AEFP 45th Annual Conference Program
Chair:, University of Virginia
Chair:, Urban Institute
Chair:, American Institutes for Research
Chair:, Florida State University
Chair:, Michigan State University
Chair:, RAND Corporation
Chair:, Stanford University
Chair:, University of Arkansas
Chair:, University of Texas at Austin
Chair:, University of Oregon
Chair:, Maryland State Department of Education
Student growth measures are widely used in state accountability systems to assess schools’ effects on student achievement based on assessments in grades 3–8. However, no states currently measure the impact elementary schools have on growth before grade 3, or the impact high schools have on student long-term outcomes such as college enrollment, college persistence, and success in the job market. Our panel will discuss measures currently being developed that could extend growth (and growth-like) measures used for school accountability to grades K–3 (in Maryland) and beyond high school graduation (in Louisiana). These measures have the potential to provide states with new information about school performance that is reflective of true school effectiveness. The proposed policy talk will include members of the research teams that are developing the new measures and state officials who are considering the ways the measures might be used.
The Maryland State Department of Education partnered with the Regional Education Laboratory Mid-Atlantic to explore whether a school-level K–3 growth measure could be constructed, and to assess its validity and precision. The study measured schoolwide student growth for reading and math using student growth percentiles based on Maryland’s Kindergarten Readiness Assessment and the grade 3 Partnership for Assessment of Readiness for College and Careers (PARCC) assessment. The study assessed validity by calculating correlations between student scores on the two assessments and compared those correlations with correlations between scores on tests in later years. To assess precision, the study constructed 95 percent confidence intervals around schools’ growth estimates. The study offers lessons to other states interested in constructing early-grade growth measures using two different assessments that are administered multiple years apart.
The Louisiana Department of Education and Mathematica are conducting a study to develop measures of each Louisiana public high school’s promotion power—the school’s impact on the long-term success of its students, as indicated by high school graduation, college or career readiness, college enrollment and persistence, and success in the job market. Measures of promotion power aim to provide data that fairly compares schools serving different populations of students. Like growth models, they are based on statistical models that identify schools’ contributions to students’ long-term outcomes separately from other factors, such as prior achievement and demographic characteristics. To assess the reliability and stability of promotion power measures, the study calculated the precision, year-to-year correlations, and across-outcome correlations for each measure. This study will be of interest to states that are considering broadening school accountability measures beyond high school graduation, and doing so in a way that separates schools’ contributions from external factors affecting student outcomes.
Some questions that may be covered include: To what extent should schools’ K–3 growth and promotion power measures be included in formal accountability systems versus used for diagnostic purposes? How should they be weighted relative to other accountability measures? How can the K–3 growth and promotion power measures be communicated to various audiences—schools, districts, teachers, parents—in a way that properly reflects the benefits and limitations of the measures? What are the policy issues that states must consider in exploring the possible adoption of such measures?