Developing and refining value-added models for school accountability

Edward Haertel, Susanna Loeb, and Anthony Bryk are conducting research using value-added models with data from California, New York and Chicago respectively. This research is supported in part under subcontract to the National Center for Research on Evaluation, Standards, and Student Testing (CRESST) at UCLA.

The model of standards-based reform enacted under the No Child Left Behind Act depends on reliable, valid, and credible measures of status and growth for schools and districts. In order to identify best practices both within and across states, researchers and policy-makers must have metrics that enable fair comparisons across jurisdictions. Research projects under the Value-Added initiative include theoretical work, simulations, and investigations using state-level data sets to guide policy and practice in the implementation of accountability systems, focusing on fair and comparable metrics for identifying schools in need of improvement.

Value-added research being done at IREPP also concentrates on developing and evaluating statistics for measuring and comparing gaps and gap trends measured using different tests, such as the National Assessment of Educational Progress (NAEP) and state accountability tests. Over the next several years, this work is expected to include comparisons of value-added models with the current successive-cohort model for setting growth targets and determining school-level achievement trends.