Juggling the Numbers

A federal spending shift spurs debate over mathematics and science education

Table: Some examples of NSF awards that seek to improve the quality of mathematics instruction in schools and colleges Table: How the federal government has been spending more money for Education Department programs in mathematics and science, and less on those of the NSF Read the transcript of a live, online discussion with Gordon A. Kingsley, the principal investigator on a National Science Foundation study of educational partnerships, about the best way to improve science and mathematics education in the public schools. Printer friendly E-mail article Subscribe Order reprints

By JEFFREY BRAINARD

For a decade, as the chancellor of Maryland's public university system, Donald N. Langenberg had pushed universities there and elsewhere around the nation to produce better-qualified schoolteachers in science and mathematics.

Then, after stepping down in 2001, he returned to the University of Maryland at College Park as a professor of physics, and began collaborating with other physicists there on how to improve undergraduate instruction in their field. Only then did it dawn on him that he'd been missing a big problem: the way physics was taught in college often bored students and force-fed them facts.

"I decided after 50 years in higher education that I taught just as I was taught by some of the nation's most distinguished physicists -- that is to say, badly," he says. He recalls that, as a graduate student, his own mentors told him that "teaching was not rocket science. Just get into the classroom and do it."

Since 2002 he has been working to rectify those shortcomings as a participant in a project financed by the National Science Foundation's Math and Science Partnership program, which supports efforts to get university scientists and mathematicians to improve teaching at both the undergraduate level and in public schools. Besides trying to bolster schoolteachers' knowledge of math and science, it seeks, through improvements in college instruction, to produce more, better-prepared schoolteachers. Maryland's university system has teamed with the Montgomery County, Md., public schools in one of the largest such projects of its kind.

However, a debate is under way among Washington policy makers and in academe about whether such large projects are the most desirable approach for getting university mathematicians and scientists to work with schoolteachers. The U.S. Department of Education has its own version of the Math and Science Partnership program, which differs in several respects from the NSF one. It largely lacks a focus on undergraduate teaching, concentrating instead on having university professors instruct practicing schoolteachers. And, lately, it has been winning more federal funds.

The NSF originally said it would issue new grants each year for five years. But, at the request of President Bush, Congress cut spending for the NSF program in this year's budget, leaving only enough money to continue existing multiyear programs like Maryland's. Congress increased funds for the Education Department's version over the past two years. And Mr. Bush's budget for 2006 would continue that disparity.

The shift in spending comes at a time when school systems are struggling to find enough cash to meet the requirements of the president's No Child Left Behind program to improve students' performance in math and other subjects. Administration officials have said that the Education Department has more connections than does the NSF with the nation's 15,000 school districts and is better suited to help.

However, university scholars, who are gaining some support in Congress, argue that the NSF program is uniquely fashioned to develop and test new ways for colleges and schools to work together. "The NSF has always had much more attention to the quality of the work in programs in math and science education than has the Department of Education," says Jere Confrey, who is a professor of mathematics education at Washington University in St. Louis and who chaired a National Academy of Sciences panel on mathematics curricula. "With its peer-review process and connection to the [scientific] disciplines, it has been a better source of continuous effort." Behind the debate are wider questions about the role that universities should play in improving schools.

Meeting the Demands

The discussion is taking place as many policy experts warn of a crisis in math and science teaching. U.S. high-school students score far lower on math tests than their peers in many other countries. Many states have significant shortages of schoolteachers who majored in math and science. And industry leaders have questioned whether colleges will produce enough graduates, especially U.S. citizens, to meet future demand in technology-driven industries.

Those alarm bells have sounded before. The Math and Science Partnership programs are the latest phase of several long-running efforts by the federal government to involve colleges in improving schoolteaching. Those efforts began in the 1960s as an outgrowth of Cold War priorities, when the NSF began to sponsor a series of summer training institutes for teachers. The agency also sponsors internships for schoolteachers to work in college researchers' laboratories and for graduate students to teach in schools.

However, the Math and Science Partnerships, created in 2002 by Congress and President Bush as part of the No Child Left Behind Act, have been seen as building on previous efforts in a new way. The push to involve more faculty members from departments of math and science was a break from some previous efforts at the NSF that had tended to draw on university schools of education and focus more on teaching methods.

The NSF version calls for universities to lead efforts with one or several school districts. Among the partnerships' many goals is to ensure that high-school curricula are better coordinated with what students will be taught at the undergraduate level. And the interactions are not meant to be one-way: College professors are expected to learn from practicing schoolteachers about how to teach better and more clearly.

Between 2002 and 2004, the NSF made three rounds of grants for 48 such ventures involving 150 colleges, 450 districts, and 140,000 teachers. The largest, for $35-million over five years, was awarded to the University of Wisconsin at Madison and the University of Pittsburgh to work with 40,000 teachers in Los Angeles, Denver, Madison, and Providence, R.I. The grants also paid for summer teaching institutes and efforts to evaluate the success of the projects.

A Growing Budget

The Education Department's budget for its version of the program started out small, just $12.5-million in 2002. But that has been changing: Congress provided the department's program $178.6-million for the 2005 fiscal year, compared with $79.4-million for the NSF version.

The Education Department program represents a different approach for grant making and projects. Almost all of its grants have been awarded through a population-based formula to state departments of education, which in turn are required to award them competitively to participants. The department has summary information so far only for grants in the 2003 fiscal year: That year the states together awarded a total of about 340 projects involving 135 colleges.

The Education Department program grants tend to be smaller than the NSF ones, with most providing a total of under $400,000. The money that reaches participating colleges can be significantly less.

Francis M. "Skip" Fennell, a professor of education at McDaniel College, in Maryland, says his college received a total of $18,000 over two years for work by him and another professor to teach summer institutes and follow-up workshops on mathematics to middle-school teachers in Howard County, Md.

That's a lot less than colleges typically expect their math and science faculty members to land as research grants. Mr. Fennell says he took on the project to help teachers and advance his research on ways to improve teaching.

Education Department grants also run for a shorter time than do NSF grants, typically less than three years. And they usually involve no more than two school districts.

The projects have a wider reach than the NSF's, since the department awards money to all 50 states. But they also have a narrower focus: Among 266 grants made in 2003, the efforts most often served schoolteachers in grades six through eight, according to an analysis by scholars at the Brookings Institution and the American Enterprise Institute. The median number of teachers participating was 46.

Participating teachers, on average, are "getting about 120 hours of professional development and content in a 12-month period," says Patricia O'Connell Ross, team leader for the Education Department's Math and Science Partnerships. That consists of both summer institutes and follow-up sessions during the academic year. "That's quite intensive for these kinds of efforts," she says.

Teachers in the NSF projects get varying amounts of instruction, with some receiving more than 200 hours a year.

Like the NSF projects, the department's projects have a research emphasis to determine which of them improve teacher knowledge and student performance, Ms. Ross notes. So far, both the department and NSF programs are too new for such results to be measured.

The NSF's director, Arden L. Bement Jr., has tried to put the best face on the shift in funds. He said at an appropriations hearing in March that educational researchers already know a lot about how to improve teaching in schools. And the Education Department has closer links to school districts than does the NSF and thus is better positioned to put some of those ideas into practice, he said.

"We're not an education agency, we're a research agency," he said this month at a meeting of the NSF's advisory committee on education and human resources. "We need to focus on those things we can do and do well."

Mr. Bement added that the squeeze on math-and-science partnerships and other educational programs at the NSF reflects pressure on the agency's budget, which fell by 2 percent this year, to $5.5-billion. The agency has tried to preserve the scientific-research programs at the core of its mission, and educational projects have taken a disproportionately large hit, he said.

Support in Congress

Now, some members of Congress are pushing for more money for both the NSF and the Education Department partnership programs. Late last month, 76 members of the U.S. House of Representatives, including 25 Republicans, signed a letter to appropriators asking that NSF spending be increased to $200-million, which would allow the program to resume making new grants. It was a rare display of support for an individual program within the NSF. A separate letter signed by members of Congress proposed an even larger increase for the Education Department partnerships, to $400-million.

Like those lawmakers, many academic officials who would like to see more spending for both programs argue that they are complementary. Advocates of the NSF program appear to have an ally in Rep. Frank R. Wolf. A Virginia Republican, he leads the U.S. House of Representatives appropriations subcommittee that oversees spending for the NSF and has publicly emphasized math and science education. But Mr. Wolf is fighting an uphill battle. Congress faces a tight budget for 2006.

Supporters of expanding the NSF program argue that not enough is known about how to improve student performance in mathematics and science. And without such understanding, giving schoolteachers more training in those subjects may not produce significant progress, argues Ms. Confrey of Washington University, who is participating in an NSF-financed math-and-science partnership there. Having large numbers of faculty members working with large numbers of teachers, as in the NSF projects, is necessary to generate a critical mass of activity and new ideas about how to help students at all levels, she and other educational researchers suggest.

In contrast, the Education Department program has a strong focus on helping teachers to prepare students to pass state tests on mathematics that are required by the No Child Left Behind Act.

Advocates also praise the NSF program's explicit focus on undergraduate instruction. It makes little sense, for example, to improve teaching of math for high-school students, if "when they get to college we promptly throw them into a class of 300 kids learning calculus," says Gordon A. Kingsley, an associate professor of public policy at Georgia Institute of Technology. Such large undergraduate classes may not give students the attention they need to develop problem-solving skills, says Mr. Kingsley, who holds an NSF grant to study the outcomes of that agency's Math and Science Partnerships.

Not everyone in academe agrees that the NSF's Math and Science Partnerships should be the focus of the agency's efforts to help colleges improve schools. Other educational-research programs within the NSF also face the budget ax. The agency's budget for Research, Evaluation and Communication, which finances academic work to study ways to improve school curricula and teaching methods, fell 10 percent for 2005, to $59.5-million. And the Bush administration has requested a further 43-percent cut for 2006.

Are Universities Good Partners?

What's more, some critics suggest that universities are not necessarily the best partners in school reform. During the 1990s, the NSF financed several multiyear projects that brought together school districts with universities, often through their schools of education. And the results of those projects, called Statewide Systemic Initiatives, were mixed.

"Universities like to do what interests them, rather than what K-12 districts need," says Daryl E. Chubin, former director of evaluation at NSF. Colleges "like to experiment," but "I wouldn't consider them close collaborators with K-12 schools," says Mr. Chubin, who now directs the Center for Advancing Science and Engineering Capacity at the American Association for the Advancement of Science.

An issue that those projects raised, and that remains an open question for the math-and-science partnerships, is how to encourage academic mathematicians and scientists to continue interacting with schoolteachers after the grant money is gone. That's a significant issue because faculty members have so many incentives to stay inside their ivory tower, highest among them the pressure to win and carry out research grants in order to gain tenure and promotion.

That pressure can be hard to surmount even though some of the Maryland institutions participating in the NSF-financed effort are offering a substantial cash stipend -- $7,000 a year -- to faculty members who participate at a certain level. "Even at Towson University, which has traditionally been a teachers college, it's been a problem," says David B. May, project director for the Maryland System. "At least one faculty member was told she wouldn't get tenure if she participated in our project."

But Lev R. Ryzhkov, an associate professor of chemistry at Towson, says it's been worth it to juggle work with schoolteachers and his research efforts.

"I want to know what content they cover," he says of the schoolteachers, "because the product of their work comes to us."