Includes bibliographical references and index.

This series of studies began with a grant proposal focused on students classified with learning disabilities which was successfully funded by the National Science Foundation. Although the main emphasis of this National Science Foundation program was science, technology, engineering, and math, we were also interested in broader educational and social psychological outcomes for these students. Previous studies had largely relied on aggregate-level data and bivariate analyses. We sought to build on these studies by analyzing data describing students, teachers, and schools with quasi-experimental techniques such as multilevel regression modeling and propensity score matching. This article describes the development of the analytical and theoretical foundations for the studies, with a particular focus on data limitations for at-risk status groups, the meaning of and need for quasi-experimental techniques, options for handling missing values and clustered data, and how disciplinary differences shape research.

Description based on XML content.

Licensed e-book