Essays on Ratings of Teachers and Teacher Applicants

Abstract

Over the past decade, states and districts across the country have reformed their teacher evaluation systems to systematically collect ratings of teacher performance using standards-based observation protocols. Furthermore, many districts have implemented efforts to improve teacher selection through the use of tools that summarize information available at the time of application into ratings of teacher applicants. Additional research is needed to provide greater insights into the information captured by ratings of teachers and teacher applicants. In the first paper, I investigate the role of race and gender dynamics between teachers and administrators in the setting of classroom observations. I find that teachers, on average, experience small positive effects from sharing race or gender with their raters on their observation scores. These effects have important implications for the estimated size of the Black-White and male-female observation score gaps. In the second paper, I examine the predictive validity of a commercial screening tool that was designed to improve teacher selection. I find that the tool’s ratings have some capacity to predict teacher performance, but they are not necessarily a substitute for the more elaborate and promising screening processes that are conducted by human resources officials in some districts. In the third paper, I examine whether teachers’ observation scores contain information about the extent to which teachers’ impacts on student test scores persist in the longer term. I find some evidence to suggest that, in English Language Arts, observation scores may contain information that explains variation in the persistence of teacher value-added to test scores. However, further research is needed to provide better insights into the relationship between classroom observation scores and the persistence of teacher effects.