Anti-bullying programme-tweaking based on prevalence estimates at school- and classroom-level in a Swedish municipality

Research topic/aim

Estimated prevalence of bullying depends on the measurement strategies involved (Bjereld, Augustine & Thornberg, 2020). While the Nordic Countries, in international comparisons, have had lower prevalence estimates, rates in Sweden have shown increased prevalence in recent years (op.cit.). Differences in measurement methods also reflect differing conceptualisations of bullying.

Theoretical framework

At least nine bully types have been described (Goldbach, Sterzing & Stuart, 2018). When “low frequency involvement” has been included, accumulated prevalence rates of up to 70% have been posited (op.cit.). In countries with lower prevalences of bullying, between-school differences are also lower (Johansson, Myrberg & Toropova, 2022). The present study uses a well-tried estimate of school-children’s exposure to intentional degrading treatment and bullying victimisation (Flygare, Gill & Johansson, 2013).

Methodological design

The analysis undertaken seeks to replicate an earlier population study (all 4th to 9th grade school-children in a Swedish municipality). In that study (Gill, Simonsson & Matton, 2017), data at school- and classroom-level for boys and girls, deemed post hoc to be victims of bullying, at two measurements within the school-year (October 2016, and March 2017), were used to “paint a picture” of the intervention challenge. The intervention programme (“Gävlemodellen”, Gill, et al., 2016) was offered to municipal schools for voluntary participation in 2013. An initial group of 6 municipal schools expanded to include all schools within the municipality by 2016/17. The present study will follow-up the 2016/2017 population study, targeting all 7200 pupils, in 29 schools, distributed among 280 classrooms. The analysis will complement the VR-financed (2023) research project “Associations between the social climate in school, classroom and school class, peer victimization, bullying and student engagement over time” (Thornberg, 2023). Specifically, patterns of distribution of victims, at 6-year-follow-up, will be compared with a view to possible revisions and developments in the intervention design.

Expected conclusions/findings

The measure used to estimate bullying is indirect and does not use an a priori prescriptive definition. Pupils are asked if certain things have happened to them, “in the last couple of months” (teased, hit, pushed, excluded, nasty messages on social media etc. Pupils are then asked for their perception of why the event (or events) occurred. Respondents are categorized (discrete measure), as bullied (victims) or not (see Flygare, Gill & Johansson, 2013, and Swedish National Agency for Education, 2011 for details). Descriptive statistics will demonstrate the importance of not basing bullying prevention on a “general case”. Victimhood varies over time, by gender, and within and between schools. The goal is to explore how differences between the 29 schools and between classrooms within schools (from no victims to gender specific clusters) can be used for “program tweaking” the intervention strategy.

Relevance to Nordic educational research

Reduced bullying prevalence in Nordic countries creates new intervention challenges. Lessons and challenges from population data in a Swedish municipality can have relevance for all 1053 Nordic municipalities.