Is Bullying on the Rise? Examining 12-Year Trends in Bullying and Related Attitudes among 4th-12th Graders

• In Event: 2-145 - Consequences of bullying and peer victimization

Integrative Statement

Introduction: Bullying is a significant concern and has received considerable attention over the past decade, leading some to believe that it is on the rise. However, there are relatively few multi-year large-scale studies of bullying to inform our understanding of such trends. This study examined the prevalence of bullying and related behaviors between 2005 and 2016, and explored whether any such changes varied as a function of school-level covariates.
Methods: Self-report data were collected on-line from 342,794 students (grades 4-12) in 109 Maryland schools on13 indicators of bullying and related behaviors across 12 years. The content for the survey was based on prior published and validated measures of bullying and perceptions of school safety. The data were weighted to reflect the school population and analyzed using longitudinal hierarchical linear modeling to examine changes over time.
Results: With regard to the prevalence of bullying, we observed general declines in the school-level rate of reported negative indicators (e.g., bullying others, victimization) and increases in the rate of reported positive indicators (e.g., feeling safe; see Table 1 for all 13 indicators). For example, the rates of self-reported frequent bullying victimization dropped from 28.5% in 2005 to 14.7% in 2016 (Cohen’s d = -0.31). Regarding bullying perpetration, it similarly dropped from 21.3% in 2005 to 7.4% in 2016 (d = -0.34). The multilevel analyses indicated 12 statistically significant slope effects. Specifically, being bullied (overall, β = -0.02, p < .001), and being victimized by being pushed (β = -0.03, p < .001), hit (β = -0.02, p < .001), threats (β = -0.02, p < .001), cyberbullying (β = -0.004, p < .001), and having rumors spread (β = -0.03, p < .001) decreased across the 12 years. Similarly, the slope of perpetrating bullying (β = -0.01, p < .001) and witnessing bullying (β = -0.03, p < .001) decreased. The slope for reporting feeling safe (β = .01, p < .001) and students reporting that adults do enough to prevent or stop bullying (β = 0.02, p < .001) all were significant (i.e., favorable). However, two iatrogenic effects emerged, whereby students reported more aggressive retaliatory attitudes (β = -0.01, p < .001) and less belonging at school (β = -0.002, p < .001). With regard to developmental trends, elementary schools reported fewer bullying concerns as compared to middle and high schools, and high schools showed significantly greater declines in witnessing (β = -0.019, p < .001) and perpetrating bullying (β = -0.007, p < .05) compared to middle schools. Reports of feeling safe in elementary schools had significantly less improvement over time as compared to middle schools (β = -0.01, p < .001). See Table 2 for additional results.
Conclusions: The covariate-adjusted models indicated that there continues to be improvement in bullying and related concerns, on 10 out of 13 indicators, with the greatest improvements occurring in more recent years (last 3-5). Additional analyses will be presented which aim to better understand these trends, and implications of these findings will be discussed.

Authors

• Catherine Bradshaw, University of Virginia
  Non-Presenting Author

• Tracy Waasdorp, Children's Hospital of Philadelphia
  Presenting Author

• Elise T. Pas, Johns Hopkins Bloomberg School of Public Health
  Non-Presenting Author