A randomized controlled trial of instructional coaching in high-poverty urban schools: Examining teacher practices and student outcomes.

Job-embedded professional development is needed to effectively and efficiently enhance teachers' use of evidence-based practices in high-poverty urban communities. This study employed a three-cohort, waitlist controlled, randomized block design to investigate the effectiveness of the Classroom Strategies Coaching Model (CSC) in 14 high-poverty urban elementary schools. The CSC Model is guided by observations of teachers' instructional and behavioral management practices as measured by the Classroom Strategies Assessment System. Primary dependent measures included teacher use of evidence-based practices, student academic engagement, and teacher ratings of class wide student academic and behavior functioning along with perceived instrumental support, emotional support, and stress. The sample included 2195 students and 106 teachers randomly assigned to CSC coaching or waitlist control. Multilevel negative binomial modeling revealed that teachers in the CSC coaching condition had significant improvements in the frequency of academic praise (used 1.74 times more frequently) and behavior praise (used 2.10 times more frequently) as compared to teachers in the waitlist control condition. Multilevel linear models revealed that, relative to the waitlist control condition, teachers in the CSC coaching condition demonstrated significant improvements in quality of instruction (d = 0.52), behavior management (d = 0.60), and class wide student academic engagement (d = 0.41). Teachers reported significant improvements in class wide student academic (d = 0.96) and behavioral functioning (d = 1.24), instrumental support (d = 0.90) and emotional support (d = 1.04). No change was found for teacher stress. Implications for research and practice are reviewed.