Evidence-Based Reform: Advancing the Education of Students at Risk

Read full report in PDF

Despite some recent improvements, the academic achievement of American students remains below that of those in most industrialized nations, and the gap between African American and Hispanic students and White students remains substantial. For many years, the main policy response has been to emphasize accountability, and No Child Left Behind has added further to this trend. There is much controversy about the effects of accountability systems, but they have had little impact on the core technology of teaching: Instruction, curriculum, and school organization.

This paper argues that genuine reform in American education depends on a movement toward evidence-based practice, using the findings of rigorous research to guide educational practices and policies. No Child Left Behind gives a rhetorical boost to this concept, exhorting educators to use programs and practices “based on scientifically-based research.” In practice, however, programs that particularly emphasize research-based practice, such as Reading First, have instead supported programs and practices (such as traditional basal reading textbooks) that have never been evaluated, while ignoring well-evaluated programs. The same is true of the earlier Comprehensive School Reform program, which was intended for “proven, comprehensive” programs but has instead primarily supported unresearched programs.

Despite these false starts, the evidence-based policy movement remains the best hope for genuine reform in U.S. education. The Institute for Education Sciences in the U.S. Department of Education, as well as NICHD, NSF, and other funders, are supporting many research and development initiatives that use rigorous randomized experiments to evaluate educational products and practices. Of equal importance, the What Works Clearinghouse (WWC) is beginning to review educational programs to identify those supported by rigorous research. These changes create the possibility that educators will soon have available a broad range of programs in which they can place confidence, just as Food and Drug Administration approval gives physicians and the public confidence in medical treatments.

This paper reviews research on programs that already have strong evidence of effectiveness. It establishes criteria for study quality like those of the WWc= Programs with strong evidence of effectiveness fell into the following categories.

1. Comprehensive school reform models, which provide professional development and materials to improve entire schools. Research particularly supports Success for All and Direct Instruction, but smaller numbers of studies support several additional models including the School Development Program, America’s Choice and Modern Red Schoolhouse.
2. Instructional technology. Research supports integrated learning systems in mathematics. Word processing has been found to improve writing achievement.
3. Cooperative learning programs engage students in small groups to help each other learn. Many studies support this strategy in elementary and secondary math, reading, and other subjects.
4. Innovative mathematics programs. The first What Works Clearinghouse report supported research on two technology-based programs, Cognitive Tutor and I Can Learn, in middle schools. Elementary programs such as Cognitively Guided Instruction and Project SEED also have strong evidence of effectiveness.
5. Innovative elementary reading programs having strong evidence of effectiveness include Success for All and Direct Instruction, as well as Reciprocal Teaching and Cooperative Integrated Reading and Composition.
6. Tutoring programs in reading, especially Reading Recovery, have rigorous evaluations showing their effectiveness.
7. Dropout prevention programs, such as the Coca-Cola Valued Youth Program and Alas, have good evidence of effectiveness.

Policy Recommendations

Clearly, much more research is needed, and current policies are not supporting use of the research-based programs that now exist. Policy recommendations are as follows.

1. Substantially increase support for research and development to at least $500 million per year.
2. Fund development of new programs.
3. Fund evaluation of existing and new programs.
4. Provide incentives for schools to participate in research.
5. Provide incentives for schools and districts to use programs validated in rigorous research.
6. Maintain the integrity of proven programs by ensuring that publishers continue to provide professional development and support like that provided in the research.
7. Encourage states to base policies on research.

Conclusion

The solutions to America’s education problems must draw on our nation’s ingenuity, inventiveness, and technology. We can solve these problems as we have solved many others, by using research, development, and dissemination of effective tools and practices.

Read full report in PDF