the starting point for R&D that is intended as one of the hallmarks of SERP: following promising program outcomes with research on the circumstances under which the results are obtained, the feasibility of the intervention in the classroom and the school context, the teacher knowledge and support required for success, and the organizational factors that influence outcomes. We refer to these as “downstream” cases because the work in these areas has already traveled some distance toward classroom usability.

2. Are there pervasive problems of practice that are widely recognized as critical, but for which the knowledge base is too weak to guide instructional practice? We refer to these as the “upstream” cases because the work is still in early stages. Since there are few promising interventions in these areas at present, an impact on practice is likely to require more time.

The panel recommends three areas for focus: reading, mathematics, and science. In both mathematics and reading, the proposed downstream work would address learning in the elementary years. In science, in contrast, the downstream work is in physics—a subject generally taught in high school—because of the strength of the research base. The upstream research proposed for the three domains focuses on reading comprehension, algebra, and the sequencing and content of science instruction across the school years.

There is an unusual degree of consensus regarding the goals of early reading instruction, as well as a fairly solid research base both on the contributors to success in achieving those goals and on assessments to predict reading difficulties. The targets of early reading instruction are therefore fairly clear.

Still, many children in U.S. schools are not learning to read well and, in many classrooms, teaching practices have not been influenced by research knowledge. In the panel’s view there is a