A Multilevel Analysis of School Influences on Student Mental Health: The Role of School and Classroom Climate

Katholiki Georgiades, Anna Kata, Laura Duncan, Michael Boyle

Introduction: Reducing inequalities in child mental health constitutes an important public health priority in Canada, and schools represent an ideal context for effective action; they are the first point of referral for many students with mental health problems and may offer mental health promotion and prevention opportunities. However, studies in Canada have yet to examine the extent to which schools affect mental health and their potential to reduce inequalities. The primary objective of our school mental health study is to examine the impact of schools on student mental health by estimating between school and classroom differences in student mental health outcomes and the extent to which school and classroom climate account for these differences.

Methods: A representative sample of 248 schools across Ontario, stratified by low, medium and high median family income at the community level, participated in school-wide surveys. Anonymous student surveys were administered to all consenting students in grades 6-8 and a random sample of secondary students (3 classrooms per grade in secondary school). All elementary teachers, the secondary school teachers of the classrooms selected, and the school principal were asked to complete surveys of school programs, practices and climate. The sample includes 31,124 students in grades 6-12; 3,373 teachers and 206 principals from 248 schools across Ontario.

Results: Multi-level models were used to calculate the variability in student mental health outcomes attributable to between classroom- and -school differences, expressed as intraclass correlations (ICCs). ICCs converted to percentages for internalizing problems were 3.1% at the classroom level and 4.0% at the school level. Comparable estimates for externalizing behavior problems were 4.1% at the classroom level and 1.6% at the school level. Adjusting for the socio-economic and demographic composition of students reduced the magnitude of the ICCs, although significant variability remained at the school and classroom level. School and classroom climate explained between 10-40% of the remaining variability.

Conclusions: Evidence arising from this work demonstrates moderate to strong associations between the school and classroom setting and student mental health outcomes, providing an empirical basis to support interventions delivered in these settings. School and classroom climate account for a significant portion of the variability in student mental health outcomes and represent important targets for preventative intervention strategies.

This abstract was submitted to the 2017 Society for Prevention Research Annual Meeting.