Valid yet Bias? An Examination of North Carolina’s Educator Effectiveness System

• In Event: Experiences in Teacher Evaluation Systems

Abstract

In the wake of national criticism of teacher evaluations based solely on value added scores, many states, through RttT funds, have developed standards-based teacher evaluation systems to provide a holistic measure of teacher effectiveness. This paper examines the predictive validity of North Carolina’s Educator Evaluation System (NCEES) by comparing teachers’ evaluation ratings with their value-added scores. I also examine whether school leaders are systematically bias in ratings they provide based on teachers’ race, gender, ethnicity, and age. The findings suggest that NCEES is correlated with teacher effectiveness in reading, math, and science. Conditional on effectiveness; White men are systematically rated lower than White women in reading and math. Black women receive lower ratings compared to White women in math only.

Author

• Shanyce L. Campbell, University of Michigan - Ann Arbor