Predictors of Fighting among Rural Elementary School Students

Abstract

Violence in the school setting is receiving increased attention by educators, parents, and the media. It is important for health educators to understand predictors and risk factors of fighting, so appropriate intervention programs can be developed and implemented. Little research has been conducted concerning fighting among rural children. This study identified predictors of fighting among rural elementary school children. In 1997, data were collected concerning several Healthy People 2000 priority health behaviors from 1,187 kindergarten to 6th grade children in southern Illinois. Reliability of the behavior subscale was .63 as estimated by Cronbach's alpha. Data were analyzed using bivariate odds ratios and logistic regression. Survey findings suggested that having been involved in a fight “where someone was punched, hit, kicked, or hurt in some way” was related to a number of risk factors. Unadjusted odds ratios (ORs) showed that using smokeless tobacco (OR = 5.5), smoking (OR = 3.6), drinking alcohol (OR = 3.3), gender (OR = 3.3), and having carried or shot a gun (OR= 3.1) were significantly related to having been involved in a fight. A logistic regression model based on significant bivariate risk factors explained approximately 44% of the variance of having been involved in a fight (Somers' D = 0.443). The logistic regression model suggested that use of smokeless tobacco, gender, and drinking alcohol were the strongest predictors of having been involved in a fight, with adjusted ORs of 3.4,2.5, and 2.0, respectively. These data help identify children at high risk for being involved in a fight. Health educators and others charged with the development of intervention programs can use these data to develop relevant programs for this target population.