Implications of the Present Findings in Distinguishing Between Reading Disabled and Non-Reading Disabled Children
The present findings have several important implications concerning the role of early identification and early intervention in distinguishing between reading disabled and non-reading disabled children and for preventing early and long-term reading difficulties in children who enter school at risk for such difficulties. First, they add to the growing body of evidence demonstrating that reading difficulties in most beginning readers can be corrected if such children are provided with early and comprehensive remedial intervention tailored to their individual needs (Clay, 1985; Foorman, 2003; Iversen & Tunmer, 1993; Pinnell, 1989; Torgesen, 2000; Torgesen, et al, 1999; Vellutino et al, 1996; Vellutino & Scanlon, 2002; Wasik & Slavin, 1993). They also extend our earlier findings (Scanlon et al, 2000, 2003; Vellutino et al., 1996; Vellutino & Scanlon, 2001) insofar as they document that children at risk for early and long term reading difficulties can be efficiently and economically identified on entry into kindergarten and that such difficulties can be prevented in most such children with relatively low cost (small group) remedial intervention implemented during their kindergarten year. Moreover, the data suggest that although some at risk children who receive kindergarten intervention will require additional remedial assistance at the beginning of first grade, most of these children will no longer require such assistance by the end of first grade. The important implication from these findings is that early identification of children at risk for reading difficulties and early intervention to institute foundational literacy skills in such children may be profitably implemented when they enter kindergarten (if not before), both to prevent them from experiencing early and long-term reading difficulties and to develop the means (e.g. normative data) for distinguishing between those who are no longer at risk and those who will continue to need some degree of remedial assistance beyond kindergarten, and in some cases, beyond first grade.
Second, the present findings provide additional support for the assertion (Clay, 1987; Vellutino et al.; 1996) that early and long-term difficulties in the majority of impaired readers are caused primarily by experiential and instructional deficits rather than biologically based cognitive deficits. Conversely, the data provide additional support for the assertion (Vellutino et al., 1996) that reading difficulties in some impaired readers are caused primarily by cognitive deficits of biological origin, especially phonological deficits that impair the acquisition of basic word level skills. Additional support for the former assertion is provided by the finding that emergent literacy deficits were remediated in most of the at risk children by the end of kindergarten and by the finding that performance levels in these children were, on several measures, comparable to those of the normal readers and stronger than those of poor readers who continued to need remedial assistance in first grade. Additional support for the latter assertion is provided by the finding that the cognitive profiles of the poor readers who received additional remedial instruction in first grade were generally weaker than those of both the normal readers and the no longer at risk children, and both sets of findings essentially replicate results obtained in the Vellutino et al. (1996) first grade intervention study. Yet, it is noteworthy that the majority of the poor readers were able to profit from the additional instruction they received in first grade and maintain their gains over time, despite the fact that their performance levels on most of the cognitive measures were not appreciably different from the performance levels of the poor readers who were unable to maintain the gains resulting from the additional instruction they received in first grade.
Collectively, these results imply that an at risk child's initial response to remedial intervention can be a reasonably good barometer of whether that child will be readily remediated or difficult to remediate, and that the profile developed from this and subsequent work with the child can aid as a "first-cut diagnostic" in determining whether his or her reading difficulties are caused primarily by experiential and instructional deficits or by cognitive deficits of biological origin. Thus, it would seem that at least some of the poor readers who continued to need remedial assistance after the intervention program ended, would be more defensibly classified as "reading disabled" (for whatever official purposes such classifications are needed) than either the children who were found to be no longer at risk following kindergarten intervention or the children whose reading difficulties were remediated after having received both kindergarten and first grade intervention. Of course using response to intervention as the initial means of making this distinction does not rule out the use of appropriate psychometric tests with well established measurement properties to cross-validate impressions gained from remedial work with the child (e.g. tests evaluating phonological awareness, verbal memory, name retrieval etc.), as we have suggested elsewhere (Vellutino & Scanlon, 2002). But, our finding, in the present study, that the Difficult to Remediate and Less Difficult to Remediate poor readers were not appreciably different on such measures suggests that response to intervention would be a more reliable criterion, not to mention its utility for developing the remedial strategies and activities that may be needed to facilitate literacy acquisition in children who prove to be more difficult to remediate.
A third implication of the results from this study is inherent in our finding that the intelligence measures did not reliably distinguish between the children who were difficult to remediate (DR) and those who were less difficult to remediate (LDR), nor between children in each of the three remedial groups (DR, LDR, and NLAR) on the one hand, and the normal readers of average intelligence on the other (AvIQNorm). This was found to be especially true in the case of the non-verbal measures of intelligence, no doubt because these measures are less vulnerable to Matthew Effects than are the verbal measures (Stanovich, 1986). In fact, whereas the DR and LDR group means were not appreciably different from the AvIQNorm group mean on the measure of non-verbal intelligence, the NLAR group mean was even somewhat larger than the AvIQNorm group mean on this measure. Moreover, the NLAR group mean was not appreciably different from the AvIQNorm group mean on the measure of verbal intelligence. And, although the AbAvIQNorm group generally performed above the AvIQNorm group on measures of word level skills as well as on the measure of reading comprehension, group differences on the word level measures diminished over the grades and were negligible on the phonological decoding and spelling measures by the end of third grade.
These results essentially replicate results from the Vellutino et al. (1996) study, which showed that performance on intelligence tests did not predict response to intervention in poor readers and did not reliably distinguish between poor and normal readers. Nor did it predict performance on tests of word level skills in average and above average IQ normal readers. Thus, the present results provide additional support for the contention that intelligence test scores are not strong or reliable barometers of reading ability, when this ability is measured by performance on tests of word level skills (Fletcher et al., 1994; Siegel, 1988; 1989; Stanovich & Siegel, 1994; Vellutino, et al., 1996; 2000). By logical extension, they further undermine the use of the IQ-achievement discrepancy to define reading disability.
Finally, in keeping with similar results from previous intervention studies (e.g. Clay, 1985; Torgesen, 2000; Vellutino et al, 1996), the present findings make it clear that some impaired readers will require continued and protracted remedial assistance, beyond that provided in the present and previous studies, if they are to eventually acquire functional literacy skills. This raises the question of how such remedial assistance should be implemented. Obviously, the question can only be answered through additional research, but it is worth reminding the reader that the children in the present study, who were found to be difficult to remediate, generally performed within the average range on the various reading measures at the end of first grade, after receiving both kindergarten and first grade intervention. However, they were not able to maintain these gains during second and third grade, and this raises questions about the adequacy of the classroom and remedial instruction they received after the project intervention ended. It is quite possible that many of these children were subsequently exposed to classroom and/or remedial instruction that did not facilitate application of the skills and strategies they had acquired through exposure to our intervention programs. Given the emerging consensus as to best practices in early literacy instruction (Adams, 1990; Foorman, 2003; Snow, Burns, & Griffen, 1998; National Reading Panel, 2000), there is a pressing need to ensure that classroom teachers and remedial personnel are able to understand and implement such practices. There is also a pressing need to explore the means for ensuring greater compatibility between classroom and remedial instruction, as the present findings strongly suggest. Thus, research addressing the quality of both classroom and remedial instruction and the relationship between these two enterprises, in preventing and remediating early reading difficulties is of critical importance.
It is worth noting, in connection with these latter points, that Fuchs and Fuchs (1997) have proposed a very promising curriculum-based "response to intervention" model that targets classroom instruction as a possible source of a child's reading difficulties, even before targeting ineptitude in the child as the primary source of such difficulties. The model includes systematic observation of classroom instruction, ongoing consultation with the classroom teacher, and a trial period of diagnostic teaching to help ascertain the best approach to individualizing both classroom and remedial instruction. Thus, it holds promise, not only for facilitating greater precision in identifying the source of a child's learning difficulties, but also for (1) fashioning the most effective approach to remediating these difficulties; (2) facilitating greater compatibility between classroom and remedial instruction; and (3) improving both classroom and remedial instruction through greater collaboration between the classroom and remedial teacher. In our opinion, this kind of model needs to be given serious consideration, along with more traditional and better established intervention models, in any discussion of possible approaches to operationalizing response to intervention in diagnosing reading disability (see Gresham, 2002 for an excellent analysis of response to intervention as an alternative to traditional approaches to identifying learning disabilities).
Previous Page | Next Page
First-Grade Intervention | Summary
