Publication Date: 2003-04-30
Here, in an Education Week Commentary, Joanne Yatvin tells the real story of The National Reading Panel Report.
April 30, 2003
A member of the National Reading Panel blasts policy makers for distorting the truth to advance political agendas.
Late in the process of putting together the National Reading Panel report, I declared my intention to write a minority view. Although, as a school-based educator on the panel, I failed to notice the scientific flaws in some sections of the report that critics have since pointed out, I could see that the report as a whole was narrow, biased, and elitist. I was troubled by the fact that it covered so few of the important issues in teaching reading today, that the ideologies of panel members had governed the choice of most of the topics, and that teachers of beginning reading had been excluded from the entire process. Moreover, I believed that because of its weaknesses, the report was dangerous in its potential for misuse. The responses of my fellow panel members to my concerns ranged from reassurances that all problems would be ironed out before the report's publication to "So what?"
Back then, in the winter of 1999, I couldn't answer that question. I couldn't predict how many false conclusions, misinterpretations of data, and simplistic judgments would be made by careless readers. I had no idea how far policymakers, bureaucrats, and special-interest groups would go in distorting the truth to advance political agendas. And, in my heart, I believed that panel members would not let those things happen; that they would speak up and tell the world, "No, that is not what we said." Now?on all counts?I know better.
In the three years since its publication in 2000, the findings of the National Reading Panel report have been used to support the research agenda of the National Institute of Child Health and Human Development and the Reading First initiative of the federal "No Child Left Behind" Act of 2001.
In addition, government officials and promoters of phonics have twisted those findings in an effort to reconfigure all school reading instruction and all teacher preparation in reading to conform with their own ideas of how reading should be taught. In the process of applying for federal funds through Reading First, states that have designed successful models of teacher training and school districts that have developed effective reading programs have been told that their plans are not sufficiently "scientific," or "systematic," and that they will have to change them. University professors of reading have been criticized for not having evidence of "knowledge of research-based methods" in their vitae. In short, any program or any educator that does not fit with today's fashionable orthodoxy is considered unfit for the teaching of reading.
In the interest of truth and fairness?not to mention educational quality?I offer my own account of what the report says and does not say. Although I still maintain, as I did in my minority view, that the NRP report is not complete or objective enough to stand as the foundation for reading instruction in America's schools, I believe that its findings, reported accurately, do provide some valuable guidance for schools and teachers.
But that accurate reporting is too seldom in evidence. Below are the major misconceptions in the public mind about the NRP report, followed by my informed recollections of the panel and its work, and explanations of what the report actually contains. I have presented these in a True-False format:
FALSE: The National Reading Panel was a diverse and balanced group of reading experts.
TRUE: Congress asked for a balanced panel, but that's not what it got.
As specified by Congress, the director of the National Institute of Child Health and Human Development was to convene a panel made up of "leading scientists in reading research, representatives of colleges of education, reading teachers, educational administrators, and parents." The NICHD read all but one of those plural nouns as singular and threw in a couple of ringers.
In the interest of truth and fairness?not to mention educational quality?I offer my own account of what the report says and does not say.
Out of the 15 people appointed, nine were reading researchers, two were university administrators with no background in reading research or practice, one was a teacher- educator, one a certified public accountant (and parent), one was a middle school teacher, and one an elementary principal (me). When one researcher resigned after the first panel meeting, the NICHD declined my request that he be replaced by an elementary-level teacher and left that position unfilled. As a result, the panel included no teacher of early reading instruction.
Moreover, the science faction of the panel could hardly be considered balanced. All were experimental scientists; all were adherents of the discrete- skills model of reading; and some of them had professional ties to the NICHD. With so many distinguished reading researchers available in the United States, it is difficult to understand why the NICHD could not find one or two involved in descriptive research or with a different philosophy of reading.
FALSE: The panel carried out a comprehensive analysis of the entire field of reading research.
TRUE: Only a small fraction of the field was considered, and only a few hundred studies were actually analyzed.
Using electronic databases, the panel estimated that over 100,000 studies of reading had been published between 1966 and 1998, when it began its review of the research literature. Because the vast majority of these studies were on topics that the panel did not choose to investigate, they were never even looked at. Thousands more were eliminated at first glance because it was obvious from their titles or abstracts that they did not fit the panel's criteria. In the end, 432 studies on nine topics were reviewed and reported on. Yet, government documents and press releases, and news articles based on those sources have continued to assert that the panel reviewed 100,000 studies. Even the subtitle that the NICHD put on the panel's report: "An Evidence-Based Assessment of the Scientific Research Literature on Reading and Its Implications for Reading Instruction," misleads readers by implying that a comprehensive review of the field of reading research was undertaken.
FALSE: The panel determined that there are five essentials of reading instruction.
TRUE: Although the NRP reported positive results for five of the six instructional strategies it investigated, it never claimed that these five were the essential components of reading.
At its first meeting, the panel divided itself into three subgroups (Alphabetics, Fluency, and Comprehension), and each subgroup selected its own first topic on the basis of its members' interests. After regional meetings with citizens and educators, two more subgroups (Teacher Education and Computer Technology) were added, and they, too, chose their own topics.
At its third meeting, the panel identified 35 additional topics that merited investigation, but soon discovered that it did not have enough time or resources to study them. In the "Next Steps" section of the summary booklet, the panel expresses its regret at not being able to examine all worthy topics, and states: "The panel emphasizes that omissions of topics such as the effects of predictable and decodable text formats on beginning reading development, motivational factors in learning to read, and the effects of integrating reading and writing, to name a few, are not to be interpreted as determinations of unimportance or ineffectiveness." Nowhere in its report does the panel assert that the strategies found effective are the "essentials" of reading instruction. That determination was made elsewhere, embodied in the No Child Left Behind Act, and then included in the guidelines for Reading First. Ultimately, references to the "five essentials of reading" appeared in state applications, media commentaries, and promotional literature for various commercial programs.
FALSE: The panel endorsed only explicit, systematic instruction.
TRUE: Only in the phonics subgroup report is "explicit, systematic" instruction called essential.
Although encouragement for some measure of "explicit" teaching also appears in the conclusions of the reports on phonemic awareness, vocabulary, and comprehension strategies, the panel acknowledges that all children acquire a portion of these skills incidentally, and that many children become quite proficient without instruction. Incidental learning is particularly emphasized in regard to vocabulary, where reading and discussion play a significant role in the acquisition of new words.
FALSE: The panel identified certain comprehensive commercial reading programs as being research-based, and concluded that teachers need one of these programs, or a comparable program, to teach children effectively.
TRUE: No comprehensive reading programs were investigated by the panel. The panel had nothing to say about whether teachers need a commercial program or can develop their own.
Any program or any educator that does not fit with today's fashionable orthodoxy is considered unfit for the teaching of reading.
The only area where any subgroup examined commercial packages of instruction was in phonics, and that was possible because enough independent research already existed on seven packages. The panel concluded that all seven were comparable and more effective than instructional designs lacking a phonics component. Assertions made by government officials in advising Reading First applicants, and claims made by publishers in advertising their products, that certain comprehensive programs are "research based" (with the implication that others and teacher-constructed programs are not) are not supported by anything in the National Reading Panel report.
FALSE: The panel identified phonics as the most important component of reading instruction throughout the elementary grades.
TRUE: The panel made no such determination.
In the report, the phonemic-awareness section comes first, phonics second, then fluency, and finally, comprehension. No explanation is given for this ordering of sub-reports, although it does conform to the sequence of skills hypothesized in the skills model of reading.
FALSE: The panel found that phonics should be taught to all students throughout the elementary grades.
TRUE: The panel found no evidence to justify teaching phonics to normally progressing readers past 1st grade.
The results of the meta-analysis done by the phonics subgroup showed that effects were greatest in kindergarten and 1st grade and declined steadily after that. They concluded that phonics "failed to exert a significant impact on the reading performance of low-achieving readers in 2nd through 6th grades." In addition, they found that phonics produced no significant gains in comprehension for normal readers above 1st grade.
FALSE: The panel's findings repudiate whole language as an approach to teaching reading.
TRUE: The panel did not investigate whole language as a topic and did not draw any conclusions about it as an approach to teaching reading.
In reviewing its six instructional strategies, however, the panel touched on aspects of reading ordinarily included in whole-language programs (cooperative-learning groups for comprehension, repeated oral reading of texts for fluency) and found that teaching them had positive effects. Furthermore, in four of the phonics studies, the only ones done in regular 1st or 2nd grade classrooms, children taught by a whole-language approach did as well (in one study, better) in overall reading achievement as those taught by phonics only or basals.
FALSE: The panel found research evidence indicating how teachers should be trained to teach reading.
TRUE: The panel found no such evidence.
The body of research on teacher education as it relates to student learning was so small and diverse that the panel was not able to answer the critical question of how teachers can be educated to teach reading effectively. The only conclusion the panel was able to draw from the data available was that the amount of in-service education teachers receive has a positive effect on the learning of their students. The panel felt that this evidence was strong enough to assume that the amount of teachers' preservice education would also have a positive effect.
These are the facts of the National Reading Panel report, not only as I remember them, but as verified by the first edition of the full report and the transcribed proceedings of the panel meetings. If others who were involved in the creation of the report wish to challenge my accuracy, I invite them to do so. So far, all that has been heard is the "official line" from individuals and groups with their own vested interests, and who, at best, were working behind the scenes while the NRP was producing its report. An honest debate between those in the know would go far to serve the public interest.
Joanne Yatvin was a member of the National Reading Panel and wrote a "minority view" of its report. At that time, she was an elementary school principal in Boring, Ore. She has since retired and is an adjunct instructor in the graduate school of education at Portland State University, in Portland, Ore.
On the Web
Dr. Yatvin's minority view is published by the National Institute of Child Health and Human Development. (Requires Adobe's Acrobat Reader.) The National Reading Panel's full report, "Teaching Children to Read," is also available from NICHD.