Decades of shifting fashion in education have disillusioned many parents, school staff, policymakers, and taxpayers. Since the early 1970s, public spending per student has increased 90 percent, and billions of dollars have been invested in initiatives touted as the cure for education ills ranging from declining test scores to rising truancy. Yet the results of these efforts have fallen well short of expectations, and little can be learned from them because most education research has failed to generate rigorous evidence about what works, what does not work, and why. It’s no wonder many have come to view reforming the nation’s education system as both expensive and futile.

Now, with the passage of the comprehensive No Child Left Behind Act of 2001, Congress and the Bush Administration have put evidence-based education policymaking at the center of the nation’s elementary and secondary school reform agenda. Under the new law, only those instructional activities that are backed by “scientifically based research” may be supported using federal funds. School districts must devote some of the funds to evaluation when no hard evidence on effectiveness is available. These new mandates offer an unparalleled opportunity for high-quality research to shape how children are schooled.

But what constitutes scientifically based research in education? From a research perspective, education poses challenges absent in fields such as medicine, welfare, or workplace safety. For one thing, control over education policy is highly decentralized, residing with the country’s 16,850 local education authorities, and only 7 percent of education funding comes from federal sources. Also, education is universal and compulsory, which means it must accommodate diverse values and needs while striving for both equity and excellence. Finally, elementary and secondary schools have been charged with meeting a wide variety of goals, from teaching students basic reading and math to preparing them for civic engagement and labor market participation as adults. Regrettably, these challenges are often invoked to justify research methods that lack rigor, leaving education stakeholders to debate the merits of the evidence instead of its implications for policy and practice.

MDRC brings to the special challenges of education a wealth of insight accumulated over nearly 30 years of investigating welfare and employment policy. Like education initiatives today, programs for the poor and unemployed were once characterized as “conceived on a whim and dismissed on an anecdote.” As described in The Politics of Random Assignment, MDRC pioneered the use of research methods — notably randomized experimentation — now considered the gold standard for gathering reliable evidence about social programs’ effectiveness. In a randomized study, each member of a research sample — whether people with a particular disease, welfare recipients, or whole institutions, such as hospitals or housing developments — are assigned by lottery to a treatment group or a control group and studied over time.

With a focus on disadvantaged students and the schools that serve them, MDRC applies its expertise in the use of random assignment to test school-based reforms such as Career Academies and KIPP Schools. The prospects for continued use of this demanding but powerful methodology in the field brightened when the U.S. Department of Education began planning an array of research and evaluation projects — under the rubric of the newly established Institute for Education Sciences — that are widely expected to rely on random assignment.

MDRC also recognizes that random assignment is not always feasible or even the most appropriate way to assess the effectiveness of education reforms. The Working Papers on Research Methodology series highlights several rigorous quasi-experimental approaches. These methods were pioneered in MDRC’s evaluation of Accelerated Schools and are being refined and applied in our studies of First Things First, Talent Development, and Project GRAD. All of our education projects also draw on implementation research to elucidate the links between an intervention and its observed effects.

Now that building knowledge about what works in education has become a top priority, it is up to policymakers, educators, and researchers to gather and interpret the evidence needed to ensure that all children receive the best education possible.