Prevalence and Predictors of Student- and School-Level Weapon-Related Behaviors

Background and Purpose

The public and academic interest focuses on a relatively small number of tragic and shocking mass shootings in schools, such as Columbine, Sandy Hook, Parkland. This focus, however, has severely constrained the conceptual and empirical discourse on other weapon-related issues facing students in schools. In this presentation, we aim to expand the discussion to include issues such as weapon-related incidents that do not result in homicide or even injury, the use of many different types of weapons in addition to guns, and students’ experiences of being threatened with a weapon or even seeing and hearing about a presence of a weapon on school grounds. We employ our theoretical framework of 'bullying and school violence in evolving contexts' to examine the prevalence of multiple types of weapon-related behaviors, their interrelationships, and what are their predictors, in both student- and school-level analyses.

Methods: 

A secondary analysis of the California Healthy Kids Survey (CHKS) 2011-2013.The survey samples students in fifth, seventh, ninth and 11^th grades biannually, and two-years data wave provides a representative sample of  students and schools of California. N= 528,436 students and 1,849 schools.

Results:

About 4% of students brought a gun to school, 8% knife, 6.8% were threatened/injured with a gun or knife, and 23.3% saw a weapon in school. On the school level, 10.2% of schools had no reports of students carrying a gun to school, whereas 3.3% of schools had 15% or more of their students report carrying a gun to school. In almost 90% of the schools, at least 15% of students reported seeing a weapon in school. The intercorrelations between the various weapon-related behaviors are substantial and are higher at the school level compared with student level. For instance, on a student level the relationships between bringing a gun to school and being threatened by a weapon is r = .43 and on the school level r = .73. Bi-variate and multivariate regressions on the student- and school levels examined predictors of weapon-related behaviors and showed high levels of explained variance (e.g., R²= .63 for carrying a gun). For instance, correlation between gang membership and carrying a gun was r=.25 on school-level, and r=.58 at the school level, showing that schools with many gang members also had many more students reporting carrying a gun to school. Victimized students were more likely to bring weapons to school.

There were many differences between student-level and school-level findings. For instance, although the race/ethnicity of individual students was associated very weakly (or not at all) with their involvement with weapons, at the school level, schools with more African American and Hispanic students also had more reports of weapon involvement.

Conclusions and Implications

A small number of schools have a very large number of weapons and require immediate attention before a student is hurt. This requires a public health approach that identify vulnerable schools and focus resources in these schools to reduce weapon-related behaviors. Future prevention and research should include all forms of weapon victimization.