the concrete reality of an event. The ideas and cognitions that operate in the heads of offenders are important as they become agents of violent events. The aim here is to bridge these different levels of analysis.

Our interests go beyond what might have caused the individual shooting and what might have prevented it. We are also interested in the incident as a cause of both a community and a policy response. Table 8-1 is a representation of the committee’s analysis. Although lacking precision, the table is meant to convey the extreme rarity of school rampage violence within the larger picture of lethal school violence, other serious youth violence, youth violence in general, and violence in general.

In reviewing the cases, the case authors and committee members developed multiple hypotheses about what might be driving the behavior and checked them against the data collected in the six cases. For example, in one case an offender had a more successful older sibling; in another the family of the offender had recently moved into town; in another the offender knew a great deal about a previous school rampage shooting, and those attending school were organized in a large number of subcultures with few overlapping members and little interaction with one another. We then considered whether there was a plausible causal connection between the feature observed and the likelihood that the violent incident would have occurred or would have taken a particular shape. If the factor showed up in any of the other cases and seemed to be working in the same way, then there was a plausible hypothesis about an important causal phenomenon. If little such evidence was found, then the hypothesis seemed less likely to be true, or less important even if true, because its impact could not be seen across the cases examined.

We also considered some common hypotheses—for example, that the offenders had been bullied and the attack was an attempt to get back at those who bullied them, or that the schools in which the incidents had occurred had allowed cliques to form and to skirmish with one another in the school—and looked at the cases for any examples. If none of the cases was characterized by such a feature, that cast significant doubt on either the truth or the importance of that factor as a contributing cause of the violence.

These methods, although not systematic enough to confirm or refute alternative causal explanations, could be viewed as an effort to undertake the preliminary scientific task of generating causal hypotheses and doing the rough work of casual inspection that helps make some hypotheses seem stronger than others. In this approach, part of the task is to reduce confidence in any particular claim by imagining the variety of things that could be true and then ordering the claims according to some rough sense of plausibility.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement