1703 North Beauregard St.
Alexandria, VA 22311-1714
Tel: 1-800-933-ASCD (2723)
8:00 a.m. to 6:00 p.m. eastern time, Monday through Friday
Local to the D.C. area: 1-703-578-9600
Toll-free from U.S. and Canada: 1-800-933-ASCD (2723)
All other countries: (International Access Code) + 1-703-578-9600
March 2004 | Volume 61 | Number 6
What Research Says About Reading
G. Reid Lyon and Vinita Chhabra
To prevent reading failure, educators must understand and act on scientific evidence.
Because students who do not learn to read will have difficulty mastering academic content, succeeding in school, and fulfilling their life potential, the schools' fundamental responsibility is to ensure that all students read proficiently. Yet that goal remains elusive. The report of the most recent National Assessment of Educational Progress (NAEP) indicates that once again, 4th, 8th, and 12th grade reading scores are abysmally low, particularly among disadvantaged students (National Center for Education Statistics, 2003).
This situation is especially distressing because we now know that the majority of students can learn to read irrespective of their backgrounds—if their reading instruction is grounded in the converging scientific evidence about how reading develops, why many students have difficulties, and how we can prevent reading failure (Lyon, 2002; Moats, 1999; Shaywitz, 2003). Unfortunately, many teachers do not have the background or training they need to access this information and implement research-based reading instruction in their classrooms.
Instead of looking to authoritative and valid evidence of what works, teachers frequently rely on experience and anecdotal information to guide their teaching. Although experience is valuable, depending on experience alone typically leaves many children behind. To provide reading instruction that will enable all students to succeed, educators must also have basic information about scientific knowledge: how it is developed and how it should guide the selection and implementation of instructional programs, strategies, and approaches (McCardle & Chhabra, 2004; Moats, 1999; Stanovich & Stanovich, 2003).
Scientific evidence must become a fundamental part of teaching. Teachers should ask hard questions about the reading programs and instructional methods they use: Does the program comprehensively cover each of the evidence-based skills that students need to read proficiently? Has the program or approach been proven scientifically to work with students like mine? How can I be sure?
From talking to principals and teachers, we know that they are aware of the current emphasis on using scientific evidence to inform instruction. However, jargon and a lack of clear guidance make it difficult for educators to develop a good understanding of the term scientific evidence.
“Research evidence is essential for identifying effective educational practice” (Reyna, 2004, p. 47). Before considering research evidence, however, we must understand the process of conducting scientific research.
Scientific research begins with clear, answerable questions. At the National Institute of Child Health and Human Development (NICHD), complex questions guide researchers as they examine instructional methods or programs for teaching reading. Researchers consider not just what works in general but also what specific instructional methods work for which students and under what conditions. Multiple researchers may study the same questions. The accumulation of research—not just the results from one study—constitutes scientific evidence.
Before beginning a study, researchers identify what previous research has already established and refine their research questions on the basis of that knowledge. Next, the researchers develop a hypothesis or a set of hypotheses that will guide the direction and methodology used in the study. Specifically, the hypotheses state what researchers predict will happen on the basis of the theoretical model they are using, with the model itself incorporating information from previously conducted, high-quality studies. The researchers then identify an appropriate method for testing the hypothesis. Finally, they collect and analyze the data and report their findings.
Even though different types of studies use different method-ologies, the scientific method provides a consistent foundation for each research study. The evidence derived from a study is only as good as the care with which the researchers ask their questions, select and implement their research methods, and analyze and interpret their data.
Research methodologies fall into two major types—qualitative and quantitative. Both types of research can be valuable and relevant to improving education. Both are equally scientific—if we align the right method with the specific question or questions being addressed (Shavelson & Towne, 2002; Wiersma, 2000).
Qualitative research is based on holistic inquiry and is context-specific, acknowledging the uniqueness of individuals and settings (Ravid, 1994). Researchers using this approach rely on observation and description of events.
In contrast, quantitative research seeks to quantify observations made on larger numbers of individuals and events; it is not context-specific. Researchers working with this approach use deductive inquiry, focus on individual variables and factors, and make comparisons (Wiersma, 2000). Quantitative researchers look for findings that they can generalize to similar settings using statistical analyses.
Quantitative research attempts to answer questions about “what causes what.” To draw reliable inferences about cause and effect—for example, to determine whether a particular instructional approach produces significant gains in reading achievement—researchers must use specific types of quantitative research known as experimental and quasi-experimental studies.
Both experimental and quasi-experimental studies attempt to establish causal relationships by manipulating the variable that the hypothesis predicts will cause a change (a particular reading program, for example) and then determining whether or not a significant change occurred while holding other variables constant. The two approaches differ in the methods they use to rule out alternative explanations for any effects. True experimental studies typically assign people randomly to the experimental condition (Stanovich & Stanovich, 2003), whereas quasi-experimental studies control for alternative explanations of effects through statistical means. Although true experimental design typically produces more reliable results than quasi-experimental design, both approaches enable us to determine cause and effect.
The type of research question asked tells us which research method (or combination of methods) to employ. Qualitative research has a role in helping to tease out the how or why in quantitative studies designed to find out what works. But qualitative research cannot produce generalizable results identifying instructional strategies that will improve reading achievement. Only quantitatively based studies—experimental or quasi-experimental—can do that. However, both quantitative and qualitative research methods are important and necessary if we are to develop the fullest understanding of which specific instructional approaches are most effective for which students and why (Shavelson & Towne, 2002).
Many teachers and principals, battered by a succession of “research-based” instructional magic bullets, have lost faith in the idea that research can inform their teaching. This should not surprise us. Education research has historically been of poor scientific quality, lacking in the authority of valid evidence, poorly communicated, and woefully impractical (Kennedy, 1997). Yet when new research findings emerge with recommendations for classroom implementation, educators have an important responsibility to ask, Can we trust the findings of this study?
What principles or criteria can teachers and administrators apply to research reports to ensure that those reports meet high standards of trust-worthiness? Major elements include appropriate methodologies, peer review, converging evidence, and practical application.
When presented with a curriculum, a program, or an approach that has been touted as “research-based,” teachers and administrators must ask, Did the researchers use the appropriate research methods to determine actual effectiveness? For example, research should not use observational measures, a type of qualitative research, to answer questions about the effects of reading interventions and the applicability of these interventions to similar classrooms. As Reyna writes,
Classroom observations under the best of circumstances (systematic and reliable observers) do not even permit generalization to other classrooms. (2004, p. 54)
Even after researchers have followed appropriate criteria for conducting research, it is crucial that an independent group of researchers who specialize in the same topic evaluate the studies:
The front line of defense for teachers against incorrect information in education is the existence of peer- reviewed journals in education, psychology, and other related social sciences. (Stanovich & Stanovich, 2003, p. 7)
Peer review is a minimal benchmark for trustworthiness. Scientific documentation for effectiveness can only be found in journals that publish original research, such as The Journal of Educational Psychology, Reading Research Quarterly, and Scientific Studies in Reading. The studies considered for publication in these journals are first reviewed by an objective group of scientists in the same field to ensure high standards for design and methodology and to confirm the veracity of the findings. Although there are journals that publish editorials or opinion pieces (Educational Leadership and Phi Delta Kappan, for example) and provide valuable information, these periodicals are not designed to evaluate and publish the results of original scientific studies.
The good news is that researchers must now adhere to rigorous standards to have their scientific research supported through federal funds provided by NICHD or the Institute for Educational Sciences within the U.S. Department of Education. Specifically, to obtain federal funding to conduct research, the studies proposed must pass a rigorous peer review process that ensures that the information gained is accurate and ready for implementation in classroom settings.
A single study's findings are not sufficient to generalize results to different populations. An accumulation of high-quality research across similar topic areas will verify the accuracy of findings and increase our confidence in those findings. For example, the National Reading Panel (NRP) found that systematic phonics instruction produced significant benefits for K-6 students and for those having difficulty learning to read (2000). The NRP also found that systematic phonics instruction was most effective when provided within the context of a comprehensive reading program that also addressed phonemic awareness, fluency, vocabulary, and comprehension strategies. These conclusions were based on a collection of similar findings replicated across many research studies conducted over several years.
New resources can help teachers locate reading programs supported by converging research evidence. The state of Florida has funded a scientific review of many reading programs used frequently in schools. Teachers and administrators can access information on these reviews from the Florida Center for Reading Research Web site (www.fcrr.org/FCRRReports/contents.htm). In addition, the U.S. Department of Education is in the process of developing the What Works Clearinghouse (www.w-w-c.org), a resource that will provide educators, policymakers, and the public with a central and independent source of scientific evidence of what works in education.
Another important question about reported research findings is whether teachers can implement those findings in their classrooms. To determine whether or not the findings have practical instructional value for their students, teachers should ask whether similar classrooms implemented the research findings with good results. In addition, teachers and administrators may want to consult recent publications that have summarized reading research findings along with practical teaching recommendations based on the findings. For instance, Put Reading First: The Research Building Blocks for Teaching Children to Read is based on an accumulation of research and offers teachers practical examples of how to teach reading skills in the classroom, as well as ways in which teachers can identify components of good reading programs (Partnership for Reading, 2003).
An expanded understanding of the principles of scientific evidence will enable educators to take advantage of the converging evidence that research now provides about learning to read. What does this research tell us?
First, scientific evidence clearly indicates that learning to read the English language is not as easy as some have suggested (Adams, 1990; Fletcher & Lyon, 1998; Liberman, 1997; Lyon, 1998; NRP, 2000; Rayner, Foorman, Perfetti, Pesetsky, & Seidenberg, 2001; Shaywitz, 2003; Snow, Burns, & Griffin, 1998). The sheer number of youngsters who fail to learn to read—particularly those from impoverished backgrounds—underscores the fact that children do not acquire reading ability naturally, easily, or incidentally. For many children, learning to read is a monumental task that requires years of effort.
As Moats (1999) points out, because of the complexity of learning to read, teaching reading is clearly a job for an expert; in fact, she compares it to rocket science. Given that reading instruction requires substantial expertise, what does the expert teacher need to know? At a minimum, teachers and administrators should be able to answer the following two questions.
Learning to read is a relatively lengthy process that begins very early in development, before children enter formal schooling. The quantity and quality of language and early literacy interactions during the preschool years profoundly affect the acquisition of the language building blocks that support skilled reading (Snow et al., 1998). As noted in all of the NAEP reading results for the past quarter of a century, reading failure is most prevalent among children from disadvantaged environments. The gap between these children and their more affluent peers begins early. Lonigan (2003) found that low-income preschool children were significantly less adept at identifying and manipulating the sound structure of language—a skill known as phonological sensitivity—than were middle- and high-income children. Low-income children also experienced significantly less growth in knowledge of phonemes, letter names, and letter sounds. Vocabulary development, phonological sensitivity, and alphabetic skills are extremely strong predictors of later reading proficiency; typically, children from low socioeconomic backgrounds lag woefully in these abilities (Snow et al., 1998).
The major reason for these gaps, particularly in vocabulary, is that children growing up in low-income environments engage in significantly fewer language and literacy interactions during the preschool years (Hersch, 2001; Snow et al., 1998). Hart and Risley (2003) found that the average child on welfare had half as much experience listening and speaking to parents (616 words per hour) as the average working-class child (1,251 words per hour), and less than one-third that of the average child in a professional family (2,153 words per hour). Unfortunately, preschool programs for low-income children, such as Head Start, rarely address these gaps (Whitehurst, 2001).
Crucial language and early reading skills can be developed in kindergarten and the elementary grades, but teachers must have a solid foundation in how to teach these abilities (Moats, 1995). Until recently, most teacher preparation programs have not provided this training (Moats, 1999).
In addition to the development of these early foundational abilities, learning to read requires continued development of phonemic awareness, phonics, reading fluency, vocabulary, and reading comprehension strategies (NRP, 2000). Reading proficiency depends on the expert and integrated teaching of these skills so that the reader learns to access print accurately and fluently and to relate what he or she reads to vocabulary and content knowledge to ensure comprehension.
The majority of children who enter kindergarten and elementary school at risk for reading failure can learn to read at average or above-average levels—if they are identified early and given systematic, intensive instruction in phonemic awareness, phonics, reading fluency, vocabulary, and reading comprehension strategies (Lyon et al., 2001; Torgesen, 2002a). Substantial research carried out and supported by NICHD indicates clearly that without this systematic and intensive approach to early intervention, the majority of at-risk readers rarely catch up. Failure to read by 9 years of age portends a lifetime of illiteracy for at least 70 percent of struggling readers (Shaywitz, 2003).
On the other hand, the identification of children at risk for reading failure coupled with the provision of systematic, comprehensive, and evidence-based reading interventions can reduce the number of students reading below basic levels to less than 6 percent (Torgesen, 2002a).
NICHD-supported research has also found that older struggling readers can indeed develop strong reading capabilities under the right instructional conditions (Lyon, 2002). However, successful remediation of reading problems among older students requires extensive, intensive instruction (see Torgesen, 2002b). Moreover, in addition to providing a comprehensive reading program, consistency and duration of instruction are crucial factors for older readers (Shaywitz, 2003).
Current instructional practices and policies are failing our most vulnerable students. How many more children need to fail before we realize that human potential and life are too precious to risk on the basis of untested beliefs and opinions about what works?
Any other helping profession would respond to a persistent problem such as pervasive reading failure—a problem that deeply affects the quality of life—by objectively and relentlessly identifying the causes of the failure and systematically developing solutions. The scientific method would typically inform this response to system failure. This method has solved many of the world's thorniest medical mysteries—mysteries just as complex as learning to read.
The scientific method is already realizing some of its potential to improve reading instruction. Current research has shown that basing our instructional practices and policies on converging scientific evidence will result in significantly more students learning to read proficiently. As more students learn to read, including those from disadvantaged backgrounds, we hope that the tendency within education to embrace instructional fads will give way to more effective practices based on solid, scientific research.
Adams, M. J. (1990). Beginning to read: Thinking and learning about print. Cambridge: MIT Press.
Fletcher, J. M., & Lyon, G. R. (1998). Reading: A research-based approach. In W. Evers (Ed.), What's gone wrong in America's classrooms (pp. 49–90). Stanford, CA: Hoover Institute Press.
Hart, B., & Risley, T. R. (2003). The early catastrophe: The 30 million word gap. American Educator, 27(1), 4–9.
Hersch, E. D. (2001). Overcoming the language gap. American Educator, 25(2), 4–7.
Kennedy, M. M. (1997). The connection between research and practice. Educational Researcher, 26, 4–12.
Liberman, A. M. (1997). Why is speech so much easier than reading? In H. C. Joshi & R. M. Joshi (Eds.), Reading and spelling: Development and disorders. Mahwah, NJ: Erlbaum.
Lonigan, C. J. (2003). Development and promotion of emergent literacy skills in children at risk of reading difficulties. In B. R. Foorman (Ed.), Preventing and remediating reading difficulties: Bringing science to scale (pp. 23–50). Baltimore: York Press.
Lyon, G. R. (1998, March). Why reading is not a natural process. Educational Leadership, 14–18.
Lyon, G. R. (2002). Reading development, reading difficulties, and reading instruction: Educational and public health issues. Journal of School Psychology, 40, 3–6.
Lyon, G. R., Fletcher, J. M., Shaywitz, S. E., Shaywitz, B. A., Torgesen, J. K., Wood, F. B., Shulte, A., & Olson, R. (2001). Rethinking learning disabilities. In C. E. Finn, R. A. J. Rotherham, & C. R. Hokanson (Eds.), Rethinking special education for a new century (pp. 259–287). Washington, DC: Thomas B. Fordham Foundation & Progressive Policy Institute.
McCardle, P., & Chhabra, V. (2004). The voice of evidence in reading research. Baltimore: Brookes.
Moats, L. C. (1995). The missing foundation in teacher preparation. American Educator, 19(9), 43–51.
Moats, L. C. (1999). Teaching reading is rocket science. Washington, DC: American Federation of Teachers.
National Center for Education Statistics. (2003). National assessment of educational progress: The nation's report card. Washington, DC: U.S. Department of Education.
National Reading Panel. (2000). Teaching children to read: An evidence-based assessment of the scientific research literature on reading and its implications for reading instruction. Reports of the subgroups. Washington, DC: National Institute of Child Health and Human Development.
Partnership for Reading. (2003). Put reading first: The research building blocks for teaching children to read. Kindergarten through grade 3. Washington, DC: Author.
Ravid, R. (1994). Practical statistics for educators. Lanham, MD: University Press of America.
Rayner, K., Foorman, B. R., Perfetti, C. A., Pesetsky, D., & Seidenberg, M. S. (2001). How psychological science informs the teaching of reading. Psychological Science in the Public Interest, 2(2), 31–74.
Reyna, V. (2004). Why scientific research? The importance of evidence in changing educational practice. In P. McCardle & V. Chhabra (Eds.), The voice of evidence in reading research. Baltimore: Brookes.
Shavelson, R. J., & Towne, L. (2002). Scientific research in education. Washington, DC: National Academies Press.
Shaywitz, S. E. (2003). Overcoming dyslexia. New York: Knopf.
Snow, C., Burns, S., & Griffin, P. (Eds.). (1998). Preventing reading difficulties in young children. Washington, DC: National Academy Press.
Stanovich, P., & Stanovich, K. (2003). Using research and reason in education. Washington, DC: Partnership for Reading.
Torgesen, J. K. (2002a). The prevention of reading difficulties. Journal of School Psychology, 40(1), 7–26.
Torgesen, J. K. (2002b). Lessons learned from intervention research in reading: A way to go before we rest. In R. Stainthorpe (Ed.), Literacy: Learning and teaching. London: British Psychological Association.
Whitehurst, G. (2001). Cognitive development during the preschool years. Paper presented at the White House Summit on Early Childhood Cognitive Development. Washington, DC: U.S. Department of Education.
Wiersma, W. (2000). Research methods in education: An introduction (7th ed.). Boston: Allyn and Bacon.
Editor's note: For a further discussion of this article's findings, see “Continuing the Discussion” on p. 86.
G. Reid Lyon (email@example.com) is Chief and Vinita Chhabra (firstname.lastname@example.org) is a Research Scientist at the Child Development and Behavior Branch, National Institute of Child Health and Human Development, Rockville, Maryland.
Copyright © 2004 by Association for Supervision and Curriculum Development
Subscribe to ASCD Express, our free e-mail newsletter, to have practical, actionable strategies and information delivered to your e-mail inbox twice a month.
ASCD respects intellectual property rights and adheres to the laws governing them. Learn more about our permissions policy and submit your request online.