Open Access
American Research Journal of Humanities and Social Sciences
ISSN (Online): 2378-7031
DOI: 10.46568/arjhss
Unmasking Risk Factors of Bullying Behaviors among Adolescents in Schools Using Logistic Regression
Abstract
The bulling behaviors among adolescents has become a serious issue in the United States. According to the U.S. Department
of Education’s National Center for Education Statistics (NCES), 22% of students from 12-18 years old are the victims of
bullying and about 28% of middle schools, 15% of high schools, and 10% of elementary schools reported bullying at least
once a week.Victims of bullying suffer from issues like anxiety, depression, and poor academic performance, while bullies
are more prone to future violence or substance abuse. Thus it is important to find the factors causing school bullying and
prevent it from happening.
In this research, we investigated possible risk factors for bullying behaviors at school among adolescents. We used the 2021 Adolescent Behaviors and Experiences Survey data with features ranging from demographic information to the adolescents’ family condition. The response variable is whether an adolescent has been bullied at school during the past 12 months.
After processing the dataset, we built a logistic regression model to identify possible factors associated with school bullying. The logistic regression model has achieved an AUROC score of 0.76, with 74.5% true positive rate (TPR) and 11.2% false positive rate (FPR). The findings reveal that students who experience electronic bullying and racial discrimination are at a heightened risk of physical bullying, while those carrying weapons or having strong social connections in school are less likely to be bullied. From the findings, the research proposes some recommendations for teachers and parents to identify and support the victims of school bullying, preventing the happening of school bullying.
In this research, we investigated possible risk factors for bullying behaviors at school among adolescents. We used the 2021 Adolescent Behaviors and Experiences Survey data with features ranging from demographic information to the adolescents’ family condition. The response variable is whether an adolescent has been bullied at school during the past 12 months.
After processing the dataset, we built a logistic regression model to identify possible factors associated with school bullying. The logistic regression model has achieved an AUROC score of 0.76, with 74.5% true positive rate (TPR) and 11.2% false positive rate (FPR). The findings reveal that students who experience electronic bullying and racial discrimination are at a heightened risk of physical bullying, while those carrying weapons or having strong social connections in school are less likely to be bullied. From the findings, the research proposes some recommendations for teachers and parents to identify and support the victims of school bullying, preventing the happening of school bullying.
