Graduate Theses and Dissertations
Iowa State University Capstones, Theses and
Dissertations
2019
Victims as aggressors: Does victim-bystander status
influence eyewitness decision-making in showup
procedures?
Ryan Edward Ditchfield
Iowa State University
Follow this and additional works at: https://lib.dr.iastate.edu/etd
Part of the Law Commons, and the Social Psychology Commons
This Thesis is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital
Repository. It has been accepted for inclusion in Graduate Theses and Dissertations by an authorized administrator of Iowa State University Digital
Repository. For more information, please contact digirep@iastate.edu.
Recommended Citation
Ditchfield, Ryan Edward, “Victims as aggressors: Does victim-bystander status influence eyewitness decision-making in showup
procedures?” (2019). Graduate Theses and Dissertations. 17001.
https://lib.dr.iastate.edu/etd/17001
Victims as aggressors: Does victim-bystander status influence
eyewitness decision-making in showup procedures?
by
Ryan E. Ditchfield
A thesis submitted to the graduate faculty
in partial fulfillment of the requirements for the degree of
MASTER OF SCIENCE
Major: Psychology
Program of Study Committee:
Max E. Guyll, Co-major Professor
Stephanie Madon, Co-major Professor
Jason C.K. Chan
Gary L. Wells
The student author, whose presentation of the scholarship herein was approved
by the program of study committee, is solely responsible for the
content of this thesis. The Graduate College will ensure this thesis is globally
accessible and will not permit alterations after a degree is conferred.
Iowa State University
Ames, Iowa
2019
ii
TABLE OF CONTENTS
Page
LIST OF TABLES ……………………………….
…………………………………………… iv
LIST OF FIGURES ………………………………. …………………………………………. vi
ACKNOWLEDGMENTS………………………………. …………………………………. vii
ABSTRACT……………………………….
…………………………………………………….. viii
CHAPTER 1
INTRODUCTION ……………………………………………………………….. 1
The Use of Showup Procedures in the United States ………………………………….. 2
Case Law Related to Showups
…………………………………………………………….. 4
Factors Influencing Eyewitness Identification ……………………………………………. 5
Estimator Variables
……………………………………………………………………………. 5
System Variables ……………………………………………………………………………… 6
Eyewitness Decision-Making ………………………………………………………………….. 6
Decision-Making at Identification ………………………………………………………. 7
Post-Identification Judgments
…………………………………………………………….. 9
Victim-Bystander Status and the Frustration-Aggression Hypothesis ……………. 10
Aggression and Eyewitness Decision-Making ……………………………………… 11
Review of Previous Victim-Bystander Research
…………………………………………. 14
Hypotheses …………………………………………………………………………………………… 17
CHAPTER 2
METHOD ………………………………………………………………………….. 20
Methods
…………………………………………………………………………………………… 20
Materials
…………………………………………………………………………………………… 21
Procedure …………………………………………………………………………………………… 26
iii
CHAPTER 3 RESULTS …………………………………………………………………………… 31
Preliminary Analyses ………………………………………………………………………………. 31
Negative Affect
………………………………………………………………………………… 31
Eyewitness Motivation Measures ……………………………………………………….. 32
Primary Analyses …………………………………………………………………………………… 34
Decision-Making at Identification ………………………………………………………. 35
Post-Identification Judgments
……………………………………………………………… 39
Anger as a Mediator of Post-Identification Judgments
……………………………. 44
CHAPTER 4
DISCUSSION ……………………………………………………………………… 46
Aggression, Victim-Bystander Status and Eyewitness Decision-Making
………. 47
Culprit Presence and Eyewitness Decision-Making ……………………………………. 55
Limitations …………………………………………………………………………………………… 60
Conclusion ……………………………………………………………………………………………. 63
REFERENCES ……………………………………………………………………………………………. 65
TABLES
……………………………………………………………………………………………… 70
FIGURES
……………………………………………………………………………………………… 88
APPENDIX A
DEMOGRAPHICS QUESTIONNAIRE
………………………………… 90
APPENDIX B BOGUS COMMUNICATION TASK
……………………………………. 91
APPENDIX C CULPRIT AND INNOCENT SUSPECT PHOTOGRAPHS …….. 94
APPENDIX D SUSPICION CHECK FORM ……………………………………………….. 95
APPENDIX E BOGUS CAMPUS THEFT REPORT ……………………………………. 96
APPENDIX F SHORTENED PROFILE OF MOOD STATES ………………………. 99
APPENDIX G
MANIPULATION CHECKS ……………………………………………….. 100
APPENDIX H
DEBRIEFING
……………………………………………………………………… 101
APPENDIX I IRB APPROVAL LETTER
…………………………………………………… 102
iv
LIST OF TABLES
Page
Table 1 Means for Self-Reported Retrospective Negative Affect …………………….… 70
Table 2 MANOVA Analysis of Effects of Manipulations on
Retrospective Negative Affect ……………….
…………………………………………….. 71
Table 3 Frequencies and Means for Self-Reported Retrospective
Motivations to Avoid Identification Errors ……………………………………………
72
Table 4 Using Logistic Regression to Test Effects of Manipulations on
Dichotomous Motivation Measure ……………. …………………………………………..
73
Table 5 MANOVA Analysis of Effects of Manipulations on
Continuous Motivation Measures ………………………………………………. 74
Table 6 Identification Rates and Signal Detection Measures……………………………. 75
Table 7 Using Logistic Regression to Test Effects of Manipulations on
Identification Rates ………. ……………………………………………………………………….
76
Table 8 Descriptive Statistics for Immediate Confidence Ratings and
Response Latency …………. ……………………………………………………………………..
77
Table 9 Two-Way ANOVAs Examining Effects of Manipulations on
Immediate Confidence Ratings and Response Latency
for Identifications and Rejections ……………………. ………………………………….
78
Table 10 Means for Post-Identification Judgments of Ability to Identify the
Culprit by Eyewitnesses Who Made a Positive Identification……………………….
79
Table 11 Means for Post-Identification Judgments of Perceived
Decision-Making Performance and Willingness to Testify
by Eyewitnesses Who Made a Positive Identification ……………….
……………..
80
Table 12 MANOVA Analysis of Post-Identification Judgments of
Ability to Identify the Culprit by Eyewitnesses Who Made
a Positive Identification ……………………. ……………………………………………….
81
Table 13 MANOVA Analysis of Post-Identification Judgments of
Perceived Decision-Making Performance by Eyewitnesses Who Made
a Positive Identification ………………. ………………………………………………………
82
v
Table 14 ANOVA Analysis of Post-Identification Willingness to Testify
by Eyewitnesses Who Made a Positive Identification …………………………. 83
Table 15 Mediation Effects of Anger on the Relationship between
Experimental Effects and Attention Paid to the Culprit …………………….… 84
Table 16 Mediation Effects of Anger on the Relationship between
Experimental Effects and Number of Specific Facial Features
of the Culprit Attended To …………………….……………………………… 85
Table 17 Mediation Effects of Anger on the Relationship between
Experimental Effects and Clarity of Culprit’s Face …………………………… 86
Table 18 Mediation Effects of Anger on the Relationship between
Experimental Effects and Willingness to Testify in Court ……………………. 87
vi
LIST OF FIGURES
Page
Figure 1 A graphical representation of identification rates across culprit presence and
victim-bystander status ……………………………………………………….
88
Figure 2 A graphic representation of victims’ and bystanders’ retrospective
mood scores on each of the six subscales in the Shortened Profile
of Mood States Scale …………………………………………………………
89
vii
ACKNOWLEDGMENTS
I would like to thank my advisors, Dr. Max Guyll and Dr. Stephanie Madon, and my
committee members, Dr. Gary Wells and Dr. Jason Chan, for their guidance and support
throughout the course of this research. In addition, I would like to thank all of the research
assistants who devoted their time to running my experiment, and those who took the time to
participant in my experiment for course credit.
I would also like to thank my partner, Tiffany Louk, and my good friends here at Iowa
State University—Anna Wehde, Tyler Knaplund, and Andreas Miles-Novelo—for their
encouragement, support, willingness to listen to my rants about research, and for reminding me
to take a break every now and then. Finally, I would like to thank my parents, Ashleigh and Jay
Fowler, for helping me to get this far and for supporting me at every step of the way.
viii
ABSTRACT
Police rely on eyewitness identifications to solve crimes, but eyewitnesses can make
mistakes. These mistakes in decision-making can have serious consequences for the criminal
justice system. Eyewitnesses can erroneously identify an innocent suspect as the culprit, which
could result in a wrongful conviction, or fail to identify a guilty suspect as the culprit, which
could result in the culprit avoiding punishment. Drawing on Berkowitz’s (1989) frustration-
aggression hypothesis, the current study tested whether eyewitnesses’ status as either a victim or
a bystander influenced their decision-making processes in a showup procedure after eliminating
attention and encoding as possible mediators of the victim-bystander status manipulation.
Although victims reported significantly higher feelings of anger than bystanders, victims did not
significantly differ from bystanders in identification rates, confidence ratings, response latency,
or motivation to catch the guilty culprit. In addition, victims did not significantly differ from
bystanders in their ability to distinguish between guilty culprits and innocent suspects. However,
analysis of post-identification judgments revealed that victims who made identified innocent
suspects reported paying significantly more attention to the culprit, recalling significantly more
specific facial features of the culprit, and having a significantly clearer image of the culprit’s face
than bystanders. In addition, victims reported being significantly more willing to testify in court
regarding their identification decision than bystanders. These findings suggest that victims may
increase the potency of their positive identifications by bolstering responses to post-identification
judgments. Consequently, victims may give the appearance of having a greater ability to
accurately identify guilty culprits than bystanders, even in the absence of true differences.
1
CHAPTER 1. INTRODUCTION
Police rely on eyewitness identifications to solve crimes, but psychological research and
DNA exoneration cases have shown that eyewitness identification evidence is often unreliable
(Innocence Project, 2019; Wells & Olson, 2003). Nevertheless, eyewitness testimony remains an
important source of evidence in the criminal justice system. Understanding how eyewitnesses
make decisions will deepen current understanding of eyewitnesses’ potential for error and allow
the field to develop best practices. Using a novel experimental paradigm in which participants
believed that a real crime had occurred, the current study aimed to test whether an eyewitness’
status as either a victim or a bystander influences decision-making in the manner predicted by
Berkowitz’s (1989) frustration-aggression hypothesis. To limit the effects under consideration to
eyewitness decision-making during and after the identification procedure, the victim-bystander
status manipulation was delayed until immediately prior to the showup procedure. This
eliminated attention to and encoding of the culprit as possible mediators of the victim-bystander
status manipulation. In addition, this study examined decision-making in the context of showups:
an identification procedure in which an eyewitness is presented with a single suspect and the
eyewitness then decides whether the person presented is the culprit (Smith, Wells, Lindsay, &
Penrod, 2017).
Using Berkowitz’s (1989) revised frustration-aggression hypothesis as a theoretical
framework, positive identifications during an eyewitness procedure can be defined as aggressive
behavior. If victims experience a criminal event as more frustrating than bystanders, victims
should experience increased anger and an increased desire to resolve the criminal event by
making a positive identification. In addition, victims’ increased aggression should lead to
2
bolstering of self-reported ability to accurately identify the culprit and self-reported decision-
making performance. Bolstering increases the potency of a positive identification by making an
identification appear more trustworthy to investigators, thereby increasing the likelihood that the
criminal event is resolved. Three objectives were developed to test this theoretical framework.
The first objective was to test whether victim-bystander status influenced decision-making at the
time of identification, specifically, identification rates, confidence ratings, and response latency.
The second objective was to test whether victim-bystander status influenced post-identification
judgments of self-reported ability to identify the culprit and self-reported certainty in positive
identifications. The third objective was to test whether the effects of victim-bystander status on
decision-making were mediated by anger and desire to resolve the criminal event.
By addressing these objectives, this research advances understanding about eyewitness
decision-making processes and examines implications for identification performance and
eyewitness testimony. The current chapter discusses case law and research relevant to police
showup procedures, outlines factors influencing eyewitness identification, and provides an
overview of eyewitness decision-making and its relation to Berkowitz’s (1989) frustration-
aggression hypothesis.
The Use of Showup Procedures in the United States
The majority of eyewitness identification studies have focused on the police lineup: a
procedure in which a criminal suspect is placed among known innocents (i.e., fillers) and is then
shown to an eyewitness to determine if the eyewitness can identify the suspect as the culprit
(Wells & Olson, 2003). This study focused on a less-studied procedure: the showup. A showup is
an identification procedure in which an eyewitness is presented with a single suspect and the
eyewitness then decides whether the person presented is the culprit (Smith et al., 2017). Recent
3
behavioral science research has demonstrated that showups are suggestive and promote worse
applied outcomes than lineups (Smith et al., 2017). It is important to note, however, that lineups
are not superior to showups because they improve eyewitnesses’ ability to distinguish between
innocent suspects and guilty culprits. Indeed, showups have been shown to produce more correct
rejections than lineups (Smith et al., 2017). Instead, the lower rate of innocent suspect
identifications in lineups can be attributed to differential filler siphoning, whereby false alarms
are spread away from the innocent suspect and toward the lineup fillers (Smith et al., 2017). As
showups have no fillers to draw away false alarms from an innocent suspect, showups result in
more false identifications of innocent suspects than lineups.
Despite this evidence, showups are still used by approximately 62% of police agencies in
the United States. This makes showups second only to photo lineups (used by over 94% of
agencies), and considerably more common than other identification procedures such as
composite sketches (35.5%), mugshot searches (28.8%), and live lineups (21.4%) (Police
Executive Research Forum, 2013). The argument by police in favor of showups is largely
practical. To be detained long enough to conduct a live lineup, suspects need to be placed under
arrest. Likewise, creating a lineup and presenting it to a witness is time consuming, even if
photos of the suspect and potential fillers already exist in police records. Showups, on the other
hand, can be conducted quickly and do not require the suspect to be detained for a long period of
time (Smith, Wells, Lindsay, Myerson, Kovera, & McAuliff, 2018). Furthermore, even a
detained, arrested suspect can agree to a showup rather than a lineup. These facts mean that
showups will likely not be abandoned by police departments in the near future. As such,
showups remain an important area of study and researchers should continue to examine showups
and develop procedures to improve their diagnostic value.
4
Case Law Related to Showups
The Supreme Court recognized the substantial risks of bias in lineups and showups as
early as the 1960s. In a series of cases known as the Wade Trilogy (United States v. Wade, 1967;
Gilbert v. California, 1967; Stovall v. Denno, 1967), the Court established the totality of the
circumstances standard for admissibility of showup and lineup evidence in court (Agricola,
2009). Notably, the Court considered the dangers of suggestion to be stronger in showups.
In the 1970s, however, the Court established a new view of the totality of the
circumstances standard in two seminal cases. In Neil v. Biggers (1972), the Court outlined five
factors to be considered when evaluating the reliability and admissibility of showups: (1)
opportunity of the witness to view the criminal at the time of the crime, (2) the witness’s degree
of attention at the time of the crime, (3) the accuracy of the witness’s prior description of the
criminal, (4) the level of certainty demonstrated by the witness at identification, and (5) the
length of time between the crime and the identification. The Court then reaffirmed this new
standard in Manson v. Brathwaite (1977), ruling that a flawed procedure did not necessarily
invalidate an identification when the procedure satisfied all or some of these guidelines. In
practice, however, the courts have rarely found a showup to be so suggestive as to warrant a
reversal of a criminal conviction; few, if any, showups are ever excluded under the current
standards (Cicchini & Easton, 2010).
Furthermore, the totality of the circumstances standard has not survived scientific
scrutiny. Research on eyewitness identification has demonstrated that, when compared to lineup
procedures, showups lead to more innocent suspect identifications and similar rates of guilty
culprit identifications (Steblay, Dysart, Fulero, & Lindsay, 2003; Wetmore, Neuschatz,
Gronlund, Wooten, Goodsell, & Carlson, 2015). This has led experts to recommend additional
5
legal reforms and has even led some experts to conclude that showups should not be allowed as
evidence in court (e.g., Agricola, 2009). Recently, New York, Massachusetts, and Wisconsin
have adopted reforms that prohibit showups as evidence unless police lacked probable cause to
arrest and detain the suspect long enough to construct a photo lineup, or exigent circumstances
prevented the use of a photo lineup (Cicchini & Easton, 2010). However, these reforms are not
intended to prevent the use of showups entirely. Although the legal system has recognized
potential problems with showups, showups will likely continue to be used under the legal
presumption that they have probative value. Accordingly, it remains important to study the
psychological processes that operate during showups.
Factors Influencing Eyewitness Identification
To better understand the processes that operate during eyewitness identification, a
taxonomy has been developed to categorize factors that influence eyewitness performance
(Wells, 1978; Wells & Olson, 2003). Factors outside the justice system’s control are known as
estimator variables; factors under the control of the criminal justice system are known as system
variables. The following section outlines relevant research on these factors.
Estimator Variables
Two categories of estimator variables can influence eyewitness accuracy: characteristics
of the eyewitness (e.g., the cross-race effect) and characteristics of the witnessed event (e.g., a
suspect wearing a disguise; Wells & Olson, 2003). Most estimator variable research has
examined the effects of these variables using lineups; few studies have examined their effects
using showups. One estimator variable that has been examined using showups is clothing bias.
Clothing bias occurs when a suspect appears in the identification procedure wearing clothing
similar to the culprit’s clothing described by eyewitnesses, and has been shown to increase false
6
identifications in lineups, mug shots, and showups (Dysart, Lindsay, & Dupuis, 2006; Lindsay,
Wallbridge, & Drennan, 1987; Lindsay, Nosworthy, Martin, & Martynuck, 1994; Yarmey,
Yarmey, & Yarmey, 1996). This demonstrates that estimator variables that negatively impact
eyewitness accuracy in lineups may also have detrimental effects in showups.
System Variables
System variables, on the other hand, can be organized into five categories: instructions to
the witness (e.g., instructing the witness that it is equally important to clear the innocent and
identify the guilty), content of the identification procedure (e.g., selection of fillers), presentation
method (e.g., simultaneous vs. sequential lineups), behavioral influence (e.g., use of double-blind
procedures), and base rates (e.g., how likely it is that a suspect is innocent) (Wells & Olson,
2003; Smalarz & Wells, 2014). For example, researchers and the Department of Justice have
recommended the use of a “may or may not be present” instruction (otherwise known as the
“standard admonition”) for nearly two decades (Clark, 2012; Technical Working Group for
Eyewitness Evidence, 1999). Employing the standard admonition in lineups has been shown to
reduce mistaken identifications in culprit-absent lineups by 41.6% while reducing accurate
identifications by only 1.9% (Steblay, 1997). Accordingly, system variables like the standard
admonition can provide relatively simple techniques for police to offset the limitations of
eyewitness procedures.
Eyewitness Decision-Making
In a showup procedure, a guilty culprit is either present or not, and an eyewitness either
identifies the guilty culprit or does not. Therefore, there are four possible outcomes: a hit (the
guilty culprit is present and the eyewitness makes a positive identification), a miss (the guilty
culprit is present and the eyewitness rejects the showup by refusing to make an identification), a
7
false alarm (an innocent suspect is present and the eyewitness makes a positive identification), or
a correct rejection (an innocent suspect is present and the eyewitness rejects the showup by
refusing to make an identification). Ideally, an eyewitness should always make a positive
identification when the guilty culprit is present and should always reject the showup when an
innocent suspect is present. Of course, eyewitnesses can and do make mistakes (Innocence
Project, 2019). Therefore, it is important to examine how eyewitness decision-making can lead to
errors. To this end, the current study examined eyewitnesses’ decision-making at the time of
identification and eyewitnesses’ post-identification judgments.
Decision-Making at Identification
Eyewitness decision-making outcomes at the time of an identification procedure include
identification rates, confidence ratings, and response latency. Identification rates reflect the ratio
of positive identifications to non-identifications in a sample of eyewitnesses. Any shift in
eyewitnesses’ identification rates will yield a corresponding tradeoff in costs and benefits
(Dobolyi & Dodson, 2013; Palmer & Brewer, 2012). Higher identification rates will increase hits
but also increase false alarms. Conversely, lower identification rates will decrease false alarms
but also decrease hits. However, it is important to note that as identification rates shift, the ratio
of correct decisions to errors remains the same. In other words, shifts in identification rates do
not correspond to changes in eyewitnesses’ ability to distinguish between guilty culprits and
innocent suspects (Palmer & Brewer, 2012). Instead, eyewitnesses’ ability to distinguish between
guilty culprits and innocent suspects is reflected by shifts in the ratio of hits to false alarms.
When hits are frequent and false alarms are rare, one can infer that eyewitnesses can easily
distinguish between guilty culprits and innocent suspects. On the other hand, when the ratio of
8
hits and false alarms are close to equal, one can infer that eyewitnesses are performing no better
than chance in a showup procedure.
Confidence is an eyewitness’ certainty in her or his identification decision. For an
eyewitness who makes a positive identification, confidence indicates her or his level of certainty
that the person she or he identified is the guilty culprit. For an eyewitness who rejects a showup,
confidence indicates certainty that the guilty culprit was not present. One can further distinguish
between immediate and delayed confidence. Immediate confidence refers to confidence ratings
given immediately following the showup procedure, whereas delayed confidence refers to
confidence ratings given in a post-identification setting, up to and including testimony in court.
Recent research has indicated that confidence ratings can shift as soon as 10 minutes after an
identification procedure (Eisen, Cedré, Williams, Jones, Kovera, & McAuliff, 2018). Therefore,
it is important to examine effects on both immediate and delayed confidence judgments.
Response latency is the time from when an eyewitness is first presented with a suspect to
when the eyewitness makes a final identification decision (Brewer, Caon, Todd, & Weber, 2006;
Weber, Brewer, Wells, Semmler, & Keast, 2004). Past research has shown that response latency
can provide useful information about eyewitness decision-making and performance. For
example, faster reaction times are reliably associated with higher accuracy, as stronger memory
traces of the culprit’s face are easier to retrieve than weak memory traces (Robinson, Johnson,
Herndon, & Murphy, 1997). However, the relationship between response latency and accuracy is
relatively weak and the time frame for accurate identifications is highly variable across
experiments and situations, ranging from 5 to 29 seconds (Brewer et al., 2006; Weber et al.,
2004). Furthermore, it is important to note that response latency is not the same as reaction time;
9
response latency is a less precise tool than reaction time and therefore has less ability to detect
effects on eyewitness decision-making (Brewer et al., 2006).
Post-Identification Judgments
In the real world, eyewitness decision-making does not end after an identification
decision is reached. Eyewitnesses may also be asked to testify in court regarding their
identification decision and the witnessed criminal event. Past research suggests that
eyewitnesses’ testimony is evaluated based on how confident eyewitnesses are, what
eyewitnesses claim about their viewing conditions, and how much attention eyewitnesses claim
they paid during the criminal event (Bradfield & Wells, 2000). When eyewitnesses are not given
feedback or influenced in any way, mock-jurors can accurately differentiate between correct and
mistaken eyewitness testimony (Smalarz, Wells, & Kovera, 2014). However, mock-jurors’
ability to distinguish between accurate and mistaken eyewitness testimony disappears when
eyewitnesses receive even minor confirming feedback about their identification decision (Eisen,
2018; Smalarz, Wells, & Kovera, 2014).
To date, eyewitness research has identified several system variables that influence post-
identification decision-making and eyewitness testimony, such as steering effects and post-
identification feedback (Eisen et al., 2018; Smalarz, Wells, & Kovera, 2014). However, it is also
possible that estimator variables can influence eyewitnesses’ perceptions regarding the witnessed
event and their ability to make an accurate identification. Indeed, there is no theoretical reason to
believe that estimator variable effects on eyewitness decision-making stop after an identification
decision is made. Therefore, it is important to examine effects on eyewitness decision-making
both during and after an identification procedure.
10
Decision-making at identification revolves around eyewitnesses’ decisions to make a
positive identification or reject the showup procedure. In contrast, decision-making in the post-
identification stage can be distilled down to two types of judgments: eyewitnesses’ perceived
ability to correctly identify the culprit and, among eyewitnesses who make a positive
identification, eyewitnesses’ perceived certainty in their identification decision. Eyewitnesses’
perceived ability to identify the culprit stems from perceived viewing conditions and memory for
the culprit at the time of the crime. Perceived certainty in a positive identification, on the other
hand, is reflected by judgments related to ease of identification, delayed confidence ratings, and
time needed to make an identification. Eyewitnesses’ judgments in both of these categories can
impact the nature and believability of their testimony.
Victim-Bystander Status and the Frustration-Aggression Hypothesis
Victim-bystander status is one estimator variable that could potentially influence
eyewitnesses’ decision-making during and after an identification. There is theoretical reason to
believe that victims are more likely to make an identification than bystanders. Under Berkowitz’s
(1989) revised formulation of the frustration-aggression hypothesis, frustrations are defined as
aversive events that people would normally seek to avoid and that give rise to aggressive
behavior. These unpleasant events produce negative affect (i.e., feelings that individuals seek to
reduce or eliminate), which in turn generates aggressive tendencies (Berkowitz, 1989). Although
Berkowitz posited that any kind of negative affect will produce aggressive inclinations, the
revised frustration-aggression hypothesis specifically highlights the role of anger. However, the
presence of an aversive event and feelings of anger are not sufficient by themselves to produce
aggressive behavior. Rather, aggression is likely to occur following aversive events if a suitable
target is nearby, if non-aggressive behaviors do not eliminate the aversive event, if restraints
11
against aggression are relatively weak, and if the aversive event is perceived as a deliberate
personal attack (Berkowitz, 1989).
Berkowitz’s (1989) frustration-aggression hypothesis has two implications for victim-
bystander status and eyewitness identification. First, crime can be viewed in the context of the
frustration-aggression hypothesis as a deliberate and unjustified aversive event. Although
bystanders may experience a criminal event as aversive, victims of a crime may additionally
experience it as a personal attack. Second, a positive identification of a suspect has potential
negative consequences for the suspect, regardless of the suspect’s guilt or innocence. Therefore,
identifying a suspect as the guilty culprit can be framed as an act of aggression. The following
section examines how the frustration-aggression hypothesis can inform predictions regarding the
effects of victim-bystander status on eyewitness decision-making, both during and after an
identification.
Aggression and Eyewitness Decision-Making
The following section considers the effects of victim-bystander status on eyewitness
aggression in terms of decision-making at identification and post-identification judgments. In
addition, this section examines two potential mediators of the relationship between victim-
bystander status and eyewitness aggression: anger and desire to resolve the aversive event.
Decision-making at identification. As noted earlier, there are typically three
components to eyewitness decision-making at the time of an identification: the identification
decision itself, immediate confidence ratings, and response latency. In terms of Berkowitz’s
(1989) frustration-aggression hypothesis, the offer of a potential identification of the culprit
allows both victims and bystanders a chance to eliminate the aversive event (i.e., by catching the
culprit). However, because the aversive event is more personal and frustrating for the victim than
12
the bystander, victims should experience greater anger and an increased desire to resolve the
aversive event. This increase in anger and desire to resolve the aversive event should lead to
victims having a higher tendency to aggress against the culprit by making a positive
identification. Therefore, victims should make more positive identifications of both guilty
culprits and innocent suspects, resulting in an increase in identification rates compared to
bystanders.
Whereas positively identifying a suspect can be considered an aggressive act under
Berkowitz’s (1989) frustration-aggression hypothesis, immediate confidence ratings and
response latency can be considered behaviors that increase the potency of the aggressive act. In
other words, among eyewitnesses who do aggress by making a positive identification, these
decision-making outcomes reflect the strength or potency of aggression. Higher confidence by
eyewitnesses who make a positive identification conveys increased certainty to investigators that
the suspect is, indeed, the culprit. Thus, expressions of greater confidence may be perceived by
eyewitnesses as enabling faster resolution of the criminal event. If victims do experience a
greater desire to resolve the aversive event than bystanders, it follows that victims should express
higher confidence than bystanders in their positive identifications, in order to increase the
likelihood that the identified suspect will be captured and punished.
It is also possible that victim-bystander status could influence response latency. One
theoretical explanation for this effect lies in Berkowitz’s (1990) cognitive neoassociative model:
victimization generates anger, which in turn activates ideas, memories, and schemas associated
with aggression. This increased activation of aggression schemas due to victimization may result
in victim-bystander status facilitating an aggressive response, thereby causing eyewitnesses to
make identifications sooner and resulting in faster response latency (Berkowitz, 1990). However,
13
it should be noted that faster response latency due to the priming effect of victim-bystander
status does not necessarily need to operate via anger, as Berkowitz (1990) suggests. Aggression
schemas could also be directly activated by victimization via semantic priming (Hutchison,
2003). With that said, both theoretical explanations suggest that victims should have faster
response latency when making an identification than bystanders.
Post-identification judgments. As with immediate confidence ratings and response
latency, post-identification judgments may also reflect the strength of eyewitnesses’ aggression.
Specifically, eyewitnesses who aggress by making a positive identification may bolster their
post-identification judgments in order to frame their positive identification decision as
trustworthy and reliable, enabling faster resolution of the aversive event. Eyewitnesses may be
particularly likely to do so when their post-identification judgments could have an impact on the
resolution of the criminal event. Therefore, an examination of post-identification judgments in
the context of the frustration-aggression hypothesis requires that eyewitnesses believe that their
post-identification judgments could have a real impact on the case.
Mediators of eyewitness aggression. Berkowitz’s (1989) frustration-aggression
hypothesis highlights two potential mediators of aggressive behavior: anger and desire to resolve
aversive events. Notably, the revised frustration-aggression hypothesis states that not all aversive
events are equally frustrating and that not all individuals will experience an aversive event in the
same way. Instead, the exact nature of the aversive event is not as important as the intensity of
the resulting negative affect, and, in particular, the intensity of the resulting feelings of anger
(Berkowitz, 1989). If victims do experience a criminal event as more aversive than bystanders,
one would therefore expect that victims would report experiencing greater anger at the time of
the aversive event.
14
Although Berkowitz primarily focused on the role of negative affect in his revised
formulation of the frustration-aggression hypothesis (1989), individuals must also experience a
desire to resolve the aversive event before an aversive event results in aggressive tendencies.
When social norms prohibit an aggressive response or when an event is simply seen as not
aversive enough to warrant frustration, individuals may not feel the need to resolve the event by
acting aggressively (Berkowitz, 1989; Cohen, 1955). In an eyewitness identification setting,
social norms do not necessarily prohibit eyewitnesses from making a positive identification; in
fact, one can consider social norms as encouraging eyewitnesses to make a positive
identification. Eyewitnesses’ motivation to catch the guilty culprit is, of course, tempered by a
motivation to avoid identifying innocent suspects. However, this does not take away from the
fact that, in order for an eyewitness to help resolve a criminal event, she or he must make an
identification decision that results in the capture of the guilty culprit. A positive identification of
the guilty culprit moves an eyewitness closer to resolving the event, whereas the rejection of an
innocent suspect does not. It follows that victims’ higher desire to resolve the aversive event
should lead to victims being more motivated than bystanders to catch the guilty culprit.
Review of Previous Victim-Bystander Research
The procedures used in this study were adapted from two studies that investigated the
effects of victim-bystander status on eyewitness decision-making. In the first study (Hosch &
Cooper, 1982), victim-bystander status manipulated whether participants believed that a
confederate had stolen a participant’s watch (victim condition), a calculator belonging to the
laboratory (bystander condition), or neither (control condition). In the second study (Kassin,
1984), victim-bystander status manipulated whether participants believed that a thief had stolen
money from them (victim condition) or their partner (bystander condition). In both studies,
15
participants were run in pairs and were presented with a single culprit-present lineup. Notably,
these studies produced conflicting findings regarding victim-bystander status. There were no
significant differences in accuracy between victims and bystanders in the first study, although
accuracy was significantly worse in the no-theft condition (Hosch & Cooper, 1982). In contrast,
victims were significantly less accurate than bystanders in the second study, with no victims
making a correct identification of the culprit (Kassin, 1984).
Five limitations of these studies reduced their utility in examining the effects of victim-
bystander status on eyewitness decision-making. First, both studies had small sample sizes: 33
pairs of participants in the Hosch and Cooper (1982) study and 15 pairs of participants in the
Kassin (1984) study. The low statistical power of these studies reduces confidence in the
estimates of effect sizes for victim-bystander status.
Second, although running participants in pairs can facilitate data collection, doing so
removes an experimenter’s ability to control for effects of the partner on the participant. It is
possible, for example, that bystanders’ or victims’ reactions to a theft may drive eyewitnesses’
decision-making, rather than the psychological impact of victimization on eyewitnesses.
Although the effect of partner reactions on eyewitness decision-making may be an important
area to address, a thorough examination of victim-bystander status should either account for
these effects or eliminate them as a possible mediator.
Third, both studies only employed culprit-present lineups, preventing a full analysis of
eyewitness performance. In experimental studies examining eyewitness decision-making, a
culprit-absent lineup or showup is necessary to calculate the diagnostic value of suspect
identifications by using the ratio of hits in culprit-present lineups to false alarms in culprit-absent
16
lineups (Wells & Olson, 2003). In other words, these preliminary studies on victim-bystander
status only had the potential to capture one half of eyewitness decision-making.
Fourth, participants in both studies were aware of their victim-bystander status at the time
of the theft. This may have led to differential attention and encoding of the culprit, in addition to
differential identification rates at identification. In other words, victims may have paid attention
to the culprit in a manner different from bystanders at the time of the crime, leading to
differences in accuracy that cannot solely be attributed to identification rates. This makes it
difficult to pinpoint whether the effects of victim-bystander status influence decision-making
during encoding, after encoding, or both.
Finally, in both studies participants were debriefed about the staged theft prior to
identification. This means that participants were aware that their identification decision would
have no actual consequences for the culprit, the victim, or the bystander. In the real world,
eyewitness identification procedures always involve consequences and risk. A missed
identification can lead to a culprit getting away while an identification of an innocent suspect can
result in wrongful conviction. Debriefing participants prior to the identification procedure
removes these consequences from an eyewitness’ decision-making calculus.
With that said, little is known about decision-making by eyewitnesses who believe that
their decisions could have real consequences. Almost all experimental eyewitness studies have
examined eyewitness decision-making in the absence of perceived real-world consequences (for
a notable exception, see Eisen et al., 2017). That is, participants in most eyewitness studies do
not believe that an incorrect identification has any additional cost beyond simply being wrong. In
addition to reducing generalizability of the findings to the real world, it is possible that
perceptions of the victimization and identification consequences as real are prerequisites for
