Collaborative Research: RealVAMS-Getting Real-World Value from Value Added Models
University Of Nebraska-Lincoln, Lincoln NE
Investigators
Abstract
Value-Added Models (VAMs) in education are increasingly being used to attempt to measure the impact of a teacher on student achievement in a way that takes into account students? and peers? prior performance and demographics. Data that typically inform a VAM rely largely on student achievement data from state testing systems. The RealVAMs project tests a multi-dimensional mechanism to include other measures of student success such as scores from different assessment instruments (SAT, ACT) and categorical responses such as whether or not a student graduated or persisted in a STEM field. The real world model using multi-dimensional data is being tested using a longitudinal database from a large school district. The project is creating an open source accessible program to facilitate the expanded measurement of VAMS across a larger scale. This project has national significance as the use of VAMs for teacher evaluation is expanding rapidly across the country in schools and districts. A major objection to the use of VAMs in teacher evaluation has been the over-reliance on the use of state testing data. Opponents of the use of VAMs argue that this reliance contributes to a narrowing of curriculum and instruction to only what can be tested on state tests. Not all subjects are tested, so the use of VAMs does not support the evidence that can be used in the evaluation of teachers in non-tested subjects or grade levels. This project informs the educational practice community of schools and districts and provides for more nuanced and richer information about the impact of teachers for policy decision making.
View original record on NSF Award Search →