University of Southern California Dissertations and Theses

Atypicality: Benefit and bane. Provocation, trigger, typicality, and the expression of aggression and generalization

(USC Thesis Other)

Atypicality: Benefit and bane. Provocation, trigger, typicality, and the expression of aggression and generalization

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films
the text directly from the original or copy submitted. Thus, some thesis and
dissertation copies are in typewriter face, while others may be from any type of
computer printer.
The quality of this reproduction is dependent upon the quality of th e
copy submitted. Broken or indistinct print, colored or poor quality illustrations
and photographs, print bleedthrough, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send UM I a complete manuscript
and there are missing pages, these w ill be noted. Also, if unauthorized
copyright material had to be removed, a note w ill indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand comer and continuing
from left to right in equal sections with small overlaps.
ProQuest Information and Learning
300 North Zeeb Road, Ann Arbor, M l 48106-1346 USA
800-521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ATYPICALITY: BENEFIT AND BANE
PROVOCATION, TRIGGER, TYPICALITY, AND THE EXPRESSION OF
AGGRESSION AND GENERALIZATION
by
Candace L. Gonzales
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment o f the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PSYCHOLOGY)
May 2002
Copyright 2002 Candace L. Gonzales
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 3073782
Copyright 2002 by
Gonzales, Candace Letitia
All rights reserved.
__ ___ __®
UMI
UMI Microform 3073782
Copyright 2003 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
The Graduate School
University Park
LOS ANGELES, CALIFORNIA 90089-1695
This dissertation, w ritten b y
Under th e direction o f Aer... D issertation
Com m ittee, and approved b y a ll its members,
has been presented to and accepted b y The
Graduate School, in partial fulfillm ent o f
requirem ents fo r the degree o f
DOCTOR OF PHILOSOPHY
..................... t .t ;................................ ................
/ C -----Bean o f Graduate Studies
D ate December 18, 2002
DISSER TA TION COMMITTEE
Chairperson
P . 0 &
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ii
Table o f Contents
List o f Figures iii
Abstract V
Chapter 1- Introduction I
Chapter 2 - Method 27
Chapter 3 - Results 32
Chapter 4 - Discussion 77
References 98
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ill
List o f Figures
Figure 1. Predicted mean effects o f aggression as a function o f typicality and
trigger under conditions o f no provocation. 23
Figure 2. Predicted mean effects o f aggression as a function o f typicality and
trigger under conditions o f provocation. 24
Figure 3. Predicted mean effects o f stereotype generalization as a function o f
typicality and trigger under conditions o f no provocation. 25
Figure 4 . Predicted mean effects o f stereotype generalization as a function o f
typicality and trigger under conditions o f provocation. 26
Figure 5. M ean effects o f aggression as a function o f typicality and trigger under
conditions o f no provocation. 36
Figure 6 . M ean effects o f aggression as a function o f typicality and trigger under
conditions o f provocation. 38
Figure 7. Mean effects o f aggression as a function o f typicality, trigger, and
provocation for Caucasians. 41
Figure 8. Low state self-esteem as a function o f typicality, trigger, and provocation
for Caucasians. 43
Figure 9. Application reliance as a function o f typicality, trigger, and provocation
for Caucasians. 44
Figure 10. Interaction o f expectancy violation and aggression toward a typical
target as a function o f trigger and provocation for Caucasians. 46
Figure 11. Interaction o f perceptions and aggression toward an atypical target as a
function o f trigger and provocation for Caucasians. 48
Figure 12. Mean effects o f aggression as a function o f typicality, trigger, and
provocation for Asians and Hispanics. 5 1
Figure 13. Low state self-esteem as a function o f typicality, trigger, and provocation
for Asians and Hispanics. 52
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
V
Abstract
Provocation (incidental negative affect) was found to moderate the effect of
typicality and triggering actions (minor target transgressions) on intergroup aggression and
stereotype generalization. Atypicality was a benefit to African American targets under
conditions o f no provocation. However, atypicality was a bane to targets under conditions of
both provocation and trigger. Atypical targets were the victims of greatest aggression after
ego threat. Assessment o f predictors revealed that for provoked Caucasians, a triggering
atypical target was the source o f greater negative affect than a typical target. Expectancy
violation, low state self-esteem, and negative affect predicted aggression. Exposure to
atypical targets also impacted stereotype generalizations. Counterintuitively, atypical targets
impacted stereotype generalizations more than typical targets. Predictors o f aggression were
found to suppress the expression o f generalization. For Caucasians, affect, low state self
esteem, and expectancy violation augmented the expression of displaced aggression as
evidenced in repeated measures analyses. The suppression o f generalization under
provocation may be due to stereotype activation and social desirability. Few moderating
effects were found for typical African American targets. Asians and Hispanics had similar
responses toward atypical targets. However, Asians and Hispani ' aggressed and generalized
more than Caucasians after exposure to typical targets than atypical targets.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
PROVOCATION, TRIGGER, TYPICALITY, AND THE EXPRESSION OF
AGGRESSION AND GENERALIZATION
Chapter 1 - Introduction
Overview
How does prior negative affect interact with characteristics and behaviors of a single
outgroup member in expressions of aggression and perceptions of the group? In other words,
after an unrelated bad expereince, do atypical African Americans still benefit from positive
associations or are they treated as negatively as their typical brethren and why? Further,
does this treatment generalize to statements made regarding the group? These are my
primary interests in this dissertation.
Purpose
Three research questions are posed in this paper: (a) How do the attributes
(typicality) and behaviors (trigger) of African Americans influence the expression o f direct
aggression (no-provocation) and displaced aggression (provocation)? (b) How do the
attributes and behaviors of a single African American influence the endorsement of
stereotype generalization? and (c) How do the underlying processes that result from
provocation and trigger influence the expression of direct aggression, displaced aggression,
and generalization?
Organization of the Dissertation
First, the dissertation reviews the general theoretical background of aggression,
beginning with its proposed antecedents, followed by a discussion o f displaced aggression
and interracial aggression. Second, the characteristics or traits o f the target as potential
moderators in aggression and generalization are examined. Third, the dissertation reviews
the literature on affect and cognition as mediators in aggression. Next, the study is presented.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The study examines how provocation, triggering action, and typicality effect displaced
aggression and stereotype generalization.
Aggression
Antecedents
In the following section I discuss the antecedents of direct aggression. Prior to a
discussion of displaced aggression, a rudimentary discussion of the proposed antecedents
and measures of aggression is warranted. Aggression is defined as engaging in an action
with the intent to injure another (Dollard, Doob, Miller, Mowrer, & Sears, 1939) Frustration,
provocation, or more generally, any negative affect is proposed to be a necessary antecedent
to aggressive responses. (Berkowitz, 1989). Negative affect is induced in the lab both
socially and non-socially, (Carlson & Miller, 1988). Socially instigated aggression, such as a
personal insult, has resulted in greater aggressive responding toward the provocateur than
non-socially instigated aggression. In addition, there is a moderate positive relationship
between socially instigated provocation and aggression toward an innocent bystander
(Marcus-Newhall & Pederson, 2000). Within socially instigated negative affect, attack upon
a person’s self has been suggested as a necessary precursor to aggression. Insult directed at a
person has been found to be more powerful in inducing displaced aggression than frustration
alone (Geen, 1968). Within non-socially instigated aggression, fhistrations that are due to
illegitimate or unexpected causes have resulted in stronger aggressive responses (Kulik &
Brown, 1979). The dissertation uses a composite of both social and non-social inductions to
motivate aggression.
Aggressive responses, as measured in social psychology, may be separated into two
types, behavioral measures and evaluative measures. Behavioral indicators o f aggression
include, intensity of shock administered (Konicni & Doob, 1972), coldness o f water immersion,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
(Berkowitz, 1989; Berkowitz, Cochran, & Embree, 1981), and intensity o f aversive noise
(Johnson & Rule, 1986). Evaluative measures include negative written evaluation, (Carlson &
Miller, 1988; Bell, 1980) and negative verbal response (Kulik & Brown, 1979). Evaluative
measures are typically used as ecologically valid measures of aggression. There is support for
the use of evaluative measures of aggression as they are highly related to more direct behavioral
forms of aggression, and they are influenced by similar antecedents (Carlson, Marcus-Newhall,
& Miller, 1989).
The current work uses an evaluative response and a more behavioral response to
measure aggression. These scales are defined as measures of aggression because the participants
are cognizant of the target’s desire for a goal, a job, and are also cognizant o f their own power
to prevent the target from obtaining the goal. Goal prevention will be harmful both financially
and psychologically to the target. This response is defined as aggressive because it fulfills the
requirements set forth by Geen (1990); there is an intentional delivery of a negative stimulus
with the express purpose to harm a target. Further, participants have the expectation that the
target will suffer as a result of their actions.
The aggression expressed in the study is both interpersonal and intergroup. The
aggression is interpersonal, as there are only one participant and one target during the course
of the experiment. It can also be conceptualized, however, as intergroup. A dyadic
interaction may be defined as intergroup if the individuals are perceived during the
interaction as group representatives (Brewer & Miller, 1995). The paradigm used in the
current work makes both individuating information and group membership salient.
Thus, the aggression in the dissertation as a result of insult and unexpected noxious
stimuli is expressed in an evaluative and a more behavioral response. In the following
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
section, I discuss aggression toward a source not entirely responsible for the aggressive
impulse, displaced aggression.
Displaced Aggression
In this section I begin with a brief history of displaced aggression research.
Following that, I describe research that illustrates factors that maximize the expression of
displaced aggression. This research provided the basic theoretical building blocks on which
the dissertation paradigm is constructed.
Aggression may be augmented toward a target if the aggressor has been previously
provoked. Displaced aggression is an action undertaken to injure a second party, or target, as
a result of a prior negative experience with a first party, or provocateur, in circumstances that
have precluded retaliation against the first party (Dollard et al., 1939: Miller, 1948).
The displacement o f aggression from a primary to a secondary source was proposed
as early as 1939 (Dollard et. al., 1939). Miller (1948) made several predictions regarding the
conditions under which displaced aggression occurs. Aggression is suggested as most likely
to be displaced onto a target that is similar to the original provoking source (Miller, 1948). If
the strength of response to the original provocateur is sufficiently strong, however, dissimilar
targets will also prompt aggressive responses. Subsequent research has supported the belief
that association with the original provocateur is not necessary for displaced aggression to
occur (Fenigstein & Buss, 1974). Innocent targets, however, are not invariably the victims of
displaced aggression. Berkowitz & Knurek (1969) found ‘a contrast effect’ in which
individuals who had been previously angered responded more favorably to innocent targets
than individuals who were not previously angered.
Worchel (1966) proposed a sensitization theory of displaced aggression. Although
aggressive responding is normally inhibited, Worchel (1966) theorized that when an
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
individual is already in a negative affective state, or provoked, the magnitude o f this
inhibitory tendency is decreased toward the displacement target. If another small threat
appears, the individual reacts to that small threat in a manner that is inappropriate to the
threat. Therefore, retaliation occurs because of the lowered threshold produced by the earlier
provocation.
A Meta-analysis o f displaced aggression identified several factors that moderate
aggressive responding (Marcus-Newhali, Pederson, Carlson, & Miller, 2000). They include
the type of provocation, the instigator of the provocation, and the target o f aggression.
Among types of provocation, participants who were both verbally attacked and experienced
a negative evaluation exhibited the greatest displaced aggression. Among types of
instigators, participants who were provoked by the experimenter exhibited the greatest
displaced aggression. With respect to targets of aggression, participants who were allowed to
aggress against an experimental assistant exhibited the greatest amount of displaced
aggression. In addition, the meta-analysis found less displaced aggression toward targets
who were similar to the provocateur.
The meta-analysis provides guidance on how to structure a paradigm in order to
maximize displaced aggression. The paradigm used in the dissertation is based on the results
o f this meta-analysis. Each aspect of the experimental setting was chosen to maximize
displaced aggression. The provocation experienced will be both negative evaluation and
verbal attack, the experimenter will provide the provocation, and the target will be an
experimental assistant. Also, the experimenter will be of a different racial background than
the target.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
Triggered Displaced Aggression
The following section considers the target of displaced aggression, previously
presented as largely blameless, as a contributing factor in aggression. I briefly review the
concept of triggered displaced aggression and introduce part of the dissertation design.
The target’s behavior, which is in no way associated with the original provoking
experience, may contribute to the expression of displaced aggression. The target of displaced
aggression has been traditionally operationalized as largely blameless (Fitz, 1976). However,
displaced aggression may be augmented in reaction to an additional negative action by a
target, however trivial (Dollard, 1938; Worchel, 1966). A slightly negative, triggering action
by a target, sans provocation, may be interpreted as ambiguous. When combined with a prior
provoking experience the triggering action, however, it may be interpreted quite negatively
and cause a counteraction from an individual that is more extreme than would occur in
response to the triggering action alone (Pederson, Gonzales, & Miller, 2000).
In triggered displaced aggression, the minor triggering action by the target follows
an earlier provocation of greater magnitude. This minor event may be interpreted as
ambiguous in the absence of prior provocation, however, in conjunction with prior
provocation, a minor trigger has been shown to elicit greater aggression, that trigger or
provocation alone (Pederson, Gonzales, & Miller, 2000).
The paradigm of the dissertation includes a 2 (provocation; present or absent) x 2
(trigger: present or absent) between subjects factorial design. I present part o f the design here
in order to introduce the reader to preliminary predictions regarding the actions by the target,
or trigger, and the experience of prior negative affect, or provocation. Based on the results of
Pederson, Gonzales, & Miller, 2000, under conditions of no provocation, there should only
be a small difference in aggression expressed against triggering targets and non-triggering
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
targets. Under conditions o f provocation, however, aggression was expected to be
augmented more against triggering targets than non-triggering targets resulting in a large
differential.
Interracial Aggression.
Prior to the presentation of aggression toward different types of African Americans,
some discussion of interracial aggression is warranted. In the following section, I discuss
interracial aggression and its links to displaced aggression. 1 also review variables that
contribute to interracial aggression in order to use them to maximize aggressive responding in
the dissertation study.
The target’s status as an outgroup member may also contribute to the expression of
displaced aggression. The mere presence of an outgroup category member may engender
aggressive responses (Wilder, 1978). One real category o f interest is race. Race is one o f the
primary categories used to classify others (Fiske & Neuberg, 1990). Race is important in
intergroup aggression research as American history is filled with instances where violent or
hostile actions were committed against others solely due to race (Mullen, 1986). Currently,
race is in the forefront of many political and domestic disputes. Displaced aggression has
been suggested as a potential basis for hostile relations between racial groups, specifically
African Americans and Caucasians (Hovland & Sears, 1940; Zadawalski, 1948).
A classic theory of displaced interracial aggression was the scapegoat theory of
racism. Although the theory is no longer a topic of research, it bears mentioning. The theory
proposed that minority groups were natural scapegoats to Caucasian’s frustrations due to
large power differentials. Zawadzki (1948) stated these aggressive responses might have
been prompted by the unfamiliar and perhaps disagreeable behaviors o f the minority groups.
In support o f a scapegoat theory of prejudice, Cowen, Landes and Shaet ( I9S9) found an
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
increase in the expression of negative attitudes towards African Americans after a frustration in
a simple pre test-post test design.
Several variables that affect aggression in general have also been found to effect
interracial aggression. These variables include participant anonymity (Donnerstein &
Donnerstein, 1973; Rogers & Prentice-Dunn, 1981), potential ingroup censure (Donnerstein &
Donnerstein, 1973), and potential target retaliation (Donnerstein & Donnerstein, 1978).
Participant anonymity has been found to increase aggressive responses. Caucasian
males were more likely to engage in direct aggression, as measured by shock intensity, against
an African American male target if they did so anonymously (Donnerstein & Donnerstein,
1973). Anonymity to both target and experimenter increased aggressive responses toward
African American targets (Donnerstein, Donnerstein. Simon. & Ditrichs, 1972). Greater levels
of deindividuation experienced by experimental participants increased aggressive responses
to outgroup members (Rogers & Prentice-Dunn, 1981).
Potential ingroup censure also decreases aggressive responses. Caucasian males were
less likely to aggress against an African American male target if an ingroup member could view
their actions, which the researchers defined as potential ingroup censure( Donnerstein &
Donnerstein, 1973).
The inability for the target to retaliate against the aggressor has been shown to
increase aggressive responses. Caucasians were likely to exhibit more aggression against an
African American target that could not retaliate than a Caucasian that could not retaliate
(Donnerstein et al., 1972).
In the current work, all three variables are taken into consideration in order to
maximize aggressive responding. Participants believe the expression of aggression is
anonymous. Participants do not believe there is potential for ingroup censure because the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
experimenter is not expected to review their responses. Finally, participants do not believe
that there is any potential for target retaliation, as participants do not expect to interact with
the target in person.
As mentioned previously, target behavior, is expected to impact the expression of
aggression. Therefore, research into the effect of an outgroup target’s behavior on intergroup
interaction, including aggression, is germane to the proposed work. An outgroup member’ s
behavior, whether positive or negative, has been shown to greatly alter responses toward that
outgroup member. A similar positive or negative behavior by an ingroup member only
slightly alters responses toward that ingroup member. This effect is referred to as response
amplification. Response amplification is the tendency for ingroup members to be more
extreme in their evaluations of and/or behaviors towards outgroup members than o f other
ingroup members (Gaertner& Dovidio, 1989; Rogers & Prentice-Dunn, 1981; Hass, Katz,
Rizzo, Bailey, & Eisenstadt, 1991). Thus, positive experiences by Caucasians with African
Americans produce more positive evaluations o f African Americans than positive
experiences with other Caucasians. In addition, negative interactions with African
Americans produce more negative responses than negative interactions with fellow
Caucasians (Hass et.al., 1991). Hass et al. (1991) proposed that these results were due to the
ambivalent feelings Caucasians have toward African Americans. The authors suggest that
because these feelings are ambivalent, they are guided by situational cues. Thus, in a
negative contact situation the African American is perceived negatively and in a positive
contact situation, the African American is perceived positively.
It is perhaps impossible to discuss responses made toward African Americans
without taking into consideration social desirability. Social desirability, the need to appear
egalitarian to the self and to others, was expected to temper interracial aggression. Social
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
desirability has been found to be negatively related to aggressive responses (Harris, 1997).
Social desirability concerns may be elevated in intergroup contexts. An individual’s desire to
appear non-prejudiced may increase the positivity of target evaluations o f African American
targets. Individuals may shy away from not recommending a target for a coveted job and
instead recommend a target to a lesser degree. As a result, target evaluations may not reach
below midpoint. This expectation is supported in the literature. Research has shown that
although Caucasians are less apt to endorse negative ratings o f African Americans, they are
also less apt to rate them positively (Gaemter & McLaughlin, 1983).
Typicality and Displaced Aggression.
In the following section, I introduce the idea that the type of African American
target may impact triggered displaced aggression. I begin with a discussion of typicality and
how it is defined in the literature. I conclude the section with predictions regarding the
impact o f typicality on aggression and the reasons for these predictions.
As mentioned previously, the differential effects o f race have been well established.
Caucasians, under certain conditions respond more negatively toward African Americans
than toward other Caucasians. As a result, comparing displaced aggression displayed toward
a Caucasian to a African American was not as interesting to me as comparing displaced
aggression displayed toward different ‘‘ types” of African Americans. My interest was in the
moderating effect that particular attributes o f racial outgroup targets have on aggressive
responses. Because of this interest I added an additional factor, typicality, to the dissertation
design. Therefore the design of the experiment in its entirety is a 2 (provocation: present or
absent) x 2 (trigger present or absent) x 2 (target: typical, atypical) between subjects
factorial design.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
II
Target attributes have been suggested as either eliciting or ameliorating aggressive
responses (Berkowitz, 1982). One attribute o f interest is how typical an outgroup target is o f
the group in general. Racial outgroup members are commonly characterized through a group
of descriptive characteristics, referred to as stereotypes (Gaertner & Dovidio, 1983). An
outgroup member who embodies most of the group’s assumed traits is referred to as
stereotypical. An outgroup member who does not embody these traits or one who has traits that
are in direct opposition to the stereotype is referred to as counterstereotypical. For the purpose
of the proposed work, the terms typical and stereotypical will be used interchangeably as will
the terms atypical and counterstereotypical.
Stereotypes are commonly confounded with valence. Stereotypes o f outgroups are
usually negatively valenced (Dovidio, Evans, & Tyler, 1986; Devine & Baker. 1991
Hewstone & Lord, 1998). Thus, typical members of a generally negatively perceived outgroup,
such as African Americans, are usually viewed negatively. Atypical members may be viewed
more positively or less negatively (Wilder 1984; Hewstone & Lord, 1998). As mentioned
previously, these effects are expected regardless o f perceivers’ personality characteristics.
Common stereotypes of African Americans are widely known and automatically activated
by prejudiced and non-prejudiced participants alike (Devine, 1989).
The degree to which a group member is typical or representative of his or her category
has been shown to moderate the behaviors toward (Wilder et. al., 1996), attitudes about (Werth
& Lord, 1992), and perceptions of (Lambert & Wyer, 1990) outgroup members. Stereotypic or
counterstereotypic traits are expected to moderate the effect of triggered displaced aggression
on African Americans.
Under conditions of no provocation, one who is more closely associated with a
negatively perceived outgroup (e.g., a typical African American) was expected to receive
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
greater aggression than one who was less associated with the outgroup (e.g., an atypical African
American).
For a typical target, there are both cognitive and affective explanations for the
aggression. First, a target may be ascribed additional negative traits also associated with the
group. Once an individual is placed into a category, he may be credited with the characteristics
of other category members (Devine & Baker, 1991). If an individual is highly associated with a
negatively stereotyped group, he may be credited with the negative characteristics believed to
typify the group. Second, the negative affect felt toward the group may be directed at a typical
group member. It is important to distinguish the ascription of negative traits from aggression.
The ascription of negative traits is the perception of the individual and aggression is the
intention of harm. In the current study, although they are assessed, the ascription of negative
traits is not the measure of aggression. Aggression may be elevated for a typical target due to
these negative perceptions and affective responses.
For an atypical target there are contrasting explanations for the predicted results. An
individual not as associated with the negatively perceived outgroup may not be victim to
adverse perceptions or feelings. An atypical target may be more individuated and therefore less
likely to be ascribed the negative traits of the group as a whole (Hewstone & Lord, 1998). An
atypical target may induce less negative affect than a typical outgroup member. Thus,
aggression may be reduced toward an atypical target as a result of positive perceptions and
affect. Finally, the differential reaction to typical and atypical outgroup members may be due to
a combination of the two effects, an elevation of aggression toward a typical target and a
suppression of aggression toward an atypical target.
This general expectation under conditions o f no provocation must be qualified by the
behavior o f the outgroup target, whose actions may be either benign or provoking. For the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
dissertation, targets either engage in benign, non-triggering behavior or more culpable,
triggering behavior. The culpable behavior was purposefully operationalized as only slightly
negative.
Under conditions o f no provocation and no trigger, equal low levels of aggression are
expected. Aggressive impulses are expected to be tempered toward the typical target due to
social desirability concerns. With no justification for aggression, automatic reactions in
response to a typical outgroup member may be consciously suppressed.
Under conditions o f no provocation and trigger, more aggression is expected toward a
typical target as compared to an atypical target. The addition of trigger is expected to reduce the
suppression of negative affect toward the target. The trigger serves two functions. First, it may
temper social desirability concerns, by providing justification for aggression. Second, the nature
of the trigger may further associate the target with the already negatively perceived outgroup by
providing confirmatory evidence, albeit weak, for the negative perception. Triggering behavior
on the part of a typical target is expected to increase aggression for either reason. However,
triggering behavior on the part of an atypical target is not expected to increase aggressive
responses. As mentioned previously, an atypical target is likely to be subtyped or only weakly
associated with the outgroup. An atypical target should not be victim to the prior negative
expectations that may increase the salience of the trigger. A triggering action of low intensity
should not be sufficient to elicit aggression. Thus, even with the addition of the triggering
behavior, under conditions of no prior provocation, significantly more aggression was expected
toward a typical target than toward an atypical target.
Under conditions of provocation and no trigger, more aggression is expected against a
typical target than an atypical target. The negative affect experienced during the provocation is
expected to reduce socially desirable responding even with the absence of justification and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
stereotype behavior confirmation. Under conditions of provocation and trigger, however,
participants are expected to aggress against typical and atypical targets equally. The increased
negative affect that results from provocation is likely to focus attention toward a targets’
outgroup status. Under conditions of provocation, a trigger provides the needed justification to
aggress against all outgroup members irrespective o f the degree to which they are typical of the
social category. The explanation for this prediction is discussed further in the affect and
cognition section.
Stereotype Generalization
Typicality and Stereotype Generalization
In the following section 1 introduce stereotype generalization as part of the
investigation. Due to the my use of typical and atypical characteristics, 1 was initially interested
if the aggression expressed toward a single outgroup member would correspond to prior
literature on perceptions of outgroup members as a whole. Examination of the literature, which
follows, suggests that generalization will not mirror direct aggression. As a result of this
interest and as a natural follow up to the typicality literature, a generalization measure was
included in the dissertation. The second research question, “how do the attributes and behaviors
of a single African American influence the endorsement of stereotype generalizations?” was
added. Following a discussion of research relevant to the predictions, I present the predictions
for stereotype generalizations and the reasons for those predictions.
The impact of typicality on stereotype change or stereotype generalization has been
studied extensively (Hewstone, Johnston & Aird, 1992; Hewstone & Lord, 1998: Werth &
Lord, 1992). Stereotype perceptions of outgroups are mostly negative (Hewstone & Lord,
1998). Brewer & Miller (1988) proposed three types of generalization; “changes in attitudes
toward the category, increased complexity of intergroup perceptions, and decategorization.”
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
The dissertation is concerned with a combination of the first and second type of generalizations:
first, changes in negative attitudes toward or perceptions of the category; second, perceiving the
group as more or less complex after exposure to a single category member. In the study, items
measuring stereotype generalization items are negatively valenced. In addition, the
generalization items suggest that a large number of African Americans possess the trait.
Endorsement of these items will reflect a negative perception of African Americans in general
and a belief that the characteristic is reflective of most African Americans. Endorsement of
scale values on stereotype generalization items will reflect less negative perceptions of African
Americans and more complex perceptions of the group.
Generalization from one exemplar o f the group is more likely when that exemplar
embodies many of the attributes associated with the group (Rothbart and John, 1985). Group
members that do not embody those attributes generally have little effect on the perceptions
of the group (Rothbart & Scott, 1988; Rothbart, 1996). Thus, exposure to extremely atypical
group members has resulted in little stereotype change (Weber and Crocker, 1983) Atypical
group members may not ordinarily impact the stereotype due to the reassignment of these
individuals into a separate category or subtype. Subtyping occurs when group members are
isolated into subgroups of '‘exceptions to the rule” (Allport, 1954; Weber & Crocker. 1983). A
single atypical group member may also be individuated, seen as a unique individual, and thus
have little bearing on group perceptions. Thus, the positive attributes of an atypical group
member do no effect group stereotypes (Hewstone & Lord, 1998).
Contrary to most findings in the area o f stereotype change and typicality, there are
circumstances when exposure to atypical group members is predicted to strengthen
prevailing negative stereotypes. Boomerang effects have been proposed as a possible
reaction to extremely deviant examples (Kunda and Olsen, 1997). Boomerang effects are
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
contact with the typical mental patient. If non-triggering interaction is interpreted as a
positive experience, when it occurs with a typical target it should result in less negative
stereotype generalizations than interaction with a non-triggering, atypical target, under
conditions of no provocation. A negative or triggering interaction should reverse this effect.
Under conditions of no provocation, exposure to non-triggering typical targets was expected
to decrease negative generalizations more than exposure to non-triggering atypical targets.
Conversely, exposure to triggering typical targets was expected to elevate negative
generalizations more than exposure to triggering atypical targets.
Research into the impact of stereotype disconfirming behaviors by typical group
members on stereotype change found that disconfirming or atypical behaviors on the part of
typical targets result in greater stereotype change or reduced stereotype generalizations (Wilder,
1996; Rothbart & John, 198S). Although, the goal of researchers in this area has been to
investigate how stereotypes change for the better, the research may be incomplete, as it has not
investigated the impact of confirming (typical) behavior by otherwise atypical targets, in other
words, how stereotypes may change for the worse. This may be due to past research that
demonstrates that atypical group members in general do not impact stereotype change, as
mentioned previously. It appears, however, to be a required investigation. Individual atypical
group members are group members after all. Although an atypical group member may be
subtyped, one negative behavior on his/her part may be sufficient to strengthen negative
attitudes directed toward the group. A more thorough investigation of this area should include a
complete 2 (disconfirming behavior, confirming behavior) x 2 (typical group member, atypical
group member) factorial design. The proposed work includes this design with confirming
behavior operationalized as triggering actions and disconfirming behavior operationalized as
non-triggering actions.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
Can aggressive impulses extend from a single incident with an individual group
member to an indictment of the group as a whole? The majority of studies concerning subtyping
and stereotype generalization have had participants experience a number of outgroup members.
Experience with one outgroup member, however, can impact a perceiver’s outgroup stereotype,
if the behavior exhibited by the outgroup member is viewed as dispositional (Wilder, Simon, &
Faith, 1996). Wilder, Simon, and Faith (1996) found that behaviors exhibited by a single typical
target affected group stereotypes if the behavior was perceived as internal and stable. If the
dispositional behavior was attributed to an atypical target, there was little effect on group
stereotypes. Regarding African Americans in particular, research has shown that even a short
interaction with a single African American can generalize to negative perceptions o f the group
(Henderson-King & Nisbett, 1996). For the proposed study, the behaviors exhibited by targets
will be largely dispositional. Thus, exposure to a single outgroup member was expected to
impact stereotype generalizations.
Intervening Processes for Aggression and Stereotype Generalization
A secondary interest o f the dissertation is the intervening variables that contribute to
the expression of aggression and generalization, thus, the third research question; '‘How do
underlying processes influence the expression o f direct aggression, displaced aggression and
generalization?” In the following section I review affective and cognitive variables that have
been proposed to impact the expression of aggression. I begin with a short review o f the
literature. Next, I discuss how social identity theory may provide an explanation for
predicted results. Measures of variables that may support the results in terms o f social
identity theory are included in the study.
Affect and cognition are both proposed to mediate aggressive responding. Affective
variables include frustration (Berkowitz, 1982), negative affect (Berkowitz, 1989), anger
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
(Fitz, 1976), and fear (Fitz, 1976). Cognitive variables include low self-esteem (Bushman &
Baumeister, 1998), ego threat (Beaumeister & Boden, 1998), and impression management
(Melburg & Tedeschi, 1989). These variables may be heightened in an intergroup context.
These variables also may interact uniquely in the expression o f intergroup displaced
aggression.
It is not only the objective nature of the negative event that engenders aggressive
responding, but also the subjective experience of that negative event (Berkowitz, 1982). As
mentioned previously, both affective and cognitive variables have been proposed to mediate
aggressive responses.
Affective variables that have been proposed to mediate aggressive responses
include, frustration (Berkowitz, 1982), negative affect (Berkowitz, 1989), anger (Fitz, 1976),
and fear (Fitz, 1976). Negative mood has been found to increase aggressive responses.
Berkowitz and Troccoli (1990) found participants who were made aware of their
experimentally induced negative affect were less likely than those who were unaware o f
their affect to behave negatively toward a neutral target. Thus, individuals who are in a
negative mood but unaware of it were expected to aggress against outgroup members to a
greater extent. Further, manipulated anger and fear have produced the greatest aggression
toward displaced targets (Fitz, 1976). Affective measures have accounted for more variance
in evaluating stereotyped outgroup members than cognitive measures (Jackson and Sullivan,
1988). Negative affect was expected to take up cognitive capacity such that individuals were
able to focus only on a limited number of target cues, primarily their group category. For the
dissertation negative affect was assessed after all conditions. The relationship between
negative affect and aggression is therefore assessed correlationally.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
Cognitive variables that have been proposed to mediate aggressive responses include
ego threat (Baumeister & Boden, 1998) and impression management (Melburg & Tedeschi,
1989). These cognitive variables have been found to increase aggressive responses. Ego-
threat, a challenge to an individual’s self-perceptions, has been found to increase aggressive
responses. Bushman and Baumeister (1998) found that individuals who had a heightened
self-regard were more aggressive when their self-impressions were threatened. Impression
management has also been found to increase aggressive responses. Melburg & Tedeschi
(1989) showed that participants delivered more shocks to individuals whose task
performance was superior to their own than to individuals whose performance equaled their
own. In response to these effects, I included a measure of ego threat.
Thus, both affective and cognitive subjective experiences o f the participant were
expected to predict aggressive responses. However, results were expected to differ
depending on which was the primary mediating factor. If affect was the primary mediating
factor, under conditions of provocation and trigger, participants would be unable to
differentiate between the typical and the atypical targets, reacting similarly to both as
outgroup members. Their ability to differentiate between the targets would be decreased due
to the cognitive capacity taken up by negative affect. As a result, participants were expected
to react to the most crucial cue, outgroup status. If cognitive variables primarily mediated the
aggressive responses, participants would still differentiate between typical and atypical
targets under conditions o f provocation and trigger. Counterintuitively, however, participants
were expected to aggress more against the atypical target. The reasons for this prediction
follow.
One way that individuals gain information about themselves is by comparing
themselves to others (Festinger, I9S4). Individuals may seek to enhance their self-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
provocation, the atypical triggering target may receive the greatest amount of aggression.
Second, the atypical non-triggering target may receive the least amount o f aggression because
there is no justification tor aggression. Third, the typical target may receive less aggression than
the atypical target under conditions of trigger because he is not a threat to the participant.
Fourth, the typical target may receive more aggression under conditions o f no trigger because
there is no justification to denigrate the atypical target.
Pre-measure
Research on stereotypical traits o f African Americans are out of date (Katz & Braly,
1933), include traits which may no longer be considered politically correct, such as lazy
(Devine, 1989), or include assessments o f African Americans in general (Devine, 1989). An
assessment of the perceived stereotypical and counterstereotypical traits of African American
college students by other college students was needed. One hundred and fifty undergraduate
students at the University of Southern California completed a scale that assessed whether they
felt a given trait was typical or atypical o f African American college males. The results of this
pre-measure were used to compose the target descriptions.
Predictions
Aggression
I. Under conditions o f no provocation. (See Figure 1)
Hypothesis I. There will be a main effect of typicality. Typical targets will be
aggressed against more than atypical targets.
Hypothesis 2. There will be an interaction between typicality and trigger. There will be
a greater disparity o f aggression between the non-triggering, typical target and the triggering,
typical target than between the non-triggering, atypical target and the triggering, atypical target
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
Figure 1: Predicted mean effects of aggression as
a function of typicality and trigger under
conditions of no provocation
10 -]
9 " -----------
8 -
e
□ Typical
B Atypical
No T rigger T rigger
2. Under conditions of provocation. (See Figure 2)
Hypothesis 1. There will be a main effect of trigger. Triggering targets will be
aggressed against more than non-triggering targets.
Hypothesis 2. There will be an interaction between typicality and trigger. A non
triggering typical target will be aggressed against more than a non-triggering atypical target.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
Figure 2: Predicted mean effects of aggression as
a function of typicality and trigger under
conditions of provocation
□ Typical
B Atypical
No Trigger Trigger
Stereotype Generalization
I. Under conditions of no provocation (See Figure 3)
Hypothesis 1. There will be an interaction between typicality and trigger.
(a) There will be greater endorsement of negative stereotype generalizations after
exposure to a non-triggering atypical target than after exposure to a non
triggering typical target.
(b) There will be greater endorsement o f negative stereotype generalizations after
exposure to a triggering typical target than after exposure to a triggering atypical
target.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
Figure 4: Predicted mean effects of stereotype
generalization as a function of typicality and
trigger under conditions of provocation
□ Typical
M Atypical
No Trigger Trigger
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 2-Method
Participants and Design
Ninety-nine non-African American University of Southern California undergraduate
students participated in a 2 (provocation: present or absent) x 2 (trigger: present or absent) x
2 (target: typical, atypical) between subjects factorial design in exchange for class credit. In
all conditions the target was an African American male. Participants were 49% Caucasian,
28% Asian, 12% Hispanic, and 9% other. Participants were randomly assigned to each cell
filling one full replication prior to beginning another replication. Participants with
incomplete data were discarded (N=7).
Procedure
Participants were told that they would be participating in two studies on problem
solving ability. They were taken into a room, which contained a tape recorder, headphones, a
timer, a television, and a VCR and asked to complete a short demographic measure that
requested information concerning their gender, class standing, race, desire to perform well, and
their predicted number of correct answers. They were then informed that the purpose of first
study was to investigate the effect of distraction on problem solving ability and led to expect to
receive fifteen anagrams, along with an answer sheet. They were told that they had five minutes
to solve as many anagrams as possible. The experimenter then turned on music and exited the
room. In the provocation condition, participants listened to mildly irritating music (Stravinsky’s
Rites of Spring) and were assigned difficult anagrams (e.g. tophhapogr = photograph). In the no
provocation condition, participants listened to mildly positive music (Brandenburg Concerto #5)
and solved easy anagrams (e.g. meit = time) After five minutes, the experimenter re-entered the
room, turned off the music, and retrieved the answer sheet The experimenter once again left
the room, ostensibly to grade the answers. Prior to leaving, the experimenter gave the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
participant a handout that indicated how a sample o f engineering students had scored on the
same anagram task during the previous semester. It showed that the engineering students did
quite well in both conditions. The experimenter re-entered the room after approximately 4
minutes.
Provocation manipulation. Participants in the provocation condition were insulted on
their problem solving ability and effort. They were informed that their scores were below
average, as compared to the sample of engineering students. In the no provocation condition,
they were told that they received an average score on their anagram task, by comparison with
the sample of engineering students. Participants were then told that the purpose of the second
study was to investigate how having to attend to both audio and visual cues effected problem
solving ability. They were informed that, by means of a videotape presentation, an assistant
would ask them 15 trivia game questions. They were instructed to correctly answer as many of
them as possible while attending to the trivia questions spoken aloud by the assistant and the
answers presented on cards by the assistant while writing down their responses. The
experimenter turned on the videotape and exited the room. The videotape began immediately.
On it, a female research assistant read trivia questions aloud while holding up multiple-choice
responses for each question. After the conclusion of the tape, the experimenter re-entered the
room and retrieved the participant’s answer sheet. The experimenter again provided a summary
sheet indicating the average score obtained by engineering students from the previous semester
and then exited the room to grade the participant’s responses. On re-entering the room after
about 5 minutes she told the participant that the research assistant on the videotape has applied
for a paid position for the following semester and the supervising faculty member had requested
feedback on her performance. Participants were informed that their feedback was completely
anonymous and confidential. They were provided with a form that requested ratings on the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
research assistant’s motivation, intelligence, work ethic, and professionalism. In addition,
participants were asked to recommend the assistant for the position. The experimenter left the
room while the participant completed these forms. To create a perception of response
anonymity, participants were explicitly asked not to put any identifying information on their
response forms, to seal them in the envelope provided, and to place their sealed envelope in a
large box containing many other such envelopes. After completing their assessment of the
research assistant, manipulation check measures regarding the anagram task and their mood
were administered. Finally, they were thanked for their participation and debriefed.
Trigger manipulation. In the trigger condition, the male triggering target read the
questions too quickly, and generally presented the questions poorly. He did not
mispronounce words because such behavior may have been confounded with typicality. He
did not mix up questions and answers, as some suspicion was found for this behavior in a
pilot study. Further, participants were told they scored poorly in comparison with
engineering students’ scores on the same task. In the no trigger condition, the research
assistant read the questions slowly, made no pronunciation errors, and correctly matched the
questions with the appropriate multiple-choice answers. Participants were told that their
score was about the same as that o f the engineering students’ average.
Typicality Manipulation. Participants were asked to read an application form after
viewing the research assistant on the videotape. The application provided information
regarding the research assistant’s hobbies, extracurricular activities, and reasons for seeking
the paid research position. For the typical applicant, the hobbies included playing basketball
and working out. The extra curricular activities included membership in the Black student
association on campus and volunteer work for the office of affirmative action. The reason
given for seeking employment was to supplement financial aid. For the atypical applicant,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
are lazy) were included but recast (e.g., African Americans who live in poor areas are not
motivated). Less blatant generalizations (e.g., African Americans feel that whites owe them for
past injustices) were also included. To decrease suspicion, items concerning African Americans
were imbedded in a larger array of items concerning numerous racial groups including Asians
and Hispanics.
Summary. Here is a short timeline of manipulations and measures. Participants were
first given the demographic form and the introduction. Second, they experienced one of the
provocation manipulations. Third, they were shown the video and given the trigger
manipulation. At the conclusion of the trigger, they rated the research assistant. Once the rating
was removed, they completed the anagram appraisal, the trivia appraisal, and the mood
assessment. Next, participants completed the diversity questionnaire. Finally, because the
questions were expected to instill the most suspicion, participants assessed target typicality.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
Chapter 3 - Results
The results are organized into three major sections, manipulation checks, aggression
results, and generalization results. First, I assessed the effectiveness o f the manipulations.
Second, I examined the results for aggression. I tested the overall interactions of provocation,
trigger, and typicality for all subjects. For further understanding of how provocation moderated
aggression, the three-way interaction was broken down into two 2x2 trigger by typicality
ANOVAs. Third, race effects were examined and found to be significant. After dividing
subjects into two types o f racial groups, Caucasians (N=49) vs. Asians and Hispanics (N=40), I
assessed cognitive load, state self-esteem, and justification as contributors to aggression. The
interactions of expectancy violation and cognition with aggression were also assessed. Next, the
overall generalization results are presented followed by race effects. Generalization results are
also presented separately for each racial group, Caucasian and Asian/Hispanic. The interaction
of expectancy violation and perceptions of typicality with generalization are presented. Finally,
the expressions of aggression and stereotype generalization were compared within subjects.
Manipulation Checks
Manipulation checks were conducted to assess the effectiveness of the provocation,
trigger, and typicality manipulations.
Provocation
Five items assessed the effectiveness of the provocation manipulation: the anagram task
was fun (reverse coded); the anagrams were difficult; I was frustrated by doing the anagrams;
there was enough time to solve the anagrams (reverse coded); and I was pleased with my
performance (reverse coded). Each item was assessed on an 8 point Likert-type scale ranging
from intensely disagree (1) to intensely agree (8). Internal consistency for these variables was
acceptable (Chronbach’s alpha = .79). Higher numbers indicated greater perceptions of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
provocation. Participants in the provocation condition indicated stronger agreement with this
composite (M = 19.05, SD = 3.15) than those in the no provocation condition (M = 12.85, SD =
3.79), t (87) = 8.39, g<. 001.
Trigger
Three items assessed the effectiveness of the trigger manipulation: there was enough
time to answer the questions; the trivia task was fun; and I was pleased with my performance.
Each item was assessed on a 6 point Likert-type scale ranging from strongly disagree (1) to
strongly agree (6). Internal consistency for these variables was acceptable (Chronbach’s alpha
= .64). Each item was reverse coded so that higher numbers indicate stronger perceptions of
trigger. Participants in the trigger condition indicated stronger agreement with this composite
(M= 10.63, SD = 2.75) than those in the no trigger condition (M = 7.68, SD= 2.18), L(9l) =
5.67, g < .001. A second manipulation check assessed the behavior of the research assistant.
Four items assessed the research assistant’s behavior: the research assistant read the questions
slowly (reverse coded); the research assistant spoke clearly (reverse coded); the research
assistant administrated the questions efficiently (reverse coded); the research assistant read the
questions correctly (reverse coded). Each item was assessed on a 6 point Likert-type scale
ranging from strongly disagree (1) to strongly agree (6). Internal consistency for these variables
was acceptable (Chronbach’s alpha = .92). Higher numbers indicated greater disapproval of the
research assistant. Participants in the trigger condition indicated stronger agreement with this
composite (M = 15.32, SD = 4.67) than those in the no trigger condition (M =6.71, SD = 2.47),
t(9 6 )= llJ 9 ,g < .0 0 l.
Typicality
Three items assessed the effectiveness of the typicality manipulation. First, to what
degree is the applicant a typical male African American College student? This item was
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
assessed on a 6-point Likert-type scale ranging from very atypical (1) to very typical (6).
Second, how similar is the applicant to other African American college students? This item was
assessed on a 6-point Likert-type scale ranging from not at all similar (1) to very similar (6).
Third, please think of the written application. How rare are the applicant’s behaviors (hobbies
and extracurricular activities) amongst other African American College students? This item
was assessed on a 6-point Likert-type scale ranging from very rare (1) to very widespread (6).
Internal consistency for these variables was moderate (Chronbach’s alpha = .55). Higher
numbers indicate the target was perceived as similar to other male African American college
students. Participants in the typical condition indicated stronger agreement with this composite
(M= 12.97. SD = 1.91) than those in the atypical condition (M = 11.80, SD = 2.62), t (65) =
2.25, p < .027.
Mood
Four items assessed negative mood. Participants were asked to indicate how strongly
they were feeling angry, frustrated, resentful, and irritated. Each item was assessed on a 6-point
Likert-type scale ranging from not at all (1) to extremely (6). Higher numbers indicate stronger
feelings o f negative mood. Internal consistency for these variables was acceptable (Chronbach’s
alpha = .77). Participants in the provocation condition indicated more negative mood (M = 9.16,
SD = 3.24) than those in the no provocation condition (M_= 6.25, SD = 2.78), t (96) = 4.75, g<
.001.
Aggression
The first analyses assessed aggression expressed by all participants, collapsing across
race. In order to assess aggression towards the target, the single item, which measured intent to
harm the target, via recommendation, was analyzed. This item, I recommend this research
assistant to be a paid intern next semester, was assessed on a 6 point Likert-type scale ranging
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
from strongly disagree (I) to strongly agree (6). The item was reverse coded for analysis so that
higher numbers reflected greater intent to harm. For the overall 2x2x2 design, more aggression
was expressed toward triggering targets than non-triggering targets, F (1,96) = 17.60, p<. 001.
More aggression was expressed toward typical targets than atypical targets, F (1,96) = 5.28, p<.
024. There was a three way interaction for trigger, typicality, and provocation, F (1,96) = 6.15,
p<. 015. For frirther understanding of how provocation moderated aggression, the three-way
interaction was broken down into two 2x2 trigger by typicality ANOVAs.
No provocation
I predicted that more aggression would be expressed toward typical targets than
atypical targets. I also predicted an interaction between typicality and trigger. These predictions
were supported. Under conditions of no provocation, more aggression was expressed toward
typical targets than atypical targets, F (1,48) = 6.206, g< .017. More aggression was also
expressed toward triggering targets than non-triggering targets, F (1,48) = 6.639, p<. 013. As
shown in Figure 5, these two main effects must be qualified by a trigger by typicality
interaction, F (1,48) = 4.533, j>< .039. Discussion of the interaction follows.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
Figure 5: Mean effects of aggression as a
function of typicality and trigger under
conditions of no provocation
3.5 -| -----------
3 -
□ Typical
1 Atypical
No T rigger T rigger
Simple effects for no provocation revealed that when participants were triggered, they
expressed more aggression toward the typical target than the atypical target (M = 3.5, SD = 138
for the yes trigger/typical target and M = 2.27, SD = .46 for the yes trigger/atypical target) F
(1312) = 7.83, p< .01. When participants were not triggered they expressed equal amounts of
aggression toward the typical target and the atypical target (M = 2.25, SD = .45 for the no
trigger/typical target and M = 2.15, SD = .98 for the no trigger/atypical target) F (1,24) = .683,
ns. For the atypical target, participants expressed no difference in aggression regardless of
trigger (M=2.27. SD= .46 for the yes trigger/atypical target and M=2.15, SD= .98 for the no
trigger/atypical target) F (1,23) = .134, ns. However, for the typical target, participants
expressed more aggression toward the typical triggering target than the typical non-triggering
target (M=3.5. SD= .138 for the yes trigger/typical target and M=2.25, SD= .45 for the no
trigger/typical target) F (1.23) = 8.87, £ > < .007.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
Thus, for all non-provoked participants, non-triggering typical and non-triggering
atypical targets were aggressed against equally. Triggering typical targets were differentiated
from triggering atypical targets, the former receiving more aggression.
Provocation
I predicted that under conditions of provocation, triggering targets would receive more
aggression than non-triggering targets. This prediction was supported. Under conditions of
provocation triggering targets also received the greatest aggression, F_(1.48) =12.18, p<.
001 .However, there was no difference in the expression o f aggression toward typical and
atypical targets, £ (1,48) = .15, g>.10.1 also predicted an interaction between typicality and
trigger. This prediction was partially supported. As shown in Figure 6, a one way analysis
revealed a significant interaction in the predicted direction, £ (1,48) = 2.80, g< .05.
Contrary to the pattern seen under conditions of no provocation, simple effects for
provocation revealed that when participants were triggered, they displayed equal amounts of
aggression regardless of typicality (M=3.0. SD= .852 for the yes trigger/typical target and
M=3.38, SD=1.38 for the yes trigger/atypical target) £ (1,24) = .683, ns. When participants
were not triggered, they displayed marginally more aggression toward the typical target than the
atypical target (M=2.45, SD= .93 for the no trigger/typical target and M=l .83, SD= .83 for the
no trigger/atypical target) ££1,22) = 2.837, p< .10. For the atypical target, participants
expressed more aggression toward the atypical triggering target than the atypical non-triggering
target (M=3.38. SD=. 1.38 for the yes trigger/atypical target and M=1.83. SD= .83 than for the
no trigger/atypical target) F_( 1 J24) = 11.23, g< .003. For the typical target, participants
expressed no difference in aggression between the typical triggering target and the typical non
triggering target (M=3.0, SD= .852 and M=2.45, SD= .93 respectively) ££1,22) = 2.144, ns.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
Figure 6: Mean effects of aggression as a
function of typicality and trigger under
conditions of provocation
□ Typical
B Atypical
No Trigger Trigger
Thus, for all provoked participants, under conditions of no trigger, typical targets were
aggressed against more than atypical targets. Under conditions of trigger, both typical and
atypical targets were aggressed against equally.
Race and Aggression
To support the assumption that any racial group would behave similarly toward an
outgroup African American; the interaction between race and aggression was examined. Race
was coded so that aggression expressed by Caucasians was compared to aggression expressed
by other minority groups, Asians and Hispanics. (Asians and Hispanics were combined as one
group because there was not enough subjects per cell to separate them) Race interacted with
aggression when entered into the three-way ANOVA. Asians and Hispanics were more
aggressive overall than Caucasians, F (1,71) = 16.04, p< .001. There was a tendency for more
aggression to be expressed after the experience of provocation, F (1,71) = 3.20, g< .08. More
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
Caucasians and Aggression
For further understanding of how provocation moderated aggression for Caucasians,
the three-way interaction was broken down into two 2x2 trigger by typicality ANOVAs. Under
conditions of no provocation there were no significant effects. Thus, unlike the overall analyses,
Caucasians displayed little or no aggression when not provoked.
Under conditions of provocation, triggering targets received more aggression than non
triggering targets, F_( 1,19) = 4.84, g< 04. Unlike the overall analyses, atypical targets received
marginally more aggression than typical targets, F_( 1,19) =2.15, g = .10. There was no trigger
by typicality interaction, F_(l, 19) = .538, p> .10. As shown in Figure 7, provoked Caucasians
aggressed marginally more toward the triggering atypical target, than against all other targets, t
(21) = 1.93, p< .066. Thus, under conditions of provocation and trigger, Caucasians displayed
more aggression toward atypical targets. These results are consistent with the overall analysis,
however, Caucasians’ expressed aggression toward the typical target under conditions of
provocation and trigger is lower than in overall analysis. What could be the cause of this effect?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
Figure 7: Mean effects of aggression as a
function of typicality, trigger, and provocation
for Caucasians
0.4
e
<
-0.6
-0.2
- 0.8
0.2
-I
0
□ Typical
1 Atypical
No Prov/No No Prov/No Prov/Trig
Trig Prov/Trig Trig
Intervening Processes
Between subjects analyses were conducted to investigate additional processes that were
predicted to contribute to aggression. In this first section, I continue with the analysis of the
results for Caucasians. 1 predicted that if atypical targets received greater aggression than typical
targets as a result of provocation and trigger, several factors would be at work. First, that there
would not be an effect of cognitive load as a result o f negative mood. Participants would still be
able to differentiate the typical and atypical targets. Second, that there would be lower state self
esteem as a result of an interaction with an exceptional outgroup member. Third, that
participants may justify their actions by disregarding the atypical target’s positive attributes and
use the triggering video as a basis for evaluation. Each of these predictions was supported.
Cognitive Load. I examined the effect of cognitive load on subjects’ ability to
differentiate targets. Participants were asked to recall the number of attributes for each job
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
candidate. These attributes were the hobbies, extracurricular activities, and the reasons given by
the research assistant for seeking employment. Overall, provocation did not effect the
participant’s ability to recall target attributes, F (1,34) = .110, p> .10. Nor did trigger impact
recall, £(1,34) = 1.12, p> .10. Participants did, however, recall a greater number of attributes of
the atypical target, F (1,34) = 4.747, p< .036.
State Self-esteem. I also examined state how state self-esteem was effected by type of
target. State self-esteem was assessed in a two-item composite. Participants were asked to
indicate how strongly they felt confident (reverse coded) or embarrassed at the conclusion of the
experiment. Each item was assessed on a 6-point Likert-type scale ranging from not at all (1) to
extremely (6). Higher numbers indicate lower state self-esteem, internal consistency for these
variables was acceptable (Chronbach's alpha = .86).
Under conditions of provocation, participants experienced lower state self-esteem after
atypical triggering targets and also after typical non-triggering targets. Overall, lower state self
esteem was felt after provocation than no provocation, F (1,41) = 16.53, g<. 001. The trigger by
typical interaction, F (1,41) = 7.90, j><. 008, and the marginal interaction of provocation, trigger,
and typical,_F (1,41) = 3.40, g<. 072 are discussed next.
Under conditions of no provocation marginally lower state self-esteem resulted from
both non-triggering targets, F_( 1,21) = 2.99, j j < . 098. Under conditions of provocation, there
was a trigger by typical interaction, F.(l, 20) = 7.95, g< .011.
As shown in Figure 8, provoked Caucasians were more likely to feel greater
embarrassment and less confidence after the experience with the non-triggering typical target
(M= -.543, SD= 1.49) than the triggering typical target (M= .87, SD= .86), t (7) = -1.67, p> .10,
although the means fell short of significance. For typical targets, state self-esteem did not
predict aggression, F (1 ,16) = .662, g> .10. beta = -206. For atypical targets, state self-esteem
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
did predict aggression. Provoked Caucasians felt greater embarrassment and less confidence
after the experience with the triggering atypical target (M= 1.10 SD= .94) than the non
triggering atypical target (M= -.007, SD= 1.01), t (13) = 232, p< .037. Thus, provoked
Caucasians felt lower state self-esteem after the triggering atypical target.
Figure 8: Low state self-esteem as function of
typicality, trigger, and provocation for
Caucasians
□ Typical
B Atypical
No Prov/No No Prov/No Prov/Trig
Trig Prov/Trig Trig
Justification. I examined participant's reliance on the application more than the video as
an assessment of justification of aggression. Justification was assessed by subtracting the
reliance on the video for recommendation from the reliance on the application for
recommendation. The item, this rating was largely based on the research assistant's trial video,
was subtracted from the item, this rating was largely based on the research assistant's
application. Higher numbers indicated greater reliance on the application. Under conditions of
no provocation, there were no significant effects. However, it appears that participants relied on
the application the most for non-triggering atypical targets.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
Under conditions of provocation, provoked participants disregarded the worthy
application of the atypical target and concentrated on the triggering video to justify aggression.
Under conditions of provocation, there was a significant trigger by typicality interaction, F
(1.20) = 4.644, g< .044. As shown in Figure 9, simple effects for provocation revealed that the
triggering atypical target’s application was disregarded more than both typical targets’
applications and the non-triggering atypical target’s application, t (22) = -3.28, g< .003. Thus,
provoked Caucasians consciously disregard a high-quality application when provided
justification (trigger) by the atypical outgroup target’s behavior.
Affect and Cognition. Because both affective and cognitive subjective experiences
were predicted to impact aggression, the effect of variables that reflected arousal and variables
that reflected cognition were examined within subjects in a repeated measures analysis.
Repeated measures analyses were conducted due to the small sample size that resulted from
Figure 9: Application reliance as function of
typicality, trigger, and provocation for
Caucasians
□ Typical
1 Atypical
No Prov/No No Prov/No Prov/Trig
Trig Prov/Trig Trig
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
separating the sample into two racial subgroups. Repeated measures analyses were employed to
test the predicted intervening variables. These analyses are appropriate when using a small
sample size as they are more sensitive than between subject analyses (Keppel, 1982).
Comparisons of affect and cognition were conducted using a three-factor, mixed model,
analysis of variance. All measures were standardized for comparison purposes. The within
subjects trial factors were aggression toward the target, entered as trial 1, and surprise at the
target’s behavior (arousal) or perception of the target’s capabilities (cognition), entered as trial
2,
Due the differential treatment expressed by Caucasians toward typical and atypical
targets within subjects analyses were conducted based on typicality. Therefore, the between
subjects treatment factors were provocation and trigger. The assumptions of normality and
homogeneity of covariances were satisfied (Greenhouse-Geisser = I), unless otherwise noted.
First, the interaction of arousal and aggressive responses was tested. Although there
was a measure of mood assessed at the conclusion of the experiment, the most appropriate
measure of arousal was one taken closest to the time of aggression. This measure was both a
measure of arousal and a measure of expectancy violation. The item, “I was surprised by the
research assistant’s behavior on the video,” was assessed on a 6 point Likert-type scale ranging
from strongly disagree (1) to strongly agree (6). Higher numbers indicated more arousal and
greater expectancy violation. In support of this item as a measure of arousal, it was correlated
with the manipulation check of negative mood given at the conclusion of the study, r = 327, g <
.003.1
For typical targets, there was a significant interaction between aggression and
expectancy violation/arousal with provocation, F (1,12) = 5.94, p<. 031. As shown in Figure 10,
although expectancy violation/arousal increased under conditions of provocation (M=-.875.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
SD= .50 for no provocation and M= -003, SD= .74 for provocation), t_( 14) = 2.903, p< .014),
aggression expressed toward typical targets remained the same (M= -.33, SD= .50 for no
provocation and M=-.54, SD= .63 for provocation), t (15) = -.751, g> .10). Thus, typical targets
were aggressed against similarly regardless of arousal.
Figure 10: Interaction of expectancy violation
and aggression toward a typical target as a
function of trigger and provocation for
Caucasains
□ Expectancy
Violation
□ Aggression
No Prov/No No Prov/No Prov/Trig
Trig Prov/Trig Trig
For atypical targets, there were no significant interactions between aggression and
expectancy violation with provocation or trigger, F_(l, 20) = .494, p> .10.
In order to assess the interaction of aggression with more cognitive processes,
aggression and perceptions of the target’s capabilities were compared. Perceptions of the
target’s capabilities were measured with a two-item composite, I consider the research assistant
qualified for the internship and the research assistant would make a successful intern. Each item
was assessed on a 6 point Likert-type scale ranging from strongly disagree (1) to strongly agree
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
(6). Both items were reversed coded such that higher numbers indicated perceptions of inferior
qualifications. (Chronbach’s alpha = .83).
For typical targets, there were no interactions of aggression and perceptions with
provocation and trigger, F (1,8) = .166, p>. 10. Caucasians’ perceptions of the typical target’s
capabilities and their aggression toward the typical target were similar.
For atypical targets, as shown in Figure 11, there was an interaction between aggression
and perceptions with provocation, F (1,27) = 9.28, p<. 005, and a marginal interaction between
aggression and perceptions with provocation and trigger, F (1,27) = 2.77, p = .10. In addition,
there was a strong interaction, under constant conditions of trigger, between aggression and
perceptions with provocation, F( 1,14) = 12.08, p<. 004. When triggered, participants expressed
the same perceptions of the individual’s capabilities, regardless of provocation (M= -.14, SD=
.85 for no provocation and M= -.14, SD= .76 for provocation), t_(14) = 00, g > .10). However,
provoked participants aggressed against the target (M = 39, SD = .87) more than non-provoked
participants did (M = -.42, SD = 33) t (14) = 2.49, p< .034. (See Figure 13, bar 3 and 4 vs. bar 7
and 8). Although Caucasian’s perceived the atypical target’s capabilities similarly, they
aggressed against him more when provoked.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
Figure 11: Interaction of perceptions and
aggression toward an atypical target as a
function of trigger and provocation for
Caucasians
0 .6 -i
□ Perception
B Aggression
No Prov/No No Prov/No Prov/Trig
Trig Prov/Trig Trig
I predicted that if provoked participants aggressed more toward triggering atypical
targets, cognition may have influenced aggression more than arousal under the specific
conditions of the experiment. Regression analyses supported cognition as a predictor of
aggression over and above arousal in the current experiment. The following analyses fail to
support cognition as a mediator of the manipulation of trigger and aggression, however. I begin
this section discussing the role of arousal and then the role of cognition as predictors of
aggression toward typical targets followed by a similar discussion for atypical targets.
For typical targets, arousal did not predict aggression. F (1.141 = .009, g>. 10, (Beta = -
.025). No further analyses were conducted on arousal due to the failure of arousal to predict
aggression for typical targets.
Also, for typical targets, cognition did predict aggression. F (1.151 = 13.91, p<. 002,
(Beta = .694). Further, mediational analyses could not be conducted due to the failure of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
manipulation o f trigger to predict aggression, _F _ ( 1,15) = .058, p>.10. (Beta = .062), and the
fai lure o f the manipulation of trigger to predict cognition, F_( 1,15)= 1.048, g>. 10 (Beta = .248).
For Caucasians, subjective cognition, beyond those instilled by the manipulation o f trigger,
predicted aggression while arousal did not.
For atypical targets, both arousal, F (1,22) = 5.672, p< .026 (Beta = .453), and
cognition. F (1.29) =37.88, g< .001 (Beta = .753) predicted aggression. When the contribution
of arousal was accounted for cognition remained a strong predictor. F (1.21) = 18.156. p> .001
(AR2 = .452, Beta=.587). When the contribution of cognition was accounted for arousal no
longer predicted aggression, F_(l,21) = 2.53, g> .10 (AR2 = .059, Beta= .258).
Mediationa! analyses could not be conducted for atypical targets because the
manipulation of trigger only moderately predicted aggression, F_( 1,29) = 3.56, p<.07. (Beta =
3 3 1), and it failed to predict cognition, £.(1,29) = .610, p>.10. (Beta = .144). Thus, the
manipulation of trigger did not predict aggression toward the atypical target either. For
Caucasians, subjective cognition of the atypical target, beyond those instilled by the
manipulation of trigger, also predicted aggression while arousal did not. Also for Caucasians,
cognition predicted aggression more than arousal for both typical and atypical targets. It did not
mediate, however, the manipulation of trigger and the expression of aggression as expected.
Asians. Hispanics. and Aggression
For further understanding of how provocation moderated aggression for Asians and
Hispanics, two 2x2 trigger by typicality ANOVAs were conducted. Contrary to Caucasians,
effects were found under conditions of no provocation. Under conditions of no provocation
more aggression was expressed toward triggering targets, F_( 1,17) = 6.44, g<. 021. However,
this main effect appears to be driven by aggression toward the triggering typical target.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
Under conditions of provocation, more aggression was also expressed toward triggering
targets, F_( 1,14) = 17.052, g<. 001. This main effect however, appears to be driven by the
aggression expressed toward the atypical triggering target. There was no difference in the
expression of aggression toward typical and atypical targets, F_(l,14) = .288, p> .10. This result
appears to be driven by the levels of aggression expressed toward the typical non-triggering
target compared to the levels of aggression expressed toward triggering atypical targets. This
interpretation is supported by a trigger by typical interaction, F_(l, 14) = 8.68, p< .011. As seen
in the figure, under conditions of provocation, contrary to Caucasians, Asians and Hispanics
aggress toward the non-triggering typical target. Asians and Hispanics also appear to aggress
more against non-triggering atypical targets than Caucasians. One result is consistent with
Caucasians elevated aggression toward triggering atypical targets after provocation. Provoked
Asians and Hispanics aggressed more toward the triggering atypical target than against all other
targets, t( 16) = 2.91, p< .01.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
Figure 12: Mean effects of aggression as a function
of typicality, trigger, and provocation for Asians
and Hispanics
□ Typical
I Atypical
No Prov/No No Prov/Trig Prov/No Trig Prov/Trig
Trig
Intervening Processes
The effect of cognitive load, state self-esteem, and justification were also tested for
Asians and Hispanics. As mentioned in the previous section, I predicted that if atypical targets
received greater aggression than typical targets as a result of provocation and trigger, several
factors would be at work. First, that there would not be an effect o f cognitive load as a result of
negative mood. Participants would still be able to differentiate the typical and atypical targets.
Second, that there would be lower state self-esteem as a result of an interaction with an
exceptional outgroup member. Third, that participants may justify their actions by disregarding
the atypical target’s positive attributes and use the triggering video as a basis forevaluation.
Each of these predictions was supported.
Cognitive Load. As with the Caucasians, the test for cognitive load revealed no effects.
Under conditions o f no provocation, participants recalled the attributes of both typical and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
Justification. Justification for aggression was also tested for Asians and Hispanics. The
reliance on the video was again subtracted from the reliance on the application. Under
conditions of no provocation there were no significant effects. As shown in Figure 14, under
conditions of provocation, there was a significant trigger by typicality interaction, F (1,15) =
4.72, p< .046. Simple effects for provocation revealed, however, that when participants were
triggered by the atypical target, they did not disregard the application more than for any other
target, t (15) = -1.02, g < . 10, although the means were in that direction. Similar to Caucasians,
simple effects did reveal that they disregarded the application more for the triggering atypical
target than the non-triggering atypical target, t (4) = -7.62., jj_< .002.
Figure 14: Application reliance as function of
typicality, trigger, and provocation for Asians
and Hispanics
□ Typical
1 Atypical
No Prov/No No Prov/No Prov/Trig
Trig Prov/Trig Trig
Affect and Cognition. The same analyses were conducted to assess expectancy
violation/arousal and cognition for Asians and Hispanics. In order to compare aggression and
arousal within subjects, a repeated measured analysis was conducted. Both measures were
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
standardized for comparison purposes. The assumptions of normality and homogeneity of
covariances were satisfied (Greenhouse-Geisser = I). The comparison of aggression and arousal
were analyzed in a three-factor, mixed model, analysis of variance. The within subjects trial
factor was aggression toward the individual, entered as trial 1, and expectancy violation/ arousal
or perception of the targets capabilities, entered as trial 2. For consistency, the effect o f arousal
and cognition on aggressive responses was also tested based on typicality. The between
subjects treatment factors were provocation and trigger.
The same measure of expectancy violation /arousal was assessed for Asian and
Hispanic participants. For typical targets, there was no significant interaction within subjects
between aggression, arousal, and provocation, F (1,14) = 2.634 p>. 10. For atypical targets,
there were no significant interactions between aggression and arousal with provocation or
trigger, £.(1,8) = . 136, g> .10. Arousal closely paralleled the expression of aggression.
In order to compare aggression with more cognitive processes for Asians and
Hispanics, aggression and perceptions of the target’s capabilities were compared. As shown in
Figure 15, for typical targets, there was a significant interaction between aggression and
perceptions with provocation and trigger, £.(1,22) = 7.13, g< .013. Under conditions of trigger,
target’s capabilities are assed as equally poor. However, under conditions of provocation he is
aggressed against less.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
Figure 15: Interaction of perceptions and
aggression toward a typical target as a function
of trigger and provocation for Asians and
Hispanics
1.5 i
-0.5 -
□ Perception
1 Aggression
No Prov/No No Prov/No Prov/Trig
Trig Prov/Trig Trig
For atypical targets, there was a marginal main effect of trial. Overall, participants
marginally aggressed against the target with more intensity than they than they felt he was
incapable, F (1,8) = 3.23, g = .10. Contrary to Caucasians, when triggered, Asians and Hispanic
participants expressed marginally more negative perceptions of the individual’s capabilities, as
a result o f provocation /M= -.50, SD= .97 for no provocation and M= -.86, SD= .91 for
provocation), t_(8) = 2.92, g < .061. The increase o f negative perceptions were consistent with
an increase o f aggression as a result of provocation. Provoked participants aggressed against the
triggering atypical target (M = 2.28, SD = .00) more than non-provoked participants did (M =
.008, SD = .54), t (8) = 5.42, p< .012. (See Figure 16, bar 3 and 4 vs. bar 7 and 8). Both negative
perceptions and aggression increased toward atypical targets as a result of provocation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
Figure 16: Interaction of perceptions and
aggression toward an atypical target as a
function of trigger and provocation Asians and
Hispanics
2 -
1.5 -
1
a
I ■
0.5 -
si
0 ■
-0.5 -
-1 -
□ Perception
B Aggression
No Prov/No No Prov/No Prov/Trig
Trig Prov/Trig Trig
Mediational analyses were conducted to compare arousal and cognition as mediators of
the manipulation of trigger on expressions of aggression for Asians and Hispanics. Contrary to
the Caucasian sample, regression analyses results were able to assess arousal and cognition as
possible mediators. As with the Caucasian sample, I begin a discussion of arousal and cognition
for the typical target and then discuss arousal and cognition for the atypical target.
For typical targets, I assessed arousal as a mediator of the effect of the manipulation of
trigger on aggression. Contrary to Caucasians, the manipulation of trigger predicted aggression
toward typical targets, F_(l,24) = 834, p<.008. (Beta = .50). The manipulation o f trigger
predicted subjective arousal, F_(l,l7) = 23.83, jj< .001, (Beta = .76). The subjective measure of
arousal also predicted aggression. F (1,24) = 8.92, g< .009, (Beta = .59). When the manipulation
o f the targets behavior was controlled for, however, subjective arousal no longer predicted
aggression expressed toward the target, F (1,15) = 1.43, p>.lO (AR2= .05). After controlling for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
arousal, the manipulation of trigger no longer predicted aggression F ( 1 ,15) = 1.13, g>. 1 0 (AR2
= .04). Thus, for Asian and Hispanic participants, arousal does not appear to mediate the
manipulation of trigger and aggression.
Mediational analyses were also conducted to assess cognition as a mediator of the
effect o f the manipulation of trigger on aggression for typical targets. The manipulation of
trigger predicted aggression toward typical targets, £.(1,24) = 8.34, p<.008. (Beta = .50). The
manipulation of trigger predicted cognition, £_( 1,25) = 9.22, p<. 006, (Beta =.52). Cognition
predicted aggression, £.(1,24) = 49.68, g<. 001, (Beta = .82). When the manipulation of the
target’s behavior (trigger) was controlled for, cognition still predicted aggression expressed
toward the target, F (1,23) = 30.22, p<.001 (AR2= .42). After controlling for cognition,
however, the manipulation of trigger was no longer predictive of aggression F (1,23) = .362,
p>.10 (AR2 = .005). Thus, for Asian and Hispanic participants, cognition and not arousal
mediated the manipulation of trigger on the aggression expressed toward the typical target.
For atypical targets, the picture is less clear. Although, the manipulation of trigger also
predicted aggression toward atypical targets, £.(1,11) = 8.41, p<.015. (Beta = .656), mediational
analyses could not be conducted for arousal because arousal did not predict aggression, £(1,10)
= 2.726, g> .10 (Beta = .463). This result was contrary to Caucasians.
Mediational analyses were also conducted to assess cognition as a mediator of the
effect o f the manipulation of trigger on aggression for atypical targets. As mentioned, the
manipulation of trigger predicted aggression toward atypical targets, £.( 1,11) = 8.41, p<.015,
(Beta = .656). The manipulation of trigger only marginally predicted cognition, £.(1,11) = 3.16,
p<.10. (Beta = 52). Cognition predicted aggression, £.(1,11) = 37.97, £K. 001, (Beta = .88).
When the manipulation of the target’s behavior was controlled for, subjective cognition still
predicted aggression expressed toward the target, F (1,10) = 27.91, p<-001 (AR2= .41). After
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
controlling for cognition, the manipulation of trigger was less predictive of aggression, F ( 1,10)
= 4.97, p<.05 (AR2 = .07), although it was still significant. Thus, for Asian and Hispanic
participants, cognition maybe at least partially mediate the manipulation of trigger on
aggression toward atypical targets. The role of arousal is unclear.
Thus, for Asians and Hispanics, arousal does not appear to mediate the manipulation of
trigger on aggression. This result is similar to the results for Caucasians where the contribution
of arousal is unclear. For Asians and Hispanics, unlike Caucasians, cognitions do mediate the
manipulation o f trigger on aggression for typical targets and partially mediate the effect of
trigger on aggression for atypical targets.
Generalization
Provocation was also predicted to moderate the effect of typicality and trigger on
stereotype generalizations. Five items assessed generalizations: African Americans feel that
Caucasians owe them for past injustices; African Americans who live in poor areas are not
motivated; Most African Americans do not speak English well; A lot of African Americans are
on welfare; A lot of African American want extra benefits because they are African American.
Each item was assessed on a 6 point Likert-type scale ranging from strongly disagree (1) to
strongly agree (6). Higher numbers indicated greater endorsement of generalization items.
Three items were based on notions o f traditional racism: African Americans who live in poor
areas are not motivated; Most African Americans do not speak English well; and A lot of
African Americans are on welfare. Two of the items were based on notions of modem racism:
African Americans feel that Caucasians owe them for past injustices; and A lot of African
American want extra benefits because they are African American.
A factor analysis was conducted to assess if traditional generalizations and modem
generalizations loaded on separate factors. Due to correlations among the factors, an Oblimin
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
rotation was used with Kaiser Normalization. A principal components analysis revealed three
factors. The first factor accounted for 44.74% of the variance (total eigenvalue^ .80). This
factor contained two items: African Americans who live in poor areas are not motivated (.839);
and Most African Americans do not speak English well (.814). These two items were combined
to form a composite measure of traditional stereotype generalizations. The second factor
accounted for 19% of the variance (total eigenvalue^. 10). This factor contained one item: A lot
of African Americans are on welfare (.975). The third factor accounted for 15.7% of the
variance (total eigenva!ue=l .72). This factor contained two items: African Americans feel that
Caucasians owe them for past injustices (.710); and A lot o f African Americans want extra
benefits because they are African American (.949). These two items were combined to form a
composite measure of modem stereotype generalizations.
Earlier I indicated that participants might be reluctant to endorse statements associated
with traditionally racist tenets. They may be more likely, instead, to endorse statements that
denote generalization yet not appear as prejudicial. This expectation was supported.
Traditional Generalizations
Overall, participants did not endorse the first traditional generalization factor (African
Americans who live in poor areas are not motivated and Most African Americans do not speak
English well). Participants in the provocation condition did not endorse these items (M = 2.69,
SD =1.11) more than those in the no provocation condition (M = 2.56, SD = .98), t (96) = .63, g
> .10.
Also consistent with expectations, for the first measure of traditional stereotype
generalizations, the overall 2x2x2 design had no effects. There was no main effect of trigger, F
(1,98) = .732, p>. 10, typicality, £.(1,98) = .640, p>. 10, or provocation. Nor was there a three-
way interaction for trigger, typicality, and provocation, F (1,98) = .121, p> .10.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
The second measure of traditional stereotype generalizations (A lot o f African
Americans are on welfare) revealed similar non-effects. Participants in the provocation
condition did not endorse this item (M = 4.24, SD = 5.96) more than those in the no provocation
condition (M = 3.18, SD = .1.10), t (96) = 1.20,p>.I0.
In addition for this item, there was no main effect of trigger, F_( 1,98) = 2.5, p>. 10,
typicality F _( U98) = .533, p>. 1 0, or provocation, F_( 1,98) = . 1.32, p>. 10. Nor was there a three
way interaction for trigger, typicality, and provocation F (1,98) = .344, p> .10.
Modem Generalizations
The measure of modem stereotype generalization did reveal significant between
subjects effects. Although, there was no main effect of trigger F_( 1,98) = 1.98, p>. 10 or
typicality F.CU98) = .014, p>. 10. The effect of provocation approached significance, F_( 1,98) =
2.57, p>. 10, suggesting that participants were more likely to endorse modem generalizations
after provocation. Finally, there was a three way interaction of trigger, typicality, and
provocation F (1,98) = 10.719, p< .002. As with aggression, for further understanding of how
provocation moderated modem stereotype generalizations, the three-way interaction was broken
down into two 2x2 trigger by typicality ANOVAs.
No provocation. I predicted that there would be greater endorsement of negative
stereotype generalizations after exposure to a non-triggering atypical target than after
exposure to a non-triggering typical target. This prediction was supported. I also predicted
that there would be greater endorsement o f negative generalizations after exposure to a
triggering typical target than after exposure to a triggering atypical target. This prediction
was not supported.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
61
Under conditions o f no provocation, there was no main effect o f trigger, F_(l,48) =
.324, p > . 10, or typicality, F_( 1,48) = . 138, p>. 1 0. There was a trigger by typicality
interaction, F (1,48) = 8.79, p< .005. (Figure 17)
Figure 17: Modern Generalizations as a function
of typicality and trigger under conditions of no
provocation
1 0 -I
9 -
□ Typical
1 Atypical
No Trigger Trigger
Simple effects for no provocation revealed that when participants were triggered, they
endorsed modem generalization items more after exposure to the atypical target than after
exposure to the typical target (M= 7.72, SD= 2.57 for the yes trigger/atypical target and
M=5.75, SD= 2.60 for the yes trigger/typical target), F (1,22) =5.77, p< .027. When participants
were not triggered they endorsed generalization items more after exposure to the typical target
than after exposure to the atypical target (M=7.54, SD= 1.75 for the no trigger/typical target and
M=5.61, SD=2.32. for the no trigger/atypical target), F (1,24) =5.09, p< .034. For the atypical
target, participants endorsed generalization items more after exposure to the triggering atypical
target than after exposure to the non-triggering atypical target (M= 7.72, SD= 2.57. for the yes
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
62
trigger/atypical target and M= 5.61, SD=2.32. for the no trigger/atypical target), F (1,23)
=4.45,p< .046. However, for the typical target, participants endorsed generalization items more
after exposure to the non-triggering typical target than after exposure to the triggering typical
target (M=7.54, SD= 1.75 for the no trigger/typical target and M=5.75, SD= 2.26 for the yes
trigger/typical target), F (1,23) = 4.46, p< .047.
Thus, contrary to prediction, under conditions of no provocation, generalizations were
endorsed more after exposure to triggering atypical targets and non-triggering typical targets.
Provocation. Under conditions of provocation there was no main effect for trigger, F
(1,48) = .125, p> .10 or typicality, F.(l,48) = .29, p>. 10. There was a trigger by typicality
interaction, F (1,48) = 4.45, p< .041. (Figure 18)
Figure 18: Modern Generalizations as a function
of typicality and trigger under conditions of
provocation
□ Typical
B Atypical
No Trigger Trigger
Contrary to the pattern seen for no provocation, simple effects for provocation revealed
that when participants were triggered, they endorsed generalization items marginally more after
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
Caucasians and Generalizations
For traditional generalizations there were no between subjects effects. For modem
generalizations, under conditions of no provocation, there was no trigger by typicality
interaction, F (1,25) = 2.10, p>. 10. However, the pattern of results was the same as all races
combined. Under conditions of provocation there were no significant results, although the
pattern was also the same as the all races combined, F (1,25) = .314, p>.!0.
Intervening Processes and Generalization
Additional analyses were conducted to assess the impact of one individual outgroup
member’s behavior on generalization toward the group. Analyses were conducted to assess if
predictors o f aggression such as low state self-esteem and expectancy violation/arousal also
predicted generalizations and to assess additional unique predictors of generalizations.
Traditional Generalizations. An exploratory analysis of low state self-esteem and
aggression demonstrated that after exposure to typical targets, low state-self esteem did not
predict traditional generalizations for factor 1, (F (1,7) = 1.07, p>.10, beta = .365 non-provoked,
and F (1,7) = .164, p>. 10, beta = -.151, provoked) or factor 2 (F (1,7) = 1.59, p>.10, beta =
.431, non-provoked, and F (1,7) = 5.042, p< 06, beta = -.647, provoked).
However, for non-provoked Caucasians, exposed to atypical targets, low state self
esteem predicted traditional generalizations for factor 1, (F ( 1,14) = 12.58, p<. 003, beta = .688),
and factor 2 (F (1,14) = 5.62, p<. 029, beta — .545). For provoked Caucasians, there was no
relationship between low state self-esteem and traditional generalizations factor 1 (F (1,13) =
.545, p>. I0,beta= -.201 or factor 2 (F (1,13) = .260, p>. 10, beta = -.140, respectively).
In summary low state selfesteem predicted traditional generalization factors only for
non-provoked participants.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
Modem Generalization. For typical targets, low state-self-esteem did not predict
modem generalization. For atypical targets, Low state self-esteem did not predict modem
generalizations for non-provoked participants, F = .935, p >.10 beta = .250. However, provoked
Caucasians’ low state self-esteem predicted the suppression of modem generalizations,_F (1,13)
= 5.24, p<. 039, beta = -.536.
Thus, exposure to typical targets had no impact on either traditional or modem
generalizations. Low state self-esteem, resulting from exposure to atypical targets, predicted the
expression of traditional generalizations for non-provoked subjects and the suppression of
modem generalizations for provoked subjects.
Affect and Traditional Generalization. Analyses were also conducted within subjects to
assess if expectancy violation or surprise at the research assistant’s behavior interacted with
either traditional generalization factors. Both the surprise item and the traditional generalization
factors were standardized. The assumptions of normality and homogeneity of covariances were
satisfied (Greenhouse-Geisser = 1). The within subjects factors were expectancy violation,
entered as trial 1, and traditional generalization factor 1, entered as trial 2. The between subjects
treatment factors were provocation and trigger. For consistency, within subjects analyses were
conducted based on typicality.
For traditional generalizations factor 1, typical targets, there was a significant
interaction within subjects between traditional generalization and arousal with provocation,_F
(1,12) = 7.34, p<. 019. Under conditions of no provocation, participants reported lower levels of
arousal and generalized more. Under conditions o f provocation, generalization closely
paralleled arousal. For atypical targets, there was no interaction between traditional
generalization factor I and arousal with trigger and provocation, F (1,20) = 308, p> .10.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
66
For traditional generalization factor 2, analyses were conducted in the same fashion.
Results for traditional generalization factor 2 closely paralleled results found for factor 1. For
typical targets, there was a significant interaction within subjects between traditional
generalization and arousal with provocation,_F (1,12) = 6.58, p<. 025. Under conditions of no
provocation, participants reported lower levels of arousal and endorsed generalizations to a
greater extent. Under conditions of provocation, generalization closely paralleled arousal. For
atypical targets, there was no interaction between traditional generalization factor 2 and arousal
with trigger and provocation, F(l,20) = .438, p> .10. Arousal and traditional generalizations
were expressed similarly.
Affect and Modem Generalizations. Analyses were conducted to assess if expectancy
violation/arousal interacted with modem generalization items. Within subjects factors were
expectancy violation, entered as trial 1. and modem generalization items, entered as trial 2. The
between subjects treatment factors were provocation and trigger.
For typical targets, there were no significant effects. For atypical targets, within
subjects, there was a marginal three-way interaction of generalization, arousal, trigger, and
provocation, F (1,18) = 4.31, p< .052. (Figure 19) For provoked participants, greater arousal
marginally predicted lower generalizations £=3.865 p< .081 (Beta=-.548).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
68
Analyses were conducted to assess if perceptions of typicality interacted with the
endorsement o f modem generalization items. The standardized statement, ‘To what degree is
the applicant a typical male African American College student?’ measured the perception of
typicality. This item was assessed on a 6-point Likert-type scale ranging from very atypical (1)
to very typical (6). Higher numbers indicated greater perceptions of typicality. Both perceptions
o f typicality and modem generalization composites were standardized. The assumptions of
normality and homogeneity of covariances were satisfied (Greenhouse-Geisser = 1). The
comparison of perceptions of typicality and modem generalization was analyzed in a three-
factor, mixed model, analysis of variance. The within subjects factor was perceptions of
typicality, entered as trial 1, and modem generalization items, entered as trial 2. The between
subjects treatment factors were provocation and trigger.
For typical targets there were no significant effects between perceptions of typicality
and modem generalizations items with provocation and trigger, F (1,14) = 1.09, p> .10.
For atypical targets, there was a marginal interaction between perceptions of typicality
and modem generalizations items with provocation and trigger, F (1,25) = 3.69, p<. 066. The
more typical the target was perceived; the less participant’s generalized. The less typical the
target, the more likely participants were to generalize. A regression analysis was conducted to
support this assumption, greater perceptions of typicality were predictive of lower
generalizations, across provocation, F =4.292 p< .048 (Beta = - .37) (Figure 20).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
69
Figure 20: Interaction of perceptions of
typicality and modern generalizations after
exposure to an atypical target as a function of
trigger and provocation for Caucasians
u - 0 .2 ■
si *°*4 -
-0.6 -
-0.8 -
□ Typicality
■ Generalizations
No No Prov/No Prov/Trig
Prov/No Prov/Trig Trig
Trig
Perceptions o f typicality also appeared to interact with provocation, under conditions of
provocation, perceptions o f typicality were greatest at yes trigger, under conditions of no
provocation, perceptions o f typicality were greatest for non-triggering targets.
Traditional and Modem Generalizations Although there were no between subjects
effects of traditional generalizations, an exploratory repeated measures analysis was conducted
to ascertain if there were within subjects effects in the endorsement of traditional and modem
generalizations. For typical targets, there were no significant effects. For atypical targets, there
was a interaction between traditional and modem generalization with provocation and trigger, F
(1,27) = 6.031, p<.021. (Figure 21). While traditional generalizations increased as a result of
provocation and trigger, modem generalizations decreased.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
Figure 21: Interaction of traditional and modern
generalizations after exposure to an atypical
target as a function of trigger and provocation
for Caucasians
0.5 •
□ Traditional
■ Modern
No Prov/No No Prov/ No Prov/Trig
Trig Prov/Trig Trig
Aggression and Modem Generalization. To assess if there was an interaction between
the treatment of an individual outgroup member and perceptions of the group as a whole, a
repeated measured analysis was conducted. Within subject's factors were aggression, entered as
trial 1 , and modem generalization, entered as trial 2. The between subjects treatment factors
were provocation and trigger.
For typical targets, there were no significant results. For atypical targets, there was a
three-way interaction between aggression and generalization with trigger and provocation, F
(1,25) = 14.031, p< .001 (Figure 22). Non-provoked participants expressed greater aggression
and lower generalization toward non-triggering atypical targets. Provoked participants
expressed greater aggression and lower generalizations toward triggering atypical targets.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
Figure 22: Interaction of aggression and modern
generalizations after exposure to an atypical
target as a function of trigger and provocation
for Caucasians
1 -
0.5 -
0
£
|
cS -0.5 - I
-1 1
-1.5
No No Prov/No Prov/Trig
Prov/No Prov/Trig Trig
Trig
□ Aggression
■ Generalizations
Asians and Hispanics and Generalizations.
Traditional Generalizations. There were no significant between subjects effects for
either traditional generalization factor.
Modem generalizations Under conditions o f no provocation, there was no trigger by
typicality interaction. Under conditions of provocation the pattern which approached
significance was the same as the all races combined, F (1,21) = 2.10, p> .10.
Intervening Processes. As with the Caucasians, analyses were conducted to assess the
affect o f low state self-esteem, expectancy violation and perceptions o f typicality on traditional
and modem generalizations.
After exposure to typical targets, low state-self esteem did not predict either traditional
generalization factor regardless of provocation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
For non-provoked Asian and Hispanics, exposed to atypical targets, low state self
esteem did not predict traditional generalizations for factor I (F (1,5) = .606, p>. 10, beta =
.329) or factor 2 (F (1,4) = 1.84, p>. 10, beta = .189). For provoked Asian and Hispanics there
was no relationship between low state self-esteem and traditional generalizations, factorl (F
(1,5) = 1.670, p>. 10, beta = -.543) or factor 2, (F (1,5) = 2.63, p>. 10, beta = -.630). However,
the lack of results appears to be caused by the small cell sizes more than the absence o f effect.
For typical targets, low state-self-esteem did not predict modem generalizations. For
atypical targets, low state self-esteem did not predict modem generalizations for non-provoked
participants, F (1,5) = 2.70, p >.10, beta = -.635. Nor did provoked Asian and Hispanic’s low
state self-esteem predicted the suppression of modem generalizations..!: (1,4) = 2.20. p>.10.
beta = -.554, although the means were in that direction.
Although low state self-esteem did not predict the expression of traditional
generalizations for non-provoked subjects and the suppression of modem generalizations for
provoked subjects, the lack of significance may be due to small sample size rather than a lack of
affect.
Affect and Traditional Generalization. Analyses were conducted to assess if
expectancy violation or arousal interacted with traditional generalization items. Within subjects
factors were expectancy violation, entered as trial 1, and traditional generalization items,
entered as trial 2. The between subjects treatment factors were provocation and trigger.
For typical targets, there was a marginal interaction between traditional generalization
and arousal with trigger. F ( 1,15) = 4.257, p<. 056. For atypical targets, there was a marginal
interaction between traditional generalization, arousal, and trigger, F (1,8) =3.85, p> .08.
Affect and Modem Generalization Analyses were conducted within subjects to assess if
expectancy violation or surprise at the research assistant’s behavior interacted with modem
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
generalization items. The within subjects factors were expectancy violation, entered as trial I,
and modem generalization items, entered as trial 2. The between subjects treatment factors were
provocation and trigger.
For typical targets, there was a significant interaction within subjects between modem
generalization and arousal with trigger,_F (1.15) = 9.33, p<. 008. Thus, regardless of the
intensity of felt arousal, modem generalizations are low (Figure 23).
Figure 23: Interaction of expectancy violation
and generalizations after exposure to a typical
target as a function of trigger and provocation
for Asians and Hispanics
□ Expectancy
Violation
■ Generalizations
No Prov/No No Prov/No Prov/Trig
Trig Prov/Trig Trig
For atypical targets, there was a marginal three-way interaction of generalization,
arousal, trigger, and provocation F (1,8) = 3.41, p = .10. The pattern of the interaction was
consistent with that seen for Caucasians (Figure 24).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
74
Figure 24: Interaction of expectancy violation
and generalizations after exposure to an atypical
target as a function of trigger and provocation
for Asians and Hispanics
1.5 - i
N -0.5 -
-I ■
-1.5 •
g 0.5 -
1 -
□ Expectancy
Violation
■ Generalizations
No No Prov/No Prov/Trig
Prov/No Prov/Trig Trig
Trig
Perceptions o f Typicality. Analyses were conducted to assess if perceptions of
typicality interacted with the endorsement of modem generalization items for Asians and
Hispanics. The within subjects factor was perceptions of typicality, entered as trial 1, and
modem generalization items, entered as trial 2. The between subjects treatment factors were
provocation and trigger.
For typical targets there were no significant effects. For atypical targets, there was not a
interaction between perceptions of typicality and modem generalizations items with provocation
and trigger, F (1,9) = .494, p> .10.
Traditional and Modem Generalizations Although there were no between subjects
effects o f traditional generalizations, a repeated measures analysis was conducted to ascertain if
there were within subjects effects in the endorsement of traditional and modem generalizations.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75
For typical targets, there was a marginal interaction between traditional and modem
generalization with trigger. F fl.23) = 2.91, p=. 10. For atypical targets, there was no interaction
between traditional and modem generalization with provocation and trigger. F (1.9) = 2.10, p>.
10.
Aggression and Modem Generalization. To assess if there was an interaction between
treatment of the individual outgroup member and perception of the group as a whole for Asians
and Hispanics, a repeated measured analysis was conducted. The within subject’s factors were
aggression, entered as trial 1, and modem generalization, entered as trial 2. The between
subjects treatment factors were provocation and trigger.
For typical targets, there was a interaction between aggression and generalization with
trigger, F_( 1 ,22) = 5.63, p< .027. There was a marginal interaction between aggression and
generalization with trigger and provocation, £.(1,22) = 3.45, p< .077.
For atypical targets, within subjects analysis revealed a marginal interaction between
aggression and generalization with provocation F (1.9) = 5.08, p< .051. There was also a
interaction between aggression and generalization with trigger, F.( 1,9) = 4.40, g< .065. Finally,
there was a significant three-way interaction between aggression and generalization with trigger
and provocation, F_(l,9) = 6.79, p< .028. The pattern of the interaction was similar to that
exhibited by Caucasians. (Figure 25)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Z-scores
76
Figure 25: Interaction of aggression and
generalizations after exposure to an atypical
target as a funtion of trigger and provocation
for Asians and Hispanics
2 -
1.5 -
1 •
0.5
0
-0.5
-1
J L
I
□ Aggression
■ Generalizations
No No Prov/No Prov/Trig
Prov/No Prov/Trig Trig
Trig
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
77
Chapter 4 - Discussion
The results o f the dissertation study support the prediction that the experience or lack of
experience o f provocation influences how trigger and typicality moderate aggression and
stereotype generalizations toward African Americans. Thus, beyond behavior and attributes,
extraneous conditions affecting mood modify the treatment of outgroup members individually
and the perception of outgroup members as a whole. For African Americans, typicality appears
to be a strong determinate of the expression or suppression of aggression and generalization.
The discussion is divided into two major sections. The first section addresses the first
research question; How do the attributes and behaviors of African Americans influence the
expression o f direct aggression and displaced aggression? In addition, it addresses the third
research question regarding the underlying processes of cognition and affect. In this section, I
also present some possible reasons why the race of aggressor was a factor. The second section
addresses the effects of the attributes and behaviors of a single African American on
generalization following with a discussion of the influence of affect and cognition.
Aggression
Overall
Because original predictions were based on all races collapsed, results for all races are
included in the initial discussion. Subsequently, the two racial groups are divided for discussion.
No provocation. As predicted, under conditions of no provocation, participants did not
aggress against typical nor atypical non-triggering targets. There are two possible explanations
for this result. First, when individuals have a neutral experience with an outgroup member,
regardless of the strength of the outgroup member’s group association, they react positively and
do not aggress. Alternatively, and perhaps a more pessimistic explanation, when there is no
justification for aggression, individuals repress initial negative responses against outgroup
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
78
members who are highly associated with a negatively perceived group, typical targets. Group
members who are not as associated with the group, atypical targets, are viewed more positively.
In addition, typical targets were somewhat negatively valenced. they had a lower gpa than their
atypical counterparts and their desire for employment was to supplement their income as
opposed to pursuing research interests. Given that these factors may have made them less
qualified for the Job all things being equal, under conditions o f no provocation and no trigger,
the fact that both targets were treated equally under these conditions further supports the
suppression of aggression toward typical targets as an explanation. Thus, the atypical and the
typical target were not differentiated due to distinct underlying motivations. Support for the
latter explanation is further found when considering the results under conditions of trigger.
Under conditions of no provocation and trigger, the typical target and the atypical target
were differentiated. Participants aggressed more against the typical target. For typical targets,
the trigger may serve as justification for aggression. With justification, participants may no
longer be held in check by social desirability concerns (Brewer and Miller, 199S). Because the
typical target is more associated with the negatively perceived outgroup only a small amount of
justification may be needed to release aggressive impulses. Further, a small offense committed
by a typical out-group member, or in this case, a typical African American, is interpreted as
quite negative as it confirms the stereotype primed with the target's attributes and is reacted to
accordingly. Thus, it appears that it may be the absence of justification that motivates the
positive treatment of non-triggering typical outgroup members.
The same small offense committed by the atypical African American was not reacted to
as negatively. This result was as expected if the atypical target is not as associated with the
negatively perceived group. His actions may not prime additional negative associations
(Hewstone, 1998). Consequently, the same small offense is given less weight in the evaluation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79
Triggering atypical outgroup members are given the benefit o f the doubt so to speak, and treated
similarly to atypical outgroup members who do not commit an offense. Therefore, under
positive or neutral conditions, atypical outgroup members are given allowances for error above
those afforded to typical targets. This result is consistent with work done by Wilder (1984) and
Hewstone and Lord (1998) mentioned in the introduction.
Provocation. Under conditions of provocation, participants have a heightened negative
affect prior to their exposure to the outgroup target. As predicted, under conditions o f no trigger,
typical and atypical targets are differentiated. The heightened negative affect appears to reduce
the need for justification or stereotype confirmation (trigger) to aggress against the typical
target. The atypical target is still given concessions and aggressed against less. Participants may
make a conscious effort to focus on the positive attributes of the application and to discount the
negative behavior o f the atypical target. Typical targets are not granted the same courtesy.
However, the difference in the expression of aggression between the two targets is tentative.
Under conditions of provocation and trigger both targets are once again aggressed
against equally. As predicted, aggression is not suppressed it is elevated. For typical targets, the
suppression o f social desirability concerns caused by negative affect, combined with
justification and stereotype confirmation allows the participant to freely aggress. However, the
addition of trigger does not change the expression of aggression toward the typical target under
conditions of provocation. Thus, justification and stereotype confirmation appear unnecessary
for aggression to occur after the experience of an extraneous negative affect. It is the experience
of the extraneous negative affect that is more indicative o f the treatment of the typical target
than the target’s behavior. This result is consistent with Dollard et. al. (1939) and Marcus-
Newhall et. al. (1999): displaced aggression is alive and well.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
What are the implications o f the current study for typical outgroup members? First, that
a small indiscretion provides sufficient motivation for outgroup actors to treat them with
aggression. Aggression was consistent across all conditions towards the typical target with the
exception of the no provocation/no trigger condition. Unless all is right with the world and the
typical African American’s behavior is above board, he will be treated with animosity.
Even more interesting is the treatment of the atypical target under conditions of
provocation and trigger. The atypical target is no longer given concessions. He is now treated
with the same animosity as the typical out-group member. There may be several explanations
for this treatment. First, cognitive capacity may be taken up by negative mood. As a result,
individuals no longer have the capability to interpret additional information regarding typicality.
The most salient information, outgroup membership, may be used to form an opinion. This
explanation was not supported by the results. Participants were still able to correctly recall the
research assistant’s attributes at the conclusion of the study regardless of provocation. Second,
participants may have the capacity to comprehend typicality information. However, under
conditions of extraneous negative affect, participant’s sensitivity to the triggering offense may
be heightened, causing increased aggression (Worchel, 1966). This explanation was partially
supported by the results. There was a strong main effect of perceptions of triggering behavior
under conditions of provocation. Finally, participants may have the capacity to comprehend the
typicality information, but heightened negative arousal causes them to focus on the perceived
negative aspects of the individual. The negative outgroup membership and negative behavior is
primarily focused on. This explanation was also supported by the results. All provoked
participants consciously ignored the triggering atypical target’s worthy application. As a result,
participants may assign atypical outgroup members who commit a minor offense the same
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81
negative perceptions as the group as a whole. After all, no matter how model an outgroup
member may be they are ultimately outgroup members.
As with the typical target, extraneous conditions of negative affect are more indicative
of treatment than the atypical target’s behavior. Under conditions o f no provocation, the small
triggering action by the atypical target is overlooked. Under conditions o f provocation it is
given much more consideration. What are the implications for atypical group members?
Although, it required both extraneous negative affect and a minor transgression, the positive or
non-negative associations enjoyed by atypical outgroup members were overridden. Atypical
outgroup members appear to enjoy the freedom of special allowances only to a certain extent,
with sufficient motivation, they too are treated with the same animosity as typical outgroup
members. Prior negative experiences or incidental negative affect is sufficient to ensure that
even the exemplary atypical group member may fall (or be pushed) off his pedestal.
These results are somewhat consistent with the interracial aggression literature
discussed in the introduction. Although conditions were favorable, i.e., there was participant
anonymity and unlikely retaliation, interracial aggression was only expressed after there was
sufficient justification for typical targets, and both sufficient justification plus prior negative
affect for atypical targets. Why do these additional variables appear necessary in the current
study? The interracial aggression literature was conducted almost thirty years ago. One would
expect, although some would argue with me, that interracial aggression is even more frowned
upon today. Thus, additional circumstances favorable to interracial aggression must be present
before said aggression is expressed. Is this a good thing? Yes, participants suppress aggression
toward outgroup members more than in the past. Is this a bad thing? Yes, the expression of
aggression against outgroup members is still alive and well under some circumstances,
suggesting that although we have made strides, we have yet to perceive each other impartially.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
82
Prior Research
In prior research, provoked participants expressed the greatest aggression toward
triggering targets and little aggression toward others. (Pederson, Gonzales, & Miller, 2000).
This research was conducted without the typicality manipulation. Aggression expressed toward
atypical targets is consistent with this prior research on triggered displaced aggression. Contrary
to expectations, aggression expressed toward typical targets is inconsistent with this prior
research. One would expect in the absence of the typicality manipulation participants perceived
the African American target as typical. If that was the case what could be the cause for the
elevated expression of aggression against typical targets in the current study? The application
may be viewed as an additional trigger due to the increased associations with a negatively
perceived outgroup. As a result, the application may activate additional negative associations
and increase negative perceptions of the typical target. The triggering behavior may then
confirm these perceptions and justify aggression.
Race and Aggression
Unexpectedly, but not surprisingly, Caucasians, Asians, and Hispanics differed in their
expression of aggression and generalization toward typical and atypical targets. Asians and
Hispanics expressed greater aggression toward typical targets under conditions of no
provocation and trigger. Racial differences in the expression of aggression can be attributed to
several causes. The dominant group, Caucasians, may be prone to more aggression as a result of
status or less aggression as a result of guilt or conscious monitoring of behavior toward African
Americans (Devine, 1989: Montieth, 1991). The later appears to be the case in the current
study. Evidence for this statement will be presented during the discussion of intervening
variables. Asians and Hispanics are both minority groups who behaved similarly toward African
Americans. It appears, due to the intensity of aggression expressed by this group over
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
83
Caucasians, that other minority group members did not feel obligated to suppress aggressive
impulses.
Prior to beginning the race subsection, I must state that due to the small sample size in
some of the cells, results broken down by race should be interpreted with caution.
Caucasians and Aggression
For the most part, Caucasians did not aggress against typical targets regardless of
provocation or trigger. Atypical African Americans receive greater aggression than typical ones
under conditions of provocation. Several suggestions were made during the introduction
regarding the underlying processes that might account for this effect.
Intervening Processes. Factors suggested in the introduction as possible contributors to
greater aggression toward atypical targets were at work. First, cognitive load, as a result of
increased negative affect, did not appear impair participants’ ability to attend to target
differences. Participants were still able to differentiate atypical and typical target attributes at the
conclusion of the experiment Caucasians were more likely to recall the attributes of the atypical
targets. This result is consistent with prior research (Srull & Wyer, 1989; Hastie, 1980). Second,
participants’ state self-esteem was more threatened by an interaction with a positively valenced,
atypical African American rather than an interaction with a negatively valenced, typical African
American. Provoked Caucasians experienced lower state self-esteem after exposure to
triggering atypical targets. Third, aggression was expressed the most only after justification
(trigger).
For Caucasians, state self-esteem as a result of typicality moderates the effect of ego-
threat on aggression. Prior research has demonstrated ego-threat predicts aggression
(Baumeister & Boden, 1998). State self-esteem results in the current study show that ego-threat
alone did not always predict aggression toward African Americans. In addition to an objective
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
85
Cognition and Affect. For typical targets, consistent with prior research comparing
cognition and affect as predictive of aggression, results were mixed. Regardless of Caucasians’
fluctuations in expectancy violation/arousal, typical targets were not aggressed against. Arousal
failed to predict aggression. There appears to be a strong motivation for Caucasians to suppress
aggressive impulses toward a typical target. What may be the cause of this effect? First,
Caucasians may be engaging in a correction stage in which negative impulses are filtered and
suppressed (Devine, 1986; Montieth, 1991). The typical target’s attributes may be quite
obvious. As a result, Caucasians may be able to correct any negative impulses that may or may
not be present. Assuming the participants in the study, young, presumably middle to upper class
college students, were well versed in socially desirable behavior, a typical target may have
triggered a well-rehearsed suppression of negative affect, or distancing. In prior research, non
prejudiced participants were shown to have this ability (Devine, 1986; Montieth, 1991). This
well rehearsed/ learned response is evident even after increased levels of arousal as a result of
ego-threat.
Further support for the idea that Caucasians are suppressing conscious or unconscious
negative impulses toward the typical target can be found in the mediational analyses or lack
thereof. For Caucasians, the manipulation of trigger was not predictive of perceptions of
(cognition) or actions toward (aggression) the typical target. This result is curious and requires
some explanation. The failure of the manipulation o f trigger to predict target recommendation
combined with the failure of the triggering video to predict perceptions of the target suggests
that participant’s evaluations were influenced by factors beyond the video. This suggestion is
consistent with the fact that the application, the typicality manipulation, was also given to
subjects as supplemental information. Thus, the application was taken into consideration in
addition to the target’s behavior, as expected. Can it be that the typical applicant’s attributes
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8 6
were more predictive of treatment than his behavior? This possibility would require further
study. It does appear however, that for Caucasians, perceptions of and treatment o f the typical
target were influenced by his typicality perhaps more so than the target’s behavior. Can I state
with certainty that it was typicality and not valence that was the influence? Unfortunately not.
As mentioned in the introduction, valence and typicality are often confounded. Further study
would have to tease apart the two influences. Regardless of the influence, typicality or valence,
the results remain intriguing.
For atypical targets, results were also mixed. Consistent with the results found for
typical targets, neither arousal nor cognition mediated aggression, again suggesting Caucasian’s
were taking the target’s attributes or typicality into consideration when making the evaluation.
Contrary to the results found for typical targets, expectancy violation/arousal predicted
aggression toward atypical targets, although not beyond cognition for most conditions.
Caucasians do not appear to have a well-rehearsed controlled reaction toward atypical targets.
They are more consistent in both their perceptions and treatment of this target, for the most part
treating and perceiving him positively. Perhaps the weak group association usually enjoyed by
atypical targets, may also weaken the link to the correction stage under stressful conditions. This
appears to be the case under conditions o f provocation and trigger.
Under conditions of provocation and trigger greater aggression was expressed toward
triggering atypical targets after provocation. In this situation, arousal may have played a greater
role in the aggression than perceptions. The lack o f a correction stage may allow participants to
be more consistent in their actions with their emotions especially after ego-threat. The
exemplary characteristics of the positively valenced atypical target (high gpa, lofty research
interests, and many extracurricular activities) may have further threatened participants egos.
Provoked Caucasians aggressed against the triggering atypical target regardless of how positive
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
87
they viewed the target’s qualifications. It appears that when the need for self-enhancement as a
result o f ego-threat wars with objectivity, negative affect and the need for self-enhancement
wins and atypical targets loose. This result is consistent with Forgas (1992) who found mood
was more predictive of treatment of atypical targets than typical ones.
In summary, Caucasians appear to suppress reactions toward typical targets under all
conditions, perhaps due to a correction or filtering stage. Atypical targets apparently do not
prime the initiation of the correction stage and as a result are victims o f their own excellence
when conditions are unfavorable to the aggressor. Because this correction stage is not enacted
toward atypical targets, Caucasians were free to react to them without reservation, once
provoked. I must mention these results cannot support either arousal or cognition as a mediator
of aggression in the current experiment. Although some results were consistent with the
expectation that cognition was the primary mediator of aggression toward typical targets, no
conclusions could be made comparing cognition or affect as mediators of trigger and aggression
due to the apparent influence of typicality on aggression.
Asians. Hispanics. and Aggression
Contrary to Caucasians, Asians and Hispanics are less inhibited in their expression of
aggression toward typical targets. This result suggests that other minority groups may not have a
well-rehearsed correction stage when faced with a typical African American. Non-provoked
Asians and Hispanics do aggress against triggering typical targets, suggesting, only a small
justification is needed to express aggression against a typical group member. Treatment o f the
atypical target was consistent with Caucasians. Under conditions of provocation, Asians and
Hispanics are more aggressive against triggering atypical targets.
Intervening Processes. Contrary to Caucasians, Asians and Hispanics do not recall the
atypical target attributes any more than the typical target attributes under either provocation
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8 8
condition. Perhaps they are not as surprised at the attributes, making them less salient. Also,
although lower state self-esteem was experienced after the atypical triggering target for
provoked participants, self-esteem was not a reliable predictor of aggression. Thus, compared to
Caucasians, Asians and Hispanics do not appear to be as motivated to strike out as a result of
perceived personal inadequacies. Consistent with Caucasians, Asians and Hispanics justified
their aggression toward atypical targets by ignoring the application.
Mediational analyses show that, for Asians and Hispanics, cognition was the primary
mediator of aggression for typical targets. Other minority groups apparently are not as
motivated to over think their evaluations o f typical targets, instead, they rely on the immediate
tangible behavior in the video for their evaluation more than his attributes. For atypical targets,
as with Caucasians the role of arousal or cognition as a mediator is unclear. The failure of the
manipulation of trigger to predict cognition, was consistent with Caucasians and suggests that
the other minority groups were taking the target’s application into consideration when making
the evaluation, perhaps searching for an explanation of behavior.
In summary, the differences in the aggression expressed toward typical targets between
Caucasians and other minority groups is apparently due to the well rehearsed correction stage
that Caucasians have toward typical African Americans illustrated by Devine (1986) and
Montieth (1991). Other minority groups were uninhibited in their responses to a typical African
American, unlike Caucasians. They apparently have no correction stage. Their treatment o f an
atypical target closely paralleled Caucasian’s treatment, providing further support that
Caucasians do not have a correction stage for atypical targets.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
89
Generalization
Overall
As predicted, the experience of prior negative affect moderated the expression of
generalizations. However, only those generalization items that corresponded to definitions of
modem racism were impacted between subjects. This may be due to several reasons. First,
participants may simply no longer believe in the largely negative sentiments of the traditional
generalizations (Gaertner & Dovidio, 1989). Second, as mentioned previously, participants may
be responding in a socially desirable manner. Traditional generalizations are strongly looked
down upon and perhaps considered racist in nature. These sentiments are blatantly negative as
well. Participants may be well trained not to endorse such items for fear of appearing racist.
Extensive social training against the expression o f prejudice should avert the effect of transitory
negative affect or unpleasant experiences with one African American (Devine, 1989).
Interesting results were found for modem stereotype generalizations between subjects.
Participants were less reluctant to endorse these items. These modem generalization items may
not be governed as much by social desirability concerns. First, these items are not as blatantly
negative as the traditional ones. Individuals may not be as well trained to reject such items. As a
result, transitory negative affect or unpleasant experiences with one African American affect
their endorsement of them. Second, individuals may feel more comfortable in endorsing beliefs
that are not viewed as blatantly racist but more politically motivated. These generalizations tap
into a perspective that African Americans want something for nothing, that they want extra
benefits and they feel that they are owed a debt from society. Both items are closely tied to
Affirmative Action, which has been a hot topic of recent political debate.
No provocation. Under conditions of no trigger, participants endorsed stereotype
generalizations to a greater extent after exposure to a non-triggering typical target than after
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
90
exposure to a non-triggering atypical target. This result is consistent with prior findings of
increased generalizations after exposure to typical targets (Wilder, 1996). Contrary to
prediction, the non-triggering or positive behavior by the typical target was insufficient to
decrease negative generalizations toward the group as compared to the atypical target. As
mentioned in the introduction, admirable behavior by a typical group member was expected to
have a positive impact on group perceptions. This was not the case in the current study. Perhaps
the non-triggering behavior was not admirable enough to obscure the negative associations
activated by the typical manipulation.
Under conditions of trigger this result is reversed, also contrary to prediction.
Participants endorsed stereotypes to a greater extent after exposure to atypical targets. The
positive attributes of the atypical target are not sufficient to override the impact of his behavior
on the group as a whole. (Although, the target himself did not suffer any adverse consequences
as a result of his triggering action.) These findings are consistent with boomerang effects found
by Kunda and Olsen (1997). What could be the possible reasons for this effect? The atypical
target’s bad behavior may prime negative associations, thus leading to greater perceptions of
group homogeneity. That is, although he looks great on paper, he behaves poorly. This poor
behavior is more associated with typical targets, thus reinforcing the perception of outgroup
homogeneity. Thus, stereotype confirming behavior on the part of an atypical outgroup member
leads to increased generalizations toward the outgroup. In other words, '‘ they really are all
alike”. This explanation was supported for both Caucasian and Asian/Hispanic samples. This
explanation is also supported through the significant increase in generalizations from non
triggering atypical targets and triggering atypical targets for the overall sample.
Participants may have a different type of boomerang response after exposure to a
highly typical target, one who is both typical and whose behavior is stereotype confirming. The
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
91
highly typical, stereotype confirming target, may cause subject to recall cases that do not
conform to the stereotype and thus reduce generalizations. In other words, ‘‘ they all can’t be that
bad.” This explanation is supported by the significant reduction in generalizations from the
typical non-triggering target to the typical triggering target. Second, participants may be
reacting with greater suspicion of a heavy-handed influence attempt. They may no longer
believe this individual is real and thus suppress any negative associations. This explanation was
not supported in debriefing sessions, where participants in this condition did not express any
more suspicions about the genuineness of this target than participants in other conditions.
Provocation. The opposite pattern is seen under conditions of provocation. When
participants are not triggered, they endorse generalization items equally after exposure to typical
and atypical targets, however, the direction of effect is in the predicted direction. The direction
of effect suggests a greater tendency to generalize after exposure to atypical targets. There are
two possible explanations for this effect. First, the typical target may prime participants to think
of stereotypes. In reaction to those thoughts, participants may consciously suppress negative
stereotype generalizations. Because individuals may be aware of their negative affect, they may
engage in a correction stage to reduce generalizations. Second, in the absence of the stereotype
prime from the atypical target and lack of a subsequent correction stage, participants may
endorse generalizations to a greater extent.
As predicted, when participants were triggered there was no difference in the
endorsement of generalization items after exposure to typical and atypical targets. However,
there is a strong tendency for participants to endorse generalizations to a greater extent after
exposure to typical targets. This result is also consistent with predictions. Exposure to the
triggering typical target was more likely to result in greater generalizations than exposure to the
non-triggering typical target. However, the direction of effect toward atypical targets was
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
92
opposite to prediction. The triggering atypical target was more likely to result in lower
generalizations than the non-triggering atypical target. Possible explanations for this result are
addressed in the following sections.
Caucasians and Generalizations
Caucasians appear again to suppress any negative expressions of generalizations
toward typical targets. However, some results were revealed with within subjects analyses.
First, the interaction of traditional generalizations and arousal masks the simple effect that
traditional generalizations remain consistent regardless of the change in arousal. Much like the
interaction of aggression and arousal for typical targets. Thus, regardless of arousal, participants
suppress generalizations.
For atypical targets, traditional and modem generalizations were expressed differently
depending on the condition. Low self-esteem and greater arousal were consistent in the
expression of traditional generalizations. This result is consistent with the expression of
aggression as more likely from individuals with low self-esteem (Bushman, et. al., 2000). Low
self-esteem and greater arousal were related to the suppression of modem generalizations for
provoked participants.
An explanation for these results for atypical targets appears to be an increase in the
perceptions of typicality, the more the target was perceived as typical, the less likely participants
were to generalize. Perhaps the perceptions of typicality activate the correction stage and
generalizations are suppressed. The less the atypical target was perceived as typical, the more
participants felt free to generalize. Once again supporting the notion that atypical targets do not
activate a correction stage.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
93
Asians/Hispanics and Generalizations
Contrary to Caucasians, low state self-esteem did not predict traditional or modem
generalizations. However, as mentioned, the lack of effect appears to be due to the small sample
size.
Regardless of arousal, Asians and Hispanics did not generalize after exposure to typical
targets. This provided some support to a correction stage for typical targets. For atypical targets,
like Caucasians, as expectancy violations increased, generalizations declined. Although,
contrary to Caucasians this does not appear to be related to perceptions of typicality. Perhaps
for Asians and Hispanics atypical targets are subtyped regardless of condition.
Caveats
Several questions were raised when interpreting the result of the study. First, the
measure of arousal taken close to the expression o f aggression was a troublesome one. Because
it measured expectancy violation, a cognition, and surprise, arousal, it is difficult to ascertain
which was the driving force behind this variable’s results. Also, the measure was taken after the
aggression response and could be simply justification on the part of participants for their
behavior. For future studies a less contaminated measure of arousal is needed at the time of
aggression. Second, provocation was both a manipulation of ego-threat and a negative mood
manipulation. Indeed the manipulation of provocation and prior manipulations of ego-threat are
similar (Schotte, 1992). Once again it is unclear if it is ego-threat or negative affect that
contributes to aggression and generalization toward atypical targets.
Finally, the race of the experimenter may have affected the reactions of the subjects.
The experimenter was an Asian female. Considering the results, there may be an effect o f race
of the research assistant. Perhaps insult by an outgroup member increased the salience of
outgroup targets for Caucasians and not for other racial groups. Thus, typical targets may have
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
95
group association more salient. Prior research has shown stereotype activation is a function of
group member salience (Oakes, Turner, & Hasiam, 1991). As a result, the correction stage
usually reserved for typical group members is activated and utilized.
Limitations of the Study
Possible alternative explanations not accounted for in the current paper that may also
explain the results must be discussed. First, perhaps the results found in the current study were
due target valence in addition to or beyond target typicality. In the introduction, I stated that
target valence was purposefully confounded with typicality due to prior theorizing that African
Americans are a negatively stigmatized group. However, the typical target’s traits were perhaps
not entirely negatively valenced, being athletic and musical have been described as positive
stereotype traits of African Americans in prior research (Kawakami, Dion, & Dovidio, 1998).
The only negatively valenced attribute of the application was the target’s GPA. In order to truly
disentangle valence from the typicality information, further study would need to use an
orthogonal manipulation of both valence and typicality information.
An additional limitation of the current work is its generalizability to theories of
prejudice, bias, and racism. Although the current work was based on treatment and perceptions
of outgroup targets, I reference bias, discrimination, or prejudice sparingly. Without the
appropriate ingroup comparison target, such conclusions would be premature. To support the
treatment of the outgroup targets as motivated by differential treatment further study would
include both ingroup and outgroup targets, holding all other characteristics equivalent, such as
valence or typicality. In addition, including ingroup category members would support the
assertion that it is superior outgroup performance and not just superior performance that
contributes to the interaction of low self-esteem with provocation and trigger.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
96
Theoretical Implications
What are the theoretical implications for the current study? First, atypical group
members are not always subtyped. Second, under certain conditions, atypical outgroup
members may impact generalizations to a greater extent than typical targets. These results are
consistent with the findings of Kunda and Olesen (1995). Contrary to the explanation for this
effect suggest by the authors, I propose the underlying motivations contributing to greater
generalizations is the absence of a correction stage. Because participants are not well rehearsed
in stereotype suppression when interacting with atypical targets, generalizations are expressed
unabated. One way to test this idea is to measure the psychopsysiological responses of
participants exposed to both typical and atypical targets. If my theory is correct, for Caucasian
participants, both targets will elicit similar levels of negative arousal. However, participants
exposed to typical targets should suppress or mask this arousal in paper and pencil measures
lowering generalizations as compared to atypical targets.
Conclusions
Overall, what is the impact of the current research above and beyond previous work?
After all, the variables shown to predict aggression, self-esteem (Bushman et. al., 2000), and
expectancy violations (Bettencourt, Dill, Greathouse, & Charlton, 1997) have been investigated
previously. The main contribution of this work is to demonstrate the differentiation of the
atypical target under conditions of provocation and trigger. Atypical targets are not always
subtyped or subgrouped and given concessions. Regardless of prior prejudicial inclinations,
highly capable African Americans may be treated with more animosity than deserved because
o f a transitory lapse in an aggressors self-esteem as a result of ego-threat or simply as a result
of heightened negative affect. Augmentation of intergroup aggression can result from prior
irrelevant negative experiences. Contrary to prior research that indicates treatment toward
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
97
atypical targets is by extraneous information (Kunda and Oleson; 1995). In the current studies,
treatment is justified through an active reliance on negative behavior while simultaneously
ignoring behavior that would justify positive treatment. Aggression appears to be rationalized
away by the aggressor. For both typical and atypical targets, extraneous conditions affecting
mood may be more indicative of treatment than the targets’ own behavior.
The work is important as it demonstrates how individuals who belong to powerful
ingroups, may express hostility toward individual outgroup members that is disproportionate to
that warranted by the outgroup member’s behavior. The work has practical implications at the
individual and the group level. For instance, consider an individual is not hired for a prestigious
job, a frustration. Perhaps the individual feels that they were not hired because racial quotas
were met with a less qualified African American. Subsequently, in a new position, they severely
admonish an African American who is late to work. On a more general level, they vote against
Affirmative Action legislation. Ultimately, atypical African Americans are subject to the same
fate as their typical brethren regardless of their attributes. After all, no matter how model an
outgroup member may be they are ultimately outgroup members.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
98
References
Allport, G.W. (1954) The Nature of Prejudice. Cambridge, Mass. Addison-Wesley.
Anderson, S., & Klatzky, R., (1987) Traits and social stereotypes: levels o f categorization in
person perception. Journal o f Personality and Social Psychology. 53, 235-246
Baumeister, R. F., & Boden, (1998). Aggression and the self: High self-esteem, low self-
control, and ego-threat. In R.Geen (Ed) and E.Donnerestein (Ed) Human Aggression:
Theories, research, and implications for social policy, (pp. 111-137). San Diego, Ca:
Academic Press Inc.
Baumeister, R.F., Bushman, B J., & Campbell, W.K. (2000). Self-esteem, narcissism, and
aggression: Does violence result from low self-esteem or from threatened egotism? Current
Directions in Psychological Science. 9, 26-29.
Berkowitz, L. (1982). Aversive conditions as stimuli to aggression. Advances in
Experimental Social Psychology. 15. 249-288.
Berkowitz, L.(I989). Frustration-aggression hypothesis: Examination and reformulation.
Psychological Bulletin. 106. 59-73.
Berkowitz, L., Cochran, S. T., & Embree, M. C., (1981). Physical pain and the goal of
aversively stimulated aggression. Journal o f Personality and Social Psychology. 4. 687-700.
Berkowitz, 1 ., & Hiemer, K. (1989). On the construction of the anger experience: Aversive
events and negative priming in the formation of feelings. Advances in Experimental Social
Psychology. 22. 1-37.
Berkowitz, L., & Knurek, D. A., (1969). Label-mediated hostility generalization. Journal of
Personality and Social Psychology. 13.3. 200-206.
Berkowitz, L., and Troccoli, B. T., (1990). Feelings, direction o f attention, and expressed
evaluations of others. Cognition and Emotion. 4. 305-325.
Brewer, M. B. & Miller, N. (1988). Contact and cooperation: When do they work? In P. A.
Katz & D. Taylor (Eds). Eliminating racism: Profiles in controversy. Perspectives in social
psychology, (pp. 315-326) NewYork, N.Y. Plenum Press.
Brown, R. J., Tajfel, H.,& Turner, J.C. (1981).Minimal group situations and intergroup
discrimination; Comments on the paper by Aschenbrenner and Schaefer. Eurpoean Journal
o f Social Psychology. 10. 399-414.
Bushman, B.J., & Baumeister, R.F. (1998). Threatened egotism, narcissm, self-esteem, and
direct and displaced aggression: Does self-love or self-hate lead to violence? . Journal of
Personality and Social Psychology. 75. 219-229.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
99
Carslon, M., Marcus-Newhall, A., & Miller. N., (1989). Evidence for a general construct of
Aggression. Personality and Social Psychology Bulletin. 15. 377-389.
Carlson, M.& Miller, N., (1988). Bad experiences and aggression. Sociology and Social
Research. 72. 155-158.
Carver, C. S., (1975). Physical aggression as a function of objective self-awareness and
attitudes toward punishment. Journal of Experimental Social Psychology. 11. 510-519.
Crosby, F., Bromley, S., & Saxe, L., (1980). Recent unobtrusive studies of black and white
discrimination and prejudice: A literature review. Psychological Bulletin. 87. 546-563.
Crocker, J. Blaine, B., & Luhtanen. (1993). Prejudice, intergroup behavior and self-esteem:
Enhancement and protection motives. In M. A. Hogg & D.Abrams (Eds) Group motivation:
Social psychological perspectives (pp. 52-67). London, England. Harvester Wheatsheaf.
Cowen, E. L., Landes. J., & Schaet, D. E., (1959). The effects o f mild frustration on the
expression of prejudiced attitudes. Journal o f Abnormal and Social Psychology. 58. 33-38.
Dengerink, H. A., 7 Myers, J. D., (1977). The effects o f failure and depression on
subsequent aggression. Journal of Personality and Social Psychology. 35.88-96.
Devine, P. G., (1989). Stereotypes and Prejudice; Their automatic and controlled
components. Journal of Personality and Social Psychology. 56. 5-18.
Devine, P., & Baker, S.M., (1991) Measurement of racial stereotype subtyping, Personality
and Social Psychology Bulletin. 17, 44-50.
Devine, P. G., Montieth, M J., Zuwerink, J.R., & Elliot, A.J. (1991). Prejudice with and
without compunction. Journal o f Personality and Social Psychology. 60. 817-830.
Dollard, J., Doob, L. W., Miller, N. E., Mower, O. H., & Sears, R. R. (1939). Frustration and
Aggression. New Haven, CT: Yale University Press.
Donnerstein, E., & Donnerstein, M., (1973) Variables in interracial aggression: Potential in
group censure. Journal o f Personality and Social Psychology. 27.143-150.
Donnerstein, E., & Donnerstein, M., (1977). Direct and vicarious censure in the control of
interracial aggression. Journal o f Personality. 46. 162-175.
Donnerstein, E., Donnerstein, M., Simon, M, & Ditrichs, R. (1972). Variables in interracial
aggression: Anonymity, expected retaliation, and a riot. Journal o f Personality and Social
Psychology. 22. 236-245.
Dovidio, J. F., Evans, N, & Tyler, R.B., (1986). Racial stereotypes: The contents o f their
cognitive representations. Journal of Experimental Social Psychology. 22. 22-37.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
100
Dovidio, J. F. & Gaertner, S. L. (2000). Aversive racism and selection decisions: 1989 and
1999. Psychological Science. 11.315-319.
Epstein, R., (1965). Authoritarianism, displaced aggression, and social status o f the target.
Journal of Personality and Social Psychology. 2. 585-589.
Fenigstein, A., & Buss, A. H., (1974). Association and affect as determinants o f displaced
aggression. Journal of Research in Personality. 7. 306-313.
Festinger, L. (1951). Informal social communication. Psychological Review. 57. 271-282.
Festinger, L. (1954). A theory of social comparison processes. Human Relations. 7. 117-140.
Fiske, S., Neuberg, S., Beattie, A., & Milberg, S., (1987) Category based and attribute based
reactions to others: Some informational conditions of stereotyping and individuating
processes. Journal o f Experimental Social Psychology. 23, 399-427.
Fitz, D.. (1976). A renewed look at Miller's conflict theory of aggression displacement.
Journal of Personality and Social Psychology. 33. 725-732.
Forgas, J.P., (1992) Affect in social judgments and decisions: A multi-process model. In
Advances in Experimental Social Psychology. 25. 227-274.
Fridlund, A. J., (1986). Guidelines for human electromyographic research.
Psychophysiology. 23. 567-586.
Gaertner, S. L., & Dovidio, J.F. (1986). The aversive form of racism. In J.F. Dovidio & S.L.
Gaertner (Eds) Prejudice, discrimination, and racism, (pp.61-89) Orlando, FL. Academic
Press
Gaertner, S. L., & Mclaughlin, J. P., (1983). Racial stereotypes: associations and ascriptions
of positive and negative characteristics. Social Psychology Quarterly. 46.23-30.
Geen, R. G., (1968). Effects of frustration, attack, and prior training in aggressiveness upon
aggressive behavior. Journal of Personality and Social Psychology. 9. 1968.
Gilbert, D. T. & Hixon, J. G., (1991). The trouble of thinking: Activation and Application of
stereotypic beliefs. Journal of Personality and Social Psychology. 60.509-517.
Gurwitz S., & Dodge, K., (1977) Effects of confirmations and disconfirmations on
stereotype based attributions. Journal of Personality and Social Psychology. 35.495-500.
Guglielmi, R.S. (1999). Psychophysiological assessment of prejudice: Past reserch, current
status, and future directions. Personality and Social Psychology Review. 3.123-157.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
101
Hantzi, A. (1995). Changes in stereotypical perceptions o f familiar and unfamiliar groups:
The pervasiveness o f the subtyping model. British Journal o f Social Psychology. 34. 463-
477.
Harris, J., A. (1997). A further evaluation of the aggression questionnaire: Issues o f validity
and reliability. Behaviour Research and Therapy. 35.1047-1053.
Harris, M. B., (1993). How provoking! What makes men and women angry? Aggressive
Behavior. 19. 199-211.
Hass, R. G., Katz, I., Rizzo, N., Bailey, J., & Eisenstadt, D., (1991). Cross-racial appraisal as
related to attitude ambivalence and cognitive complexity. Personality and Social Psychology
Bulletin. 17. 83-92.
Hastie, R. & Kumar, PA. (1979). Person memory: Personality traits as organizing principles in
memory for behaviors. Journal of Personality and Social Psychology. 37. 25-38.
Henderson-King, E. I., & Nisbett, R. E.. (1996). Anti-black prejudice as a function of exposure
to the negative behavior of a single black person. Journal o f Personality and Social
Psychology. 71. 654-664.
Heathertone, T.F., & Polivy, J.. (1991). Development and Validation of a scale for measuring
state self-esteem. Journal of Personality and Social Psychology. 60.895-910.
Hewstone, M., Aird, P., & Johnston, L., (1992), Cognitive models of stereotype change:
perceptions o f homogeneous and heterogeneous groups, European Journal of Social
Psychology. 219-234.
Hewstone, M. Hopkins, N., & Routh, D., (1992), Cognitive models of stereotype change: 1.
Generalization and subtyping in young peoples views o f the police, European Journal o f
Social Psychology. 22, 235-249
Hewstone & Lord (1998). Changing intergroup cognitions and intergroup behavior The role
of typiclaity. In J. E. Schopler & C. E. Sedikides (Eds), Intergroup Cognitions and
Intergroup Behavior, (pp. 367-392). Mahaw, NJ: Erlbaum.
Hovland, C.I. & Sears, R. R. (1940). Minor studies of aggression: Correlation o f lynchings
with economic indices. Journal of Psychology. 9. 301-310.
Hynan, D. J., & Grush, J. E., (1986). Effects of impulsivity, depression, provocation, and
time on aggressive behavior. Journal of Research in Personality. 20.158-171.
Jackson, L. A., (1988). Cognition and affect in evaluations o f stereotyped group members.
The Journal o f Social Psychology. 129.659-672.
Jackson, LA .. & Sullvan, L. A. (1989). Cognition and affect in evaluations of stereotyped
members. Journal o f Social Psychology. 129. 659-672.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
102
Johnson, T. E. & Rule, B. G. (1986). Mitigating circumstance information, censure, and
aggression. Journal of Personality and Social Psvchologv.SO. 537-542.
Johnston, L., & Hewstone, M., (1992) Cognitive models of stereotype change: subtyping and
the perceived typicality o f disconfirming group members. Journal of Experimental
Psychology. 28. 386.
Jussim, L., Coleman, L.M., & Lerch, L. (1987). The nature of stereotypes: A comparison
and integration o f three theories. Journal o f Personality and Social Psychology. 52. 536-546.
Katz, D., & Braly,. K., (1933). Racial stereotypes in one hundred college students. Journal of
Abnormal and Social Psychology. 28, 280-290.
Katz, I., Hass, G. R., & Wackenhut, J., (1986). Racial ambivalence, value duality, and
behavior. In Prejudice. Discrimination, and Racism, eds. Dovidio J. F., & Gaertenr,
Academic Press, chapt 2.
Kawakami, K., Dion, K.L., & Dovidio, J. F. (1999). Implicit stereotyping and prejudice and
the primed Stroop task. Swiss Journal of Psychology. 58. 241 -250.
Koneci, V. J., & Doob. A. N., (1972). Catharsis through displacement of aggression..
Journal of Personality and Social Psychology.23. 379-387.
Kulik, J. A., & Brown, R., (1979). Frustration, attribution of blame, and aggression. Journal
of Experimental Social Psychology. 15. 183-194.
Kunda, Z., & Oleson, K. C., (1997) When exceptions prove the rule: how extremity of
deviance determines the impact of deviant examples of stereotypes. Journal of Personality
and Social Psychology. 72. 965-979.
Lambert, A. J., 7 Wyer, R.S., (1990) Stereotypes and social judgments: The effects of
typicality and Group Heterogeneity. Journal of Personality and Social Psychology. 59.676-
691.
Linville, P, & Fischer, G., (1986) Stereotyping and perceived distributions o f social
characteristics: an application to ingroup-outgroup perception. In J.F. Dovidio & S. Gaertner
(Eds), Prejudice. Discrimination, and Racism, (pp. 165-195) Orlando, FI: Academic Press.
Marcus-Newhall, A., Pederson, W.C., Carlson, M., & Miller, N. (2000). Displaced
aggression in alive and well: A meta-analytic review. Journal o f Personality and Social
Psychology. 78.670-689.
McConahav, J. B., & Hough, J.C. (1976). Symbolic racism. Journal of Social Issues. 32.23-
45.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
103
Melberg, V. & Tedeschi, J. T. (1989). Displaced aggression: Frustration or impression
management? European Journal of Social Psychology. 19. 139-145.
Miller, N. E., (1948). Theory and experiment relating psychoanalytic displacement to
stimulus response generalization. Journal o f Abnormal and Social Psychology. 43. 155-178.
Miller. N. & Brewer, M.B. (1986) Categorization effects on ingroup and outgroup
perception. In J.F. Dovidio & S.L. Gaertner (Eds) Prejudice, discrimination, and racism.
(pp.209-230) Orlando, FL. Academic Press.
Montieth, M J. (1993). Self-regulation of prejudiced responses: Implications for progress in
prejudice-reduction efforts. Journal o f Personality and Social Psychology. 65.469-485
Oakes, P.J., Turner. J. C., & Haslam, S. A. (1991). Perceiving group members: The role of
fit in the salience o f social categorizations. British Journal of Social Psychology. 30. 125-
144.
Pederson, W. C., Gonzales, C. & Miller, N. (2000). The moderating effect o f trivial
triggering provocation on displaced aggression. Journal o f Personality and Social
Psychology. 78.913-927.
Rogers, R. W„ & Prentice-Dunn, S., (1981). Deindividuation and anger mediated interracial
aggression: Unmasking regressive racism. Journal of Personality and Social Psychology. 41.
63-73.
Rothbart M., & John, O. P., (1985). Social categorization and behavioral episodes: A
cognitive analysis o f the effects of intergroup contact. Journal o f Social Issues. 41. 81-104.
Rothbart, M., (1996). Category-exemplar dynamics and stereotype change. International
Journal o f Intercultural Relations. 20. 305-321.
Rothbart, M., & Lewis, S. (1988). Inferring attributes form exemplar attributes: Geometric
shapes and social categories. Journal of Personality and Social Psychology. 55. 861-872.
Schotte, D. E. (1992). On the special status of ego threats: Comments on Heatherton,
Herman, and Polivy (1991).. Journal of Personality and Social Psychology. 62. 798-800.
Smith, C. A., & Ellsworth, P. C., (1985). Patterns of cognitive appraisal in emotion. Journal
of Personality and Social Psychology. 48. 813-838.
Smith, E. R., (1993). Social identity and social emotions: Toward new conceptualizations of
prejudice. Affect Cognition, and Stereotyping: Interactive Processes in Group Perception.
Chapt. 13.
Srull. T. K., & Wyer, J. C. (1989). Person memory and judgement. Psychological Review.
96.58-83.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
104
Tajfel, H. (1974). Social identity and intergroup behavior. Social Science Information. 13.
65-93.
Tassinary, L. G., Cacioppo, J. T., & Geen, T. R., (1989). A psychometric study of surface
electrode placements for facial electromyographic recording: I. The brow and cheek muscle
regions. Psvchophvsiologv. 26.1-16.
Taylor, S., A categorization approach to stereotyping, In D, Hamilton (Ed) Cognitive
processes in Stereotyping and Intenzroup Behavior, (pp 83-113) Hillsdale, N.J. Erlbaum.
Turner, J.C., Brown, R.J., & Tajfel, H. (1979). Social comparison and group interest and
ingroup favoritism. European Journal of Social Psychology. 45, 961-977.
Vanman, E., Paul,B., Ito, T., & Miller, N. (1997). The modem face of prejudice. Journal of
Personality and Social Psychology. 73.941-959.
Weber, R., & Crocker, J., (1983), Cognitive processes in the revision of stereotypic beliefs,
Journal of Personality and Social Psychology. 45, 961-977.
Werth, J. K., & Lord, C. G., (1992). Previous conceptions o f the typical group member and
the contact hypothesis. Basic and Applied Social Psychology. 13. 351-369.
Wilder, D. A. (1978). Reduction of Intergroup discrimination through individuation of the
out-group. Journal of Personality and Social Psychology. 36. 1361-1374.
Wilder, D. A. (1984). Intergroup contact: The typical member and the exception to the rule.
Journal of Experimental Social Psychology 20.177-194.
Wilder, D. A., Simon, A. F., & Shapiro, P. (1989). Effects o f anxiety of impression
formation in a group context: An anxiety-assimilation hypothesis. Journal of Experimental
Social Psychology 25.481 -499.
Wilder, D. A. & Shapiro, P. (1989). Role of competition induced anxiety in limiting the
beneficial impact of positive behavior by an out-group member. Journal o f Personality and
Social Psychology. 56. 60-69.
Wilder, D. A., & Simon, A. F. (1996).Incidental and integral affect as triggers of
stereotyping. In R.M. Sorrentino & Higgins, E.T. (Eds) Handbook of motivation and
cognition, (pp. 397-4l9).New York, N.Y., Guilford Press.
Wilder, D. A., Simon, A. F., & Faith, M., (1996). Enhancing the impact of
counterstereotypic information: Dispositional attributions for deviance. Journal of
Personality and Social Psychology. 71.276-287.
Wilson, L., & Rogers, R. W., (1975). The fire this time: Effect o f race of target, insult, and
potential retaliation on Black aggression. Journal o f Personality and Social Psychology.
32.857-864.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
105
Wood, J.V., & Taylor, K.L. (1991). Serving self-relevant goals through social comparison. In J.
Sulls & T. A. Wills (Eds), Social comparison. Contemporary theory and research, (pp. 23-
49) Hillsdale, NJ: Lawrence Erlbaum Associates.
Worchel, P., (1966). Displacement and the summation of frustration. Journal of
Experimental Research in Personality. 1.256-261.
Zillman, D., & Bryant, J., (1974). Effect of residual excitation on the emotional response to
provocation and delayed aggressive behavior. Journal of Personality and Social Psychology.
30.782-791.
Zillman, D., & Cantor, j. R., (1976). Effect o f information about mitigating
circumstances on emotional responses to provocation and retaliatory behavior.
Journal o f Experimental Social Psychology. 12. 38-55.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Displaced aggression as a function of target power

PDF

Emotion regulation as a mediator between family-of-origin aggression and marital aggression

PDF

Alcohol's impact on triggered displaced aggression

PDF

An integrative analysis of racism and quality of life: A comparison of multidimensional moderators in an ethnically diverse sample

PDF

A cluster analysis of the differences between expert and novice counselors based on real time training

PDF

Associations and mechanisms among attention deficit hyperactivity disorder symptoms, cognitive functioning, and drinking habits

PDF

Affection and conflict in family relationships

PDF

An evaluation of a brief HIV /AIDS prevention intervention using normative feedback to promote risk reduction among sexually active college students

PDF

Collective efficacy, anxiety, creativity/innovation and work performance at the team level

PDF

Effects of parenting style and ethnic identity on European American and Asian Indian adolescents' academic competence and self esteem

PDF

A context for timing, conditioning and modification of cerebellar function

PDF

Integration of science and practice: A collective case study of scientist -practitioner programs in counseling psychology

PDF

Cognitive predictors of intention -behavior consistency in safer sex practices

PDF

Childhood videotaped neuromotor and social precursors of schizophrenia: A prospective investigation

PDF

Asymmetries in the bidirectional associative strengths between events in cue competition for causes and effects

PDF

Ideological threat, social identity, and consensus estimation

PDF

Development and validation of the Cooper Quality of Imagery Scale: A measure of vividness of sporting mental imagery

PDF

Individual motivation loss in group settings: An exploratory study of the social -loafing phenomenon

PDF

Hedonic aspects of conditioned taste aversion in rats and humans

PDF

Family functioning, intergenerational conflict, and psychological symptomology of Asian American college students

Asset Metadata

Creator Gonzales, Candace Letitia (author)

Core Title Atypicality: Benefit and bane. Provocation, trigger, typicality, and the expression of aggression and generalization

School Graduate School

Degree Doctor of Philosophy

Degree Program Psychology

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

Tag OAI-PMH Harvest,psychology, social

Language English

Contributor Digitized by ProQuest (provenance)

Advisor [illegible] (committee chair), [illegible] (committee member), Brekke, John (committee member), Davison, Gerald C. (committee member), Read, Stephen (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-218938

Unique identifier UC11334869

Identifier 3073782.pdf (filename),usctheses-c16-218938 (legacy record id)

Legacy Identifier 3073782.pdf

Dmrecord 218938

Document Type Dissertation

Rights Gonzales, Candace Letitia

Type texts

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

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA