RESUMO
BACKGROUND: Bullying is a widespread phenomenon that has captured attention from mental health researchers. Several studies have assessed bullying prevalence with some methodological concerns. OBJECTIVES: Preliminary, we analyzed the psychometric properties of two bullying scales for victimization (the multidimensional peer victimization scale - MPVS) and for perpetration (the bully subscale of the Illinois bully scale - IBS-B); then, we estimated bullying prevalence; finally, we evaluated the effect of gender and classroom on the phenomenon. PARTICIPANTS AND SETTING: 2959 students from the metropolitan city of Naples constituted the sample. METHODS: Data collection was obtained using a multi-assessment approach that included both single-item questions and intensity scales in order to compare the two methods. RESULTS: The two scales resulted valid and showed good reliability. The MPVS displayed a 1-factor second order model. The IBS-B had a mono-factorial structure. Both showed full invariance for gender and classroom. Prevalence of victimization was 37% whereas that for perpetration was 21%. As expected we obtained several bullying prevalence results depending on the specificity of questions and in particular repetitiveness of episodes. There was a good correspondence between results of single-item questions and multi-item scales. Finally results demonstrated several differences for gender and classroom attended. CONCLUSION: In this epidemiological study the multi-assessment approach identified different but complementary features of bullying phenomena. The use of the two measurement approaches allowed us to obtain more precise and exhaustive information on bullying prevalence and compare it with previous findings.