Helping Children with Reading Difficulties in Grades 1 to 3
As indicated in Chapters 6 and 7, many children learn to read with good instruction, but some do not. And many children have problems learning to read because of poor instruction. In all cases, the question is what kinds of additional instruction (usually called "interventions" because they are not part of the regular school reading instruction) are likely to help.
The purpose of providing extra instructional time is to help children achieve levels of literacy that will enable them to be successful through their school careers and beyond. It is not simply to boost early literacy achievement. Given the focus of this volume, we restrict our discussion to the primary grades; however, it is likely that children who have had interventions in the primary grades will need additional supplementary experiences in the upper grades as well. We know that the literacy demands are of a different nature for older children; as children proceed through the grades, they are expected to learn from informational text with which they may have had few experiences in the primary grades (see Fisher and Hiebert, 1990); they are expected to use text independently; and they are expected to use text for the purpose of thinking and reasoning.
Long-term follow-up studies reveal that even very powerful early interventions often require "booster" sessions.
We begin by discussing some interventions that are specific to reading, targeting the training of phonological skills. We then proceed to discuss individual tutoring and supplementary small-group efforts provided by professionals with specialties in reading that have been designed to provide comprehensive supplementary literacy instruction. We continue with information on computer support for reading instruction, retention in grade, and special education for children with learning disabilities. Although the latter two are not specific to reading, they have often been introduced in response to reading failure. The chapter ends with a brief mention of some controversial therapies for reading problems.
TRAINING IN PHONOLOGICAL AWARENESS
Phonological awareness, the appreciation of speech sounds without regard for their meaning, is critical to discovering the alphabetic principle (the idea that letters generally represent the small speech segments called phonemes). The theoretical importance and strong empirical relationship of phonological awareness to success in learning to read was discussed in Chapters 2 and 4, and the demonstrated benefits of phonological awareness training for children who have not yet begun formal reading instruction were reviewed in Chapters 5 and 6. Here we examine evidence of the effectiveness of such training for two groups of children: beginning students at risk for reading difficulties and schoolchildren with existing reading difficulties (whose achievement is unacceptably low after two or more years of instruction).
Phonological Awareness Training for Kindergartners at Risk
Many children at risk for reading difficulties enter school with little or no phonological awareness. Does explicit instruction and practice in attending to and manipulating the sounds within spoken words facilitate these children's reading acquisition? Evidence is accruing that indeed such training can be of particular benefit to
youngsters at risk due to socioeconomic disadvantage and/or weak initial preparedness in reading-related skills.
As was the case in the kindergarten research reviewed in Chapter 6, some training and intervention programs for at-risk youngsters have emphasized phonological awareness exclusively (Bentin and Leshem, 1993; Hurford et al., 1994), whereas others have combined phonological awareness activities with instruction in letter identification and letter-sound correspondences (e.g., Ball and Blachman, 1991; Felton, 1993; Smith et al., 1993; Torgesen et al., 1992, 1997). It is therefore important to point out that even those with the more narrowly focused programs have observed gains in reading skills (word recognition), as well as in phonological awareness itself, relative to control groups. This suggests that the effectiveness of the more broadly focused studies does not rest solely on the inclusion of early reading instruction, but also benefits from lessons that draw the child's attention to the sounds within spoken words.
How effectively has phonological awareness training (alone) benefited word identification? In a sample of 431 children who had not yet received formal reading instruction, 99 had been designated as at risk on the basis of a screening battery (Hurford et al., 1994). Half of the at-risk group received individual tutoring in phonological awareness for a total of about 10 to 15 hours over a 20-week period, during which time regular classroom reading instruction also commenced for all participants. Prior to training, there was a substantial difference (13 to 14 points) between mean standard scores of the not-at-risk children and each at-risk group on the word identification measure. After the training period, this large gap remained for the untrained at-risk group, but the trained group's post-test mean was 7 points below that of the controls who were not at risk.
Another study compared the effects of phonological awareness training with an alternative kind of language training (in vocabulary and sentence skills) as well as with a no-training condition for children at risk on the basis of their initial skill levels (Bentin and Leshem, 1993). Compared with the performance of not-at-risk classmates, the at-risk groups who received no training or alternative language training scored about 40 points lower on two post-tests. In contrast, those who had received training in phonological segmentation scored
30 to 34 points higher, on average, than the other at-risk groups, and within 10 or fewer points of the not-at-risk group's mean.
In the studies in which training has also included instruction in letters and letter-sound relationships, similar patterns of results have generally been found (e.g., Ball and Blachman, 1991; Felton, 1993; Smith et al., 1993; Torgesen et al., 1992, 1997). Modification of the standard Reading Recovery program (described in a later section) so as to include an additional phonologically oriented component has also been shown to be effective; when researchers compared a group of at-risk first graders who participated in the standard program with a matched group in the modified Reading Recovery training, the latter group reached criterion for successful completion significantly faster (Iverson and Tunmer, 1993).
Torgesen and his colleagues (1992, 1997) have also explored the question as to what degree of explicitness in such instruction is most effective for kindergartners with weak letter knowledge and phonological awareness skills when they begin school. At-risk kindergartners were assigned to one of four conditions: a highly explicit and intensive phonologically oriented instruction; a less explicit phonologically oriented instruction delivered in the context of meaningful experiences with reading and writing text; regular classroom support; or no treatment. The group receiving explicit phonologically oriented instruction scored highest on word identification, but only the 12-point difference with the no-treatment group was statistically significant. A similar pattern of means favoring the explicit phonologically oriented instruction group was obtained for reading comprehension, but these smaller group differences were not significant. These data are consistent with those of Foorman et al. (1998, discussed in Chapter 6), suggesting that greater intensity and explicitness of early phonological training may reap greater gains in reading acquisition for at-risk youngsters.
One reason that statistical significance has sometimes been difficult to achieve in these training and intervention studies (with their relatively small sample sizes) has been the considerable variability among children within groups in their responses to treatment. It is clear that a majority of at-risk children who receive training in phonological awareness show strong gains in awareness itself, but a
minorityperhaps a quarter (Torgesen et al., 1997)gain little or no insight into the structure of spoken words, much less into reading, by the end of training. Typically, these children are among the very weakest at the outset in their phonological awareness (and other linguistic) abilities. For these children to reap the benefits of training, it is likely that many more hours of or a different type of special instruction are needed than have typically been provided in studies to date.
The fact that the effects of phonological awareness training have not been found to include gains in reading comprehension in the early grades is not particularly surprising. As discussed previously, reading comprehension depends not just on mastery of word recognition skills but also on a host of other factors, including vocabulary, background knowledge, memory skills, and so forth. Children assigned to the at-risk groups have typically been weaker than classmates in their overall cognitive and linguistic preparedness, and training in phonological awareness is not designed to strengthen other skills that contribute to comprehension. In short, the goal of phonological training is limited to facilitating the acquisition of word-decoding abilities, which are necessary but not sufficient for the development of skilled comprehension.
Taken together, these studies indicate that training in phonological awareness, particularly in association with instruction in letters and letter-sound relationships, makes a contribution to assisting at-risk children in learning to read. The effects of training, although quite consistent, are only moderate in strength, and have so far not been shown to extend to comprehension. Typically, a majority of the trained children narrow the gap between themselves and initially more advanced students in phonological awareness and word reading skills, but few are brought completely up to speed through training, and a few fail to show any gains at all. Hence, it is unrealistic to think of phonological awareness training as a one-shot inoculation against reading difficulties for children at risk. Rather, its greater demonstrated value is as the first of many aggressive steps that can be taken in an ongoing effort to intensify all facets of reading instruction for schoolchildren who need it.
Training for Children with Reading Disabilities
Because most children who are identified as being poor readers are also weaker than their classmates in phonological awareness skills, providing training in awareness has been thought to be helpful for ameliorating these children's reading difficulties. To date, several studies have examined the efficacy of this approach to remediation, with somewhat mixed results.
One of the earliest studies of phonological awareness training for disabled readers focused on phoneme analysis, blending, and phonological decoding of text for students ages 7 to 12 with serious reading difficulties (Williams, 1980). Compared with similarly low-achieving children who did not receive training, the trained group earned significantly higher scores on several measures of phoneme awareness, reading of nonsense words, and reading of regular three-letter words that had not been used in the training materials.
A computer-based training program provided supplemental small-group phonological instruction for children in grades 2 to 5 who were in the bottom 10 percent in word recognition skills (Wise and Olson, 1995; Olson et al., 1997). One group was first trained in phoneme awareness and phonological training and then progressed to reading stories on the computer. The comparison group's training focused on comprehension strategies, beginning with small-group instruction and then reading stories on the computer. In all, the comparison group spent more than twice as much time reading stories as the other group.
In contrast to untrained control groups in previous research that have consistently shown no improvement, both groups made gains in word recognition over the training period. As would be expected, the first group improved significantly more than the comparison group in phoneme awareness and phonological decoding of pseudo words, and these differences were maintained for a year beyond training. The comparison group, which spent much more time actually reading on the computer, scored higher on speeded word recognition. These findings suggested that the phonologically trained group's better decoding skills worked to their advantage when they had unlimited time to apply them, but that they were not yet suffi-
ciently automatized in decoding to do well under time-limited conditions. Although it was hypothesized that with further reading experience the phonologically trained group's word recognition skills would become more automatized, there was no evidence for this (Olson et al., 1997).
Another study also demonstrated that word recognition skills of severely disabled readers can be substantially improved through intensive supplementary training (Lovett et al., 1994). Training was given to each of three groups in explicit instruction in phoneme awareness and letter-sound mappings, training in using common orthographic patterns and analogies to identify unfamiliar words. and study skills training (the control condition). The phonological program produced greater improvement in phoneme awareness and phonological decoding, but the two trained groups showed similar gains in word recognition compared with the controls. Recent analyses on an expanded sample indicate that the two training conditions are about equally effective for older (grades 5 and 6) and younger (grades 2 and 3) children with reading disabilities (Lovett and Steinbach, 1997).
A final study compared immediate and long-term outcomes for groups of children with severe reading disabilities who had received one of four types of training: phonological awareness training alone; reading instruction alone, based on the Reading Recovery approach, but with no coverage at all of letter-sound relationships; training in both phonological and reading skills in combination; and no treatment controls (Hatcher et al., 1994). Performance of the group that received the combined training consistently exceeded that of the control group on both immediate and delayed post-tests, but scores of the other trained groups did not differ significantly from those of the controls. This pattern of results was seen for word recognition, nonword reading, text reading accuracy, and reading comprehension. Immediately after training, the combined training group was six months ahead of the control group, on average, in both accuracy and comprehension of text reading; nine months after the cessation of training they remained four months ahead in accuracy and eight months ahead in comprehension. Despite these considerable gains
relative to the progress of other severely disabled readers, their reading levels remained one to two years below age norms.
These studies indicate, first, that intensive training, even over relatively short periods of time, can substantially improve the word-reading skills of children with serious reading disabilities and that these positive outcomes are maintained over months or years after the cessation of training. Whether a continuation of such training over longer periods would lead to a fuller remediation of these children's difficulties remains unknown, however. In particular, fluency and automaticity of word recognition, which may be required for skilled reading comprehension, may require much more or different types of training and extensive practice.
Second, it is clear that phonologically oriented training programs are not the only type of intervention that can facilitate word recognition, although this approach produces the strongest gain in phonemic awareness and phonological decoding when combined with training in other reading skills. Other, more orthographically oriented approaches have been of equivalent benefit for improving word reading in this population, many of whom have already acquired some decoding skills (although these may be minimal) before training. Finally, although most children with reading disabilities are characteristically deficient in phonological abilities (both oral and written), they may also have, in part due to limited print exposure, deficits in oral vocabulary, language comprehension, and background knowledge (Stanovich and West, 1989). Dealing with these problems is clearly beyond the scope and aims of the training programs we have reviewed in this section.
In this section, we describe supplementary interventions that take the form of tutoring. They were selected for review because they have received more sustained research attention than other tutoring programs. Like the training studies in phonological awareness reviewed above, they approach the provision of extra time in reading instruction by tutoring children individually.
Reading Recovery, which is singled out for a relatively extensive review, has garnered significant attention in the United States. It requires extensive training of teachers, as well as intensive one-on-one instruction with children, rendering it quite costly. The program was designed by Marie Clay for the purpose of intervening with young children in New Zealand identified as having reading problems. For complete descriptions of the instructional program, the reader is referred to Clay (1985) and Pinnell et al. (1988).
The program has a particular framework for providing instruction to the tutees. For the initial 10 days of a child's participation in Reading Recovery, the teacher gathers information about the child's current literacy strategies and knowledge. Following this period, referred to as ''roaming the known," each lesson includes (a) engaging the child in rereadings of previously read books; (b) independent reading of the book introduced during the previous lesson (during which the teacher takes a running record to assess fluency); (c) letter identification exercises, if necessary; (d) writing and reading his or her own sentences, during which the child's attention is called to hearing the sounds in words; (e) reassembling the child's sentence which is not cut up into individual words; (f) introduction to a new book; and (g) supported reading of the new book. These activities occur in a 30-minute block of time on a daily basis. One feature of Reading Recovery is time on reading of familiar bookssheer on-task, engaged learning time for students.
Teacher support provided during each of these activities is designed to enhance what are referred to as children's self-extending systems; that is, children are encouraged to use multiple sources of information while reading and writing and to engage in literacy activity using a problem-solving approach, monitoring for the effectiveness with which they are making sense of the text. The short books used by the children have been sequenced on the basis of teacher judgment of difficulty.
Once the child has achieved the level of functioning that matches (within a .5 standard deviation) the competence demonstrated by a
randomly selected group of first graders drawn from the child's school on the same tasks, the child is discontinued from the program. Typically, this translates into 60 30-minute sessions over a 12- to 16-week period. Typically, teachers conduct Reading Recovery lessons with four children a day and spend the remainder of their day as first-grade teachers. During the course of a school year, about 8 to 11 children per Reading Recovery teacher generally complete the program successfully and another 27 percent of children are dismissed from the program without having successfully reached criterion performance.
By most professional development standards, the preparation of Reading Recovery teachers is quite extensive. Following 30 hours of training before the beginning of the school year, Reading Recovery teachers participate in weekly sessions in which the central activity is the observation and discussion of two lessons that are conducted by Reading Recovery teachers (working behind one-way viewing windows) with one of their students. The observations are guided by a teacher-leader, who focuses the group's attention on the activity of both the teacher and the child.
There are now a number of publications asking the question, "Does Reading Recovery work?" These include publications by the implementers of Reading Recovery in the United States, including DeFord et al. (1987), Pinnell et al. (1994), and Pinnell et al. (1995). In addition, a number of thoughtful syntheses and reviews have been reported by others, including Center et al. (1995), who also report an empirical study of their own using Reading Recovery, Hiebert (1994a), Rasinski (1995), and Shanahan and Barr (1995). In fact, it appears that the data available through these reviews exceed the data available through firsthand published investigations of Reading Recovery; that is, the reviewers have included in their syntheses technical reports and unpublished documents that have not been disseminated by the Reading Recovery organization.
Clay's own research regarding Reading Recovery in New Zealand (Clay, 1985) has been criticized, in particular by Nicholson (1989) and Robinson (1989). These authors point out that, although Clay provides clear evidence that children improve on measures that she has designed, there is no evaluation for transfer to
other reading measures. Perhaps more troubling is their finding that the results reported by Reading Recovery are only for children who have successfully been discontinued from the program, excluding about 30 percent of the participants. Because children are not randomly assigned to Reading Recovery or an appropriate control group, the question is raised whether the growth demonstrated in Reading Recovery might not be explained simply in terms of normal development. Finally, maintenance measures comparing the performance of students successfully graduated from Reading Recovery with other low-progress students who did not receive Reading Recovery tutoring indicate that 12 months after the intervention there are very small differences between the reading achievement of Reading Recovery children and the other low-progress children (Glynn et al., 1992). This finding regarding the failure of the low-progress children to respond to Reading Recovery in the long run was replicated in a reanalysis of Pinnell et al.'s (1988) data on U.S. participants in Reading Recovery, once again indicating that 30 percent of the original sample of low-progress children who were enrolled in Reading Recovery failed to benefit from the program (Center et al., 1995). Similar analyses and conclusions have been presented by Hiebert (1994a) and Shanahan and Barr (1995).
In a study of Reading Recovery conducted by Pinnell et al. (1994), including random assignment of participants to one of five groupsReading Recovery, three other early intervention programs (differing from one another in group size, amount of teacher training, and whether or not they adhered to Reading Recovery instructional plans), and a control groupthe results indicated that following 70 days of program intervention the students in the Reading Recovery clearly outperformed the students in the other three intervention programs on an array of measures of reading achievement. The study being described here contained high amounts of familiar book reading time for the reading recovery group and for one additional intervention group compared to much less time for the other groups. The group that equaled Reading Recovery method in time spent reading familiar books equaled Reading Recovery in outcome data. However, after three months, post-tests using standardized measures did not reveal any statistical differences among the treat-
ment groups, although the Reading Recovery group continued to maintain its gains12 months lateron those measures that are specific to Reading Recovery (Clay's concepts of print and dictation tasks).
In their own research investigating Reading Recovery, Center et al. (1995) included an analysis of the individual cases of three groups of students participating in Reading Recovery and reported an important finding. They divided their Reading Recovery instructional groups into children who were totally "recovered" versus those who were unsuccessful and examined the profiles of these children in terms of their pretest measures. They reported that the recovered group was markedly superior to the unrecovered group in terms of their pretest metalinguistic knowledge, as determined by assessment of phonemic awareness, word attack, and cloze comprehension (that is, a method of systematically deleting words from a prose selection and then evaluating the success a reader has in accurately supplying the words deletedMcKenna, 1980). Center et al. conclude that children with poor metalinguistic knowledge are less likely to be successful in Reading Recovery. This hypothesis received support from the instructional research of Iverson and Tunmer (1993), who conducted a study including a condition in which they modified Reading Recovery to include explicit code instruction involving phonograms (common elements in word families, such as the letter sequence, "at" in "bat, cat, sat"). Children who were assigned to the modified condition achieved criterion performance more quickly than children in the standard condition.
Despite the controversies regarding the efficacy of Reading Recovery, a number of intervention programs owe their design features to it, and it offers two important lessons. First, the program demonstrates that, in order to approach reading instruction with a deep and principled understanding of the reading process and its implications for instruction, teachers need opportunities for sustained professional development. Second, it is nothing short of foolhardy to make enormous investments in remedial instruction and then return children to classroom instruction that will not serve to maintain the gains they made in the remedial program.
Book Buddies is a supplementary intervention in which selected children received one-on-one tutorials twice a week in addition to classroom reading instruction, using highly qualified community volunteers as tutors (Invernizzi et al., 1997). These tutors received continuous on-site training and supervision in the delivery of a four-step lesson designed by reading specialists. The four-part plan consists of repeated reading of familiar text to enhance reading fluency, word study (phonics), writing for sounds, and reading a new book. The word study portion of Book Buddies lessons is derived principally from research on developmental spelling; hence instruction initially focuses on beginning consonants, proceeds to beginning and ending consonants, and finally goes to full phoneme representation of consonant-vowel-consonant words, at which point the child has stable speech-to-print concepts and the beginnings of a sight vocabulary.
Although not all Book Buddy children start at the same point, the basic program proceeds through alliteration in whole words to onset-rime segments to individual phonemes. Children are explicitly taught basic letter-sound correspondences and how to segment and manipulate beginning consonants in the onset position of simple words. As they achieve a stable concept of words and begin to acquire a sight word vocabulary, they are encouraged to segment and manipulate the rime unit. Finally, when the corpus of known words is larger and the child begins to read, medial short vowel sounds are examined. The use of known words, gathered from context and then analyzed in isolation (for instance, with the use of word bank cards), provides an opportunity to transfer phonological awareness training and grapheme-phoneme practice from text to automatic reading of sight words.
The third component is writing for sounds. Children are allowed to write in invented spelling, but they are held accountable for those phonics features already taught. The rationale for this activity is that the act of segmenting speech and matching letters to sounds is a rigorous exercise in phonemic awareness. Furthermore, there is
substantial research demonstrating that invented spelling can enhance children's memory for words, at least in the beginning stages (Ehri and Wilce, 1987).
The fourth component of each lesson is the introduction of a new book, which includes focusing the child's attention on the sequence of events and assessing the child's related background knowledge. Finally, reading comprehension is fostered throughout the reading of the new book through predictions, discussions, and opportunities to write about the new story.
In summary, this supplementary intervention has four driving principles: children learn to read by reading in meaningful contexts; reading instruction should be differentiated based on the diagnosis of learner need; phonics instruction should be systematic and paced according to a child's developing hypotheses about how words work; and reading, writing, and spelling develop in synchrony as children interact with others who assist their learning and development.
Evaluations of Book Buddies included three cohorts of 358 first and second graders. The first graders were in the bottom quartile of each school's Title I referral list. There were 15 tutors, each of whom was supervised by a university faculty member who made assessments and wrote lesson plans. The cost was estimated to be one-sixth that of Reading Recovery. The effect size was 1.29 for word recognition, which is considerably higher than effect sizes reported for other tutorial programs and is indeed comparable to that found with professionally trained teachers. However, it is important to note that the tutors were carefully prepared, were supervised on a daily basis, and were provided guidance, feedback, and support.
Reading One-One uses trained and managed paraprofessionals (college students, community residents, teacher aides) to deliver three to five one-on-one tutoring sessions to low-performing readers on a weekly basis throughout the school year (Farkas and Vicknair, 1996). The program aims to serve the lowest-performing readers in elementary school grades 1 to 6, including children with limited English proficiency and children living in poverty. Teachers recommend
students who "need help the most" or "could most profit from help."
Prospective tutors are tested on their English-language skills and interviewed. If they pass this stage, they are trained and tested again on the tutor manual used in the program. Each school has an on-site coordinator. Expert staff sit with each tutor during actual tutoring sessions and fill out an appraisal form, which is then used to provide feedback to the tutor. This continues until the tutor has met the program's standard, at which point they are certified.
The curriculum combines explicit instruction on decoding skills with the use of small books that are ranked by difficulty level (see descriptions of these types of books in Chapter 6). They include fiction and nonfiction and range in level from emergent literacy through fluency. After assessment, each child is placed into one of three curricula: alphabet, word-family, or reading-ready. The first of these is for children who are still learning their letters and sounds, the second is for children still learning the most basic decoding skills, and the third is for children who are able to read at least the easiest-level books on their own. Each tutoring session allows for about 30 minutes of instruction.
For all three curricula, the session involves both book reading and explicit instruction on skills related to reading. This is organized as follows:
· for alphabet studentsreview of previous letters/sounds, new letter/sound instruction, reading (reading to the student and/or assisted reading), assisted creative writing;
· for word-family studentsreview of previous word families, new word family instruction, reading (reading to the student and/or assisted reading), and creative writing; and
· for reading-ready studentsrereading, new reading, high-frequency words practice, and creative writing.
In general, children are assessed every fifth session. For reading-ready children, a running record is taken every session. The child also creates and maintains letter and/or word banks, a copy of which goes home with them.
An evaluation of the Reading One-One program examined the amount of improvement in relation to the number of sessions of tutoring received, which varied unsystematically as a result of varying logistical circumstances (Farkas and Vicknair, 1996). Another evaluation showed that 70 Reading One-One sessions (taking about four to six months) typically raised a child's grade-equivalent score by about half a year (Farkas and Vicknair, 1996).
COMPREHENSIVE LITERACY-ORIENTED EFFORTS WITH SMALL GROUPS OF CHILDREN
Early Intervention in Reading
Early Intervention in Reading is an intervention that took place in regular first-grade classrooms and was directed at improving the reading achievement of the lowest-performing five to seven readers in each class (Taylor et al., 1994). This research was conducted over a four-year period in diverse school settings (rural to inner city). Selection for the intervention was made by identifying the children with the lowest scores on tasks that require them to produce individual sounds in words and to blend sounds together to form words. The lessons were planned in three-day cycles and began with the reading of a picture book to the whole first-grade class. The teacher engaged the class in a retelling of the story, which was printed on large chart paper so that children could read the retelling together over the three days. Also included in the instructional cycle was a writing activity in which the teacher selected three short phonetically regular words, which the children were asked to write. Also, the children engaged in assisted sentence writing about the story. A final component of each instructional cycle was individual reading by each child, using either the retelling or an appropriate book. Throughout the course of the school year, longer trade books were introduced. Although the majority of children participating in this early intervention program were indeed reading by the end of first grade, only one-third to one-half of them were reading at grade level.
Restructured Chapter I
In Chapter 7 we discussed congressional efforts to help disadvantaged children through Title I of the Elementary and Secondary Education Act of 1965, which was reauthorized as Chapter I. Here we have selected an intervention to restructure Chapter I services to illustrate an alternative model for using these resources (Hiebert et al., 1992). The restructuring began by reducing the numbers of students the teacher instructed at any one time from eight to three. This was made possible through the use of teacher aides who were funded by Chapter I. The second step was to work closely with the Chapter I teachers in the design of curriculum and instruction that would enable first graders to achieve grade-appropriate reading skills. Toward this end, there were three activities around which each lesson was organized: reading, writing, and word study (phonemic awareness).
To ensure that the children were engaged in sufficient reading of text, repeated reading of predictable text was selected as the primary oral reading activity, during which children were taught to track the print as they read aloud. Selected books were also brought home, and parents were asked to verify that their children had read at home. The reading of 10 books at home resulted in the award of a trade book. Writing activity included maintaining a personal journal (during which children received guidance in the use of phonetically plausible invented spellings) and constructing sentences around word patterns to which the children had been exposed in the reading activity. Finally, the word study (phonemic awareness) portion of each lesson consisted of two activities: one, designed to heighten awareness of phonemic identity, engaged the child in selecting rhyming words from among a list read aloud, and the other, designed to heighten awareness of phonemic segmentation, called for the child to listen carefully as a word was pronounced with elongated sounds and to move a chip as she or he heard each new sound in the word.
The effectiveness of the restructured program was evaluated in multiple ways. First, the participating students' achievement was compared with an absolute level of achievement (proficient grade-
level reading). Second, end-of-year achievement of the participating Chapter I students was compared with nonparticipating students (from district programs that were not enrolled in the restructuring effort.) Finally, the researchers compared the end-of-year performance of the participating Chapter I students with nonidentified classmates who had begun the year with higher performance on reading assessments. All three forms of assessment revealed significant differences in favor of participation in the restructured program, in terms of primer-level fluency, first-grade text fluency, and performance on a standardized reading assessment.
COMPUTER SUPPORT FOR READING INSTRUCTION
Recent advances in computer technology offer new support for reading instruction. Digitized and high-quality synthetic speech has been incorporated into programs focusing on phonological awareness and issues related to emergent literacy, letter-name and letter-sound knowledge, phonological decoding, spelling, and support for word decoding and comprehension while reading and writing stories. Computer speech, along with interesting graphics, animation, and speech recording, has supported the development of programs that are entertaining and motivating for both prereaders and beginning readers.
Talking books, widely distributed on CD-ROM, are among the most popular programs that claim to improve children's reading. Book pages are presented on the computer screen, and children can select the whole text or specific words and phrases to be read aloud by the computer. The most popular books include many clever animations that are highly entertaining to children, perhaps so much so that they distract from the task of reading; children can often access the animations without paying any attention to the print.
Storybook software displays storybooks on the screen. The programs come not only with software but also with ordinary printed material available for use without a computer. Some are stand-alone titles, such as Living Books and Discus books. Others are parts of larger sets, such as IBM's Stories and More and Josten's Dragontales.
Multimedia writing tools engage children in oral language about their composing acts and final compositions. Children integrate previously prepared background illustrations, their own drawings, and writing into either stand-alone ''papers" or multimedia slide shows.
The development of comprehensive literacy software for preprimary and primary-grade literacy has been accelerating, together with the more recent surge in the power/cost ratio of desktop computers. IBM's Writing to Read program set the stage for classroom use of comprehensive literacy software programs for use in beginning reading instruction. Comprehensive literacy software programs that have been developed more recently and for which systematic evaluation has begun include Foundations in Learning by Breakthrough, Early Reading Program by Waterford, and the Little Planet Literacy Series by Young Children's Literacy Project.
Although the promise of new computer technology is real, it is still only a promise by any large-scale measure of effectiveness to address reading instruction. First, the availability of serviceable technology in U.S. schools remains unevenly distributed across school districts and is generally low. Second, for schools that have or are given hardware and software, studies repeatedly report implementation difficulties (Cuban, 1986; Sandholtz et al., 1997; Schofield, 1995).
Finally, even if current computing and networking resources were universally and easily available and practitioners were universally prepared to use them in their classrooms, their potential educational value depends on the quality of the software itself. Software can promote learning only to the extent that it engages students' attentionyet software that engages students' attention may or may not promote learning. The features and dynamics of software that determine its educational efficacy are subtle and, despite developers' best intentions, are often absent or mismanaged (Papert, 1996). As computing resources become more available, software that is well marketed, adequately engaging, and superficially appropriate may be purchased and used for educational purposes regardless of its real educational value in improving students' reading performance. To date, a great deal of educational software design is a commercial art
rather than an instructional science: it needs to be both. An analytic base is urgently needed for properly guiding and evaluating future educational software offerings.
In summary, with the availability of technology, quality software, and well-prepared practitioners, there is the potential for students to benefit. The materials described in this section were designed to offer distinct instructional strategies for learning to read; evaluation of each has revealed successful literacy growth and development in children (Sharp et al, 1995; Heuston, 1997; Zimmerman, 1997). Yet the use of educational technology and software is not available for all children; low and uneven distribution of technology places low-income and minority school districts at a disadvantage. Many schools do not have enough computers or have outdated nonfunctioning equipment. They may even lack the technical support and knowledge needed to maintain the use of computers in classrooms. Ultimately, constant evaluation and development of these resources will increase the value of technology in education.
RETENTION IN GRADE
In recent years, some schools have raised their kindergarten entrance age and have adopted the use of screening tests to determine school readiness (Cannella and Reiff, 1989). Some parents, hoping to avoid early school failureor to increase the likelihood of having a child who excels in comparison to classmateshave responded to the increased academic demands of kindergarten by holding their young children out of school for an extra year before kindergarten. This practice is sometimes referred to as "buy a year" or BAY (May and Welch, 1984). One effect of this growing practice is that the gap between the most and least advanced children in kindergarten and first grade has widened, making it more likely that children at the low end of their classes initially will appear even less successful when compared with older classmates.
In order to accommodate the perceived needs of at-risk children, schools have turned increasingly to providing them with an extra year of school. In addition to retention, or repetition of a grade, some school districts' extra-year programs, variously known as
prekindergarten or transitional first-grade or developmental first-grade classes, serve to extend the school career of many youngsters, as does retention at a grade level. Provisions for extra time in school are also secured through full-day (as opposed to half-day) kindergarten classes and through extending the length of the school year. Generally, the purpose of such options is to allow children the time and appropriate experiences needed for future school success.
Across the nation, the children most likely to be retained in early grades are those who are younger than their classmates, boys, children from low socioeconomic backgrounds, and ethnic or linguistic minorities (Meisels and Liaw, 1993). In the early grades, failure to achieve grade-level expectations in reading is the primary reason for retention. Retention has many supporters among teachers, administrators, and the public, but there is little evidence that retention practices are helpful to children (Shepard and Smith, 1990).
It is also important to note that few of these studies distinguish between children who are merely retained and those who are retained and receive special assistance. One such study found more favorable longitudinal results for achievement for children who did receive special services in the year following a grade retention.
A frequently cited effect of retention is the significantly higher school dropout rate for students who have experienced grade retention (Roderick, 1994). Other research indicates that dropping out is not a one-time, one-moment phenomenon. Students begin dropping out long before they are actually considered dropouts for data collection purposes. Clearly, we need to learn more about the social, emotional, and cognitive factors that precede dropping out. Furthermore, in the absence of better research, it is probably unwise to suggest, as some have, that the practice of retention in kindergarten and first grade should be entirely banned. It is certainly possible that for some children repeating a grade with services from a reading specialist or related service provider may produce more positive results than merely repeating the same sequence of instruction without any modifications, or moving on to the next grade with or without support. Nevertheless, the value of retention as a practice for preventing reading difficulties has not yet been amply demonstrated.
SPECIAL EDUCATION FOR LEARNING DISABILITIES
One response to the problems of children with reading difficulties is placement in special education programs, primarily services for children identified as learning disabled. In this section we discuss some factors that have limited the delivery of special education services to children with reading difficulties in the primary grades, as well as ways to maximize the benefits of reading instruction in these programs.
Federal legislation, notably Public Law (P.L.) 94-142 in 1975 and its amendments in 1986, was enacted to ensure the basic right to appropriate education for all children with disabilities, including specific learning disabilities in reading and writing. Congress intended that special education should address the problem of identifying and treating reading disabilities during the early school grades.
However, the law contained a definition of specific reading disability that has often contributed to an unfortunate delay in identification and treatment: to be eligible for special education placement, children must exhibit a severe discrepancy, typically 1.5 standard deviation units, between standardized tests of their reading achievement and their general intellectual ability. Schools are often hesitant to use standardized tests of reading achievement or IQ before the third grade, in the belief that most children with early reading problems will grow out of them. Longitudinal studies have shown, however, that most children who are substantially behind at the end of first grade remain behind in the later grades (Juel, 1988). When the disparity between achievement and IQ is finally noted in the later grades, it may be much more difficult for remedial instruction to counteract the emotional and educational consequences of early reading failure.
A second problem with the aptitude-achievement discrepancy criterion is that basic reading deficits and responsiveness to intervention have not been shown to be significantly different in children who meet or do not meet this criterion (discussed in Chapter 3). For example, a child with a standard reading score of 75 and an IQ of 90 is likely to show similar benefits from remedial instruction when compared with a child who has a reading score of 75 and an IQ of
100, but only the latter child would have a sufficient aptitude-achievement discrepancy to be eligible for special education services in most states. The learning disabilities field is acutely aware of the problems created by an arbitrary discrepancy criterion for special education services (see Lyon, 1995). The 1997 reauthorization of P.L. 94-142, however, still includes the earlier discrepancy criterion for specific learning disabilities.
In addition to the need for earlier intervention with less emphasis on aptitude-achievement discrepancy, there are a number of other complexities involved in considering the role of special education for young children with reading difficulties. The 1997 reauthorization of P.L. 94-142 discussed several concerns that needed to be addressed. These included the assurance of quality instruction in the regular classroom to reduce the number of students needing special education services, the use of proven methods and well-trained teachers in special education programs, greater attention to the effective integration of special education and regular classroom instruction, and the maintenance of high expectations for the achievement of children with learning disabilities.
An important component of the 1997 reauthorization of P.L. 94142 is its detailed agenda for additional research aimed at improving special education. Specified areas of research include the design of assessment tools to more accurately determine the specific needs of children with reading disabilities, longitudinal studies such as the one by Englert et al. (1995) to determine the optimal methods and intensity of instruction, and studies of effective practices for preparing teachers to provide services to children with learning disabilities. The knowledge gained from this research and its dissemination throughout the nation's teaching colleges and primary schools will help special education programs increase their contribution to the early prevention and remediation of reading disabilities.
Although many current special education programs for children with reading disabilities may fail to address some or all of the above concerns, there are some well-documented examples of successful programs. In one, the Early Literacy Project (ELP), special educators worked in collaboration with university educators to devise an approach that would be meaningful and beneficial for students with
mild disabilities in primary special education classrooms (Englert and Tarrant, 1995).
The principles of the Early Literacy Project include embedding literacy instruction in meaningful and integrated activities that span the disparate areas of the literacy curriculum (reading, writing, listening, speaking), guiding students to be self-regulating in their learning activity, and responsively instructing students. Activities involve the reading of connected text (using choral and partner reading to enhance word attack and fluency) and writing connected text (using emergent writing principles as well as strategy instruction in composition), interwoven through the use of a thematically based curriculum and teaching. Students in the Early Literacy Project also continue to receive instruction in Project Read, a systematic approach to phonics instruction that was in place in the participating schools. The comparison children for a study of the effectiveness of the ELP program were students in special education settings who were receiving Project Read instruction only.
The outcomes indicated that the average gain of children in the Project Read condition was .5 years on measures of word reading. The growth on the part of ELP students ranged from .7 years for those students whose teachers were in the project one year to 1.3 years for those students whose teachers were in their second year of the project. Furthermore, of the 23 students who received two or three years of instruction in their original teachers' classrooms, 19 were reading at or above grade level by the end of the second or third year and only four students continued to read below grade level.
This research is significant in several respects. First, it illustrates how curriculum and instruction can be designed and conducted in special education settings to advance children's literacy learning. Second, the finding regarding the more significant gains made by children whose teachers were more experienced in this form of instruction points out the important role that teacher expertise plays in maximizing their effectiveness with students who have significant reading problems. Finally, in the push for the inclusion of all children in the general education classroom, regardless of disability condition, it is important that we not lose sight of the intensive assistance that many of these students need in order to achieve at grade
level; assistance that will be very difficult to provide in a classroom context in which there is a ratio of 1 teacher to 25to 30 children.
Perhaps because of the serious consequences that a history of reading difficulties poses for children, or perhaps because of the intractable nature of some of these reading problems, the area of reading and learning disabilities has seen more than its fair share of therapies. These therapies are controversial in the sense that they are not supported by either contemporary theoretical understandings of the causes and nature of reading problems, nor are they supported by an empirical base. The therapies range from psychological to pharmaceutical to neurophysiological interventionsalthough clearly not all such therapies are controversial.
A number of reviews provide examples of controversial treatments that have garnered the attention, typically of the news media, and in turn, of parents and professionals as well (Hannell et al., 1991; Kavale and Forness; 1987; American Optometric Association, 1988; Worral, 1990; Silver, 1987). Those interventions for which, currently, there are no confirmed or replicated research findings that have nevertheless been touted to address reading and learning disabilities include: (a) neurophysiological retraining, which includes "patterning," optometric visual training, cerebellar-vestibular stimulation, and applied kinesiology; (b) nutritional therapies, such as megavitamin therapy and elimination (of synthetic flavors and colors) diet therapies; (c) the use of tinted lenses to correct for color sensitivity and thereby cure dyslexia; and (d) educational therapies, such as modality testing and teaching.
The consequences of the proliferation of quick fixes have an ethical dimension. As desperate parents cling to the hope for a miracle cure for their child's learning problem, more efficacious solutions are ignored. The disappointments add to the stresses already experienced by the parents of children with reading problems. A number of these therapies are a financial burden. Clearly parents need guidance and children need the best interventions for which we can develop evidence as to their efficacy and feasibility.
Several important themes have been stressed in this chapter. First, each literacy intervention must be considered in light of available resources, including financial, instructional, cultural, timing, and time required. Second, it is imperative to assess the existing external factors or characteristics before simply adding an intervention. Consideration must be given to the adequacy of existing instructional practices before deciding to implement any intervention. Third, the process of determining appropriate interventions must take into account the characteristics of students who are at risk for failure. For example, if an entire school is at risk, it might be wiser to begin an intervention that includes school-wide restructuring, as presented in Chapter 7, than to devote resources on an isolated turtoring technique.
Furthermore, a close examination of the successful supplementary interventions described in this chapter reveals a number of common features across these studies:
· Duration of the interventiongenerally occurring on a daily basis for the duration of a school year or a good portion of the school year.
· The amount of instructional timeall successful interventions involve more time in reading and writing than for children not at riskbut extra time is not sufficient in itself.
· In each case, there is an array of activities that generally consist of some reading (and rereading) of continuous text. In addition, each intervention features some form of word study. In some cases, specific strategies for decoding are incorporated.
· In all cases, writing is an important feature. However, the writing activity is not simply support while engaging in invented spelling; it is typically conducted in a more systematic manner.
· Although materials vary among the interventions, in each case there is careful attention paid to the characteristics of the material used, whether they are characterized as predictable, patterned, sequenced from easy to more difficult, or phonologically protected.
There is a focus on using text that children will find interesting and engaging.
· Each program includes carefully planned assessments that closely monitor the response of each child to the intervention.
Professional development of teachers, teachers aides, and professional or volunteer tutors were integral to each programthere is an important relationship between the skill of the teacher and the response of the children to early intervention. Effective intervention programs pay close attention to the preparation and supervision of the teachers or tutors.