Psychology, Crime & Law, 2013
http://dx.doi.org/10.1080/1068316X.2013.770855
Plea bargaining and appraisals of eyewitness evidence by prosecutors
and defense attorneys
Kathy Pezdek* and Matthew O’Brien
Department of Psychology, Claremont Graduate University, Claremont, CA, USA
Downloaded by [matthew o'brien] at 18:35 10 April 2013
(Received 12 May 2012; final version received 6 November 2012)
A sample of defense attorneys and prosecutors from matched California counties
participated in a two-part study. Study 1A reports the results of a survey
regarding how influential each of the 17 eyewitness factors is in affecting the
accuracy of real eyewitness identifications. Generally, both attorney groups
considered all eyewitness factors to be influential; on only 6 of the 17 factors were
defense attorneys more likely than prosecutors to provide higher importance
ratings. In Study 1B, the attorneys answered questions regarding their willingness
to plea bargain after reading each of four scenarios in which (1) same- versus
cross-race identification and (2) whether the perpetrator was familiar were
experimentally manipulated. Both eyewitness factors influenced plea bargaining
decisions, and effects were generally consistent for both attorney groups. Results
confirm that plea bargaining decisions at least by defense attorneys are made ‘in
the shadow of the trial,’ and that appraisals of the strength of eyewitness evidence
play a significant role in these decisions.
Keywords: eyewitness identification; eyewitness memory; plea bargaining; ‘in the
shadow of the trial’
On 21 March 2012, Supreme Court of the United States, ruling on two cases,
Missouri v. Frye and Lafler v. Cooper, determined that defendants have a
constitutional right to effective assistance of counsel during plea negotiations. In
the majority opinion, Justice Kennedy wrote, ‘Criminal justice today is for the most
part a system of pleas, not a system of trials.’ In this ruling, the Supreme Court
determined that not only do all defendants, including indigent defendants, have a
right to effective counsel at trial as guaranteed by the 6th Amendment to the
Constitution, but also, in light of the fact that the large majority of all cases filed are
settled by plea bargaining, defendants have the right to effective counsel during plea
negotiations as well.
It has been estimated that in the USA, approximately 90% of cases are
resolved through plea bargaining (Libuser, 2001). Regarding federal criminal trials,
Higginbotham (2004) reported that the number of federal criminal trials has dropped
from 14 to 15% of all criminal dispositions in 1976 to about 2% in 2002, and
according to the Bureau of Justice Statistics (2005) between October 2004 and
September 2005, 86% of all federal criminal cases filed were resolved with a guilty
plea. So significant is this trend that the effect has been referred to as ‘the
*Corresponding author. Email: Kathy.Pezdek@cgu.edu
# 2013 Taylor & Francis
Downloaded by [matthew o'brien] at 18:35 10 April 2013
2
K. Pezdek and M. O’Brien
disappearing trial’ (Higginbotham, 2004). Given this trend, and in light of the recent
Supreme Court decision, it is increasingly more important to assess how attorneys
resolve criminal cases rather than how juries do so. This two-part study focuses on
how prosecutors and defense attorneys weigh the evidence in eyewitness identification cases (Study 1A) and specifically assesses how decisions relevant to plea
bargaining are affected by two of these eyewitness factors (Study 1B).
Eyewitness evidence is critical for solving crimes, and it is often the sole source of
evidence for determining the perpetrator’s identity. However, studies consistently
report that eyewitness misidentifications are the leading cause of erroneous
convictions (Huff, 1987; Huff, Rattner, & Sagarin, 1996; Penrod & Cutler, 1999);
eyewitnesses frequently identify the wrong individual, or they fail to identify the
correct individual. There is a wealth of scientific research on the psychological
factors that affect the accuracy of eyewitness memory. Reviews of this work include a
meta-analysis by Shapiro and Penrod (1986), a two volume edited tome by Toglia,
Read, Ross, and Lindsay (2006), recent articles by Wells, Memon, and Penrod (2006)
and Wells and Olsen (2003), and chapters by Pezdek (2007, 2009). Most of this
research has examined the factors that affect eyewitness memory and how this
information influences jurors’ decision-making. The present study compared
appraisals of the strength of eyewitness evidence by prosecutors and defense
attorneys and assessed how decisions relevant to plea bargaining are affected by
two of these eyewitness factors (Study 1B). There are important public policy
implications of this research because in fact, it is attorneys who estimate the strength
of the eyewitness evidence in real criminal cases and determine which cases will go to
trial.
It is generally assumed that parties strike plea bargains ‘in the shadows of
expected trial outcomes,’ a concept that dates back at least as early as Landes (1971).
In keeping with this view, Burke (2007) and Pritchard (1986) reported that attorneys’
decisions regarding whether to plea bargain a case are largely based on their
perceptions of the strength of the evidence against the defendant. This is consistent
with both decision theory accounts of plea bargaining and economic models of plea
bargaining (Covey, 2007; Kramer, Wolbransky, & Heilbrun, 2007). When perception
of the evidence is weak, prosecutors are more likely to offer a plea bargain; when
perception of the evidence is strong, defense attorneys are more likely to recommend
a plea bargain. However, in a prominent article in the Harvard Law Review, Bibas
(2004) argued that information deficits and cognitive biases also play a significant
part of plea bargaining. In one such example, Goodman-Delahunty, Granhag,
Hartwig, and Loftus (2010) reported that attorneys tend to be overconfident in their
predictions of trial outcomes.
Relevant to the present study, the presence of an eyewitness identification has
been reported to reduce prosecutors’ desire to plea bargain and increase defense
attorneys’ desire to plea bargain (McAllister, 1990). More recently, Flowe, Mehta,
and Ebbesen (2011) reviewed archival data on dispositions in 725 felony cases of a
California District Attorney’s Office and found that a positive eyewitness identification was present more often in accepted than rejected cases (although only for
acquaintance but not stranger identifications). These findings are important because
eyewitness evidence is one of the more frequently encountered types of evidence in
criminal cases. Wells et al. (1998) estimated that each year in the USA, eyewitness
Downloaded by [matthew o'brien] at 18:35 10 April 2013
Psychology, Crime & Law
3
evidence is the primary or sole evidence against the defendant in at least 77,000
criminal trials.
What do we know about how accurately attorneys can determine variations in the
strength of the eyewitness evidence? This question is even more important in light of
the 2012 Supreme Court decisions because the Court’s decision dictates that
attorneys will have an even greater role in the plea bargaining decisions of
defendants. In prior cases, courts have assumed that attorneys understand the
factors that influence the fairness of identification procedures (Kirby v. Illinois, 1972;
United States v. Wade, 1967). Both of these cases concerned the rights of defendants
to have defense counsel present at a live lineup in light of the possibility that pretrial
identifications can suggestively influence subsequent identifications at trial. However, the results of research studies that focused on attorneys’ knowledge of the
factors that influence the fairness of identification procedures specifically, and
eyewitness identification more generally, are less convincing. A number of studies
have examined how well attorneys understand the specific factors that relate to the
accuracy of eyewitness evidence (Brigham & WolfsKeil, 1983; Lindsay, MacDonald,
& McGarry, 1990; Stinson, Devenport, & Kravitz, 1996). However, none of these
studies included the scope of eyewitness factors now known to be significant, nor did
these studies compare the knowledge and beliefs of prosecutors with defense
attorney. On both points, a more recent study by Wise, Pawlenko, Safer, and Meyer
(2009) is more useful.
Wise et al. (2009) had prosecutors and defense attorneys with extensive criminal
trial experience complete a survey that primarily focused on 13 statements assessing
knowledge of eyewitness factors. The major finding was that prosecutors responded
correctly to significantly fewer of the 13 eyewitness questions (M !6.07) than did the
defense attorneys (M !10.10). Prosecutors were also less skeptical of eyewitness
evidence than defense attorneys and less skeptical of jurors’ knowledge of eyewitness
testimony. Based on these findings, we posit that if prosecutors (and defense attorneys
as well) had more accurate knowledge about when eyewitness evidence is more likely to
be weak versus strong, this would positively influence plea bargaining decisions and
better protect the rights and liberties of individuals on both sides of the bar.
This research includes a study with two parts " a survey and an experiment " to
compare appraisals of the strength of eyewitness evidence by prosecutors and defense
attorneys (Study 1A) and assess how decisions relevant to plea bargaining are
affected by two of these eyewitness factors (Study 1B). A sample of 93 defense
attorneys and 46 prosecutors from matched counties in California participated in
both Study 1A and 1B. The attorneys had extensive experience practicing law and
trying felony criminal cases in Superior Court. Study 1A reports the results of a
survey of the attorneys regarding how influential they thought each of 17 eyewitness
factors was in affecting the accuracy of eyewitness memory and identification in realworld cases. This is a different question than that posed by Wise et al. (2009), who
asked about the truth value (‘generally true,’ ‘generally false,’ or ‘I do not know’) of
statements regarding specific eyewitness factors. In Study 1B, the attorneys were
presented four scenarios in which two specific eyewitness factors " (1) same- versus
cross-race identification and (2) prior contact or not " were experimentally
manipulated in a factorial design. After reading each scenario, they were asked
four questions regarding whether they would plea bargain the case, the lowest/highest
plea bargain they would offer/accept, and their estimate of the probability that the
4
K. Pezdek and M. O’Brien
defendant was guilty, and the probability that they would win the case if it went to
trial. Study 1B experimentally assessed how these typical decisions related to plea
bargaining are influenced by variations in the strength of two eyewitness factors, and
the whether this pattern of results differs for prosecutors versus defense attorneys.
Downloaded by [matthew o'brien] at 18:35 10 April 2013
Study 1A
Methods
Participant sample
The data for this program of research were collected from county offices of the District
Attorney and Public Defender in the state of California over a period of 18 months.
Because the questionnaire in Study 1B included questions about sentencing, and
sentencing guidelines vary by state in the USA, it was necessary to restrict data
collection to one state. In the data collection phase, over a period of 18 months, email,
telephone, and postal mail were used to contact repeatedly, the head District Attorney
and Public Defender in each of the 58 counties in California, to request the
participation of the Deputy District Attorneys and Deputy Public Defenders in their
county. Only the data from counties in which there were completed test materials from
both the offices of the District Attorney and the Public Defender were included in this
study. This was done to match the two samples, at least in terms of the counties in
which they served. These included eight counties, contributing a total of 93 Deputy
Public Defenders and 46 District Attorneys. Within each county, it was simply
required that each participant had had felony trial experience. Participants were given
the opportunity to complete the test materials online or using hard copies; most used
the hard copy option.
Power analyses were conducted using methods suggested by Faul, Erdfelder,
Lang, & Buchner (2007). The number of participants exceeded that required to
detect effects with an estimated effect size of 0.25, a !0.05 and power !0.80.
On the questionnaire, each participant was asked four background questions.
These questions, and the mean response for each by all prosecutors and defense
attorneys, are presented in Table 1. On none of these four questions was there a
Table 1. Background data demonstrating equivalence between prosecutors and defense
attorneys.
Question
(1) How many years have you been a
trial attorney?
(2) How many cases do you estimate
you have tried?
(3) In a typical year, how many of
your cases involve eyewitness
evidence?
(4) In the last 5 years, what percent of
your cases typically are settled
through plea bargaining?
Prosecutors
Defense attorneys
mean rating (SD) mean rating (SD)
Tests of
significance
14.24 (9.28)
12.82 (10.98)
t(135) !0.75
56.79 (50.56)
54.75 (78.10)
t(134) !0.16
15.35 (29.04)
19.51 (28.97)
t(94) !0.65
86.23 (19.95)
86.23 (19.22)
t(132) !0.01
Note: There were no statistically significant differences between prosecutors and defense attorneys in this
sample.
Psychology, Crime & Law
5
significant difference between the responses of these two groups, all t-values B1.00.
This suggests that differences in the pattern of responses to the questions regarding
assessments of eyewitness evidence cannot be attributed to differences in felony trial
experience between the two groups.
Downloaded by [matthew o'brien] at 18:35 10 April 2013
Questionnaire
Each participant was provided one packet of test materials including those for both
Study 1A and 1B. The questionnaire for Study 1A was placed at the end of the test
materials to reduce the extent to which Study 1A responses sensitized respondents to
the factors varied in the Study 1B scenarios. The questionnaire for Study 1A included
a list of 17 eyewitness factors (see Table 2). The 17 factors were selected from the list
of factors included in the meta-analysis of face recognition research by Shapiro and
Penrod (1986) with the addition of several significant eyewitness factors reported
more recently by Pezdek (2007). Participants were instructed to read each factor and,
‘based on your general knowledge and experience, indicate how influential you think
each factor is in affecting the reliability of eyewitness memory and identification in
real-world cases.’ They responded to each factor on a 1 (minimally influential) to 7
(most influential) scale and were encouraged to use the full range of the response
scale.
Results and discussion
The mean responses to each of the 17 eyewitness factors by prosecutors and defense
attorneys are provided in Table 2. First, the mean rating of how influential each
factor was considered to be were generally high. On the 7-point scale, all but three of
the 28 mean ratings presented in Table 2 were above 4.00, the mid-point of the scale.
The three exceptions were two ratings by prosecutors of the influence of not
following a double blind procedure (#15) and post-event information (#16), and one
rating by defense attorneys regarding the confidence expressed by the eyewitness
(#10).
The mean responses per item by prosecutors and defense attorneys were next
compared with a multivariate analysis of variance (MANOVA). Results of the
MANOVA revealed a significant multivariate main effect for responses given by the
two attorney groups on the 17 eyewitness factors, Wilks’ l !0.46, F(17, 109) !7.65,
pB0.001, and h2 !0.54. Post hoc comparisons (two-tailed) were conducted using a
Bonferroni adjusted alpha level on each of the 17 eyewitness factors. On 8 of the 17
items, there was a significant difference between ratings of how influential the factor
was considered to be by prosecutors versus defense attorneys. On two of these
factors, the prosecutors thought that the factor was more influential than did the
defense attorney. These two factors are confidence expressed by the eyewitness (#10)
and the number of independent eyewitnesses who do identify the defendant (#11).
On six of these factors, the prosecutors thought that the factor was less influential
than did the defense attorney. These six factors are cross-race identification (#4),
time delay to identification (#7), composition of the lineup used and possible lineup
bias (#14), double blind procedure (#15), post-event information (#16), and
contamination introduced in a police interview with eyewitness (#17). Responses
to question 14 are especially interesting in light of court findings that attorneys
6
K. Pezdek and M. O’Brien
understand the factors that influence the fairness of identification procedures (Kirby
v. Illinois, 1972; United States v. Wade, 1967). Responses to question 14 suggest that
attorneys do recognize the influence of fair identification procedures, but that
defense attorneys rate the potential biasing influence of these identification
procedures significantly higher than do prosecutors.
Table 2. Assessments of prosecutors and defense attorneys in Study 1A regarding how
influential each of the 17 eyewitness factor is in affecting the reliability of eyewitness memory
and identification in real-world cases, scale from 1 (minimally influential) to 7 (most
influential).
Downloaded by [matthew o'brien] at 18:35 10 April 2013
Statement
(1) Exposure time to the perpetrator’s
face
(2) Distraction of multiple suspects
present
(3) Visible weapon
(4) Cross-race identification of the
eyewitness and perpetrator
(5) Perpetrator was wearing a hat
and/or sunglasses
(6) Distance, lighting and possible
obstructions to view
(7) Time delay from the incident to
the time of the first ID
(8) High level of stress during the
incident
(9) Whether the eyewitness had
previously seen the perpetrator
(10) The confidence expressed by the
eyewitness
(11) Number of independent
eyewitnesses who do ID defendant
(12) Number of independent
eyewitnesses who do not ID
defendant
(13) Whether the eyewitness is
intoxicated/sober
(14) Composition of the lineup used
and possible lineup bias
(15) Officer administering lineup or
showup knows who the suspect is
(16) Post-event information " from
other witnesses, media, etc.
(17) Contamination introduced in
police interview with eyewitness
Prosecuting
attorney mean
rating (w/SD)
Defense
attorney mean
rating (w/SD)
5.95 (0.96)
5.54 (1.18)
t(131) !2.03
4.86 (1.05)
5.21 (1.23)
t(131) !1.62
4.89 (1.33)
4.20 (1.67)
5.26 (1.46)
5.42 (1.30)
t(131) !1.42
t(131) !4.59*
4.98 (1.50)
5.10 (1.37)
t(131) !0.48
5.68 (1.29)
5.75 (1.22)
t(131) !0.31
5.00 (1.57)
5.80 (1.18)
t(130) !3.27**
4.66 (1.61)
5.48 (1.63)
t(130) !2.73
6.35 (1.09)
5.91 (1.13)
t(129) !2.12
4.80 (1.59)
3.38 (2.10)
t(131) !3.93*
6.25 (.99)
5.14 (1.81)
t(130) !3.79*
4.70 (1.97)
5.07 (1.78)
t(131) !1.07
5.75 (1.38)
5.91 (1.03)
t(131)!.75
4.68 (1.54)
5.47 (1.21)
t(131) !3.24**
2.66 (1.64)
5.10 (1.44)
t(130) !8.77*
3.77 (1.72)
5.39 (1.21)
t(131) !6.28*
4.86 (1.70)
5.70 (1.27)
t(130) !3.17**
Note: Bonferroni correction applied to p-values to protect against alpha inflation.
*p B0.001; **pB0.05.
Tests of
significance
Downloaded by [matthew o'brien] at 18:35 10 April 2013
Psychology, Crime & Law
7
These results are not directly comparable to those of Wise et al. (2009) because in
the previous study, attorneys were asked about the truth value (‘generally true,’
‘generally false,’ or ‘I do not know’) of statements regarding specific eyewitness
factors, and the results were reported in terms of accuracy rates comparing the
results reported in the relevant research literature. In that study, defense attorneys
were generally more knowledgeable than were the prosecutors, and Wise et al.
concluded that differences in prosecutors’ and defense attorneys’ knowledge and
beliefs about eyewitness testimony ‘may even be greater today’ (p. 1277).
In the current study, we specifically did not access accuracy. Rather, attorneys
were asked based on their general knowledge and experience, how influential each
factor was in affecting eyewitness memory in real-world cases. There are two principal
findings. First, with only the three exceptions mentioned above, generally, both
prosecutors and defense attorneys considered all of the eyewitness factors to be
influential in affecting eyewitness memory in real-world cases. Second, on the
majority of the 17 items, ratings by prosecutors and defense attorneys did not differ.
However, when there were differences between the ratings of prosecutors and those
of defense attorneys, higher ratings of influence were more likely to be reported by
defense attorneys than prosecutors.
Several previous studies have compared prosecutors with defense attorneys on
their knowledge and beliefs about the role of various factors in eyewitness
identification accuracy. Comparing their more recent findings with those of Brigham
and WolfsKeil (1983) a quarter of a century prior, Wise et al. (2009) concluded that
differences in prosecutors’ and defense attorneys’ knowledge and beliefs about
eyewitness testimony ‘may even be greater today’ (p. 1277). In both studies, it was
reported that prosecutors were significantly less knowledgeable than defense
attorneys on almost every issue. The results of Study 1A are more encouraging;
prosecutors and defense attorneys were more consistent in their ratings of the role of
eyewitness factors. Both prosecutors and defense attorneys considered most of the 17
eyewitness factors to be influential in affecting eyewitness memory in real-world
cases, and, when there were differences between the ratings of prosecutors and
defense attorneys, higher ratings of influence were only modestly more likely to be
reported by defense attorneys than prosecutors.
Study 1B
Study 1B extends the findings on perceptions of the strength of eyewitness evidence
to other relevant decisions in the criminal justice process. Whereas Study 1A assessed
how influential prosecutors and defense attorneys considered various eyewitness
factors to be in affecting the accuracy of eyewitness memory and identification in
real-world cases, Study 1B experimentally assessed how typical decisions regarding
plea bargaining by prosecutors and defense attorneys are influenced by variations in
the strength of two of these eyewitness factors. These two factors are (1) same- versus
cross-race identification and (2) whether the eyewitness had had prior contact with
the perpetrator (i.e. familiar) or not (i.e. unfamiliar). After reading each scenario,
attorneys answered four questions regarding their estimate of the probability that the
defendant was guilty (question 1), the probability that they would win the case if it
went to trial (question 2), whether they would plea bargain the case (question 3), and
the lowest/highest plea bargain they would offer/accept (question 4). ‘Winning’ was
Downloaded by [matthew o'brien] at 18:35 10 April 2013
8
K. Pezdek and M. O’Brien
defined as getting a guilty verdict for prosecutors and an acquittal for defense
attorneys. Only questions 3 and 4 specifically ask about plea bargaining. However,
we questioned whether the defendant was considered to be guilty and whether they
would win the case if it went to trial to assess the extent to which plea bargaining
decisions were made ‘in the shadow of the trial.’
The cross-race effect is one of the strongest factors associated with identification
accuracy (Kassin, Tubb, Hosch, & Memon, 2001; Meissner & Brigham, 2001).
Meissner and Brigham (2001) reviewed 39 studies on cross-race identification and
reported that eyewitnesses were 1.4 times more likely to identify correctly previously
viewed own- than other-race faces. Furthermore, selection of the wrong suspect was
1.56 times more likely with other- than same-race individuals. The cross-race effect
has also been observed to be consistent across a wide age range (Pezdek, BlandonGitlin, & Moore, 2003).
The second eyewitness factor that was manipulated in the Study 1B scenarios
was whether the eyewitness had previously seen the perpetrator. It would seem a
matter of common sense that eyewitnesses would be more likely to identify
correctly someone they had seen before than someone they had not seen before.
And, in fact, at least one court has held the exclusion of an eyewitness expert
harmless when the witness claimed to have been familiar with the defendant and
seen him on a daily basis for ‘well over a year,’ Hagar v. United States. In the
familiar condition in Study 1B, the scenario included this sentence: ‘The clerk was a
22-year-old Hispanic male who, at the scene of the crime, told the police that he
had seen the robber a couple of times in the neighborhood last summer.’ Study 1B
assesses the extent to which prosecutors and defense attorneys consider this
statement of casual familiarity indicative of actual prior contact and predictive of
eyewitness identification accuracy.
Relevant to Study 1B is the question of whether individuals can assess the
accuracy of eyewitness statements from brief written scenarios and whether
findings from written scenarios generalize to real criminal cases. Lindholm
(2008) had police detectives, judges, and laypeople to judge the accuracy of
eyewitness statements from their videotaped responses or from transcripts.
Participants were actually more accurate evaluating the eyewitness statements
when this information was in writing rather than videotaped. Together with results
reported by Pezdek, Avila-Mora, and Sperry (2010), these findings suggest that the
use of brief written crime scenarios in Study 1B does not restrict the generalizability
of the findings to real criminal cases.
Methods
Participants and design
The same 93 Deputy Public Defenders and 46 District Attorneys who participated in
Study 1A also participated in Study 1B. Study 1B is 2 (prosecutor vs. defense
attorney) #2 (familiar vs. unfamiliar suspect) #2 (same-race vs. cross-race identification) mixed factor design with only the first factor varied between subjects. Four
different versions of a crime scenario were drafted in which the conditions of
familiarity (the eyewitness had seen the suspect previously or this was not mentioned)
Psychology, Crime & Law
9
and cross-race condition were varied. The four scenarios were identical except for the
slight wording changes required to vary these conditions.
Downloaded by [matthew o'brien] at 18:35 10 April 2013
Procedures and experimental materials
Each attorney read four versions of a crime scenario. The order of presenting the
scenarios was counterbalanced across participants; four of the possible 24 orders
for arranging the four scenarios were randomly selected and approximately equal
numbers of participants received each presentation order. After responding to each
scenario, instructions directed attorneys to turn the page and not return to
scenarios they had completed. The scenarios described a store robbery in which
identification by one eyewitness was the only evidence against the defendant. One
version of the store robbery scenario is presented in the Appendix; this is the crossrace condition with a familiar suspect. Because in 1977 California enacted a
Statutory Determinate Sentencing Law, it was necessary to construct scenarios that
specifically allowed for a discrete range of plea bargaining options (e.g. no weapon,
no physical harm). After reading each scenario, attorneys were asked to respond to
four questions in light of the facts presented. The four questions are specified in
Table 3.
Results and discussion
Initial analyses were conducted to assess whether the order of presenting the four
scenarios significantly affected responses to each question. Individual ANOVAs were
conducted on each of the three dependent variables (responses to questions 1, 2, and
3) using Bonferroni adjusted alpha levels. In none of these analyses was there, a
significant main effect of order or interaction of order by attorney group. The fact
that none of the findings varied as a function of the order of presenting the scenarios
minimizes concerns that the findings were affected by demand characteristics
resulting from the use of a within subjects design.
Following each scenario were the same four questions. Separate analyses were
conducted on responses to questions 1, 2, and 3 as a function of the 2 (prosecutor vs.
defense attorney)#2 (familiar vs. unfamiliar suspect) #2 (same-race vs. cross-race
identification) mixed factor design of the study, using Bonferroni adjusted alpha
levels. Post hoc simple effects tests followed all significant interactions using Tukey’s
HSD tests. The means (and standard deviations) for all conditions for all four
questions are presented in Table 3.
Question 1: In the above scenario, what do you think is the probability that the
defendant is guilty? Specify on a scale from 0 to 100%.
In the 2#2#2 ANOVA conducted on responses to question 1, all three main
effects were significant and all in the direction consistent with expectations.
Prosecutors (M !0.85, SD !0.15) were more likely to think that the defendant was
guilty than were defense attorneys (M !0.37, SD ! 0.21), F(1, 134) !231.07, pB
0.001, and h2 !0.63. Estimates of the probability that the defendant was guilty were
higher in the same-race (M !0.63, SD !0.18) than cross-race condition (M !0.60,
SD !0.18), F(1, 134) !14.42, p B0.001, and h2 !0.02. And, estimates of the
probability that the defendant was guilty were higher with familiar (M !0.66,
Cross-race
Same-race
Cross-race
Same-race
Cross-race
Familiar
Same-race
Cross-race
Unfamiliar
In the above scenario, what do you think is the probability that the defendant is guilty? Specify on a scale from 0 to 100%
89.28 (13.50)
88.61 (12.68)
81.28 (16.72)
81.15 (17.05)
45.32 (22.13)
39.53 (21.00)
35.46 (20.63)
29.28 (20.01)
In the above scenario, what is the probability that you would win this case if it went to trial? Specify on a scale from 0 to 100%
80.62 (19.47)
78.55 (18.86)
71.44 (21.22)
69.59 (23.63)
45.09 (20.48)
47.09 (20.55)
50.51 (21.80)
52.82 (21.94)
In the above scenario, what is the probability that you would offer (prosecutor) /recommend (defense attorney) any plea bargain to the
defendant? Specify on a scale from 0 to 100%
69.23 (36.22)
70.83 (36.62)
78.26 (29.50)
77.80 (33.11)
50.43 (27.59)
46.45 (26.42)
42.39 (24.68)
36.73 (27.22)
From the list of seven potential plea bargain offers specified below, circle the very lowest offer you would offer (prosecutor)/highest offer
you would recommend (defense) to the defendant in this scenario
0.39
0.41
0.30
0.26
0.44
0.39
0.31
0.25
Same-race
Unfamiliar
Defense attorneys
a
Values presented for Question 4 represent the proportion of prosecutors and defense attorneys who indicated that they would offer a plea bargain above the mode for their
attorney group.
4a
3
2
1
Question
Familiar
Prosecutors
Table 3. Mean response (with SD) for each of the four questions in Study 1B, for prosecutors and defense attorneys in the same- and cross-race
conditions for familiar and unfamiliar suspects.
10
K. Pezdek and M. O’Brien
Downloaded by [matthew o'brien] at 18:35 10 April 2013
Downloaded by [matthew o'brien] at 18:35 10 April 2013
Psychology, Crime & Law
11
SD !0.17) than unfamiliar defendants (M !0.57, SD !0.19), F(1, 134) !61.21,
p B0.001, and h2 !0.19. There was one significant interaction in the analyses of
results to this question, that is the interaction between attorney group and race, F(1,
134) !10.85, p B0.01, and h2 !0.02. For prosecutors, estimates of the defendant’s
guilt were identical in the same-race (M !0.85, SD !0.15) and cross-race conditions
(M !0.85, SD !0.15), t(45) !0.61, r !0.01. However, for defense attorneys
estimates of the defendant’s guilt were higher in the same-race (M !0.40, SD !
0.21) than cross-race condition (M !0.34, SD !0.21), t(89) !5.20, p B0.01, r !
0.14. The interaction of attorney group and familiarity was not significant;
prosecutors and defense attorneys were equally likely to think that the defendant
was guilty in the familiar and unfamiliar conditions.
Question 2: In the above scenario, what is the probability that you would win this
case if it went to trial? Specify on a scale from 0 to 100%.
In the 2#2#2 ANOVA conducted on responses to question 2, prosecutors
(M !0.75, SD !0.21) were significantly more likely than defense attorneys (M !
0.49, SD !0.21) to think they would win the case if it went to trial, F(1, 135) !
57.76, p B0.001, and h2 !0.30. There was a significant interaction of attorney group
and the cross-race condition, F (1, 135) !7.57, pB0.01, and h2 !0.01. Whereas
prosecutors indicated that the probability that they would win the case (i.e. get a
guilty verdict) if it went to trial was significantly higher in the same-race (M !0.76,
SD !0.20) than cross-race condition (M !0.74, SD !0.21), t(45) !2.20, p B0.05,
r !0.05, for defense attorneys, estimates of the probability that they would win the
case (i.e., get a not guilty verdict) if it went to trial were higher in the cross-race (M !
0.50, SD !0.21) than same-race condition (M !0.48, SD !0.21), t(90) !2.26, pB
0.05, and r !0.05. The interaction between the attorney group and the familiarity
condition was also significant, F(1, 135) !35.54, pB0.001, and h2 !0.14. Consistent with expectations, whereas prosecutors indicated that the probability they
would win the case if it went to trial was significantly higher in the familiar (M !
0.80, SD !0.19) than unfamiliar condition (M !0.71, SD !0.22), t(45) !4.28, pB
0.01, and r !0.22, defense attorneys indicated that the probability they would win
the case was significantly higher in the unfamiliar (M !0.52, SD !0.22) than
familiar condition (M !0.46, SD !0.20), t(90) !4.11, pB0.01, and r !0.14. No
other main effects or interactions were significant.
Question 3: In the above scenario, what is the probability that you would offer
(prosecutor) /recommend (defense attorney) any plea bargain to the defendant?
Specify on a scale from 0 to 100%.
In the 2#2#2 ANOVA conducted on responses to question 3, prosecutors
(M !0.74, SD !0.34) were significantly more likely to offer a plea bargain to the
defendant than were defense attorneys to recommend one (M !0.44, SD !0.26),
F(1, 135) !38.83, pB0.001, h2 !0.22. There was also a significant interaction of the
attorney group by familiarity condition, F(1, 135) !29.65, pB0.001, and h2 !0.09.
Prosecutors were more likely to offer a plea bargain to an unfamiliar (M !0.78,
SD !0.31) than a familiar defendant (M !0.70, SD !0.36), t(45) !2.79, p B0.01,
and r !0.12. However, defense attorneys were more likely to recommend a plea
bargain to a familiar (M !0.48, SD !0.27) than an unfamiliar defendant (M !
0.40, SD !0.26), t(90) !5.31, pB0.01, and r !0.17. No other main effects or
interactions were significant.
12
K. Pezdek and M. O’Brien
1
Proportion of responses
0.9
0.8
Prosecutors
0.7
0.55
0.6
Defense attorneys
0.5
0.36
0.4
0.3
0.1
0.21
0.20
0.2
0.10
0.04
0.13
0.15
0.07
0.06
0.08
0.00
0.05
0.00
0
No plea
No strike No strike
probation 16 months
One strike
probation
One strike
2 years
One strike
3 years
One strike
5 years
Downloaded by [matthew o'brien] at 18:35 10 April 2013
Plea bargain offer
Figure 1. The distribution of responses to question 4 in Study 1B, averaged across the four
scenarios, by prosecutors and defense attorneys, with the modal response for each attorney
group indicated. Within each attorney group, the modal response was the same for each of the
four scenarios. Question 4: From the list of seven potential plea bargain offers specified below,
circle the very lowest offer you would offer (prosecutor)/highest offer you would recommend
(defense) to the defendant in this scenario.
Question 4: From the list of seven potential plea bargain offers specified below,
circle the very lowest offer you would offer (prosecutor)/highest offer you would
recommend (defense) to the defendant in this scenario. Please respond to this question
but use only these seven options:
No plea
No strike
probation
No strike
16 months
One strike
probation
One strike
2 years
One strike
3 years
One strike
5 years
This response scale represents the seven options available in the four versions of
the scenario presented in this case under the California Statutory Determinate
Sentencing Law.1 This response scale is not an interval scale, and thus the options for
the statistical tests of significance are limited. Thus, the results in response to
question 4 will be presented only descriptively. The distribution of responses,
averaged across the four scenarios (varied within subjects), by prosecutors and
defense attorneys is presented in Figure 1, with the modal response for each attorney
group indicated. In each of the four scenarios, the modal response by prosecutors
was the same; their lowest offer was ‘one strike probation,’ the response selected 36%
of the time by prosecutors. The modal response by defense attorneys was also the
same in each of the four scenarios; the highest offer they would recommend was ‘no
strike probation,’ the response recommended 55% of the time by defense attorneys.
In general, then prosecutors were seeking a higher plea bargain than were defense
attorneys.
Although within each attorney group the modal response was the same for each
of the four scenarios, the distribution of responses did vary across conditions. An
additional analysis was conducted to assess how representative the mode was as an
indication of the response to question 5 by each attorney group. In this analysis, we
13
Downloaded by [matthew o'brien] at 18:35 10 April 2013
Psychology, Crime & Law
examined for each condition, the proportion of prosecutors and defense attorneys
who indicated that they would offer/recommend a plea bargain above the mode that
was reported for their attorney group in that condition. These proportions are
reported in the bottom row of Table 3. In these descriptive statistics, it can be seen
that the modal plea offer was more representative of the responses of both attorney
groups in the unfamiliar than the familiar condition. When the perpetrator was
familiar compared to unfamiliar, prosecutors were more likely to offer a plea that was
higher than the modal offer for prosecutors in this condition. Similarly, when the
perpetrator was familiar compared to unfamiliar, defense attorneys were more likely
to recommend a plea that was higher than the modal offer for defense attorneys in
this condition. This finding suggests that both prosecutors and defense attorneys
considered identifications of familiar perpetrators to be more compelling than
identifications of unfamiliar perpetrators, and were likely to adjust their plea
bargaining offers/recommendations accordingly. Consistent with the results reported
for questions 3 and 4, on question 5, the cross-race factor had less impact on the
distribution of plea bargaining decisions especially for prosecutors.
To assess whether plea bargaining decisions are made ‘in the shadow of the trial,’
Pearson product-moment correlation coefficients were computed for each attorney
group between responses to question 3 (willingness to plea bargain) and responses to
(1) question 1 (probability that the defendant is guilty) and (2) question 2
(probability of winning the case). These correlations were computed separately for
each of the four scenario types as scenario type was a within-subject factor.
Bonferroni adjusted alpha levels were used. As can be seen in Table 4, these
correlations are all significant for defense attorneys; none were significant for
prosecutors. For defense attorneys, for each scenario type, they were more likely to
recommend a plea bargain if they thought the defendant was guilty and if they
thought they were less likely to get an acquittal, thus plea bargaining in the shadow
of the trial. The absence of significant relationships between responses to question 1
and questions 2 and 3 for prosecutors is likely because of the ceiling effect and
restricted range of responses by prosecutors to questions 2 and 3 (see Table 3),
although additional research is necessary to test this interpretation.
Table 4. Summary of Pearson product-moment correlation coefficients (r) for each attorney
group in Study 1B, examining the relationship between (1) willingness to plea bargain and
belief that the suspect was guilty and (2) willingness to plea bargain and belief that the
attorney would win the case if it went to trial.
Prosecutors
Defense
Scenario
Plea bargain
and guilt
Plea bargain
and wina
Plea bargain
and guilt
Plea bargain
and wina
Cross-race familiar
Same-race familiar
Cross-race unfamiliar
Same-race unfamiliar
r ! $0.153
r ! $0.147
r ! $0.251
r ! $0.089
r! $0.241
r! $0.276
r! $0.336
r! $0.324
r !0.665*
r !0.701*
r !0.695*
r !0.670*
r ! $0.513*
r ! $0.647*
r ! $0.519*
r ! $0.446*
Note: Bonferroni correction applied to p-values to protect against alpha inflation.
a
‘Winning’ was defined as obtaining a guilty verdict for prosecutors and an acquittal for defense attorneys.
*p B0.001.
Cross-race
Same-race
Cross-race
Familiar
Unfamiliar
Prosecutors
Familiar
b
Unfamiliar
Defense attorneys
In the above scenario, what do you think is the probability that the defendant is guilty? Specify on a scale from 0 to 100%.
85.28 (15.11) !84.88 (14.87)
40.09 (21.61) !34.42 (20.60)
88.94 (13.09)
81.22 (16.89)
42.15 (15.11)
32.36 (20.34)
In the above scenario, what is the probability that you would win this case if it went to trial? Specify on a scale from 0 to 100%.
76.03 (20.35) !74.07 (21.25)
47.82 (20.92) B50.01 (20.95)
79.59 (19.17) !70.51 (22.43)
46.00 (20.31) B51.83 (21.56)
In the above scenario, what is the probability that you would offer (prosecutor) /recommend (defense attorney) any plea bargain to the
defendant? Specify on a scale from 0 to 100%.
73.75 (32.86)
74.32 (34.87)
46.25 (25.97)
41.79 (26.76)
70.03 (36.42) B78.03 (31.31)
48.13 (27.14) !39.90 (25.59)
Same-race
Defense attorneys
With question 1, although the main effect of familiarity was significant, the attorney group x familiarity interaction was not significant.
With question 3, only the attorney group x familiarity interaction was significant.
a
3b
2
1a
Question
Prosecutors
Table 5. Summary of mean responses (with SD) for the interactions of attorney group by cross-race condition and attorney group by familiarity
condition in study 1B.
14
K. Pezdek and M. O’Brien
Downloaded by [matthew o'brien] at 18:35 10 April 2013
Psychology, Crime & Law
15
Downloaded by [matthew o'brien] at 18:35 10 April 2013
General discussion
This study compared appraisals of the strength of eyewitness evidence by prosecutors
and defense attorneys (Study 1A) and assessed how decisions relevant to plea
bargaining cases are affected by two of these eyewitness factors (Study 1B). The
results of both studies present a generally consistent pattern of responding for
prosecutors and defense attorneys. First, in Study 1A, both prosecutors and defense
attorneys generally considered all of the eyewitness factors to be influential in
affecting eyewitness memory in real-world cases, and, on the majority of the 17 items,
ratings by prosecutors and defense attorneys did not differ. However, when there
were differences between the ratings of prosecutors and those of defense attorneys,
higher ratings of influence were reported by defense attorneys than prosecutors.
Regarding the role of the cross-race eyewitness factor, in Study 1A, defense
attorneys considered this factor to be more influential on the accuracy of real-world
eyewitness identifications than did prosecutors (Table 2, question 4). The central
issue in Study 1B is addressed by the interactions of (1) attorney group #cross-race
factor and (2) attorney group#familiarity factor in responses to questions 1, 2, and
3. In Table 5, are the means for these first-order interactions involving attorney
group. As can be seen in the left half of Table 5, defense attorneys, but not
prosecutors, responded that the defendant was more likely to be guilty in a same-race
than cross-race condition. However, the responses to questions 2 and 3 showed a
consistent pattern of responses for the two attorney groups. Although judgments
regarding winning the case (question 2) in same- and cross-race conditions did
interact with attorney group, the substantive direction of the effect was actually the
same for both attorney groups given that ‘winning the case’ was defined as getting a
guilty verdict for prosecutors and getting an acquittal for defense attorneys. That is,
prosecutors thought they were more likely to get a guilty verdict and defense
attorneys thought they were less likely to get an acquittal in the same- than cross-race
condition. Willingness to plea bargain (question 3) in same- and cross-race
conditions did not interact with attorney group. Together, these results suggest
that the effects of the cross-race eyewitness factor on anticipated trial outcomes and
plea bargaining were generally consistent for prosecutors and defense attorneys.
Regarding the role of the familiarity eyewitness factor, in Study 1A, defense
attorneys and prosecutors considered this factor to be equally influential in affecting
the accuracy of real-world eyewitness identifications (Table 2, question 9). As can be
seen in the right half of Table 5, on question 1 in Study 1B, the interaction of the
familiarity factor with attorney group was not significant; both attorney groups
thought the defendant was more likely to be guilty when the perpetrator was familiar
than unfamiliar. Although judgments regarding winning the case (question 2) in
familiar and unfamiliar conditions did interact with attorney group, the direction of
the effect was actually the same for both attorney groups given that ‘winning the
case’ was defined as getting a guilty verdict for prosecutors and getting an acquittal
for defense attorneys. That is, prosecutors thought they were more likely to get a
guilty verdict and defense attorneys thought they were less likely to get an acquittal
in the familiar than unfamiliar condition. Willingness to plea bargain (question 3) in
familiar and unfamiliar conditions did interact with attorney group but the direction
of the effect was actually the same for both attorney groups given that when evidence
against the defendant is strong, prosecutors are less likely to offer a plea and defense
Downloaded by [matthew o'brien] at 18:35 10 April 2013
16
K. Pezdek and M. O’Brien
attorneys are more likely to recommend one. Accordingly, on question 3, prosecutors
were less likely to offer a plea and defense attorneys were more likely to recommend
one when the perpetrator was familiar to the eyewitness. Together, these results
suggest that the effects of the familiarity eyewitness factor on anticipated trial
outcomes and plea bargaining were generally consistent for prosecutors and defense
attorneys.
Regarding the main effects of attorney group in Study 1B, on each of the three
questions, prosecutors provided significantly higher ratings than defense attorneys.
On question 1, prosecutors (M !0.85) were more likely to think that the defendant
was guilty than were defense attorneys (M !0.37). On question 2, prosecutors (M !
0.75) were more likely than defense attorneys (M !0.49) to think they would win the
case if it went to trial. And, on question 3, prosecutors (M !0.74) were more likely
to offer a plea bargain to the defendant than were defense attorneys to recommend
one (M !0.44). These results suggest that prosecutors feel they are more in control
of what is likely to happen in a trial and thus are generally more likely to offer a plea
bargain " although a less lenient one " than are defense attorneys. These results are
important for defense attorneys to know as they enter into plea bargain negotiations
for their clients. These findings are also consistent with the fact that prosecutors have
an ethical obligation not to prosecute a charge that they know is not supported by
probable cause (see Brady v. Maryland, 1963). Our findings suggest that prosecutors
are generally abiding by this principle.
Caveats to consider
There are several caveats. First, although every office of the District Attorney and
Public Defender in California was repeatedly contacted to participate, the final
sample size included only 93 Deputy Public Defenders and 46 Deputy District
Attorneys from eight counties. Anyone doing research with practicing attorneys is
aware of the difficulty getting them to participate. In fact, a recent study by Edkins
(2011) included as participants a total of only 101 defense attorneys and no
prosecutors for this specific reason, and the archival analysis by Flowe, Mehta, and
Ebbesen (2011) included data from the District Attorney’s office in only one
California county. However, although the sample size in the current study was small,
the total number of participants exceeded that required to detect effects with an effect
size of 0.25, a !0.05 and power !0.80. Nonetheless, the generalizability of these
findings may be restricted by the low response rate.2 Furthermore, given the
difficulty obtaining volunteers for this study, and the fact that it was not possible to
randomly select participants, there may have been biases in the obtained samples.
However, based on the data obtained regarding each participant’s trial experience
(see Table 1), the two samples do not differ. Thus, although we may not be able to
generalize these findings to all felony trial attorneys, it is reasonable to assume that
we can generalize these findings to the population of attorneys who have relatively
more felony trial experience, especially with cases that involve eyewitness evidence.
Second, because this study was a within-subject design, and each attorney
responded to all four scenarios, there is the possibility that some priming occurred
and the various responses were not independent. However, the absence of any effects
of the order of presenting the scenarios minimizes these concerns. Third, although
four of the six interactions involving the eyewitness factors and attorney group were
Psychology, Crime & Law
17
significant (see Table 5), the effect sizes for two of these interactions were small (h2 !
0.01 and 0.02), one was medium (h2 !0.09) and one was large (h2 !0.14), Cohen
(1988). Thus, although these interactions are statistically significant, the practical
value of these effects in the real world is mixed. Fourth, Study 1B included only two
eyewitness factors and only one type of crime. Additional research is necessary to
determine the eyewitness factors and types of crimes most likely to yield similarities
between prosecutors and defense attorneys and those likely to yield differences. Also,
because the scenarios used were each only one-page in length, research on this topic
using more complex case details should be conducted.
Downloaded by [matthew o'brien] at 18:35 10 April 2013
Policy implications
The purpose of plea bargaining is to reduce the number of lengthy criminal trials.
Plea bargaining does this by allowing defendants to plead guilty to a particular
charge in return for some concession from the prosecutor to avoid the risk of
conviction on a more serious charge at trial. In light of the recent Supreme Court
rulings of Missouri v. Frye and Lafler v. Cooper, attorneys " especially defense
attorneys " will play an increasingly important role in counseling defendants in plea
negotiations. Currently, the plea bargaining process is typically done behind closed
doors, often with only oral negotiations leading to defendants’ sentences. The recent
Supreme Court decisions put a greater onus on attorneys involved in plea
negotiations because defendants who reject a favorable plea offer due to bad legal
advice are heretofore protected. Attorneys’ knowledge of the importance of
eyewitness factors will become even more salient, as their lack of knowledge could
lead to an unfair plea bargain and rulings of ineffective counsel. The findings from
this study confirm that plea bargaining decisions are made ‘in the shadow of the trial’
at least for defense attorneys. It will thus be increasingly important for attorneys to
understand the factors that actually influence the accuracy of eyewitness memory.
Although the courts assume that attorneys understand the factors that influence
the fairness of identification procedures (Kirby v. Illinois, 1972; United States v.
Wade, 1967), Wise et al. (2009) suggested that inaccurate appraisals of eyewitness
testimony by prosecutors have led to ‘25 more years of needless wrongful convictions
from eyewitness error’ (p. 1280). The results of the present study are more
encouraging; prosecutors and defense attorneys were more consistent in their ratings
of the role of eyewitness factors. Both prosecutors and defense attorneys considered
most of the eyewitness factors listed to be influential in affecting eyewitness memory
in real-world cases, and, when there were differences between the ratings of
prosecutors and defense attorneys, higher ratings of influence were only modestly
more likely to be reported by defense attorneys than prosecutors. These results
suggest that significant differences between ratings of eyewitness factors by
prosecutors and defense attorneys may be subsiding in recent years, perhaps because
of an increase in general knowledge on the topic.
These findings suggest that defendants will benefit and legal counsel will be more
effective if additional training is available regarding factors that affect the accuracy
of eyewitness memory and identification. Information about the factors that affect
the accuracy of eyewitness identification should be included in clinical practice
courses at both law school and bar admissions, and should be regularly available in
continuing education courses. Also, given the variability in responses across
18
K. Pezdek and M. O’Brien
attorneys in this study (see Figure 1), prior to considering pretrial settlement,
attorneys would do well to obtain third-party views on the likelihood that the
eyewitness evidence in specific cases is accurate. These suggestions are especially
important in the climate of ‘the disappearing trial.’
Acknowledgments
This research was supported by Grant # 2009-IJ-CX-0019 to the first author from the
National Institute of Justice. The authors thank Attorneys John Brock and Jeri Polen for their
help in developing and pilot testing the materials for both studies, Edward P. Teyber for
assistance with legal research, and the District Attorneys and Public Defenders who allowed
us to distribute materials to their deputy attorneys.
Downloaded by [matthew o'brien] at 18:35 10 April 2013
Notes
1. In 1994, Proposition 184, commonly known as the ‘Three Strikes Law,’ was approved by
California legislators and voters. As its name suggests (in reference to the fact that in
baseball, three strikes and you are out) the law requires, among other things, a minimum
sentence of 25 years to life for three-time repeat offenders with multiple prior serious or
violent felony convictions. Relevant to the present study it is important to note that in
considering plea bargains, accepting a strike can have considerable consequences.
2. We have no way of knowing the exact response rate for either sample because it is not a
matter of public record how many felony attorneys there are in each Public Defenders
Office and Prosecutors’ Office nor do we have any way of knowing first hand whether our
contact person in each office actually distributed the materials to all felony attorneys in that
office. Nonetheless, given the low number of returns that we received, it is clear that the
percent return rate was quite low.
References
Bibas, S. (2004, June). Plea bargaining outside the shadow of trial. Harvard Law Review, 17,
2463"2547. doi:10.2139/ssrn.464880
Brady v. Maryland, 83 S.Ct. 1194 (1963).
Brigham, J. C. & WolfsKeil, M. P. (1983). Opinions of attorneys and law enforcement
personnel on accuracy of eyewitness identifications. Law and Human Behavior, 7, 337"349.
doi:10.1007/BF01044736
Bureau of Justice Statistics. (2005). Sourcebook of criminal justice statistics. Washington, DC:
U.S. Department of Justice.
Burke, A. S. (2007). Prosecutorial passion, cognitive bias, and plea bargaining. Marquette Law
Review, 91, 183"212.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed). Hillsdale, NJ:
Erlbaum.
Covey, R. (2007). Reconsidering the relationship between cognitive psychology and plea
bargaining. Marquette Law Review, 91, 213"248.
Edkins, V. A. (2011). Defense attorney plea recommendations and client race: Does zealous
representation apply equally to all? Law and Human Behavior, 35, 413"425. doi:10.1007/
s10979-010-9254-0
Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power 3: A flexible statistical
power analysis program for the social, behavioral, and biomedical sciences. Behavior
Research Methods, 39, 175"191. doi:10.3758/BF03193146
Flowe, H. D., Mehta, A., & Ebbesen, E. B. (2011). The role of eyewitness identification
evidence in felony case dispositions. Psychology, Public Policy, and Law, 17, 140"159.
doi:10.1037/a0021311
Downloaded by [matthew o'brien] at 18:35 10 April 2013
Psychology, Crime & Law
19
Goodman-Delahunty, J., Granhag, P. A., Hartwig, M., & Loftus, E. F. (2010). Insightful or
wishful: Lawyers’ ability to predict case outcomes. Psychology, Public Policy, and Law, 15,
133"157. doi:10.1037/a0019060
Hagar v. United States, 856 A.2d 1143 (D.C. Cir. 2004).
Higginbotham, P. E. (2004). The disappearing trial and why we should care. Rand Review.
Retrieved from http://www.rand.org/publications/randreview/issues/summer2004/28.html
Huff, C. R. (1987). Wrongful conviction: Societal tolerance of injustice. Research in Social
Problems & Public Policy, 4, 99"115.
Huff, C. R., Rattner, A., & Sagarin, E. (1996). Convicted but innocent: wrongful conviction and
public policy. Thousand Oaks, CA: Sage Publications.
Kassin, S. M., Tubb, V. A., Hosch, H. M., & Memon, A. (2001). On the ‘‘general acceptance’’
of eyewitness testimony research: A new survey of the experts. American Psychologist, 56,
405"416. doi:10.1037/0003-066X.56.5.405
Kirby v. Illinois, 406 U.S. 682 (1972).
Kramer, G. M., Wolbransky, M., & Heilbrun, K. (2007). Plea bargaining recommendations by
criminal defense attorneys: Evidence strength, potential sentence, and defendant preference.
Behavioral Sciences and the Law, 25, 573"585. doi:10.1037//0003-066X.56.5.405
Lafler v. Cooper, No. 10-209, 2012 BL 67236 (U.S. Mar. 21, 2012).
Landes, W. (1971). An economic analysis of the courts. Journal of Law and Economics, 14, 61"
108. Retrieved from http://dx.doi.org/10.1086/466704
Libuser, M. R. (2001). The investigation of case severity and plea bargaining decisions in rapes
and robberies using archival and survey data (Ph.D. dissertation). University of California,
San Diego, California. Retrieved from Dissertation & Theses: Full Text database.
Lindholm, T. (2008). Who can judge the accuracy of eyewitness statements? A comparison of
professionals and lay-persons. Applied Cognitive Psychology, 22, 1"14. doi:10.1002/acp.1439
Lindsay, R. C. L., MacDonald, P., & McGarry, S. (1990). Perspectives on the role of the
eyewitness expert. Behavioral Sciences and the Law, 8, 457"464. doi:10.1002/bsl.2370080411
McAllister, H. A. (1990). Effects of eyewitness evidence on plea-bargain decisions by
prosecutors and defense attorneys. Journal of Applied Social Psychology, 20, 1461"1473.
doi:10.1111/j.1559-1816.1990.tb01487.x
Meissner, C. & Brigham, J. C. (2001). Thirty years of investigating the own-race bias in
memory for faces: A meta-analytic review. Psychology, Public Policy, and Law, 7, 3"35.
doi:10.1037/1076-8971.7.1.3
Missouri v. Frye, No. 10-444, 2012 BL 67235 (U.S. Mar. 21, 2012).
Penrod, S. D. & Cutler, B. (1999). Preventing mistaken convictions in eyewitness identification
trials. In R. Roesch, S. D. Hart, & J. R. P. Ogloff (Eds.), Psychology and law: The state of the
discipline (pp. 89"118). New York: Kluwer Academic.
Pezdek, K. (2009). Content, form and ethical issues concerning expert psychological testimony
on eyewitness identification. In B. L. Cutler (Ed.), Expert testimony on the psychology of
eyewitness identification (pp. 29"50). New York: Oxford University Press. doi:10.1093/
acprof:oso/9780195331974.003.012
Pezdek, K. (2007). Expert testimony on eyewitness memory and identification. In M.
Costanzo, D. Krauss, & K. Pezdek (Eds.), Expert psychological testimony for the courts
(pp. 99"117). Mahwah, NJ: Erlbaum. doi:10.1093/acprof:oso/9780195331974.003.002.
Pezdek, K., Avila-Mora, K., & Sperry, K. (2010). Does trial presentation medium matter in
jury simulation research?: Evaluating the effectiveness of eyewitness expert testimony.
Applied Cognitive Psychology, 24, 673"690. doi:10.1002/acp.1578
Pezdek, K., Blandon-Gitlin, I., & Moore, C. (2003). Children’s face recognition memory:
More evidence for the cross-race effect. Journal of Applied Psychology, 88, 760"763.
doi:10.1037/0021-9010.88.4.760
Pritchard, D. (1986). Homicide and bargained justice: The agenda setting effect of crime news
on prosecutors. Public Opinion Quarterly, 50, 143"159. doi:10.1086/268971
Shapiro, P. N. & Penrod, S. (1986). Meta-analysis of facial identification studies. Psychological
Bulletin, 100, 139"156. doi:10.1037/0033-2909.100.2.139
Stinson, V., Devenport, J. L., & Kravitz, D. A. (1996). How effective is the presence-of-counsel
safeguard? Attorney perceptions of suggestiveness, fairness, and correctability of biased
lineup procedures. Journal of Applied Psychology, 81, 64"75. doi:10.1037/0021-9010.81.1.64
20
K. Pezdek and M. O’Brien
Toglia, M. P., Read, R. D., Ross, D. F., & Lindsay, R. C. L. (Eds.). (2006). The handbook of
eyewitness psychology. Mahwah, NJ: Lawrence Erlbaum Associates.
United States v. Wade, 388 U.S. 218 (1967).
Wells, G. L., Memon, A., & Penrod, S. D. (2006). Eyewitness evidence: Improving its probative
value. Psychological Science in the Public Interest, 7, 45"75. doi:doi:10.1111/j.15291006.2006.00027.x
Wells, G. L. & Olsen, E. A. (2003). Eyewitness testimony. Annual Review of Psychology, 54,
277"295. doi:10.1146/annurev.psych.54.101601.145028
Wells, G. L., Small, M., Penrod, S. D., Malpass, R. S., Fulero, S. M., & Brimacombe, C. A. E.
(1998). Eyewitness identification procedures: Recommendations for lineups and photospreads. Law and Human Behavior, 22, 603"647. doi:10.1023/A:1025750605807
Wise, R. A., Pawlenko, N. B., Safer, M. A., & Meyer, D. (2009). What US prosecutors and
defence attorneys know and believe about eyewitness testimony. Applied Cognitive
Psychology, 23, 1266"1281. doi:10.1002/acp.1530
Downloaded by [matthew o'brien] at 18:35 10 April 2013
Appendix. Sample scenario from Study 1B with questions worded for defense attorneys
Important: In each scenario, you should assume that a very thorough investigation of the case
has uncovered no more information besides what is presented in the scenario, and there was no
video operating in the store. Also, the defendant is not a minor, and he has no priors. He
denies involvement but has no solid alibi.
A neighborhood convenience store in East Los Angeles was robbed on 3 February 2009.
The robber approached the clerk from behind and although no weapon was present, he forced
the clerk to empty the cash drawer into his gym bag. As the robber ran from the store, the clerk
saw him across the counter, face-to-face, for a couple of seconds; this was his only opportunity
to see the robber. The clerk was a 22-year-old Hispanic male who, at the scene of the crime,
told the police that he had seen the robber a couple of times in the neighborhood last summer.
He described the robber as a Black teenager dressed in dark clothing. Three weeks later the
clerk called the police because he thought he saw the suspect at the neighborhood gas station.
When the police arrived they apprehended the defendant after the clerk told them, ‘yes, that’s
the guy.’ One count of robbery has been filed against the defendant.
1. In the above scenario, what do you think is the probability that the defendant is guilty?
Specify on a scale from 0 to 100% ___________.
2. In the above scenario, what is the probability that you would win this case if it went to
trial?
Specify on a scale from 0 to 100% ________.
3. In the above scenario, what is the probability that you would recommend any plea
bargain to the defendant? Specify on a scale from 0 to 100% _________.
4. From the list of seven potential plea bargain offers specified below, circle the very highest
offer you would recommend to the defendant in this scenario.
No plea
No strike
probation
No strike
16 months
One strike
probation
One strike
2 years
One strike
3 years
One strike
5 years