Thematic density and methodological trends in peer bullying theses in Turkey: A text mining study

In this study, English abstracts of 424 master's and doctoral theses on peer bullying in Turkey, which included the term peer bullying in their abstracts, were examined using Latent Dirichlet Allocation, a text mining method. In order to conduct an in-depth examination of theses written on peer bullying, English stopwords, general concepts on peer bullying, and academic research concepts were excluded from the analysis and text mining was applied. Four Topics were reached as a result of the analysis. Topic-1, which constitutes approximately 32% of the studies, consists of studies focusing on emotional problems of individuals and cyberbullying. Topic-2, which constitutes approximately 26% of the studies, is qualitative studies investigating bullying based on ethnic and cultural differences and their solutions in the classroom and on the basis of school administrators. Topic-3, which constitutes approximately 17% of the theses, is studies focusing on clinical/psychological problems and risks that cause peer bullying and occur as a result of bullying. Topic-4, which constitutes approximately 25% of the studies, is the studies focusing on the differences in demographic dimensions of peer bullying. Correlation analysis was conducted to examine the relationships between the Topics and positive relationships were reached. The results obtained show that peer bullying is a concept that should be examined not only individually but also socially.