Value-Added Modeling Alternatives That Reflect Classroom Composition

• In Event: 25.068 - Roundtable Session 8
  In Roundtable Session: 25.068-10 - Toward Equal Opportunity: Seeking Best Practices for Incorporating Value-Added Measures in the Teacher Evaluation Process

Abstract

Administrators, school personnel and policy makers are all invested in finding valid and useful ways of evaluating teacher performance in the classroom. To date, the average treatment effect on the treated (ATT) has been proposed as an alternative to the average treatment effect (ATE) for the estimation of teacher effects on student achievement gains as it better reflects the teacher’s responsibility to teach the students actually assigned. While value-added models estimate the ATE, several alternative approaches estimate the ATT. These include linear model-based methods, propensity score-based approaches, and regression tree-based approaches. This project compares estimates of the ATT using these various approaches. Results suggest that several statistical approaches should be used conjointly when making high stakes determination about teachers.

Authors

• Kimberlee Callister Everson, Western Kentucky University

• Erika Feinauer, Brigham Young University