Federal Law Will Require Research-Based Programs

Published April 1, 2002

The teaching of reading could become the first area of education to benefit from a requirement, repeated more than 100 times in the new K-12 education law, that “scientifically based research” be the linchpin of all school practices receiving federal subsidies.

Education has been notorious for flitting from fad to fad with nothing more substantive than clever marketing techniques to justify curricular changes. Indeed, the federal government has been part of the problem.

In congressional testimony in 1999, Maris Vinovskis, a senior research scientist at the University of Michigan’s Institute for Social Research, noted Washington has been collecting, analyzing, and disseminating education statistics for more than 130 years. Over time, the focus has shifted from data-gathering to research on effective ways to educate children at the state and local levels.

However, added Vinovskis, who worked in the federal Office of Educational Research and Improvement (OERI) in 1992-93, “the quality of work … has varied greatly. As a result, educational research and development usually has not been held in high esteem by most academics and policy-makers in the twentieth century.”

In February, a well-attended U.S. Department of Education (DoEd) conference looked at what the research reference in the No Child Left Behind Act means. One indication DoEd is taking the new mandate seriously is its establishment of a What Works Clearinghouse, which will draw on independent contractors to determine whether touted educational products have legitimate research behind them.

Wanted: Sound Scientific Research

Leading off the conference, Valerie Reyna, senior research advisor in DoEd’s OERI, noted that while anecdotal evidence is often entertaining, “we can’t base practice on it, as is well known in medicine.” She expressed the department’s hope that “the use of scientific research as a basis for educational practice will become routine.”

The strongest form of research experiment, known as the “gold standard,” is the randomized clinical trial, in which pure chance determines into which of two comparison groups the subjects are assigned. However, even the results from these studies can be compromised when something less than pure chance goes into the assignment of subjects.

Sound research can be conducted without randomized clinical trials, participants in the DoEd conference noted. But when researchers are weighing evidence by use of statistics or theories, they should take care to consider before announcing conclusions that there may be factors for which they have not controlled.

While some will charge science is “soulless and heartless” when applied to education, Reyna said it is a false dichotomy to pit science against values, or science against emotion. Relying on scientific research doesn’t take the heart out of education. Instead, “facts give the students a chance for success. Without facts, we might make the wrong decision.”

Reading Research Findings

Reading appears to be one area of education where practice finally is catching up to research.

Illinois reading consultant Eunice Greer said that 20 years ago, a common line of reasoning about beginning reading was, “a lot of different things work for different kids. If one doesn’t work, just try another.”

In sharp contrast, she said, today “there is a converging body of research” about the essential elements of teaching all children to read.

What brought the picture into focus was the April 2000 report of the National Reading Panel, which examined 100,000 studies of reading instruction done since 1966. Convened by the National Institute of Child Health and Human Development, the panel identified five essential elements of reading instruction: phonemic awareness, phonics, fluency, vocabulary, and text comprehension.

Phonemes are the smallest units of spoken language; for instance, the words “go” and “she” each consist of two phonemes. Acquisition of phonemic awareness, best taught by “working with sounds–making noise, not worksheets,” is one of the best predictors of whether a child will learn to read during the first two years of instruction.

In teaching phonics–the relationship between written language and sounds–research shows that “explicit, systematic instruction is more effective than non-systematic instruction.” Moreover, phonetic instruction “is most effective when begun in kindergarten or first grade,” Greer said.

Fluency–the ability to read rapidly and with expression–is the “most neglected” aspect of reading instruction, Greer noted. Repeated, monitored oral reading is the surest pathway.

With vocabulary, the larger a teacher’s vocabulary, the easier it is for children to comprehend. Children learn only 8 to 10 new vocabulary words taught directly in a week; they learn many more indirectly, Greer noted.

Text comprehension is the ultimate goal. This means the development of “purposeful and active readers.” Teachers can begin to develop this systematically by cultivating listening and story comprehension at a very early stage.

Math Research Findings

Mathematics researchers have a way to go before they reach a consensus as firm as what has emerged with reading, said Russell Gersten of the University of Oregon’s Eugene Research Institute.

Gersten did, however, cite promising results from studies of low-achieving pupils who are tested every four weeks to determine where they are in math so they can be brought up to speed quickly with tutoring. A weak math student is paired up with a strong one for one-on-one peer tutoring. With just two students in a tutorial setting, the teacher can easily monitor what’s going on; not so in a group of four or five, noted Gersten.

For the future, he said a useful focus of research could be an evaluation of early prevention programs and an investigation of what sort of valid predictors there are for math achievement. Some children just seem to have “number sense,” a natural ability to grasp numeric concepts. A challenge for researchers is to investigate how that might be turned into teaching strategies to reach students with less natural “number sense.”

Robert Holland is a senior fellow at the Lexington Institute, a public policy think tank in Arlington, Virginia. His email address is [email protected].