Responsiveness-to-Intervention Symposium

Application of RTI Models to Identification of Reading Disabilities in Older Students

In thinking about the application of RTI models to the identification of reading disabilities in older children, it may also be of help to identify what is known about rates of growth in reading skills that are typically achieved by current interventions. The interventions we have most direct knowledge of are not what is typically referred to as "secondary interventions" but rather are interventions that have been applied to children with serious to moderate reading difficulties at any point from 3^rd grade and higher. Table 1 reports outcomes from a number of recent studies using a common growth metric. This metric is calculated by dividing the amount of gain in standard score units by the number of hours of instruction that are provided, so rate of growth is expressed as the number of standard score points gained per hour of instruction. Of course, this metric depends on the common use across studies of standardized measures that have the same standard deviation, but there are a number of studies that have used measures similar enough to allow rough comparisons. Table 1 reports these growth rates for phonemic decoding (Ph. Dec.), word reading accuracy (Acc.), and passage comprehension (Comp.), along with other characteristics of the samples and the interventions they received. Not all scores are represented for each study, because standardized measures were not provided in all three areas of reading skill for all samples.

Table 1: Gains in standard score points per hour of instruction for three measures of reading skill

Several aspects of the data reported in Table 1 are worthy of specific discussion. First, there is remarkable consistency in the rates of growth for phonemic decoding skills, word reading accuracy, and passage comprehension skills reported across the studies. It should also be noted that these growth rates are far higher than is typically achieved in public school special education settings (Hanuschek, 1998). Second, growth rates for phonemic decoding skills are consistently higher than they are for word reading accuracy and passage comprehension. Even students who have failed to acquire functional word decoding skills by third grade can achieve rapid growth in these skills if taught with intensity and skill. Although not enough studies included standardized measures of reading fluency to include estimates of growth in this area, studies that do allow calculation of this metric for fluency indicate that it shows the slowest rate of growth (when compared against the fluency of non-disabled children of the same age) of any area of reading skill. For example, for samples 1,2,7,8,9,and 10 (numbered from top) in Table 1, the standard score gains in fluency per hour of instruction were .06, .01, .25, .14, .17, and .00. Rate of growth in the fluency area (in standard score terms) was directly related to the level of development in word reading ability the children attained before intervention began . Moderately impaired children, as in samples 7,8, and 9 in Table 1 show the best growth, while students with very severe word level reading skills after third grade, although they do become more fluent readers in an absolute sense, do not typically "close the gap" in reading fluency to a significant degree (Torgesen, Alexander, & Rashotte, 2001).

The third thing to note from our experience in calculating growth rates across reading skills for older children with reading difficulties is that, apart from the level of impairment of the children, several factors influence estimates of growth rates and outcome levels. For example, in our studies, we find that estimates of word reading accuracy are consistently higher when a measure of text reading accuracy (such as the Gray Oral Reading Test) is used rather than a measure of single word reading accuracy (such as the Woodcock Reading Mastery Test). The particular test used to assess word reading accuracy affects the estimate of final status more than it does the estimate of growth rate. The higher scores for the Gray undoubtedly reflect the student's ability to use passage level context as an aid to more accurate identification of words (Share & Stanovich, 1995).

Another factor that is likely to influence the estimate of growth rate obtained within any single study is the number of hours of intervention that were provided. Truch (2003) has recently documented that rate of gain may decelerate quite rapidly for intensive interventions after the first 12 hours of the intervention. In his study, 80 hours of intensive instruction using the Phono-Graphix method (McGuinness, McGuinness, & McGuinness, 1999) were provided to 202 students ranging in age from 6 years old to over 17 years old. For students ranging in age from 10-16, the average gains per hour of instruction for single word reading accuracy was .74 standard score points per hour of instruction for the first 12 hours of instruction. For the next twelve hours, the rate was .11, and for the final 56 hours, it was .10 standard score points per hour. Although this study did not calculate standard scores for their phonemic decoding measure, the findings were similar, but expressed in terms of grade level units per hour of instruction. For phonemic decoding, the growth rate for the first 12 hours of instruction was .25 grade level units per hour of instruction, for the next 12 hours it was .07, and for the final 56 hours, it was .04. This deceleration in growth rate across time within intensive interventions is probably part of the explanation for the particularly low growth rates observed in the 133 hour intervention study reported by Torgesen, et al., (2003).

Previous Page | Next page
(Third Grade) | (Conclusion)