When Pat Morgan, mathematics coordinator for Oklahoma's Moore Independent School District, considered changing the middle school algebra curriculum, she knew she needed hard evidence to show teachers that any new program would really make a difference. So Morgan conducted an experiment in which 224 students switched to the Carnegie Learning Corporation's Cognitive Tutor Algebra I, which Morgan knew had been favorably rated by the U.S. Department of Education, and 220 students continued to receive traditional algebra instruction with the standard district-issued textbook.

Knowing a little bit about research design, Morgan had all the participating algebra teachers teach both types of classes, thus controlling for any effects that teacher style or quality might have on student achievement. Using a standardized algebra post-test from the Educational Testing Service with both groups, Morgan found the Carnegie course to have significant benefits over the regular curriculum. Now she says all Algebra I classes in the Moore district, with the exception of honors sections (which were already succeeding under the original curriculum), study with Cognitive Tutor, and all algebra teachers in the district are enthusiastic supporters of the new program.

"It was particularly rewarding to find that some students were succeeding who had previously been deemed 'unable to handle algebra.' It was refreshing to find more students actually looking forward to coming to algebra class," Morgan says.

A thousand miles away, in the quintessential college town of Ann Arbor, University of Michigan researcher Annemarie Palincsar is trying to figure out how to provide teachers with professional development around Reciprocal Teaching (RT), a program she and Anne Brown developed in the 1980s to improve children's reading comprehension.

In RT, teachers model four comprehension-building strategies: generating questions, summarizing paragraph by paragraph, clarifying content that poses comprehension problems, and making predictions. Once students learn these strategies, they coach each other in their use. The ultimate goal is for students to internalize the strategies and to use them when they read on their own.

RT has shown promising results, especially for students who have adequate decoding skills but difficulty comprehending what they read. In one study of 4th- and 7th-grade students, for example, participating children significantly outperformed their peers on standardized reading comprehension post-tests and daily assessments. Research has also shown that the benefits of RT hold up over time, and that many children are able to transfer the text interpretation skills they learn in "RT dialogues" to their own reading.

Palincsar has written a facilitator's manual and produced a videotape that includes interviews with teachers and students who have successfully used RT. But, she says, because she has only been able to provide professional-development support to a fraction of the many teachers who have requested it, she has no way of knowing whether the versions of RT some schools are using are consistent with the original model. "Most of us [university researchers] get support to do the initial experimental work, but it's hard to get the funds to see that work through to practice on a large scale," Palincsar says.

What if the details of Morgan's study could be provided to educators beyond her district so they could build upon previous research knowledge in algebra instruction? What if the support existed for RT to be piloted in a range of schools and, if the initial findings bore out, it could become a widely used strategy for teaching reading to students with poor comprehension skills? What if funding were provided for large-scale professional-development experiences to share best practices like RT?

Linking Research and Practice

Such links between education research and school practice probably seem natural to outsiders—and utopian to anyone who has followed the disjointed relationship between research and practice over the years. Yet these are among the collaborations envisioned in an ambitious new proposal for a Strategic Education Research Partnership (SERP), published this fall in a report by a special committee of the National Research Council.

Developed by a diverse team consisting of teachers, administrators, researchers, foundation officers, business leaders, and others with an interest in improving K-12 education, the SERP plan calls for nothing less than a national "education research and development" network, the efforts of which would inform teaching and learning all over the country. Main structural elements of the SERP proposal include:

• A central organization or headquarters, where directors would maintain program coherence; oversee "quality control" in education research; handle communication among educators, researchers, policymakers, and other constituencies; and work toward a variety of long-term research programs aimed at improving teaching and learning in U.S. schools. SERP would maintain a cumulative research knowledge base so that studies that probe various questions in education could build more systematically upon prior work, rather than being conducted in isolation.
• A network of research and development teams, representing university schools of education as well as private and public research organizations, that would focus on the issues most central to improving student achievement and classroom practice—especially in reading comprehension and mathematics.
• A set of field sites—schools and school districts around the country—where educators and researchers would conduct classroom-based research, pursuing "key questions and puzzles" about what works best in schools and why. Modeled after teaching hospitals, these sites would be chosen to represent a diversity of schools and districts.

Suzanne Donovan, director of the SERP project, said recently that most education research is currently not made useful or meaningful enough to the people who need it most: K-12 teachers and administrators. Speaking at a presentation about SERP at the Harvard Graduate School of Education, Donovan also noted that research findings are not made easily accessible to educators and that, conversely, educators' best practices are not studied systematically enough, if at all: "[The SERP model] is intended to build capacity for education R&D and cooperative relationships between practitioners and researchers."

Who Pays—and How Much?

As might be expected, a plan as ambitious as SERP comes with an equally ambitious price tag. SERP planners estimate the startup costs for the program to be in the neighborhood of $500 million over seven years. According to Donovan, the plan's launch is envisioned as a "public- private partnership," with foundations, businesses, and federal agencies among those who would share initial costs. Once the program is up and running, the SERP plan calls for states—which spend an average of one-third of their annual budgets on education—to enter into an "interstate compact" that would be the primary source of operating funds.

SERP planners acknowledge that the case for public funding may be especially tough to make in lean economic times, but they also say states have some strong incentives to buy into the partnership. For one thing, accountability initiatives such as the federal No Child Left Behind Act have led states and school districts to look for ways they can meet benchmarks such as "adequate yearly progress." A coordinated, targeted research effort might help school systems—and ultimately states—discover strategies that lead to the required gains, SERP proponents say.

They also point out that if each state allocated just one-half of one percent of its annual education expenditures to the endeavor, approximately $800 million per year would be available for education R&D. Such expenditures are routine in other fields, the planners note, and are usually viewed as investments that yield greater returns down the road. "States now spend over $300 billion per annum on K-12 education delivery. In other sectors—medicine, agriculture, transportation, communication—it is assumed that an investment of several percent of the 'production' budget should go to R&D," they write in a summary statement of the SERP proposal. "As in other sectors, we would expect the R&D investment to yield a return in education—in the form of fewer repeat grades, fewer special education placements, higher graduation rates, [and a] reduction in teacher time required."

Obstacles to Overcome

Could such an interstate compact ever really get off the ground? In the mid-1960s, states joined forces to form the Education Commission of the States, an organization dedicated to interstate knowledge sharing to help inform state education policies. Former Wyoming Gov. James Geringer, who spoke at the recent SERP forum at Harvard, says some governors and other state officials may well be ready for another cooperative venture in the 21st century since they, like many others, are hungry for solutions to the problem of how to raise student achievement. "We don't know what's working and what isn't working without some kind of research partnership," Geringer says. "There's a feeling [among lawmakers] that we need to do something."

Still, even with the buy-in—both philosophically and financially—of a critical mass of states, additional issues will need to be addressed to satisfy those who are skeptical that partnerships like SERP can succeed.

The politics of education research. As any observer of education research knows, politics plays a large role in determining what issues are studied—and, some would even argue, what conclusions are drawn—by many researchers. As a result of highly politicized debates on issues such as class size and bilingual instruction, educators have often been caught in a crossfire of reports that, when viewed collectively, lead to nothing but confusion about what approaches work best. Would the SERP model change all this, or merely fan the flames of the same old disagreements?

Predictably, SERP's proponents argue the former. Despite the dependence of the model on state funding, the SERP plan calls for researchers to function independently and for the SERP director to make "scientific and programmatic decisions" without influence from the compact's member states. "If a strong SERP existed, some of the heat could be removed from the education debates that presently tend to cripple progress in education," the report states. "A SERP enterprise one step removed from politics could take on questions about theories of learning or the efficacy of different instructional approaches (basic skills versus inquiry-based, phonics versus whole language) and subject such questions to systematic study. These questions are answerable."

The working lives of teachers. A key goal of the SERP plan, one that distinguishes it from many other efforts to link education research and practice, is to harness the discoveries that teachers are making in their classrooms and to make these findings available to a larger audience. Another is to ensure that educators actually employ the aspects of instruction that convincing bodies of research identify as effective. Yet the structure and pace of the typical classroom teacher's work day—with its 20-minute lunch, 45 minutes of prep time, and severely limited professional development opportunities—raise several questions about whether SERP can truly have an effect on classroom instruction on the national scale envisioned by its planners. Given their current working lives, can classroom teachers become true partners in a research collaborative? If so, what kind of professional development would they need to be able to use and conduct research effectively? And how would school systems that are already strapped for resources pay for the instruction and the teacher release time that would be required?

Palincsar, who teaches at the University of Michigan's School of Education, says educators need to be afforded the time, professional-development opportunities, and other resources to integrate research-based work into their day-to-day school lives if all the goals of SERP are to be met: "I think SERP is one way to address the research-practice [disconnection] problem, but it's naïve to assume that SERP would solve the entire issue on its own. If we want teachers to engage in research-based teaching practice, we have to give them the support to do that."

K-12 practitioners like Pat Morgan echo Palincsar's concern about the limits of time. "The greatest obstacle in using data to guide our instruction is the significant amount of time that it takes to collect, organize, and analyze the data," Morgan says. "We have so many other responsibilities and commitments in the daily implementation and assessment of curricular programs that it is difficult to devote that much time to experimental studies."

The SERP proposal calls for teacher-release time and new roles for teachers in field sites and at SERP headquarters. But how such a vision can become a reality is an open question.

The working lives of researchers. Another culture that would have to undergo change for a research-practice partnership to work to its full potential is that in university schools of education, says Catherine Snow, SERP committee vice chair and a professor at the Harvard Graduate School of Education. The current reward systems for junior faculty at many institutions, Snow says, value individual achievements and publication over collaboration. "It's almost as though collaboration is something you can only afford once you've got tenure," Snow says. Snow adds, however, that there are signs that this culture is beginning to change at some institutions, and that more leaders at major research universities are recognizing the value of education research that has a direct impact on classroom practice.

Stay Tuned for SERP?

How close are teachers and principals to seeing SERP-inspired projects in their own classrooms? A lot depends on funding, SERP planners say. Right now, the committee is conducting a major push for foundation support that could provide the critical startup funds needed for launch in a small network of sites. Without such funding in the next half year, progress on SERP could be stalled indefinitely.

Regardless of how the short-term challenges play out, Snow says the dialogue about the need to provide better links between education research and practice is sure to continue. "Even if SERP itself didn't happen, the ideas behind SERP—that education research needs to be done differently and that there needs to be a new and better-organized relationship between research and practice—aren't going to go away," says Snow. "Currently, the way education research is done doesn't do much for schools."