Social relationships in multicultural schools: Bullying and victimization

The study investigated victimization and bullying in Native Austrians and three groups of immigrants living in Austria. The sample comprised 280 pupils (133 girls, 147 boys) in school grades 5-8. Data were collected via self-assessments and peer nominations. For data analysis, in the first step, a variable centred approach using MANOVAs was applied to analyse whether there are gender and cultural group differences in levels of victimization and bullying. Native Austrian pupils were found to score higher in victimization and bullying according to self-assessments and they were also found to be more often nominated as victims by their peers than the three immigrant groups. However, the correlations between self-assessments and peer nominations were quite low for both victimization (r = .38, p = 5.01) and bullying (r = .20, p = 5.01) indicating that only a small number of pupils could be identified as "true'' victims or "true'' bullies on the basis of both measures. Thus, a person-centred approach using configural frequency analyses (CFAs) was applied to investigate whether first- or higher-order relationships exist between self-assessment (as bully or victim), peer-nomination (as bully or victim), gender, and cultural group. For victimization, two typical variable patterns could be identified: More Austrian boys and more Turkish boys than expected by chance were found to be "true'' victims (identified by self-assessment and peer nomination). For bullying, one type could be identified. More Austrian boys than expected by chance were found to be "true'' bullies. In addition, these victim/bully subgroups were compared for positive peer relations and self-esteem.