Systematic analysis on school violence and bullying using data mining

Catherine Joy Escuadra, Krizia Magallanes, Sunbok Lee, Jae Young Chung

Research output: Contribution to journalArticlepeer-review

Abstract

Background: School violence and bullying public issues that have been related to several risk factors and negative consequences. These have been the subject of several research, education, and policy. An increased number of related publications were found from the past years. However, published narrative, systematic, and meta-analysis reviews are limited only to the synthesis and analysis of studies to certain numbers due to the manual and tedious process involved. Objective: To identify prevalent school violence and bullying research, and to describe the temporal trends of topics within the last three decades using data mining. Methodology: This study systematically mined abstracts from Web of Science and Scopus using the keywords “school violence” or “school bullying” to characterize the relevant literature by an efficient and effective approach. Pre-processing, word frequency and co-occurrence analysis, topic modeling, and trend analysis were done to detect semantic patterns and explore the yearly development of research themes. R packages like udpipe, stopwords, topicmodel were utilized to perform all data mining processes. Results: Primary categories of research topics were related to definition, forms, consequences, factors, and strategies of school violence and bullying. The most prevalent school violence and bullying research are related to “bullying phenomenon,” “physical school environment,” “sexual violence,” “physical violence” and “influencing factors.” While the least prevalent school violence and bullying research are related to “monitoring and evaluation,” “nature of conflict,” “victimization and self-identification,” “psychological violence and bullying” and “social context of bullying.” Overall, except for the topic on monitoring and evaluation (p 0.06), all school violence and bullying research topics showed an increasingly significant trend (p < 0.05) from 1995 to 2021. Conclusion: The use of data mining to understand the research publications about school violence and bullying revealed that there is a large amount of research since the 1990s. However, some topics of school violence and bullying, like monitoring and evaluation, still need further study.

Original languageEnglish
Article number107020
JournalChildren and Youth Services Review
Volume150
DOIs
StatePublished - Jul 2023

Bibliographical note

Publisher Copyright:
© 2023 Elsevier Ltd

Keywords

  • Data mining
  • School bullying
  • School violence

Fingerprint

Dive into the research topics of 'Systematic analysis on school violence and bullying using data mining'. Together they form a unique fingerprint.

Cite this