Catherine E. Snow, a Harvard University education researcher, tells a story about a vocabulary curriculum for Spanish-speaking students that she tested a decade ago in a public school.
Satisfied that the results had been a great success, she left the teaching materials with the school and returned a year later to ask teachers how the program was going.
鈥� 鈥極h, that project is over,鈥� 鈥� Ms. Snow said the teachers replied. 鈥淚t wasn鈥檛 something researchers were doing for them. It was something researchers were doing to 迟丑别尘.鈥�
Producing work that educators actually want to use is the raison d鈥櫭猼re for an ambitious national research-and-development program known as the Strategic Education Research Partnership, or SERP.
Inspired by a series of reports from the National Research Council from 2001 to 2003, the nonprofit partnership was created as a vehicle to improve educational research through studies aimed at solving practical problems.
At its start in 2003, the project鈥檚 鈥済rand vision鈥� called for raising up to $700 million for a 15-year effort enlisting states to support and set an agenda for large-scale, sustained collaborations involving practitioners and researchers.
Four years later, SERP鈥檚 founders have fallen far short of that initial fundraising target. The organization鈥檚 current annual operating budget is closer to $1.3 million, its directors say. But the SERP vision of research as a collaborative enterprise, grounded in practice, is quietly soldiering on.
69传媒 as Partners
This year, with support from private foundations, as well as volunteer and in-kind contributions from researchers and practitioners, dozens of nationally known researchers are working under the SERP banner in schools in Boston, San Francisco, Evanston, Ill., and a handful of other districts across the country. (The foundations include the Chicago-based Spencer Foundation, which also supports coverage of education research in Education Week.)
鈥淚t鈥檚 a completely different way of doing business,鈥� said Ms. Snow, who is helping to lead the national network. 鈥淲e aren鈥檛 using schools like test tubes for our ideas. We鈥檙e using schools like partners in developing questions.鈥�
The idea for a national infrastructure to promote and guide usable research in education was the brainchild of Bruce A. Alberts, a biochemist and a former president of the National Academy of Sciences, a private, congressionally chartered group that advises the federal government.
His model was highway research. As in education, studies in that field were once badly underused and underfunded. The field of research got a second life, though, after a national scientific panel crafted a 10-year strategy for improving it in the 1980s.
Early on, SERP attracted a few high-profile champions, including then-Govs. Mark R. Warner of Virginia and Mike Huckabee of Arkansas. Even so, states balked at ponying up to join the effort.
鈥淭his kind of work that is focused on how you meet the challenges of daily practice is a bit too distant from what鈥檚 pressing on states,鈥� said M. Suzanne Donovan, the executive director of the SERP Institute, the partnership鈥檚 Washington-based headquarters. 鈥淪chool districts are really the ones who feel that the problem is pressing.鈥�
The first test case was in Boston, where the district鈥檚 superintendent at the time, Thomas W. Payzant, had been a member of the national panel that developed the original SERP idea. On Mr. Payzant鈥檚 recommendation, the research group in 2005 set its sights on improving middle school literacy in the 57,00-student district. 鈥淲e had huge numbers of students coming into 9th grade with literacy skills鈥攁nd also math skills鈥攖hat were several years behind where they needed to be,鈥� said Mr. Payzant, now a professor of practice at Harvard University鈥檚 graduate school of education. 鈥淧lus, there are good university people here in the Boston area, at Harvard and Boston University and Wheelock College, who were interested in literacy as well.鈥�
To address that issue, the researchers devised diagnostic tests to help middle schools pinpoint students鈥� strengths and weaknesses in reading and writing.
鈥淢ost existing measures give an overall comprehension score, and what we鈥檝e been finding is that students have multiple difficulties,鈥� said John P. Sabatini, a senior research scientist for the Princeton, N.J.-based Educational Testing Service who is developing the new test with scholars from Boston University.
Boston teachers also told researchers that students coming into the city鈥檚 middle schools were stumbling on academic language not normally taught at that level.
鈥淔or example, if you tell kids 鈥榩hotosynthesis is a process of conversion of solar energy,鈥� they don鈥檛 know what 鈥榩rocess鈥� and 鈥榗onversion鈥� means,鈥� said Ms. Snow, who also created a practicum at Harvard to draw young scholars into the SERP work.
The researchers proposed a program called Word Generation, which introduces students to vocabulary through texts on such topics as stem-cell research and athletes鈥� salaries. They learn five words a week in daily 15-minute lessons spread across the curriculum among multiple teachers.
Two middle schools pilot-tested the program last year, and teachers provided weekly feedback to the researchers. This year, six schools will road-test the program.
鈥淭hat鈥檚 the beauty of the researcher-practitioner piece,鈥� said Kathleen Murphy, a literacy coach at one of the schools, the 550-student Rogers Middle School in Boston鈥檚 Hyde Park neighborhood. 鈥淭he researchers develop the program, and they have the expertise and knowledge, but we鈥檙e the ones who see where the roadblocks are in the practical application.鈥�
With another Harvard researcher, Richard F. Elmore, the vocabulary lessons are being piggybacked with schoolwide surveys intended to measure schools鈥� 鈥渋nternal coherence鈥濃�攊n other words, the degree to which staff members trust one another, share a vision, and take responsibility for student learning, and the staff鈥檚 capacity to respond to pressure for improvement. The idea is to find out whether changes such as Word Generation have more sticking power in schools with high scores for internal coherence.
鈥淭he SERP group really listens to the questions the district wants answered, and not the questions that researchers think would be interesting to answer,鈥� said Ellen C. Guiney, the executive director of the Boston Plan for Excellence, an independent group that promotes school improvement efforts in the city. 鈥淎nd it combines research with the implementation of the proposed solution.鈥�
Work in California, Illinois
In the 55,000-student San Francisco public schools, where the SERP work is just getting under way, the focus in on middle school math and science. Researchers from those disciplines and scholars on teaching English-language learners are trying to come up with a program for teaching academic language to ELLs and testing ways to help students with word problems in math.
鈥淚f the district hadn鈥檛 pushed us in this direction, the language people and the math and science people would never have been collaborating,鈥� said Philip Daro, the University of California, Berkeley, researcher who is leading the SERP work in San Francisco schools.
The 3,000-student Evanston district in Illinois comes into the partnership as part of the Minority Student Achievement Network, a national collaboration of suburban districts working to narrow achievement gaps among their diverse student populations. The other districts in the network that are working with SERP are in Arlington, Va.; Madison, Wis.; and Shaker Heights, Ohio.
Laura A. Cooper, the assistant superintendent for public instruction for Evanston Township High School, said the partnership there focuses on students about to enter 9th grade but who have not yet taken algebra. In particular, researchers are exploring the role that motivational and social issues play in students鈥� success in math.
鈥淲e have a lot of students walking into high school who say, 鈥業 can鈥檛 do math,鈥� 鈥� Ms. Cooper said. Where SERP still comes up short, said Frederick A. 鈥淔ritz鈥� Mosher, a senior consultant to the Center for Policy Research in Education, a consortium based at the University of Pennsylvania in Philadelphia, is in establishing a large-scale structure to produce research that will stand the test of time.
鈥淭he question is how you get stuff funded over time and at a scale to give you sufficient evidence that you鈥檝e got a real solution to the problem,鈥� said Mr. Mosher, who is based at Teachers College, Columbia University, and is not part of the SERP effort.
Ms. Donovan said that goal may be years away, given current funding levels. In the meantime, use of some of the partnership products generated in Boston is spreading to other field sites.
鈥淚鈥檓 happy we鈥檙e small,鈥� Mr. Daro added. 鈥淲e鈥檙e still figuring out how to do this, and the last thing we should do is try to have everything in dozens of sites all of a sudden.鈥�
