In the wake of the Columbine school shootings, it is of utmost importance for psychiatrists to be aware of the role they can play in preventing violence and bullying in our schools. What programs have been tried and how have they fared? What are the elements for a successful program?
Although serious physical violence in schools has decreased measurably since 1993 (U.S. Department of Education and U.S. Department of Justice, 1999), social and emotional violence appears to be on the rise. Bullying has now reached epidemic proportions in schools. Nansel et al. (2001) reported that more than 16% of U.S. schoolchildren said they had been bullied by other students during the current term, and approximately 30% of sixth- through 10th-grade students reported being involved in some aspect of moderate-to-frequent bullying, either as a bully, the target of bullying or both. Victims of bullying can develop serious posttraumatic and depressive problems, and bullies may develop a range of antisocial behaviors. Both groups often suffer academically (Olweus et al., 1998).
A serious challenge facing violence prevention efforts comes from research suggesting that, as children grow throughout middle school and high school, they become hardened to the victimization of others, with decreasing empathy and helpfulness: 10% to 20% of children admitted vicarious satisfaction in seeing others hurt (Jacobs et al., unpublished data).
Various Approaches to Preventing School Violence
There is growing consensus that primary and secondary prevention of violence require three overlapping processes:
The clinician will most likely be consulted because of teachers' concerns that students' threatening communications may indicate that they might become violent. Teachers make these referrals to seek reassurance that these children are not going to become physically violent to themselves or others. Sometimes, the clinician may make a psychiatric assessment that does not address this particular need. Thus, useful assessment of children who threaten others with violence must occur in the context of close collaboration with teachers and law-enforcement personnel, along with home visits to assess issues such as computer activity and family dynamics.
Accurate threat assessment requires evaluating risk factors and protective factors for conduct disturbances. Risk factors include poverty, overcrowding, disadvantaged school settings, difficult temperaments, inadequate parenting, and poor prosocial skills and school performance (Kazdin, 1995; Rutter et al., 1998). Known protective factors include high IQ; easy disposition; ability to get along well with parent, siblings, teachers and peers; ability to do well in school; and being competent in social problem-solving (Rutter et al., 1970).
The clinician should take these and other factors into account in making a threat assessment model that can assist schools in a practical fashion. The Figure depicts a possible algorithm. (Due to copyright concerns, this figure cannot be reproduced online. Please see p62 of the print edition--Ed.)
In our own work, we reviewed major social factors that seem to promote violence in children who felt persecuted, bullied or misunderstood in the school environment (Twemlow et al., 2002). Some of the items summarized in the Figure represent our direct collaboration with the Federal Bureau of Investigation's psychiatrically informed model (O'Toole, undated), with information from the Secret Service Safe School Initiative's behavioral/case-management approach (Vossekuil et al., 2000).
Issues that should alert the clinician include previous warning(s) of problems, ambiguous threatening messages, availability of guns, victimization at school, concern expressed by adults or peers, interest in hate and weaponry Web sites, out-of-character changes in emotion and interests, and families with low emotional closeness and little knowledge of their child's life.
A dizzying array of curriculum-based interventions address essential skills such as conflict resolution, peer mediation, anger management, improved communications, social skills, decision making and impulse control, among others. Other programs emphasize recreational arts, mentoring, character development and after-school activities. These efforts--especially in high schools--are often linked with increased metal detectors and security, video surveillance, and collaboration with local police.
The connection between drug and alcohol abuse and violence has been well-established, and a variety of programs, including Drug Abuse Resistance Education (D.A.R.E.), have attempted to educate children about drug and alcohol abuse. However, they have had disappointing results (Ringwalt et al., 1994). Other innovative models such as D.A.R.T. (Development and Risk Together) promote wellness, with links to healthy development and prevention of violent behaviors (Guerra, 2003).
Complex interventions aimed at altering the social climate of the school are difficult to evaluate and replicate, partly because of the inherent complexity of social interventions. This creates difficulties in study integrity and problems with quantifying and standardizing data. The Blueprints for Violence Prevention system <www.colorado.edu.cspv/blueprints/> has a listing of various evidence-based programs for violence prevention. Replication is a requirement for the highest classification of effectiveness.
Very few programs ever meet these criteria. Various programs have had relatively disappointing results in the United States and elsewhere (Smith and Ananiadou, 2003). Table 1 lists some Blueprints model programs and one of our programs that have had promising results (Greenberg et al., 1998; Henggeler et al., 1998; Olweus et al., 1998; Sussman et al., 2002; Twemlow et al., 2001b, 1999; Webster-Stratton et al., 2001).
Wilson et al. (2003) reported on a well-designed meta-analysis of 221 school-based interventions designed to reduce aggressive behaviors. Eighty percent of the interventions were research-based demonstration programs and 90% of the studies were conducted in the United States. Very few studies of non-research routine interventions were available. However, effect sizes in those non-research studies were much lower than those for research interventions, where efforts were made to increase buy-in and fidelity. Respectable overall effect sizes of 0.25 for research programs translated into about a 7% reduction in fighting in schools. Specific intervention strategies had effect sizes grouping in the 0.25 to 0.33 range, with behavioral classroom management techniques and counseling showing the strongest effects. These were followed closely by improved academic services and programs promoting social skills, conflict resolution and communication skills.
Although consulting psychiatrists may feel slightly out of their element with interventions that do not focus on at-risk or ill children, the pervasiveness of the problem means that schools will often ask for a professional opinion (see Shafii and Shafii  for a summary of current research). Smith and Ananiadou (2003) summarized worldwide efforts to assess large-scale, school-based anti-bullying interventions. Efforts to prevent bullying have been moderately successful, and are more successful the younger the children involved are. Although most systemic interventions have addressed middle school and high school students, some have addressed programs for elementary schools (Twemlow et al., 2001b, 1996).
Many anti-violence approaches address a common power dynamic often seen in violent schools (Twemlow et al., 2001a). Power dynamic refers to a conscious or unconscious coercive pattern in which an individual or group controls the thoughts and acts of others by repeated verbal and/or physical humiliation. This dynamic may be subtle and unconsciously motivated, but the school climate usually reveals high levels of disciplinary referrals and unrest, with low teacher morale. Besides the bully and victim, systemic interventions must also address the issue of bystanders. The management of such pathology requires the coordinated effort of all school personnel and parents--such programs have been shown to improve academic performance (Fonagy et al., unpublished data).
Core elements include consultation with school staff to promote awareness of the power dynamic and to draw staff out in a non-blaming atmosphere. As part of the assessment of the school climate, several questions for students and teachers may be able to tease out a variety of dynamics. Table 2 provides some specific examples (Scott, 1999).
It is important to develop a non-coercive classroom-management or discipline plan, instead of consequence-based punishments. Innovative approaches promote insights into the bully, victim and bystander roles. For example, a teacher can define infractions as incidents involving the entire class in various roles, thus minimizing blame and maximizing insight.
Positive climate campaigns suggest alternatives to bully, victim and bystander behavior, while encouraging self-reflection, helping and understanding others. Individualized and creative use of buttons, magnets, posters, jokes, classroom banners and other accessories can often increase the awareness of children, teachers and parents.
Using peer and adult volunteers is an established method to assist child development in schools (Sprinthall et al., 1992). There is a need for guidance in managing bully-victim-bystander conflicts because--in contrast to traditional peer mediation where the two parties start from an equal position--there is already a power differential.
Physical education programs allow role playing and discussion, with specific techniques to deal with bullying such as defensive martial arts and/or a focus on competitive sports (Bell and Suggs, 1998; Twemlow et al., 1996). Guided discussion illustrates that win-win dynamics preserve the dignity and value of the losing party, whereas in win-lose situations, the loser is humiliated and often vengeful.
The overall goal of the psychiatric consultation is to soften coercive power dynamics by equipping victims with coping skills and guiding bystanders into helpful reflective roles. This reduces the power of bullies and reduces scapegoating of bullies by emphasizing that everyone occupies all these roles at different times and that each person involved is a part of both the problem and its solution.
Vernberg and Gamm (2003) noted that all people in the community need to be united by a core set of values, beliefs and attitudes that encourage more peaceful relationships in order for any anti-violence program to be successful. Most people believe that children should be spared from coercive power dynamics. However, coerciveness has long been used for social control in rituals such as union blackballing, college hazing, excommunication and corporal punishment. Victims are often felt to have done something to cause their own problems and that they are getting what they deserve when they are bullied. Public policies can create false divisions between educational outcomes (a school issue), mental health concerns (a community mental health issue), safety and health concerns (public health and social welfare issues), and legal matters (juvenile justice). These divisions compound the problem of attempting to integrate services and are fueled by the feeling of many teachers and administrators that the social and emotional needs of children are not their concerns.
Because successful prevention programs emphasize the social and emotional needs of children, strong support is required from school principals and superintendents who can then champion staff training by making resources and personnel time available. Gamm et al. (unpublished data) reported that in a sample of 60 teachers, 10% saw little or no value in such training and, in a number of instances, tried actively to undermine the anti-bullying program. We have found that about 75% of the teachers need to buy into a program for it to be effective. School anti-violence research is a developing field and the jury is still out on which programs are most effective. Attention to the issues and teacher commitment to facing and solving problems create powerful effects, with specific interventions still having detectable effects if properly implemented.
In sum, all violence prevention programs are defined by relationships: the ability to listen to ourselves, empathize with other people's experiences, use this information to solve mutual problems and learn to live together. Because schools are mirrors of the community from which children come, the partnerships of school, home and community are essential for the success of any school violence prevention program. A school psychiatrist who can comfortably enhance primary and secondary preventive models can be invaluable as a leader in guiding schools through choosing prevention programs based on their psychological merits.