Expert Commentary

Violence in schools: Research findings on underlying dynamics, response and prevention

2014 literature review of academic studies relating to various aspects of mass shooting and murder incidents in United States schools.

School crime scene (iStock)

According to the Centers for Disease Control and Prevention (CDC), 7.4% of high school students in 2011 reported being threatened or harmed with a weapon on school grounds. The National Center for Education Statistics notes that between 1992 and 2009, there were between 14 and 34 homicides among children ages 5 to 18 at school each year. The CDC puts these levels of violence in statistical perspective: “Approximately 1% of all youth homicides in 2008-2009 occurred at school, and the percentage of all youth homicides occurring at school has been less than 2% since the 1992-1993 school year. There was approximately one homicide or suicide of a school-age youth at school per 2.7 million students enrolled during the 2009-2010 school year.” Further, the CDC notes that 5.4% of students “reported carrying a weapon (gun, knife or club) on school property on one or more days in the 30 days preceding the [2011] survey.”

Many schools have implemented strategies to eliminate or mitigate the damage and loss of life in possible future attacks; these range from improving mental health care and implementing violence prevention curriculum to enhanced security and communication protocols. Preparation levels vary by setting, according to 2009-10 National Center for Education Statistics data: 58% of suburban schools perform safety drills according to a specified plan, compared with 49% of urban schools and 48% of rural schools.

As researchers have noted, media outlets have sometimes overreacted to school shootings, bringing about what some scholars have called the construction of “moral panic” — an inflated sense of alarm over a perceived threatening trend. But it is unclear if school shootings can be sufficiently distinguished and defined from other violence in America. Hype and exaggeration may distract from formulating effective school policies, some commentators believe, even as horrifying incidents such as the school shootings in Littleton, Colo. and Newtown, Conn., trigger a national conversation about “cultures” of school violence. It is worth noting, too, that more mundane (and routine) forms of physical violence plague some schools, and non-gun related violence — such as the mass stabbings at Franklin Regional High School in Murrysville, Penn. — remains a significant issue.

Below are studies and articles that bring a research perspective to these issues. For more information, see other research reviews on rampage violence, firearms-related violence, global comparisons of firearms and homicides and the relationship between violence and psychotic behavior. Additional useful resources include the Poynter Institute’s primer on what journalists should know about school shootings and guns, as well as the FBI’s list of resources on school violence.


“School Intervention Related to School and Community Violence”
Lisa H. Jaycox; Bradley D. Stein; Marleen Wong. Child and Adolescent Psychiatric Clinics of North America, Volume 23, Issue 2, April 2014, Pages 281–293. doi:

Summary: “Schools offer the opportunity for intervention efforts related to student trauma, capitalizing on the natural environment and reduced stigma to leverage recovery. Such programs can be divided into different phases: crisis intervention, early intervention, and selective interventions for longer-term recovery. Although best practices for crisis intervention have been developed and rolled out nationally, interventions for early and longer-term recovery tend to be disseminated in a piecemeal fashion without the benefit of organized training and funding. Dissemination has most typically occurred following specific community-wide or school crises, or in inner city schools where violence exposure is endemic. Evaluations of program effectiveness have focused on the longer-term recovery interventions, and those that contain cognitive behavior elements have been the best tested to date. Future directions include research that measures educational gains, improvements in attendance, and reduction in behaviors associated with expulsion and suspension.”


“Perceived Injustice and School Violence: An Application of General Strain Theory”
Katie James; Jackson Bunch; Jody Clay-Warner. Youth Violence and Juvenile Justice, February 2014, doi: 10.1177/1541204014521251.

Abstract: “We examine the effect of perceived school fairness on one’s likelihood of participating in school violence and how social support influences this relationship. General strain theory (GST) and procedural justice theory suggest that when students perceive unfairness in school rules or treatment from teachers, they will be more likely to participate in violence. GST proposes that the strength of these relationships may be reduced by social support. Data from the 2009 School Crime Supplement of the National Crime Victimization Survey show that students who perceive unfair treatment from teachers are more likely to bring a weapon to school and fight at school than are students who believe that their teachers are fair. Students who perceive that rules are unfair are more likely to bring a weapon to school than are students who believe rules are fair. Perceived support from adults at school reduces the impact of teacher/rule unfairness on school violence.”


“Report to the President on Issues Raised by the Virginia Tech Tragedy”Leavitt, Michael O. U.S. Department of Justice. June 2007, Publication No. NCJ 218828.

Abstract : “Education officials, health-care providers, law enforcement personnel, and others were not sharing critical information on individuals who were likely to be a danger to themselves or others. Further, State laws and practices do not uniformly ensure that information on persons prohibited from possessing firearms is appropriately collected and made available for the National Instant Criminal Background Check System. Also, parents, students, and teachers are not sufficiently informed about the warning signs for potentially severe violent behavior and how to intervene appropriately for the person’s benefit. The resources and accessibility of mental health services for mentally ill persons living in communities are often insufficient in providing care, privacy, and safety. Recommendations pertinent to each of the aforementioned findings are divided into those to be implemented by State and local institutions and Federal institutions. The recommendations focus on interagency, intergovernmental, and public-private cooperation in integrating preventive and intervention services as well as information-sharing. Attention is given to integrated planning and implementation of plans through practice and effective communication.”


“Rampage Shootings in American High School and College Settings, 2002-2008”
Newman, Katherine; Fox, Cybelle. American Behavioral Scientist, May 2009, Vol. 52, No. 9, 1286-1308. doi: 10.1177/0002764209332546.

Findings: “Because [college shooters] are not bent on integration, their motives for the shooting are different from the protagonist in the middle and high school cases, who broadcast threats in advance to try to reverse their negative reputations and become notorious for the purpose of attracting friends… we found that many high school shooters were ambivalent about following through on their threats, hoping the threat alone might help improve their image in the eyes of their peers. The college shooters are bent on notoriety, but they appear less ambivalent about the attack. This is critical because it impacts the lead-up to the shooting. Because high school shooters are looking to attract attention, they let off warning signals that could—at least potentially—enable those who are within earshot to notify authorities who can intervene. Indeed, post- Columbine, our data (Newman et al., 2004, p. 51) suggest that once high school students recognized what the warnings could presage, they came forward and helped to reduce the number of completed attacks. But college shooters have no reason to let off warnings, and they generally do not. They may even delight in an interior, secretive knowledge of what is to come, while their victims are in the dark.”

“Crime in Schools and Colleges: A Study of Offenders and Arrestees Reported via National Incident-Based Reporting System Data”
Federal Bureau of Investigation, November 2007.

Excerpt: “This study, over the 5-year period [2000-04], found that 3.3 percent of all incidents reported via NIBRS involved school locations. The number of crime in school-related incidents was highest in October. Offense records were also most likely to include the use of personal weapons (hands, fists, feet, etc.), while reports of the offender’s use of alcohol, computers, and/or drugs were minimal. Reported offenders of crime in schools were most likely 13-15 year old white males who the victims reportedly knew; however, there was nearly an equally large number of 16-18 year old reported offenders. More than half of the arrestees associated with crime at school locations were arrested for simple assault or drug/narcotic violations. Arrestees had similar characteristics to the reported offenders, most likely being reported as 13-15 year old white non-Hispanic males who were residents of the community of the school location where the incident was reported.”


“A Coordinated Mental Health Crisis Response: Lessons Learned From Three Colorado School Shootings”
Crepeau-Hobson, Franci, et al. Journal of School Violence, July 2012, Vol. 11, Issue 3. doi: 10.1080/15388220.2012.682002

Excerpt: “Our experiences responding to the three school shootings in Colorado have provided us with a wealth of knowledge that can be used to inform responses to other major school-based crises. It is hoped that the information contained in this article will complement and add to existing crisis response models and published guidelines while perhaps filling in some gaps. Following is a summary of important considerations…. Crisis preparedness is key to an effective response. Schools and districts must have crisis teams and crisis plans in place. Teams should be organized functionally and be aligned with the National Incident Management System ICS structure. All crisis responders on these teams need training in school-based crisis response. In addition, each team should know what additional resources are available in their respective communities. Crisis plans ought to include potential reunification and safe haven sites, means for monitoring and following up with impacted individuals, and steps for evaluating the response….”


“Tragedy and the Meaning of School Shootings”
Warnick, Bryan R., et al. Education Theory, July 2010, Vol. 60, Issue 3. doi: 0.1111/j.1741-5446.2010.00364.x

Excerpt: “It is useful, first, to think about school shootings in terms of their apparent ritual meaning, particularly their meaning as a form of inquiry, a problem and proposed solution, about broken communities. Such an approach might with time not only lead to a better understanding of the cause of school shootings, it may also suggest ways of dealing with school shootings that involve the construction of counterrituals that offer alternative solutions to the ritual problem. We have also suggested that the lack of hospitality within the schools of the “neoliberal security state” may be one of many factors influencing the rise of targeted school shootings in the past twenty years, implying that school shootings might be addressed (in part) by revising some of those policy initiatives. Finally, we have tried to rethink the role of the teacher if we take seriously the advice given to teachers about “what to look for” in potential shooters, and have warned about where this might lead. These proposals are mostly exploratory and conjectural in nature, but they at least seem worthy of more extended consideration.”


“Cops and Cameras: Public School Security as a Policy Response to Columbine”
Addington, Lynn A. American Behavioral Scientist, June 2009, Vol. 52, No. 2, 1426-1446. doi: 10.1177/0002764209332556.

Findings: “[In the wake of the Columbine shootings], school administrators turned to visible security measures to demonstrate that they were “doing something.” Measures such as security cameras and SROs were appealing choices given financial support from the government and marketing efforts by companies. The outstanding question, however, concerns the effectiveness of these policy decisions. It is unclear whether these security measures work and to what extent they might generate negative consequences for students and schools… To more effectively respond to events like Columbine, school administrators and public officials should have a better appreciation for public reactions to highly publicized acts of extreme violence, especially with regard to increased fear and risk assessment. In this situation, better communication may be needed to express convincingly that effective policies might not embody the most visible changes. In addition, officials need information about the actual costs and benefits of a policy so that informed decisions can be made. Although such incidents are rare, they do occur. School administrators and other public officials must be able to quickly respond in an appropriate manner and not misdirect scarce resources from effective remedies.”


“Rampage School Shooters: A Typology”
Langman, Peter. Aggression and Violent Behavior, January-February 2009, Vol. 14, No. 1, 79-86. doi:

Abstract: “A number of researchers have sought to identify the features that school shooters have in common in terms of family life, personalities, histories, and behaviors. This article examines the cases of 10 rampage school-shooters in an effort to find out not only how they are alike, but also how they differ. Based on available information, these youths are categorized into three types: traumatized, psychotic, and psychopathic. Out of the 10 shooters discussed, three were traumatized, five were psychotic, and two were psychopathic. The three traumatized shooters all came from broken homes with parental substance abuse and parental criminal behavior. They all were physically abused and two were sexually abused outside of the home. The five psychotic shooters had schizophrenia-spectrum disorders, including schizophrenia and schizotypal personality disorder. They all came from intact families with no history of abuse. The two psychopathic shooters were neither abused nor psychotic. They demonstrated narcissism, a lack of empathy, a lack of conscience, and sadistic behavior. Most people who are traumatized, psychotic, and psychopathic do not commit murder. Beyond identifying the three types of rampage shooters, additional factors are explored that may have contributed to the attacks. These include family structure, role models, and peer influence.”


”School Shootings: Making Sense of the Senseless”
Wike, Traci L.; Fraser, Mark W. Aggression and Violent Behavior, May-June 2009, Vol. 14, No. 3, 162-169. doi: 10.1016/j.avb.2009.01.005.

Abstract: “School shootings have altered the patina of seclusion and safety that once characterized public and higher education. Callous and brutal, school shootings seem to make no sense. However, case comparisons and anecdotal reports are beginning to show patterns that provide clues for understanding both the individual factors motivating shooting events and the characteristics of schools where shootings have occurred. We describe these factors and characteristics as the bases for six prevention strategies: (a) strengthening school attachment, (b) reducing social aggression, (c) breaking down codes of silence, (d) establishing screening and intervention protocols for troubled and rejected students, (e) bolstering human and physical security, and (6) increasing communication within educational facilities and between educational facilities and local resources.”


“Mass Murder Goes to College: An Examination of Changes on College Campuses Following Virginia Tech”
Fox, James Alan; Savage, Jenna. American Behavioral Scientist, June 2009, Vol. 52, No. 10, 1465-1485. doi:10.1177/0002764209332558.

Abstract: “Although incidents at middle schools, high schools, and institutions of higher education all fall under the umbrella of “school shootings,” there are, in fact, several characteristics of shootings at college campuses that make them unique. As will be discussed, a major distinction surrounds the motivation and age of the shooter, which are important differences for consideration when fashioning appropriate steps for prevention. In addition, differences in the school environment—that is, a single building versus a sprawling campus—means that a reasonable response to shootings at middle or high schools (for example, lockdown) may not be feasible at colleges and universities. Thus, college administrators should not hastily adopt preventive measures simply because they appear effective at lower level schools. In fact, administrators should think very carefully before adopting many measures that are popular among colleges and universities these days, as some common strategies are not necessarily productive and may even potentially have negative consequences.”


“Mental Health System Transformation After The Virginia Tech Tragedy”
Bonnie, Richard J.; Reinhard, James S.; Hamilton, Phillip; McGarvey, Elizabeth L. Health Affairs, May/June 2009, Vol. 28, No. 3, 793-804. doi: 10.1377/hlthaff.28.3.793.

Abstract: “On 16 April 2007, a deeply disturbed Virginia Tech student murdered thirty-two fellow students and faculty and then shot himself. Less than one year later, the Virginia legislature improved the emergency evaluation process, modified the criteria for involuntary commitment, tightened procedures for mandatory outpatient treatment, and increased state funding for community mental health services. The unanswered question, however, is whether the necessary political momentum can be sustained for the long-term investment in community services and the fundamental legal changes needed to transform a system focused on managing access to scarce hospital beds to a community-based system of accessible voluntary services.”


“Mass Murder at School and Cumulative Strain: A Sequential Model”
Levin, Jack; Madfis, Eric. American Behavioral Scientist, May 2009, Vol. 52, No. 9, 1227-1245. doi: 10.1177/0002764209332543.

Findings: “It is important to emphasize that many school shootings take place during a period of less than 15 min (Vossekuil et al., 2004), so reactive measures (such as resource officers, emergency plans, and even armed faculty members) can ultimately accomplish little. Resource officers were on guard at Northern Illinois University and Columbine High School, but they were not able to reach the mass killers in time to avert disaster. From a routine-activities perspective, increasing the number and effectiveness of capable guardians and engaging in target-hardening tactics to diminish their suitability and easy access does nothing to diminish the third and most vital of Felson’s (1994) factors, the motivation of offenders. To this end, the focus must also be on long-term prevention techniques to ensure that students do not develop the desire to engage in a school massacre in the first place…. Our analysis suggests that incidents of multiple-victim shootings aimed at students and teachers might be deterred early on by reducing the chronic strains experienced by students who are likely to turn violent. Frequently, there are important warning signs — bullying, serious acts of animal abuse, lack of friendships — to identify students who have suffered prolonged frustration in school and/or at home and are in urgent need of assistance from supportive adults. The problem is that teachers, school psychologists, and counselors do not always react to troubled students until they become troublesome and are seen as a threat to others.”


“The Impacts of the Virginia Tech and Northern Illinois University Shootings on Fear of Crime On Campus”
Kaminski, Robert J. Journal of Criminal Justice, January-February 2010, Vol. 38, No. 1, 88-98. doi: 10.1016/j.jcrimjus.2009.11.011.

Abstract: “Recent campus shootings appear to be having wide-ranging impacts on colleges and universities across the country. Many campuses have expanded their emergency communication systems using multiple notification routes, such as text, e-mail, and phone alerts (Hamblen, 2008), though a recent survey of five hundred campuses using a particular emergency alert system found that only about 40 percent of students had registered for the service (Mark, 2008). Other initiatives in place or under consideration include the use of campus lockdowns, increasing security personnel, student profiling, and allowing students, faculty, and staff to carry concealed firearms on campus (Fox, 2008). As Fox (2008) notes, however, these responses may not be consistent with the actual threat to members of the university community and they may have detrimental consequences for the life of college campuses. Instead of making students feel safer about their surroundings, these new policies may, in fact, make students more fearful and less engaged in their college campuses. It is unclear whether the trend in campus shootings will continue to increase, level off, or decline, but clearly research on the effects of campus shootings is warranted.”


“Gun Carrying by High School Students in Boston, Mass: Does Overestimation of Peer Gun Carrying Matter?”
Hemenway, David, et al. Journal of Adolescence, October 2011, 34(5), 997-1003. doi: 10.1016/j.adolescence.2010.11.008.

Abstract: “This paper investigates: (1) whether high school students overestimate gun carrying by their peers, and (2) whether those students who overestimate peer gun carrying are more likely to carry firearms. Data come from a randomly sampled survey conducted in 2008 of over 1,700 high school students in Boston, MA. Over 5% of students reported carrying a gun, 9% of boys and 2% of girls. Students substantially overestimated the percentage of their peers who carried guns; the likelihood that a respondent carried a gun was strongly associated with their perception of the level of peer gun carrying. Most respondents believed it was easier for other youth to obtain guns than it was for them. Social marketing campaigns designed to lower young people’s perceptions about the prevalence of peer gun carrying may be a promising strategy for reducing actual gun carrying among youth.”


“Mass Shootings in Schools: The Worst Possible Case for Gun Control”
Kleck, Gary. American Behavioral Scientist, June 2009, Vol. 52, No. 10, 1447-1464. doi:10.1177/0002764209332557.

Abstract: “The most frequent policy lesson drawn following the Columbine school shootings was the need for more gun controls. Review of the details of both Columbine and other contemporary school shootings indicates, however, that the specific gun control measures proposed in their aftermath were largely irrelevant and almost certainly could not have prevented the incidents or reduced their death tolls. These measures included restrictions on gun shows, child access prevention laws mandating locking up guns, and bans on assault weapons. Ironically, exploitation of school shootings for the advocacy of irrelevant gun controls may have obscured the genuine merits of various gun control measures for reducing ‘ordinary’ gun violence. Thus, mass school shootings provided the worst possible basis for supporting gun control.”


“Tragedy at Virginia Tech: Trauma and Its Aftermath”
Flynn, Christopher; Heitzmann, Dennis. The Counseling Psychologist, May 2008, Vol. 36, No. 3, 479-489. doi:10.1177/0011000008314787.

Abstract: “While college campuses are relatively safe environments, the promise of safety and security on campus was shattered by a single gunman on April 16, 2007. Seung-Hui Cho, a senior at Virginia Tech, shot 49 students and faculty, killing 32, before killing himself. The authors are psychologists and directors of university counseling centers; they examine the many implications of this tragedy on mental health counseling. The assailant’s significant psychological disturbances and previous contact with mental health professionals are critical to understanding how he was able to act out his murderous rage. The mental health response to a traumatized community of families, friends, colleagues, and peers is reviewed. Out of the tragedy, there have emerged many issues that challenge the role of counseling centers within the university including the development of threat assessment teams, the potential conflicts between client confidentiality and crisis prevention/management, and the on-going education for the university community regarding suicide prevention, mental illness and support for potentially marginalized students.”


Keywords: research roundup, crime, safety, guns, mental health, higher education, youth

About The Author