Reducing High School Suspension Rates and Improving School Climate: Evaluating Wisconsin's Safe and Supportive Schools Program

Introduction:  This paper reports on the dissemination, implementation and effectiveness of strategies to reduce high suspension rates and to improve school climate in 55 high schools.   Under US Department of Education  funding, the Safe and Supportive Schools program has the goal of improving conditions for learning by increasing school safety, school climate and student engagement.  Schools have been selected on the basis of high rates of suspension and disruptive drug and violence related incidents.  The primary grantee is the state Department of Public Instruction (DPI); DPI is providing funding, professional development and technical assistance to high schools to develop, implement and evaluate interventions at the district and building levels.  Evidence-based interventions including modified administrative policies and student oriented programs are being implemented.  The most common interventions are Positive Behavior Intervention Supports (PBIS), bullying prevention, Restorative Justice, and pupil non-discrimination policies. 

Methods:  Our evaluation is using annual on-line surveys of 9^th and 11^th grade students and accessing school-level data on suspensions, behavioral incidents, and program implementation.  Student surveys include measures of perceived school climate, experience of violence, bullying and harassment, personal alcohol and drug use, commitment to school work, perceived safety, and perceived rule enforcement.   Implementation data come from school consultants and annual school reports.  Data on school-level suspensions, graduation rates, and standardized test scores come from DPI reporting systems.  The design is multi-level, with 4 waves of sequential cross-sectional student data (annual n= 22,000) nested within schools (n=55).  Comparison high schools are also being examined for school-level variables, using propensity score adjustments to increase comparability.

Results: To date we have found significant reduction in suspension rates within schools, but minimal student-level change in risk behaviors and perceptions of school safety or climate.  Further analyses regarding the effect of variations in  implementation and fidelity, and the relationship over time between student level variables and school-level administrative actions (e.g., suspension rates), will be presented.

Conclusions: Significant reductions in suspension rates reflect changes in administrative behaviors of school officials and district policy which have been influenced by this large scale dissemination initiative.  Further analyses will examine whether administrative changes have had an impact on student perceptions and behaviors, which appear to be more resistant to change than administrative behaviors.  Lessons learned regarding dissemination and effective implementation of preventive interventions in multiple high schools will be discussed.