The question, posed at a recent conference here, was this: Is the federal government’s call for “scientifically based research” in education a mandate or just strong advice?
Nearly 26 months after President Bush signed the federal No Child Left Behind Act, that question still looms.
Although the law is known best for its bold testing and accountability requirements, it has helped put the phrase “scientifically based research” on the lips of educational decisionmakers from coast to coast. It calls on educators to use research to guide a wide range of school practices, including reading instruction, professional development, technical assistance, and anti-drug programs.
Part of the problem is that the legislation itself contains two different definitions of the phrase—one for Reading First, President Bush’s flagship reading initiative, and one to cover the rest of the law’s provisions.
Since the law’s passage in 2001, the U.S. Department of Education has crafted regulations spelling out what it means by scientifically based research with regard to the reading program.
But no such regulations exist for the rest of the law—mostly because federal policymakers view the research base as weaker in those areas. As a result, even some of the law’s primary architects and enforcers are reluctant to call the phrase a mandate.
“It certainly is not a druthers situation. It is the law,” said Robert Sweet, who helped define “scientifically based research” in the law as a staff member on the House Committee on Education and the Workforce. “In practice, is everyone going to be held to a strict scientifically based research definition?” he continued. “Probably not.”
The question about scientifically based research came up at a March 11 conference on the topic organized by three Washington-based organizations: the congressionally chartered National Academies of Science, the National Education Knowledge Industry Association, a research trade group, and the Progressive Policy Institute, a centrist think tank.
At the meeting, both Mr. Sweet and Grover J. “Russ” Whitehurst, the director of the Education Department’s Institute of Education Sciences, said no “scientifically based research police” would be coming around inspecting schools. In due time, though, they hope educators will come to rely on such evidence.
States, on the other hand, are taking the federal law at its word, according to Ted Sanders, the president of the Education Commission of the States, a Denver-based research group for state policymakers. “I don’t think there’s any state in the country that looks at this as advisory,” he told conference-goers.
In fact, in a survey published this month by the Center on Education Policy, a national advocacy group based here, 64 percent of the districts polled said they require schools participating in the federal Title I program for disadvantaged students to select programs from a list of curricula or instructional programs grounded in scientifically based research.
Washington state, for example, last year turned down half the experts who applied to present at state-sponsored professional-development meetings for lack of scientific evidence, according to Dawn Billings, as assistant superintendent in that state’s education department.
“We have to do things differently,” she said last week. “We can’t go higgledy-piggledy doing this reform and that. We are taking this very seriously.”
Even so, Jack Jennings, the center’s director, thinks the numbers his survey turned up could be inflated.
“I think there’s a difference in people’s minds about whether they’re trying to use scientifically based research in improving education and whether they’re strictly following the requirements of federal law,” he said in an interview last week.
Both of the definitions of scientifically based research under the No Child Left Behind law call for studies that apply “rigorous, systematic, and objective” procedures and have been subjected to peer review.
But the definition that applies specifically to all programs outside of reading goes a few steps further. It expresses a clear preference for either randomized experiments or quasi-experiments, in which the group that receives the experimental treatment can be compared with another group that does not.
The definition also says studies should be detailed enough so that their findings can be replicated.
Though the Reading First definition is the more flexible of the two, states appear to be interpreting it very narrowly, partly because the Education Department sent many states’ program applications back for numerous revisions. (Feb. 4, 2004.)
Whether from a desire to appease federal reviewers or confusion over how to meet the new requirements, some states modeled their programs on those of the successful applicants that went before them.
In the process, some reading programs that are widely considered to have solid research track records were sidelined.
Success for All, a school improvement program established by researchers at Johns Hopkins University in Baltimore, is a case in point. Rated in 1999 by the American Institutes of Research as one of a handful of programs with “positive” evidence of effectiveness, the program got left off some of the earliest approved-program lists that states put together for Reading First.
“Daily, we’re getting reports from one state or another where somebody is being told Success for All is not a legitimate use of Reading First funds,” Robert E. Slavin, the program’s founder, said in an interview. “There’s an undercurrent that, if you want to be safe, you’re better go with one of the big [commercial] series.”