Disentangling the Impact of a Classroom-Based Randomized Controlled Trial: The Moderating and Mediating Roles of Teacher Stress and Classroom Quality

Background and Purpose: Research has suggested teachers' personal and work-related stress and classroom quality can affect the effectiveness of classroom-based interventions. Efforts have been made to disentangle the black box of causal mechanism on how students can benefit from classroom-based interventions, particularly those of teacher-focused. This study uses data from the Chicago School Readiness Project (CSRP) to examine whether the impact of CSRP on children's social-behavioral and academic outcomes is moderated by the initial measures of teachers' stress and classroom quality at baseline, and is mediated by these measures after CSRP intervention.

Methods: The CSRP aimed to help teachers improve effective classroom management and reduce their stress as a way to support the development of self-regulation and opportunities for learning of low-income minority preschoolers. Using a clustered randomized controlled trial (RCT) design in Head Start programs, the CSRP provided one-year multifaceted services, including teacher training on behavior management strategies, the placement of mental health consultants in classrooms to coach teachers in implementation, and stress-reduction workshops for teachers. The CSRP included a total of 90 teachers who taught 602 children in 35 classrooms from 18 Head Start sites.

Teacher stress and classroom quality were measured at baseline in the fall and after the intervention in the spring of Head Start year. Teachers' self-reported stress was assessed by the Child Care Worker Job Stress Inventory (e.g., job control, demands, and rewards) and the K6 on psychological distress (e.g., feeling nervous, hopeless, restless, and worthless). Classroom quality was assessed by trained observers using the Classroom Assessment Scoring System (e.g., emotional climate and teacher sensitivity) and the Early Childhood Environment Rating Scale-Revised. Children's outcomes include social-behavioral and academic measures that had been continuously assessed in Head Start year, kindergarten, and third grade, including teacher-student relationship (e.g., closeness and conflict) using the Student-Teacher Relationship Scale, internalizing and externalizing behavior problems using the Caregiver-Teacher Report Form, and academic outcomes (e.g., language, literacy, and mathematical thinking) using the Academic Rating Scale.

Hierarchical linear modeling (HLM) is employed to account for the multilevel structure of data, in which children were nested within classrooms, and classrooms were nested within Head Start sites. The covariates include baseline characteristics of children, families, teachers, and classrooms, as well as site-fixed effects.

Results: Previous analyses have shown the beneficial effects of CSRP on reducing teachers' stress and improving classroom quality and child outcomes in Head Start year. Some evidence suggests the CSRP effects on children's outcomes vary by the levels of teacher stress and classroom quality at baseline, and the CSRP has some indirect effects through changes in teacher stress and classroom quality in Head Start year on children's outcomes in Head Start, kindergarten, and third grade.

Conclusions and Implications: Findings provide new evidence on the moderating and mediating roles of teacher stress and classroom quality in classroom-based early interventions on low-income children's outcomes in the short and long terms. Implications can help policymakers and practitioners identity the buffering and inhibiting factors in program implementation, and improve the effectiveness of classroom-based interventions.