High stakes test-based accountability systems primarily rely on aggregates and derivatives of scores from tests that were originally developed to measure individual student mastery of content specifications. Current validity models do not explicitly address this use of aggregate scores to measure the performance of teachers, administrators, and…

The Every Student Succeeds Act requires states to identify schools with low-performing student subgroups for Targeted Support and Improvement or Additional Targeted Support and Improvement. Random differences between students' true abilities and their test scores, also called measurement error, reduce the statistical reliability of the performance…

This Snapshot highlights key findings from a study that used Bayesian stabilization to improve the reliability (long-term stability) of subgroup proficiency measures that the Pennsylvania Department of Education (PDE) uses to identify schools for Targeted Support and Improvement (TSI) or Additional Targeted Support and Improvement (ATSI). The…

The "Stabilizing Subgroup Proficiency Results to Improve the Identification of Low-Performing Schools" study used Bayesian stabilization to improve the reliability (long-term stability) of subgroup proficiency measures that the Pennsylvania Department of Education (PDE) uses to identify schools for Targeted Support and Improvement (TSI)…

Authentic assessments used in response to accountability demands in higher education face at least two threats to validity. First, a lack of interchangeability between assessment tasks introduces bias when using aggregate-based scores at an institutional level. Second, reliance on written products to capture constructs such as critical thinking…

We use data from a large-scale, school-level randomized experiment conducted in 2010-2011 in public schools in Indiana. Our sample includes more than 30,000 students in 70 schools. We examine the impact of two interim assessment programs (i.e., mCLASS in Grades K-2 and Acuity in Grades 3--8) on mathematics and reading achievement. Two-level models…

Oklahoma is one of 16 states electing to use an A-F letter grade as an indicator of school quality. On the surface, letter grades are an attractive policy instrument for school improvement; they are seemingly clear, simple, and easy to interpret. Evidence, however, on the use of letter grades as an instrument to rank and improve schools is scant…

Value-added scores from tests of college learning indicate how score gains compare to those expected from students of similar entering academic ability. Unfortunately, the choice of value-added model can impact results, and this makes it difficult to determine which results to trust. The research presented here demonstrates how value-added models…

Drawing on national and state assessment datasets in reading and math, this study tested "external" versus "internal" standards-based education models. The goal was to understand whether and how student performance standards work in multilayered school systems under No Child Left Behind Act of 2001 (NCLB). Under the…

Public-private partnerships are being increasingly supported and advocated for, ideologically and financially, as an approach to educational reform in many countries across the world. Proponents suggest that non-state involvement in the education sector has the potential to bolster international Education for All efforts, improve school…