How to improve Reading First
April 18, 2007
Tomorrow, the House Education and Labor Committee will hold a hearing to consider the future of the much-discussed Reading First program, a key component of NCLB. While the hearing's title promises a focus on alleged "mismanagement and conflicts of interest" within the program, members of Congress would be better advised to concentrate on the future of federal policy in the domain of primary reading.
To consider the future, we're always well advised to start with a bit of history. Where did Reading First come from?
It came, above all, from mounting concern with the educational plight of far too many of our children. Despite honorable intentions, for years our governments, our educators, our media, and our scientists have been letting down kids, particularly poor kids and their families. When over 50 percent of underprivileged children keep failing in school and dropping out, something is not working.
Any effort to address that problem must begin with reading. An enormous proportion of young Americans cannot read well enough to learn about history, math, or science. Most such kids come from disadvantaged environments and many of their parents cannot read, either. Yet when it comes to educating these children, we continue to engage in practices and programs that have had no discernible effect on improving their reading capabilities.
For far too many years, the mainstays of instructional practices were superstition, tradition, and untested assumptions about how kids learn to read. Scientific research has rarely been applied to identifying effective instructional practices despite the fact that making responsible decisions about what is effective in classrooms and good for students requires scientific evidence. Until awfully recently, the practice of education resembled the practice of medicine a century back: virtually any treatment that could be thought up was tried out.
Though pioneering experts such as the late Jeanne Chall were on this case by the mid-sixties, it wasn't until the mid-1990s that the nation's leaders actually began to consider looking at different ways of addressing reading failure. When President Clinton included the rates of such failure in his 1996 State of the Union Address, it was the first time that reading instruction was recognized as a major issue by the Federal government. That mention provided the context for the development of the Reading Excellence Act, the antecedent of Reading First.
In 1997, I received a call from the office of the chairman of the House Committee on Education and the Work Force--the very same committee that is conducting tomorrow's hearing. Chairman Bill Goodling had learned that NIH's National Institute of Child Health and Human Development was conducting research on reading development, reading difficulties, and reading instruction, and had been doing so for years. He asked me to brief him on the findings. Goodling was surprised that NICHD had studied and supported research involving over 44,000 children and adults, many for more than 20 years, and that the findings of that research were validated and published in peer-reviewed journals. He was, frankly, shocked by the amount of evidence I showed him.
I explained that the phonics vs. whole-language debate was largely a waste of time, that reading was far too complex to place into such a binary straight jacket. In fact, I noted, reading requires the development of a number of complex skills that have to be integrated and practiced constantly.
After that meeting, the Reading Excellence Act was drafted and for the first time the phrase Scientifically Based Reading Research (SBRR) was introduced into law. Unfortunately, the 1998 Act provided little professional development or technical assistance and no systematic monitoring. Many states and local school districts received funds and spent them on whatever methods of reading instruction they were already using, scientific or not.
That federal funding should be contingent on evidence of effectiveness of a reading program or instructional strategy was not considered until 2001. It was then that my colleague Bob Sweet and I recommended such a framework to the new Bush administration. The idea caught on, and Reading First was born.
Unfortunately, as the program made its way through Congress, it was watered down. Rather than funding only programs with demonstrated effectiveness, Congress opted for the much broader and looser category of programs "based on scientific research." There are many possible reasons for this change including the fact that only a handful of programs had been "proven" effective. It's possible that implementing Reading First with so few programs would not have proved practical. However, we also knew that members of Congress were heavily lobbied by developers of non-proven programs who did not want to be excluded. Many were lured by the promise of federal dollars.
This dilution by Congress has had significant negative consequences. For example, some vendors of reading programs simply changed the language in their promotional materials to create the impression that they are "based on scientific research"--without making any real changes (see here).
Even with its flaws, Reading First remains incredibly important. It encourages reading instruction that is comprehensive, that is based on scientific research, and that is taught using direct and systematic instructional principles.
How to make it better? Congress should make two key changes. First, federal funds should only be used for those programs and instructional models that have been found to be effective using experimental research designs that can determine their causal impact on student achievement in reading. "Scientifically-based" programs should be replaced with "scientifically-proven" ones.
Second, Congress should stop dancing around the "local control" issue and simply ask a federal agency to vet the reading research and to determine, on a regular basis, which reading programs make the "scientifically-proven" cut. In other words, Congress should create for reading (and perhaps other subjects where scientific research can be done) the equivalent of an FDA for education to ensure that states and school districts only spend their Reading First funds on interventions that have been conclusively shown to work. (The "What Works Clearinghouse" might serve as a model.)
Five years after Reading First became law, evidence is beginning to show that it's starting to move publishing companies and diagnostic assessment creators towards a higher standard (see here and here). Moreover, the intent and the language of Reading First are now contained in discussions about reading in state departments of education, school districts, and individual schools. And the educational community has been energized by Reading First and its promise that all students should be provided with research-based instruction. It is difficult to visit any school and not hear administrators and teachers discussing whether particular instructional programs and strategies have sufficient evidence of effectiveness. Several publishing companies and program developers have now invested in research initiatives to test the effectiveness of their own programs. Five years back, few programs had done anything of the sort.
It would be irresponsible to allow medical practitioners to revert to superstition, anecdote, and snake oil in the 21st century. Why would we have teachers do that? Rather than moving backwards to the day when federal dollars flowed to any program under the sun, regardless of effectiveness, we need to continue moving forward to the day when education is truly a research-based enterprise. The life prospects of millions of children depend on it.