The Effect of Teacher Coaching on Instruction and Achievement: A Meta-Analysis of the Causal Evidence

Matthew A. Kraft,David Blazar,Dylan Hogan
DOI: https://doi.org/10.3102/0034654318759268
2018-02-22
Review of Educational Research
Abstract:Teacher coaching has emerged as a promising alternative to traditional models of professional development. We review the empirical literature on teacher coaching and conduct meta-analyses to estimate the mean effect of coaching programs on teachers’ instructional practice and students’ academic achievement. Combining results across 60 studies that employ causal research designs, we find pooled effect sizes of 0.49 standard deviations (SD) on instruction and 0.18 SD on achievement. Much of this evidence comes from literacy coaching programs for prekindergarten and elementary school teachers in the United States. Although these findings affirm the potential of coaching as a development tool, further analyses illustrate the challenges of taking coaching programs to scale while maintaining effectiveness. Average effects from effectiveness trials of larger programs are only a fraction of the effects found in efficacy trials of smaller programs. We conclude by discussing ways to address scale-up implementation challenges and providing guidance for future causal studies.
education & educational research
What problem does this paper attempt to address?