Association for Public Policy Analysis & Management

Teacher preparation has the potential to profoundly improve the instructional quality and corresponding student outcomes. To date, we have only limited evidence of how to realize this potential. There are steep returns to classroom experience early in a teacher’s career, suggesting that teachers learn a great deal “on the job” (Atteberry et al., 2015; Kraft & Papay, 2014). Most novices get considerably better their first few years in the classroom, learning a lot with real children in their charge. An important policy question then becomes how can we trace some of those lessons learned early in a teacher’s career into the pre-service period? Practicing teaching in a simulated environment is one possible approach to building pre-service teachers’ skills, but we know little about the way how to leverage this technology. This randomized control trial examines the extent to which coaching pre-service teachers in a simulated environment improves their teaching.

Context

Many professional training programs feature simulations (Konge & Lonn, 2016; Lateef, 2010; Shapiro et al., 2010). In theory, simulations provide practice opportunities with 1) immediate feedback from teacher educators, and 2) no harm done to real children. During a general teaching methods course, all prospective teachers at public university (N=105) participated in a series of simulations focused on practicing various teaching skills. Candidates were randomly assigned to one of two conditions: 1) Self-reflection between simulation sessions or 2) Expert coaching between sessions. Candidates span both elementary and secondary education programs and are broadly representative of the teaching workforce (predominantly white and female).

Methods and Data Sources

We used a balanced randomized block design to assign all participants to treatment, stratifying by course section. Simulation sessions were recorded and coded on a set of standardized rubrics targeting feedback within a text-based discussion. Sessions were additionally coded using text-as-data methods to account for student and teacher turn-taking as well as the proportion of open-ended questions. Additionally, coaches were surveyed on candidates’ receptiveness to feedback and coaching sessions were transcribed for analysis of treatment heterogeneity.

Results

Preliminary results suggest large, statistically significant effects of coaching on teacher candidates’ classroom practice. Participants who received coaching demonstrated higher overall quality performance than participants who self-reflected. Additionally, participants who were coached provided a substantively larger portion of high-quality, descriptive feedback relative to their peers assigned to self-reflection.

Significance

Teacher education programs need more robust evidence about how to prepare novices who are ready to provide effective instruction their first day in the classroom. This question is doubly imperative in a broader context of dire teacher shortages, and low enrollment in teacher preparation programs across the country. This experiment provides some of the first causal evidence about how to improve teaching skills in a teacher education course. As in the professional development literature, we find that coaching can be a powerful treatment for improving teaching in the pre-service period (Allen et al., 2011; Kraft & Blazar, 2017).