The role of school-based interventions in mitigating polyvictimization: Evidence from a national sample of Tunisian...

Polyvictimization, when children experience multiple forms of victimization, is a prevalent global health problem associated with multiple forms of negative developmental outcome. The paper argues that in conservative cultural contexts like that of Tunisia, school responsiveness to sexual harassment may be a critical indicator of a school's formal and informal interventions to protect children. Hence, in study 1 the paper develops and validates a measure of school social control of sexual harassment, and in study 2 tests the association between school-level social control of sexual harassment and child polyvictimization in schools. Study 2 also hypothesizes that school level social control of sexual harassment may weaken the link between polyvictimization at home and polyvictimization at school, and that individual secrecy about victimization exacerbates in-school polyvictimization.

Examine a new measure of school social control of sexual harassment. Test for a relationship between school level social control of sexual harassment and in-school polyvictimization in a national sample of Tunisia. Test for an interaction between school-level social control of sexual harassment and polyvictimization at home. Test for individual and school-level relationships between victimization secrecy and in-school polyvictimization.

The sample is a representative, stratified, random, four-stage probability proportional to size (PPS) cluster sample of 3098 preparatory and secondary students aged 12-18 in 40 schools, and 571 staff from the from the same 40 schools in Tunisia.

Factor analysis and criterion validity are explored for the new social control of sexual harassment scale. The school staff sample was randomly divided into a training sample and a holdout sample. EFA and CFA models were run on the training sample; CFA was run on the holdout and full sample. Based on these analyses maximum likelihood factors were extracted from the CFA. Multilevel models are used to test the hypotheses, including relationships between dimensions of social control of sexual harassment and child polyvictimization in Tunisian schools. Survey weights are used to provide national prevalence estimates of school polyvictimization in Tunisia.

Polyvictimization prevalence in Tunisian schools is high, with 63.8% of students reporting at least one victimization in school. The 8 item Social Control of Sexual Harassment Scale broke into two dimensions; one factor involving the school's local responses to sexual harassment and a second factor involving escalation and reporting to authorities responses. School-level local responses were associated with less in-school polyvictimization. Sensitivity analysis revealed that the main driver of the significant finding was staff reporting of sexual harassment to victims' parents. There was a positive interaction between reporting to parents and polyvictimization at home, indicating school reporting sexual harassment to parents appears to be more effective in protecting children who are not victimized at home. Individual secrecy about victimization was associated with more polyvictimization. Distance to school, attending an urban school, and attending a school with older staff were all risk factors for in-school polyvictimization. Female gender and attending a school with a higher percentage of female staff were protective factors.

Social control of sexual harassment, particularly telling parents, may be at the forefront of protecting children from polyvictimization, both via direct intervention and signaling that violence against children will not be ignored. However, school responses appear to better protect children who are not victimized at home. Victimization secrecy is associated with more polyvictimization. Schools should work to improve local and reporting responses to all forms of polyvictimization, and help children to more readily disclose victimization.