Educational research is often poorly funded as well, and the federal government and foundations often lean toward what is new, rather than what is proven effective.
"The gap between the research community and the practitioner community is much wider than what you'd find between practicing engineers and physicists," says John Bruer, president of the St. Louis-based James S. McDonnell Foundation, which supports educational and biomedical research.
Whether early readers should be taught whole language or phonics is a case in point. Last March, the National Research Council released a landmark report that announced a truce, or "pax lectura," in the nation's "reading wars." The report urged an end to take-no-prisoner swings from one method to another: Good reading instruction includes both, it argued.
"People in the field of reading are very passionate about correcting the errors that they see their predecessors as having made. The field looks faddish because people have gone too far down a reasonable road," says Catherine Snow, who chaired the National Research Council's report, "Preventing Reading Difficulties in Young Children."
"Many state boards of education had been losing faith in publishers and the education establishment for telling them what to do," she adds.
A key recommendation of this report is that schools require that publishers provide data to support claims of the effectiveness of their products.
"The reading wars had eased off even before publication of this report. If they hadn't we could never have reached a consensus," Ms. Snow notes.
Barbara Grohe, this year's National Superintendent of the Year, lived through two decades of the reading wars as a former reading teacher. "There has been a tendency to go from one end of the continuum directly to the other end without stopping in between to find a balance," she says. "It's almost as if to make your point that one approach doesn't work, you need to go to the exact opposite."