ARTICLES
What Have We Learned from Columbine:
The Impact of the Self-System on Suicidal
and Violent Ideation Among Adolescents
Susan Harter
Sabina M. Low
Nancy R. Whitesell
ABSTRACT. This study first examined a model of the predictors and
mediators of both suicidal ideation and violent ideation, particularly
since both types of ideation were highly correlated (r = .55) among our
young adolescent participants. The variables in the model were those
Susan Harter is Professor and Developmental Area Head, Department of Psychology,
University of Denver, 2155 S. Race St., Denver, CO 80208 (E-mail: sharter@frontier.
psy.du.edu).
Sabina M. Low is a PhD Candidate, Department of Psychology, University of Denver,
2155 S. Race St., Denver, CO 80208.
Nancy R. Whitesell is a Research Associate, Department of Psychology, University of
Denver, 2155 S. Race St., Denver, CO 80208.
This research was supported by a grant from NIH to the first author and by a National
Research Service Award from NIH to the second author.
Address correspondence to: Susan Harter, Department of Psychology, University
of Denver, 2155 S. Race Street, Denver, CO 80208.
Journal of School Violence, Vol. 2(3) 2003
http://www.haworthpress.com/store/product.asp?sku=J202
 2003 by The Haworth Press, Inc. All rights reserved.
10.1300/J202v02n03_02
3
4
JOURNAL OF SCHOOL VIOLENCE
identified in media accounts of the histories of the 10 high-profile school
shooting cases, and that have also been included in our earlier model of
the predictors of suicidal ideation, which we expanded to include violent
ideation. Perceptions of competence or adequacy and social support
from peers and parents predicted two mediators in the model, an adjustment/depression composite (self-esteem, hope to hopelessness, and affect, along a continuum from cheerful to depressed) and anger-induced
physical aggression. These variables, in turn, predicted both suicidal and
violent ideation. Secondly, we examined adolescents’ reactions to vignettes that described harassing events, at the hands of peers and teachers simulating the type of audience-observed ridicule experienced by the
school shooters. We were particularly interested in the emotional reaction of humiliation, an effect that has received surprisingly little attention
in the emotion literature. Humiliation was related to other emotions such
as anger and depression, consistent with the co-occurrence of internalizing and externalizing symptoms demonstrated in the overall model. Next
we compared those who reported that they would respond violently to
the events in vignettes compared to those reporting nonviolent reactions.
Examining the variables in the general model, those who reported violent ideation reported markedly lower perceptions of adequacy, lower
levels of social support, lower self-esteem, greater hopelessness, depressed affect, and anger-induced aggression. The violent ideators also
reported higher levels of both homicidal and suicidal ideation. Discussion focused on the fact that thoughts of violence toward others as well as
toward the self have similar psychosocial origins that should be considered in prevention and intervention efforts. [Article copies available for a
fee from The Haworth Document Delivery Service: 1-800-HAWORTH. E-mail
address: <[email protected]> Website: <http://www.HaworthPress.
com> © 2003 by The Haworth Press, Inc. All rights reserved.]
KEYWORDS. Suicide, violent ideation, adolescents, Columbine, self-esteem, hopelessness
Columbine has become a metaphor for the now ten high-profile school
shootings to occur in the past six years, where white, male adolescents have
fired into crowds of classmates, killing some and wounding others (see Table 1 for a list of these cases). Our own interest in this topic stems from the
fact that many of the predictors in our theoretically-driven model of self-esteem and depressive reactions, leading to suicidal ideation, could be identi-
Harter, Low, and Whitesell
5
fied in the media accounts of these boys’ histories. In this paper, we will
first describe that model. We will then give examples from the accounts of
the school shooters’ own lives. Attention will then shift to evidence on the
co-occurrence of violence against others and violence toward the self. As a
result, we have expanded our model to include not only depressive reactions and suicidal ideation but also aggressive reactions and violent
ideation. Evidence for this revised model will be presented.
A common feature in the histories of the school shooters has been that
they each had a history of being humiliated by peers, a romantic other, or a
teacher. In certain cases, the actual precipitating event was such victimization, leading to revenge. To address this issue, we crafted hypothetical
TABLE 1. List of the Ten Major School Shootings Since 1996
1. Feb. 2, 1996. Moses Lake, Washington. 14-year old BARRY LOUKAITIS fires on an Algebra Class at his middle school, killing 3 and wounding 1.
2. Feb. 19, 1997. Bethel, Alaska. 16-year old EVAN RANSEY shoots and kills one student
and the principal of his high school.
3. Oct. 1, 1997. Pearl, Mississippi. 16-year old LUKE WOODMAN stabs his mother, and
then goes to his high school parking lot killing a classmate who was dating his ex-girlfriend.
4. Dec. 1, 1997. Peducah, Kentucky. 14-year old MICHAEL CARNEAL shoots and kills
three students, wounding 5 others, firing on a prayer circle outside Heath High School.
5. March 24, 1998. Jonesboro, Arkansas. Two boys, ANDREW GOLDEN, age 11, and
MITCHELL JOHNSON, age 13, shoot and kill one teacher and four girls, wounding 10 others, during a false fire alarm at their middle school.
6. April 24, 1998. Edinboro, Pennsylvania. A 14-year old shoots a science teacher to death
in front of students at their eighth-grade graduation.
7. May 21, 1998. Springfield, Oregon. 15-year old KIP KINKEL kills both parents and at
school shoots and kills two teenagers, wounding 22 others.
8. May 29, 1998. Fayetteville, Tennessee. Three days before graduation, an 18-year-old
opens fire in his high school parking lot, killing a classmate who was dating his ex-girlfriend.
9. April 20, 1999. Littleton, Colorado. ERIC HARRIS and DYLAN KLEBOLD kill 1 teacher
and 11 students, wounding many others, after a history of being teased and rejected by
peers, and then kill themselves.
10. May 20, 1999. Conyers, Georgia. 16-year old T.J. SOLOMON shoots and wounds six
Heritage High School students, after having been spurned by a girlfriend. He threatened
to take his own life but was disarmed by a teacher.
6
JOURNAL OF SCHOOL VIOLENCE
events that simulated the types of humiliation that the school shooters experienced. Young adolescents from a normative sample were asked to put
themselves in the position of the victim and indicate what emotional and
cognitive reactions they would experience, as well as what action they
would take, along a continuum from doing nothing to planning serious
harm or violence. Finally, we have identified violent and non-violent
ideators based on the hypothetical actions they would take. These two
groups were compared with regard to potential differences in the predictors
of our general model, predictors that represent predisposing factors in their
background. Moreover, we have documented differences that represent
more behavioral causes of violence that have been hypothesized, namely,
interest in violent media (movies, TV, music), in violent video games, as
well as in weapons and bombs. Through these efforts, we have identified a
profile of characteristics that differentiate violent from nonviolent ideators.
We explore the implications for identifying those at risk for violent and suicidal ideation, including suggestions for appropriate prevention and intervention efforts.
A MODEL OF THE CAUSES, CORRELATES,
AND CONSEQUENCES OF GLOBAL SELF-ESTEEM
A major goal of our past program of research (see Harter, 1999) has
been the construction of a theoretically-driven model of the causes, correlates, and consequences of self-esteem, defined as perceptions of one’s
global worth as a person. The primary correlates have been self-reported
depressed affect and hopelessness. In the original model, we focused on
one particular outcome, suicidal ideation. In identifying potential causes of
self-esteem, we drew upon the historical formulations of James (1892) and
Cooley (1902). For James, self-esteem was a product of the relationship
between one’s successes and one’s aspirations for success. Thus, success in
domains deemed important would lead to high self-esteem. Conversely, if
one falls short of one’s ideals, low self-esteem would ensue. In contrast to
James, symbolic interactionists such as Cooley (1902) and Mead (1934)
considered the self to be a social construction, crafted through linguistic
exchanges with significant others (e.g., parents and peers). In Cooley’s
construct of the “looking glass self” and Mead’s concept of the “generalized other,” one looked to the opinions that such others held about the self,
gradually incorporating or internalizing these judgments into a sense of
one’s self-esteem.
Harter, Low, and Whitesell
7
Considerable evidence from participants ages eight and older (see
Harter, 1999; Harter, Marold, & Whitesell, 1992; Harter & Marold, 1993)
provides support for a model in which competence or inadequacy in domains judged important, where the potential domains are perceived physical appearance, peer likability, and athletic competence (which form a
higher-order factor), not only impacts self-esteem directly but is partially
mediated by peer approval. Thus, those who value these domains but express their inadequacy, will also report lower levels of peer support that, in
turn, leads to lower self-esteem. Two additional domains, perceived scholastic competence and behavioral conduct (that also form a second higherorder factor) not only directly impact self-esteem but also have an indirect
effect on self-esteem that is mediated by parental approval. Those who report weaknesses in these latter two domains will not garner the support of
parents, and both these perceived inadequacies and associated lack of parental approval will erode one’s self-esteem.
With regard to the correlates of global self-esteem, we have demonstrated (see Harter, 1999) that it bears strong empirical relationships
(r’s from .70 to .82) to two other constructs, affect (along a continuum of
cheerful to depressed) and hope (hopeful to hopeless) about one’s future
(see also Kovacs & Beck, 1977; 1978). These three constructs that have
been combined into an adjustment/depression composite, in turn, are
highly predictive of suicidal ideation, the initial outcome in our model.
There are limitations of this earlier model, however. First, it focuses on
only one outcome, suicidal ideation. Secondly, it identifies only one mediator, represented by the adjustment/depression composite. The psychological literature, as well as the media accounts of the school shooters, informs
us that feelings of inadequacy or lack of competence as well as the media
accounts of the school shooters, informs us that feelings of inadequacy or
lack of competence as well as perceived lack of peer and parental approval
can also resort in angry and aggressive responses, and in violent ideation,
namely, the intent to harm others. Antecedents in our model, specifically,
feelings of inadequacy and lack of peer and parental support, have been
found to be associated with physical aggression (Dubow & Reid, 1994;
Learny, Shreindorfer, & Haupt, 1995; Simons, Paternite, & Shore, 2001),
and with externalizing problems (Booth, Rosen-Krasnor, McKinnon, &
Rubin, 1994; Hawkins & Lishner, 1987; Patterson & Bank, 1989). Moreover, within the clinical literature, increasing attention has been paid to the
co-occurrence of profound sadness or depression and anger (Bridewell &
Chang, 1997; Capaldi & Stoolmiller, 1999; Boergers, Spirato, & Donaldson,
1998; Curran, 1987; Defferbacher & Swaim, 1999; Gispert, Davis, Marsh, &
Wheeler, 1987; Khan, 1987; Renouf & Harter, 1990; Rutter, 1987). In
8
JOURNAL OF SCHOOL VIOLENCE
addition to the fact that depression has been found to be predictive of suicidal ideation (Carlson & Cantwell, 1982; Harter, 1999; Hillbrand, 2001;
Pfeffer, Zuckerman, Plutchik, & Mizruchi, 1984; Pfeffer, Solomon,
Plutchik, Miruchi, & Weiner, 1982; Velaz & Cohen, 1988), suicidal behavior has also been linked to aggression (Bach-Y-Rita & Veno, 1985;
Inamdar, Lewis, Siomopolous, Shanok, & Lamelea, 1982; Shafi, Carrigan,
Willinghil, & Derrick, 1995). This pattern of findings is consistent with
the literature revealing that on symptom checklists, internalizing and
externalizing symptoms are highly correlated (see Achenbach & Edlebrock,
1983).
Given these relationships, we expanded our model adding anger-driven
aggression and violent ideation, retaining the adjustment/depression composite and suicidal ideation. The revised model is particularly timely, given
an analysis of the media accounts of the histories of the white, male adolescents who have been responsible for the major school shootings since
1996. Many of theses boys were described as inadequate in terms of their
social skills, physical appearance, or athletic ability, those domains most
likely to impact peer approval in our model. In fact, many were teased,
taunted, or harassed by peers whose active rejection was very distressing to
the eventual school shooters. “I killed because people like me are mistreated everyday,” said pudgy, bespectacled Luke Woodham, age 16, from
Pearl, Mississippi, who murdered two students, injuring seven others. “My
whole life I felt outcasted, alone.” In Peducah, Kentucky, 15-year-old Michael Carneal was tired of being teased and picked on by his schoolmates.
Another shooter, Mitchell Johnson, observed that “Everyone that hates me,
everyone I don’t like, is going to die.” In his internet manifesto, Eric Harris,
age 18, from Columbine High School in Littleton, Colorado described how
classmates, primarily the jocks, “ridiculed me, chose not to accept me, and
treated me like I am not worth their time.” A surviving member of the
Trenchcoat Mafia (the name given to Harris, Klebold, and friends by the
jocks at Columbine), described how he, as well as Harris and Klebold–the
second shooter–were constantly “cornered, pushed day after day, being
ridiculed or bashed against lockers.”
From the media accounts, there has also been evidence that many of the
shooters felt abandoned, rejected, or neglected by their parents. Andrew
Golden, age 11, from Jonesboro, Arkansas, who shot down one teacher and
numerous students, was raised primarily by his grandparents since his parents worked long hours. In Moses Lake, Washington, 14-year-old Barry
Loukaitis fired on his middle school Algebra class after learning that his
mother announced that she was divorcing his father. Others had parents
who made them feel inadequate, often in comparison to other siblings. Kip
Harter, Low, and Whitesell
9
Kinkel, the 15-year-old school shooter from Springfield, Oregon, felt rejected by his parents who compared him to his “perfect” older sister. Michael
Carneal, from Peducah, Kentucky, was also made to feel incompetent in
comparison to his successful sister. In other cases, the neglect was more benign, as unknowing parents did not monitor their son’s activities (e.g.,
bomb-making in the garage, long amounts of time engaged in violent computer games). As Garbarino (1999), author of Lost boys: Why our sons turn
violent, observes, such rejection and neglect leads to anger as well as depression, culminating in acts of violence towards peers, parents, or both.
The media accounts also describe how feelings of inadequacy and lack
of support led to low self-esteem, depression, hopelessness and helplessness, and in some cases suicidal ideation and suicidal behaviors. With regard to depression, several of the boys were actually on medication. (Both
Michael Carneal and Kip Kinkel had clinical diagnoses of depression, prior
to their attacks on their classmates.) Eric Harris (from Columbine) was described as depressed and was on medication. T. J. Solomon (the 15-yearold from Conyers, Georgia, who shot and wounded six classmates and then
threatened to take his own life) was described by an examining psychiatrist
as suffering from “major depression with severe psychotic features.” Two
of the boys, when apprehended, pleaded that they be shot. Of course, Harris
and Klebold successfully suicided.
With regard to helplessness, Garbarino has described how, with the exception of their violent acts, many of these boys were quite passive, reacting as victims of the system, the objects of harassment and ridicule.
Michael Carneal’s principal found journal entries that revealed a child who
felt “weak and powerless,” with an angry desire to lash out at the world.
That their anger fueled their violent acts is also evident in the reports of
many. For example, Luke Woodham, in his last will and testament recorded on cassette, claimed; “I do this to show society, push us and we will
push back.”
Although all of the variables in our expanded model could be identified
in the lives of some if not many of these boys, those involving paths to and
from peer rejection seem to be present in all cases. These boys were the
loners, the outcasts, the neglected and rejected, victims of a peer system in
which popularity and membership in the cliques with status would appear
to be increasingly critical in the contemporary school culture (Harris, 1998;
Rubin, Bukowski, & Parker, 1998). The inability to perform adequately in
the domains most highly related to peer acceptance or rejection, namely,
failure to meet punishing peer standards in the areas of physical attractiveness, athletic prowess, and social skills, seem to be particularly evident in
the cases of many of the school shooters, all of whom obviously acted vio-
10
JOURNAL OF SCHOOL VIOLENCE
lently. Thus, it seemed appropriate to add anger-induced aggression and violent ideation to the model, given that many of the predictors in the original
model would appear to lead directly to these outcomes as well as to depressive features and suicidal ideation.
Evidence for an expanded model. To examine the fit of our expanded
model, we administered our original battery of measures, plus two new
measures that tapped anger-caused physical aggression and violent
ideation in the form of homicidal thoughts about killing others. The original battery of measures consisted of the Self-Perception Profile for Children
(Harter, 1985), the Social Support Scale for Children (Harter, 1988), and
the Dimensions of Depression Scale (Harter, Nowakowski, and Marold,
1985). From the Self-Perception Profile, we included four self-concept
subscales, Physical Appearance, Peer Social Likeability, Behavioral Conduct, and Scholastic Competence. From the Social Support Scale, two
subscales were administered, Peer/Support (to which was added an additional four peer rejection items, given its relevance to school violence) and
Parent Support. From the Dimensions of Depression instrument, four
subscales were administered, the three that comprise the depression/adjustment composite, Global Self-Esteem (how much one values one’s worth as
a person), Affect (along a continuum of depressed to cheerful), and Hope
(hopeless to hopeful), as well as Suicidal Ideation. All of these instruments
and their respective subscales have been demonstrated to show excellent
psychometric properties in numerous past studies as well as the current
study. The anger-induced physical aggression scale was adapted from
items contained in the Past Feelings and Acts of Violence Scale (Plutchik
and van Pragg, 1989). Items tapped the extent to which one attempts to
“physically injure people whom they are angry with,” “get in fights and
hurt people whom they are angry with,” “get angry at people and physically attack them.” A new homicidal ideation subscale was designed for
the purpose of this study and was embedded in the Dimensions of Depression Scale. Items tapped “thoughts about killing someone else,” “time
spent thinking about killing another person,” harboring thoughts “about
ending another person’s life.” Reliabilities for these two new susbscales
were .81, and .74, respectively.
The model was tested employing a sample of 313 6th, 7th, and 8th grade
middle school students (175 males, 138 females). The school is approximately 1/3 European-American, 1/3 Hispanic, and 1/3 African-American,
from lower middle-class to lower-class families. The self-report battery
was group-administered in participants’ classrooms.
Figure 1 presents the model that was examined through path modeling
techniques employing Lisrel VIII (Joreskog & Sorbom, 1993). Preliminary
Harter, Low, and Whitesell
11
analyses found no differences in the model for males and females and thus
their results have been combined in Figure 1 which provides path coefficients for the entire sample. Consistent with the original modeling, the domains of Physical Appearance and Peer Likeability contributed to the
Depression/Adjustment Composite through a direct path. However, their
impact on this composite was also partially mediated by Peer Support/Rejection. Findings similar to the original modeling were also obtained given
a direct path from the combination of Scholastic Competence and Behavioral Conduct to the Depression and Adjustment Composite, as well as an
indirect route that was mediated by Parent Support. Thus, young adolescents who feel that they have inadequacies with regard to their appearance
and their peer likeability report both low peer approval as well as depressive reactions in the form of low self-esteem, depressed affect, and hopelessness. Those acknowledging low scholastic competence, and poor
behavioral conduct also reported both low parent support as well as depressive reactions. The path from the Depression/Adjustment composite to
Suicidal Ideation is also consistent with the original modeling. There are
also direct paths from peer and from parent support to suicidal ideation, revealing that those adolescents who lack supportive social relationships are
more likely to engage in suicidal thinking.
What is novel about the extended model are links to and from anger-induced physical aggression as well as to and from homicidal ideation, the
two new constructs added in light of our interest in thoughts of violence
among adolescents. Physical aggression was predicted by perceptions of
behavioral conduct and scholastic performance, as well as by parental approval. Thus, those evaluating their conduct and school performance negatively were more likely to report that they were physically aggressive and
that they lacked parental support. Homicidal ideation, not surprisingly, was
predicted by anger-induced physical aggression and by peer support/rejection. Thus, those adolescents who report high levels of physical aggression
toward those that provoke their anger are more likely to have thoughts of
killing others, as do those adolescents who experience rejection at the
hands of peers. Interestingly, the depression/adjustment composite was
also predictive of homicidal ideation. Adolescents who experience the
combination of low self-esteem, depressed affect, and hopelessness are
more likely to turn their thoughts not only to suicide but to killing others.
Moreover, the fact that homicidal ideation and suicidal ideation were correlated at .55, also attests to the co-occurrence of ideation directed at harming others and harming the self.
Troubled adolescents, therefore, with a history of perceived inadequacies as well as low peer and parent support, not only experience depressive
12
JOURNAL OF SCHOOL VIOLENCE
FIGURE 1. Structural Equation Model of the Predictors of Violent and Suicidal
Ideation
Violent Ideators
Nonviolent Ideators
4
3
2
1
Affect
Self-Worth
Hope
Suicidal
Ideation
Homicidal
Ideation
Physical
Aggression
Low scores represent more negative characteristics.
and aggressive reactions but become introspective about terminating both
their own lives and those of others, putting them in double jeopardy. There
is obvious value, therefore, in developing models that address the predictors of both homicidal and suicidal ideation, given the common pathways
to each of these outcomes. The findings represented in this model give credence to the challenges expressed by experienced clinicians who often, in
treating adolescents with such histories, find it difficult to predict whether
they will become violent toward others, violent toward the self, or toward
both, as we saw in the case of the Columbine shootings.
VIOLENT REACTIONS TO HARASSMENT
AND THE ROLE OF HUMILIATION
The model presented identifies psychosocial factors in the histories of
youth that may predispose them to violent or homicidal ideation. However,
what events in the day-to-day lives of adolescents may provoke violent behavior? An analysis of the media accounts of the ten high-profile school
Harter, Low, and Whitesell
13
shootings since 1996 reveals that in every case, the shooters described how
they had been ridiculed, taunted, teased, harassed or bullied by peers (because of their inadequate appearance, social or athletic behavior), spurned
by someone in whom they were romantically interested, or put down, in
front of other students, by a teacher or school administrator, all events that
led to profound humiliation. In certain cases (e.g., Harris and Klebold from
Columbine), such a history culminated in violent revenge. In Harris’ manifesto, written days before the shooting incident, he described how peers
constantly ridiculed him. As one of the members of the Trenchcoat Mafia
told reporters, “Tell people that we were harassed and sometimes it was
impossible to take; eventually someone was going to snap.” He noted that
the torment often became vicious. He described waking on school days
with a knot in his stomach, dreading to face the continual humiliation. Central to the events that the boys described was the presence of an audience
who witnessed the harassment, often laughing or joining in the mockery.
Given these observations, we sought to determine what the literature offered in terms of an understanding of humiliation. Surprisingly, in the literature on emotion, in general, and in the literature on adolescence, in
particular, there is scant attention to humiliation. In two excellent edited
volumes on contemporary emotion research (see Lewis and Haviland,
1993; Tangney and Fischer, 1995) and in a more recent volume by
Tangney and Dearing (in press), humiliation does not appear in the index
nor is it discussed at any length in the chapters. This is especially surprising
given the fact that adolescence is a psychologically fragile period of development, where youth have heightened levels of self-consciousness and a
vulnerable sense of self (see Harter, 1999), making them prone to reactions
of humiliation. Garbarino (1999), in his analysis of violence, based on interviews with incarcerated youth, is one of the few to mention the humiliation of boys who have committed violent acts. However, it is always linked
directly with shame, as if the two emotions were synonymous. Pollack
(1998), in his book Real Boys, also discusses the role of shame in a slightly
different context, namely, the active socialization of boys who are shamed
into not crying or expressing fear, in the service of hardening boys to become men that meet traditional standards of masculinity. Pollack also observed that boys are made to feel guilty about feelings of weakness,
vulnerability, fear, and despair.
It is curious that in these treatments of the emotions experienced by boys
in our culture, as well as in the more theoretical and empirical literature
(see Tangney & Fischer, 1995), attention has not been devoted to humiliation, despite the fact that it seems so prevalent in the lives of children and
adolescents. The closest emotional reaction to appear in the literature is a
14
JOURNAL OF SCHOOL VIOLENCE
shame-rage cycle (see Tangney & Dearing, in press) where the experience
of being shamed can lead to angry revenge. However, humiliation itself
has not been explored in any depth, either from a theoretical or an empirical
perspective. Moreover, our own recent research (Harter, Kiang, Whitesell, &
Anderson, 2002) has revealed that young adults, link humiliation to anger
and embarrassment, not to shame. It should be noted that there is a growing
literature on the topic of bullying (see Olweus, 1997; Ross, 2002) that can
clearly be inferred to be a cause of humiliation given threats to the self.
However, bullying constitutes an event that is inflicted on someone, it does
not capture the emotional reaction of the victim in the face of such harassment or abuse.
Thus, in the second phase of our research, we explored young adolescents’ reactions to hypothetical events that simulated the types of harassment experienced by the school shooters, albeit in somewhat milder form.
We constructed vignettes that portrayed incidents in which the target character (same gender as the study participant) is ridiculed or put down by
peers, in two cases, by a teacher, or by someone in whom they were romantically interested. A sample peer vignette reads as follows: “Jason is coming to school and is walking toward the school building. A small group of
kids come toward him and start to tease him about what he is wearing and
how he looks. As he tries to get away, they shove him and make loud insulting comments that all of the other kids can hear. (This wasn’t the first
time something like this had happened to Jason). Imagine you are Jason.”
In the second peer vignette, Brian is the only one who doesn’t get a party
invitation that is being passed out in class. The distributor insults Brian by
calling him a loser and all of the kids in the class start laughing. In a third
vignette, the teacher intercepts a note that the target adolescent is passing
and reads it out loud to the entire class, provoking the students to laugh. In
the fourth vignette, the target adolescent is at a dance, where his (or her)
dance partner, one of the most popular kids at school, tells him (her) that he
(she) dances like a dweeb, in front of everyone who laugh and comment
“Looks like you got dissed!” as the popular adolescent leaves the target adolescent standing alone on the dance floor.
Two features about these vignettes are notable. First, there is a peer audience who either laughs at the victim or perpetuates the harassment. Secondly, each vignette ends with a statement to the effect that this is not the
first time something like this has happened to the target child, thus building
in the implication of a history of insult. These features characterized the harassment experienced by the actual school shooters. With regard to the role
of the audience, our most recent research examines a prototype approach to
humiliation with college students as participants, addressing commonali-
Harter, Low, and Whitesell
15
ties (or prototypic features) of humiliation (Harter, Kiang, Whitesell, &
Anderson, 2002). Findings reveal that every student reported that an audience was an important contributor to the humiliation experience. Moreover, 89% indicated that the more people that observed the humiliating
event, the greater the level of humiliation. In addition, 61% spontaneously
mentioned that the onlookers would laugh at the victim. Thus, the vignettes
appear to capture some of the key causes leading one to feel humiliated.
Emotional, cognitive, and behavioral reactions. At the end of each vignette, the participant was asked to imagine that he were victim. What followed were questions, all on a four-point scale, about (a) emotional reactions
(how humiliated you would feel in front of others; how angry at the perpetrator, how angry toward the self, and how sad or depressed); (b) cognitive
reactions (how unfair was the act, and how difficult would it be to get the
incident out of your mind); and (c) behavioral reactions (along a continuum from doing nothing to, at the extreme, planning a way to get back at
the perpetrator by seriously harming him/her and thoughts of harming anyone, it wouldn’t matter who they were. This second form of harm was included given the fact that the school shooters engaged in firing into a crowd
of students, randomly.
We were first interested in the level of humiliation reported as well as
the relationships among the initial emotional and cognitive reactions to the
vignettes, combined. (Preliminary analyses revealed that both emotional
reactions and cognitive reactions were highly correlated across the four vignettes.) As anticipated, the mean level of humiliation was 3.0 (on a
four-point scale) indicating that for the sample as a whole, the harassing
events did provoke a relatively high level of humiliation. The entire range
of emotional and cognitive reactions, as can be seen in Table 2, were found
to be highly related to each other, forming a constellation of psychological
responses to potentially humiliating events. Noteworthy is the fact that humiliation is significantly associated with feelings of anger, both toward the
perpetrator (r = .41) as well as toward the self (r = .28), and with depression
(r = .51). Moreover, anger at the perpetrator (an externalizing reaction) is
also significantly associated with two internalizing reactions, anger at the
self (r = .41) and depression (r = .51). These findings are consistent with
those in the general model and with the literature on the co-morbidity of
externalizing and internalizing reactions.
Humiliation is also related to the two more cognitive reactions, namely
how unfair the harassment is (r = .40) and how difficult it is to get the event
out of one’s mind (r = .51). In fact, each of the emotions associated with
humiliation, anger at the perpetrator, anger at self, as depression are also
predictive of judgments of unfairness (r’s of .46, .21, and .38, respectively)
16
JOURNAL OF SCHOOL VIOLENCE
TABLE 2. Correlations Among Emotional and Cognitive Reactions to Vignette
Scenarios
Humiliation
Anger/Other
Anger/Other
Anger/Self
Depressed
Unfair
.41
Anger/Self
.28
.41
Depressed
.56
.40
Unfair
.40
.46
.21
.47
Difficulty/
Out of mind
.51
.49
.45
.48
.38
.47
All correlations significant at p < .01 or higher
and the difficulty of getting the event out of one’s mind (r’s of .49, .45, and
.48, respectively). Thus, there would appear to be a network of understandable correlates of humiliation that reflect the complexity of reactions to harassment that is consistent with the research on multiple emotions (see
Harter, 1999) and the attribution literature (see Dodge, 1980, 1986).
Violent versus non-violent ideators. A major goal of this research was to
develop a profile of those who, in response to hypothetical humiliating
events, indicated that they would respond violently (plan a way to get back
at the perpetrator by seriously harming him/her in some way and think of
seriously harming anyone, it wouldn’t matter who they were) versus those
who would not respond violently (rather, they reported that they would do
nothing or would attempt a constructive solution (e.g., try to talk to the person). Among the violent group, we identified 22 males and 11 females.
Among the non-violent group, we identified 43 males and 65 females. One
goal was to determine how these groups differed on the potential predisposing factors included in the general model, described earlier.
Because the vignettes described events that were specifically designed
to provoke feelings of humiliation (put-downs, harassment, threats to the
self, in front of an audience) we did not anticipate that the two groups
would differ significantly on humiliation, nor did they. However, the violent ideators, compared to the nonviolent ideators, reported greater anger at
the perpetrator (Ms = 3.40 versus 3.15), greater anger at the self (Ms = 2.45
versus 2.12), more blame toward the perpetrator (Ms = 3.33 versus 2.86)
and more blame toward the self (Ms = 1.98 versus 1.69), all of which were
significant at p < .05. These findings further address the co-occurrence of
internalizing and externalizing reactions.
Of particular interest was the finding that the violent ideators reported
that if the event had happened to them, they would have greater difficulty
Harter, Low, and Whitesell
17
getting the event “out of their mind,” compared to the nonviolent ideators
(Ms = 2.8 versus 2.3, respectively, p < .01). Moreover, an additional question at the end of each vignette asked how often has such an event actually
happened to them, in their own life. The violent ideators reported that such
events were experienced more frequently (Ms of 2.0 versus 1.4, respectively, p < .01). It could be inferred, therefore, that if such events occurred
more often in one’s life, coupled with the fact that one was ruminating
more about these experiences, one would be more likely to react violently.
To examine this hypothesis directly among the majority of the entire sample (77%) who viewed the event as highly humiliating, we conducted a 2
(frequency of the event) ⫻ 2 (level of rumination) ANOVA where level of
violence was the dependent measure. Consistent with our expectation, we
found that those high on the frequency of the event in their own lives who
also had high levels of rumination reported significantly (p < .001) higher
levels of violence (M = 2.6) than those low on the frequency of the event
coupled with a low level of rumination (M = 1.3). The other two groups fell
in between.
Group differences for the predisposing factors identified in the model.
We were also interested in whether the violent ideators differed from the
nonviolent ideators on the predisposing factors identified in the original
model. Figure 2 presents significant differences (p’s < .01 or better) associated with components of the depression composite, as well as suicidal
ideation, homicidal ideation, and physical aggression, where lower scores
represent more negative reactions. The violent ideators reported greater depressed affect, lower self-worth, and greater hopelessness. In addition, the
violent ideators reported greater suicidal ideation, greater homicidal ideation,
and higher levels of anger-induced physical aggression. Thus, the pattern is
consistent with our framework emphasizing the co-occurrence of internalizing and externalizing reactions, where violent ideators report higher levels of both.
With regard to the predictors of these reactions, drawn from the model,
we compared perceptions of competence/adequacy as well as social support across the violent and nonviolent ideators. On our four-point scales,
we found marked and highly significant (p’s between .01 and .001) differences between the violent and nonviolent ideators for the self-concept domains of peer likeability (M’s = 2.49 versus 3.24), physical appearance
(M’s = 2.22 versus 2.86), and scholastic competence (M’s = 2.40 versus
2.85). With regard to social support, the other key predictor in the model,
the violent ideators also showed significantly less (p’s between .01 and
.001) support than the nonviolent ideators from both parents (M’s = 2.66
versus 2.85) and from peers (M’s = 2.62 versus 3.12). A separate measure
18
JOURNAL OF SCHOOL VIOLENCE
FIGURE 2. Differences in Predisposing Factors in the Lives of Violent and
Nonviolent Ideators
Peer
Support/
Rejection
*
.31
.34*
.1
6*
1*
.2
.26*
Peer
Likeability
.16*
*
.22
Physical
Appearance
Depression/
Adjustment
Composite
.12*
.56*
Parent
Support
Behavioral
Conduct
.13*
.32*
Suicidal
Ideation
.28*
.38*
Scholastic
Competence
.42*
Homicidal
Ideation
AngerInduced
Physical
Aggression
Expanded path model ␹ (6) = 8.27, p > .05, NFI = .99, NNFI = .99, Critical N = 507.
2
of peer rejection also markedly differentiated the two groups where the
lower score reflects more negativity or, in this case, greater rejection (M’s =
2.32 versus 3.19). Thus, the factors that predispose young adolescents to
depression, aggression, homicidal ideation and suicidal ideation in the general model also discriminate between those who report violent ideation in
reaction to the hypothetical events in the vignettes involving harassment
and humiliation.
MORE BEHAVIORAL PREDICTORS OF VIOLENT IDEATION
In addition to those predisposing factors identified in our general model,
media analyses and the psychological research literature suggest two other
variables that contribute to violence, (a) preoccupation with violent media
(movies, TV, music, video games) and (b) interest in and access to weapons and bombs. Many of the school shooters were attracted to violent media. All, obviously, had access to weapons and many were preoccupied
with guns, some with bomb-making.
Harter, Low, and Whitesell
19
With regard to interest in violent media, there is a growing body of data
suggesting significant relationships between violent film viewing and aggressive behaviors (Huessman & Eron, 1986; Livingstone, 1996; Rutter,
Giller, & Hagell, 1998). A meta-analysis (Redecki, 1990) revealed that
75% of 1,000 research studies found that there were negative consequences associated with exposure to violent TV. Possible explanations for
the influence of violent TV on behavior include desensitization or
disinhibition of aggressive tendencies as well as the incorporation of violence into one’s psyche (Rutter et al., 1998). The link between desensitization and violent media exposure is not new. The military has long been
employing violent video games as training devices to desensitize soldiers
to killing. There has been less research on the effects of violent video
games, per se. However, there is accumulating evidence that exposure to
violent video games is followed by more negative and violent behavior in
both children (see Silvern & Williamson, 1987) and adolescents (Dominick,
1984). Others have analyzed the content of popular video games for boys.
For example, a recent study by Dietz (1998) found that 79% of the games
involved some type of aggression or violence and 21% portrayed aggression or violence directed specifically at women. However, there are caveats
in terms of the role of preoccupation with violence in the media. Does such
an interest fuel violence or, conversely, do violent boys express a special
interest in violent media? Regardless of the directionality of effects, interest in violent media would appear to be a risk factor in developing a profile
of predictors of adolescent violent ideation.
With regard to interest in and access to weapons, recent treatments of
school violence have pointed out that our youth have ample access to
weapons (see Garbarino, 1999; Pollack, 1998). Statistics indicate that approximately 40% of households in this country have guns of some type
(Kingery, Coggeshall, & Alford, 1998). In addition, young males, in particular, indicate that there is easy access to getting guns on the street, including legally. (Shortly after the Columbine killings, adolescents interviewed
by reporters of the Denver Post claimed that you could get a gun faster than
you could have a pizza delivered.) Thus, those boys preoccupied with violence who entertain thoughts of violent actions against others typically
have the means to act on their intentions.
Based on these potential predictors, we developed an instrument (what I
do in my spare time) where we included items that tapped interest in violent media, asking, on a four-point scale, how much do they engage in activities such as watching violent movies, violent TV shows, listening to
gangster rap, heavy metal, and playing violent video games. To assess access/interest in weapons, items tapped time spent playing with guns or
20
JOURNAL OF SCHOOL VIOLENCE
other weapons, hunting or practice shooting, and making or playing with
explosives and/or bombs. The alpha for items tapping engagement in violent media was .82; for interest in/access to weapons, .73. (The complete
scale included non-violent activities for balance, such as “hanging with
friends,” “working at a job,” “watching comedy or family TV shows,”
“watching comedy movies,” “listening to popular music,” and so forth).
A comparison of violent and nonviolent ideators (based on their self-reported responses to what they would do in response to harassment) revealed that the violent ideators were significantly (p < .01) more likely to
be engaged in violent media (M = 3.12) compared to the nonviolent
ideators (M = 2.49). The two groups also differed significantly (p < .01) on
interest in and access to weapons and bombs. Thus, these two variables are
risk indicators for violent ideation. What we do not know from this study is
the extent to which violent ideation will translate into violent behavior.
Moreover, since the incidence of actual school violence is relatively low,
evidence of violent ideation cannot directly predict this type of violent action. Yet the identification of violent ideation among children and adolescents can alert responsible adults (teachers, mental health workers, parents)
to the fact that those harboring violent or suicidal thoughts are at risk for a
variety of psychological adjustment problems. They are obviously unhappy, for good reason according to their own self-reported life experiences, and the depressive and aggressive content in their introspections
undoubtedly interferes with more constructive energies devoted toward
school success, extra-curricular activities, positive social interactions, and
rewarding family life in which they can be validated for who they are as
people. It is unlikely that many adolescents, harboring suicidal or homicidal ideation, are sharing these thoughts with the significant adults in their
lives, an important question for further research. While they may be sharing such thoughts with close friends, pacts of secrecy and loyalty will typically prevent peers from bringing such ideation to the attention of
responsible adults.
CAN SCHOOL OFFICIALS IDENTIFY THOSE
AT RISK FOR POTENTIAL VIOLENT BEHAVIOR?
School officials have become increasingly sensitive to the need to deal
with issues of both suicide and violence, and there is a growing literature on
just what actions schools can take to deal with these growing problems (see
Elliot, Hamberg, & Williams, 1998). Our own research has specifically
dealt with ideation, both about suicide and more recently violent aggres-
Harter, Low, and Whitesell
21
sion toward others, and the fact that each of these forms of thinking about
harming the self and others does co-occur. School officials typically have
access to records documenting overt behavior, for example, evidence of
delinquency, conduct problems, acting out at school, gang behavior, aggressiveness with peers, and involvement with the law. However, in the
case of the ten school shooters, most had no such history and a few had minor, normative scrapes with the law but nothing that would identify serious
violent tendencies. This suggests that educators and clinicians we need to
think differently about the potential violent acts of those secretly harboring
fantasies that they may or may not act on, those who have not come to the
attention of school personnel or community officials, those who do not fit
the profile of the delinquent, the diagnosed conduct-disordered youth, the
known gang member. The next school shooter may have never come to the
attention of responsive school officials because he does not, overtly, manifest any signs of his intentions to such officials.
In our study, we asked teachers to rate students’ behavioral conduct (acting out, getting in trouble, misbehaving). Teachers are relatively accurate
in making such judgments. However, we found that of those whom teachers rated as well-behaved, 92% were low on violent ideation. However, 8%
reported high levels of violent ideation. While this percentage is low, it
suggests that there is a small percentage of adolescents who are harboring
very serious violent thoughts but who have gone unrecognized because
their manifest behavioral conduct is commendable in the eyes of observing
adults.
Self-report measures have come under increasing criticism in recent
years. Yet our findings suggest that adolescents can and do report on both
suicidal and violent ideation, and that school officials cannot necessarily
detect these pernicious forms of thought, nor should that necessarily be
their job. We feel that psychologists, such as our own team, and others, can,
through sensitive self-report screening devices, help to identify students
that report on these debilitating thoughts, toward the goal of referring them
to professionals that may help to improve the quality of their lives, and perhaps prevent incidents of suicide and homicide.
There also needs to be research on different pathways to violent ideation
and action. The research on school and community violence tends to focus
on a “one model fits all” (see Elliot, Hamburg, & Williams, 1998). The
school shooters did not seem to be cast out of the same mold as delinquents, gang members, or those with clinical diagnoses of conduct-disordered behavior. They were not inner-city youth. They were all white males
from suburban America, from middle class families. They did not target a
single victim to eliminate, nor did they do a drive-by shooting or a
22
JOURNAL OF SCHOOL VIOLENCE
school-yard extermination of one particular individual. Rather, they
splayed bullets into crowds of students, seemingly randomly in many
cases. This would appear to be a very different population of killers than
those who have populated the landscape of violent youth in the literature.
Our own research to date cannot make these discriminations. It only alerts
us to the need to think about different pathways to violent and suicidal
ideation, and to move to strategies to identify those who are not at obvious
risk, given the observations of responsible adults.
CONCLUSIONS
Our efforts, based on a theoretically-derived model of global self-esteem, its predictors, correlates, and consequences has revealed that perceived competence or adequacy in domains of importance to the self, as
well as perceived social support from peers and parents, represent predisposing factors leading to both depressive and anger-induced aggressive reactions. These, in turn, predict both homicidal and suicidal ideation,
revealing that far from being independent reactions to psychosocial factors
in one’s history, thoughts of violence against others as well as against the
self co-occur in the minds of those with a negative history.
Our findings, based on responses to vignettes simulating the harassing
events in the lives of the school shooters, have also illuminated our understanding of the role of humiliation within the school context and how it can
provoke a complex constellation of related reactions, including anger toward the perpetrator, anger at self, and depression. For a small percentage
of adolescents, humiliation and its related effects fuels thoughts of violent
revenge. Those entertaining such thoughts, in turn, report general histories
characterized by perceptions of inadequacy, lack of peer and parental support, low self-esteem, depressed affect, hopelessness and anger-induced
aggression. Such individuals also report higher levels of both suicidal and
homicidal ideation. Attempts to identify students at risk, particularly given
that a minority are unlikely to have come to the attention of school officials,
is critical. Self-report measures would appear to be a useful tool for the
identification of such students.
Current school violence programs reveal a range of approaches from institutional safety precautions, for example, gun safety, increased police
presence, metal detectors (Mercy & Rosenberg, 1998; Northrup &
Hamrick, 1990), to anti-bullying curricula (Olewus, 1996), to the development of prosocial skills, conflict management, and anger control (Catalano &
Hawkins, 1996; Hawkins, Farrington, & Catalano, 1998). While these are
Harter, Low, and Whitesell
23
worthy efforts, our own findings suggest that prevention programs should
also attend to the psychosocial factors we have illuminated, namely, perceptions of inadequacy (and their causes), lack of social support, low
self-esteem, depressed affect, hopelessness, homicidal and suicidal ideation,
as well as the emotional concomitants of harassment and bullying, including humiliation. The challenges are legion, but well worth the effort in our
attempts to protect and promote a valuable societal commodity, our youth.
REFERENCES
Achenbach, T.M., & Edelbrock, C.S. (1983). Manual for the Child Behavior Checklist and
Revised Child Behavior Profile. Burlington, VT. University of Vermont.
Bach-Y-Rita, G., & Veno, A. (1974). Habitual violence: A profile of 82 men. American
Journal of Psychiatry, 131, 1015-1017.
Boergers, J., Spirato, A., & Donaldson, D. (1998). Reasons for adolescent suicide attempts:
Associations with psychological functioning. Journal of the American Academy of
Child and Adolescent Psychiatry, 37, 1287-1293.
Booth, C. L., & Rose-Krasnow, L., McKinnon, J., & Rubin, K. H. (1994). Predicting
social adjustment in middle childhood: The role of preschool attachment security
and maternal style. From family to peer group: Relations between relationship systems. Social Development, 3, 89-204.
Bridewell, W. B., & Chang, E. C. (1997). Distinguishing between anxiety, depression, and
hostility: Relations to anger-in, anger-out, and anger control. Personality and Individual Differences, 22, 189-204.
Capaldi, D.M., & Stoolmiller, M. (1999). Co-occurrence of conduct problems and depressive symptoms in early adolescent boys: III. Prediction to young-adult adjustment. Development and Psychopathology, 11, 59-84.
Carlson, G.A., & Cantwell, D. P. (1982). Suicidal behavior and depression in children and
adolescents. Journal of the American Academy of Child Psychiatry, 21, 361-368.
Catalano, R.F., & Hawkins, J.D. (1996). The social development model: A theory of antisocial behavior. In J.D. Hawkins (Ed.), Delinquency and Crime: Current Theories (pp.
149-197). New York: Cambridge University Press.
Cooley, C. H. (1902). Human nature and the social order. New York: Charles Scribner’s
Sons.
Deffenbacher, J.L., & Swaim, R.C. (1999). Anger expression in Mexican American and
White non-Hispanic adolescents. Journal of Counseling Psychology, 46, 61-69.
Dietz, T. (1998). An examination of violence and gender role portrayals in video games: Implications for gender socialization and aggressive behavior. Gender Roles, 38, 425-442.
Dodge, K.A. (1980). Social cognition and children’s aggressive behavior. Child Development, 51, 162-170.
Dodge, K. A. (1986). A social information processing model of social competence in children. In M. Perlmutter (Ed.), The Minnesota Symposium on Child Psychology (Vol. 18,
pp. 77-125). Hillsdale, NJ: Erlbaum.
24
JOURNAL OF SCHOOL VIOLENCE
Dominick, J.R. (1984). Videogames, television violence, and aggression in teenagers.
Journal of Communication, 34, 136-147.
Dubow, E. F., & Reid, G. J. (1994). Risk and resource variables in children’s aggressive
behavior: A two-year longitudinal study. In R. L. Huesmann (Ed.), Aggressive behavior: Current Perspectives. Plenum series in social/clinical psychology (pp. 187-211).
NY: Plenum Press.
Dykeman, C., Daehlin, W., Doyle, S., & Snow Flamer, H. (1996). Psychological predictors of school-based violence: Implications for school counselors. The School Counselor, 44, 35-47.
Elliot, D. S., Hamburg, B. A., & Williams, K.R. (Eds.) (1998). Violence in American
Schools: A New Perspective. New York: Cambridge University Press.
Garbarino, J. (1999). Lost boys: Why our sons turn violent and how we can save them. NY:
The Free Press.
Gispert, M., Davis, M. C., Marsh, L., & Wheeler, K. (1987). Predictive factors in repeated
suicide attempts by adolescents. Hospital and Community Psychology, 38, 390-393.
Goetz, T.W., & Dweck, C.A. (1980). Learned helplessness in social situations. Journal of
Personality and Social Psychology, 39, 246-255.
Harter, S. (1985). The Self-Perception Profile for Children. Manual, University of Denver,
CO.
Harter, S. (1985). The Social Support Scale for Children and Adolescents. Manual, University of Denver, CO.
Harter, S. (1999). The Construction of the Self. New York: Guilford Press.
Harter, S., Kiang, L., Whitesell, N.R., & Anderson, A. (2002). An emotion prototype approach to humiliation. Society for Research in Child Development, poster submission,
for the 2003 meetings in Tampa, Florida.
Harter, S., & Marold, D.B. (1993). The directionality of the link between self-esteem and
affect: Beyond causal modeling. In D. Cicchettit & S. L. Toth (Eds.), Rochester Symposium on Developmental Psychopathology: Disorders and Dysfunctions of the Self
(Vol. 5, pp. 333-370). Rochester, NY: University of Rochester Press.
Harter, S., Marold, D. B., & Whitesell, N. (1992). A model of psychosocial risk factors
leading to suicidal and ideation in young adolescents. Development and Psychopathology, 4, 167-188.
Harter, S., Nowakowski, M., & Marold, D. B. (1985). The Dimensions of Depression Profile for Children and Adolescents: Revision. Manual, University of Denver.
Hawkins, J. D., Farrington, D. P., & Catalano, R. F. (1998). Reducing violence through the
schools. In D. S. Elliot, B. A. Hamburg, & K. R. Williams (Eds.), Violence in American
schools: A new perspective. New York: Cambridge University Press.
Hawkins, J. D., & Lishner, D. M. (1987). Schooling and delinquency. In E. H. Johnson
(Ed.), Handbook on Crime and Delinquency Prevention (pp. 179-221). New York:
Greenwood Press.
Hillbrand, M. (2001). Homicide-suicide and other forms of co-occurring aggression
against self and against others. Profession Psychology: Research and Practice, 32,
626-635.
Huessman, L.R., & Eron, L.D. (1986). The development of aggression in American children as a consequence of television violence viewing. In L.R. Huessman & L.D. Eron
Harter, Low, and Whitesell
25
(Eds.), Television and the Aggressive Child: A Cross-National Comparison (pp.
45-80). Hillsdale, NJ: Lawrence Erlbaum Associates.
Inamdar, S. C., Lewis, D. O., Siomopoulous, G., Shanok, S. S., & Lamela, M. (1982). Violent and suicidal behavior in psychotic adolescents. American Journal of Psychiatry,
139, 932-935.
James, W. (1892). Psychology: The Briefer Course. NY: Henry Holt & Co.
Jorekskog, K. G., & Sorbom, D. (1998). LISRELVIII. Chicago: Scientific Software International, Inc.
Khan, A. U. (1987). Heterogeneity of suicidal adolescents. Journal of the American Academy of Child and Adolescent Psychiatry, 26, 92-96.
Kingery, P.M., Coggeshall, M.B., & Alford, A.A. (1998). Violence at school: Recent evidence from four national surveys. Psychology in the School, 35, 247-258.
Kovacs, M., & Beck, A. T. (1977). An empirical-clinical approach towards a definition of
childhood depression. In J. G. Schulterbrandt & A. Raskin (Eds.), Depression in Childhood: Diagnosis, Treatment, and Conceptual Models. NY: Raven Press.
Kovacs, M., & Beck, A. T. (1978). Maladaptive cognitive structures in depression. American Journal of Psychiatry, 135, 525-533.
Leary, M., Shreindorfer, L. S., & Haupt, A. L. (1995). The role of low self-esteem in emotional and behavioral problems: Why is low self-esteem dysfunctional? Journal of Social and Clinical Psychology, 3, 297-314.
Lipschitz, A. (1995). Suicide prevention in young adults (18-30). In M. Silverman and R.
Maris (Eds.), Suicide Prevention Toward the Year 2000. New York: Guilford Press.
Livingstone, S. (1996). On the continuing problem of media effects. In J. Curran & M.
Gurevich (Eds.), Mass Media and Society, pp. 305-324. London: Arnold.
Mead, G. H. (1934). Mind, self, and society from the standpoint of a social behaviorist.
Chicago: University of Chicago Press.
Mercy, J. A., & Rosenberg, M. L. (1998). Preventing firearm violence in and around
schools. In D. S. Elliot, B. A. Hamburg, & K. R. Williams (Eds.). Violence in American Schools. A New Perspective. New York: Cambridge University Press.
Olweus, D. (1996). Bullying at school: Knowledge base and an effective intervention program. In C. Ferris and T. Grisso (Eds.) Understanding Aggressive Behavior in Children, Annals of the New York Academy of Sciences, v. 794. The New York Academy of
Sciences: New York.
Olweus, D. (1997). Bully/victim problems in school: Facts and intervention. European
Journal of Psychology of Education, 12, 495-510.
Patterson, G. R., & Bank, C. L. (1989). Some amplifying mechanisms for pathologic processes in families. In M. Gunnar & E. Thelan (Eds.), Systems and Development: Symposia on Child Psychology (pp. 167-210). Hillsdale, NJ: Erlbaum.
Pfeffer, C. R., Solomon, G., Plutchick, R., & Mizruchi, M. S. (1984). Suicidal behavior in
normal school children: A comparison with child psychiatric inpatients. Journal of the
American Academy of Child Psychiatry, 23, 416-423.
Plutchik, R. & van Praag, H. (1989). The measurement of suicidality, aggressivity and
impulsivity. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 13,
S23-S34.
Pollack, W. (1998). Real Boys: Rescuing Our Sons from the Myths of Boyhood. NY: Random House.
26
JOURNAL OF SCHOOL VIOLENCE
Radecki, T.E. (1990). Television promotes teen violence. In J. Rohr (Ed.), Violence in
America: Opposing Viewpoints, San Diego, CA: Greenhaven Press.
Reinecke, M.A. (1995). Co-morbidity of conduct disorder and depression among adolescents: Implications for assessment and treatment. Cognitive and Behavioral Practice,
2, 299-326.
Renouf, A.G., & Harter, S. (1990). Low self-worth and anger as components of the depressive experience in young adolescents. Development and Psychopathology, 2, 293-310.
Rigby, K., & Slee, P. (1999). Suicidal ideation among adolescent school children, involvement in bully-victim problems, and perceived social support. Suicide and Life Threatening Behavior, 29, 119-130.
Ross, D. M. (2002). Bullying. Mahwah, NJ: Lawrence Erlbaum Associates.
Rutter, M. (1987). Psychosocial resilience and protective mechanisms. American Journal
of Ortho-Psychiatry, 57, 316-331.
Rutter, M., Giller, H., & Hagell, A. (1998). Antisocial Behavior by Young People. New
York: Cambridge University Press.
Shafii, M., Carrigan, S., Whilinghil, L., & Derrick, A. (1985). Psychological autopsy of
completed suicide in children and adolescents. American Journal of Psychiatry, 142,
1061-1064.
Silvern, S.B. & Williamson, P.A. (1987). The effects of video game play on young children’s aggression, fantasy and prosocial behavior. Journal of Applied Developmental
Psychology, 8, 453-462.
Simons, K. J., Paternite, C. E., & Shore, C. (2001). Quality of parent/adolescent attachment
and aggression in young adolescents. Journal of Early Adolescence, 21, 182-203.
Tangney, J.P., & Dearing, R.L. (in press). Shame and Guilt. New York: Guilford Press.
Tangney, J.P., & Fischer, K.W. (Eds.). (1995). Self-Conscious Emotions: The Psychology
of Shame, Guilt, Embarrassment, and Pride. New York: Guilford Press.
van Praag, H.M., Plutchik, R., & Apter, A. (Eds.) (1990). Violence and Suicidality: Perspectives in Clinical and Psychobiological Research. New York: Brunner/Mazel.
Velaz, C. N., & Cohen, P. (1988). Suicidal behavior and ideation in a community sample
of children: Maternal and youth report. Journal of the American Academy of Child and
Adolescent Psychiatry, 27, 349-356.
RECEIVED: October 18, 2002
ACCEPTED: December 23, 2002