NCJ Number
248362
Journal
Journal of Youth and Adolescence Volume: 43 Issue: 9 Dated: September 2014 Pages: 1409-1420
Date Published
September 2014
Length
12 pages
Annotation
Our knowledge on adolescents' bullying behavior has rapidly increased over the past decade and it is widely recognized that bullying is a group process and, consequently, context-dependent.
Abstract
Only since recently, though, researchers have had access to statistical programs to study these group processes appropriately. The current 1-year longitudinal study examined the interplay between adolescents' bullying and likeability from a social network perspective. Data came from the evaluation of the Finnish KiVa antibullying program, consisting of students in grades 7-9 (N = 9,183, M age at wave 1 = 13.96 years; 49.2 percent boys; M classroom size = 19.47) from 37 intervention and 30 control schools. Perceived popularity, gender, and structural network effects were additionally controlled. Longitudinal social network analysis with SIENA revealed that, overall, the higher the students' level of bullying, the less they were liked by their peers. Second, students liked peers with similar levels of bullying and this selection-similarity effect was stronger at low levels of bullying. This selection effect held after controlling for selection-similarity in perceived popularity and gender. Third, students were likely to increase in bullying when they liked peers high on bullying and to decrease in bullying when they liked peers low on bullying. Again, this influence effect held after controlling for the effects of perceived popularity and gender on changes in bullying behavior. No significant differences between control and intervention schools appeared in the effects. The results are discussed in light of their theoretical and methodological implications. (Published Abstract)