Triangulating School Climate: Areas of Convergence and Divergence in Multilevel Perspective

Jessika H. Bottiani, Sarah Lindstrom Johnson, Katrina J. Debnam, Catherine Bradshaw

Introduction: School stakeholders, including students, staff, and parents, often differ in their perceptions of school climate, making a multiple informant approach to the measurement of climate critically important. Yet, relatively little research has utilized a fourth perspective with potential added value – that of trained, independent observers. The present study seeks to understand areas of convergence and divergence between students’ and independent observers’ viewpoints of school climate within a three-level model. This perspective offers the opportunity to understand the measurement and influence of climate at both the classroom- and school-level.

Methods: Students in 50 high schools participating in the Maryland Safe and Supportive Schools Initiative (MDS3) were surveyed using a measure of school climate (Bradshaw et al., 2014) consistent with the U.S. Department of Education’s Safety, Engagement, and Environment framework. Independent observers were trained to reliably assess student and teacher interactions in the classroom using the Assessing School Settings: Interactions of Students and Teachers (ASSIST; Rusby, Taylor, & Milchak, 2001) observational measure. Data at the school level were collected via the School Assessment for Environmental Typology (SAfETy), an observational measure of the school physical environment. A stepwise approach to build the three-level model in Mplus was used, adding one variable and one level at a time in order to be sensitive to the stability and significance of findings (Raudenbush & Bryk, 2002).

Results: The final model demonstrated adequate fit, χ2(40) = 1172.35, p < .001, CFI = .93, RMSEA = .041. Adjusting for student and school characteristics (e.g., race/ethnicity and socioeconomic status), we found observed classroom teacher responsiveness was significantly associated with higher student report of engagement and environment, but not safety. Indicators of disorder in the physical environment (e.g., graffiti) were significantly associated with lower student-rated safety and environment. However, school- and classroom-level observations of the schools’ use of positive behavioral expectations were not significantly associated with each other, suggesting potential differences between school- and classroom-level implementation of positive behavior support.

Conclusions: The findings enhance our understanding of areas of convergence and divergence of independent observations of climate in relation to students’ viewpoints and in multilevel context. Implications for future integration of observational measures of climate in comprehensive school climate assessments will be discussed.

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