Abstract
A large and still-growing body of literature has established that the quality of a student’s teacher can have considerable effects on that student’s outcomes, both short- and long-term—and the extent to which individual teachers drive academic achievement can vary substantially. This expanding literature on teacher effectiveness has in part been the impetus for significant developments in the evaluation of teachers nationally, with the goal of identifying and developing quality teaching. One of the most prominent, high-stakes, and arguably more successful examples of the recently expanded use of evaluation is the program employed by the Washington, DC public school (DCPS) system—IMPACT, which serves as the basis for each of the three chapters in this dissertation. The features of IMPACT are designed not simply to shift the overall distribution of teaching quality in the DCPS upward, but also to narrow variation in performance across the district, thereby lessening differences in access to quality teachers across students. In the three chapters of this dissertation, I explore how teacher evaluation can drive teachers’ performance and retention decisions in high-stakes settings, and how observation-based measures of teachers’ performance may glean insights into teachers’ development over time. Together, these chapters indicate that teacher evaluation in DCPS can provide useful insight into teacher’s’ development—and that the process of evaluation in DCPS may drive this development in and of itself. The research presented in this dissertation suggests that teachers’ performance is sensitive to the contexts in which they teach, as chapter 1 demonstrates that changing standards and expectations can significantly influence teachers’ practice—and can do so in potentially deleterious ways. It also demonstrates (in chapter 2) that, even in mature evaluation systems, low-performing teachers make decisions about whether to remain in the profession when faced with consequences of their inadequate practice, and these teachers improve when they choose to stay. Finally, the third chapter demonstrates that improvements do not simply occur at consequential performance thresholds, but that new teachers in general, who are in the process of learning and developing the skills that will make them better teachers, make substantial performance gains though their early years in the classroom. The variation in these gains across teachers and across practices, likewise, highlights areas where targeted professional development might facilitate steeper performance trajectories for DCPS teachers overall. These questions are particularly meaningful given the low-income, high-minority population the District serves. Students with demographic characteristics similar to those of the student body of DCPS are more likely to have underqualified, less experienced, and possibly less-effective teachers. Meanwhile, the strongest outcomes associated with teacher effectiveness and development are often concentrated among high-poverty or low-achieving schools, classrooms, and students, and among new and low-performing teachers, suggesting that policies to facilitate teacher development might be more meaningful in these contexts and could serve to make a sizeable dent in the academic achievement gap.
