Bullying victimization among school-attending adolescents in Pakistan.

M. Shaikh
2013-09-01
Abstract:Madam, bullying among adolescents is a serious public health problem. Several studies have reported its association with age, sex, loneliness, inability to sleep, absence in classes, and suicide consideration.1-4 There are no studies on adolescent bullying in Pakistan. To describe factors associated with self-reported bullying and its prevalence, we used data from the two-stage cluster sample-based nationally representative Global School-Based Health Survey for Pakistan, which was conducted in 2009 by the Ministry of Health in collaboration with the World Health Organization and Centers of Disease Control and Prevention, United States.5 Pakistani students of class 8-10 who had been victimized in the past 30 days were included in the study. Design-based analysis using STATA-12 was done using bivariate and multivariate logistic regression. Factors found statistically significant at p<0.10 level on bivariate logistic regression were used for multiple logistic regression. The overall prevalence of bullying victimization in the past thirty-days was 41.3% [95% confidence interval (CI): 34.4%, 48.3%] (n = 4676). In male students, the prevalence was 45.1% [95% CI: 39.3%, 51.0%], while among females the prevalence was 35.5% [95% CI: 21.7%, 49.3%] Table 1 provides the description of the respondent's demographic and psychosocial factors, as well as the association with bullying victimization in bivariate and multivariate analyses. Age and sex were not found to be statistically significant in the bivariate logistic regression model, and hence were not included in the final multivariate logistic regression model. Final model included having felt lonely in the past twelve months, worrying to the point of not being able to sleep in the past twelve months, having missed classes without permission in the past thirty days, and having seriously considered attempting suicide in the past twelve months. Results of the goodness-of-fit test concluded that this model was a good fit for the data.
What problem does this paper attempt to address?