| Purpose. This study is designed to address significant gaps in the medical and public health literature with regard to the study of adolescent weapon carrying behaviors through the secondary analysis of two waves of public-use data from The National Longitudinal Study of Adolescent Health. The specific aims of this study include the following: (1) to explore differences between individuals who do not carry weapons, those who carry weapons but do not use them, and those who both carry and use weapons; (2) to examine the role of school-connectedness on weapon carrying behavior; and (3) to examine transitions in weapon-related behavior as a function of transitions in predictors across multiple ecological domains.;Methods. The data from this study was obtained from Wave 1 (N=6,024) and Wave 2 (N=4,884) of the Add Health public-use data. Analysis of weapon carrying behavior was conducted using multinomial logistic regression of a three-level outcome variable (No Weapon Carrying, Carry-only, Carry-use). Analysis of transitions in weapon carrying behavior was conducted using multinomial logistic regression of a separate four-category outcome (No change-no weapon carrying either wave (Reference), No change-weapon carrying/use both waves, Worsening, Improvement). A fixed-effects models was utilized to explore the influence of transitions in selected predictor variables on transitions in the outcome variable.;Results. In Wave 1, 6.3% of respondents reported weapon-carrying without use and 10.3% reported weapon-carrying with use. In Wave 2, 2.5% of respondents reported weapon-carrying without use and 9.5% reported weapon-carrying with use. School-connectedness was significantly associated with weapon carrying behavior in cross-sectional analysis of both Wave 1 and Wave 2. Wave 1 school-connectedness was also significantly associated with transitions in weapon-carrying behavior from Wave 1 to Wave 2. Fixed effects models did not reveal a significant association between transitions in school-connectedness and transitions in weapon-related behavior. Gender was significantly correlated with weapon-related behavior in all analyses.;Conclusion. This study demonstrated subtle but important differences between weapon-carriers and weapon-users. Violence-related indicators were strongly correlated with weapon-related behaviors across all analyses highlighting potential targets for prevention and intervention. These results underscore the need for more comprehensive research on adolescent weapon-carrying behavior through longitudinal studies. |
