Response to Reynolds and Shaywitz: Let’s Not Go Back to the Good Old Days before RTI
In a recent paper1, the prominent researchers Cecil Reynolds and Sally Shaywitz express many reservations and concerns about Response to Intervention (RTI) models and implementation. Chief among those concerns is that RTI models as originally conceived are not working to the advantage of students with dyslexia and other learning disabilities. They characterize RTI as a "fad” that lacks sufficient research support or implementation guidance from federal, state, or local agencies. As a consequence, they assert, some states and districts are using response to instruction as the only criterion for determining whether or not a student has a handicapping condition, and comprehensive evaluations are being avoided or denied to frustrated parents and teachers. Reynolds and Shaywitz defend the importance of IQ tests and discrepancy criteria in determining the existence of dyslexia and other learning disabilities and express concern that those students with higher IQs are not recognized as having learning disabilities if low achievement (non-responsiveness to instruction and intervention) is the only criterion used to identify LD in bright students. They argue that current implementations of RTI are steering diagnosticians away from comprehensive evaluations that attempt to measure the processing deficits of LD students and that, in their view, provide valuable insight into a student’s strengths, weaknesses, and instructional needs. They conclude that RTI should not be used to determine eligibility for special services; that comprehensive assessment is necessary before intervention is implemented; and that teachers should be required to use evidence-based practices before RTI can be a legitimate and effective approach to serving students with LD.
Although Reynolds and Shaywitz raise some legitimate concerns with which we agree, we also believe their arguments place too much credence in failed policies of the past and too little emphasis on effective implementations of RTI. Rather than a “wait to fail” model, RTI’s main purpose is to promote early identification of students at risk, immediate delivery of instruction and intervention, and prevention of academic failure. When RTI is used effectively, students at risk, including those with dyslexia and other forms of LD, receive evidence-based instruction and interventions even before determination of eligibility for special education.
Federal funding policies in the revision of IDEA 2004 permit districts to use a small portion of their federal special education funds (up to 15%) for preventive screening and intervention, prior to their use of a special education eligibility standard. This policy change – a move away from dependence on comprehensive evaluations to justify supportive or intensive instruction – came about for several reasons established by decades of research. Evidence has accrued for many years showing that a) students at risk for reading failure can be identified in preschool or kindergarten with simple, inexpensive screening assessments; b) intervention is most effective and least costly at the kindergarten and first grade levels; c) early intervention, outside of special education, can be sufficient to prevent or ameliorate potential learning disabilities in many students; and d) regular classroom and supplemental instruction are key factors in improving student outcomes.
Simultaneously, evidence is strong that special education laws and policies, prior to 2004, rewarded schools for retaining students instead of teaching them, delaying identification of special needs, and spending limited resources on testing rather than instruction. After the third grade, when special education referrals tend to increase dramatically, special education placement is not, on average, enabling students to make significant gains in relative standing. Special education placement is seldom associated with more effective instruction or greater rates of gains for students with LD once they are identified.
Several meta-analyses and longitudinal studies of reading development and disabilities have also shown that students with IQ-achievement discrepancies are not different from students without IQ-achievement discrepancies in the nature or expression of their learning disability (Fletcher, Lyon, Fuchs, and Barnes, 2007). In other words, students who are poor readers, who score below the 20th percentile in reading, cannot be distinguished on the basis of their IQ. The manifestations of their reading disability, their cognitive profile of strengths and weaknesses, their pattern of growth in reading skill over time, and their typical response to instruction do not vary by IQ. Intelligence or IQ may matter in other ways, such as student aspirations, interests, and academic coping skills, but the presence or absence of a discrepancy should not determine whether or not the student is taught using appropriate methodology. These findings have been replicated many times and the justification for using IQ-achievement discrepancies to identify and treat reading disabilities has been seriously challenged in the research community for over 20 years.
That said, we agree with Reynolds and Shaywitz that some practices now adopted by states, districts and schools are not in keeping with research or with the intention of well-conceived, well-implemented and sustained RTI frameworks. The core activity in an RTI model should be informed problem solving around student needs and the use of universal screening data to find and serve students at risk. RTI requires participation from an entire school faculty (with students and families) and purposeful use of various kinds of data. Data must be interpreted to group children for interventions focused on specific academic and/or behavioral weaknesses. Progress-monitoring is necessary to know who is responding to instruction and who is not. All of this requires that a faculty be educated to understand how students learn to read and write, why the essential components of instruction are necessary, how to teach the content (e.g., the structure of language), and how to give and interpret assessments.
Those core activities, however, in no way are intended to supplant comprehensive evaluations if students are referred by teachers or if parents request an evaluation. Federal law states unequivocally that, at any point, students have a right to an individual, multi-disciplinary, comprehensive evaluation if requested by parents or teachers. Learning disabilities are manifested in various ways, but students who are struggling to learn from day to day, who are confused and anxious in spite of sound instruction, and who have unusual difficulties learning any essential subject matter should be carefully watched and referred for more in-depth diagnostic assessment. Comprehensive assessment should consist of research-validated tests and protocols that are directly relevant for deciding how to teach the student.
One of the Reynolds and Shaywitz criticisms of RTI is that smaller scale research, on which many RTI practices have been developed, cannot be “scaled up” successfully and that malpractice is more prevalent than good practice. Our collective experience with implementations of RTI is that they represent the continuum from poor to highly successful. The existence of poor implementation, however, does not mean that the framework or model is wrongly conceived or that we should do away with it and go back to the good old days of discrepancies, late identification, delayed and expensive evaluations, and little accounting for instruction. Rather, we need to redouble efforts to ensure that administrators and teachers are much better educated to carry out their responsibilities with fidelity and sustainability. In addition, the field needs to clarify for itself some good, consensual answers to recurring questions such as these:
- What does the term RTI refer to? Can we agree?
- What constitutes an adequate “response” to instruction? Is this a matter of a measurement formula, observation, judgment, or all of the above?
- Why is a system-wide model necessary? Why is RTI a whole-school framework?
- Why is RTI a paradigm and not a model? That is, why is it a way of thinking about students, programs, and resources rather than a formula or prescription for matching programs to students?
- What is the role of curriculum-based measurement in relation to traditional, standardized, diagnostic tests?
- How should we address the problem of over-identification of students at risk that occurs with many screening tests?
- How should we locate and respond to the needs of high-IQ, struggling students, even if they are not eligible for special education services in an RTI framework?
- What is the relationship between “problem-solving” and the “standard treatment protocol” approach to RTI? Don’t all good RTI implementations require continuous, student-focused problem solving?· Don’t all good RTI implementations require some level of use of standard protocols?
- What knowledge and practice standards must be met for administrators, coaches, and teachers to use the RTI framework successfully? How much professional development time is required for those standards to be met?
- Who should administer screening, diagnostic, and progress-monitoring assessments? What should the teacher’s role be?
- If a student receives a comprehensive evaluation, what should be measured and who makes that decision?
- How do we help administrators evaluate the quality and fidelity of implementation of instruction and intervention in their school?
We still have a long way to go before the tenets of RTI and the problem-solving required by it are more universally understood and wisely applied. However, there should be no going back. Education life as we knew it, with unjustifiable practices, was not serving the needs of students with LD or other students who were dependent on informed instruction to limit or even eliminate the need for special education. At least we have turned the corner toward research-driven ideas and have a moderately clear path to realizing them.
1Reynolds, C. R. Shaywitz, S.E. (2009) Response to intervention: Ready or not? Or, from wait-to-fail to watch-them-fail. School Psychology Quarterly, 24 (2), 130-145.
Fletcher, J., Lyon, R., Fuchs, L., and Barnes, M. (2007) Learning Disabilities: From Identification to Intervention. New York: Guilford.
Note: Many thanks to Jim Wendorf of the National Center for Learning Disabilities for his helpful editorial comments.
Back To Top