Investigating the use of value-added models for student achievement : does using multiple value-added measures lead to stronger conclusions about teacher effectiveness?

In the quest to achieve better academic outcomes for all students, the focus in education has shifted to a model of accountability. The most recent trend in the accountability movement is a focus on the effect of teachers in promoting student achievement. Research has found that teachers have the most significant school level impact on student achievement, and increases in teacher effectiveness could have major implications for the learning outcomes of students across the nation. Much of the current focus in teacher evaluation reform centers on methods through which teachers can be more accurately evaluated based on their contributions to student learning. In the push towards greater accountability for teachers, the development of measures that are both fair for teachers and lead to stronger outcomes for students are critical to seeing long-term improvements in the education system.

This report explores variability and stability of value-added measures over time by looking in depth at the methods, assumptions, limitations, and implementation of the most commonly used value-added models across the country and the research about the correlations of these measures over time. This research is followed by a case study of a de-identified large urban school district implementing a teacher evaluation system that uses both a commercially produced value-added measure and an alternative student-growth measure to make high stakes decisions about teacher effectiveness. The findings from this case study show correlations that do not differ significantly from the prior research on the year-to-year variability in teacher value-added measures, but urge for continued evaluation of these measures over time, especially in high-stakes decisions. Ultimately, value-added measures are only as useful as their effectiveness in influencing the core outcomes of teaching and learning, and therefore these measures must be carefully integrated into and validated against holistic assessments of teacher effectiveness in order to truly impact student outcomes.