Discussion
Children were selected for more intensive intervention if they fell below the cut-off scores on the phonological measures (O'Connor & Jenkins, 1999), WRMT-R, or oral reading fluency (Good et al., 2001) that have successfully identified children with reading disability profiles in these grades. Although data were collected by university personnel, teachers ultimately had control over how the data were used. For example, teachers determined whether students would receive Tier 2 as in-class or pull-out services. As teachers learned new ways to group for reading and to provide appropriate practice, some teachers were able to orchestrate these opportunities without pullout assistance from project personnel, although about 80% of Tier 2 and all of Tier 3 were provided by the research team. Thus, opportunity for children to receive targeted intervention prior to becoming eligible for special education interfaced with teachers' use of performance data.
Tiers 2 and 3 were not "one size fits all" interventions. The small group activities were tailored to the needs of the two, three, or four children in the group. As needs changed, groups were rearranged. Some groups focused on strategies for reading words children encounter in print and do not immediately know. Instruction in these groups included phonic generalizations (e.g., an 'e' on the end of a word often signals a long vowel) and orthographic patterns (e.g., -ain makes the same sound in rain, stain, and terrain) that commonly occur in written English (Gough & Walsh, 1991; Moats, 2000). Several groups in Grades 2 and 3 included comprehension strategies as part of their intervention, however, we found no children in either school with grade appropriate decoding and fluency, but poor comprehension.
Nine children identified as at risk in kindergarten caught up sufficiently in that year to need no additional intervention. Four of the 18 first graders who continued with Tier 2 intervention improved to the average range by the beginning of second grade. By third grade, only 9 of the 31 children identified in kindergarten continued to need interventions in addition to general class instruction. Of the children who received Tiers 2-3 for all four years, 7 became eligible for special education, two identified as first graders, 3 in Grade 2, and 2 in Grade 3.
Using Tiers of Intervention to Identify LD
The data in our study do not show identification as cleanly as those in Vaughn's work. Although all of the children eventually identified with LD were among the broad net we cast in kindergarten, we found more "in and out" movement, in which children appeared to be at risk at some time points, reached the average range (with Tier 2 or 3 interventions) for short periods of time, and then fell behind again when supports were removed and/or reading tasks became more difficult. Figure 1 demonstrates this messiness.
If we compare descriptive statistics for the children who were identified as LD with those at risk who became average readers, we see no differences in kindergarten scores, and only small trends at the end of first grade (e.g., about 2 raw score points on each subtest of the WRMT). Dot plots show considerable overlap among groups for these scores and also for oral reading fluency, although we also found a group of students who had participated in Tier 2 whose oral reading at the end of first grade exceeded 40 words per minute (the benchmark in the work of Good and colleagues), and these students needed no additional assistance through third grade. In our data, it was not until the end of second grade that the scores of the two groups (those with and without LD) became clearly differentiated, and even then, only in fluency and comprehension.
Achieving levels of fluency sufficient for the development of reading comprehension is clearly a stumbling block for students with RD, even as discrete skills, such as decoding words, improve. We saw changes between control and treatment years in the amount of time teachers in general classes devoted to fluency practice, particularly in the second grades. Although this increase in emphasis improved outcomes for the second graders overall, it was still not enough to eliminate the fluency bottleneck for the children with disabilities. Among the control group, for example, 23 of the 101 second graders read fewer than 50 words per minute in grade level text, a rate associated with reading failure by some researchers (e.g., Good et al., 2001). Children who participated in the tiered interventions fared better, however 9 children still read fewer than 50 words per minute at the end of 2nd grade. By the end of 3rd grade, 5 children in Tier 3 read fewer than 66 wpm, despite four years of intervention.
Our sample was small, however, our measurement system allowed us to track changes and growth spurts over time. We noticed a trend in the data between first and second grades and second and third that suggested that children eventually identified for special services tended to drop in reading scores over the summer months, while those without disabilities--even those at risk students who received Tiers 2 and 3--tended to grow. Future studies could be designed to explore this trend with larger samples, and to determine whether summer growth and decline might be another avenue for early identification and intervention in RD.
Rates of placement in special education. By assessing the effects and outcomes of each stage of intervention, we hoped that a decreased referral rate to special education would occur. Although the rates of placement in special education appeared to decrease, rates are not necessarily consistent in a school from one year to the next, therefore, we are not confident that this decrease is a result of the early intervention provided through this project.
We were encouraged with the reading progress of children with LD, however, children reading near the average range were still referred to special education during second or third grade. One possible reason for these "nearly normal" referrals was the discrepancy between the reading performance of children eventually labeled with disabilities, and the average reading achievement of children in their classrooms. When we examined the spread of scores during the treatment and control years, this possibility warranted attention. Although scores of the top 25% of students across years appear to be about the same, the scores of "average" children rose about a half standard deviation. Children diagnosed with reading disability had composite standard scores around 93, versus classmates with standard scores averaging 112. Thus, children diagnosed as RD were approximately 1.3 SD's below the mean of their peers, which is about the same difference between children with and without RD in the Control group.
A model of tiered interventions appears to help children who struggle with reading acquisition, and could help to document alternative instruction before a student is referred for special education services. By first teaching children in whole group lessons, then intervening in smaller groups, and then--only when necessary--with pullout instruction, we implemented a technique feasible in general and special education settings. By second and third grade, however, it became clear to us that the instruction in general education classrooms might not be a good match for the needs of children who had failed to thrive in Tier 2. The increasing difficulty of words and the expanding range of text in these grades made word study and fluency practice at appropriate levels difficult for general class teachers to manage.
Previous Page | Next Page
Results | References
