New Directions in Research
RTI (Response to Intervention): Rethinking Special Education for Students With Reading Difficulties (Yet Again)

Russell Gersten, Instructional Research Group, Signal Hill, California

Joseph A. Dimino, Instructional Research Group, Signal Hill, California

About the authors

The field of special education seems to lurch forward in a seemingly never-ending series of reforms and initiatives. Legislation often has been influenced by the aspirations, hopes, and dreams of family members of those with disabilities coupled with the theoretical models and aspirations of researchers. Rarely is there a strong empirical basis for the proposed reform.

Although this trend is true for many fields of human service, the speed of these reforms in the field of special education seems extraordinary. More recently, every shift or change in special-education policy or procedure has had dramatic repercussions for the field of reading instruction. These reforms invariably have a profound effect on students with reading difficulties, because the largest groups of special-education students are those with LD, and the vast majority of these students demonstrate serious difficulties in reading. Response to Intervention (Fuchs & Fuchs, 1998; Vaughn & Fuchs, 2003), or RTI, is the latest of such innovations.

The purpose of RTI is not only to provide early intervention for students who are at risk for school failure but also to develop more valid procedure for identifying students with reading disabilities. As the authors of the accompanying articles in this volume explain, RTI allows teachers to judge which students need special education instruction in reading based on whether or not the student can respond to either typical classroom instruction, or the type of support that is possible in a typical classroom (e.g., brief but intensive small-group intervention on key skills). It seems much more direct and logical than relying on discrepancies between IQ scores and reading achievement scores. Another appealing feature is the fact that it is a form of dynamic assessment (e.g., Campione & Brown, 1987). Reading researchers have been trying to operationalize that concept for several decades, with minimal success until recently.

Unlike the innovative concept of the resource room in the 1970s, and prereferral interventions in the 1980s, this model is being scaled up based on a body of controlled scientific research. Admittedly, this body of research is small; as Fuchs and Fuchs (in this issue) note, there is considerable hype surrounding RTI. Having watched the problems with the resource-room model, prereferral intervention, and the consultation model in actual implementation, we feel it is important to raise issues and concerns as large-scale implementation begins to unfold. In this article, we discuss aspects of RTI that appear to be promising for the field of reading, and those that we view as problematic. Unlike many earlier models of innovation in special education service delivery, there is some research indicating promise. Most of that research has been conducted with students in the first two or three years of schooling (e.g., O'Connor, 2000; Speece, Case, & Molloy, 2003; Vaughn, Linan-Thompson, & Hickman, 2003), so RTI in early intervention will be the focus of this article.

The way we were: Understanding why early intervention in reading did not always seem optimal

With the perspicacity of hindsight, it is easy to condemn the so-called “wait and fail” method commonly used by school districts since it seems to senselessly delay help for students who require assistance. As commonsensical as the RTI concept seems in 2005, in the 1980s and 1990s the concept of early identification of students with reading difficulties seemed cruel to many. Typically, districts wait until the end of second or beginning of third grade before determining that a student has significant disabilities in reading.

Until recently, identification was linked to referral for special education services, and the thought of labeling a 5- or 6-year-old as learning disabled was deemed improper. Special educators at that time failed to provide little but resource-room services for students with a label of LD. It seemed more humane to provide students with a chance to mature before making the serious determination that special education placement was necessary. Ironically, as Juel noted as early as 1988 in her longitudinal research, students who do not learn to read by the end of first grade almost invariably remain poor readers.

Many of the commonly used early screening measures had not yet been developed; the few phoneme awareness measures that were developed were not well known until the early 1990s along with the wide dissemination of Adams's (1990) Beginning to Read: Thinking and Learning About Print. The typical readiness tests from the 1970s and 1980s, such as the Metropolitan Readiness Test, possessed a predictive validity of close to zero. The reading field learned from Adams's synthesis that neither language proficiency measures nor IQ predicted the ability to learn to read (i.e., successfully and fluently decode at the first- and second-grade level). In fact, contemporary research on English learners continues to confirm these findings (e.g., Lesaux & Siegel, 2003). However, it was not until awareness that both phonological processing measures and measures of naming the letters of the alphabet were reasonably valid predictors of future reading ability that the field had some basis for making determinations with any type of precision. (Precision of many of these measures for kindergartners remains far from perfect.)