University of Southern California Dissertations and Theses

The moderating effect of triggering provocation on direct aggression

(USC Thesis Other)

The moderating effect of triggering provocation on direct aggression

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

THE MODERATING EFFECT OF TRIGGERING PROVOCATION
ON DIRECT AGGRESSION

by
Linli Yang

A Thesis Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF ARTS
(PSYCHOLOGY)

December 2009

Copyright 2009 Linli Yang

ii
Table of Contents
List of Tables iii
List of Figures iv
Abstract v
Chapter 1: Introduction 1
Direct Aggression and Displaced Aggression 1
Triggered Displaced Aggression and Triggered Direct Aggression 3
The Cognitive-Neoassociationistic Model 6
Chapter 2: The Current Research 9
Chapter 3: Method 15
Participants and Design 15
Procedure 15
Chapter 4: Results 23
Chapter 5: Discussion 36
References 42

iii
List of Tables
Table 1: Means and Standard Deviations for the relevant 4 cells in previous studies 11
Table 2: Mean percentage score of the word completion task as a function of the
Order to administer the task either before or after the aggression measure 39

iv
List of Figures
Figure 1: Mean percentage score of the word completion task as a function of the
order to administer the task either before or after the aggression measure. 28

Figure 2: Mean percentage score of total words in the word completion task as a
function of the Time 1 provocation and the Time 2 trigger. 33

Figure 3: Mean percentage score of action words as a function of the Time 1
provocation and the Time 2 trigger. 33

Figure 4: Mean percentage score of arousal words as a function of the Time 1
provocation and the Time 2 trigger. 34

Figure 5: Mean percentage score of affect words as a function of the Time 1
provocation and the Time 2 trigger. 34

v
Abstract
Consistent with previous research on Triggered Displaced Aggression and the predictions
of the Cognitive-neoassociationistic model, the current research found that a major provocation
followed by a minor triggering event resulted in disjunctive escalation of aggression which
exceeded the level that was predicted by the additive combination of the independent
aggression-eliciting effects of the initial provocation and the subsequent triggering event. Thus, it
was demonstrated that the synergistical interaction between the Time 1 provocation and Time 2
trigger occurs not only in displaced aggression, but also in direct aggression. Mediation analyses
showed that for participants who had been provoked, both negative affects and attributional
distortions of the triggering event mediated the effect of the trigger on aggression.

1
Chapter 1: Introduction
Tom makes the same mistake again at work. His angry boss severely reprimands him,
stating that the company does not need such stupid workers. For the entire day, Tom keeps
thinking about the reprimand but he cannot retaliate against his boss because he fears losing his
job. He arrives home later that day, only to be greeted by the greeting barks of his dog. In
response, he angrily kicks the poor dog away.
John steps into the conference room with a cup of hot coffee one morning. As he is
entering, one of his coworkers, Kim bumps into him, spilling hot coffee onto John's favorite shirt
– a gift from his beloved girlfriend, and quickly darts away. Though furious, John does not
display it toward Kim because all of his bosses are present and the conference is about to start.
Hours later after the conference is finished, Kim borrows a pen from his desk without permission.
This time, overreacting, John harshly berates Kim for being presumptive and inappropriate.
Direct Aggression and Displaced Aggression
Aggression is typically defined as behavior aimed at causing harm or pain to another
person (Berkowitz, 1993). However, human aggression is complex and cannot be defined as a
unitary phenomenon. Aggression can be direct or indirect, active or passive, physical or verbal,
etc. In animals, at least seven distinct subtypes have been identified (Moyer, 1976). The
identification and validation of qualitatively different forms of aggression has also been
attempted in human research in the last decades. Numerous attempts have been made to
categorize different forms of aggression based on differences in physiological, psychological, and

2
neurochemical processes that are associated with these behaviors. A hostile, angry reaction to a
perceived threat or frustration (Berkowitz, 1969) has been differentiated from an
offensive/instrumental aggressive response (Bandura, 1973). Whereas the former aggression type
(defined as defensive, affective, or impulsive aggressive response) is instigated by current
emotions like anger, the latter (non-impulsive, planned or instrumental aggression) is instigated
by an anticipated goal. Direct aggression, consisting of straightforward attacks that are often
visible and disruptive, has been differentiated from indirect aggression, which has been defined
as noxious behavior in which the target person is attacked in circuitous ways (such as trying to
make others dislike the target person by gossiping, revealing secrets, etc.) (Lagerspetz &
Bjorkqvist, 1994). Likewise, physical aggression has been differentiated from verbal aggression.
In line with this theoretical differentiation of aggression subtypes, Marcus-Newhall,
Pedersen, Carlson, and Miller (2000) make a distinction between direct and displaced forms of
aggression, differentiating them in terms of the targets being aggressed against
1
. Direct
aggression occurs when an individual is provoked and, in retaliation, imposes harm on the
original provocateur. Displaced aggression is thought to occur when a person is provoked but is
unwilling or unable to retaliate against the original provocateur, and instead, subsequently
aggresses against a seemingly innocent target. This differentiation between displaced and direct
aggression attained scientific prominence in the work of Dollard and his colleagues 70 years ago
(Dollard, Doob, Miller, Mowrer, & Sears, 1939), but the interest in it soon waned. Nonetheless,

1
Direct aggression here is different from direct aggression in the direct vs. indirect classification proposed by
Lagerspetz and Bjorkqvist (1994).

3
meta-analytic findings show that although aggression frequently is directed toward the agent
perceived to have provoked it, displaced aggression is also a reliable phenomenon
(Marcus-Newhall et al., 2000).
Triggered Displaced Aggression and Triggered Direct Aggression
In the original definition of displaced aggression proposed by Dollard et al. (1939), an
individual is initially provoked and restrained from retaliating against the provocateur, and in
response, subsequently aggresses against a completely innocent target. However, in most
instances wherein displaced aggression occurs in our daily life, the target of aggression is not
totally innocent but, instead, has provided some minor or trivial instigation. For instance, in the
first opening vignette, although he usually views them as friendly, Tom perceived the dog's
greeting barks as annoying and hostile and then responded with a disjunctively escalated
aggressive response toward the poor dog. To differentiate it from pure displaced aggression with
a completely innocent target, this type of displaced aggression has been termed Triggered
Displaced Aggression. Pedersen, Gonzales, and Miller (2000) present the first experimental
evidence showing that such triggered displaced aggression can elicit disjunctively escalated
retaliation. In the typical triggered displaced aggression paradigm, following an initial
provocation, the target emits a minor provocation (the trigger), which prompts the displacement
of aggression. As illustrated in the first scenario, the reprimand from the boss is the Time 1
provocation and the greeting bark from the dog is the following Time 2 trigger.
Triggered displaced aggression is of special theoretical interest because the resulting

4
aggression directed toward the target can exceed that which is expected based on norms of
reciprocity and tit-for-tat matching rules (Axelrod, 1984; Gouldner, 1960) that governs most
social interactions. Specifically, the initial Time 1 provocation and the subsequent Time 2 trigger
can synergistically interact to augment aggressive responding, resulting in a disjunctively
escalated aggressive response towards the triggering target. The level of such aggressive
retaliation exceeds that which is predicted by the additive combination of the independent
aggression-eliciting effects of the initial provocation and the triggering event (Miller &
Marcus-Newhall, 1997).
However, as experimentally shown by Vasquez, Denson, Pedersen, Stenstrom, & Miller
(2005), a disjunctively augmented aggressive response is constrained to circumstances wherein
the intensity of the trigger is relatively trivial. A strong Time 2 provocation simply invokes the
tit-for-tat matching rule rather than leading to unexpected disjunctive escalation. As shown in two
previous studies (Worchel, 1966; Baron & Bell, 1975), when very strong Time 2 triggering
events were employed, the interesting potential characteristic of triggered displaced aggression –
the hypothesized interaction between provocation and trigger – was lost. Thus, for disjunctive
augmentation to occur, the triggering event must instead be minor and ambiguous in its meaning
and intention, which allows attributional distortion of the trigger when preceded by a moderately
strong initial provocation (Miller, Pedersen, Earleywine & Pollack, 2003). In effect, a previously
provoked person is more likely to make the attribution that the minor trigger was an intentional
provocation, while one who was not been provoked before may perceive the minor trigger as a

5
non-provoking event and hence, may simply ignore it without any further aggressive reactions.
In sum, the triggered displaced aggression process can cause trivial transgressions to lead
to unexpectedly severe retaliation. Consequently, it is important to understand its causes and
"catalyzers", as well as the way to control it. A number of studies have addressed the task of
identifying such mediators and moderators, examining similarity between the participant and the
aggression target (Pedersen, Bushman, Vasquez, & Miller, 2008), salience of aggressive cues
(Denson, Aviles, Pollock, Earleywine, Vasquez, & Miller, 2008), rumination (Bushman, Bonacci,
Pedersen, Vasquez, & Miller, 2005), alcohol (Aviles, Earleywine, Pollock, Stratton, & Miller,
2003; Denson et al., 2008), group membership of the aggression target (Vasquez, Ensari,
Pedersen, Tan, & Miller, 2007; Pedersen et al., 2008), the public or private nature of the setting,
and other attributes of the target (Pedersen et al., 2008).
At this point it is useful to draw attention back to direct aggression. As previously noted
and well-established, aggression typically is directed toward the agent perceived to have
provoked it. Displaced aggression only occurs when certain situational factors constrain direct
aggression. Therefore, although recent researchers have provided strong evidence establishing
that displaced aggression is a reliable phenomenon, direct aggression is still the most common
form of aggression in real-world settings. Considering the robust and interesting
aggression-escalating effects of triggered displaced aggression and the ecological significance of
direct aggression, the current research aimed at extending the triggered displaced aggression
paradigm to direct aggression.

6
The current study retains all the key features of the triggered displaced aggression
paradigm: the Time 1 provocation, the Time 2 triggering event, and the final assessment of
aggression toward the triggering person. In the current study, however, the individual who emits
the Time 1 provocation is also both the source of Time 2 trigger and the target of aggression. This
stands in contrast to prior triggered displaced aggression studies wherein, although the individual
who is the source of trigger is also the target of the aggression, this target person is not the
original provocateur. As illustrated in the second opening anecdote, Kim is both the original
provocateur and the source of trigger. Her coffee-sloshing behavior is the initial Time 1
provocation and her borrowing of the pen without permission is the subsequent Time 2 trigger.
Also, Kim is the target of aggression. Because John perceived the behavior of borrowing pen
without permission as presumptive and inappropriate, he harshly berates Kim in response to it. In
this anecdote and in the current study, this type of aggression can be labeled Triggered Direct
Aggression. The purpose of this study was to demonstrate that disjunctively escalated triggered
aggression is not limited to displaced aggression, but also occurs in direct aggression.
The Cognitive-Neoassociationistic Model
In the current study, the expected interactive effect of the initial provocation and the
subsequent trigger can be explained within the framework of Berkowitz's (1989, 1990, 1993)
Cognitive-Neoassociationistic Model (CNA) model of aggression. This model proposes that
aversive events (any type of aversive event, including but not limited to provocations) lead to
aggression as a consequence of the negative affect they generate. In turn, such affect activates

7
various thoughts, memories, physiological responses, and expressive-motor reactions contained
within an associative network. According to this model, these related feelings, thoughts, and
memories are linked together in a cognitive network; once a construct is processed or stimulated,
activation spreads out along the network links and primes (activates) associated or related
constructs. Thus, any prior priming of a part of the cognitive network (such as aggressive
cognitions related to the initial provocation in triggered displaced aggression paradigm) will
make other related cognitions (such as further aggressive cognitions related to the subsequent
triggering event) more readily available and accessible.
The CNA model proposes a multistage process in reaction to negative events. During the
first stage, the negative event produces negative affect that in turn activates associated thoughts,
memories and expressive-motor reactions. At the same time, two sets of reactions are activated
simultaneously: fight (associated with rudimentary anger) and flight (associated with rudimentary
fear) responses. The second stage, higher-order processing begins and high-level cognitions
emerge, elaborating, intensifying, or suppressing the initial primitive emotional reactions by
high-order appraisal and attributional processes. In this stage, relatively basic emotions of anger
or fear are differentiated into more subtle variations, such as vexation, sadness, annoyance,
anxiety, and so forth. Moreover, the CNA model does not claim that the second stage will
necessarily occur. This theoretical position allows it to account for instances of aggression in
which attributions are important while simultaneously accounting for those in which they do not
come into play.

8
In the current study the theoretical framework concerning triggered direct aggression is
consistent with the predictions of the CNA model. Following this model, the initial provocation,
being obviously aggressive, activates associated aggressive thoughts and primes their related
cognitive constructs within the associated networks. This increased activation increases the
likelihood that the Time 2 trigger will be perceived as a more negative and provoking event than
it is, although in the absence of the initial provocation it may only be interpreted as trivial and
harmless. Subsequently, when confronted with the minor trigger, individuals respond with a
disjunctively escalated aggressive reaction. In sum, the CAN model provides a context in which
the expected feature of disjunctive escalation in the triggered direct aggression paradigm can be
explained.

9
Chapter 2: The Current Research
As indicated, previous triggered aggression research has been limited to displaced
aggression. The goal in the current research was to examine the effects of replacing the displaced
target with the original provocateur, who emits both the Time 1 provocation and the Time 2
trigger. Aside from the replacement of the target of aggressive retaliation, the current research
adheres very closely as previous triggered displaced aggression studies. Imagine these two
opening scenarios again: in the first one, Tom receives a severe reprimand from his boss, but, in
response to his dog's greeting barks he vents all his pique on a poor dog instead of his boss. In the
second anecdote, Kim messes up John's favorite shirt, but he suppresses his anger because of the
conference setting. However, after Kim borrows a pen from his desk without permission, John
pours out all his anger by harshly berating Kim for being presumptive and inappropriate. Thus,
triggered displaced aggression and triggered direct aggression, respectively, are well illustrated
by these two scenarios.
As readily seen, these two types of triggered aggressions have the same key features: a
major Time 1 provocation, a minor Time 2 trigger, and a final aggression part. The only
difference between them is that triggered displaced aggression involves three people – the actor
and two provocateurs – whereas the triggered direct aggression paradigm only involves the actor
and a single provocatuer, albeit one who provides two provocations.
Because the current study retains all the key features of the triggered displaced
aggression paradigm, the same mechanisms should operate. Hence, I expected the same

10
disjunctive aggression-escalation effect from the current research on triggered direct aggression
following a minor Time 2 provocation as was seen in the prior work on triggered displaced
aggression that also used minor Time 2 provocations. Because the disjunctive
aggression-escalation effect has been well demontrated by many previous triggered displaced
aggression studies (Pedersen et al., 2000; Miller et al., 2003; Aviles et al., 2005; Bushman et al.,
2005), it makes sense to test the expected form of interaction within the 2 × 2 factorial design
manipulating presence or absence of both provocation and trigger by applying a theory-driven
planned contrast to the cell means. Based on outcomes for the four cells in the existing literature
on triggered displaced aggression following a minor trigger (see Table 1), aggression is always
strongest in the provocation/trigger cell. By comparison, aggression in the trigger only cell and
aggression in the control cell are both very low and do not differ from each other in most studies
because the trigger itself is very trivial and not viewed as intentional harm. Aggression in the
provocation only cell is worth more discussion. In some studies (e.g., Bushman et al., 2005),
Aggression in this cell is as low as that in the two no provocation cells because the target for
aggression has emitted no provocation. In other studies (e.g., Pedersen et al., 2000), aggression in
the provocation alone cell was the lowest among the four cells. Although on first thought
anomalous, this result can be explained by a judgmental contrast effect: a person who provides a
subsequent non-triggerring neutral evaluation will be perceived more positively when compared
with a person who moments previously had provided a rather strong initial provocation; when,
instead, the actions of the target person in the no provocation/no trigger cell have not been

11
preceded by one who harshly provokes the participant, he/she is perceived less positively.
Table 1
Means and Standard Deviations for the relevant 4 cells in previous studies
Reflecting these differences in the provocation only cell, two types of contrast weights

Citation

n

Dependent
variable
P/T P/No T No P/T No P/No T
M SD M SD M SD M SD
1. Pedersen,
Gonzales, &
Miller (2000)
Study 1

2. Pedersen,
Gonzales, &
Miller (2000)
Study 2

3. Bushman,
Bonacci,
Pedersen,
Vasquez, & Miller
(2005) Study 2

4. Vasquez,
Denson, Pedersen,
Stenstrom&
Miller (2005)

5. Pedersen,
Bushman,
Vasquez, & Miller
(2008) Study 1

6. Pedersen,
Bushman,
Vasquez, & Miller
(2008) Study 2
64

49

347

59

305

96

Evaluation of
the assistant's
qualification
for a RA
position

Impression
and attitude
toward the
confederate

Hot sauce
paradigm

Cold water
paradigm

The intensity
and duration
of noise

Cold water
paradigm
0.52

32.50

5.44

6.57

0.47

56.73
0.09

12.19

N/A

N/A

N/A

N/A
-0.285

17.17

2.78

3.83

-0.39

14.69
0.64

4.20

N/A

N/A

N/A

N/A
0.002

23.00

2.44

3.35

-0.15

38.80
0.72

6.18

N/A

N/A

N/A

N/A
-0.047

22.54

2.59

2.48

-0.20

19.20
0.75

9.54

N/A

N/A

N/A

N/A

12
can be derived from previous triggered displaced aggression studies for testing the interaction
between presence or absence of both provocation and trigger: 3 (provocation/trigger), -1
(provocation only), -1 (trigger only), and -1 (no provocation/no trigger) for these studies without
contrast effect; and 4 (provocation/trigger), -2 (provocation only), -1 (trigger only), and -1 (no
provocation/no trigger) from those studies wherein contrast effect has occurred. Although
research has not isolated the critical factor responsible for the presence or absence of judgmental
contrast within the Triggered Displaced Aggression paradigm, neither of these two patterns of
contrast weights applies to the current study because the original provocateur is the same as the
source of the trigger and the aggression target, thereby precluding a contrast effect. Moreover, the
aggression towards the original provocateur in the provocation/no trigger condition is likely to be
more intense in the current study than that in previous triggered displaced aggression studies
because the target for aggression and the person who emitted the Time 2 trigger is the same as the
person who provided the Time 1 provocation.
Thus, when the target initially provokes the participant, even if no subsequent trigger is
emitted by the provocateur, the participant is likely to display substantial aggression toward the
provocateur because of the harsh initial provocation. Yet even stronger retaliatory aggression is
expected when the provocateur both provokes and triggers the participant. By contrast with these
conditions, very little aggression is expected not only in the no provocation/no trigger cell, but
also in the no provocation/trigger cell. This latter expectation assumes that in the absence of a
strong prior provocation that primes aggression, the trigger will be interpreted as benign and as

13
insufficient to exceed the threshold for the expression of aggression. Therefore, more reasonable
contrast weights for describing the interaction between presence or absence of both an initial
provocation and a trigger that is expected in the current study would be 3, 1, -2, and -2 for the
provocation plus trigger, provocation alone, no provocation but trigger present, and neither
provocation nor trigger cells respectively.
In the current study, the cognitive accessibility of aggressive constructs was also our
interest. Aggressive constructs here refer to the aggression-related cognitions that comprise the
elements of the cognitive-neoassociationistic network. According to the CNA model, high levels
of accessibility of these aggression-related cognitions indicate that aggression is on a person's
mind, even if one is not consciously aware of them (Anderson, Carnagey, & Eubanks, 2003;
Anderson, Carnagey, Flanagan, Benjamin, Eubanks, & Valentine, 2004). Also, higher
accessibility to these aggression-related cognitions within this network will result in higher levels
of aggressive inclination. Thus, in the current study, it was expected that there would be a
relationship between the activation of the cognitive-neoassociationistic network and triggered
direct aggression.
In sum, according to the predications of the CNA model and in comparison with previous
triggered displaced aggression studies, the current triggered direct aggression study aimed at
testing two main hypotheses: First, the Time 1 provocation and Time 2 trigger would
synergistically interact to augment aggressive responding, and result in a disjunctively escalated
aggressive response in the provocation/trigger condition in comparison to that found in the other

14
three conditions (provocation alone, trigger alone, and the control); Second, according to the
CNA model, higher levels of cognitive accessibility of aggression-related constructs will be
associated with higher levels of aggression. Thus, I expected that the cognitive accessibility of
aggressive constructs would display a similar pattern (though not identical)
2
to the main
dependent measure of aggression: participants who experienced both the major provocation and
the minor trigger would have higher levels of cognitive accessibility to aggression-related
constructs than those in the provocation alone, the trigger alone or the control condition.

2
The difference between the pattern expected for the word completion measure from that expected for the aggression
measure is discussed in the results section.

15
Chapter 3: Method
Participants and Design
Participants were fifty-five undergraduate students (41 females and 14 males) from the
University of Southern California who received extra course credit for their voluntary
participation. Data from 7 participants were discarded because of suspicion or other reasons: 4 of
them had participated similar studies before, 1 of them had heard about the experiment from their
friends and the other two just figured out the real purpose of this study. Therefore, the final
sample consisted of 48 participants, 34 females with 6 in the provocation/trigger condition, 7 in
the provocation/no trigger condition, 10 in the no provocation/trigger condition and 11 in the no
provocation/no trigger condition; and 14 males with 6, 5, 2 and 1 in the provocation/trigger,
provocation/no trigger, no provocation/trigger and no provocation/no trigger condition
respectively. In each experimental session, a confederate, who was trained prior to the study,
always pretended to be the second participant and participated the study together with the real
participant. The study was a 2 (provocation manipulation: Provocation vs. No provocation) × 2
(trigger manipulation: Trigger vs. No trigger) between-subject factorial design. Each participant
was randomly assigned to one of the four conditions.
Procedure
Participants were tested individually. When they arrived for experiment, the experimenter
led them into the experimental room and introduced himself/herself. As part of the cover story,
the experimenter told the participants that the study was concerned with problem solving ability

16
and would have three different tasks. In individual sessions participants were led to believe they
would be interacting with another participant (actually a confederate). After the participants were
seated in the experimental room, the experimenter told them that this experiment actually
involved a second participant who would be run at the same time in a different room. The
experimenter then excused himself/herself to check whether the other participant had arrived.
After 2 or 3 minutes, the experimenter returned to tell the participants that the other participant
was there and they could start. Three experimenters (two males and one female) and three
confederates (1 male and two females) helped with this study. Each experimenter always paired
up with the same confederate .
Time 1 Provocation. After they read and signed the informed consent sheet, the
experimenter collected it and then told the participants that he/she needed to check whether the
other participant had completed it. After 1 minute, the experimenter returned to tell the
participants that the first part of the study was a general measure of cognitive ability. In reality, it
served as the manipulation of provocation. Participants received 15 anagrams and were told to
solve as many of them as possible within 4 minutes. The experimenter then turned on the loud
music and left. After 4 minutes, he/she returned, turned off the music, took the anagram answer
sheet from the participants, and told the participants that he/she would bring the other participant
here for scoring and exchanging evaluation, because the researchers believed such mutual
evaluation would be helpful to improve their performance in the next task. After leaving again,
the experimenter re-entered with the other participant (the confederate), had the confederate sit

17
opposite the participant, and gave their answer sheets back to them, along with a correct answer
sheet and a handout indicating how previous students had performed on the same anagram task.
Then the experimenter told them that they needed to exchange their answer sheets and score each
other's answers according to the correct answer sheet. When they finished scoring, the
experimenter told them to take back their own answer sheet and look through it as well as the
handout. After they finished reading, the experimenter asked them to briefly evaluate each other's
performance and give some suggestions to each other. The participants were always asked to give
their evaluation first.
In the provocation condition, participants received a list of 15 difficult anagrams to solve
(e.g., nvtnimereon, tophhapogr, and iosunttinti) whereas in the no provocation condition they
received very easy anagrams (e.g. meit, dre, and stet). Also, in the provocation condition,
participants listened to loud, distracting, and mildly irritating background music (Stravinsky's
Rites of Spring played at 80 dB) while working on those difficult anagrams, whereas those in the
no provocation condition listened to quiet and mildly positive music (the opening movement of
Bach's Brandenberg Concerto No. 3). In addition, in the provocation condition, the confederate
insulted the participants with a harsh evaluation of their performance. All confederates used the
same script to insult the participants: "Well, your performance is not so good, er…maybe I should
say poor. About suggestions, umm…I don't know what kind of suggestion could be helpful to
you, maybe…if I were the experimenter, I would ask you to do this task all over again." Then the
confederate added in a contemptuous tone, that he/she was afraid that it would just turn out to be

18
a waste of time. In order to justify confederates' superiority and reduce participants' suspicion
levels, the confederate always correctly answered 11 anagrams while the average level in the
previous studies was approximately 5. (The experimenter always prepared the confederate's
answer sheet before the experiment). This made the confederate seem more intelligent and
justified the subsequently harsh insult to the participant. In the no provocation condition, the
participants were not insulted. The confederate only said in a normal tone: "Well, your
performance is ok; look at the printout; you have received an average score compared with the
other people (or "your score is above average" if the participant had correctly answered most of
the anagrams)."
Time 2 Trigger. After receiving their evaluations, the experimenter asked the confederate
to return to his/her original experimental room. Then, the experimenter told the participants that
they now would be participating in the second part of the experiment, the NASA task, which was
concerned with how to use their imagination to solve problems. Actually this task served as the
context for the trigger manipulation. Also the participants were informed that this time they and
the other participant (the confederate) would evaluate each other's work indirectly rather than
face-to-face.
The experimenter then left to "get the other participant started first." After 2 minutes,
he/she returned and gave the participant the task instruction for the NASA task in an envelope.
Participants were instructed to think of five good characteristics that would be helpful to an
astronaut, including traits, beliefs, and qualities. The experimenter left the participants alone to

19
complete the task in 3 minutes. Upon returning, he/she asked the participant to place their work
in an envelope and allegedly took the participant's work to the other participant for evaluation.
One minute later, the experimenter returned with the confederate's alleged completed work and
an evaluation form for the participants to fill out, told the participant to evaluate their partner's
work using the evaluation form, and informed the participant that the other participant was
already evaluating their work. The evaluation consisted of ratings on the following six
dimensions: (1) the originality of the confederate's answers; (2) the quality of the answers; (3) the
effort put into the task; (4) the variety of answers; (5) the degree to which the answers made
sense; (6) the confederate's overall performance on the task. Ratings were made on a 7-point
Likert-type scale, ranging from 1 (not good at all) to 7 (extremely good). In addition, space was
available for participants to indicate additional comments. After the participant completed the
evaluation of the confederate, the experimenter ostensibly took their evaluation to the confederate
and returned with a bogus evaluation of their own work, allegedly completed by the confederate.
In the trigger condition, the participant's performance on each of the dimensions was
rated as 3, 4, 3, 3, 4, and 4, with an average overall rating of 3.5. In addition, in the space that
was provided for additional comments, the following statement was written: "I think the
performance could have been better." In the no trigger condition, the participant's performance
was rated as 6, 5, 6, 5, 5, and 5, with an average overall rating of 5.5. In addition, the confederate
ostensibly stated that the participant's performance was fine. All the participants were given
approximately 3 minutes to look over the evaluation from the confederate. Though slightly

20
positive in an absolute sense, this evaluation was designed to be psychologically neutral in that
people exhibit a positivity bias and evaluate even totally unknown strangers in a slightly positive
manner (Lau, Sears, & Centers, 1979; Rook, Sears, Kinder, & Lau, 1978; Sears, 1983).
Measure of Aggression. The experimenter returned with a cup containing two pieces of
paper and explained that the next portion of the study would investigate how physical distraction
affects a person's problem solving abilities. Participants then drew one of the two folded pieces of
paper to determine their distraction condition (Distraction or No-distraction). In reality, both
pieces of paper contained the no-distraction condition, thereby always assigning them to that
condition. The experimenter then left the room ostensibly to see which distraction condition the
other participant (the confederate) received. While waiting for the experimenter, half of the
participants were asked to do a word completion task, which assessed the cognitive accessibility
of aggressive action, arousal, and negative affect cognitions.
Approximately 2 minutes later, the experimenter returned with a bucket of cold water
(10°C, 50°F) and the dependent measures. He informed the participants that the other participant
had been assigned to the distraction condition. Participants were told that they would determine
how long the other participant would immerse one of their hands in the bucket of cold water
while performing a cognitive task. Participants were instructed to place one hand in the bucket of
water for about 5 seconds, to guide their decision about the length of time that their partner
would be distracted. Next, they were instructed to circle the amount of time that the other
participant should be distracted on a 9-point scale starting at "1 = 0 second/no distraction at all"

21
to "9 = 80 seconds/very strong distraction" with 10-sec. intervals for each point. This served as
the dependent measure of physical aggression. Participants were asked to slide the aggression
sheet under the door so that the experimenter could administer the task to the other participant.
Participants then complete the remaining dependent measures at their own pace.
Measure of Cognitive accessibility of aggression constructs and Manipulation checks.
Participants completed the remaining measures in a packet. The first one is a word completion
task that was actually meant to measure the cognitive accessibility of aggressive constructs. This
pen-and-paper word completion task consisted of a series of 34 incomplete words, with some
letters provided and some missing. Each was specifically selected so that participants could fill in
letters to form either an aggressive or a non-aggressive word. For instance, "prov_ _e" could be
completed as "provoke" or "provide", "m _ d" could be completed as "mad" or "mud", and "ang _
_" could be completed as "anger" or "angle". Participants were instructed to complete this task as
fast as they can, and use the first word that comes to mind. This instrument is Pedersen et al.'s
(unpublished manuscript) modification of a measure created by Anderson et al. (2003, 2004).
Whereas the original version only included words in the action category, the revision added affect
and arousal words. The measure was used successfully by Petersen et al. to assess differences in
the cognitive accessibility of aggressive constructs. Note that half of the participants had already
completed the word completion task before, right after they drew the slips of paper and when
they were waiting for the experimenter. Only the other half completed the task at this point.
The second measure in this packet was a modified Mood Adjective Check List (Nowlis,

22
1965) to report how the participants were affected by the initial provocation. Participants were
instructed to describe how they felt after performing the anagram task and rate the degree to
which they experienced each of 26 emotions in the list. Each emotional descriptor was rated on a
4-point scale ranging from 1 (definitely did not feel this way) to 4 (definitely did feel this way).
The third measure in the packet was the manipulation check for the trigger. Specifically,
participants were asked to rate how "happy, complimented, pleased, annoyed, irritated, and
angry" they felt about the confederate's evaluation of their NASA task. Each item was rated on a
7-point Likert-type scale (1 = not at all, 7 = extremely so).
Six additional questions assessed the attributional distortion. Participants were asked to
rate the degree to which their partner provided a good, unfair, useless, and unbiased evaluation.
In addition they also rated the degree to which their partner meant to provide a negative
evaluation and a valuable evaluation. All items were rated on a 9-point Likert-type scale (1 =
strongly disagree, 9 = strongly agree).
Finally, all participants were thanked for their participation, probed for suspicion with a
funnel debriefing, and then dismissed.

23
Chapter 4: Results
I withdrew 7 participants (6 females and 1 male) from the analyses because of suspicion.
Fisher's exact test was performed to assess whether the removed suspicious participants were
differently associated with specific conditions. Results indicated that there was no differential
attrition of participants across the four experimental conditions (p = .27). Thus, 48 participants
(12 in each cell) were included in the following analyses (34 females, 14 males; 23 White, 13
Asian, 4 Hispanic, 2 African American and 6 who self-identified as "other"). The mean age for
the sample was 20.19. There were no main effects or interactions involving gender for any of the
measures. Thus, the data for men and women were combined for subsequent analyses.
Provocation manipulation checks
To assess the effectiveness of the provocation manipulation, participants were asked to
recall how they felt when they completed the anagram task. They indicated their mood by
responding to the MACL (Nowlis, 1965). On an a priori basis, six adjectives that describe an
irritated or annoyed mood were analyzed, including angry, irritable, defiant, annoyed, grouchy,
and frustrated. Internal consistency among these items was acceptably high (Cronbach's α = .85).
Thus, these items were combined into a reliable negative mood composite ranging from 1
(minimum anger) to 4 (maximum anger). As expected, those in the provocation group (M = 2.95,
SD = .87) felt more angry than those in the no provocation group ( M = 1.45, SD = .63), t(46) =
6.85, p < .001, d = 2.02, indicating an effective manipulation.
Trigger manipulation checks

24
To assess the effectiveness of the trigger manipulation, participants rated their emotional
response to the feedback they received from the confederate for the NASA task (i.e., how
irritated, complimented, happy, angry, or offended, pleased, and annoyed they felt). After the
positive items were reverse scored (complimented, happy, and pleased), these items were formed
into a composite measureof negative affect (α = .91). As expected, triggered participants (M =
4.81, SD = 1.02) felt more negative affect after they received the evaluation than did
non-triggered participants (M = 3.05, SD = .93), t(46) = 6.212, p < .001, d = 1.83. Also,
participants were asked to rate the degree to which their partner provided a good, unfair, useless,
negative, valuable, and unbiased evaluation. After the positive items (good, valuable, and
unbiased) were reverse coded, these six items were combined into a composite measure of
attributional distortion (α = .87). As expected, triggered participants (M = 5.66, SD = 1.53) were
more likely to perceive the confederate's evaluation as more unfair and biased than did
non-triggered participants (M = 3.42, SD = 1.16), t(46) = 5.69, p < .001, d = 1.68. These analyses
indicated a successful trigger manipulation.
Analysis of the trigger manipulation-check data for the subset of participants in the
no-provocation conditions produced similar results. Specifically, separate t tests for trigger
manipulation check questions of negative affect and attributional distortion showed reliable
differences in the predicted direction (t(22) = 4.11, p < .001, d= 1.75; t(22) = 2.21, p < .04, d
= .94, respectively), indicating that under conditions of no prior provocation, participants in the
trigger condition were significantly more angry, less happy, and thought the evaluation was less

25
fair and more biased than did those who were not triggered. Thus, the effect obtained on the
overall analyses of the manipulation check for the trigger manipulation was not only due to those
participants who had experienced prior provocation.
Aggression
The aggression measure consisted of the assigned duration for which participants
required their partners to immerse their hand in ice water. Three pairs of experimenters and
confederates ran the study. Each pair conducted 15, 17, and 16 experimental sessions respectively.
To examine whether there was any experimenter/condeferate effects on the dependent variable
measure of aggression a (three pairs of experimenters/confederates) × 2 (provocation/no
provocation) × 2 (trigger/no trigger) ANOVA was performed on the aggression measure. The
experimenter/confederate variable failed to produce any significant main effect (F(2, 36) = .01, p
= .99, partial η
2
= .001). Therefore, there was no indication that participants in our study
responded differently to any of the three pairs of experimenters/confederates in comparison to the
others . Also, there was no significant interaction found between the experimenter variable and
provocation (F(2, 36) = 1.23, p = .28, partial η
2
= .033)/trigger (F(2, 36) = .41, p = .67, partial η
2
= .022). The three-way interaction also was not significant (F(2, 36) = 2.37, p = .13, partial η
2
= .116).
Next, the expectation of disjunctive escalation was confirmed by inspecting the means:
When the participant had not experienced previous provocation, the triggering evaluation given
by the confederate had no moderating effect on aggressive responses toward him/her (M = 3.00,

26
SD = 1.91 for the no provocation/no trigger cell) when compared with the baseline no
provocation/no trigger cell (M = 2.92, SD = 1.73). For those participants who were initially
provoked, the trigger displayed the expected moderating effect on aggressive retaliation (M =
5.58, SD = 2.11 for the provocation/trigger cell compared with M = 4.00, SD = 1.91 for the
provocation/no trigger cell). Thus, compared with the baseline of 3.00, the provocation/trigger
cell deviated by 2.58, which exceeds the sum of the independent deviations from the baseline
found in the provocation-only and trigger-only cells (.98).
Direct t-test comparisons of the provocation/trigger cell with each of the other three cells
further confirmed the hypothesis of disjunctive escalation - the cell in which disjunctive
escalation was expected to occur differed from the no provocation/trigger cell (t(22) =3.39 p
= .003, d = 1.44) and no provocation/no trigger cell (t(22) = 3.15, p = .005, d = 1.34). The
difference between provocation/trigger cell and provocation/no trigger cell did not reach but
approached significance (t(22) = 1.93, p = .06, d = 0.82). This effect differed from the typical
findings in triggered displaced aggression studies, wherein the provocation/no trigger cell always
displayed very little aggression or even the least amount of aggression among the four cells due
to a judgmental contrast effect. However, in our study, this difference was expected because the
person who was the target for final aggression had emitted both the Time 1 provocation and the
Time 2 trigger. In this case, when the participant was previously provoked by the target, even
though there was no subsequent trigger emitted by the target, the participant would also display
substantial aggression (but still less than those who were both provoked and triggered) toward the

27
target because of the initial harsh provocation. By contrast, in previous triggered displaced
studies the target for final aggression was the same as the person who provided the Time 2 trigger,
but was different from the person who emmitted the Time 1 provocation. In this latter situation,
after being provoked by another person first, if the participant did not receive the trigger from the
target, then the participant would not show any aggressive response toward the target because the
target did nothing wrong and there was no justification for aggressive response toward him/her.
Because I wanted to test the specific model predicted by exsiting theory, as explained in
the Introduction, an a priori test was applied to test the hypothesized interaction using the
contrast weights 3 (provocation/trigger), 1 (provocation only), -2 (trigger only), and -2 (no
provocation/no trigger), which also received support from the t-tests above. As expected, when
compared to the no provocation/no trigger cell, aggression in the provocation/trigger cell was
higher (approximately two times higher according the model) compared to the provocation/no
trigger cell, while aggression in the latter cell itself was stronger than the other two cells, F(1, 44)
= 14.36, p < .001, partial η
2
= .246 (see Figure 1). The residual variance was not significant, F(1,
44) = .71, p > .40, partial η
2
= .016, indicating that the model fully explained the systematic
variance between the groups.

28
Figure 1. Mean score of the distraction duration for the confederate as a function of the
Time 1 provocation and the Time 2 trigger. Higher score indicates a longer distraction
time and greater aggression.

Also, I expected that the aggression in the provocation conditions would be significantly
higher than that in the no provocation conditions. Thus, contrast weights 1 (provocation/trigger),
1 (provocation only), -1 (trigger only), and -1 (no provocation/no trigger) were applied to test this
main effect of provocation. As expected, aggression in these two provocation condition differed
from that in the two no provocation conditions, F(1, 44) = 10.93, p < .002, partial η
2
= .199.
Finally, to test the expectation of highest aggression in the provocation/trigger cell among the
four cells, an a priori test was applied using the contrast weights 3 (provocation/trigger), -1
(provocation only), -1 (trigger only), and -1 (no provocation/no trigger). The results also
confirmed this hypothesis, F(1, 44) = 12.65, p < .001, partial η
2
= .223. In sum, these results
supported the notion that in the presence of initial provocation, the trigger would synergistically
interact with the provocation to produce disjunctive escalated aggression, and the augmented

29
aggression in the provocation/trigger condition would be the strongest among the four conditions.
According to the CNA model, a link between provocation induced negative affect, which
was assessed by the modified MACL (Nowlis, 1965) in the postmeasure packet, and the
subjective affective responses to the triggering evaluation among the participants who were in the
provocation conditions is to be expected. Correlational analyses revealed that provocation
induced negative affect was indeed associated with subjective affective responses as assessed by
the trigger manipulation checks (r = .33, p = .02), and in turn, these negative affective responses
to the trigger were associated with a stronger level of aggressive response (r = .38, p = .007).
Thus, stronger provocation-induced negative affect resulted in more negative affective responses
to the triggering evaluation; and stronger negative affective responses to the trigger in turn
elicited greater aggressive response to the target. This pattern of results is consistent with the
CNA model which argues that the initial provocation activates the related representations of
negative affect within the cognitive-neoassociationistic network, these cognitive representations
in turn primes greater negative affective responses to a minor trigger, and then, these negative
affective responses elicit stronger aggressive responses to the target.
Mediational analyses
To further explore the hypotheses that the negative affective responses to and
attributional distortions of the Time 2 trigger elicits aggression toward the target, we conducted
two separate mediational analyses on negative affective responses to and attributional distortions
for participants in the provocation condition following Baron and Kenny's (1986) procedures.

30
First, mediational analyses were performed on negative affective responses to the Time 2
trigger. I first regressed the level of final aggression on the trigger manipulation. The results
suggest that the trigger manipulation did predict the level of aggression in the hypothesized
direction, b = .75, p = .04 (one-tail). Next, I regressed negative affective responses on the trigger
manipulation. The results confirmed that the trigger manipulation reliably predicted participants'
negative affective responses, b = 1.76, p <.001. Then, the amount of aggression was regressed on
negative affective responses. The results confirmed that a more negative affective state led to a
higher level of aggression, b = .73, p <.01. Finally, when negative affective responses were
controlled, trigger manipulation no longer predicted aggression, b = .28, p = .47, but negative
affective responses remained as a significant predictor of aggression when trigger manipulation
was controlled, b = .69, p <.01. Thus, the four criteria suggested by Baron and Kenny (1986) for
establishing mediation were fulfilled. The Sobel test confirmed that this pattern of mediation was
statistically significant (z = 2.43, p = .015).
Second, mediational analyses were performed on attributional distortions of the Time 2
trigger. Baron and Kenny (1986)'s four criteria were also fulfilled: first, trigger manipulation
predicted the level of aggression in the hypothesized direction, b = .75, p = .04 (one-tail); second,
trigger manipulation also reliably predicted attributional distortions,b = 2.34, p < .001; third,
attributional distortions of the trigger predicted the level of aggression, b = .67, p < .01; finally,
trigger manipulation no longer predicted aggression when attributional distortions were
controlled, b = .19, p = .52, but attributional distortions remained as a significant predictor of

31
aggression when trigger manipulation was controlled, b = .57, p = .01. The Sobel test confirmed
that this pattern of mediation was also statistically significant (z = 2.35, p = .019).
In sum, these results provide support for our hypothesized mediation: for participants in
the provocation conditions, both negative affective responses to the Time 2 trigger and
attributional distortions of the Time 2 trigger mediated the effect of trigger evaluation on
participants' final aggressive response toward the target.
Cognitive accessibility of aggressive constructs
The word completion task we used contained thirty-four items designed to test the
cognitive accessibility of aggressive constructs. There were three components of aggressive
constructs and three corresponding categories of words in this measure: a) behavioral intention
(action words), consisting of fourteen items including such words as "hit" and "attack", b) arousal
(arousal words), consisting of nine items such as "excited" and "intense", and c) affect (affect
words) words, consisting of eleven items such as "anger" and "irate". As not all participants
completed every word in this task, percentage score was created for each of these three categories
as well as the entire measure.
As previously noted, this measure was administered to half of the participants before the
final aggression measure and half after the aggression measure. This allowed assessment of
whether participants in conditions wherein they first expressed their anger by letting their partner
(the confederate) experience longer distraction, had decreased cognitive accessibility to
aggressive constructs. To examine this hypothesis, a one-way ANOVA was applied to the total

32
scores across all three categories of words, as well as to scores for each category across order
conditions. The results revealed that the main effect for order was not significant for any of the
above measures (ps > .70). Thus, the two conditions of order were collapsed into one in the
following analyses.
I expected the cognitive accessibility of aggressive constructs to display a similar pattern
to the main dependent measure of aggression: the provocation and the trigger would
synergistically interact to increase cognitive accessibility of aggression-related constructs,
resulting highest cognitive accessibility in the provocation/trigger condition among the four
experimental conditions. However, although I expected both the strong provocation and the
minor trigger to activate aggression-related thoughts, affects, and arousals, nonetheless, even
though these aggression-related constructs were activated, they might not necessarily elicit an
aggressive action. A certain threshold must be exceeded in order for aggressive action to occur.
Thus, the following contrast weights were used in the a priori test: 3 (provocation/trigger), 2
(provocation only), 0 (trigger only), and -5 (no provocation/no trigger). The results supported my
hypothesis, F(1, 44) = 7.94, p = .007, partial η
2
= .154 (see Figure 2), and the residual variance
was not significant, F(1, 43) = .40, p > .50, partial η
2
= .009, signaling that the model did explain
significant amount of systematic variance between the groups. Next, I separately applied the
same contrast weights to each of these three categories of words. The results supported the
hypothesized interaction between provocation and trigger for the action words, F(1, 44) = 13.64,
p < .001, partial η
2
= .238 (see Figure 3), with non-significant residual variance, F(1, 43) = 1.86,

33
p > .10, partial η
2
= .041. The results for the affect and arousal words, however, failed to support
the predicated interaction, F(1, 44) = .124, p > .70, partial η
2
= .003 (see Figure 4); F(1, 44) =
2.73, p > .10, partial η
2
= .059 (see Figure 5), respectively.
Figure 2. Mean percentage score of total words in the word completion task as a function of
the Time 1 provocation and the Time 2 trigger. Higher score indicates that among all
the words that were completed, a larger portion of words was related to aggression.

Figure 3. Mean percentage score of action words as a function of the Time 1 provocation
and the Time 2 trigger. Higher score indicates that in the category of action words that were
completed, a larger portion of words was related to aggression.

34
Figure 4. Mean percentage score of arousal words as a function of the Time 1 provocation
and the Time 2 trigger. Higher score indicates that in the category of arousal words
that were completed, a larger portion of words was related to aggression.

Figure 5. Mean percentage score of affect words as a function of the Time 1 provocation
and the Time 2 trigger. Higher score indicates that in the category of affect words that
were completed, a larger portion of words was related to aggression.

35
Thus, responses on the word completion task partly supported the hypothesis that the
provocation and the trigger would interact to increase cognitive accessibility of
aggression-related cognitions, especially the cognitions about aggressive behavioral intentions.
However, the results for the affect and arousal words were contrary to my expectation.
Comparing the mean for the affect word percentage scores in the provocation plus trigger cell (M
= .21, SD = .12) to the means in the other three cells ( M = .22, SD = .10 for the provocation
alone cell; M = .22, SD = .16 for the trigger alone cell and M = .20, SD = .13 for the no
provocation/no trigger cell), the provocation/trigger cell deviated from the baseline (the no
provocation/no trigger cell) by .01, which was smaller than the sum (.04) of the deviation from
the baseline in the provocation alone and trigger alone cells. Thus, the trigger did not interact
with provocation to produce much greater cognitive accessibility to aggression-related affect than
the provocation alone, trigger alone, or the control condition, which means that we failed to find
the expected disjuntively escalation of cognitive accessibility to aggressive constructs. The same
outcome was found for comparison of arousal word percentage scores: the provocation/trigger
cell (M = .28, SD = .19) deviated from the baseline (the no provocation/no trigger cell, M = .19,
SD = .07) by .09, which was smaller than the sum (.14) of the deviation from the baseline in the
provocation alone (M = .25, SD = .15) and trigger alone (M = .27, SD = .13) cells.

36
Chapter 5: Discussion
As predicted, the results from the current study supported our primary hypothesis: the
Time 1 provocation and the Time 2 trigger would interact with each other to produce
disjunctively escalated aggression toward the target. When there was no Time 1 provocation, the
Time 2 trigger had no moderating effect on the final aggression: the aggression level in the
trigger alone condition was almost the same in the no provocation/no trigger condition. This
indicated that the trigger itself was really trivial and would likely be ignored by the participants
as an unintentional accident. However, when the Time 1 provocation was present, the Time 2
trigger displayed significant moderating effect on aggression: when participants experienced both
the major provocation and the minor trigger, they behaved more aggressively than those who
only experienced the initial provocation or subsequent trigger or neither. Moreover, their level of
aggression was even more than the sum of the aggression for those in the provocation alone and
trigger alone condition.
The current study extended Berkowitz's Cognitive-Neoassociationistic Model and the
triggered displaced aggression experimental paradigm to the direct aggression situation wherein
the target for aggression was the same as the source of the initial provocation and the subsequent
trigger. Our study demonstrated that when a major provocation is followed by a minor trigger,
disjuntively escalation of aggression occurs in direct, as well as displaced aggression. Although I
did not obtain irrefutably convincing evidence for the CNA model, certain aspects of the results
from our study were still consistent with the predication of the CNA model. For example,

37
although the results from the current study did not support the synergistical interaction of
provocation and trigger on cognitive accessibility to aggression-related arousal and affect, they
did support such an interaction on cognitive accessibility of aggression-related behavioral
intentions and general aggression-related constructs. Also, I found that provocation induced
negative affect was associated with negative affective responses to the triggering evaluation; and
stronger negative affective responses to the trigger in turn elicited greater aggressive response to
the target. This pattern of results is consistent with the CNA model: The initial provocation
activates the related representations of negative affect within the cognitive-neoassociationistic
network, these cognitive representations in turn prime greater negative affective responses to the
minor trigger, and then a stronger aggressive response was elicited by these negative affective
responses.
Although the results shown in our study closely resembled those of previous trigger
displaced aggression research, there was a noticeable difference exist in the provocation alone
cell of the current study: the level of aggression in this cell was also substantially higher
compared to that in the trigger alone and control cells. This was different from the finding
obtained in triggered displaced aggression research, wherein the level of aggression was lower in
the provocation alone cell compared to that in the trigger alone and the no provocation/no trigger
control cell. However, this pattern of difference was expected in the current study because the
target for final aggression and the person who emitted the Time 2 trigger was the same as the
person who emitted the Time 1 provocation. In this case, when the participant was previously

38
provoked by the target, even when there was no subsequent trigger emitted by the target the
participant would also display substantial aggression toward the target because of the initial harsh
provocation (although this level would still be less than that exhibited by those who was both
provoked and triggered). As noted, in previous triggered displaced aggression research, although
the target for final aggression was the same person who provided theTime 2 trigger, he/she was
not the person who provided the Time 1 provocation. In this situation, after provoked by another
person first, if the participant did not receive the trigger from the target, then the participant
would consciously inhibit any aggressive response toward the target because the target did
nothing wrong and there was no justification for an aggressive response toward him/her.
Limitations
The current study has limitations such as a relatively small sample size, but the major
limitation was the hypothesis about the cognitive accessibility to aggressive constructs did not
receive systematic support from the current study. As previously stated, the measure of cognitive
accessibility was administered to half of the participants before the final aggression measure and
half after the aggression measure. According to the CNA model as well as the General Affective
Aggression Model (GAAM) (Lindsay & Anderson, 2000), I expected that when participants
expressed their anger out, the cognitive accessibility to aggressive constructs would be
significantly decreased. However, I did not find the expected main effect of the order condition
(see Table 2). Moreover, although the results indicated that the cognitive accessibility of general
aggression-related cognitions and the cognitions about aggressive behavioral intentions were

39
indeed disjunctively increased, as seen in the synergistic interaction between provocation and
trigger, I did not find the expected disjuntively escalation effect of the trigger on the cognitive
accessibility to aggression-related affect and arousal: it did not interact with the initial
provocation to produce a higher level of cognitive accessibility in the provocation/trigger
condition, as compared with the other three conditions.
Table 2
Mean percentage score of the word completion task as a function of the order to administer the
task either before or after the aggression measure. Overall score is shown as well as the score
for each category. Higher score indicates that among all the words, or in each category of
words, that were completed, a larger portion of words was related to aggression.

Order
Mean percentage score of words
Overall Action Arousal Affect
Before aggression
After aggression
0.26
0.27
0.30
0.32
0.25
0.25
0.20
0.22

This might be due to the small sample for the current study, but it was more likely due to
a flaw in the methodology of the current study: the participants were given too much time to
complete this task, especially for the half who were assigned to do this task after the final
aggression measure. The task has 34 words in total, most of them are relatively easy. Therefore, 2
to 3 minutes should be a reasonable time limit for the participants to work out all these items "as
fast as you can". However, in the current study, although participants were told to work on this
task as fast as they can, they did not have a time limit for this task. Most of the participants took
approximately 5 mins to complete this task at their own pace, not following the direction. I
emphasized "as fast as you can" in the directions because I only wanted the participants to write
down the first word, if there was one, that came to their minds, -- not to figure out the answer

40
through hard thinking. Because there was no time limit in the current study, participants had
enough time to figure out the answer, especially for the participants in the provocation only
condition, whose aggression-related constructs were also activated but not as strong as those of
the participants in the provocation plus trigger condition. Thus, these participants were more
inclined and also more likely to figure out aggression-related words when they did not have the
answer at the first glance. This might result in participants in the provocation only cell
completing almost the same number of aggression-related words as participants in the
provocation/trigger cell. If so, this would undermine any systematic interaction between
provocation and trigger on cognitive accessibility to all the aggression-related constructs.
Additionally, the failure to find disjuntively escalation effect of the trigger on the
cognitive accessibility to aggression-related affect and arousal might also reflect a flaw in the
measure I used in the current study. The measure was a modified version of Anderson et al. (2003,
2004) measure by Pedersen et al. (unpublished manuscript). In this word completion task, all of
the items in this task could be made into several possible words. But I noticed the
aggression-related and non aggression-related answers are different in their frequencies
especially for the words in the arousal and affect categories. For example, the item "disg _ _ _
ed" have two possible answers, one is "disgusted", which is considered as aggression-related
affect word, and the other is "disguised", which is not related to aggression. However, it seems
that the aggression-related word "disgusted" is more frequently used than the non
aggression-related word "disguised". More examples could be found as "tense" vs. "terse" for the

41
item "te _ _e", "excited" vs. "expired" for the item "ex _ _ _ed", and "hate" vs "hale" for the item
"ha _ e". Thus, it is possible that the word frequency effect has led to the failure to find
disjuntively escalation effect of the trigger on the cognitive accessibility to aggression-related
affect and arousal. Future research should take the word frequency effect and the time limit into
consideration, and I believe that it will be helpful to reveal the expected significant interaction of
provocation and trigger on the cognitive accessibility to all the aggression-related constructs.

42
References

Anderson, C. A., Carnagey, N. L., & Eubanks, J. (2003). Exposure to violent media: The effects
of songs with violent lyrics on aggressive thoughts and feelings. Journal of Personality and
Social Psychology, 84, 960-971.

Anderson, C. A., Carnagey, N. L., Flanagan, M., Benjamin, A. J., Eubanks, J., & Valentine, J. C.
(2004). Violent video games: Specific effects of violent content on aggressive thoughts and
behavior. Advances in Experimental Social Psychology, 36, 199-249.

Aviles, F., Earleywine, M., Pollock, V., Stratton, J., & Miller, N. (2005). Alcohol’s effect on
triggered displaced aggression. Psychology of Addictive Behaviors, 19, 108-111.

Axelrod, R. (1984). The evolution of cooperation. New York: Basic Books.

Bandura, A. (1973). Aggression: A social learning analysis. Englewood Cliffs, NJ: Prentice-Hall.

Baron, R. A., & Bell, P. A. (1975). Aggression and heat: Mediating effects of prior provocation
and exposure to an aggressive model. Journal of Personality and Social Psychology, 31,
825-832.

Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social
psychological research: Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51, 1173-1182.

Berkowitz, L. (1969). The frustration-aggression hypothesis revisited. In L. Berkowitz (Ed.),
Roots of aggression: A re-examination of the frustration-aggression hypothesis (pp. 1-28).
New York: Atherton Press.

Berkowitz, L. (1989). Frustration-aggression hypothesis: Examination and reformulation.
Psychological Bulletin, 106, 59-73.

Berkowitz, L. (1990). On the formation and regulation of anger and aggression: A
cognitive-neoassociationistic analysis. American Psychologist, 45, 494-503.

Berkowitz, L. (1993). Aggression: Its causes, consequences, and control. New York:
McGraw-Hill.

43
Bushman, B. J., Bonacci, A. M., Pedersen, W. C., Vasquez, E. A., & Miller, N. (2005). Chewing
on it can chew you up: Effects of rumination on triggered displaced aggression. Journal of
Personality and Social Psychology, 88, 969-983.

Denson, T. F., Aviles, F., Pollock, V., Earleywine, M., Vasquez, E. A., & Miller, N. (2008). The
effects of alcohol and the salience of aggressive cues on triggered displaced aggression.
Aggressive Behavior, 34, 25-33.Dollard, J., Doob, L. W., Miller, N. E., Mowrer, O. H., &
Sears, R. R. (1939). Frustration and aggression. New Haven, CT: Yale University Press.

Gouldner, A. W. (1960). The norm of reciprocity: A preliminary statement. American
Sociological Review, 25, 161-78.

Lagerspetz, K., & Bjorkqvist, K. (1994). Indirect aggression in boys and girls. In R. Huesmann
(Ed.), Aggressive behavior. Current perspectives (pp. 131-150). New York: Plenum Press.

Lau, R. R., Sears, D. O., & Centers, R. (1979). The "positivity bias" in evaluations of public
figures: Evidence against instrument artifacts. Public Opinion Quarterly, 43, 347-358.

Lindsay, J. J., & Anderson, C. A. (2000). From antecedent conditions to violent actions: A general
affective aggression model. Personality and Social Psychology Bulletin, 26, 533-547.

Marcus-Newhall, A., Pedersen, W. C., Carlson, M., & Miller, N. (2000). Displaced aggression is
alive and well: A meta-analytic review. Journal of Personality and Social Psychology, 78,
670-689.

Miller, N., & Marcus-Newhall, A. (1997). A conceptual analysis of displaced aggression. In R.
Ben-Ari & Y . Rich (Eds.), Enhancing education in heterogeneous schools: Theory and
application (pp. 69-108). RamatGan, Israel: Bar-Ilan University Press.

Miller, N., Pedersen, W., Earleywine, M., & Pollock, V. (2003). A theoretical model of triggered
displaced aggression. Personality and Social Psychology Review, 7(1), 75-97.

Moyer, K. E. (1976). The psychobiology of aggression. New York: Harper & Row.

Nowlis, V. (1965). Research with the mood adjective checklist. In S. S. Tomkins & C. E. Izard
(Eds.), Affect, cognition, and personality (pp. 352-389). New York: Springer.

Pedersen, W. C., Gonzales, C., & Miller, N. (2000). The moderating effect of trivial triggering
provocation on displaced aggression. Journal ofPersonality and Social Psychology, 78,
913-927.

44
Pedersen, W. C., Bushman, B. J., Vasquez, E. A., & Miller, N. (2008). Kicking the (barking) dog
effect: The moderating role of target attributes on triggered displaced aggression.
Personality and Socical Psychology Bulletin, 34(10), 1382-95.

Rook, K. S., Sears, D. O., Kinder, D. R., & Lau, R. R. (1978). The "positivity bias" in evaluations
of public figures: Evidence against interpersonal artifacts. Political Methodology, 5,
469-499.

Sears, D. O. (1983). The person-positivity bias. Journal of Personality and Social Psychology, 44,
233-250.

Vasquez, E. A., Denson, T. F., Pedersen, W. C., Stenstrom, D. M., & Miller, N. (2005). The
moderating effect of trigger intensity on triggered displaced aggression. Journal of
Experimental Social Psychology, 41, 61-67.

Vasquez, E.A., Ensari, N., Pedersen, W. C., Tan, R.Y ., & Miller, N. (2007). Personalization and
differentiation as moderators of triggered displaced aggression towards outgroup targets.
European Journal of Social Psychology, 33, 297-319.

Worchel, P. (1966). Displacement and the summation of frustration. Journal of Experimental
Research in Personality, 1, 256-261.

Abstract (if available)

Abstract Consistent with previous research on Triggered Displaced Aggression and the predictions of the Cognitive-neoassociationistic model, the current research found that a major provocation followed by a minor triggering event resulted in disjunctive escalation of aggression which exceeded the level that was predicted by the additive combination of the independent aggression-eliciting effects of the initial provocation and the subsequent triggering event. Thus, it was demonstrated that the synergistical interaction between the Time 1 provocation and Time 2 trigger occurs not only in displaced aggression, but also in direct aggression. Mediation analyses showed that for participants who had been provoked, both negative affects and attributional distortions of the triggering event mediated the effect of the trigger on aggression.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Frustration within the triggered displaced aggression paradigm

PDF

Children's triggered displaced aggression

PDF

The angry brain: neural correlates of interpersonal provocation, directed rumination, trait direct aggression, and trait displaced aggression

PDF

The impact of naturally-occurring rumination on aggression

PDF

Implicit theory ideology about human attributes and aggression

PDF

Projection on outgroups: complementary projection in aversive intergroup contexts

PDF

Mediators and moderators in the link between maternal psychological control and peer victimization for Hong Kong Chinese boys

PDF

The effect of massed versus spaced arguments on persuasion

PDF

Emotion regulation as a mechanism linking parents’ marital aggression to adolescent behavioral problems: a longitudinal analysis

PDF

Investigating the effects of personality factors on goal shielding

PDF

Marital conflict and parenting: moderating effects of ethnicity and parent sex

PDF

Examining the longitudinal relationships between community violence exposure and aggressive behavior among a sample of maltreated and non-maltreated adolescents

PDF

Spouse aggression, depression, and physical health: a multivariate longitudinal study of midlife couples

PDF

Observations and modeling of dynamically triggered high frequency burst events

PDF

Stigma-based peer aggression and social status in middle adolescence: the unique implications of weight-related aggression

PDF

Effects of racial exclusion from media

PDF

Effects of nicotine abstinence on orienting, executive function, arousal and vigilance

PDF

The effects of familism and cultural justification on the mental and physical health of family caregivers

PDF

Accuray of subjective cognitive complaints in a longitudinal context: the effect of depression and dementia status

PDF

Social identification and the treatment of in-group deviants: the black sheep effect or the devil protection effect

Asset Metadata

Creator Yang, Linli (author)

Core Title The moderating effect of triggering provocation on direct aggression

School College of Letters, Arts and Sciences

Degree Master of Arts

Degree Program Psychology

Publication Date 11/11/2009

Defense Date 10/21/2009

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag aggression,OAI-PMH Harvest,provocation

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Miller, Norman (committee chair), Schwartz, David (committee member), Wood, Wendy (committee member)

Creator Email icy_1121@hotmail.com,linliyan@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m2733

Unique identifier UC1279923

Identifier etd-Yang-3368 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-270780 (legacy record id),usctheses-m2733 (legacy record id)

Legacy Identifier etd-Yang-3368.pdf

Dmrecord 270780

Document Type Thesis

Rights Yang, Linli

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu