the Psychology of Eyewitness Testimony, 66 J. APPLIED
PSYCH.79 (1977); see also Shapiro & Penrod, Meta-analysis of
Facial Identification Studies, 100 PSYCH. BULL. 139 (1986).
42. 388 U.S. 218 (1967).
43. 388 U.S. 263 (1967).
44. 388 U.S. 293 (1967).
45. 406 U.S. 682 (1972).
46. 413 U.S. 300 (1973).
47. 409 U.S. 188 (1972).
48. 432 U.S. 98 (1977).
49. 409 U.S. at 199-200.
50. 432 U.S. at 114.
51. See generally Cutler & Penrod; Loftus; Ross, et al., supra note 1.
in Neil v. Biggers47 and
Manson v. Brathwaite.48 In
evaluating the admissibility
of the identification, the
Court considered whether,
under the totality of circum-
stances, the identification
was reliable, even though the
confrontation procedure may
have been suggestive. The
Court established in Neil49
and reaffirmed in Manson50
five factors that should be taken into account in evaluating the
reliability of an identification: (1) the witness’s opportunity to
view the criminal during the crime; (2) the length of time
between the crime and the subsequent identification (retention
interval); (3) the level of certainty demonstrated by the witness
at the identification; (4) the (apparent) accuracy of the wit-
ness’s prior description of the criminal; and (5) the witness’s
degree of attention during the crime.
In the Neil and Manson cases, the Court’s emphasis appeared
to shift from a concern with suggestivity, as demonstrated in the
Wade, Gilbert, and Stovall decisions, to an overriding concern
with the reliability of an identification, even if it was obtained
under suggestive circumstances. At the time of the Neil deci-
sion, little published scientific research on eyewitness memory
existed. The Supreme Court could, therefore, make only “edu-
cated guesses” about the factors that might affect eyewitness
accuracy. However, scientific research conducted in the ensu-
ing years permits a systematic evaluation of the validity of the
five criteria enumerated by the Court. We will briefly examine
the validity of each of the five factors that the Court believed
were related to eyewitness accuracy, as established by subse-
quent empirical research.
Research findings indicate that only two of the five Neil fac-
tors are clearly related to accuracy in the way that the Court
assumed. First, as the Court suggested and as we noted earlier,
witnesses with a better opportunity to observe the criminal
(e.g., better lighting, closer view, longer viewing time) are more
likely to make accurate identifications.51 (But recall also that
being told that one’s identification was “correct” can signifi-
cantly bias one’s memory for how good the opportunity to
observe was.) Second, the length of the retention interval (i.e.,
the time between the crime and the identification) is generally
related to accuracy, with longer retention intervals yielding
poorer accuracy. But research shows that this relationship is
not always simple. Other factors such as race or stress may
witness and the suspect involves what has been termed the
own-sex bias in face recognition for women.40 Several studies
have demonstrated that female participants tend to outperform
male participants in remembering female faces. Curiously,
though, male and female participants do not consistently differ
in their ability to remember male faces.41
Some Research-Based Conclusions on
Eyewitness Memory
Overall, the extensive research on eyewitness memory in
recent decades has demonstrated the great range of instances in
which an erroneous identification of the suspect might occur.
Face recognition is an inherently difficult task under the most
optimal conditions. When factors at the crime scene distract
the attention and cognitive capacities of the eyewitness, or
when questioning or lineup procedures used by law enforce-
ment officials are overly suggestive, the difficulty of this task
increases immensely. Most researchers and memory experts
would agree that the “weight” assigned to eyewitness evidence
should be viewed with great caution. Given the known prob-
lems with its accuracy, the most appropriate use of a positive
eyewitness identification is not as definitive evidence of guilt,
but rather as an indication to law enforcement officials of a
potentially valuable direction in which to search for more reli-
able forms of forensic evidence.
III. U. S. SUPREME COURT RULINGS ON
EYEWITNESS EVIDENCE
The first U. S. Supreme Court decisions that specifically
addressed eyewitness evidence issues were a trio of 1967 cases:
United States v. Wade,42 Gilbert v. California,43 and Stovall v.
Denno.44 The gist of these legal rulings was to determine rights
to counsel during identification proceedings, standards regard-
ing suggestibility within identification procedures, and laws
regarding in-court identifications if the original identification
procedure was determined to be highly suggestive. The Wade
decision granted a suspect the right to an attorney during a live
lineup. However, five years later the Supreme Court reversed
the Wade ruling in Kirby v. Illinois,45 limiting the right to coun-
sel only after the initiation of criminal proceedings. Finally, in
United States v. Ash,46 the Supreme Court ruled that there is no
right to counsel at any photographic identification procedures.
It was believed that since a photo lineup could be reconstructed
and subsequently analyzed for suggestivity, counsel was not
necessary at the time of the identification.
The U. S. Supreme Court addressed the admissibility of eye-
witness identification obtained under suggestive circumstances
Face recognition
is an inherently
difficult task
under the most
optimal
conditions.
52. See generally id.
53. Bothwell, et al., Correlation of Eyewitness Accuracy and
Confidence: Optimality Hypothesis Revisited, 72 J. APPLIED PSYCH.
691 (1987).
54. Lindsay, Read, & Sharma, Accuracy and Confidence in Person
Identification: The Relationship Is Srong When Witnessing
Conditions Vary Widely, 9 PSYCH. SCIENCE 215 (1998); Read, et al.,
The Relationship Between Accuracy and Confidence in Eyewitness
Identification Studies: Is the Conclusion Changing?, in EYEWITNESS
MEMORY: THEORETICAL AND APPLIED PERSPECTIVES 107 (1997).
55. Hastie, et al., Eyewitness Testimony: The Dangers of Guessing, 19
JURIMETRICS J. 1 (1978); Wells, et al., The Tractability of Eyewitness
Confidence and its Implications for Triers of Fact, 66 J. APPLIED SOC.
PSYCH.688 (1981); Wells & Bradfield, supra note 29.
56. See Piggott & Brigham, Relationship Between Accuracy of Prior
Description and Facial Recognition, 70 J. APPLIED PSYCH. 547
(1985); Pigott, et al., A Field Study of the Relationship Between
Description Accuracy and Identification Accuracy, 17 J. POLICE
SCIENCE & ADMIN.84 (1990); Wells, Verbal Descriptions of Faces
from Memory: Are They Diagnostic of Identification Accuracy?, 70 J.
APPLIED PSYCH.619 (1985) .
57. Meissner, supra note 22.
58. Deffenbacher, The Influence of Arousal on Reliability of Testimony,
in EVALUATING WITNESS EVIDENCE 235-251 (1983).
59. United States v. Jordan, 924 F. Supp. 443 (W.D. N.Y. 1996); United
States v. Stevens, 935 F. 2d 1380 (3d Cir. 1991); State v. Chapple,
660 P. 2d 1208 (Ariz. 1983); People v. Campbell, 847 P. 2d 228
(Colo. App. 1992); People v. McDonald, 690 P. 2d 709 (Cal.
1984); Weatherred v. State, 963 S.W. 2d 115 (Tex. Ct. App. 1998).
60. See United States v. Norwood, 939 F. Supp. 1132 (D.N.J. 1996);
People v. Jackson, 210 Cal. Rptr. 680, 164 Cal App. 3d 224
(1985), People v. Lewis, 520 N.Y.S. 2d 125 (N.Y. Co. Ct. 1987).
61. 612 N.E. 2d 124, (Ind. Ct. App. 1993).
interact with the length of the
retention level and affect eye-
witness accuracy.52
The importance of the other
three factors for estimating
accuracy — witness certainty,
description accuracy, and
degree of attention — have
received mixed support by
researchers. With respect to
witness certainty, results of
thirty-five staged-event stud-
ies showed that there is only a very weak relationship between
witnesses’ degree of certainty and identification accuracy.53
However, several recent studies have demonstrated that when
witnessing conditions are varied to make later identification eas-
ier or more difficult (e.g., by shortening or extending the time of
encoding), a rather substantial relationship can be found
between identification accuracy and confidence of the eyewit-
ness.54 To further complicate things, research has demonstrated
that witnesses may become more certain of the identification as
time passes. Such “confidence hardening” is likely to occur
whenever people have publicly committed themselves to their
identification, or when they are told that their identification was
correct.55 Overall, then, one’s self-reported confidence in the
accuracy of their identification, especially when it is given a con-
siderable time after the identification was made, is not a good
indicator of accuracy.
With respect to the apparent quality of a witness’s initial
description of the suspect, the Court opined that the accuracy of
that description would be related to the probable accuracy of the
identification. But contrary to the Court’s assumption, research
has consistently demonstrated that accuracy of description is not
generally related to accuracy of identification. Further, the
apparent “completeness” of a description, the number of attrib-
utes that are recalled, also is not related to identification accu-
racy.56 However, there may be one aspect of descriptions that is
related to identification accuracy. A recent series of studies in
our lab showed that when one looks only at the number of incor-
rect facial features that are recalled, a significant relationship
between this aspect of description accuracy and identification
accuracy actually does exist.57 To clarify, when people generate
incorrect features while giving a description, this appears to lead
to later misidentifications. (Unfortunately, this finding is not
particularly helpful to law enforcement investigators, since in an
actual case one can never be sure which described features are
inaccurate descriptions of the perpetrator.) Additional research
will be valuable to further specify the precise relationship
between characteristics of an eyewitness’s description and the
accuracy of his or her later identification of the suspect.
Considering the final Neil factor, degree of attention, research
has found that eyewitnesses who pay a moderate degree of atten-
tion to a situation are likely to be more accurate when compared
to those who did not pay attention, or to those who were dis-
tracted because they were in a stressful crime situation.58 Even
if someone is trying to be attentive, high fear or stress (if pre-
sent) is likely to interfere with memory and impair the accuracy
of subsequent identifications. The perceptual situation is made
even more difficult if a weapon is involved, because the per-
ceiver is likely to focus his or her attention on the weapon
(weapon focus) rather than on the face of the person holding the
weapon. As a consequence, the person does not acquire a strong
representation of the suspect in memory.
In several recent cases, courts have held that it was reversible
error not to have allowed expert testimony pertaining to several
factors outlined in the Neil decision, in addition to other factors
found to influence eyewitness memory. Specifically, a few cases
addressed the research finding that there is not a scientifically
significant correlation between confidence and accuracy.59 The
courts felt this information was particularly relevant not only
because the research directly refuted one of the criteria laid out
in the Neil decision, but also because many jurors believe the
opposite to be true: they believe that a strong sense of confi-
dence portends great accuracy. Other cases have addressed fac-
tors such as the lack of a significant correlation between descrip-
tion accuracy and identification accuracy, and the effects of
stress or weapons on one’s ability to remember details of the per-
petrator.60
While some cases have utilized the Neil criteria as a basis for
admitting eyewitness expert testimony, the reverse was true in
Farrel v. State.61 There, the court decided to exclude eyewitness
expert testimony because the victim had both adequate lighting
[O]ne's self-
reported
confidence in the
accuracy of ...
identification ...
is not a good
indicator ...
18 Court Review - Summer 1999
62. 293 F. 1013 (D.C. Cir. 1923).
63. 509 U.S. 579 (1993).
64. See Cutler & Penrod, supra note 1.
65. 509 U.S. at 594-95.
66. 119 S. Ct. 1167 (1999).
67. 488 F. 2d 1148 (9th Cir. 1973).
68. See United States v. Brown, 501 F. 2d 146 (9th Cir. 1974); United
States v. Brown, 540 F. 2d 1048 (10th Cir. 1976).
and an extended period of time to view the defendant’s appear-
ance. In addition, the eyewitness provided police with a rather
detailed description of the defendant immediately following the
incident.
IV. CASE LAW PERTAINING TO EXPERT
TESTIMONY ABOUT EYEWITNESS EVIDENCE
The Admissibility of Scientific Expert Testimony
Generally
Concerned about the conviction of innocent persons on the
basis of erroneous eyewitness identifications, the courts have
struggled in recent years to balance the rights of defendants
threatened by the specter of incorrect eyewitness identification
with the need to prosecute cases based upon disputed eyewit-
ness identification evidence. As summarized above, thousands
of empirical studies have investigated factors that may affect the
reliability of eyewitness identification. One could argue that the
wealth of general scientific information that these studies have
yielded might be very helpful to decision-makers whose cases
involve eyewitness identification. However, this information
has not been readily accepted by the court system. The intro-
duction of any new type of expert evidence is typically met with
skepticism and challenge, and not necessarily for imprudent rea-
sons. But does the exclusion of expert testimony pertaining to
eyewitness identification, a topic in which common beliefs are
not always accurate, add to the problem of innocent persons
being convicted? Below we will address several of the issues that
have faced the courts, and describe how they have been handled
through the years.
There is a long history pertaining to the admissibility of
expert scientific testimony. The leading case on the admissibil-
ity of “novel” scientific evidence is Frye v. United States.62 The
Frye test is premised on the “general acceptability” rule, in
which the scientific evidence to be presented to the jury must be
considered good science, i.e., generally accepted within the rel-
evant scientific community. The purpose of the Frye test was to
screen out unreliable scientific evidence.
In the recent landmark case of Daubert v. Merrell Dow
Pharmaceuticals, Inc.63 the U.S. Supreme Court found that “gen-
eral acceptance,” as stated under the Frye test, was not a neces-
sary precondition to the admissibility of scientific evidence under
Rule 702 of the Federal Rules of Evidence. The Court opined
that the Rule 702 assigned to the trial judge the tasks of ensuring
that the expert’s testimony was both reliable and relevant to the
case at hand, and that the expert is proposing to testify to scien-
tific knowledge that would assist the trier of fact to understand
or determine a fact in issue. This “helpfulness” standard of the
Daubert ruling has been seen as less stringent than the Frye test
for determining the admissibility of expert testimony.64
In explicitly rejecting the Frye test, Justice Blackmun wrote
for the unanimous majority in Daubert that “a rigid ‘general
acceptance’ requirement would be at odds with the ‘liberal’
thrust of the Federal Rules and their general approach of relax-
ing the traditional barriers to
‘opinion testimony.’” The Court
stressed that the “overarching
subject is the scientific validity”
of the research in question, rather
than its general acceptance
within the relevant scientific
community.65 Thus, trial judges
were assigned the role of gate-
keeper, whose task is to decide, in
effect, whether the proposed tes-
timony represents methodologi-
cally sound research or is “junk science.” In Kumho Tire v.
Carmichael,66 the Supreme Court recently reaffirmed this aspect
of the Daubert decision, ruling that trial judges should be
granted broad latitude in determining which factors are applied
in assessing the reliability of a given expert’s testimony. The
Court also extended Rule 702 to include all expert testimony, be
it “scientific,” “technical,” or “other specialized” knowledge.
We should note that the scientific research on factors that
affect eyewitness accuracy, which we have very briefly reviewed
above, most certainly would meet any reasonable criterion of
“good science.” The research is published in highly selective,
peer-reviewed scientific journals, most of which reject (usually
on methodological grounds) about eighty percent of the manu-
scripts that are submitted to them.
The Admissibility of Expert Testimony on
Eyewitness Evidence
While Frye and Daubert deal with expert testimony in gen-
eral, our focus is on expert testimony pertaining specifically to
the reliability of eyewitness identification. There have been
both federal and state court decisions dealing with the admissi-
bility of expert testimony on this subject. In many circum-
stances, due to contradictory rulings, the overall determination
of whether expert testimony on this subject is admissible has
not been definitively answered.
Two decades before the Daubert ruling, the issue of helpful-
ness had been raised in United States v. Amaral.67 The main
inquiry in this case pertaining to the admissibility of eyewit-
ness expert testimony was whether the jury would receive
“appreciable help” from the proffered expert testimony. Four
guidelines were set out in Amaral to determine the helpfulness
of expert testimony: (1) whether the expert is deemed quali-
fied; (2) whether the testimony proffered is a proper subject
for expert testimony, meaning that it will provide information
that is not already part of jurors’ common knowledge and will
not invade the province of the jury; (3) whether the testimony
given conforms to a generally accepted explanatory theory;
and (4) whether the probative value of the testimony out-
weighs its possible prejudicial effect. Most decisions based on
the Amaral ruling have come down against the admittance of
expert testimony on the reliability of eyewitnesses.68
Three general types of appellate decisions have predominated
[T]his
information
has not been
readily
accepted by
the court
system.
Summer 1999 - Court Review 19
69. 50 F. 3d 1335 (5th Cir. 1995).
70. 660 So. 2d 341 (Fla. 1998).
71. See State v. Chapple, 660 P. 2d 1208 (Ariz., 1983); People v.
McDonald, 690 P. 2d 709 (Cal., 1984); State v. Moon, 726 P.2d
1263 (Wash. App. 1986); United States v. Stevens 935 F.2d 1380,
1400-01 (3d. Cir. 1991).
72. 631 N.Y.S. 2d 926 (N.Y. App. Div. 1995). See also United States
v. Purham, 725 F. 2d 450 (8th Cir. 1984); United States v. Larkin,
978 F. 2d 964 (7th Cir. 1992); Johnson v. State, 393 So. 2d 1069
(Fla. 1980).
73. See Amarl, 488 F.2d 1148;United States v. Daniels, 64 F. 3d 311
(7th Cir. 1995); State v. Kemp, 507 A. 2d 1387 (Conn. 1986);
People v. Anderson, 630 N.Y.S. 2d 77 (N.Y. App. Div. 1995).
74. State v. Chapple, 660 P. 2d 1208 (Ariz. 1983); People v.
McDonald, 690 P. 2d 709 (Cal. 1984).
75. United States v. Watson, 587 F. 2d 365, (7th Cir. 1978); United
States v. Downing, 753 F. 2d 1224 (3rd Cir. 1985).
76. See United States v. Downing, 753 F. 2d 1224 (3rd Cir. 1985);
United States v. Jordan, 924 F. Supp. 443 (W.D. N.Y. 1996); State
v. Moon, 726 P. 2d 1263 (Wash. App. 1986); Skamarocius v. State,
731 P. 2d 63 (Alaska App. 1987).
77. See United States v. Moore, 786 F. 2d 1308 (5th Cir. 1986); State
v. Buell, 489 N.E. 2d 795 (Ohio 1986).
78. Devenport, et al., Eyewitness Identification Evidence: Evaluating
Commonsense Evaluations, 3 PSYCH., PUB. POLICY, & LAW 338
(1997).
the case law pertaining to
expert testimony on eyewit-
ness evidence. First, some
courts, especially in decisions
before the 1980s, ruled that
expert testimony about eye-
witness memory was per se
inadmissible. A reason some-
times given for this decision
was that there was not a suffi-
ciently large body of research
on which to base scientific
expert testimony (a violation of the third Amaral criterion).
This would have been a legitimate concern up until the late
1970s, by which time a substantial amount of good eyewitness
research had been published. Judicial opinions also sometimes
said that expert testimony on this issue would invade the
province of the jury to evaluate evidence (the second point in
Amaral), or that information regarding factors affecting eyewit-
ness identification was part of jurors’ common knowledge,
thereby constituting an improper subject matter for an expert
(also part of the second point in Amaral).
A second approach, the most common appellate view in
recent years, is for the trial judge to use his or her discretion in
admitting or excluding such expert testimony. In coming to
these decisions, some appellate courts have expressed pes-
simism that the expert testimony will be of assistance, while
other decisions have noted that under some conditions such
expert testimony may be relevant and helpful. For example, in
United States v. Jackson,69 it was decided that the court could
have, as a matter within its discretion, admitted expert testi-
mony on eyewitness reliability if such testimony had been
offered. Similarly, in McMullen v. State,70 the court held that
when the sole issue in a criminal case is one of identity and the
sole incriminating evidence is eyewitness testimony, the admis-
sion of expert testimony upon factors that affect the reliability
of eyewitness identification is within the discretion of the trial
judge. To us, this situation represents one in which expert tes-
timony is the most important and would make the greatest con-
tribution.
Finally, a third set of decisions has ruled that the exclusion
of expert testimony about eyewitness evidence constitutes a
reversible error by the trial court.71 Again, courts have ruled
this way mostly when the sole evidence against a defendant has
been the eyewitness identification. Below, we review decisions
that have established criteria for the admission or exclusion of
expert testimony about factors that affect the accuracy of eye-
witness memory.
Case Law and Research on Jurors’ “Common
Knowledge”
The belief that factors affecting the reliability of eyewitness
identification are common knowledge to the lay juror has often
been cited as a reason for the exclusion of expert testimony on
this matter. For example, in People v. Kelly,72 the court felt that
the reliability of eyewitness identification was not beyond the
ken of the average juror. It is often believed, therefore, that
admittance of opinion and expert testimony on information
already known to the jury is a usurpation of the jury’s fact-find-
ing province. Several court rulings, both federal and state, have
opined that the introduction of expert testimony on eyewitness
reliability would, in fact, invade the province of the jury.73
However, there have also been cases in which courts have ruled
otherwise, suggesting either that the admission of the expert
testimony did not invade the province of the jury,74 or that,
although the expert testimony may invade the province of the
jury, the Federal Rules of Evidence do not preclude its admis-
sion into testimony.75 The basis for this last caveat is that the
jury has the wherewithal to accept or reject the expert opinion
and afford it the weight it deems appropriate. As noted previ-
ously, admittance of expert testimony on eyewitness reliability
has been considered especially important when the case against
the defendant rests solely on eyewitness identification, and no
other physical evidence exists.76 However, when other physical
evidence is available (e.g., fingerprints, DNA), the exclusion of
expert testimony on eyewitness reliability has often been con-
sidered harmless error.77
Since many legal decisions have been based on the notion
that ideas and testimony proffered by an expert witness are
already common knowledge to jurors, it seemed an important
task for researchers to determine exactly what the lay person
knows about factors affecting eyewitnesses. There have been
three basic methodologies used to investigate this information:
(1) surveying jury eligible citizens as to their knowledge and
beliefs; (2) assessing jurors’ ability to predict the outcome in an
eyewitness identification experiment; and (3) using mock trials
to assess the influence of trial techniques such as cross-exami-
nation.78
Survey studies, conducted by administering questionnaires
such as the Knowledge of Eyewitness Behavior Questionnaire
[T]he most common
appellate view ... is
for the trial judge
to use... discretion
in admitting or
excluding such
expert testimony.
20 Court Review - Summer 1999
79. Deffenbacher & Loftus, Do Jurors Share a Common Understanding
Concerning Eyewitness Behavior?, 6L
AW & HUMAN BEH. 15 (1982).
80. Brigham & Bothwell, The Ability of Prospective Jurors to Estimate
the Accuracy of Eyewitness Identifications, 7 LAW & HUMAN BEH. 19
(1983).
81. Wells, et al., Accuracy, Confidence, and Juror Perceptions in
Eyewitness Identification, 64 J. APPLIED PSYCH. 440, (1979). See
also Lindsay, et al., Can People Detect Eyewitness Identification
Accuracy Within and Across Situations?, 66 J. APPLIED PSYCH. 79
(1981).
82. Lindsay, et al., Mock-juror Evaluations of Eyewitness Testimony: A
Test of Metamemory Hypotheses, 16 J. APPLIED SOC. PSYCH. 447
(1986).
83. Cutler, et al., Jury Decision Making in Eyewitness Identification
Cases, 12 LAW & HUMAN BEH. 41 (1988).
84. Cutler, et al., Juror Sensitivity to Eyewitness Identification
Evidence, 14 LAW & HUMAN BEH. 185 (1990).
85. DEVENPORT, ET AL., DOES EXPERT PSYCHOLOGICAL TESTIMONY
IMPROVE JUROR SENSITIVITY TO LINEUP SUGGESTIVENESS? (unpu-
bished manuscript 1996).
86. See United States v. Norwood, 939 F. Supp. 1132 (D.N.J. 1996);
People v. Jackson, 210 Cal. Rptr. 680 (Cal. App. 1985), People v.
Lewis, 520 N.Y.S. 2d 125 (N.Y. Co. Ct. 1987).
(KEBQ), assess beliefs about factors that affect the accuracy and
reliability of eyewitness identification.79 The KEBQ consists of
fourteen multiple-choice scenarios describing crime scenes, dif-
fering in aspects such as retention interval, training, age of the
witness, prior photo array identification, and cross-racial iden-
tification. Results of such studies have demonstrated that while
respondents are sensitive to the influences of cross-race and
prior photo array identifications, they appear less sensitive to
the effects of age (young or old) and retention intervals on eye-
witness reliability. Furthermore, participants tended to believe,
contrary to research findings, that training could improve iden-
tification accuracy.
A second type of study commonly used is the “post-diction”
study, in which participants read written summaries of identifi-
cation experiments and then guess the accuracy rates that
occurred in the experiments. Results indicate that participants
usually predict higher accuracy rates for the original subjects
than were actually obtained, suggesting that people often
believe witnesses to be much more accurate in their judgments
than they truly are. For example, Brigham and Bothwell found
that more than eighty percent of the registered voters they stud-
ied overestimated the accuracy rate of eyewitness research sub-
jects.80 Overall, participants in post-diction studies seem to be
insensitive to the influence of crime seriousness, instruction
bias, and cross-racial identification. Additionally, contrary to
research findings, participants seem to believe that confidence
is an important variable.
Finally, researchers have used the “mock trial” as a method
for assessing jurors’ commonsense knowledge about factors
affecting eyewitness reliability. These studies involve manipu-
lating different factors known to influence eyewitness identifi-
cation accuracy (e.g., observation conditions) and those shown
to have little influence on identification accuracy (e.g., witness
confidence). Participants are typically asked to assume the role
of a juror as they are introduced to the summary of a trial via
written transcript, or audio or videotape. Finally, the partici-
pants are asked to complete questionnaires assessing their ver-
dicts. For example, Wells, Lindsay, and Ferguson81 found that
witness confidence correlated significantly with whether a juror
believed an eyewitness, even though, as noted earlier, research
has demonstrated a very weak relationship between witness
confidence and eyewitness accuracy.
Researchers have also studied the effects of individuals’
awareness of the conditions surrounding a crime scene at the
time of the identification, such as lighting, time of day, and
duration of viewing time. For example, in one study
researchers created a tran-
script stating that the crime
occurred at either 9 a.m.
on a sunny day, or at 1
a.m., sixty feet from the
closest street light.
Further, the length of time
the witness saw the event
was varied between five
seconds and thirty min-
utes. Results demonstrated that jurors displayed a lack of sen-
sitivity to witnessing conditions that may affect identification
accuracy, as the conviction rates for groups of subjects who
heard about the different conditions did not differ statisti-
cally.82
Cutler, Penrod, and Stuve83 investigated the commonsense
knowledge of jurors by manipulating ten factors known to
influence eyewitness accuracy to varying degrees. Variables
studied included the presence or absence of a weapon,
whether the perpetrator was wearing a disguise, and whether
the crime was violent. Furthermore, the length of the reten-
tion interval, the presence or absence of instruction bias and
foil bias during identification, and the level of the witness’s
confidence were all manipulated. Unfortunately, results
demonstrated that jurors seemed insensitive to the factors that
should have called the validity of the identification into ques-
tion. Additionally, participants relied heavily on the expres-
sions of confidence from the eyewitness. A follow-up study
demonstrated that college students and jury-eligible citizens
were equally insensitive to these important factors.84
Researchers have also examined whether jurors can identify
factors that would render a lineup biased. Findings suggest
that while jurors do have the commonsense knowledge in
identifying foil and instruction bias, they have difficulty in
applying this knowledge as demonstrated by their verdicts.85
Several courts have decided that these sorts of research find-
ing are not necessarily within the jurors’ common knowledge,
and that the jury might be missing out on information that
might assist them in determining the facts at issue, especially
when the eyewitness identification was the sole evidence
against the defendant.86
Probative v. Prejudicial Value of the Expert
Testimony
As of 1990, psychologists had testified as expert witnesses
on the reliability of eyewitness evidence in more than 450 cases
[P]eople often
believe witnesses
to be much more
accurate ... than
they truly are.
Summer 1999 - Court Review 21
87. FULERO, EYEWITNESS EXPERT TESTIMONY: ANOVERVIEW AND
ANNOTATED BIBLIOGRAPHY, 1931-1988 (unpublished manuscript
1993).
88. See United States v. Collins, 395 F. Supp. 629 (M.D. Pa. 1975);
United States v. Fosher, 590 F. 2d 381 (1st Cir. 1978); United
States v. Watson, 587 F. 2d 365 (7th Cir. 1978); United States v.
Thevis, 556 F. 2d 616 (5th Cir. 1982).
89. Hosch, et al., Influence of Expert Testimony Regarding Eyewitness
Accuracy on Jury Decisions, 4LAW & HUMAN BEH.287 (1980);
Loftus, Imact of Expert Testimony on the Unreliability of Eyewitness
Identification, 65 J. APPLIED PSYCH.9 (1980); Maass, et al.,
Testifying on Eyewitness Reliability: Expert Advice Is Not Always
Persuasive, 15 J. APPLIED SOC. PSYCH. 207 (1985).
90. 833 F. 2d 1296, (9th Cir. 1987).
91. 208 S.E. 2d 850 (Ga. 1974).
92. 122 F. 3d 1355 (11th Cir. 1997).
in 25 states.87 Unlike clinical
expert testimony on factors
such as insanity, competency
to stand trial, or child cus-
tody, eyewitness experts typi-
cally do not seek to interview
or analyze the individual eye-
witness. Rather, the expert
serves an educational func-
tion for the jury, presenting
the general factors that increase or decrease the likelihood that
the average eyewitness will be correct in particular situations.
One could suggest that this is like being “a tutor for the jury,”
that the eyewitness expert resembles in some ways a judge giv-
ing jury instructions. Some courts have ruled that such a role
is unnecessary or inappropriate, stating that jurors’ everyday
“commonsense knowledge” is sufficient and hence they do not
need such “tutoring.” However, it can be argued that such help
is both necessary and appropriate, due to the unique status of
eyewitness evidence. Research has shown that issues sur-
rounding eyewitness memory are significantly more difficult
and counterintuitive than many other issues confronted by
jurors.
The function of an expert witness is not to tell the jury what
to believe or to imply that a particular witness is either correct
or incorrect. The expert does not know whether a particular
eyewitness is correct; often, even the eyewitness cannot know
this with certainty. What the expert can do is provide the jury
members with a factually based frame of reference within which
to interpret the eyewitness evidence, along with all the other
evidence, in reaching a verdict. Awareness of the error factors
most relevant to the eyewitness identification in a particular
case may cause jury members to weigh the eyewitness evidence
more heavily or, conversely, to give it less emphasis than they
otherwise would.
This brings us to the fourth component in the Amaral deci-
sion: the comparison of the probative versus prejudicial value
of expert testimony. According to the Amaral decision, the
probative value of the testimony must outweigh any prejudi-
cial effects. Many courts have felt that the “impressive creden-
tials” an expert brings to the courtroom are enough to create
an overwhelming prejudicial effect, thus compromising any
ability a juror may have to appropriately weigh the testimony
given.88 Such thinking cripples the justice system, in that it
seems to assume the jurors are unable to make decisions based
on weighing evidence when it comes from an expert, as
opposed to when it comes from any other witness.
Interestingly, it is often believed that police officers carry a
larger amount of persuasive power with juries simply because
of their authoritative position, yet their testimony is not usu-
ally considered prejudicial.
Some critics have asserted that expert evidence regarding
eyewitness reliability may adversely affect jury deliberations by
making proper convictions that much harder to obtain.
However, research with mock juries has demonstrated this to
be untrue, indicating that while jurors do become more skep-
tical after hearing expert testimony pertaining to eyewitness
reliability, it does not result in blanket skepticism. In fact,
expert testimony appears to make mock jurors more skeptical
in situations where it is appropriate: situations in which the
conditions faced by the eyewitness are greatly associated with
high error rates. Conversely, in strong cases, where other
physical evidence is also introduced, mock juries are just as
likely to convict after hearing the expert testimony than after
not hearing it. Furthermore, the introduction of expert testi-
mony seems to increase the care with which jurors analyze all
the evidence, and not just the eyewitness evidence, perhaps
encouraging a more deliberate examination of the facts in the
case.89
V. TRADITIONAL TRIAL SAFEGUARDS FOR
DEFENDANTS
Cross-examination
In justifying the refusal to allow expert testimony on the reli-
ability of eyewitness identification, courts often cite the ability
or opportunity of defense counsel to use cross-examination as
an effective tool in casting doubt upon an eyewitness’s identifi-
cation of the defendant. Traditionally, skillful cross-examina-
tion of opposing witnesses has been seen as the strongest safe-
guard against mistaken convictions and has been referenced in
case law in answer to constitutionality arguments against the
ruled inadmissibility of proffered expert testimony. For exam-
ple, in United States v. Christophe,90 skillful cross-examination of
eyewitnesses, along with jurors’ common sense and experience,
were deemed sufficient to alert jurors to specific conditions that
would render particular eyewitness identifications reliable. An
earlier case, Jones v. State,91 concluded that the defendant’s
rights of due process were not violated by excluding the prof-
fered testimony because both of the eyewitnesses were sub-
jected to considerable cross-examination concerning their
means and opportunity of observing the perpetrator. This rea-
soning was repeated more recently in United States v. Smith,92
which concluded that the proffered expert testimony regarding
the reliability of eyewitness identification was inadmissible
under the helpfulness prong of Daubert, because the jury could
[T]he expert
serves an
educational
function for the
jury.
22 Court Review - Summer 1999
93. See Dyas v. United States, 376 A. 2d 827 (D.C. App. 1977); United
States v. Larkin, 978 F. 2d 964 (7th Cir. 1992); United States v.
Langford, 802 F. 2d 1176 (9th Cir. 1986); People v. Hurley, 157
Cal. Rptr. 364 (Cal. App. 1979); Moore v. Tate, 882 F. 2d 1107
(6th Cir. 1989); Jackson v. Ylst, 921 F. 2d 882 (9th Cir. 1990);
Garth v. State, 536 So. 2d 173 (Ala. App. 1988).
94. Lindsay, et al., Mock Juror Belief of Accurate and Inaccurate
Eyewitnesses: A Replication and Extension, 13 LAW & HUMAN BEH.
333 (1989); Lindsay, et al., Can People Detect Eyewitness
Identification Accuracy Within and Across Situations?, 66 J.
APPLIED PSYCH.790 (1981); Wells, et al., Accuracy, Confidence and
Juror Perceptions in Eyewitness Identification, 64 J. APPLIED PSYCH.
440 (1979).
95. 469 F. 2d 552, 558-9 (D. C. Cir. 1979).
96. 409 U.S. 188 (1972).
97. Greene, Judge’s Instructions on Eyewitness Testimony: Evaluation
and Revision, 18 J. APPLIED PSYCH.252 (1988),
98. Cutler, et al., Nonadversarial Methods for Improving Juror
Sensitivity to Eyewitness Evidence, 20 J. APPLIED SOC. PSYCH.1197
(1990); Greene, supra note 97; Katzev & Wishart, The Impact of
Judicial Commentary Concerning Eyewitness Identifications on Jury
Decision Making, 76 J. CRIM. LAW & CRIMINOLOGY 733 (1985).
99. 660 P. 2d 1208 (Ariz. 1983).
100. 928 S.W. 2d 550 (Tex. Crim. App. 1996).
determine the reliability of the eyewitness identification with
the assistance of cross-examination.93
Contrary to these opinions, there are two general reasons
why cross-examination, no matter how skillfully conducted,
cannot be fully effective in illuminating the accuracy of eyewit-
ness evidence. First, in order to effectively cross-examine, the
attorney would need to have the opportunity to identify the fac-
tors that were likely to affect the identification, be aware of their
influence, and be able to inform the judge and jury of these
effects. This is highly unlikely, if not impossible, in most
instances. Second, lawyerly skill in questioning an eyewitness
may be insufficient to distinguish between an eyewitness who
is honestly mistaken and one who is accurate. If a witness were
lying, it is possible that cross-examination could make the lie
apparent. But when a person is telling the truth as he or she
knows it, cross-examination will not necessarily determine
accuracy. Furthermore, research studies in which eyewitnesses
have been cross-examined by experienced lawyers have shown
that mock jurors who view the cross-examination cannot dis-
tinguish accurate from inaccurate eyewitnesses.94
Cautionary Instructions to Jurors
A second traditional safeguard for defendants is the use of
cautionary instructions to jurors. Some courts have utilized
special judicial instructions about eyewitness identifications
for this purpose. Probably the most widely utilized special
instructions are those developed in 1972 by the U.S. Court of
Appeals for the District of Columbia in United States v.
Telfaire.95 These instructions focus on the previously outlined
factors listed in Neil v. Biggers.96 Although the Telfaire instruc-
tions have been employed in many cases, a survey of fifty-two
judges found that most of them (seventy-eight percent) did not
think that they were proper instructions to give to a jury.97
There have been several scientific studies of the effect of these
instructions on juror decision-making.98 Results indicated
that the instructions do not adequately enhance jurors’ sensi-
tivity to potential problems in eyewitness identification evi-
dence. Because the Telfaire instructions were developed from
legal precedents, rather than being based on scientific research
findings, they do not address several areas that research has
shown are important, such as cross-race identifications, stress,
unconscious transference, lineup bias, and weapon focus.
Additionally, the instructions emphasize witness certainty
(confidence), which is not a
strong predictor of accuracy,
according to research find-
ings. In general, then, exist-
ing cautionary judicial
instructions about eyewit-
ness evidence have two
major shortcomings: they are
seen as improper by many
judges, and they are ineffec-
tive in informing jurors
about the factors that have
been shown to affect eyewitness accuracy.
VI. ON ATTEMPTS TO INFUSE EMPIRICAL
RESEARCH FINDINGS AT TRIAL: ROADBLOCKS
TO THE ADMISSIBILITY OF EXPERT TESTIMONY
While there has been ample case law that demonstrates a
certain willingness to allow scientific expert testimony into the
courtroom, there are many cases in which the courts have ruled
otherwise. Part of the problem exists because of the inherent
differences between scientific research and the law. Whereas
legal cases are to be taken each as an individual entity, research
results are often the compilation and average of effects across
many individuals. In essence, the analogy of comparing apples
to oranges may hold true when comparing legal issues to
research issues. That is not meant to imply, however, that
research into the reliability of eyewitness testimony is not help-
ful or important.
Another problem stems from the contradictory rulings com-
ing from different state courts pertaining to scientific research
itself. For example, in State v. Chapple,99 the court ruled that
the “generality” of the psychologist’s testimony was a factor
favoring admission. In Jordan v. State,100 the court stated that
“too narrow a definition of ‘fitting’ the case goes beyond the
requirements of helpfulness under Rule 702. An expert should
not have to address every conceivable factor that might affect
eyewitness identification.” In contrast, however, much expert
testimony on eyewitness identification has been excluded in
other cases for this same reason, namely the belief that research
findings are too general, or that the testimony given would not
help the trier of fact by not enhancing the jury’s ability to
deduce whether a specific eyewitness was able to make an accu-
[E]xisting
cautionary
instructions about
eyewitness
evidence have
two major
shortcomings.
Summer 1999 - Court Review 23
101. See United States v. Fosher, 590 F. 2d 381 (1st Cir. 1978); United
States v. Downing, 609 F. Supp. 784 (3rd Cir. 1985); United
States v. Dowling, 855 F. 2d 114 (3rd Cir. 1988); People v.
Plasencia, 223 Cal Rptr. 786 (Cal. App. 1985); Pankey v.
Commonwealth, 485 S.W. 2d 513 (Ky. 1972).
102. 690 P. 2d 709 (Cal. 1984)
103. See Echavvaria v. State, 839 P. 2d 589 (Nev. 1992); State v.
Whaley, 406 S.E. 2d 369 (S.C. 1991).
104. 587 F. 2d 365 (7th Cir. 1978). See also People v. Plasencia, 223
Cal. Rptr. 680 (Cal. App. 1985).
105. 884 F. 2d 1016 (7th Cir. 1989).
106. 410 N.E. 2d 252 (Ill. App. 1980). See also State v. Lawhorn, 762
S.W. 2d 820 (Mo. 1988).
107. See notes 22, 23 and 24 supra.
108. 390 U.S. 377 (1968).
109. 736 F. 2d 1103 (6th Cir. 1984).
110. See State v. Calia, 514 P. 2d 1354 (Or. App. 1973).
111. 438 So. 2d 774, (Fla. 1983).
rate identification.101 Several
additional areas in which
expert testimony has been
proposed will be briefly
addressed below.
Cross-Racial
Identification
Many experts have prof-
fered testimony regarding the
problems associated with cross-racial or cross-cultural identifi-
cations. The basic premise of this issue, supported by research,
is that it is generally easier to recognize or identify a person of
one’s own race, than of another race. The California Supreme
Court, in People v. McDonald,102 recognized that this finding
may be contrary to most jurors’ intuitions. Two aspects of the
research findings that the court listed specifically as outside the
common knowledge of jurors are (1) that white witnesses who
are not racially prejudiced are just as likely to be mistaken in
making a cross-racial identification as those who are preju-
diced, and (2) that white witnesses who have had considerable
social contact with blacks may be no better at identifications
than those who have not.103
Contrary to the ruling in McDonald, however, much of the
case law demonstrates a belief that the findings pertaining to
cross-racial identifications are actually common knowledge. In
United States v. Watson,104 the court ruled that proffered testi-
mony on cross-racial identification was inadmissible because it
would not be of probative value to the jury. A similar ruling was
made in United States v. Hudson,105 where the court believed that
“this issue is one which the jury is already aware.” And in
People v. Dixon,106 the proffered testimony of the psychologist
mentioned that “there is some truth to the folk notions that to
whites, all blacks look alike.” Based on this comment, the court
felt that allowing the testimony to be admitted would only serve
to verify an already existing belief, and that the proffer would
not go beyond the common knowledge of the experience of the
average juror.
Unconscious Transference and Lineup Bias
The phenomenon of unconscious transference has been stud-
ied extensively, as noted earlier.107 One example of this phe-
nomenon is when an eyewitness remembers seeing a person
from a criminal incident, when he or she may actually be
remembering the face from a previous photo lineup (perhaps a
biased lineup) or some other contact. It also includes the
effects of post-event misinformation on memory.
While there is a great deal of research pertaining to problems
and biases arising from poorly constructed lineups, there is lit-
tle case law demonstrating a willingness to allow expert testi-
mony regarding this matter. The importance of the issue was
noted in Simmons v. United States,108 in which the Supreme
Court wrote that “a conviction based on an eyewitness identifi-
cation at trial following a pretrial identification by photograph
will be set aside on that ground only if the photograph identifi-
cation was so impermissibly suggestive as to give rise to a very
substantial likelihood to irreparable misidentification.” On this
basis, the court ruled in United States v. Smith109 that an expert’s
testimony should have been admitted. In Smith, the defendant
asserted that the identification from the lineup was actually a
transference made from the showing of the photospread four
months earlier. However, based on other physical evidence, the
refusal to allow the testimony was considered harmless error.110
Yet, despite cases like Simmons and Smith, few courts are
willing to allow testimony specific to biased photo lineups. For
example, in Johnson v. State,111 the defendant was the only
blond member in the lineup and the only lineup member with
a slightly, but noticeably, different color blue shirt. Still, the
court ruled against admitting expert testimony regarding the
dangers of biased lineups.
VII. CONCLUSIONS: WHAT SHOULD BE DONE?
Overall, the wealth of research on eyewitness memory has
identified a host of conditions that may increase the chances
that an erroneous identification of an innocent suspect may
occur. As a form of forensic evidence, eyewitness evidence is
severely limited in its degree of diagnosticity, its precision in dis-
tinguishing guilt from innocence. Indeed, its level of precision
and accuracy does not nearly approximate that of other scien-
tifically validated forms of physical evidence obtained from the
crime scene (e.g., fingerprints, DNA, etc.). Nevertheless, eye-
witness testimony is consistently touted by both prosecutors
and appellate courts as a valid form of evidence. Furthermore,
it is believed that jurors possess much “common knowledge”
regarding eyewitness evidence, including an awareness of its
limitations and possible inaccuracy. However, scientific
research has failed to support either of these assertions. In light
of such findings, it seems imperative that eyewitness testimony
be viewed with caution in the courtroom, and that steps be
taken to protect defendants who are being tried solely or largely
on the basis of this fallible class of evidence.
Research has demonstrated that eyewitness expert testimony
can be extremely beneficial to the judicial system for several
important reasons. First, based on the thousands of empirical
studies on memory and on the factors that influence eyewitness
perception, researchers have found that there are specific con-
[E]yewitness
expert testimony
can be
extremely
beneficial ...
24 Court Review - Summer 1999
112. See text supra at note 89 and note 89.
113. YARMEY, THE PSYCHOLOGY OF EYEWITNESS TESTIMONY 227 (1979).
ditions, situations, and personal characteristics that may cause
an identification to be inaccurate. Some of these factors include
the witnessing conditions, the presence of a weapon, stress, sug-
gestive post-event information, unconscious transference,
lineup bias, witness confidence, and cross-racial identifications.
While many courts have felt that jurors have sufficient common
knowledge to evaluate the influence of these factors without the
advent of expert testimony, research has clearly indicated that
jurors are often insensitive to the effects of many of these factors
and overly sensitive to other factors (e.g., witness confidence).
Without proper instruction on how each of these factors may
effect a witness’s perceptual ability, jurors are left to rely on their
often incorrect, intuitive beliefs about how memory works. To
be sure, our intuitive beliefs usually serve us well in dealing
with the world, providing us with a largely accurate view of how
things work. But the area of eyewitness memory has been
shown to be different. Here, these usually reliable beliefs are not
accurate enough to ensure fair treatment under the law.
Second, some courts have worried that the opinion testi-
mony from an impressive expert may carry more weight with
the jury than it should, having a prejudicial impact. But, con-
trary to this supposition, studies have concluded that defen-
dants in cases containing strong evidence against them are not
convicted less when the jurors heard expert testimony, when
compared with cases in which expert testimony was excluded.
In fact, the jurors spent more time reviewing all of the evidence
when faced with expert evidence on eyewitness reliability,
demonstrating a willingness on the part of jurors to carefully
weigh all testimony equally.112 This type of deliberate inspec-
tion of all relevant material only strengthens the jury system.
Furthermore, fewer appeals pertaining to eyewitness identifica-
tion would have to be heard.
Even with the advent of sound, reliable research methodol-
ogy, and an overall general acceptance within the scientific com-
munity at large, the wholehearted acceptance of empirical
research into the legal field has been slow in coming. The jux-
taposition of psychological research and the law is a difficult
one, as the two disciplines have vastly different foundations —
one having an individualistic focus, and the other being an
empirical, aggregated focus. However, this does not preclude
the possibility that the two can build a strong working relation-
ship in which each may benefit from the knowledge and wis-
dom of the other.
As one researcher pointed out, “What one generation of
lawyers prefer to understand as ‘common sense’ often depends
upon the theory and findings of the previous generation of [sci-
entific] investigators.”113 The challenge faced by attorneys, legal
scholars, and researchers involved with disputed eyewitness
identifications is to persuade the courts of the relevance of the
research findings and the possible benefits of infusing such
research (via expert testimony or some other means) into the
judicial system. Courts in recent years have become increas-
ingly cognizant of the benefits (and sometimes the necessity) of
utilizing scientific findings, often introduced via expert testi-
mony, for issues to which they are relevant. It seems to us that
the issue of eyewitness evidence is one for which such testi-
mony is particularly important and relevant. Indeed, such tes-
timony would only seem to strengthen the criminal justice sys-
tem by reducing the frequency with which innocent persons,
victims of the frailties of human perception and memory, have
the misfortune of being falsely accused and erroneously con-
victed of a crime.
John C. Brigham is a professor of psychology
at Florida State University, where he has
taught since 1969, attaining the rank of full
professor in 1981. He has been involved in
the study of eyewitness identification for
more than two decades, having written
numerous published articles and having pro-
vided expert testimony in several states. In
addition to his own writing, including books
on social psychology, he serves as a consult-
ing editor and manuscript reviewer for several publications,
including Law and Human Behavior. Dr. Brigham, who earned
his Ph. D. in psychology from the University of Colorado in 1969,
served in 1997-98 as the national president of the American
Psychology-Law Society.
Adina W. Wasserman, a graduate research
assistant to Dr. Brigham at Florida State
University when this article was written,
received her master’s degree in 1997 and her
Ph.D. in 1999 from FSU in social and cogni-
tive psychology. From 1994 to 1998, she was
a research assistant in an FSU psychology
lab devoted specifically to studies on psy-
chology and the law. She had previously
worked for one year as a research assistant
in a litigation psychology firm, assisting attorneys with case
preparation and strategy; during graduate school, she worked
with the Public Defender’s Office in Tallahassee writing voir dire
questions and assisting in jury selection. Wasserman received her
undergraduate degree in psychology from the University of
Michigan in 1993.
Christian A. Meissner is a Ph.D. candidate in
the cognitive & behavioral sciences program
at Florida State University. Meissner
received his undergraduate psychology
degree from Pfeiffer University in 1996,
where he graduated magna cum laude. He
completed his master’s degree at FSU in 1999
under the supervision of Dr. Brigham.
Meisner’s current psycholegal research inter-
ests involve the various cognitive and social
influences on an individual’s memory for human faces.
Summer 1999 - Court Review 25
... Eyewitness misidentifications have been recognized as the single most common cause of wrongful convictions (Huff, Rattner, & Sagarin, 1986;Wall, 1965). However, eyewitness testimony also can be accurate and, unlike the early admonishments made by Hugo Münsterberg (1908), contemporary eyewitness researchers have not advocated that the courts submit to an alternative authority (Yarmey, 2001a), or suggest that eyewitness evidence be excluded at trial (Brigham et al., 1999). Significant advancements in the empirically based scientific understandings of the psychology of eyewitness testimony began to be made in the 1970s (see Clifford & Bull, 1978;Loftus, 1979;Yarmey, 1979). ...
... In 1967, the Supreme Court of the United States in a trilogy of cases, United States v. Wade, Gilbert v. California, and Stovall v. Denno, recognized that the lineup is a critical stage vis-à-vis the constitutionality of police practices and procedures in obtaining eyewitness identifications. Thus, the problematic nature of eyewitness reports was explicitly acknowledged by the U.S. Supreme Court (see Brigham et al., 1999). Note, however, that the Court's concern and recommendations were made without any substantial contribution from psychological science. ...
... Again, these criteria were established with little or no consultation with psychological experts on perception and memory. Subsequent research has shown that only two of the factors listed in Neil are related to the estimation of the accuracy of identification (i.e., opportunity to observe and the length of the retention interval, Brigham et al., 1999). However, research also shows that their influence on eyewitness accuracy may be moderated by other variables, such as race or stress. ...
Article
Full-text available
Innocent people are convicted each year because of mistaken eyewitness identification. This paper addresses both historical and more contemporary perspectives on the possible limitations of eyewitness identification. The primary purpose of this article is to describe some of the attempts made to protect the fact-finding process from misidentifications. The development and nature of guidelines and recommendations for the design and administration of identification lineups in the United States and in Canada are described. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
... 4 http://www.nytimes.com/2015/05/15/nyregion/witness-accounts-in-midtown-hammer-attack-show-the-power-of-falsememory.html?_r=0 suggests that 64% of convictions that relied primarily on eyewitness testimony were incorrect (Gross et al., 2005). A variety of explanations for eyewitness error have been recognized, including fear and high-stress incidents (Brigham et al., 1999;Wise et al., 2005), the presence of a weapon (Stebley, 1992;Brigham et al., 1999), the amount of time a person can focus on an event as it plays out (Loftus and Hoffman, 1989;Loftus, 1992;Loftus, 2005), and the time between the end of the event and when a person can document their observations (Loftus and Hoffman, 1989;Candel and Merckelbach, 2004). ...
... 4 http://www.nytimes.com/2015/05/15/nyregion/witness-accounts-in-midtown-hammer-attack-show-the-power-of-falsememory.html?_r=0 suggests that 64% of convictions that relied primarily on eyewitness testimony were incorrect (Gross et al., 2005). A variety of explanations for eyewitness error have been recognized, including fear and high-stress incidents (Brigham et al., 1999;Wise et al., 2005), the presence of a weapon (Stebley, 1992;Brigham et al., 1999), the amount of time a person can focus on an event as it plays out (Loftus and Hoffman, 1989;Loftus, 1992;Loftus, 2005), and the time between the end of the event and when a person can document their observations (Loftus and Hoffman, 1989;Candel and Merckelbach, 2004). ...
Article
Full-text available
Questions about police officer behaviour, particularly officer-involved shootings, have contributed to the argument that street-level police officers should wear body cameras. The assumption is that a body camera will provide an objective reality of what occurred during an encounter. Absent from the discussion is the notion of perceptual distortion, misinformation, and the development of false memories. This article provides an examination of how these psychological dimensions can impact a police officer’s decision-making as well as their ability to accurately recall the details of an incident. It is argued here that while a body camera video may provide accurate documentation of an event, it is reasonable to argue that what the officer sees may not match objective reality because of perceptual distortions. Further, deviations between objective reality and officer recall does not equate to lying, a cover-up, or a ‘rogue’ officer.
... The ubiquity of Closed Circuit Television (CCTV) cameras and the ready capture of images from affordable digital and video cameras and mobile devices, has led to a plethora of sources of facial image evidence available for use in criminal investigation and in court. In this context it is paradoxical that, whilst eyewitness facial identification has become recognised by both scientists and courts as error prone (1,2), technological approaches to courtroom identification of facial images have largely escaped legal and scientific scrutiny. This paper considers current approaches to facial identification in relation to criteria for courtroom admissibility of expert evidence in the United States and England and Wales. ...
Article
Full-text available
Much contemporary debate in forensic science concerns validity and admissibility of scientific evidence in court. In this paper, three current approaches to facial identification-image superimposition, photogrammetry, and morphological analysis-are considered with regard to criteria for scientific evidence in the United States, and England, and Wales. The aim of the paper is to assess the extent to which facial image comparison meets criteria of admissibility in these jurisdictions. The method used is a comparative evaluation of the methods of facial image comparison and their underlying premises against the range of admissibility criteria reported in court rulings and relevant judicial and scientific inquiries in the United States and the United Kingdom. While the techniques of facial image comparison are generally accepted within their practitioner communities, they are not tested, and their error rates are unknown. On that basis, the methods of facial image comparison would appear not to meet the anticipated standards. They are, nevertheless, admitted in court in the United States, and England, and Wales. This paper concludes that further research in science and law will be necessary to more definitively establish admissibility of facial image comparison evidence, as it will for other nascent and novel methods that are potentially influential in court proceedings.
Chapter
Full-text available
Eyewitness memory represents an inherently applied research problem, wherein scholars have increased public awareness of the problem of mistaken eyewitness identification and successfully developed policies and procedures that will increase the diagnostic value of an identification. At the same time, a tension has long existed between those that have urged the field to adopt this applied research focus and those that have advocated for a more theoretically informed research focus. In the current chapter, we offer a process perspective that engages psychological theories of memory, face recognition, social influence and decision processes that have been shown to influence eyewitness identifications. We propose that the eyewitness context affords scholars a 'middle road' to engage in the development and refinement of such theoretical frameworks. Greater attention to such a process perspective, rooted in the rich theoretical backdrops of cognitive and social psychology, is more likely to enhance our understanding of eyewitness decisions and lead to novel insights that leverage core processes.
Article
Citation: Brigham, J., & Grisso, T. (2003). Forensic Psychology. In D.K. Friedheim & I.B. Weiner (eds.), Comprehensive handbook of psychology. Vol. 1: The history of psychology, (pp. 391-411). New York: John Wiley and Sons. ISBN 0471383201, 9780471383208. Limited preview available via Google Books.
Article
Full-text available
The current article reviews the own-race bias (ORB) phenomenon in memory for human faces, the finding that own-race faces are better remembered when compared with memory for faces of another, less familiar race. Data were analyzed from 39 research articles, involving 91 independent samples and nearly 5,000 participants. Measures of hit and false alarm rates, and aggregate measures of discrimination accuracy and response criterion were examined, including an analysis of 8 study moderators. Several theoretical relationships were also assessed (i.e., the influence of racial attitudes and interracial contact). Overall, results indicated a "mirror effect" pattern in which own-race faces yielded a higher proportion of hits and a lower proportion of false alarms compared with other-race faces. Consistent with this effect, a significant ORB was also found in aggregate measures of discrimination accuracy and response criterion. The influence of perceptual learning and differentiation processes in the ORB are discussed, in addition to the practical implications of this phenomenon.
Article
Full-text available
The current article reviews the own-race bias (ORB) phenomenon in memory for human faces, the finding that own-race faces are better remembered when compared with memory for faces of another, less familiar race. Data were analyzed from 39 research articles, involving 91 independent samples and nearly 5,000 participants. Measures of hit and false alarm rates, and aggregate measures of discrimination accuracy and response criterion were examined, including an analysis of 8 study moderators. Several theoretical relationships were also assessed (i.e., the influence of racial attitudes and interracial contact). Overall, results indicated a "mirror effect" pattern in which own-race faces yielded a higher proportion of hits and a lower proportion of false alarms compared with other-race faces. Consistent with this effect, a significant ORB was also found in aggregate measures of discrimination accuracy and response criterion. The influence of perceptual learning and differentiation processes in the ORB are discussed, in addition to the practical implications of this phenomenon. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Article
We compared what 160 U.S. judges, 57 law students, and 121 undergraduates know and believe about factors affecting the accuracy of eyewitness testimony. Judges were no more knowledgeable than were undergraduates, and both groups were less knowledgeable than were law students. For all 3 groups, increased knowledge of eyewitness factors was associated with beliefs that might reduce wrongful convictions. Participants in all 3 groups underestimated what potential jurors know about eyewitness testimony. The results suggest that increasing judges' knowledge of eyewitness testimony might help them to reduce wrongful convictions and to more accurately assess when eyewitness experts are needed. The results also suggest that law schools need to do a better job of educating law students about eyewitness testimony.
Article
Full-text available
Sixty-three experts on eyewitness testimony were surveyed about their courtroom experiences and opinions on various issues. There was a strong consensus indicated by an agreement rate of at least 80% that the data on the following topics are reliable enough to present in court: the wording of questions, lineup instructions, misleading postevent information, the accuracy-confidence correlation, attitudes and expectations, exposure time, unconscious transference, showups, and the forgetting curve. Over 70% of the experts also endorsed lineup fairness, the cross-race identification bias among White witnesses, and the tendency to overestimate the duration of events. Although most eyewitness experts who have testified have done so on behalf of criminal defendants, they were just as likely to consent for the prosecution as for the defense; moreover, they were more likely to agree to testify in civil cases than in criminal. Concerning their role in court, most respondents indicated that their main objective is to educate the jury, and that juries are more competent with the aid of experts than without. The results are discussed in relation to the "general acceptance" provision of the Frye test and the limitations of this test for determining the admissibility of expert testimony.
Article
Full-text available
Inasmuch as a completely satisfactory estimate of effect size for the eyewitness accuracy-confidence relation does not exist, we conducted a meta-analysis of 35 staged-event studies. Estimated r = .25 ( d = .52), with a 95% confidence interval of .08 to .42. Sampling error accounted for 52% of the variation in r, leaving room for measurement error and possibly moderator variables to account for the remaining variation. Further analysis identified duration of target face exposure as a moderator variable, providing support for Deffenbacher's (1980) optimality hypothesis. When corrected for the attenuating effect of sampling error in the accuracy-confidence correlations, the correlation of exposure duration and the accuracy-confidence correlation was .51: Longer exposures allowed for greater predictability of accuracy from confidence. Even through correlation for unreliability in the confidence measure produces a higher estimate of the population correlation of accuracy and confidence, .34, one must be cautious in assessing the utility of confidence for predicting accuracy in actual cases. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Article
Full-text available
Thefts were staged 108 times for as many witnesses who were subsequently given a photo lineup for identifying the thief. The thefts were staged under conditions designed to yield low (33%), moderate (50%), or high (74%) proportions of correct identifications of the thief. Corroborating past research, the relationship between witnesses' identification accuracy and witnesses' confidence was negligible within conditions. There was no evidence that the confidence–accuracy relationship changed across conditions or that witness confidence changed across theft conditions. A representative sample of 48 witnesses (8 accurate-identification and 8 false-identification witnesses from each of the 3 theft conditions) was cross-examined. 96 undergraduates viewing the cross-examinations showed no ability to detect accurate- from false-identification witnesses within conditions as measured by Ss' belief of witnesses. Although Ss changed their rate of belief of witnesses as a function of the theft conditions (62, 66, and 77%, respectively), the rate at which Ss discounted witnesses' testimony was insufficient across conditions. Ss were especially overbelieving of witnesses when the rate of witness accuracy in that condition was low. (21 ref) (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Article
Full-text available
Although eyewitness identifications are among the most common forms of evidence presented in criminal trials, both archival studies and psychological research suggest that eyewitnesses are frequently mistaken in their identifications (B. L. Cutler & S. D. Penrod, 1995). In recognition of this problem, the legal system has established a number of safeguards to protect defendants from erroneous convictions resulting from mistaken identifications. These safeguards are based on assumptions regarding attorney, judge, and juror commonsense knowledge of the factors influencing eyewitness identification accuracy. This article addresses the validity of these assumptions by examining the role of commonsense knowledge in attorney, judge, and juror evaluations of eyewitness identification evidence. It concludes that, although these safeguards may not be as effective as the legal system intended them to be, there are a number of practices and policies that may be implemented to safeguard defendants further. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Article
We hypothesized that both accuracy and confidence in suspect identifications depend, in part, on participants' ability to identify the target, and that both accuracy and confidence therefore tend to be higher under conditions that lead to good memory for the target than under conditions that lead to poor memory for the target. Furthermore, we hypothesized that a substantial correlation between accuracy and confidence will be observed if, because of variations in conditions, there is considerable variability across participants in ability to identify the target. Consistent with these hypotheses, manipulations that affected accuracy also affected confidence in the same direction, and when data were collapsed across conditions, the accuracy-confidence correlation was substantial (mean r = .59).
Article
This study investigates the impact of different types of expert testimony regarding the unreliability of eyewitness identification. In two hypothetical court cases involving eyewitnesses, expert testimony was presented that was either sample-based (presenting the results of a research program on eyewitness identification) or person-based (presenting information about the particular eyewitness under consideration); the expert either offered causal explanations for his unreliability claim or failed to do so. Two additional control groups (with and without eye-witness identification) were not presented with any expert testimony. The results indicate that subjects who had been confronted with an expert statement made more lenient judgments about the offender but did not discount the eyewitness identification completely. Sample-based information had a moderate impact on the subjects' judgments, regardless of whether or not causal explanations were given. Person-based testimony was the most influential type of expert advice when a causal explanation was provided but the least influential one when no reasons were given. The practical (international differences in admissibility of expert testimony) and theoretical implications (processing of base-rate information) of these findings are discussed.
Article
La contribution des psychologues au developpement des programmes d'etudes et a la formation de l'Academie de Police de British Columbia est analysee d'une maniere critique. Deux themes de recherche ont ete poursuivis en collaboration avec des stagiaires de cette institution. Un ensemble d'etudes a demontre la valeur de la formation des policiers lorsqu'ils interviennent dans des conflits familiaux. Le deuxieme groupe d'etudes a examine les effets de la formation et de l'experience au point de vue de la capacite des policiers en tant que temoins oculaires. Les policiers avec formation ont surpasses leurs collegues sans formation vis-a-vis de la qualite des temoignages et de leur quantite