Response to Intervention: Considerations for Research and Systems Change

Abstract

This invited commentary on the series of articles regarding the state of the extant research and practice concerning the implementation of response to intervention (RTI) will focus on the following issues: (a) what constitutes a comprehensive service delivery system that uses RTI as its driving principle, (b) the dependent measures by which the effects of RTI should be evaluated, (c) how treatment fidelity can be meaningfully facilitated in real school environments, (d) what can be done with students who are not responsive to even robust interventions, and (e) which students should be identified as having specific learning disabilities. Throughout the discussion of these issues, implications for the practice of using RTI in the schools and research directions will be presented. The Multitier Service Delivery Model A worthwhile place to start is to attempt to offer some clarity about what is being proposed as the multitier service delivery model that uses RTI as its recurrent scheme to design interventions, monitor progress, and make both small-scale and large-scale decisions about students. In many descriptions of RTI, both here (Glover & DiPerna, 2007) and elsewhere, it appears to be assumed that the projects that began in the 1990s, including Iowa's problem-solving model (Tilly, 2003), Pennsylvania's instructional support teams (Kovaleski & Glew, 2006), Rosenfield and Gravois's (1996) instructional consultation teams, and others are being presented as intact RTI models. However, although these precursors used problem solving and progress monitoring as core components of their models, none of them embodied all of the components of current conceptualizations of a multitier service delivery system (e.g., Batsche et al., 2006). These pioneering approaches to intervention support models, which originated before the advent of the National Reading Panel (2000), standard protocol interventions (Wanzek & Vaughn, 2007), school-wide positive behavior supports (Sugai & Horner, 2002), the Dynamic Indicators of Basic Early Literacy Skills (DIBELS; Good & Kaminski, 2005), and other research-based initiatives, generally assumed (or did not address) the effectiveness of the general education program and endeavored to customize interventions for individual students who were not successful in it. (It is acknowledged that some of the beginning tiers or phases of these models involved consultation with general education teachers, but not for the purpose of making fundamental changes to the core program.) They were, in effect, single-tier approaches, and in today's terms might be understood somewhere in the middle of most contemporary multitier RTI models. Simply stated, RTI cannot be equated with the problem-solving process or with problem-solving teams. Nonetheless, as late as 2003, Fuchs, Mock, Morgan, and Young were posing the problem-solving process and standard protocol interventions as alternative approaches to RTI. In contrast, current depictions of a multitier service delivery model (Batsche et al., 2006) incorporate both of these approaches as fundamental components. Furthermore, in an effort to harvest the research efforts of the last 10 years, other features such as instituting a scientifically based core curriculum; universal screening with DIBELS, AIMSweb (Shinn & Germann, 2006), and other instruments; data-analysis teaming; differentiated instruction in general education; and school-wide positive behavior supports have been proposed as fundamental components of a comprehensive model (Kovaleski, 2007). The most important distinction between the precursor and current models is that the effectiveness of the core academic and behavior programs for all students are not taken for granted. Rather, in the face of large numbers of students who continue to fall below proficiency levels, the current models directly address the source of the problem: the inability of many schools to bring large percentages of their students (80% according to Batsche et al. …