A Cluster-Randomized Controlled Trial of the CharacterStrong SERVE Model: Effects on Student-Reported and Teacher-Reported Outcomes

Middle school is an opportune developmental period for school-based prevention. Teacher relationship-focused practices have significant potential to influence critical enablers to school success, such as belonging and academic engagement. However, there are few programs that operationalize relationship practices teachers implement to promote students’ sense of belonging and academic engagement. The purpose of this experimental study was to evaluate the effects of CharacterStrong’s SERVE Model of Adult Relationship Practices on middle school students’ sense of belonging, academic engagement, and behavior. This study took place in two large urban middle schools serving a diverse population of students. Thirty teachers were randomly assigned to either intervention (SERVE Model; ntreatment teacher= 15) or attention control condition (ncontrol teacher= 15). An average of 6 to 8 students were randomly recruited from each teacher’s classroom (nstudent total= 185). Four student outcomes assessed at baseline and posttest: belongingness, academic engaged time, truancies, and behavior discipline. Three-level general and generalized linear mixed models were used to estimate the effects of SERVE Model on student outcomes and to probe treatment effect heterogeneity across student subgroups based on baseline status on each outcome. The SERVE Model was associated with significant improvements in belongingness, academic engaged time, behavior discipline, and truancies.