This report describes the value-added models (VAMs) created for the Pittsburgh Public Schools and the Pittsburgh Federation of Teachers. Pittsburgh's VAMs use not only state assessments but also course-specific assessments, student attendance, and course completion rates, aiming to produce...

This report presents the results of exploratory quasi-experimental analyses that use a regression discontinuity design to examine the relationships between certain features of No Child Left Behind accountability and subsequent student achievement in Title I schools in two states and three school...

A quasi-experimental analysis of data from the Pittsburgh Public Schools finds no difference in the effects of two elementary-level math programs—Everyday Mathematics and Harcourt Math—on student achievement. The absence of a difference in achievement effects may be related to the...

Vouchers and charters are two of the most ambitious attempts to improve school performance through parental choice and privatization of governance. The updated edition of this book (which was first released in 2001) summarizes new empirical evidence on the achievement impacts of voucher programs...

This report describes Mathematica's work in developing value-added models (VAMs) that estimate teachers’ and schools’ contributions to the achievement of their students, and assisting the Pittsburgh Public Schools in developing performance-based pay plans for teachers and principals.

A quasi-experimental analysis of data from the Pittsburgh Public Schools finds no difference in the effects of two elementary-level math programs—Everyday Mathematics and Harcourt Math—on student achievement. The absence of a difference in achievement effects may be related to the...

This report summarizes the evidence on measures of student achievement growth used in teacher evaluation that do not rely on traditional annual state assessments, and that instead use commercially available assessments, locally developed common assessments, and teacher-developed student learning...