Professional Documents
Culture Documents
Out
Out
A Descriptive and Comparative Study of Students at Risk for Targeted School Violence
and Students Who Perpetrated Targeted School Violence
A dissertation submitted
by
Maria L. Martinez
to
in partial fulfillment of
the requirements for the
degree of
Doctor of Philosophy
in
Clinical Psychology
with emphasis in
Depth Psychology
Pro Q ue st 10140662
Pub lishe d b y Pro Q ue st LLC (2016). Co p yrig ht o f the Disse rta tio n is he ld b y the Autho r.
Pro Q ue st LLC.
789 Ea st Eise nho we r Pa rkwa y
P.O . Bo x 1346
Ann Arb o r, MI 48106 - 1346
RISK FOR TARGETED SCHOOL VIOLENCE ii
Copyright by
Maria L. Martinez
2016
RISK FOR TARGETED SCHOOL VIOLENCE iii
Abstract
A Descriptive and Comparative Study of Students at Risk for Targeted School Violence
by
Maria L. Martinez
This study examined a cohort of 55 students at risk for Targeted School Violence (TSV)
ethnicity, academic performance, family composition, bullying, target site and selection,
socialization pattern, and pre-attack and planning strategies, all of which were examined
in relation to TSV. The data gathered for this study were examined through the use of
multiple linear regression analyses. Findings in this study suggest that history of violence
is a significant factor in determining risk for TSV. Living with a single parent places a
student at a more risk for violence. Living with a foster mother was found to be a
protective factor in this study. Being a victim of bullying increased risk in both sexes,
and, finally, gender was not a factor for TSV. Preventive measures were discussed,
between law enforcement, education, and mental health professionals. The development
of a safety net and the identification of protective factors among students at risk for TSV
Acknowledgements
To my mother Angela, whom !"# $%&'()* + "&# ,'-. /# �#&1 2."3
you for the times you stayed up all night with me to comfort and guide me. I know you
are watching from above, and it helps to know that our love is forever.
To Dr. Tony Beliz, my mentor in the Department of Mental Health. Thank you for your
insights, supervision, and support. You helped me understand others and myself in a way
that helped me develop my personal and professional self. Your passion for the work we
do and our profession came at a time when I was not sure about my next step.
To Dr. Oksana Yakusko, my Chair. Your guidance and support during difficult times
made this possible. Thank you for listening and being there for me. You helped me
realize my dream of becoming a psychologist.
To my colleague Bob Martin. I am forever grateful for your willingness to share with me
your wealth of knowledge and experience about threat assessment and MOSAIC.
Finally, to Dr. Enrique Lopez and Dr. Tomas Martinez. Your compassion, support, and
guidance throughout this process will be forever appreciated. I was humbled by your
willingness to help and your availability despite your busy schedules.
All of you touched my heart and provided with me the role models and style that I hope
to incorporate fully as I continue my personal and professional journey. My heart will
always carry your wisdom and style.
RISK FOR TARGETED SCHOOL VIOLENCE v
TABLE OF CONTENTS
Suicidality .......................................................................................................... 29
Homicidality ...................................................................................................... 29
Bullying/victimization ....................................................................................... 34
Protective Factors........................................................................................................... 36
Resiliency ........................................................................................................... 36
Community ........................................................................................................ 38
School ................................................................................................................ 38
Reflexivity...................................................................................................................... 55
Family Composition....................................................................................................... 70
Gender ............................................................................................................................ 71
Bullying.......................................................................................................................... 72
Recommendations .......................................................................................................... 77
Conclusions .................................................................................................................... 81
References .................................................................................................................................. 83
The style used throughout this dissertation is in accordance with the Publication Manual
of the American Psychological Association (6th Edition, 2009), and Pacifica Graduate
(2015-2016).
RISK FOR TARGETED SCHOOL VIOLENCE
Chapter 1
Introduction
Colorado. These communities became widely known after the much publicized and
tragic shootings at Isla Vista, Sandy Hook Elementary School, Virginia Tech, and
Columbine High School (Shultz, Muschert, Dingwall, & Cohen, 2013). Each of these
well-known tragedies resulted in loss of life, severe injury, and both acute and long-term
2007). School shooting rampages have been an issue of concern since the mid-1990s. At
this point school violence in America as a whole was on the decline, but shooting events
began to occur in suburban and rural communities that people up to that point felt were
(Burns & Crawford, 1999; Reddy, Borum, Berglund, Vossekuil,
Fein, & Modzeleski, 2001; Temlow , & Vernberg, 2002). In the
9-year-period between 1992 and 2001, there were 35 incidents of violence involving
their peers and teachers (National Research Council and Institute of Medicine, 2003).
Although it is a fact that mass shootings on campuses are rare and account for only
0.04% of national firearm homicides, it is also true that the impact of these shootings is
far wider, due to the amount of public attention and media coverage they garner
(Hawdon, Oksanen, & Rasanen, 2012; Meloy, Hempel, Mohandie, Shiva, & Gray, 2001;
Shultz et al., 2013). The number of incidents together with intense media coverage has
led to increasing calls for strategies to deal with the threat of targeted school violence
(TSV), where the school setting is a key component of the attack rather than incidental to
RISK FOR TARGETED SCHOOL VIOLENCE 2
it (Fein, Vossekuil, Pollack, Borum, Modzeleski, & Reddy, 2002; Krauss, 2005).
the characteristics of students who engaged in TSV (McGee & DeBernardo, 1999; Meloy
et al., 2001; ). The United States Secret Service and the United States
Department of Education worked together to create a joint task force that performed a
detailed analysis of 37 TSV events (Vossekuil, Fein, Reddy, Borum, & Modzeleski,
2004). The joint task force study provides the most complete analysis of actual school
shooters where the attacker was a current or recent student and where the school was
In May of 2002, the joint task force published a report entitled Final Report and
Findings of the Safe School Initiative: Implications for Prevention of School Attacks in
the United States (U.S. Secret Service and Department of Education, 2002). This report
found that the majority of the school shooters profiled exhibited a pattern of
and managing threats and other behaviors of concern. Particular importance in this report
is placed on the emphasis on using the threat assessment model devised by the U.S.
Secret Service (Borum, Fein, Vossekuil, & Berglund, 1999; Fein, Vossekuil, & Holden,
Los Angeles, the School Threat Assessment Response Team (START) program was
initiated in 2008 as a joint partnership between the Los Angeles Police Department in
RISK FOR TARGETED SCHOOL VIOLENCE 3
partnership and the Los Angeles County of Mental Health. This partnership has been
characterized as one of the most intensive efforts in the nation to prevent school
local and federal law enforcement agencies, educational institutions, threat management
teams, hospitals, and other agencies to prevent TSV in the city of and county of Los
licensed mental health clinicians that include clinical psychologists, psychiatric nurses,
marriage and family therapists, clinical social workers, and case managers. The START
team receives intensive and specialized training in violence threat risk assessment and
interviewing techniques, and members of the team are also certified under the
they are a danger to themselves, others, or gravely disabled due to a mental disorder. In
2009 the Department of Mental Health Emergency Outreach Bureau expanded the
reach of START to the County of Los Angeles (Goode, 2013). START has three main
objectives: (1) the prevention of targeted school violence in Los Angeles County, (2)
provision of support and assistance to students of concern, and (3) development of the
partnerships required to accomplish the previous two points. In order to meet these
institutions, and mental health systems of care, parents, and campus threat management
program.
thousands of incidents where law enforcement officials, school authorities, and other
school, high school, and college campuses. In the 2012-2013 fiscal year alone the
previous fiscal year. Whereas only 55 students in the past five years have been
considered moderate to high risk, the immediate and long-term consequences of just one
In September 2012, START was recognized by the Ash Center for Democratic
through their Bright Ideas in Government program. START emphasizes prevention and
early intervention through partnerships with law enforcement agencies, school districts,
mental health programs, and campus threat management teams. START has participated
District, the Los Angeles Unified School District School Police, and the Los Angeles
Police Department to collaborate on students of concern who reside in the city of Los
Angeles. This partnership is the first of its kind in the country that is specifically focused
students designated by START as moderate to high risk for TSV. The students were
identified in the course of more than 8,000 school threat assessments conducted by
START and the Department of Mental Health's Psychiatric Mobile Response Teams
RISK FOR TARGETED SCHOOL VIOLENCE 5
(PMRT) between March 2009 and February 2014. The cohort of 55 cases may represent
one of the largest single caseloads of moderate to high-risk cases available for review.
approach that relies on clinical expertise, a thorough diagnostic interview, and familiarity
with relevant risk and protective factors. This study sought to improve this data-driven
assessment process by identifying specific risk factors and protective factors. The
As of this writing, there have been a total of 74 instances of target school violence
in the United States in 2014 (Every Town for Gun Safety, 2015). The frequency of these
events has not decreased and warrants further study about risk factors, shooter
characteristics, and associated school violence prevention programs. This research aimed
to compare prior findings about actual school shooters to current at-risk students in the
hopes of developing further insight into the salient characteristics and behaviors of
potential school shooters. Furthermore, this research will be able to provide threat
management teams and others with further insights about risk and protective factors, as
throughout this study as a potential model for others as programs are developed to
Researchers have sought to identify the features that school shooters have in
common in terms of family life, personalities, histories, and behaviors (Langman, 2009).
RISK FOR TARGETED SCHOOL VIOLENCE 6
Although rampage school shootings are statistically rare, the magnitude of the events, as
well as the mystery of what causes them, results in widespread speculation about the
perpetrators. Further, existing studies are largely focused on the forces that drive an
individual to carry out violent targeted attacks (Blume, 1996; Blumstein, 2000; Preti,
2008; White & Lauritsen, 2010). However, there is a scarcity of data and no known
publication of existing caseloads of individuals who have been identified as at-risk for
TSV.
Purpose of Study
students designated by START as moderate to high risk. To date, the cohort of 55 cases
may represent one of the largest single caseload of moderate to high-risk individuals
available for review. The students were identified in the course of more than 8,000 school
threat assessments conducted by START and the Department of Mental Health s PMRT
between March 2009 and February 2014. In addition, this study explored the following
hypotheses and compared its findings with those outlined in May 2002 joint task force
publication entitled The Final Report and Findings of the Safe School Initiative:
Implications for the Prevention of School Attacks in the United States (U.S. Secret
Service and Department of Education, 2002). The specific research questions, and the
RQ1. A history of violence will not be a significant factor in determining risk for
The following null and alternative hypotheses were developed from RQ1:
RQ2. Coming from a single-parent home will increase risk for targeted school violence
The following null and alternative hypotheses were developed from RQ2:
Ho2: Coming from a single-parent home will not predict MOSAIC Scores.
RQ3. Males will have higher risk for targeted school violence as demonstrated in
The following null and alternative hypotheses were developed from RQ3:
Furthermore, the goal of this project is to provide threat management teams and
other mental health practitioners working with this type of population with further
insights about risk and protective factors, as well as associated case management
between mental health, education, and law enforcement, is discussed and emphasized
throughout this study in the hopes that other programs will adopt this collaboration in
Personal Interest
Supervising Psychiatric Social Worker for the School Threat Assessment Response Team
Outreach Bureau. The goal and program objective is to prevent an incident such as
RISK FOR TARGETED SCHOOL VIOLENCE 8
Columbine, Newtown, and Isla Vista by identifying students who are on a pathway to
violence with the resources required to stabilize and complete their education. As an
presentations for students, parents, and mental health professionals to create situational
awareness about targeted school violence, violence threat risk assessments, and
appropriate intervention strategies. Intensive case management and long term monitoring
are utilized to mitigate or eliminate any potential threat. Additionally, I have a personal
interest in this because of my upbringing. I migrated to the U.S. when I was 6 years old,
and I have first-hand experience of the cultural and socioeconomic issues faced by some
of the immigrants of first generation students who considered violence as a remedy. I was
raised in South Los Angeles in a single-parent home with three siblings. My mother did
not speak English and raised us by working low paying menial jobs. There were times in
my life when food was scarce and housing inadequate. These experiences occurred
within the context of a very difficult environment where gang violence, drugs, and
bullying were very common. I was not only a victim of bullying but also saw how other
classmates were impacted by the lack of a healthy learning environment, poverty, and
overall limited resources. During my childhood and adolescence I did not have anyone to
go to for help or assistance, but I was very fortunate to stay out of trouble and concentrate
instead on getting a career and having a better future. Looking back on my childhood and
adolescence experiences, I realize that prevention programs did not exist and how helpful
they would have been for my fellow students and myself. As I grew older and matured, I
became very interested in helping others who are also struggling at school, home, and in
their neighborhood. As I developed my career in social work I experienced first hand the
RISK FOR TARGETED SCHOOL VIOLENCE 9
a school threat assessment response team, I have seen students benefit significantly from
prevention programs. One of the main features of START is prevention. The program has
improved the lives of many students and not only stopped them from engaging in school
violence but also helped them channel their energies towards completing school and
Chapter 2
Literature Review
One of the most comprehensive studies on targeted school violence is the Final
Report and Findings of the Safe School Initiative: Implications for the Prevention of
School Attacks in the United States (U.S. Secret Service and Department of Education,
2002). The Safe Schools Initiative (SSI) report reviewed the case studies of 37 targeted
school attacks over the past 25 years involving a total of 41 attackers (Vossekuil et al.,
2002). Several operational variables were developed, pilot tested, revised, and used in
compared and reconciled any disparate ratings as part of the report. The SSI study found
that school shooters were exclusively male and that guns were nearly always the weapons
of choice. The researchers also found that the attacks were rarely, if ever, impulsive acts.
The majority of the attackers had a plan at least two days prior to the incident, and in
some cases, the planning had gone on for up to a year. Revenge was more often then not
a motive for the attack, as more than three quarters of the attackers held a grievance
against particular individuals or the school itself at the time of the attack.
When these attacks took place, initial reports indicated that the violence came
without warning. However, the majority of attackers communicated their ideas or plans
before the incident. In more than three quarters of the incidents, attackers told someone
about their ideation in planning an attack at the school. Frequently, attackers told friends
or other peers or acquaintances. In more than half of the cases, multiple people knew
about plans for the attack prior to its occurrence. Although communications about the
attack were frequent, the target(s) of the threat rarely received advanced notice
RISK FOR TARGETED SCHOOL VIOLENCE 11
Dwyer, Osher, and Wagner (1998) found that the majority of the students who
committed deadly violence in the schools showed signs of needing help prior to the
incident. In almost every case, the attacker engaged in behavior that caused others to be
concerned. The authors found that in three-quarters of the incidents, an adult had
expressed concern about the attacker. A large majority of the shooters had difficulty
coping with a major loss, and other individuals including parents, counselors, and peers
were aware. Nearly 75% of these adolescents had previously threatened to commit
suicide or had tried to, and more than half had a history of depression or hopelessness and
desperation.
In a majority of cases, bullying played a key role in justifying the attack (Nansel,
Overpeck, Haynei, Ruan, & Scheidt, 2003). Attackers often felt persecuted, bullied,
threatened, attacked, or injured by others prior to the incident (Nansel et al., 2003). In
most instances, these students experienced chronic and severe harassment. These
prevention programs (Cornell & Sheras, 2006). A more recent joint report issued by the
U.S. Secret Service, the U.S. Department of Homeland Security, the U.S. Department of
Education, and the Federal Bureau of Investigation entitled Campus Attacks: Targeted
2010) reiterated the findings of the 2002 report, cautioned against profiling, and
discouraged the use of checklists or reliance on a list of warning signs. Both studies
A noteworthy finding in the Drysdale et al. (2010) report is that there are common
RISK FOR TARGETED SCHOOL VIOLENCE 12
elements between targeted school violence and campus violence in general. An analysis
with other types of murders in the larger community (Bromley, 2005). As in other types
of murders, the components in the campus murderers are domestic, intimate, and
workplace violence. In the majority of the cases, there was a relationship between the
perpetrator and the victim. These findings indicate that the subject identified a specific
(2010) report also looked at the age of the subject. The youngest perpetrator was 16 years
old and the oldest was 62 years of age. With respect to gender, 32 perpetrators were male
and only one was female. Factors that motivated or triggered the directed assaults were
obsession with the target, reaction to academic stress, failing grades, acquaintance or
al. (2010) report suggests that there exist common factors among the perpetrators of
Although the 2010 report provided additional insights into targeted school
violence, the report did not address the ethnicity of the perpetrators, family composition,
health history. Additionally, the authors cited that their data analysis came from open-
source reporting, email and media reports, and review of published documents, all of
Statistics on Violence
The World Health Organization (2004) reports that approximately 44,000 people
die every day due to intentional acts of self-directed, interpersonal, or collective violence
(Krug, Dahlberg, Mercy, Zwi, & Lozano, 2002). Many thousands more victims are
injured, experience other nonfatal health consequences, or witness violent acts (Stretesky
& Hogan, 2001). Additionally, tens of thousands of lives and families are destroyed or
The World Health Organization 2004 report on violence and health states that
public health experts worldwide have substantially devoted their attention to the
prevention of violence since the 1970s. Since 1970 publications on violence have risen by
550%, from 2711 publications in the 1970s to more than 8000 in the 1990s (World
al., 2002) has developed other activities related to violence, particularly in the area of
data collection and services for victims. Public health experts have consolidated their
efforts to place violence on the global public health agenda due to the increase of violent
It is estimated that 1.6 million people died from violence in 2000, which
corresponds to 28.8 per 100,000 population (Haider-Markel & Joslyn, 2001; Lawrence &
Birkland, 2004). According to the World Health Organization (2004) report, half of these
deaths were suicides. Nearly a third were homicides, and a fifth were war related. Rates
vary significantly between and within countries (Krug et al., 2002). A 2002 study of 48
countries around the world found that 10% to 69% of women reported having been
RISK FOR TARGETED SCHOOL VIOLENCE 14
physically assaulted by an intimate partner during their lifetime; about 20% of women
and 5 to 10% of men reported having been sexually abused as children; and the few
population-based studies on elder abuse show that 4% to 6% of the elderly are abused in
some form in their homes (DeJong, Epstein & Hart, 2003; Harding, Metha, & Newman,
Delinquency Prevention, 2015), which collects information on nonfatal crimes for the
U.S. Department of Justice, reported statistical data on crimes against youth ages 12 to 17
from 1994 to 2010. Between 1994 and 2010 the overall rate of serious violent crime
against youth declined by 77%, from 61.9 victimizations per 1,000 youth ages 12 to 17 in
1994 to 14.0 per 1,000 in 2010. Among serious violent crimes against youth, the rate of
rape or sexual assault declined by 68%, robbery declined by 77%, and aggravated assault
declined by 80%. The overall rate of simple assault declined by 83% during the same
period to 125.1 victimizations per 1,000 in 2010. Declines in simple assault against youth
were similar from 1994 to 2002 (down 61%) and from 2002 to 2010 (down 56%).
Declines in serious violent crime were greater from 1994 to 2002 (down 69%) than from
Although male youth victimization rates were twice as large as female rates in
1994, male and female youth were equally likely to be victims of serious crimes in 2010
(Bureau of Justice Statistics, 2010). Male victimization rates reflected a larger decline
than the female rate. Differences between the rates of serious violent crime against male
youth ages 2 to 17 declined by 89% from 1994 to 2010, whereas the rate against female
RISK FOR TARGETED SCHOOL VIOLENCE 15
youth declined 69% during the same period. Across both sexes, serious violent crime
declined more rapidly from 1994 to 2002 than from 2002 to 2010 (Bureau of Justice
Statistics, 2010).
According to Blumstein (2000), the rate of violence between 1995 and 2000
fluctuated in the United States. Much of the fluctuation is attributable to the increase in
violence committed by young people primarily against other young people. Starting in
1985, the rates of homicide and robbery committed by people under the age of 20 started
to increase dramatically, as did the use of handguns to commit those crimes. This
increase of violence peaked in the early 1990s and proceeded to decrease significantly by
The homicide rate in the United States between 1970 and 1995 fluctuated
continuously from between 8 to 10 incidents per 100,000 in the population. In 1980, the
rate peaked at 10.2 murders per 100,000 in the population, and by 1985 the homicide rate
had fallen to 79.9 homicides but then climbed a full 24% to each a peak of 9.8 in 1991,
and has been declining markedly since then, reaching 5.5 in 2000. The last change
represents a drop of 44% since 1991, a level that is lower than any annual rate since 1965
(U.S. Census Bureau, 2010). In 2010, there were 14,748 homicides in the United States
66,160 murders from 1960 to 1996. In 2004, there were 5.5 homicides for every 100,000
persons, roughly three times as high as Canada and six times as high as Germany (Xu,
victimization, have received a great deal of deserved publicity in recent years (Fox &
RISK FOR TARGETED SCHOOL VIOLENCE 16
Zaw
1999). According to the FBI data, 1,789 persons under 18 were victims of homicide in
1999 (Fox & Zawitz, 2001). The homicide rate of 2.6 deaths per 100,000 juveniles, or
more than 5 juveniles per day, ranks the United States first among developed countries
and is nearly double the rate of the country with the next highest rate (Adekoya, 2009;
National Center for Health Statistics, 2010; Singh & Yu, 1996; U.S. Census Bureau,
2010). The rate at which juveniles are murdered in the United States is higher than in any
!" ! ! #
that young children (i.e., those age 5 and younger) face an elevated risk of homicides,
although under different conditions from adults (Boudreaux, Lord, & Dutra, 1999;
Cooper & Eaves, 1996; Hanfland, Keppel, & Weis, 1997). FBI data show 700 homicide
victims under age 6 in 1997, a rate of 2.6 per 100,000 (Federal Bureau of Investigation,
1998; Snyder & Sickmund, 1999). The homicide rate for young children was for many
years equal to the rate for teens prior to the latest recent rise (Xu et al., 2010). In spite of
the rise, more girls under age 6 were homicide victims than girls ages 12 to 17 (320
versus 230) and White children under age 6 were victims of homicide 75% as often as
White teens (1.8 per 100,000 and 2.4 per 100,000, respectively) in 1997 (Federal Bureau
The 1999 tragedy in Littleton, Colorado brought public attention to the killing of
children in schools. Media reports have created the impression that such killings are more
commonplace than they actually are (Aitken, 2001; Chyi & McCombs, 2004; Consalvo,
2003; Muschert, 2007). FBI statistics do not categorize episodes of homicide by place of
RISK FOR TARGETED SCHOOL VIOLENCE 17
occurrence. However, the 2000 Annual Report on School Safety (U.S. Department of
Justice & U.S. Department of Education, 2000) provides school homicides statistics and
describes findings from the School-Associated Violent Deaths Study. The School-
Associated Violent Deaths Study conducted by the Centers for Disease Control and
Prevention (2012) indicates that less than 1% of the children nationwide murdered in the
first half of the 1998-99 school year were victims of school-associated murders (i.e.,
policy agenda of national, state, and local authorities (Roberts, Zhang, Truman, &
Snyder, 2011). The extent and complexity of the problem and the fact that the juvenile
homicide rate in the United States continues to be substantially higher than in other
modern countries suggest that much remains to be done (Roberts et al., 2011).
Causes of Violence
Violence cannot be attributed to a single factor (Dwyer et al., 1998), as its causes
are complex and occur at different levels (Chen & Weikart, 2008). Many biological
factors have been found to be inhibitors of violence (Neiman, Murphy, Swaim, Thomas,
and blood abnormalities like hypoglycemia are only a few contributors to violence
interaction between biological factors and social environment may have some significant
influences on child development (Alvarez, Roura, & Oses, 2011; Brunner, Hence, &
RISK FOR TARGETED SCHOOL VIOLENCE 18
Prazer, 2010). The exact nature of these influences remains largely unknown (Etain,
Henry, & Bellivier, 2008). Child development researchers have found links between
According to Gratz and Roemer (2004), these include (1) toxic materials found in the
environment (e.g., lead paint), (2) traumatic head injury (e.g., as the result of child abuse
or accident), (3) dietary deficiencies, particularly prenatal, (4) alcohol and drug ingestion
by the mother during critical fetal developmental stages, and (5) birth trauma. Once the
deficits occur, attempts to remove or remedy the biological cause may include active
supportive and competent social environment has also been found to counteract the
effects that these biological factors exert on any propensity toward violence (Felittii et al.,
1998).
learns patterns of thinking, behavior, and feeling from his or her early life experiences.
Furthermore, children can learn as much from observing significant or admired others in
their environment as from their own experiences. Research also indicated that aggressive,
antisocial, and violent behaviors are often learned from significant others (e.g., TV,
movie, or fictional characters) and are held in reserve for response to specific social
Cognitive elements refer to the ideas, beliefs, and patterns of thinking that emerge
& Bartholow, 2007). Research has revealed that violent individuals have different ways
RISK FOR TARGETED SCHOOL VIOLENCE 19
in others when there is no hostility. These individuals are also less efficient at thinking of
nonviolent ways to solve social conflicts and disagreements. Violent individuals also tend
manner. Some young males, especially members of violent groups or gangs, have
adopted the belief that it is acceptable to react to every perceived or imagined sign of
disrespect with aggression. Aggressive children and adolescents have more antisocial,
aggression in others, that encourage or engender violent behavior. Almost any aversive
situation, such as continuous loud noise, unpleasant smells, and crowded, unpleasant
living condition can provoke aggression and violence in those persons submitted to such
schools, family, and peers can all be conducive to the development of violent behavior.
The presence of weapons increases the chances that the conflict will occur in the first
place and that it will have lethal consequences once it does occur (Carnagey et al., 2007).
poverty, frustration, and hopelessness are prevalent are at much greater risk for later
involvement in violence than other children (Fox & Zawitz, 2001). Childhood aggression
can predict adult violence in some individuals. Research has discovered that
the majority of violent crimes (Blume, 1996). During their childhood, violent individuals
exhibit aggression, disobedience, and disruptions at home and in the school, are disliked
RISK FOR TARGETED SCHOOL VIOLENCE 20
and avoided by peers, are neglected by parents and teachers, and are likely to fail in
school, eventually dropping out (McCabe & Martin, 2005). Unsupervised and susceptible
aggressive, and violent young adults (Muschert & Carr, 2006). They are likely to become
involved in abusive spousal relationships, and they often abuse their own children
(Martin, 2001). However, not every child growing up under these conditions follows this
destructive path, and the example of children who have not become violent has provided
valuable insights into how to design prevention programs (Ogle, Eckman, & Amoroso,
2003).
Gender Differences
The statistical data regarding violent crimes in relation to gender vary (Neroni,
2000). Researchers at the University of Arizona (Herrera & McCloskey, 2000) conducted
a 5-year study looking in the gender differences in the risk for delinquency among youth
or in the number of referrals to juvenile offenses. The study included 147 females and
152 boys between the ages of 6 to 12 at that time the study was conducted with ages
ranging from 11 to 18 years when the 5-year study concluded. The predominant ethnicity
of the children reflected a composition specific to the county where the study was
American (4%), Native American (4%), and Asian or Pacific Islander (2%) comprise the
study sample. The study concluded that girls exposed to domestic violence and child
abuse were arrested for violent offenses more than boys with similar histories, but the
context of violent offenses differed dramatically by gender. Most referrals for a violent
RISK FOR TARGETED SCHOOL VIOLENCE 21
offense for girls were for domestic violence. The study concluded that although boys and
girls share similar family risk factors for delinquency, girls are more likely than boys to
be arrested for violent offenses in the aftermath of child physical abuse. These findings
suggest that it takes more severe abuse to prompt violence in girls than is necessary to
Although youth violence has declined in recent years, a rash of school shootings
in the late 1990s generated significant public concern and attention from policy makers.
The first shooting was on October 1, 1997 in Pearl, Mississippi, where a student shot and
killed two students (Sullivan & Guerrette, 2003). That incident was followed by
students and 1 teacher killed), Edinboro, Pennsylvania (1 teacher killed), and Springfield,
Oregon (2 students killed) (Harding et al., 2003). National concern peaked on April 20,
1999 with the shootings in Littleton, Colorado, where 14 students (including the 2
gunmen) and 1 teacher were killed (Larkin, 2007; Muschert, 2007). Although these five
incidents resulted in only 28 deaths (as compared with the roughly 3,000 annual murders
of teenagers over the same period), they were unique and disturbing because they were
random quality that increased public concern about the universality of potential risks and
The majority of students will complete their education without facing or being
impacted by school violence (Fein & Vossekuil, 1998). Nevertheless, during the last
decade, cases of school shootings have changed the image of school settings as safe and
secure environments (Fein et al., 2002). A report by the Centers for Disease Control and
RISK FOR TARGETED SCHOOL VIOLENCE 22
Prevention (2004) noted that between 1991 and 2003 a high number of students did not
schools. Statistics reflect that very few of these students will experience a serious violent
act in their schools. The U.S. Secret Service and United States Department of Education.
(2002) reported 37 incidents of school-based attacks through out the United States
Targeted school violence has been defined as incidents where both the perpetrator
and targets are identified or identifiable prior to the incident (Borum et al, 1999; Fein &
Vossekeuil, 1998; Vossekuil, Reddy, & Fein, 2000). The difference between targeted
school violence and other incidents of violence that children face out side of their school
is that the perpetrator selects a target prior to the attack. Incidents of school violence are
rare, but when they happen, the consequences are devastating (Fein et al., 2002;
According to the Center for Disease Control and Prevention (2004), the frequency
of homicide and suicide for students is much lower at school than outside of school. For
1,513 homicides outside of school (Singh & Siahpush, 2006). For the 1992-2006 period,
student homicide victims ranged in age from 6 to 18, with a mean and median of 15 years
(U.S. Department of Health and Human Services, 1999). Victims
tended to be males, students in senior high schools, and students in central cities (White
& Lauristen, 2010). Notably, homicide rates were not significantly different in rural
RISK FOR TARGETED SCHOOL VIOLENCE 23
versus urban areas or public versus private schools. These findings contradict the myth
Notably, the series of school shootings in the late 1990s that ended with the 1999
According to data from the National Crime Victimization Survey (Bureau of Justice
Statistics, 2006), the annual victimization rate for students ages 12 through 18 declined
approximately 60%, from 13 violent crimes at school per 1,000 students in 1994 to 5
violent crimes per 1,000 students in 2000. Other indices in juvenile crimes throughout the
United States reflect a decline. For example, arrests of persons under age 18 for homicide
declined 74%, from 3,102 in 1994 to 806 in 2000 (Federal Bureau of Investigation,
2001).
where the school was a site of opportunity for the attack (Sandler & Alpert, 2000).
School shootings like those occurring at Columbine High School in Colorado, Pearl High
School in Mississippi, and Heath High School in Kentucky, represent a particularly rare
subset of school-related violent homicides. These incidents and the student responsible
was the case that the attacker at the school and their victims just happened to be in the
Accessibility of firearms in the United States has the highest rate among all
nations, and the highest number of deaths out of all of the 26 industrialized nations
combined. In the United States children are 12 times more likely to be killed by a firearm
than children in all the 26 nations combined (Centers for Disease Control, 2014). The
number one cause of death for children ages 1 to 19 is by a firearm. Access to weapons in
the home has been identified as a high risk factor for suicide and homicide (Miller, 2004).
More concerning is that these weapons are finding their way onto school campuses.
Between 1996 and 1997, a total of 5,724 students were expelled for bringing a gun onto
the school campus, and 3,930 students were expelled the following school year for the
same reason (U.S. Department of Education, National Center for Educational Statistics,
2000). These statistics reflect only the students who are caught. Although there appears to
be a reduction in students carrying a handgun, the FBI reports that 100,000 students carry
a gun to school each day (Federal Bureau of Investigation, 2001). The FBI report is likely
an overestimate, but most experts in criminology agree that there are many more guns on
There are many important ways in which the media contributes to the problem. One is
sensationalizing and placing too much emphasis on the perpetrators (Carnagey et al.,
2007). This may stimulate individuals who already have fantasies about committing
similar acts, thereby creating antihero identification (Carnagey et al., 2007). The
extensive information and the variety of channels available through the media increased
access to information on violence (Samuels, 2000). Access to the Internet and violent
imagery through movies, video games, and reality shows encourage already angry or
RISK FOR TARGETED SCHOOL VIOLENCE 25
Wiedman, & Bissel, 2003). The way in which violence has penetrated American culture
subtly shapes expectations and coping strategies, and it also adds to the perception that
violence is acceptable (Mohandie, 2000; Samuels, 2000; Scharrer et al., 2003). Media
troubled youth thinking about perpetrating a similar act of violence. Prior incidents of
school shootings serve as a playing field for achieving success, superiority, and notoriety.
1994). Separation, high divorce rates, failure of nuclear families, and lack of teaching and
modeling contribute to the lack of existent positive role models and values to our children
(Blume, 1996; Elliott, 1994; Mohandie, 2000). Teachers often play a major role in the
development of our next generation and as a safety net for children (Finley, 2007).
Schools and their staff members are often underpaid and understaffed, facts that make it
difficult to provide the attention necessary to each of their students (Mohandie, 2000;
Poverty may also be a contributing factor in certain cases (Gottesman & Brown,
1999). Such is the case with the incident in of Flint, Michigan, where a 6-year-old
kindergarten student used a .32 caliber Davis and a knife to kill his target. The .32 caliber
Davis is known for its small compact size that makes it easy to conceal. The 6-year-old
boy, Deidric Owens, had had problems with another 6-year-old girl, Kayla Roland, prior
to the incident. Deidric had also been previously suspended for stabbing a classmate with
a pencil. After the fatal incident police went to the 6-year-
RISK FOR TARGETED SCHOOL VIOLENCE 26
stolen guns and a cache of drugs. Deidric was left under the care of an uncle while his
mother had to go to work and was working two jobs in order to make ends meet
(Cameron, 2009; Mohandie, 2000; Moore, Petrie, Braga, & McLaughlin, 2003).
Even though statistics reflect that males have committed most of the school
shootings in this nation, females have also been implicated in committing violence,
though not in significant numbers (U.S. Department of Justice and U.S. Department of
suggests that girls were involved in more violent crime than they were in the late 1980s;
their murder arrest rate was up 64%, but violent crimes accounted for only 3.4% of girls
arrests in 1994 (Chesney-Lind & Brown, 1999). One reason that girls commit more
violent crimes may be changes in the way girls are charged. For example, a girl shoving
her parents in self-defense as she tries to run away is now likely to be arrested for
criminal assault as opposed to the lesser offense of running away (Chesney-Lind &
Shelden, 1998).
committed by males under 18. Males in this age range are two to three times more likely
to use guns, whereas females in this age range are more likely to use knives than guns
(Webster, Gainer, & Champion, 1993). Females are more likely than boys to fight with
family members, and are also more likely to murder someone as a result of this sort of
conflict. Boys are more likely to resort to murder in the process of committing another
crime (Heimer, 1995). Data suggest that girls are still less likely to be arrested for violent
crimes (i.e., homicide, forcible rape, aggravated assault) and serious property offenses
The Safe Schools Initiative study (U.S. Secret Service and United States
physical fighting, bullying, and weapon carrying are carried out by males, and the targets
are also males. Although nearly 18% of the males carry a weapon to school, only 5% of
where large numbers of males carried weapons, there is a parallel high rate of girls with
weapons, and although the boys may carry guns, the girls carry knives (Webster et al.,
1993).
class, race, ethnicity, and culture interact to cause young females to behave violently
temporarily escape a dismissal future (Campbell, 1991; Chesney-Lind & Shelden, 1996).
Indeed, young females from poor families whose college aspirations cannot be realized or
who fail to gain status through schooling or careers may engage in violent behavior to
! "d, 2004; Schiele,
2001). Young females from disadvantaged families often embrace traditional gender role
expectations for the future: these include marriage, support by a man, a large family, and
work in stereotypical female jobs. Young females believe that men should be strong and
assertive and women passive and nonviolent. Such beliefs may hold young women in
abusive relationships and raise their risk of engaging in delinquent and violent acts
delinquency is school failure. School failure affects males and females in general;
however, whereas high grades and positive self-esteem seem to decrease young
In summary, the literature described above provides vast information about the
causes of violence and the differences and similarities for males and females under 18 in
general terms. This study focuses on targeted school violence, which involves a more
specific and defined student population. What is less known is what protective factors
are required for an effective management strategy and what risk enhancers mobilize a
Longitudinal research on adolescent violent behavior has looked into risk and
protective factors as potential areas for preventive intervention (Coie et al., 1993;
Hawkins " # $%% # 1992; Hawkins, Herrenkohl,
T., Farrington, D., Brewer, D., Catalano, R., & Harachi, 1998). Risk factors are
characteristics, variables, or hazards that, if present for a given individual, make it more
likely that an individual presenting with these characteristics will consider violence as a
adolescent violent behavior. Protective factors such as community, family, and peer
support reduce the probability of problematic behavior (Hawkins et al., 1992; Rutter,
Suicidality.
Nock, Green, and Hwang (2013) found correlations between suicidality and
violence. Factors associated with violence included the presence of a recent history of
violent behavior, past history of violence, the presence or absence of supportive social
network, the motivation of the patient to be open and to participate in mental health
treatment, substance abuse, and neurological and medical illnesses. Ethological issues
suggested that aggressive impulses underlie both suicide and violent behavior and that
these aggressive impulses either reinforce or magnify suicide and violent behavior.
Homicidality.
illness, Joyal (2013) found a significant association with schizophrenia, mood disorders,
and anxiety. Several studies found that significant substance abuse and anti-social
unbearable emotions of hate, low self-esteem, humiliation, and alienation were found to
correlate to homicidal ideation. In the cases of mass killings and targeted school violence,
studies propose that mental illness in some case was a contributing factor (Hawkins et al.,
1998). Other factors include environmental factors such as high levels of stress,
skills, and a pattern for opting to violence to solve problems (Masten & Coatsworth,
1998).
RISK FOR TARGETED SCHOOL VIOLENCE 30
The Menninger Triad (Menninger, 1938), which is defined as the wish to die,
wish to kill, and wish to be killed, is a phenomenon frequently seen among individuals
who commit suicide. In other words, they provoke an outside agent to end their life.
in which a suicidal individual attempts to provoke law enforcement to kill him or her.
ate that part of the motivation for an individual to attempt to
are considered instrumental and expressive, in which the individual has the ultimate goal
!"
Access to weapons.
# $ ! % Centers
for Disease Control and Prevention, 2012; Eaton et al., 2012; Okoro, Nelson, Mercy,
Balluz, Crosby, & Mokdad, 2005). Although gun violence in schools is rare, the majority
of youth murdered in school are killed with a firearm, and nearly half of youth suicide
deaths involve the use of a gun. According to Roberts, Zhang, and Truman (2012), less
than 1% of student homicides and suicides take place at school, on the way to or from
During the 2009-2010 school year, the odds of a student 5 to 18 years of age
being the victim of a school-associated homicide was 1 in 2.5 million. In comparison, the
16,000 (Centers for Disease Control and Prevention, 2012). Further, most school-
RISK FOR TARGETED SCHOOL VIOLENCE 31
associated student homicides involve a firearm and a single victim and offender
(Modzeleski, Feucht, Rand, Hall, Simon, & Butler, 2008). In 80% of school-associated
firearm-related homicides and suicides, the weapons used were obtained from the home
In relation to the use of guns and homicidal threats, Bovasso (2014) conducted a
represented sample. One finding of the study indicated that psychotic symptoms
increased the odds of an individual reporting gun threats versus those who never
experienced psychotic symptoms (76% vs. 3.3%). The analysis also found that
threats. At the same time, the data analysis also revealed that individuals experiencing
Choice of weapon.
weapons used in school-related homicides indicate that handguns were used in 68% of
homicides, knives and other cutting objects were used in 13% of homicides, and personal
weapons such as hands, fists, and feet were used in 6% of the cases. Finally, blunt objects
such as clubs, or baseball bats were used by 4%, and strangulation or asphyxiation by 1%
of the population.
RISK FOR TARGETED SCHOOL VIOLENCE 32
Handheld weapons.
Knives are the most common handheld weapons used by children in suburban and
rural schools, and they are considered the first weapon of choice. Krouse (2012) reported
that some of the reasons knives are the first choice of weapon is that they are easy to
obtain, conceal, and wield. Baseball bats are also known to be easily accessible on school
premises (Hemenway & Miller, 2000; Hepburn & Hemenway, 2004; Kellermann, Somes,
Rivara, Lee, & Banton, 1998; Miller, Azrael, & Hemenway, 2002).
According to the Center for Disease Control and Prevention (2012), 34% of
children in the United States, representing more than 22 million children in 11 million
homes, reside in homes with at least one firearm. In 69% of the homes with children,
more than one firearm is present. In addition, the CDC study found that the probability
that families will own a firearm varies by the family's composition, income level,
race/ethnicity, region of the country, and other factors. For example, families that include
an adult male are more likely to have firearms than families with only adult women (41%
versus 12%); Caucasian families are more likely than other ethnic groups to legally own
firearms (43% versus 16 % for African Americans, 15% for Hispanics, and 16 % for all
other ethnic groups); and families in the South are more likely to have firearms than those
in other parts of the country. Rifles are the most common firearms in homes with
children, followed by shotguns, handguns, and all other types of firearms combined.
Handguns are also present in acts of violence. According to Miller et al. (2002),
children access handguns through two major sources: from gangs that are active in the
neighborhood where the child lives, or at home. Assault rifles, such as AK-47s and M4
Carbines, are two types of assault rifles used by the most notorious school shootings.
RISK FOR TARGETED SCHOOL VIOLENCE 33
Semiautomatic Glock pistols were the preferred weapons in recent mass shootings
(Modezeleski et al., 2008). Being able to conceal them is what attracts those bent on a
high body count. Glocks and other handguns are legally sold in gun shops and online.
May 2014, Isla Vista, California. Elliott Roger owned three 9mm semiautomatic
handguns, including two Sig Sauer P226 pistols and one Glock 34 pistol;
Lanza used a Glock and Sig Sauer semiautomatic pistols along with a .223-caliber
semiautomatic rifle;
July 2012, Aurora, Colorado. James Holmes used a .40-caliber Glock handgun,
April 2007, Virginia Tech University, Blacksburg, Virginia. Seung-Hui Cho used
July 2011, Oslow, Norway. Anders Behring Breivik used a Glock 17.9 mm pistol
Substance abuse.
drugs, and violence are closely related. NIDA (2003) found that 94% of teens who are
violent use alcohol, compared to 80% of all high school seniors who have tried it, and
22% who were recent users. A total of 55% of teens who are violent, or who have been
violent, use multiple illegal drugs. NIDA (2003) also found that teen drug addiction or
alcohol abuse greatly increases the risk for suicide. According to NIDA, there may be
RISK FOR TARGETED SCHOOL VIOLENCE 34
several reasons for the link between teen substance abuse and teen violence. Substance
the violent act. Teens who are using drugs may also engage in violent acts for money or
to support their substance abuse. Some teens are influenced by their environment or
personality characteristics and resort to a range of risky behaviors such as drinking, drug
abuse, or random acts of violence. Underlying mental health problems may make some
adolescents more likely to use drugs and therefore put them at risk for suicide or
violence. Overlapping risk factors between teen substance abuse and teen violence
Bullying/victimization.
! ! "
consequence of bullying may often incubate violent tendencies or trigger a retaliatory and
violent act. Bullied children are humiliated, teased, or shamed on a regular basis.
Psychologists have long known that there is a strong correlation between shame and
violence (Olweus, 2013). Bullied children also face alienation from their peers. This
(1998), tolerance to torment and humiliation can only last for so long before children who
that included 41 school shooters in 37 separate incidents, Vossekuil et al. (2002) found
that two-thirds of the shooters had been bullied by classmates and that their attacks were
motivated by a will to seek revenge. It is also important to note that the shooters did not
limit their attacks to students. Teachers and school administrators were also frequently
targeted (Vossekuil et al., 2002). Although most of these incidents involve junior high
and high school kids, it can also occur in grade school. In Flint, Michigan, a 6-year-old
male student used a handgun to shoot and kill a 6-year-old female classmate. To date, this
is the youngest child known to engage in a targeted case of school violence. The shooting
occurred in reaction to an unresolved conflict that had occurred between the two students
Emotional trauma.
Conner (1984) posited that emotional trauma has a significant negative impact on
physical, sexual, emotional, or mental abuse. Children who are physically abused and
treated violently often become violent. Children who are neglected often have difficulties
forming relationships and emotional attachment, and in turn often have difficulties
give rise to overcontrolled hostility, meaning that the individual internalizes his anger as
a coping style. Unfortunately, when this defense fails, outbursts of anger, including acts
Protective Factors
identifying risk factors. Compared to risk factors, protective factors have not been studied
Department of Health and Human Services (2001) preliminary research that identified a
list of protective factors. These include individual, family, peer, and general system
supports. The findings of the research suggest that these are important factors critical to
mitigating or eliminating violent acts. The following are examples of protective factors.
Resiliency.
helped them overcome adversity and avoid falling into a pathology paradigm. These
and situations, and the ability to select and establish more positive relationships with
others.
Children who were found to be resilient had problem solving skills that included
the ability to think abstractly and reflectively. This flexibility allowed for alternative
solutions for both cognitive and social problems (Hawkins, Lishner, & Catalano, 1985).
Furthermore, as with social competence, these skills are found in the early childhood
communities and must continuously negotiate the demands of their environment or risk
power, high self-esteem, self-efficacy, self-control, and the ability to exert some control
RISK FOR TARGETED SCHOOL VIOLENCE 37
Intelligence quotient.
suggests that low intelligence quotient (IQ) is a predictor for criminal behavior (Moffitt,
Lynam, & Silva, 1994; Ward & Tittle, 1994). A longitudinal study of high-risk Danish
involvement, whereas individuals with lower IQs tended to have longer criminal careers
targeted school violence, the majority of shooters did not have low IQ and many, in fact,
had graduated from high school and were enrolled in college or graduate school. This
finding suggests that although IQ may serve as a protective factor, it does not necessarily
prevent someone from engaging in targeted school violence. Thus, average to above
average IQ may help explain why past school shooters have successfully, albeit not
Studies have shown that clients who receive and participate in mental health
treatment experience positive behavioral changes (Miller & Rollnick, 2002). Studies have
also shown that there are four common factors that work together to create positive
change with all interventions. These findings include client factors, relationship factors,
hope and expectancy, and the model technique employed (Clark, 2001). Relationship
factors are the strength of the alliance between the client and therapist, which contributes
warmth, trust, and encouragement for self-expression are among the factors that initiate
Family dynamics.
The Department of Health and Human Services (2001) study on protective factors
within the family include the following: intolerant attitude toward deviance, ability to
discuss problems with parents, frequent and shared activities with parents, involvement in
activities, and the consistent presence of
parents. The presence of parents must occur during at least one of the following: when
awakening, when arriving home from school, at evening mealtime, or going to bed.
Community.
According to Levin (1988), the challenge for communities, as well as for schools,
is to find venues or activities that offer meaning and purpose to students in part to
connect them to their community. Whereas no studies exist on communities that provide
youth opportunities to serve and provide needed services such as academic tutoring,
literacy training, child care, or elder care, anecdotal evidence from hundreds of youth
service programs in different communities across the U.S. bear witness of the power of
this approach. Engaging youth to develop a sense of worth and belonging in their
School.
In the last decade, the literature on the ability of the educational experience to
influence the outcome for children from high-risk environments has increased (Conner,
1984). The level of caring and support within schools is as important as the family
RISK FOR TARGETED SCHOOL VIOLENCE 39
environment. According to Williamson et al. (2003), only a few studies have explored the
role of teachers as protective factors in the lives of children who overcome adversity.
Colman and Hoffer (1987) found that when schools serve as an agent for families with
the primary emphasis of caring for the child by providing attention, personal interest,
high involvement, and some level of intimacy, the child begins to develop persistence
and continuity of effort and develops the self-conception they need to succeed in school.
Peer support.
encourage teens to conform to healthy behavior. Peers can act as positive role models and
demonstrate appropriate social behaviors. Peers often listen to, accept, and understand the
frustrations, challenges, and concerns associated with being a teenager. Children and
youth are heavily influenced by opinion of their peers and often carry more weight than
literature review provides a brief look at the scope of past and current research on the
causes of violence among children and youth and although many studies have focused on
identifying early signs and predictors of violence, few have focused on the early
prevention of these, particularly in the area of targeted school violence. Targeted school
violence and threat assessment are fairly new fields that require ongoing study in order to
gain depth understanding; therefore, the current study will benefit mental health
understanding about critical factors impacting or preventing students from acting out
violently.
RISK FOR TARGETED SCHOOL VIOLENCE 41
Chapter 3
Methodology
This study examined a cohort of 55 students at risk for Targeted School Violence
(TSV). This study analyzed 16 independent variables to see what impact these variables
had on the Method for Objectively Selecting Areas of Inquiry Consistently (MOSAIC)
scores, which served as the dependent variable for this study. The 16 independent
variables consist of demographic data such as age, gender, ethnicity, educational level,
living arrangement, whether the respondent lives with family, DCFS status, autism
diagnosis, substance abuse history, the Menninger Triad score, educational ability (i.e.,
weapons in the home, bullying history, and psychiatric hospitalization history. These
variables and their coding schemes are described in a later section of this chapter. Data
analysis was achieved via the calculation of descriptive statistics, the computation of
Ritchey (2008) notes, multiple linear regression provides the ability to study the
relationship between a single dependent variable and one or more independent variables.
A Method: Stepwise algorithm was used to enter the various independent variables into
the regression equation. As Allison (1999) notes, the use of stepwise regression is to
generate a parsimonious prediction equation that only retains those predictor variables
that have statistically significant relationships with the dependent variable. Regression
was used to examine the following research questions and associated hypotheses:
RQ1. A history of violence will not be a significant factor in determining risk for
The following null and alternative hypotheses were developed from RQ1:
RQ2. Coming from a single-parent home will increase risk for targeted school violence
The following null and alternative hypotheses were developed from RQ2:
Ho2: Coming from a single-parent home will not predict MOSAIC Scores.
RQ3. Males will have higher risk for targeted school violence as demonstrated in
The following null and alternative hypotheses were developed from RQ3:
Participants/Data Sources
The final cohort of fifty-five (55) students were identified through direct referrals,
existing contacts, and community calls made to the Los Angeles County Department of
Mental Health (DMH) Emergency Outreach Bureau (EOB) School Threat Assessment
Response Team (START). Requests for START services began shortly after the
inception of the program in April 2008. Beginning in May 2008 and continuing
institutions, campus threat management teams, law enforcement agencies, and mental
health providers. Los Angeles County is divided into eight Service Planning Areas
(SPAs). START provided at least two trainings per SPA in order to maximize the number
RISK FOR TARGETED SCHOOL VIOLENCE 43
hospitals, and outpatient mental health providers attended the trainings, and as a direct
result, the situational awareness about TSV prevention and START generated numerous
as well as local and state conferences. Trainings were also held for parents and students,
and these trainings also generated referrals. As a result of these trainings and conferences,
START received over 8,000 student referrals between January 2009 and April 2014.
Referrals came from various sources. The primary source was DMH, as they
serves as an entry point to the mental health system of care in Los Angeles County.
ACCESS received requests for evaluations of students considered at risk for violence
from school personnel, school districts, colleges and universities, families, mental health
providers, campus threat management teams, local and federal law enforcement agencies,
referral, they would document the contact and forward the information to START for
triage. START may contact the referring party for further information and determine
Another referral source was the DMH EOB Psychiatric Mobile Response Team
(PMRT). PMRT responds to Los Angeles County residents who are experiencing a
mental health crisis. PMRT averages 24,000 calls annually, of which approximately 50%
involve minors. Given the nexus between suicidal/homicidal ideation and TSV (Borum,
Cornell, Modzeleski, & Jimerson, 2010; Vossekuil et al., 2004), PMRT staff received
ongoing training on TSV and violence threat risk assessment. Beginning in early 2009,
RISK FOR TARGETED SCHOOL VIOLENCE 44
PMRT began conducting preliminary violence threat risk assessments on every student
referred to them. Thus, in a given fiscal year, PMRT assessed approximately 12,000
students, and this sample was also a part of the 8,000 students forwarded to START for
triage.
known as a resource in the community for TSV assessment, intervention, and monitoring,
local and federal law enforcement agencies requested START assistance on calls
received through their 911 or other referral systems. Referrals came from the Los
ment, the Federal
Bureau of Investigation, other local law enforcement agencies, and campus police or
security firms.
Psychiatric hospitals also made direct requests for START consultation. These
Psychiatric Emergency Teams (PET) and suspected of posing a threat to their school.
educators, school counselors, campus threat management teams, and mental health
providers also made direct referrals to START for consultation including assessment or
clinical consultation and supervision, and one licensed clinical supervisor responsible for
RISK FOR TARGETED SCHOOL VIOLENCE 45
include training, supervising, and managing nine licensed clinical staff or license-eligible
staff at the Masters in Social Work (MSW) or doctoral level in psychology. START is
comprised of three psychologists, one licensed marriage and family therapist, five
licensed clinical social workers and one associate social worker. All clinicians are
Lanterman-Petris-Short (LPS) Act, Welfare and Institutions Code 5150 and 5585. LPS
the individual presents with suicidal or homicidal ideation or behavior or if the individual
is unable to provide food, clothing, or shelter due to a mental disorder. In order to obtain
LPS designation, a clinician must attend an eight-hour training and pass a fifty-question
examination. It should be noted that START members are ethnically and gender diverse
and include males or females from Asian, Hispanic, African-American, and Caucasian
descent. The number of years of clinical experience among the clinicians varies from 5 to
14 years.
START Training
Trainings include the following: a 40-hour course on Advanced Threat Assessment and
on Surviving Violent Encounters training by the Center for Personal Protection and
Bureau of Investigations (FBI) and Joint Regional Intelligence Center (JRIC). In the
recent past, leading experts in violence threat risk assessment, such as Kevin Cameron of
the Canadian Center for Threat Assessment and Trauma Response, have also provided 16
staff consult with the Los Angeles Office FBI Threat Assessment Regional Evaluation
Team (TARGET), the Joint Regional Intelligence Center, and local law enforcement
agencies. START clinicians meet every week for clinical case consultation with the
START supervisor to present new and ongoing START cases. Finally, the clinicians meet
weekly with the START supervisor and senior clinical managers for individual
Procedure/Data Collection
approach: (1) The comprehensive interview of the student and subsequent clinical
structured method of data gathering; and (4) Team staffing and consultation with senior
clinical staff.
pathway to violence were assigned to START clinicians for further assessment as stated
above. The in-depth clinical assessment included interviews with the student, parents,
teachers, therapists, and others with knowledge of the situation or the student. The data
RISK FOR TARGETED SCHOOL VIOLENCE 47
driven assessment consisted of reviewing social media postings, journals, and access to
weapons. Backpacks, lockers, and bedrooms were also reviewed with consent of the
costuming. The MOSAIC was scored and interpreted as previously outlined. These three
data sets were then presented in a team meeting to a senior clinician, a Program Head or a
Students identified as moderate to high risk were selected for inclusion in the
cohort. The remaining students were for routine outpatient mental health services when
indicated. Examples of students who were referred out included students dealing with
domestic violence, loss of a parent through divorce or separation, a death in the family,
this study represent less than 1% of the total student population that underwent a risk
Independent Variables
The age of the student was determined via direct inquiry and cross-referenced
with other available sources of information. The student was asked to state his age and
provide his date of birth. His age was then calculated using the date that the evaluation
took place as a benchmark. The age and birthday were then verified through interviews
with a parent, legal guardian, school administrator, or law enforcement officer. The final
verification included a cross-check of the age and birthday reported by the student with
reporting persons and documentation. The variable age was then broken into a five-
RISK FOR TARGETED SCHOOL VIOLENCE 48
category variable along the following coding lines: 1 = ages 5-12; 2 = ages 13-18; 3 =
school records. In cases where gender identity issues were apparent, the student's
anatomical state and physical appearance, in combination with the input of parental or
significant others, were utilized to determine gender. Transgender issues did not surface
in this population. The variable gender was then coded in the following manner: 1 = Male
and 0 = Female.
Ethnicity was established through self-report and family history. The student's
establish ethnicity. This information was cross-referenced with data present in school or
ethnic composition of Los Angeles County. However, given its skewed distribution, the
Other.
t was established by obtaining and verifying the
student's address of record. Additionally, START conducted home visits for 99% of the
cases as part of the threat assessment protocol. Categories for this variable included living
alone, living with immediate family, living with extended family, living with foster
family, living in a residential facility with six or more people, living in a residential
facility with 7 or more people, or living in a juvenile detention center. Given the nominal
level nature of this variable, the decision was made to create a series of dummy variables
RISK FOR TARGETED SCHOOL VIOLENCE 49
Whom the respondent lives with was determined as a function of the student's
identified family system. Categories for this variable included two biological parents, two
adoptive parents, adoptive great aunt and uncle, adoptive mom, alone, in DCFS custody,
father only, girlfriend, a grandparent(s), a mom and stepdad, a single mom, or a student
host mother. Given the nominal level nature of this variable, the decision was made to
create a series of dummy variables to encapsulate all possible choices for this variable.
Services (DCFS) status was determined by self-report and cross-checked with a DCFS
database. Categories for this variable included active, closed, pending, or inactive cases.
Children with a DCFS active case had physical abuse, sexual abuse, emotional abuse, or
neglect. The coding scheme for this variable was as follows: Yes = 1 or No = 0.
s, and 2 for Autism. Educational ability
and coded 1), average or grade level student (no special education or advanced
placement, coded as 2), and gifted (honor roll, advanced placement, or exceptional
capabilities as per the educational record coded as 3). Substance abuse was defined as the
persistent or severe use of drugs or alcohol; this variable was coded as 1 for yes and 0 for
no. A history of violence was categorized as either yes or no, with yes responses which
were coded as (1) indicating that the individual had a history of attacking others with
lethal means, inflicting serious injury on another person, or having maliciously destroyed
RISK FOR TARGETED SCHOOL VIOLENCE 50
variable was coded as 0. Weaponry took into account what weapons the student
personally owned or had ready access to a weapon; this variable was coded as 1 for yes
and 0 for no. Bullying was measured by self-report, school observations, and the
chronic emotional distress and torment inflicted by others, which may or may not include
physical tormenting such as teasing, ridiculing, pushing, tripping, or hitting. This variable
was coded as 1 = Bully perpetrator, 2 = Bully Victim, and 3 = Both. The default category
of none was coded as 0. A history of hospitalization took into account whether or not the
respondent was hospitalized. This variable was coded as 1 for yes and 0 for no. Finally,
clinical diagnosis was coded as one of six different diagnoses: these included adjustment
mood diagnosis, and psychotic diagnosis. Given the nominal level nature of this variable,
the decision was made to create a series of dummy variables to encapsulate all possible
Several other variables were also collected as part of the dataset. For example,
Menninger first wrote about these homicidal and suicidal desires more than 70 years ago
(Menninger, 1938). The concept was used to describe Eric Harris and Dylan Klebold, the
adolescents responsible for the Columbine High School shooting in 1999 as described in
! " # $%%%&
Menninger Triad, and their corresponding numeric codes, were assigned as follows: 1 =
Wish to die; 2 = Wish to kill; 3 = Wish to be killed; 4 = Wish to die and Wish to kill; 5 =
RISK FOR TARGETED SCHOOL VIOLENCE 51
Wish to die and Wish to be killed; 6 = Wish to kill and wish to be killed; 7 = Wish to die,
wish to kill and wish to be killed. Given the nominal level nature of this variable, the
decision was made to create a series of dummy variables to encapsulate all possible
Dependent Variables
The level of risk was identified by the MOSAIC protocol.. MOSIAC stands for
Method for Objectively Selecting Areas of Inquiry Consistently for the initial six areas of
inquiry: Menace, Object, Subject, Intensity, Affect, and Clarity for Assessment of
Student Threats. The Method for Assessment of Student Threats (MAST) -
assisted assessment method that organizes and expresses research and expert opinion in
a quality score with a maximum value of 200 that is an indication of whether there is
enough information about the case for MOSIAC to compare it to other known cases. The
authors of MAST, as well as other risk assessment experts, caution that the MAST score
weekly staffing meeting, where all of the data were discussed and presented to a senior
clinician. In order to optimize the regression equations, MOSAIC scores were modeled
with two dependent variables. The first measures MOSAIC: Assessment scores (which
ranged from a low score of 4 to a high score of 8), whereas the second measured
MOSAIC: Quality scores, which ranged from a low score of 116 to a high score of 200.
Data for this study were collected using the MOSAIC MAST protocol developed
RISK FOR TARGETED SCHOOL VIOLENCE 52
by Gavin De Becker and Associates. The MOSAIC was not developed in the traditional
research approach and is not founded on sample size, reliability, or validity studies. It is
not a predictive tool, but rather a combination of consensus and researched based
principals. Experts, law enforcement officers, mental health practitioners, victims and
instrument used by experts and other professionals in conducting violence threat risk
assessments. The MOSAIC has been utilized in several studies. For example, the U.S.
Department of Justice Intimate Partner Risk Assessment Validation Study, Final Report
predictability in diagnosing risk and lethal violence and extreme dangerousness. DV-
MOSAIC was used in this study to assess domestic violence cases in terms of the
likelihood of escalation, including homicide. The study concluded that the DV-MOSAIC,
when compared to other instruments, scored highest in correctly classifying women that
were ultimately re-assaulted. Conversely, women with lower scores were not
Other instruments that were considered but not selected include the Workplace
and Campus Violence Risk (WAVR-21) protocol (Meloy & White, 2007), HCR-20
Violence Risk Assessment Scheme (Douglas, Guy, & Reeves, 2008), Guidelines for
Responding to Student Threats (Cornell & Sheras, 2006), and Risk Assessment Guideline
The WAVR-21 was not selected for this study because the author noted that the
protocol focuses on work place violence, intimate partner relationships and stalking but
RISK FOR TARGETED SCHOOL VIOLENCE 53
not students in a campus setting. The HCR-20 is a general violence risk assessment
protocol not specific to campus violence. The study sample consisted of adult patients
residing in a short-term inpatient unit. The Guidelines for Responding to Student Threats
was considered but not selected because it provides guidelines without identifying levels
of risk. Similarly, the RAGE-V focuses primarily on inquiry and does not provide risk
assessment ratings.
Data Analysis
This study analyzed the collected data using descriptive statistics, bivariate
correlation, and multiple linear regression. Descriptive statistics are those statistics that
provide detail on the basic patterns within the data (Ritchey, 2008). For continuous data,
means and standard deviations were calculated; for categorical data, percentages and
frequencies were calculated. Percentages and frequencies were calculated for the
MOSAIC scores and 15 of the 16 independent variables listed above. Only the variable of
education level was analyzed via the computation of a mean and standard deviation, as
utilized. Multiple linear regression was selected over simple linear regression because
the latter analyzes only one independent variable, whereas the former analyzes two or
commonly used in data analysis to establish whether or not there exists a relationship
between a dependent variable and one or more independent variables (Freedman, 2005;
Ritchey, 2008). For this study, multiple linear regression is an appropriate statistical
method because there are 16 independent variables and a single dependent variable (i.e.,
RISK FOR TARGETED SCHOOL VIOLENCE 54
Regression analysis was used to predict which of the independent variables are
statistically related to the dependent variable, and to articulate the forms of these
recommended as the correlation between the dependent variables and the independent
variables does not imply causation (Armstrong, 2012). In this investigation, multiple
Ethical Considerations
Ethical considerations to protect the privacy of the information collected and used
for this study were consistent with The American Psychological Association (2010)
submitting a research proposal prior to conducting all data analysis. Because the data
collected are archival, no informed consent was required. However, participants provided
verbal and written consent to the evaluating clinician at the time of the assessment.
tables, and grids used to analyze the data did not contain any identifiable data. For this
study, face-to-face interviews were not conducted. Nevertheless, the researcher ensured
that throughout the course of the research, there were no interruptions of mental health
The information used for this study remained confidential and stored in a locked
location for the entire time of the research and for a period of 3 years after the conclusion
RISK FOR TARGETED SCHOOL VIOLENCE 55
of the study. Lastly, the researcher took the necessary steps to protect the dignity and
welfare of all participants. Thus, the subjects were advised that they could withdraw their
consent at any time. The researcher was available by email or telephone to answer
Reflexivity
My personal interest in this topic stems from three factors. The first one is my
personal experience. I was born and raised in an impoverished and at times unforgiving
I was born in Mexico and raised by my mother who had a high school education. My
mother could not provide for my brother and two sisters and decided to come to the
United States when I was 6 years old. Although I was only 6 years old, I can still
remember how difficult it was for me to leave Mexico and adjust to a new culture and
environment. We grew up in South Central Los Angeles, where there were Hispanic and
African-American gangs, poverty, and limited resources. The few parks that were
available were usually occupied by gangs, drug abusers, and prostitutes. My mother
struggled to provide housing. There were times when we lived in cramped quarters, and
my mother could not afford to buy us new clothes. Gifts were rare for Christmas and
birthdays. Fortunately, I was able to use those experiences to get ahead. Unfortunately,
most of the subjects in this study have not been as successful. Their hopelessness and
despair remind me of the times in my life when I also felt there was no hope. At work,
part of our intervention is to provide hope and inspiration for the subjects.
The second factor is that my personal growth and development were due in large
part to the individuals in my life who took the time to listen and guide me. This is another
RISK FOR TARGETED SCHOOL VIOLENCE 56
level of intervention that I practice with the individuals in the sample. Role models were
critical for me and appear to be equally important for the individuals who are in this
program, which sends students from the inner city to suburban communities. I attended
middle and high school 30 miles from my house in an upper middle-class community.
This exposure offered me hope for the future. The students in this study either have not
had the opportunity to see success or believe it is not something they can achieve. I
provide a role model for them and try to give them hope for the future. I can identify with
the children and youth who are referred to the START program. Many of them lack
positive role models, resources, and family support and see themselves with little hope
for the future and few opportunities for success. I have been fortunate to see some of the
students who were outreached and impacted positively. They have stayed in school and
channeled their energies in a positive and constructive manner. Several in this study are
Worker for the School Threat Assessment Response Team (START), County of Los
Angeles, Department of Mental Health, Emergency Outreach Bureau. The goal and
identifying students on a pathway to violence and providing these students with the
resources required to stabilize and complete their education. As an early intervention and
parents, and mental health professionals to create situational awareness about targeted
school violence, violence threat risk assessments, and appropriate intervention strategies.
RISK FOR TARGETED SCHOOL VIOLENCE 57
Intensive case management and long term monitoring are utilized to mitigate or eliminate
the threat.
RISK FOR TARGETED SCHOOL VIOLENCE 58
Chapter 4
Results
The quantitative data were collected from the review of clinical archives from a
community mental health agency. The subjects of this study are comprised of students
metropolitan city and county. The cohort of 55 students was identified from a pool of
over 8,000 students evaluated over a 5-year period for danger to self or others and
comprehensive violence threat risk assessment process that included clinical and
interview protocol.
This section begins with an analysis of the null hypotheses, followed by a section
containing the findings on 19 variables and their correlation to MOSAIC scores. The
Associates in the early 1980s to assess and evaluate verbal and media or other generated
threats. The MOSAIC stands for Method for Method for Objectively Selecting Areas of
Inquiry Consistently. The MOSAIC consists of 47 items used to assess risk, intent, and
lethality. The items are unique for different situations, including threats and fear in the
workplace, threats by students, threats against judges, threats against public figures and
public officials, and spousal abuse situations. The enhanced MOSAIC method is used by
the U.S. Supreme Court Police to assess threats to the Justices, the U.S. Marshals Service
for screening threats to judicial officials, the U.S. Capitol Police for threats against
Members of Congress, multiple police agencies protecting the governors of eleven states,
RISK FOR TARGETED SCHOOL VIOLENCE 59
many large corporations, several universities, and the Los Angeles County Department of
The MOSAIC used for this study is the Method for Assessment of Student
Threats (MAST), which blends research and expert opinion in threat analysis. MOSAIC
scores produced from the MAST are an overall single-item MOSAIC score for
Assessment, where the single-item measurement metric ranges from a low score of 4 to a
high score of 8, and an overall single-item MOSAIC score for Quality, where the single-
item measurement metric ranges from a low score of 116 to a high score of 200.
The specific research questions, and the associated hypotheses, that were
RQ1. A history of violence will not be a significant factor in determining risk for
The following null and alternative hypotheses were developed from RQ1:
and MOSAIC score: Assessment scores. This relationship was also found as part of a
multiple linear regression. On the basis of the evidence, there is support for Research
RQ2. Coming from a single-parent home will increase risk for targeted school violence
Ho2: Coming from a single-parent home will not predict MOSAIC Scores.
RISK FOR TARGETED SCHOOL VIOLENCE 60
A statistically significant bivariate correlation was found between living with a single
parent and MOSAIC score: Quality scores. This relationship was also found as part of a
multiple linear regression. On the basis of the evidence, there is support for Research
RQ3. Males will have higher risk for targeted school violence as demonstrated in
No relationship was found between being male and either MOSAIC score: Assessment
scores or MOSAIC Scores: Quality scores. On the basis of the evidence, there is no
support for Research Question 3. The statistical decision is to fail to reject the null
hypothesis.
Descriptive Statistics
classified as category 1 and 2 on the Menninger Triad (i.e., wish to die and wish to be
killed). Nearly half of the respondent (49.1%) were Hispanic. Among respondents, the
majority (56%) had a mood disorder diagnosis. Only one in seven respondents (16.4 %)
had a substance abuse problem. Slightly more than half of respondents (56.4%) had
weapons in the home. The majority of respondents (61.8%) noted that they were victims
of bullying. Three out of every five respondents (60.0%) were of average educational
ability. Eight out of every ten respondents (83.6%) have been in a psychiatric hospital.
The majority of respondents (78.2%) do not have an Autism Spectrum Disorder. The
RISK FOR TARGETED SCHOOL VIOLENCE 61
majority of respondents (63.6%) live with immediate family members. With respect to
living arrangements, 21.8% live with a single mom. Three out of every ten respondents
(30.9%) have an open case with the Los Angeles County Department of Children and
Family Services (DCFS). Nine out of every ten respondents (89.1%) were male.
Table 2
Means and Standard Deviations, Study Variables
Ritchey (2008) noted that for continuous variables, means and standard deviations
are the appropriate descriptive statistics to utilize. Each mean and standard deviation can
be interpreted as a function of the measurement metric. For example, for the MOSAIC
score: Assessment measure, the measurement metric ranges from a low score of 4 to a
high score of 8, with 6 as the midpoint. For the MOSAIC score: Quality measure, the
measurement metric ranges from a low score of 116 to a high score of 200, with 158 as
the midpoint. For education level, the mean can be interpreted as the grade level of a
student. The average MOSAIC score: Assessment is over the midpoint of the scale (M =
6.84). The average MOSAIC score: Quality is under the midpoint of the scale (M =
statistically significant relationships exist at the bivariate level between the dependent
RISK FOR TARGETED SCHOOL VIOLENCE 62
variables of MOSAIC Score: Assessment and MOSAIC Score: Quality and all of the
the Appendix have been flagged within the correlation table in the following manner: A
single star (*) denotes a significant correlation at the p =. 05 alpha level. A double star
(**) denotes a significant correlation at the p = .01 alpha level. No stars indicates that the
correlation is not statistically significant, and that there is no relationship among the two
variables in question.
Among the correlations within Table 2 in the Appendix, the following were of
interest. First, MOSAIC Score: assessment was positively correlated with a history of
violence (r = 0.286, p < .05). This result suggests that respondents with a history of
violence will obtain higher MOSAIC Score: assessment scores. MOSAIC Score:
assessment was negatively correlated with living with both biological parents (r = -0.271,
p < .05). This result suggests that respondents who live with both biological parents will
have lower MOSAIC Score: assessment scores. MOSAIC Score: quality was negatively
correlated with being a perpetrator of bullying (r = -0.360, p < .01). This result suggests
that respondents who were perpetrators of bullying will have lower MOSAIC Score:
quality scores. MOSAIC Score: quality was positively correlated with living with a
single mother (r = 0.422, p < .01). This result suggests that respondents who live with a
As Ritchey (2008) notes, multiple linear regression is the correct method to use
when one has a single dependent variable that is continuous in nature and multiple
These conditions are met in the current scenario. A Method: Stepwise algorithm was
used to enter the various independent variables into the regression equation. As Allison
equation that only those predictor variables that have statistically significant relationships
with the dependent variable. In other words, the stepwise algorithm requires that all
algorithm can consider which variables to retain and which to reject from the regression
where
Table 3a
Multiple Linear Regression of MOSAIC Score: Assessment onto the
Independent Predictors
Variable B SE(B) p
Constant 6.538 0.189 0.000
History of violence 0.565 0.260 0.034
N 55
F 4.376 0.034
2
R 0.082
RISK FOR TARGETED SCHOOL VIOLENCE 64
Table 3b
Multiple Linear Regression of MOSAIC Score: Quality onto the
Independent Predictors
Variable B SE(B) p
Constant 155.206 2.730 0.000
Respondent lives with a single mom 18.572 5.698 0.002
Respondent is a bullying perpetrator -18.777 6.220 0.004
Respondent lives with a student host mother -39.206 16.337 0.020
N 55
F 9.134 0.000
R2 0.363
In the first regression equation, the MOSAIC score: Assessment was used as the
dependent variable. A total of 45 independent variables were entered into the regression
equation as per the stepwise algorithm procedure outlined by Allison (1999). A single
variable was retained as statistically significant predictors of the dependent variable. The
decomposition of effects within the regression model can proceed. The coefficient of
determination, also known as the R2 value, is .082. This value shows that 8.2% of the
variables in the equation. Thus it is the case that 91.8% of the variance in the equation is
due to error. The equation found that a history of violence (B = 0.565, p = .034) is a
As Ritchey (2008) notes, multiple linear regression is the correct method to use
when one has a single dependent variable that is continuous in nature and multiple
These conditions were met in the current scenario. A Method: Stepwise algorithm was
used to enter the various independent variables into the regression equation. As Allison
equation that only those predictor variables that have statistically significant relationships
with the dependent variable. In other words, the stepwise algorithm requires that all
algorithm can consider which variables to retain and which to reject from the regression
where
In the above regression equation, the MOSAIC score: Quality was used as the
dependent variable. A total of 45 independent variables were entered into the regression
equation as per the stepwise algorithm procedure outlined by Allison (1999). Only three
Therefore, decomposition of effects within the regression model can proceed. The
RISK FOR TARGETED SCHOOL VIOLENCE 66
coefficient of determination, also known as the R2 value, is .363. This value shows that
36.3% of the variation in MOSAIC Score: Quality can be explained by the independent
variable in the equation. Thus it is the case that 63.7% of the variance in the equation is
due to error. The above equation found that living with a single mom (B = 18.572, p =
.002) is a positive predictor of MOSAIC Score: Quality. In other words, living with a
single mother increases MOSAIC Score: Quality scores. The above equation also found
that being a bullying perpetrator (B = -18.777, p = .004) and living with a student foster
other words, living with a single foster mother and being a perpetrator of bullying both
Chapter 5
Discussion
In 2015 there were a total of 45 school shootings in the country (World Health
Organization, 2015). According to the Center for Disease Control and Prevention (2008),
between the years of 2003 to 2010, a total of 116 school-associated homicides occurred
among students (an average annual homicide rate of 0.03 per 100,000 students) that were
associated with 109 events. Approximately 78% of these deaths occurred on a school
educators, mental health providers, and law enforcement. Much of the recent published
literature addressing TSV appears to focus on factors and characteristics of actual school
shooters rather then students presenting risk factors for potentially engaging in TSV.
This study examined the risk factors present in a cohort of 55 students at risk for TSV
and compared this cohort to students who perpetrated school violence. The findings in
this research revealed similarities and differences with the Safe School Initiatives
gender, ethnicity, educational level, living arrangement, family unit, DCFS status, Autism
Spectrum Disorder status, substance abuse history, Menninger Triad score, educational
ability (i.e., special education, special learning disability, or gifted), clinical diagnosis,
access to weapons, bullying history, and psychiatric hospitalization history. The specific
hypotheses analyzed in this study were as follows: (1) whether or not a history of
relationship between elevated MOSAIC SCORES and history of violence; (2) whether or
not single parent status places a student more at risk for TSV; and (3) the relationship
The Final Report and Findings of the Safe School Initiative: Implications for the
Prevention of School Attacks in the United States (U.S. Secret Service and United States
Department of Education, 2002) report reviewed the case studies of 37 targeted school
attacks over the past 25 years and found that school shooters were exclusively male, that
guns were nearly always the weapon of choice, that most attackers ranged in age from 13
to 18, that 75% of the attackers were Caucasian, that 66% of the attackers came from
two-parent families, and that few were in foster care or living with a legal guardian.
Furthermore, the SSI report suggested that the attackers performed well academically,
were considered mainstream students, and 66% had never exhibited behavioral problems
or had legal problems. Almost 75% of the attackers were victimized by identified
bullying behavior. Prior mental health diagnoses and substance abuse were not found to
be prominent risk factors, but a history of depression and desperation was present. The
SSI report also found that over 59% of the attackers had an interest in violence via
movies, video gaming, and publications. Finally, most attackers did not have a prior
Similarities between the current investigation and the SSI study include the fact
that all of the students in the SSI study did not have a prior history of violence, were not
involved in the criminal justice system, did not abuse substances, and were victimized by
bullying. Differences between the current project and the SSI study were that six out of
55 students in this study were female, the age of the student population ranged from 9 to
RISK FOR TARGETED SCHOOL VIOLENCE 69
39 (as compared to 13 to 18 in the Safe School Initiative study; see Vossekuil et al.,
2004), and only 51% in the current study were Caucasian. Unique results to the current
study included the finding that living with a single mother increased the MOSAIC
Quality Score, and that being a bully and living with a foster mother decreased or
History of Violence
factor in identifying students at risk for TSV. This is a significant contribution to the
risk assessment should include a detailed investigation about prior acts of violence.
Additionally, those students with a history of violence obtained a moderate to high score
(6 to 8) on the MOSAIC. This finding highlighted that a high score on the MOSIAC
was found between a history of violence and the MOSAIC Assessment score. This
relationship was also found as part of a multiple linear regression. Huesmann, Moise-
Titus, Podolski, and Eron (2003) suggested that early onset of violence and delinquency
is associated with more serious and chronic violence later in life. However, it is important
to note that specific correlations between a history of violence and TSV have not been
found in actual incidents of school shootings. Indeed, The Final Report and Findings of
the Safe School Initiative: Implications for Prevention of School Attacks in the United
State (U.S. Secret Service and United States Department of Education, 2002) found that
Family Composition
and therefore be at risk for TSV. A statistically significant bivariate correlation was found
between living with a single parent and MOSAIC Quality score. This relationship was
also found in a multiple linear regression. Posner and Vandell (1994) found that a high
Fathers typically offer economic stability, provide positive role models for boys, reduce
stress for mothers, and provide household security (Blume, 1996). Additional studies
(e.g., Chen & Weikart, 2008) indicate that contemporary societal demands in which both
parents work make it difficult for parents to provide adequate supervision and meet the
emotional needs of their children. Strong parent-child relationships can potentially serve
more children today are being raised by single parents, including adolescent mothers, and
some children are potentially subject to neglect, physical or sexual, and substance abuse
Living with a foster mother was found to lower the MOSAIC quality score in the
current investigation. This finding suggest that placement in a foster home may serve as a
protective factor because it may help to remove the child from an abusive environment.
This is consistent with studies that show that children living in abusive environments or
who have witnessed abuse are prone to depression, learn to solve problems with violent
RISK FOR TARGETED SCHOOL VIOLENCE 71
behavior, tend to abuse substances, feel hopeless, and ultimately think about suicide (Oh,
Park, & Choi, 2008). This finding is also consistent with the results of the SSI study in
that foster care status is not necessarily linked to TSV. That said, further research in this
area is recommended, as no locatable studies address the relationship between TSV and
Gender
The current study found that gender is not a factor in targeted school violence.
MOSAIC scores were not statistically significant based on gender, a result suggesting
that gender is not a discriminating factor. Typically males have been considered the more
aggressive sex (Bukowski, Sippiola, & Newcomb, 2000). In hundreds of studies, research
on aggression has found that as a group, boys exhibit significantly higher levels of
aggression than girls (Coie & Dodge, 1998). Historically, many studies on aggression
have excluded girls from their sample, particularly in studies of targeted school violence
(Bukowski et al., 2000). For example, Leary, Kowalski, Smith, and Philips (2003) found
that the majority of school shooters are male. As a result of this observation, researchers
have sought to link gender to the construct of school shootings, highlighting that men are
more likely to act violently as compared to women. Findings in the current research
potential. This tendency is especially evident when one considers that in recent years
there have been incidents of active shooting events in which females have taken part. For
example, the recent mass shooting incident in San Bernardino California featured a 29-
year-old female, Tashfeen Malik, who was one of the two shooters (Lewis, 2015).
RISK FOR TARGETED SCHOOL VIOLENCE 72
Bullying
Being a victim of bullying increases MOSAIC scores in both sexes. This finding
aligns with prior research findings that victimization may be predictive of targeted school
violence. For example, a study by Vossekuil et al. (2004) of 41 school shooting events
between 1974 and 2000 found that one commonality among the shooters was that they
were victims of bullying. Psychological and emotional factors such as anger, low self-
esteem, anxiety, and depression are common to victims of bullying. A victim may be
more prone to repress anger over time and react violently later (Bosworth, Espelage, &
Simon, 1999). Depression has been found to be a common mental health symptom
experienced by male and female victims of bullying (Callagan & Joseph, 1995; Kaltiala-
Heino, Rimpela, Marttunen, Rimpela, & Rantanen, 1999; Neary & Joseph, 1994).
Furthermore, Craig (1998) found higher depression levels for girls in comparison to boys
bullying. Clinically elevated depression levels have been found for both male and female
students who bully their peers (Austin & Joseph, 1996; Slee, 1994). Bully-victims, which
are those students who bully and have been bullied, have also been found to have higher
rates of depression than bullies (Austin & Joseph, 1996). In other studies bully-victims
report higher depression levels than their victims (Swearer & Doll, 2001).
Anxiety is also a salient mental health concern for bullies, victims, and bully-
victims. There is a paucity of research conducted specifically on anxiety and bullying; the
research that is available has yielded inconsistent findings. Some studies find that victims
of bullying have higher rates of anxiety than bullies (Craig, 1998), whereas other studies
find that bullies and victims report similar levels of anxiety (Duncan, 1999). The
RISK FOR TARGETED SCHOOL VIOLENCE 73
relationship between anxiety and TSV within the context of bullying has not been fully
It can also be argued that perpetrators of bullying had low MOSAIC scores, in
part because they did not have access to weapons or a history of substance abuse. Most
school shooters tend to have narcissistic personality features and fragile identities
(Loeber, 1982). School shooters try to negate their hidden vulnerability by presenting
themselves in a manner that suggests confidence and mastery of their environment. The
similar vein, bullies seek to gain notoriety among their peers by targeting vulnerable
students. Once an unequal relationship between perpetrator and victim is established, the
submissive, helpless, or fearful behavior of the victim reinforces the negative behavior of
the perpetrator. In other words, the signs of pain and submission signal successful
domination and control for the perpetrator. Thus, mechanisms on the dyadic level as well
as on the whole group become important, given that bullies often single out victims who
The study of TSV is still in its early stages. The Final Report and Findings of the
Safe School Initiative: Implications for the Prevention of School Attacks in the United
States (U.S. Secret Service and United States Department of Education, 2002) was the
first comprehensive study of TSV and the first attempt to identify and describe critical
characteristics of school shooters. The current study is one of the first studies that
attempted to build on the findings of the SSI study by comparing SSI results with the
cohort of 55 students described in this research project. To this end, the current project
RISK FOR TARGETED SCHOOL VIOLENCE 74
described a cohort of 55 students at risk for Targeted School Violence (TSV) and
compared them to the cohort of 37 students who committed targeted school violence and
were described in the May 2002 United States Secret Service Department of Education
Study. The current study investigated whether 19 independent variables impact the
which served as the dependent variable. The data used in the current study were not
randomly selected; as such, it is important to note that there are limitations associated
with the generalization of the current findings. Even though there are limitations
associated with the current study, it is my hope that this study generates further research
focusing on early identification of students at risk for TSV, and most importantly, in
counselors, I valued the guidance and support I received, which enabled me to further my
the challenges and struggles similar to those faced by the students in this study.
Replacing risk with protective factors minimizes vulnerability and provides hope for
This study was unable to determine whether the findings apply to students not
RISK FOR TARGETED SCHOOL VIOLENCE 75
considered at risk for targeted school violence. The use of a control group would allow
for an analysis of the similarities or differences between students posing a risk and the
general student population. Indeed, a control group would perhaps better assist in
commonalities and differences might yield valuable data in early detection of troubled
youth. Additional scholarship should also focus on gender-related issues and should
include an analysis of motives that may incite females to participate in targeted school
violence. Examining gender and mental health status and gender specific issues should be
included in the assessment for TSV, as educating parents, teachers, and school officials
about early signs and risk factors need to be part of early intervention and prevention
efforts.
Extensive research has been done on school shootings; this research has produced
have focused on the aftermath of the shooting rather than preventive measures. The
current study attempted to remediate this gap in the knowledge base by identifying risk
factors and examining differences and similarities between students who perpetrated TSV
and students at risk of TSV. Ultimately, this project sought to provide data that would
The cohort of students used in this study is not a representative sample of all
students in Los Angeles County and therefore does not easily allow for generalization of
the data findings to the larger student population in Los Angeles County or the country at
RISK FOR TARGETED SCHOOL VIOLENCE 76
validity refers to the findings component of a study and if the findings of such could be
generalized to other populations (Heppner, Kivlinghan, & Wampold, 1999). The sample
of students for this study was not randomly selected. The lack of randomization does not
allow for external validity, and should therefore be considered a limitation of the current
study.
There were also significant differences between the SSI sample and this cohort.
The SSI sample of respondents actually committed acts of TSV, whereas the sample used
in the current study did not. Further research should also include a homogeneous rather
than heterogeneous sample with respect to current key demographic characteristics. For
example, the SSI sample age range was from 12 to 18 years of age, whereas the age range
for the 55 students analyzed in this study was from 8 to 39 years old. The SSI sample was
confined to people who were in a K-12 educational environment, whereas the current
study included elementary school through college students. Research controlling for age
and grade level could have allowed for better comparison between the groups. The aspect
of weaponry is another area that warrants further assessment. For example, is there a
specific student characteristic that predisposes the student towards a specific weapon? Is
variable limited a more robust data analysis because I did not design the MOSAIC, and as
such, more specific questions or areas of inquiry were not possible. For example,
questions about the quality of the student-parent, and more specifically, student-father
relationship were not a part of the MOSAIC. Cultural issues were not addressed by the
RISK FOR TARGETED SCHOOL VIOLENCE 77
MOSAIC, which is significant omission because a large percentage of the cohort were
not Caucasian. Performance anxiety was also not a part of MOSAIC, even though this
students at significant risk for TSV. The current study is one of the first (and may be the
high risk for TSV. The fact that this sample is considered by subject matter experts as a
cohort of students who fell just short of carrying out their violent attack makes for a
unique sample difficult to find in other parts of the country. This is also one of the only
studies that also identified clinical, social, situational, and environmental factors to
consider in the assessment of students at risk for TSV. For example, clinical diagnoses,
positive role models, and a positive home environment serve as protective or risk factors
for TSV, depending on the analysis performed. Ideally these and other factors will be
addressed in future investigations. Nevertheless, I hope that the current study will inspire
future researchers to expand on risk and protective factors as well as the other variables
previously noted.
Recommendations
The findings in this study suggest that prevention of TSV is greatly improved
through partnerships that include law enforcement, education, and field based,
specialized programs such as the School Threat Assessment and Response Team
bullying/victimization and the Menninger Triad was also reviewed and suggests that
high-risk students are more likely to endorse homicidal (52%) rather than suicidal (27%)
ideation. Mentorship and positive male role models seemed to be effective in minimizing
risk for students from single-parent families, as they serve as a protective factor.
enforcement, educators, and mental health providers. START is a unique program that
uses a behaviorally driven violence threat risk assessment process to identify behaviors
that are typically described as pre-incident indicators of targeted school violence. In this
study, bullying, gender, a history of violence, and family composition were found to
variables unique to school shooters. This study suggests that younger age children
(beginning at age 8) may be at risk for considering targeted school violence as a remedy
to their problems. Prevention efforts should therefore focus on children exhibiting early
warning signs of TSV. Bullying may have serious consequences for the perpetrator,
are not addressed appropriately. Prevention should therefore include teacher education
and situational awareness, anger management programs for the bully, and parenting
students lacking in this area, and a regimen of healthy play and regular exercise activities
may also be of utility when seeking to head off TSV. Finally, schools should endorse
RISK FOR TARGETED SCHOOL VIOLENCE 79
Gender was not found to be significantly related to TSV in the current study. This
finding suggests that risk assessment specialists should not focus exclusive on males and
should recognize that females are also at risk for targeted school violence. Because TSV
may not be gender specific, prevention programs and campus threat management teams
should not dismiss females as potential perpetrators. Future studies may want to examine
as a remedy to their problems. The current study suggests that family composition serves
a protective factor in that a two-parent family lowers the risk of TSV. Furthermore,
family composition with respect to the presence of a foster parent was found to serve as a
protective factor by virtue of extricating the student from a severely dysfunctional home
An important area for further inquiry is the exploration of the unconscious and the
internal roots of violence. Carl Jung (1912/1916) emphasized the significance of what he
viewed as a spilt-off and acknowledged the dark side of human nature, or what he called
the shadow. Jung referred to the shadow as an unconscious aspect of the personality that
the conscious ego does not recognize in itself. Jung added that one tends to reject or
remain ignorant of the least desirable aspects of personality. The shadow is largely
negative, or the entirety of the unconscious: it is everything of which a person is not fully
conscious. Jung stated that there are positive aspects which may also remain hidden in
RISK FOR TARGETED SCHOOL VIOLENCE 80
one's shadow. For example, people with low self-esteem are typically not conscious of
the shadow but display or act it out through behaviors and interpersonal relationships.
The shadow can include everything outside the light of consciousness and may be
1938
embodied in the individual's conscious life, the blacker (p. 93). It may be
that are repressed during early
mass, and there emerges a raging monster; and each individual is only one
tiny cell in the monster's body, so that for better or worse he must
Having a dark suspicion of these grim possibilities, man turns a blind eye
In trying to understand the morphosis of violent behavior, we need to understand the dark
shadow we all carry inside of us. Modern American civilization may be naïve in regard to
the dark side of human nature, as individuals seek to reason about what causes someone
RISK FOR TARGETED SCHOOL VIOLENCE 81
to kill others. Individuals and communities often search for external reason such as guns
or poverty as a cause, but we fail to look beyond external conditions. Indeed, violence has
existed from the beginning of history. Biblical stories narrate the practice of sacrifices
and bestowing that were seen as gifts of acts of leadership (Coogan, 2008). Sacrificing a
son or daughter was rewarded and seen as an act of loyalty to the gods. However,
psychology, also emphasized that the cultural tendency to split off violence and project it
toward vulnerable communities and individuals is another form of collective shadow that
problematically influences not only the entire society but especially those most affected
by exposure to violence.
psychological strategy in which the player obtains points and rewards for killing as many
people in a game. Acts of terrorism and suicide bombers are promised acceptance,
heaven, and feminine eternal companionship (Lewis, 2015). Perhaps further research
about the absence of conscious and morality and violent behavior can shed light for
clinicians working with students at risk for TSV. As Freud (1923) wrote, the root of
Conclusion
deceased, and others traumatized by the violence are left with lifelong impressions and
reactions. Instances of TSV have not decreased and continue to be an area of concern. In
RISK FOR TARGETED SCHOOL VIOLENCE 82
a recent tactical guide, the Federal Bureau of Investigation (2014) documented the fact
that the frequency of these events has doubled in the past six years. The increasing
number and frequency of TSV events suggests that further research is essential if we are
to improve our understanding of this phenomenon and develop more precise assessment
In general, the findings of this study support the published reports indicating that,
although there is no accurate profile for the perpetrators of these events, there may be
factors that, when grouped together, increase the risk or likelihood of violence. Careful
assessment and monitoring can lead to safe interventions and therefore eliminate TSV.
! " # $%
where students were receiving services from START. It is my sincere hope that this
study raises awareness of not only TSV, but also of the effectiveness of prevention
References
Aitken, S. (2001). Schoolyard shootings: Racism, sexism, and moral panics over teen
from http://www.apa.org./ethics/
Anderson, M., Kaufman, J., Simon, T. R., Barrios, L., Paulozzi, L., Ryan, G., . . . School-
Blumstein, A. (2000). Youth violence, guns and the illicit-drug industry. Journal of
Borum, R., Cornell, D. G., Modzeleski, W., & Jimerson, S. R. (2010). What can be done
39(1), 27-37.
Borum, R., Fein. R., Vossekuil, B., & Berglund, J. (1999). Threat assessment: Defining
an approach for evaluating risk of targeted violence. Behavioral Sciences and the
BSL349>3.0.CO;2
Bosworth, K., Espelage, D. L., & Simon, T. (1999). Factors associated with bullying
Boudreaux, M., Lord, W., & Dutra, R. (1999). Child abduction: Aged-based analysis of
Bovasso, G. (2014). Assessing the risk of threats with guns in the general population.
.1037/tam000010
Justice Association.
Brunner, R., Henze, R., & Parzer, P. (2010). Reduced prefrontal and orbitofrontal gray
Burns, R., & Crawford, C. (1999). School shootings, the media, and public fear:
Ingredients for a moral panic. Crime, Law and Social Change, 32(2), 147-168.
Callagan, S., & Joseph, S. (1995). Self-concept and peer victimization among school
Cameron, K. (2009). Assessing violence potential: Protocol for dealing with high-risk
Campbell, A. (1991). The girls in the gang: A report from New York City. Cambridge,
MA: Blackwell.
Carnagey, N. L., Anderson, C. A., & Bartholow, B. D. (2007). Media violence and social
Centers for Disease Control and Prevention. (2004). Violence-related behaviors among
high school students. United States, 1991-2003. Morbidity and Mortality Weekly
Centers for Disease Control and Prevention. (2012). Injury prevention and gun control:
http://www.cdc.gov/injury/wisgars/fatal_injury_reports.html
Centers for Disease Control and Prevention. (2014). Source of firearms used by students
in school-associated violent deaths United States, 2003-2010. MMWR:
Chen, C., & Weikart, L. (2008). School background, school climate, school disorder, and
Chesney-Lind, M., & Shelden, R. (1996). Girls, delinquency, and gang membership. In
C. R. Huff (Ed.), Gangs in America (pp. 185-204). Thousand Oaks, CA: Sage.
Chesney-Lind, M., & Shelden, R. (1998). Girls, delinquency, and juvenile justice (2nd
Chou, C., Montgomery, S., Pentz, M., Rohrbach, L., Johnson, C., Flay, B., & Mackinnon,
Chyi, H., & McCombs, M. (2004). Media salience and the process of framing: Coverage
Coie, J. D., & Dodge, K. A. (1998). Aggression and antisocial behavior. In W. Damon
(Series Ed.). & N. Eisenberg (Vol. Ed.), Handbook of child psychology: Vol. 3.
Social emotional and personality development, 5th ed. (pp. 779-862). New York,
NY: Wiley.
Coie, J. D., Watt, N. F., West, S., Hawkins, J. D., Asarnow, J. R., Markman, H. J.,. . . &
doi: 10.1037/0003-066X.48.10.1013
Colman, J., & Hoffer, T. (1987). Public and private high schools: The impact of
Consalvo, M. (2003). The monsters next door: Media constructions of boys and
10.1080/1468077032000080112
Coogan, M. (2008). The Old Testament: A very short introduction. New York, NY:
Cooper, M., & Eaves, D. (1996). Suicide following homicide in the family. Violence and
Cornell, D., & Sheras, P. (2006). Guidelines for responding to student threats of violence.
Cunningham, M., & Sorensen, J. (2007). Predictive factors for violent misconduct in
10.1177/0032885507303752
https://www.mosaicmethod.com/
DeJeong, W., Epstein, J., & Hart, T. (2003). Bad things happen in good communities:
Understanding lethal school violence (pp. 70-100) Washington, DC: The National
Academic Press.
Department of Health and Human Services. (2001). Youth violence: A report of the
www.surgeongeneral.gov/library/youthviolence/toc.html
Douglass, K., Guy, L., & Reeves, K. (2008). Violence risk assessment and management:
Drysdale, D., Modezeleski, W., & Simons, A. (2010). Campus attacks: Targeted violence
Department of Justice.
Duncan, R. D. (1999). Maltreatment by parents and peers: The relationship between child
55.
10.1037/h0080360
Dwyer, K., Osher, D., & Wagner, C. (1998). Early warning, timely response: A guide to
Eaton, D. K., Kann, L., Kinchen, S., Shanklin, S., Flint, K. H., Hawkins, J., & Wechsler,
H. (2012). Youth risk behavior surveillance United States, 2011. Morbidity and
Elliott, D. S. (1994). Youth violence: An overview. Boulder, CO: Center for the Study and
Etain, B., Henry, C., & Bellivier, F. (2008). Beyond genetics: Childhood affective trauma
5618.2008.00635.x
Every Town for Gun Safety. (2015). Analysis of recent mass shootings. Retrieved from
http://everytownresearch.org/documents/2015/09/analysis-mass-shootings.pdf
Federal Bureau of Investigation. (1998). Uniform crime reports: Crime in the United
Federal Bureau of Investigation. (2001). Uniform crime reports: Crime in the United
Federal Bureau of Investigation. (2014). Active shooter events from 2000 to 2012.
Fein, R., & Vossekuil, B. (1998). Protective intelligence and threat assessment
Justice.
Fein, R. A., Vossekuil, B., & Holden, G. A. (1995, September). Threat assessment: An
Action, 1-7.
Fein, R. A., Vossekuil, B., Pollack, W. S., Borum, R., Modzeleski, W., & Reddy, M.
Schools Program and U.S. Secret Service, National Threat Assessment Center.
RISK FOR TARGETED SCHOOL VIOLENCE 91
Felitti, V. J., Anda, R. F., Nordenberg, D., Williamson, D. F., Spitz, A. M., Edwards, V., .
14(4), 245-258.
0562/90/0200-0014406.00/0
Fox, J., & Zawitz, M. (2001). Homicide trends in the United States. Washington, DC:
Freud, S. (1923). The Ego and the Id, Standard Edition 19. London, England: Kamac
Books.
Glassner, B. (1999). The culture of fear: Why Americans are afraid of the wrong things.
Goode, E. (2013, March 14). Focusing on violence before it happens. The New York
Times, p. A13.
RISK FOR TARGETED SCHOOL VIOLENCE 92
Gottesman, R., & Brown. R. (Eds.). (1999). Violence in America: An encyclopedia (2nd
Haider-Markel, D., & Joslyn, M. (2001). Gun policy, opinion, tragedy, and blame
Hanfland, K., Keppel, R., & Weis, J. (1997). Case management for missing children
Harding, D., Metha, J., & Newman, K. (2003). No exit: Mental illness, marginality, and
Hawdon, J., Oksanen, A., & Rasanen, P. (2012). Media coverage and solidarity after
Hawkins, J. D., Herrenkohl, T. I., Farrington, D. B., Catalano, R. F., Harachi, T. W., &
Department.
Hawkins, J., Herrenkohl, T., Farrington, D., Brewer, D., Catalano, R., & Harachi, T.
Farrington (Eds.), Serious and violent juvenile offenders: Risk factors and
Hawkins, J., Lishner, R., & Catalano, R. (1985). Childhood predictors and the prevention
Monograph Vol. 56: Etiology of drug abuse: Implications for prevention (pp. 75-
Peterson (Eds.), Crime and inequality (pp. 140-153). Stanford, CA: Stanford
University Press.
Hemenway, D., & Miller, M. (2000). Firearm availability and homicide across 26 high-
income countries. Journal of Trauma-Injury Infection and Critical Care, 49, 985-
988.
Hepburn, L. M., & Hemenway, D. (2004). Firearm availability and homicide: A review
10.1016/s1359-1789(03)00044-2
Herrera, V., & McCloskey, L. (2000). Gender differences in the risk for delinquency
among youth exposed to family violence. Child Abuse and Neglect, 25(8), 1037-
1051.
Huesmann, L. R., Moise-Titus, J., Podolski, C. L., & Eron, L. D. (2003). Longitudinal
201-221.
Joyal, S. (2013). Relationship violence and suicidality and homicidality among high
http://digitalcommons.uri.edu/cgi/viewcontent.cqi?article=2809&context=disserta
tions
(B. M. Hinkle, Trans.). New York, NY: Moffat, Yard and Company. (Original
Jung, C. G. (1938). Psychology and religion. New Haven, CT: Yale University Press.
Jung, C. G. (1953). Two essays in analytical psychology, second edition. New York,
Kaj, B., & Pirkko, N. (1992). Of mice and women: Aspects of female aggression. San
Kaltiala-Heino, R., Rimpela, M., Marttunen, M., Rimpela, A., & Rantanen, O. (1999).
Kaminski, J. (2013). Weapons used in mass shootings: OLR Research Report. Retrieved
from http://www.reuters.com/article/2008/06/26/us-usa-guns-court-shootings-
idUSN2631008420080626
Kandel, E., Mendick, S. A., Kirkegaard-Soresen, L., Hutchings, B., Knop, K., Rosenberg,
R., & Schulsinger, F. (1988). IQ as a protective factor for subjects at high-risk for
Kellermann, A. L., Somes, G., Rivara, F. P., Lee, R. K., & Banton, J. G., (1998). Injuries
and deaths due to guns in the home. Journal of Trauma, Injury, Infection, and
Krouse, W. J. (2012). Federal regulation of firearms. In Gun Control Legislation (pp. 14-
Krug, E., M., Dahlberg, A., Mercy, J., Zwi, A., & Lozano, R. (Eds.) (2002). World report
Langman, P. (2009). Why kids kill: Inside the minds of school shooters. New York, NY:
Palgrave McMillan.
Press.
RISK FOR TARGETED SCHOOL VIOLENCE 96
Lawrence, R., & Birkland, A. (2004). Guns, Hollywood, and school safety: Defining the
school-shooting problem across the public arenas. Social Science Quarterly 85(5),
Leary, M. R., Kowalski, R. M., Smith, L., & Phillips, S. (2003). Teasing, rejection, and
Lewis, P. (December 3, 2015). San Bernardino Shootings: What we know about the
news/2015/dec/03/san-bernadino-shooting-suspects-syed-rizwan-farook-tashfeen-
malik
Loeber, T. (1982). The stability of antisocial and delinquent child behavior: A review.
Masten, A., & Coatsworth, J. (1998). The development of competence in favorable and
McCabe, K., & Martin, G. (2005). School violence, the media, and criminal justice
McGee, J. P., & DeBernardo, C. R. (1999). The classroom avenger. The Forensic
Meloy, J. R., Hempel, A. G., Mohandie, K., Shiva, A. A., & Gray, B. T. (2001). Offender
719-728.
Behavioral Sciences and the Law Behavior Science Law, 30, 256 279. Wiley
Meloy, R., & White, S. (2007). WAVR-21: A structured professional guide for the
workplace assessment of violence risk (2nd ed.). San Diego, CA: Specialized
Training Services.
Miller, A. (2004). Violence in U.S. public schools: 2001 school survey on crime and
Education.
Miller, M., Azrael, D., & Hemenway, D. (2002). Rates of household firearms ownership
Miller, W., & Rollnick, S. (2002). Motivational interviewing: Preparing people for
Modzeleski, W., Feucht, T., Rand, M., Hall, J., Simon, T., & Butler, L. (2008). School-
Moffitt, T. E., Lynam, D. R., & Sylva, P. A. (1994). Neuropsychological tests predicting
9125.1994.tb01155.x
educators, law enforcement, and mental health professionals. San Diego CA:
Moore, M., Petrie, C., Braga, A., & McLaughlin, B. (2003). Deadly lessons:
Muschert, G. (2007). The Columbine victims and the myth of the juvenile surperpredator.
10.1177/1541204006296173
Muschert, G., & Carr, D. (2006). Media salience and frame changing across event:
Nansel, T. R., Overpeck, D., Haynie, D., Ruan, J., & Scheidt, P. (2003). Relationship
between bullying and violence among U.S. youth. Archives of Pediatrics and
National Center for Health Statistics. (2010) Health, United States, 2009: In-brief-
Services.
RISK FOR TARGETED SCHOOL VIOLENCE 99
National Crime Prevention Council. (1988). Be safe and sound. A case-study evaluation
National Institute on Drug Abuse (NIDA). (2003). Preventing drug abuse among
http://ncadistore.samhsa.gov/CatalogNIDA/Pub_Details.aspx?ItemID=16616
Neary, A., & Joseph, S. (1994). Peer victimization and its relationship to self-concept and
186.
Neiman, S., Murphy, C., Swaim, N., Thomas, T., & Parmer, R. (2011). School survey on
crime and safety: 2009-10. Washington, DC: Institute of Education Sciences, U.S.
Department of Education.
Neroni, H. (2000). The men of Columbine: Violence and masculinity in American culture
and film. Journal for the Psychoanalysis of Culture and Society, 5(2), 256-63.
Newman, K. (2004). Rampage: The social roots of school shootings. New York, NY:
Basic Books.
RISK FOR TARGETED SCHOOL VIOLENCE 100
Nock, M. K., Green, J. G., & Hwang, I. (2013). Prevalence, correlates, and treatment of
10.1001/jamapsychiatry.2014.1760
crime-victimizatioon-survey
Ogle, J., Eckman, M., & Amoroso, C. (2003). Appearance cues and the shootings at
Oh, H. A., Park, Y. R., & Choi, M. H. (2008). The effects of parent-adolescent
Okoro, C. A., Nelson, D. E., Mercy, J. A., Balluz, L. S., Crosby, A. E., & Mokdad, A. H.
states and the District of Columbia: Findings from the Behavioral Risk Factor
456.
hostile intentions. Journal of the American Academy of Psychiatry and the Law,
36, 544-550.
Reddy, M., Borum, R., Berglund, J., Vossekuil, B., Fein, R., & Modzeleski, W. (2001).
threat assessment, and other approaches. Psychology in the Schools, 38(2), 157-
172.
Ritchey, F. (2008). The statistical imagination: Elementary statistics for the social
Roberts, S., Zhang, J., & Truman, J. (2012). Indicators of school crime and safety: 2011
http://bjs.ojp.usdoj.gov/content/pub/pdf/iscs1.pdf
Roberts, S., Zhang, J., Truman, J., & Snyder, T. (2011). Indicators of school crime and
safety. Washington, DC: National Center for Education Statistics & U.S.
ullivan, C., Webster, D., & Campbell, J. (2005). Intimate partner violence
Rutter, M., Giller, H., & Hagell, A. (1998). Antisocial behavior by young people. New
Sandler, P., & Alpert, J. (2000). Violence and group dynamics in the high school: The
Scharrer, E., Wiedman, L., & Bissel, K. (20003). Pointing the finger of blame: News
Schiele, J. (2001). When white boys kill: an Afrocentric analysis. Journal of Human
Shultz, J. M., Muschert, G. W., Dingwall, A., & Cohen, A. M. (2013). The Sandy Hook
Elementary School shooting as tipping point. Disaster Health, 1(2), 65-73. doi:
10.4161/dish.27113
10.2105/AJPH.92.7.1161
doi: 10.1093/ije/dyl083
RISK FOR TARGETED SCHOOL VIOLENCE 103
Singh, G. K., & Yu, S. M. (1996). Trends and differentials in adolescent and young adult
86(4), 560-564.
Slee, P. T. (1994). Situational and interpersonal correlates of anxiety associated with peer
Snell, C., Bailey, C., Corona, A., & Mebane, D. (2002). School crime policy changes:
Snyder, H., & Sickmund, M. (1999). Juvenile offenders and victims: 1999 National
Stretesky, P., & Hogan, M. (2001). Columbine and student perceptions of safety: A
10.1016/S0047-2352(01)00100-3
Sullivan, M., & Guerette, R. (2003). The copycat factor: Mental illness, guns, and the
Temlow, S. W., Fonagy, P., Sacco, F. C., O'Toole, M. E., & Vernberg, E. (2002).
U.S. Census Bureau. (2010). Statistical abstract of the United States (129th ed.).
U.S. Department of Education, National Center for Education Statistics. (2000). Digest of
U.S. Department of Justice and U.S. Department of Education. (2000). Annual Report on
School Safety. Washington, DC: U.S. Department of Justice and the U.S.
Department of Education.
U.S. Secret Service and United States Department of Education. (2002). The final report
and findings of the safe school initiative: Implications for the prevention of school
Vossekuil, B., Fein, R. A., Reddy, M., Borum, R., & Modzeleski, W. (2004). The final
report and findings of the Safe School Initiative: Implications for the prevention
of school attacks in the United States. Washington, DC: U.S. Secret Service and
Vossekuil, B., Reddy, M., & Fein, R. (2000). Safe school initiative: An interim report on
Ward, D. A., & Tittle, C. R. (1994). IQ and delinquency: A test of two competing
Watkins, M., & Shulman, H. (2008). Towards psychologies of liberation. New York,
Webber, J. (2003). The martyrs of Columbine: Faith and the politics of tragedy. New
Webster, D. W., Gainer, P., & Champion, H. (1993). Weapon carrying among inner-city
White, N., & Lauritsen, J. (2010). Violent crime against youth (2nd ed.). St Louis MO:
Mosby.
Williamson, D., Biermaher, B., & Ryan, N. (2003). The stressful life events schedule for
doi: 10.1016/S0165-1781(03)00134-3
World Health Organization. (2004). Annex 2: Death, causes, sex and mortality stratum in
WHO regions, estimates for 2002. In The world health report 2004-Changing-
http://www.who.int/whr/2004/annex/topic/en/annex_2_en.pdf?ua=1
World Health Organization. (2015). January 2015 in the WHO Mortality Database.
Xu, J., Kochanek, K. D., Murphy, S. L., & Tejada-Vera, B. (2010). Deaths: final data for
2007. National vital statistics reports: from the Centers for Disease Control and
Prevention, National Center for Health Statistics, National Vital Statistics System,
58(19), 1-19.
RISK FOR TARGETED SCHOOL VIOLENCE 107
APPENDIX
Tables
Table 1
Percentages and Frequencies, Study Variables
Frequency Percent
Menninger Triad
Category 1 only 2 3.6%
Category 2 only 13 23.6%
Category 3 only 0 0.0%
Category 1 and 2 29 52.7%
Category 1 and 3 0 0.0%
Category 1 and 3 3 5.5%
Category 1, 2 and 3 8 14.5%
Respondent race/ethnicity
Hispanic 27 49.1%
Other 28 50.9%
Clinical diagnosis
Adjustment diagnosis 5 9.1%
Anxiety diagnosis 3 5.5%
Attention Deficit diagnosis 1 18.0%
Disruptive Behavior diagnosis 7 12.7%
Mood diagnosis 31 56.4%
Psychotic diagnosis 8 14.5%
Substance abuse
Yes 9 16.4%
No 46 83.6%
History of violence
Yes 29 52.7%
No 26 47.3%
Weapons in the home
Yes 31 56.4%
No 24 43.6%
Bullying
None 10 18.2%
Bullying perpetrator 8 14.5%
Bullying victim 34 61.8%
Bullying victim and perpetrator 3 5.5%
Educational ability
Learning disabled 15 27.3%
Average 33 60.0%
RISK FOR TARGETED SCHOOL VIOLENCE 108
Gifted 7 12.7%
Hospitalized
Yes 46 83.6%
No 9 16.4%
Autism diagnosis
None 43 78.2%
Asperger's 5 9.1%
Autism 7 12.7%
Living arrangement
Lives alone 6 10.9%
Immediate family 35 63.6%
Extended family 4 7.3%
Foster family 4 7.3%
Residential facility, 6 or more people 4 7.3%
Residential facility, 7 or more people 1 1.8%
Juvenile detention center 1 1.8%
Respondent lives with...
2 adoptive parents 2 3.6%
2 biological parents 15 27.3%
adoptive great aunt and uncle 1 1.8%
adoptive mom 1 1.8%
Alone 6 10.9%
in DCFS custody 4 7.3%
father only 1 1.8%
Girlfriend 1 1.8%
a grandparent(s) 4 7.3%
a mom and stepdad 7 12.7%
a single mom 12 21.8%
a student host mother 1 1.8%
DCFS status
Yes 17 30.9%
No 38 69.1%
Age of respondent
5-12 10 18.2%
13-18 33 60.0%
19-25 7 12.7%
26-35 1 1.8%
36+ 4 7.3%
Gender of respondent
Male 49 89.1%
Female 6 10.9%
N 55 100.0%
RISK FOR TARGETED SCHOOL VIOLENCE 109
Table 2
Variables 1 2 3 4 5 6 7 8 9 10 11 12
1 1.0
2 .316* 1.0
3 -.165 -.080 1.0
4 -.211 -.118 -.108 1.0
5 .101 .086 -.205 -.588** 1.0
6 .121 -.202 -.047 -.134 -.254 1.0
**
7 .121 .162 -.080 -.230 -.436 -.099 1.0
8 .070 .136 -.136 -.240 .190 -.128 .176 1.0
9 .015 .168 -.191 .139 .056 -.076 -.096 -.087 1.0
10 .181 .068 -.061 .122 .046 -.076 -.130 -.063 -.058 1.0
11 -.041 -.176 -.047 -.134 .067 .295* -.099 -.128 -.076 -.076 1.0
12 .023 .006 -.026 -.076 .129 -.033 -.056 .109 -.134 -.043 -.033 1.0
13 -.047 -.035 .217 .173 -.185 -.092 -.003 -.350** .062 -.121 -.092 -.052
14 -.072 .092 -.025 -.201 .048 .050 .155 .066 -.016 -.359** -.273* -.155
15 .016 -.039 -.080 .135 -.023 -.099 -.024 .331* .111 -.130 -.099 -.056
16 .123 .047 -.086 -.130 .025 -.106 .236 .305* .057 -.140 -.106 -.060
*
17 .286 .132 -.011 -.159 -.094 .227 .184 .080 -.090 .173 .067 -.144
18 .151 .191 -.025 -.115 -.099 .211 .155 .029 .204 .023 -.112 -.155
**
19 .016 -.360 -.080 .013 -.023 .128 -.024 -.083 -.096 .228 .355** -.056
*
20 .059 .071 -.047 -.267 .230 .024 .006 .045 .098 -.142 -.141 .107
NOTE: * < p .05; ** < p .01; *** < p .001, two-tailed tests.
RISK FOR TARGETED SCHOOL VIOLENCE 110
Variables 13 14 15 16 17 18 19 20 21 22 23 24
1
2
3
4
5
6
7
8
9
10
11
12
13 1.0
14 -.434** 1.0
15 -.158 -.469** 1.0
16 -.021 .191 -.043 1.0
17 -.075 .048 -.126 .222 1.0
18 .116 -.109 .155 .092 .121 1.0
19 .152 -.261 -.170 -.043 .081 -.053 1.0
20 -.037 .063 .112 -.057 -.144 -.012 -.525** 1.0
NOTE: * < p .05; ** < p .01; *** < p .001, two-tailed tests.
RISK FOR TARGETED SCHOOL VIOLENCE 111
Variables 1 2 3 4 5 6 7 8 9 10 11 12
**
21 -.122 .179 .381 .055 -.093 -.058 -.099 -.128 -.076 .203 -.058 -.033
*
22 -.017 .162 -.092 .293 -.215 -.113 .073 .094 .009 -.149 -.113 -.064
23 -.180 .148 -.087 -.263 .202 -.161 .180 .701** -.120 -.226 -.134 .122
24 -.039 .061 .046 -.008 .013 .057 -.070 -.071 -.063 -.028 .187 -.189
*
25 -.173 .174 .086 -.332 .172 .106 .043 .187 -.057 -.202 .106 -.308*
26 .109 -.033 .043 -.092 .157 -.120 -.057 .242 -.279* .025 .111 .325*
27 .235 .119 -.068 -.057 -.019 -.084 .186 .689** -.227 .092 -.084 -.048
28 -.164 .146 .147 -.024 .041 -.151 -.010 -.379** .213 -.024 -.151 .103
29 .117 .005 -.054 -.156 -.015 .241 .083 -.056 .005 -.089 -.067 -.038
30 -.025 -.091 -.054 .174 -.015 -.067 -.116 -.196 .005 .155 .241 -.038
31 -.095 -.225 -.054 .174 -.156 .241 -.116 -.056 -.135 -.089 .241 -.038
32 -.115 -.044 -.026 -.076 .129 -.033 -.056 .382** .139 -.043 -.033 -.019
33 .161 -.145 -.026 -.076 .129 -.033 -.056 -.027 -.134 -.043 -.033 -.019
34 .032 .019 -.038 .121 -.011 -.047 -.080 -.136 .004 .276* -.047 -.026
*
35 -.271 -.006 -.119 -.052 .089 -.147 .095 -.082 .215 -.194 .033 .222
36 .023 -.138 -.026 -.076 -.144 .567** -.056 -.027 -.134 -.043 .567** -.019
37 -.115 -.030 -.026 .245 -.144 -.033 -.056 -.027 -.134 -.043 -.033 -.019
38 .235 .119 -.068 -.057 -.019 -.084 .186 .689** -.227 .092 -.084 -.048
39 -.166 -.125 -.054 .009 .125 -.067 -.116 -.126 -.135 -.089 .241 -.038
40 .161 -.145 -.026 -.076 .129 -.033 -.056 -.027 -.134 -.043 -.033 -.019
NOTE: * < p .05; ** < p .01; *** < p .001, two-tailed tests.
RISK FOR TARGETED SCHOOL VIOLENCE 112
Variables 13 14 15 16 17 18 19 20 21 22 23 24
21 -.092 .050 -.099 .110 .227 -.112 -.099 -.306* 1.0
22 -.039 .130 .073 .046 -.026 .130 -.194 -.600** -.113 1.0
** *
23 -.391 .196 .319 .265 -.153 .072 -.219 .112 -.161 .154 1.0
*
24 .002 -.089 .097 .105 -.283 .209 .181 -.307* -.073 .265 .245 1.0
25 .021 .007 .182 -.070 .172 .205 -.236 .259 -.110 -.046 .148 -.025
26 -.112 -.090 .018 .063 -.001 -.196 -.057 .068 -.004 -.031 .086 -.053
**
27 -.134 -.163 .352 .161 .215 -.045 .021 -.085 -.084 .137 .413** -.012
**
28 .175 .097 -.224 -.176 -.413 .021 -.224 .106 .182 -.036 -.158 .006
29 -.107 .105 .083 .065 .265 .105 .083 .076 -.067 -.132 -.054 .066
30 .103 -.177 -.116 .065 .265 -.177 .282* -.212 -.067 .050 -.243 -.048
31 -.107 .105 -.116 -.124 .125 -.036 .083 -.068 -.067 .050 -.031 -.048
*
32 -.052 -.155 .330 -.060 -.144 .120 -.056 .107 -.033 -.064 .122 -.189
*
33 -.052 .120 -.056 .308 -.144 .120 -.056 .107 -.033 -.064 .122 .253
34 -.074 -.025 -.080 -.086 -.205 .171 -.080 -.047 -.047 .160 -.152 .204
*
35 .011 .127 -.137 -.161 -.320 -.202 -.253 .229 -.147 .029 .042 -.187
36 -.052 -.155 -.056 -.060 .129 .120 .330* -.173 -.033 -.064 .031 .253
37 -.052 .120 -.056 -.060 -.144 -.155 -.056 -.173 -.033 .289* -.061 .032
**
38 -.134 -.163 .352 .161 .215 -.045 .021 -.085 -.084 .137 .413** -.012
39 .103 -.036 -.116 .065 .265 -.177 .083 -.068 -.067 .050 -.101 -.161
*
40 -.052 .120 -.056 .308 -.144 .120 -.056 .107 -.033 -.064 .122 .253
NOTE: * < p .05; ** < p .01; *** < p .001, two-tailed tests.
RISK FOR TARGETED SCHOOL VIOLENCE 113
Variables 25 26 27 28 29 30 31 32 33 34 35 36
21
22
23
24
25 1.0
26 .079 1.0
27 .155 .246 1.0
28 -.130 -.005 -.463** 1.0
29 .124 -.140 -.098 -.370** 1.0
**
30 -.065 .062 -.098 -.370 -.078 1.0
**
31 .124 -.140 -.098 -.370 -.078 -.078 1.0
32 .060 -.068 -.048 -.180 -.038 -.038 -.038 1.0
*
33 -.308 -.068 -.048 -.180 -.038 -.038 -.038 -.019 1.0
34 -.177 .183 -.068 .147 -.054 -.054 -.054 -.026 -.026 1.0
**
35 .050 .225 -.214 .463 -.171 -.171 -.171 -.083 -.083 -.119 1.0
36 .060 -.068 -.048 -.180 -.038 -.038 .486** -.019 -.019 -.026 -.083 1.0
37 .060 -.068 -.048 -.180 -.038 -.038 .486** -.019 -.019 -.026 -.083 -.019
** **
38 .155 .246 1.000 -.463 -.098 -.098 -.098 -.048 -.048 -.068 -.214 -.048
**
39 .124 .062 -.098 -.370 -.078 .730** .191 -.038 -.038 -.054 -.171 -.038
40 -.308* -.068 -.048 -.180 -.038 -.038 -.038 -.019 1.000** -.026 -.083 -.019
NOTE: * < p .05; ** < p .01; *** < p .001, two-tailed tests.
RISK FOR TARGETED SCHOOL VIOLENCE 114
Variables 37 38 39 40 41 42 43 44 45 46 47
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37 1.0
38 -.048 1.0
39 -.038 -.098 1.0
40 -.019 -.048 -.038 1.0
NOTE: * < p .05; ** < p .01; *** < p .001, two-tailed tests.
RISK FOR TARGETED SCHOOL VIOLENCE 115
Variables 1 2 3 4 5 6 7 8 9 10 11 12
41 -.115 -.044 -.026 -.076 .129 -.033 -.056 .382** .139 -.043 -.033 -.019
42 .117 .005 -.054 -.156 -.015 .241 .083 -.056 .005 -.089 -.067 -.038
43 .119 -.250 .217 .173 -.185 -.092 -.003 -.350** .062 .069 -.092 -.052
**
44 .132 .422 .133 -.087 .059 .067 -.093 -.062 .098 .139 -.127 -.072
*
45 -.253 -.283 -.026 .245 -.144 -.033 -.056 -.027 -.134 -.043 -.033 -.019
46 .071 -.155 -.130 .184 -.076 .186 -.164 -.331* -.027 .062 .186 -.091
47 -.117 .043 .068 -.080 .136 -.173 -.021 .129 -.240 -.092 .084 .048
NOTE: * < p .05; ** < p .01; *** < p .001, two-tailed tests.
RISK FOR TARGETED SCHOOL VIOLENCE 116
Variables 13 14 15 16 17 18 19 20 21 22 23 24
*
41 -.052 -.155 .330 -.060 -.144 .120 -.056 .107 -.033 -.064 .122 -.189
42 -.107 .105 .083 .065 .265 .105 .083 .076 -.067 -.132 -.054 .066
** *
43 .509 -.324 -.003 -.169 -.075 .116 .307* -.037 -.092 -.180 -.336* .002
*
44 -.202 .287 -.218 .123 .147 .110 -.093 -.129 .455** -.021 -.014 .053
*
45 -.052 -.155 .330 -.060 -.144 -.155 -.056 .107 -.033 -.064 .077 .032
* *
46 .217 -.046 -.276 -.083 .318 .033 .170 -.041 .013 -.111 -.471** -.225
47 -.041 .045 -.021 .155 .136 .045 -.021 -.035 .084 .014 .215 .012
NOTE: * < p .05; ** < p .01; *** < p .001, two-tailed tests.
RISK FOR TARGETED SCHOOL VIOLENCE 117
Variables 25 26 27 28 29 30 31 32 33 34 35 36
41 .060 -.068 -.048 -.180 -.038 -.038 -.038 1.000** -.019 -.026 -.083 -.019
** **
42 .124 -.140 -.098 -.370 1.000 -.078 -.078 -.038 -.038 -.054 -.171 -.038
43 -.126 -.190 -.134 .062 -.107 .103 .103 -.052 -.052 -.074 -.234 -.052
**
44 -.123 -.200 -.185 .399 -.148 -.148 -.148 -.072 -.072 -.103 -.323* -.072
45 .060 -.068 -.048 .103 -.038 -.038 -.038 -.019 -.019 -.026 -.083 -.019
46 .083 -.163 -.234 -.230 -.036 .419** .419** -.091 -.091 -.130 -.233 .203
47 .161 .174 .122 -.265 .098 .098 .098 .048 .048 .068 -.048 .048
NOTE: * < p .05; ** < p .01; *** < p .001, two-tailed tests.
RISK FOR TARGETED SCHOOL VIOLENCE 118
Variables 37 38 39 40 41 42 43 44 45 46 47
41 -.019 -.048 -.038 -.019 1.0
42 -.038 -.098 -.078 -.038 -.038 1.0
43 -.052 -.134 -.107 -.052 -.052 -.107 1.0
44 -.072 -.185 -.148 -.072 -.072 -.148 -.202 1.0
45 -.019 -.048 -.038 -.019 -.019 -.038 -.052 -.072 1.0
**
46 .203 -.234 .419 -.091 -.091 -.036 .217 .028 -.091 1.0
47 .048 .122 .098 .048 .048 .098 -.041 -.239 .048 -.145 1.0
NOTE: * < p .05; ** < p .01; *** < p .001, two-tailed tests.
RISK FOR TARGETED SCHOOL VIOLENCE 119
Key to correlation tables 17. History of violence 32. Living Arrangement: Residential
Facility, 7 or more people
1. MOSAIC Score: Assessment 18. Weapons
33. Living Arrangement: Juvenile
2. MOSAIC Score: Quality 19. Bullying perpetrator detention center
3. Menninger Triad Category 1 only 20. Bullying victim 34. Lives with 2 adoptive parents
4. Menninger Triad Category 2 only 21. Bullying victim and perpetrator 35. Lives with 2 biological parents
5. Menninger Triad Category 1 and 2 22. No bullying 36. Lives with adoptive great aunt &
uncle
6. Menninger Triad Category 2 and 3 23. Education level
7. Menninger Triad Category 1, 2 37. Lives with adoptive mom
24. Educational ability
and 3 38. Lives alone
25. Hospitalized
39. Lives in DCFS custody
26. Autistic diagnosis
9.Race or ethnicity: Hispanic 40. Lives with father only
27. Living Arrangement: Lives
10. Adjustment diagnosis Alone 41. Lives with girlfriend
11. Anxiety diagnosis 28. Living Arrangement: Immediate 42. Lives with a grandparent(s)
Family
12. Attention Deficit diagnosis 43. Lives with a mom and stepdad
29. Living Arrangement: Extended
13. Disruptive Behavioral diagnosis Family 44. Lives with a single mom
14. Mood diagnosis 30. Living Arrangement: Foster 45. Lives with a student host mother
Family
15. Psychotic diagnosis 46. DCFS status
31. Living Arrangement: Residential
16. Substance abuse 47. Gender of respondent
Facility, 6 or more people