Adolescent Literacy—Research Informing Practice: A Series of WorkshopsSummary of the Adolescent Literacy Workshop: State of the Science and Research NeedsThis workshop was the first of two designed for the purpose of developing specific recommendations for a program of research on adolescent literacy. Together these workshops will draw on the knowledge and experiences of researchers and practitioners who work with adolescents and their reading challenges. The March workshop brought together primarily researchers, but also some practitioners, to review and summarize the knowledge base and gaps relevant to adolescent literacy and to develop a research agenda. The workshop was jointly sponsored by government and private sector entities: the National Institute of Child Health and Human Development, the American Federation of Teachers, the American Speech-Language-Hearing Association, the International Reading Association, the National Education Association, the National Institute for Literacy, and four Offices of the U.S. Department of Education: Vocational and Adult Education, Educational Research and Improvement, Elementary and Secondary Education, and Special Education and Rehabilitation Services. The format of the workshop was a combination of presentations by speakers, comments from panels of respondents, and open discussion from the audience. On the third day, all attendees at the workshop participated in one of four small groups with the charge to identify the research questions of highest priority. Rationale for the Workshop: Research Needs in Adolescent Literacy Significant advances have been made in understanding the abilities young children must acquire to develop beginning reading skills and the conditions under which they are most effectively taught, but very little evidence is available on how these abilities are best acquired and taught during adolescence. Specifically, it is well known that in learning to read, kindergarten and elementary school-aged children must develop adequate alphabetic reading skills (including phonemic awareness and phonics abilities) and the ability to apply these word-reading skills fluently to both decoding and text-reading activities, and they must develop background knowledge, vocabulary, and reading-comprehension strategies to facilitate their understanding of what they read. But there are obvious critical influences that age and experience bring to the process of learning to read and write, particularly if basic reading skills are not developed prior to third grade. However, we do not know to what extent what is known about beginning reading instruction applies to older students who fail to acquire the building blocks of reading. First, it has been frequently stated that learning to read may be more difficult after age 9. However, the factors that might explain this decreased learning ability are not well understood. Do older students make little progress even when provided intensive, explicit, and systematic instruction in the building blocks of reading that they have not yet acquired? If so, does this failure to progress reflect a biologically based critical period for the acquisition of reading-related abilities prior to age 9? Does protracted failure in learning to read in kindergarten and elementary school lead to declines in self-esteem, increases in negative peer interactions, and other emotional factors that reduce motivation? Are the necessary quality, duration, and intensity of reading instruction simply not available in the upper-elementary, middle, and high school years? Or are combinations of these and other factors responsible for this apparent resistance to reading intervention by middle and high school students? Second, it is not well understood which specific reading abilities are most predictive of difficulties in learning to read in adolescence. Do the relationships between phonemic awareness, phonics, fluency, vocabulary, and reading comprehension that predict age-appropriate reading development from kindergarten through grade 3 apply to older students? For example, while it is known that phonemic awareness, letter and number naming, and print awareness predict word-reading skills in beginning readers, longitudinal studies indicate that older disabled readers' literacy limitations are better predicted by fluency and naming tasks. Is it the case that adolescents who have difficulties learning to read have mastered a threshold amount of word-reading skills but are unable to negotiate the demands of content-area reading because of limitations in rate, background knowledge, and vocabulary knowledge? Or could it be that they require sustained interventions over time that are not available to them? Third, it is not well understood how adolescents can be motivated to learn to read, particularly when they have endured many years of failure. Thus, there is little information for middle and high school teachers about how best to present reading concepts and have students practice them in the environments typical to classrooms at these levels. Even if it were known which reading abilities are most critical for reading mastery during adolescence, research has not documented the most effective approaches to teaching those abilities.
Welcoming Remarks In welcoming remarks to the workshop participants, Dr. Carol D'Amico, assistant secretary of the Office of Vocational and Adult Education, U.S. Department of Education, described the issue of adolescent literacy as an urgent national challenge. Not only does illiteracy carry a personal burden for individuals but also it threatens national economic security. Dr. D'Amico also expressed the commitment of the current administration to addressing the national problem of low educational achievement and the gaps in achievement between income and ethnic groups. She presented a number of alarming statistics:
Dr. D'Amico pointed out that the lack of educational preparation resurfaces later as a lack of preparation for the next step of an individual's life. Though 65 percent of high school graduates actually make it into college, about a third do not finish the first year. Many of these young adults spend the next 10 years floundering from job to job. They may start to get serious around age 28 or 29, and some try to go back to an adult or a secondary education program because it is indeed a ticket to a good job. The problem is that there will be more jobs than workers in the coming years when the baby boomers exit the labor market. This has implications for the Nation's standard of living and economic security. Dr. D'Amico urged the workshop participants to identify a process, a context, or materials that will make a difference, and a plan to get those into the system.
Profiling the Struggling Adolescent Reader: Dr. Patricia Alexander, professor in the Department of Human Development at the University of Maryland, presented a theoretical model on competencies, based on 6 years of research profiling the adolescent learner. First, she proposed seven guiding premises on which to build theories and a research agenda:
Dr. Alexander's model posits three stages of learning: acclimation, when the individual is beginning in a new domain; competence, which most should acquire between kindergarten and grade 12; and proficiency or expertise. What distinguishes development across time is the changing relationship between knowledge, interest, and strategic processing. During the acclimation stage, for example, knowledge is piecemeal, but becomes broader and deeper in the later stages. Interest in reading can be situational, that is, momentary, established by the time and place, by a teacher who motivates the children at the beginning; on the other hand, it can be individual, that is, what is personal to the individual. A student who knows little about an academic domain is not personally invested in it and depends on others, such as the teacher, to make it interesting. In the competence stage, individual interest carries one forward toward expertise. Strategic processing also develops in stages, with surface- and deep-processing strategies playing more or less important roles. For instance, surface strategies, which involve extracting meaning from text, are more indicative of readers in acclimation. By comparison, highly competent and expert readers rely increasingly on deep-processing strategies that entail the critical and analytical analysis of textual content. Ideally, the three elements of knowledge, interest, and strategic processing enhance one another throughout reading development. Dr. Alexander suggested that longitudinal studies are needed to identify the many combinations of stages and elements that occur and change over time among students. She proposed a model of six possible profiles of adolescent readers.
Discussion Period During the open discussion session, several themes were expanded:
Adolescent Reading Development: Dr. Sally Shaywitz, professor of pediatrics at Yale University School of Medicine, presented a summary of her recent research on brain activity during reading tasks as monitored by functional magnetic resonance imaging (fMRI). She expressed the importance of using strategies and tools to understand reading at fundamental levels as a prerequisite to studying the cognitive factors involved with literacy. Characterizing the basic neurobiologic factors that influence reading and how those factors and behavior change over time will enable researchers to identify behavioral and neurobiologic targets for intervention. A fundamental developmental question is whether reading disability persists in an individual. What is the nature of reading disability in children? In adolescents? What is the role of phonological analysis in adolescent readers? An important tool for research studies is the prospective, longitudinal design. Such studies can assess differential outcomes and how they may be associated with behavioral factors. Dr. Shaywitz presented data from the Connecticut Longitudinal Study, which uses an epidemiologic sample of children followed from kindergarten in 1983 to grade 12 and to age 23 in 2002. The data from the study answered a number of questions.
With fMRI, the brain at work can actually be observed and recorded. It is a tool for understanding what is responsible for phonological difficulties, for understanding the neurobiology of reading and reading disability. During fMRI scanning, stimuli are presented visually; the subject responds by pressing a button. Functional MRI studies of adults revealed that the parietal-temporal region of the brain is activated during reading. Dyslexic readers, however, do not activate the parietal region but rather the front region of the brain. In fMRI studies, 144 children, ages 7-18, showed the same patterns. Furthermore, reading skill levels tested outside the fMRI scanner corresponded to the scan results. Many children with reading disabilities improve in reading accuracy but remain slow readers, lacking automaticity. The scans show that readers with reading disabilities activate the inferior frontal gyrus on both sides; they overactivate the front of the brain. Thus, their accuracy improves because they develop compensatory pathways, but they cannot read fluently or rapidly because the pathway remains disrupted. Dr. Shaywitz hypothesizes that there might be two types of poor readers. In the study, both types did well with familiar words—but one type had a lower ability to read unfamiliar words. The implication is that in reading familiar words well, the latter were not doing phonological analysis but instead were using memory. In nonimpaired readers, the left region was well connected to the parietal and to the frontal left. In the persistently poor reader, it was connected to the right frontal region, which serves memory. The persistently poor readers had lower verbal ability and came from disadvantaged schools. The others had higher verbal ability and came from less disadvantaged schools. Dr. Shaywitz suggests that about one-third of the problem of reading disability is genetic, and two-thirds environmental. If so, this would be a strong argument for early and systematic instruction. Response to Sally Shaywitz by Michael Kamil In his commentary, Dr. Michael Kamil, professor of psychological studies in education at Stanford University, restricted his comments to two basic findings presented by Dr. Shaywitz: (1) that there is a group of persistently poor readers who can be identified early and who continue to have difficulty reading throughout school, and (2) that these readers can be differentiated in adolescence on the basis of phonological measures, but not on comprehension. Dr. Kamil stated that the two observations introduce a dilemma. The report of the National Reading Panel found the effects of phonologic instruction to be limited in both effect and time. That is, phonemic awareness training is most effective in kindergarten and grade 1, but the effects fall off after about 25 total hours of instruction. Phonics is most effective in grade 2 and below. Above these grades, the effect of phonics instruction appears to fall off dramatically. The Shaywitz research puts phonological processing at the base of all reading skills and suggests that persistently poor readers need some sort of phonological remediation, but other research indicates that such instruction has little effect for the adolescent populations under consideration. As an alternative, Dr. Kamil suggested looking at subgroups of readers in a conceptual matrix of four cells: early identified good readers (or statistically "normal") and poor readers as one dimension, and good or poor readers (ultimate success) as the other dimension. In this matrix, some poor readers succeeded; that is, they were identified as poor readers and were either misidentified or overcame whatever problem they had. There are also poor readers who remained so. In addition, some good readers later came to be in the poor group, and finally one group of readers remained good readers all along. The critical groups are those two who switch from good to poor or poor to good, but all four groups must be characterized. A longitudinal study with many interim measures would provide the necessary information. Those measures should include types of instruction that might have made the difference. The most promising of these instructional approaches or interventions should be evaluated in new contexts, with new groups of students. At that point, the data should indicate what is applicable to the problem of persistently poor readers. Dr. Kamil raised two other issues. First, in the Shaywitz study, the readers were not differentiated on comprehension ability, an essential element of reading. This points to the need for the field to be very explicit about what is meant by reading in the context of adolescent reading. What does it include and what does it exclude? Such precision will be crucial for future studies. Second, the Shaywitz studies, and many other similar studies, do not include in their design second-language populations; those students also are of urgent concern and research including or targeting such groups is important. Response to Sally Shaywitz by Hugh Catts In his comments, Dr. Hugh Catts, professor in the Department of Speech-Language-Hearing at the University of Kansas, acknowledged that Dr. Shaywitz's finding on the persistence of reading disabilities is consistent with numerous other studies. This body of research indicates that about 60 percent of children with early reading disabilities have reading deficits at follow-up several years later. This is a high rate of persistence given that another 20 percent of these children may be borderline poor readers at follow-up. Research also confirms that individual differences in phonological awareness are highly predictive of subsequent reading achievement. However, Dr. Catts argued that once children have begun school, the best predictor of reading achievement is reading itself. In a longitudinal study of approximately 600 children, he and his colleagues showed that grade 2 word-recognition abilities accounted for about 80 percent of grade 4 word-recognition abilities. Grade 2 word-recognition abilities also accounted for a large amount of variance in grade 4 reading comprehension. However, grade 2 abilities in listening comprehension explained considerable additional variance in reading comprehension. In other words, as children progress through school and are required to read more difficult text, their ability to comprehend language becomes more critical. Dr. Catts argued that there is also good reason to believe that language abilities play a similar if not greater role in adolescence. As a result, to adequately address reading problems in adolescence, attention should be given to improving language comprehension. A focus on language comprehension might also help explain late-emerging poor readers. Dr. Catts and his colleagues found that about 20 percent of poor readers in grade 4 were not poor readers in grade 2. In other words, these children might be experiencing the "4th grade slump." Further examination showed that these children had good word-decoding skills, both in terms of speed and accuracy, but had poor listening comprehension. Dr. Catts and his group are now following these children in grades 8 and 10. Their work is focusing on factors that might be related to deficits in listening comprehension. Discussion Period During the open discussion period, the following comments were made:
Motivation for Development in Adolescent Literacy The field of reading is a broad mosaic and can be studied from many angles. Highlights of the research literature presented by Dr. John T. Guthrie, professor of human development at the University of Maryland, show that in addition to low achievement, struggling adolescent readers have low motivation for reading, that is, they are disengaged from literacy. The lack of motivation to read is a serious and multifaceted problem among adolescents. The factors involved include not only personal attributes of the students, but also instructional attributes and practices of teachers. In middle school, struggling readers who are unmotivated tend to have the following characteristics: low confidence in their reading ability (low self-efficacy), low confidence in their ability to improve their reading, extrinsically rather than intrinsically motivated (they respond to rewards and incentives), unlikely to read for their own enjoyment or curiosity, feel socially marginalized and disrespected and therefore uncomfortable in school (they want to avoid looking bad and find ways to protect their egos), may nevertheless be interested in other subjects. It is a challenge to engage the struggling reader, but a number of instructional practices can motivate reading. Empirical research substantiates the following observations:
Teachers make the following observations about what facilitates motivation for reading in students:
Dr. Guthrie presented as his opinion that optimization is everything, but research does not indicate how much or what level of autonomy is optimal. It is important to give students choice, but too little choice is deadening and too much is frustrating. He also stated that knowledge about performance is needed: what the teachers and students are doing. Dr. Guthrie identified a number of barriers to reading motivation based on a mixture of practitioner observations and research:
Research studies and practitioner observations on reading motivation indicate the need for studies based on an engagement hypothesis that links cognitive strategies with motivation. Variables that would be included are subject matter knowledge, questioning, graphic organizing, summarizing, comprehension monitoring, searching, intrinsic motivation, integrated learning goals, self-efficacy, performance-approach goals, and social goals. Dr. Guthrie suggested that, in experimental study designs, these variables be taken into account in developmental and descriptive studies of cognitive and motivational strategies. Each strategy should be studied as a dependent variable, and developmental benchmarks for each strategy are needed. Response to John Guthrie by Donna Alvermann Although concurring with Dr. Guthrie that experimental studies are needed, Dr. Donna Alvermann, UGA Distinguished Research Professor in the Department of Reading Education at the University of Georgia, emphasized that experimental studies alone are not sufficient. Understanding the data produced by such studies requires doing qualitative research as well (e.g., case studies). Qualitative research enables one to look in depth at the various contexts in which literacy occurs naturally outside of formal school learning. Such research is essential because outside the school setting, children often exhibit a number of literacies that extend and enrich the more narrowly based print literacy associated with school learning. While not in the formal learning environment, people are still in situations requiring literacy. In those situations, where they do not have to acquire knowledge, their levels and sources of motivation for reading may be very different from those in the classroom. Dr. Alvermann said that it is important to think of children and adolescents as they actually are today. Her own research studies of children after school, in such settings as the library, showed that these students of the Net Generation are quick to find Internet sites and understand complex materials; for example, they make predictions and inferences from Japanese cartoons. These children, who scored in the lowest 25 percentile on a standardized reading test, can read some materials when they are motivated to do so. They often use strategies that differ from those introduced in formal school settings. She recommended that to reach resistant or reluctant readers, including English-language learners, teachers incorporate students' interests from out-of-school settings. Response to John Guthrie by Donna M. Ogle Dr. Donna M. Ogle, professor in the Reading and Language Development Department of National-Louis University, commented that it is important to disaggregate data to understand what the needs of different groups and subgroups are and what more must be made clear with research. There are two important sources of data that can be analyzed—the 1998 National Assessment of Education Progress Survey and the 2000 Program for International Student Assessment (PISA). Disaggregated data give different profiles of students than do aggregated data. For example, PISA data show that in Korea students are not interested in reading but do nevertheless score high in reading, and they are high achievers in mathematics and science. Aggregated data rank Germany quite low, but disaggregated data show an effect of recent immigrants educated in other countries. Immigrants beyond the first generation score higher, as a result of schooling in Germany. Finland, in contrast, has a homogeneous culture and is at the top of the list. Dr. Ogle suggested disaggregating data by several factors, including culture, first language, family groups, and social expectations. Response to John Guthrie by Douglas Buehl Douglas Buehl, a reading specialist at Madison East High School in Madison, Wisconsin, commenting from the point of view of the practitioner, noted that self-identity relates to motivation. He said that it has always been socially acceptable for adolescents to admit they struggle with math and science (because these are considered difficult subjects) or with history (because it is considered boring). However, it is not acceptable for students to admit that they do not read well. The student who does not read well may not be motivated to try. As a result, many struggling adolescent readers lack sufficient practice in reading because they avoid reading whenever possible. Therefore, teachers must help students understand why they need to learn something. One approach would be to find the bridge between other subject areas and literacy activities—for example, to teach students to read and communicate about science—yet there is typically a mismatch between the reading ability of the struggling reader and the textbooks used for the subject area. Discussion Period During the open discussion period, the following comments were made.
Critical Factors in Instruction-What Should We Be Teaching? Dr. Donald D. Deshler, professor of special education and director of the Center for Research on Learning of the University of Kansas, began by describing four major challenges surrounding literacy instruction for underachieving adolescents:
Dr. Deshler recommended reframing the question to be more inclusive than merely "what" we should be teaching. In addition to the question of "what," researchers must consider (1) the effects of the complex context of secondary schools, the effects of adolescent development, students' learning histories, and roles of secondary teachers, (2) how to teach skills, strategies, and content to adolescents with literacy problems, and (3) what constitutes effective instructional environments for promoting adolescent literacy in high school settings. Regarding what to teach, Dr. Deshler presented a framework that researchers at the University of Kansas Center for Research on Learning have used for conceptualizing adolescent literacy instruction. This framework, the Content Literacy Continuum, considers both student needs and the realities of secondary schools. Additionally, the five levels of this continuum acknowledge the different views regarding content literacy and adolescent literacy.
Content Literacy Continuum Dr. Deshler delineated three major domains of instructional variables that have been found to characterize effective instruction and should be the defining attributes of how-to skills, strategies, and content to promote adolescent literacy:
Effective instructional environments for promoting adolescent literacy consist of at least three factors: (1) clarifying the roles and expectations of teachers, (2) supporting collaboration among general education teachers and support staff and across teachers, and (3) creating an environment that strongly values literacy and strategic approaches to learning. Dr. Deshler asserted that any research agenda on adolescent literacy should focus on at least four outcome measures in the process of validating instructional interventions: (1) statistical significance, (2) social significance, (3) palatability of the intervention for practitioners, and (4) scalability of the intervention. The following are the pressing research needs he presented.
Response to Don Deshler by Sharon Vaughn Dr. Sharon Vaughn, Mollie V. Davis Professor and Director of the Texas Center for Reading of the University of Texas, presented insights gained from two studies and from working with students. In helping students who are behind, the center provides three types of services: tutoring to get them through subjects, providing them with strategies as tools to enhance their access, and providing instruction in basic reading skills. In one of the studies conducted last year, the researchers worked with 20 middle school students whose reading was below the grade 3 level and who had been identified as learning disabled. At the beginning, the students seemed disengaged and not interested in reading, but later their drive to learn to read became apparent. Dr. Vaughn noted that they would do anything to learn as long as no one else knew—confidentiality was prime. It would be too humiliating to have others know. They actually craved learning to read and perceived it as essential to their future success. They could not identify a single book, poem, or article that they enjoyed reading or having read to them. Dr. Vaughn was impressed with the vulnerability of these students over time, and with how willing they were to tackle the job of learning to read. In the second study, the researchers worked with 22 high school students below grade 3 reading level who had been identified as learning disabled. An intensive intervention program was conducted—one group learned to spell and learned words primarily through writing, one group learned to spell and learned words primarily through tracing, one group was a control group. There were 23 sessions of 25 minutes duration each. With the intervention, the writing and reading groups made noticeable gains in reading words compared to the control group. Response to Don Deshler by Cheryl Scott Dr. Cheryl Scott, a visiting professor in communication sciences and disorders at Northwestern University, emphasized the need to design interventions that take into account individual differences. Adolescents fail as readers for different reasons and for different combinations of reasons. Readers must attend, read accurately and fluently, parse sentences, construct propositional text, construct mental models, generate inferences, and monitor comprehension. Syntax of sentences is Dr. Scott's specialty area and she noted that the processing of written sentential syntax is a challenge distinct from listening because of a variety of structural differences between the two modalities. Two particular features in written sentences were highlighted:
Using a case example, Dr. Scott illustrated that an adolescent reader might have a distinctive problem parsing and thus understanding these types of written sentences. She noted that equally distinctive types of problems could occur for other components of the reading process, and that a "one size fits all" approach to intervention is unlikely to succeed. The development of true clinical markers for specific difficulties with component parts of the reading process is a necessary prerequisite for designing meaningful interventions. Response to Don Deshler by Annemarie Palincsar Dr. Annemarie Palincsar, Jean and Charles Walgreen Professor of Reading and Literacy at the University of Michigan, recommended reframing the research question regarding instruction. Instead of "What constitutes effective environments for literacy instruction?" ask "What constitutes effective instructional environments for literacy learning?" and "What constitutes effective instructional environments for promoting deep understanding of subject matter and literacy knowledge and skills?" She based her rewordings on the assumption that reading, writing, viewing, and oral language are all integral to the process of developing deep understanding of subject matter. Dr. Palincsar urged that future research on adolescent literacy include the study of adolescent literacy learning in enriched learning environments, and suggested that such environments have the following characteristics:
She reported on research conducted as part of the Research Institute to Accelerate Content Learning through High Support for Students with Disabilities in Grades 4-8 (REACH). The research began with the preparation of case studies of identified students in inclusion classrooms engaged in inquiry-based science instruction. The classroom teachers then read and interpreted the case studies for the purpose of identifying ways of addressing the print literacy, cognitive, social/relational, and attention challenges experienced by the students. After identifying appropriate interventions, a 2-year study revealed that all students—typically achieving, low achieving, and identified—profited from the interventions, as measured by assessments of conceptual understanding and scientific reasoning. Discussion Period During the open discussion, the following comments were made.
Designs to Study Adolescent Readers Dr. David J. Francis, professor of psychology and director of the Texas Institute for Measurement, Evaluation, and Statistics at the University of Houston, presented guidelines and principles of research design to keep in mind when designing a research agenda and developing a grant proposal. Several overarching considerations are that it is helpful to have a view that is expanded over the traditional view of conditions of observation—think about who are the subjects, how they get selected, and how they get assigned to conditions. Studying adolescents does not preclude any designs, but interest in certain phenomena might limit the number of useful designs. Scientific questions should drive all the aspects of design. Several criteria apply to a good study design. First, the design should have internal validity—can cause and effect be inferred, and to what extent does the study reduce uncertainty about cause and effect? When feasible, randomization is preferred; although it does not guarantee strong internal validity, it does eliminate common threats. For descriptive studies, qualitative study design might be the best approach. Second, external validity is the extent to which one can generalize beyond the specific situation of the study—to other settings, at other times in history, at different times in the school year, in a different context. Dr. Francis warned that it is difficult to maximize internal and external validity in the same study—this is a constant trade-off requiring careful decisions regarding design, which must be made with the research aims in mind. Early studies might try to identify common characteristics, and later studies would push for internal validity. Third, statistical conclusion validity refers to whether or not the evidence implies a true relationship (not necessarily a cause-and-effect relationship) between independent and dependent variables. Have the data been analyzed correctly? It is important to understand the nature and limits of the statistical methods used. Good design is a challenge of making difficult choices. The following are a number of questions Dr. Francis presented to consider in designing literacy studies. The specific answers to any of these questions will depend on many factors, but the answers should be driven by the aims of the research:
|
Institute Home|
National Institute for Literacy | National Institute of Child Health and Human Development | US Department of Education
|
Home
|
About Us
|
Staff
|
Employment
|
Contact Us
|
Questions
|
Site Map
Last updated: Friday, 22-Sep-2006 09:51:18 EDT |