Exploring Mechanisms of Effective Teacher Coaching: A Tale of Two Cohorts From a Randomized Experiment

Teacher coaching is considered a highquality professional development opportunity that emphasizes job-embedded practice, intense and sustained durations, and active learning (Desimone, 2009; Russo, 2004). Generally, coaches observe teachers in their classes and then provide targeted feedback aimed at improving these practices. Coaching is closely related to teacher mentoring, which also targets instructional improvement through one-on-one relationships between a novice and more veteran teacher; however, mentoring often focuses on providing general advice rather than responding directly to observed classroom practices (Wildman, Magliaro, Niles, & Niles, 1992). To date, experimental evidence on teacher coaching has been largely positive (Allen, Pianta, Gregory, Mikami, & Lun, 2011; Campbell & Malkus, 2011; Neuman & Cunningham, 2009; Powell, Diamond, Burchinal, & Koehler, 2010; Sailors & Price, 2010). This is particularly noteworthy compared with mixed results on the effectiveness of school workshops and trainings that characterize much of the professional development offerings provided to teachers (DarlingHammond, Wei, Andree, Richardson, & Orphanos, 2009; Yoon, Duncan, Lee, Scarloss, & Shapley, 2007), as well as more intensive development and mentoring opportunities (Garet et al., 2008; Garet et al., 2011; Glazerman et al., 2008). At the same time, this research base is new, and little is known about the effectiveness of specific design features or practices of different coaching models. Understanding which program features are critical for success is an important line of inquiry for the continued improvement of teacher professional development. This is especially true given the substantial costs of coaching (Allen et al., 2011) and the fact that coaching is 579487 EPAXXX10.3102/0162373715579487Blazar and KraftExploring Mechanisms of Effective Teacher Coaching research-article2015