Evaluation of a Principal Training Program to Promote Safe and Civil Schools

Statement of the Problem: Considerable research has shown that school principals play a central role in creating safe and civil school environments, yet no evidence-based training programs exist to support principals in developing these skills. The purpose of this proposal is to conduct an efficacy study of Safe and Civil School Leadership (SCSL) plus START, a professional development program designed to equip school principals with skills for fostering positive school climate and safety. Rooted in sound theory and practices, SCSL plus START provides principals with explicit training and resources to create effective disciplinary structures and student supports using data. Although SCSL plus START was developed over a decade ago, its effects on school safety and climate have never been evaluated in a controlled trial.

Subjects: Sixty principals from secondary schools in the Puget Sound Educational Service District (PSESD) with high rates of disruptive behaviors will be recruited to participate in a randomized control trial. The PSESD is composed of 35 school districts in King and Pierce county servicing 397,000 students (39% of the entire K-12 population in Washington state). PSESD includes districts from rural, suburban, and urban settings with a wide range of student sociodemographic characteristics.
Partnerships: The PSESD is a key partner. Their letter of support indicates strong agreement with the importance of the project and a clear understanding of the commitment involved.

Research Design and Methods: The efficacy of SCSL plus START will be evaluated with a group randomized trial comparing 30 schools implementing the SCSL program with 30 control schools. We will collect school records and principal, teacher, and student ratings of school safety, climate, student compliance, and leadership behavior over a two-year period. Assessments will occur prior to intervention and six months, twelve months, and eighteen months after intervention.

Analyses. We will use a two-level hierarchical linear model (i.e., a random intercept model) representing students nested within schools to compare the two conditions on school safety (e.g., prevalence of teasing and bullying, student victimization) and student behavior outcomes as assessed post-intervention with regression adjustment for baseline assessments. Student and school level covariates will be included in the final models. Multilevel SEM will be used to test mediation hypotheses based on our theory of change.

Products, Reports, Archiving: Results will be summarized and widely disseminated through peer-reviewed publications, conference presentations, and a digital repository. We have a detailed plan to ensure data archiving and open access.
ca/ncf