What Do Classroom Observation Scores Tell Us About Student Success? Capturing the Impact of Teachers Using At-Scale Classroom Observation Scores

Abstract

Classroom observations are the most widely used K-12 teacher evaluation measure in the United States. However, far less is known about the properties of observation scores relative to student test score-based measures such as teacher value-added (VA). Using statewide administrative data from Tennessee between 2006-07 to 2017-18, I apply quasi-experimental methods to examine (1) the extent to which classroom observation scores capture teacher impacts on students K-12, post-secondary, and early career labor outcomes and (2) how the magnitude of observation score effects compares to analogously-estimated effects using teacher VA. I find that changes in students’ exposure to teacher quality, as measured by observation scores, are significantly linked to changes across a multitude of student outcomes, particularly test scores, absences, suspensions, post-secondary enrollment, and post-secondary completion. The effects of teacher quality, across outcomes, are strongest for students during the high school years. Results were robust across multiple identification strategies and methods for accounting for measurement error. In relation to teacher VA, observation score effects were comparably sized to analogously-estimated VA effects. Results from this dissertation provide promising evidence on the use of administrator-led classroom observations as an outcomes-linked method for evaluating teacher effectiveness.