(Guest post by Reid Lyon)
Jay’s analysis of Response to Instruction, also known as Response to Intervention (RTI), and the need for additional reforms to help ensure effectiveness is provocative and informative. I agree with a good deal of his analysis but feel the need to expand the discussion a bit.
First off, his piece offers a wake-up call to those who are prone to fall in love with magic bullets that will close the achievement gap and reduce referrals to special education. One of my dreams is to provide policymakers and educational leaders two free tattoos to be placed anywhere they want on their persons that read: (1) “Necessary but not sufficient,” and (2) “Great policy idea, but implementation is a bitch.” Jay’s points address both tattoos.
RTI is a noble and well researched concept. One of its major goals is to reduce referrals to special education by documenting that the student’s learning difficulties are not because of inadequate instruction in general education classrooms but because of a disability. Years ago, S. Allen Cohen provided us with a more interesting term for lousy teaching which he called “dyspedagogia” (I believe this was tongue in cheek).
But the fact is most kids identified for special education and labeled as having a Learning Disability (LD) are not LD but achieve poorly because of “dyspedagogia”. In fact, our research over the past 20 years has taught us that scientifically based early reading intervention provided through a tiered approach to instruction can reduce the percentage of LD from upwards of 22% to between 2% and 10% in some states and LEAs. This is a very good thing given that LD referrals and placements constitute about 50% of all referrals to special education, and reading disabilities comprise about 80% of kids identified with LD.
Those working on the development of RTI knew that classroom teachers and special educators are from different planets, with entirely different backgrounds in curriculum and instruction and entirely different professional languages. This makes collaboration and the provision of coherent and systematic instruction impossible. In essence, RTI was developed, in part, to underscore the need to develop a common assessment and instructional language between well intentioned teachers with different specializations so that instruction was not so fragmented, and so kids would not feel like ping-pong balls when receiving entirely different types of instruction on the same day from classroom teachers and special educators. There was no other alternative – typical colleges of education would not provide the professional development necessary to accomplish the level of shared training and collaboration skills essential for effective and differentiated instruction.
Related to this, RTI as a concept was also developed to take advantage of the converging research that prevention through early identification and intervention provides a more effective approach to reducing academic failure than hanging out waiting until the kid chucks his books through the window in the third or fourth grade. RTI procedures can differ across LEAs but they typically screen all students in kindergarten or early first grade to determine which students require enhanced instruction in the classroom or more intensive instruction in smaller groups, while still under the direction of the regular classroom teacher. If the kid demonstrates little academic (or behavioral) growth following more intense instruction, then special education becomes a possibility. It is critical to remember that being placed in special education is also a tiered process. The questions are: Does the student have a disability? (RTI contributes to answering this question); and, if yes, does the student require special education? (RTI contributes to answering this question).
So we have two elements in RTI that we know can increase achievement (at least in reading): (1) collaboration/common language and (2) prevention is where it’s at. These elements are inseparable. But we know that these two elements cannot work the way they are supposed to unless the folks who run the system know all of the potential barriers to the implementation of RTI and ensure that essential conditions to support the initiative are in place. There are examples of districts implementing successful RTI initiatives that increase students’ reading achievement and reduce referrals to special education. There are other examples where the initiative has not resulted in changes in achievement or referral outcomes.
So what is going on? It will be important to figure out what works in some districts and not others. Susan Hall has recently published a very user-friendly book that lays out what districts and schools did in implementing RTI in a way that resulted in substantial reductions in referrals and significant increases in reading achievement: A Principal’s Guide: Implementing Response to Intervention. I am not hawking this book, as there are others that lay out the conditions that are essential for implementing RTI and scaling it (see Dianne Haager et al. for specific evidence of effectiveness).
The bottom line is that effective implementers of RTI have had to do a tremendous amount of study and planning to ensure that the initiative actually makes a difference. Jay lays out some the barriers, including problems with implementing a program that actually takes funding away from your district or schools, and persuading educators to replace programs and procedures that they have used for some time.
But there are others as well. For example, how do you overcome the fact that intervention services in schools are often funded by separate entitlement programs, especially Title I and IDEA, that have specific eligibility criteria that make it difficult to co-mingle funds to support school-wide programs? How do you implement programs that have been typically isolated from general education? And how can school leaders and teachers avoid the mistakes that result in limited or no effectiveness, not to mention that the excitement for change and increased morale will be crushed?
If you did a factor analysis of all the crap that can derail the implementation of RTI, these are the most common errors:
–Focusing Too Many Resources on Administering and Collecting Assessment Data Rather Than Ensuring that Staff Use the Data to Inform Instruction
-Viewing Purchased Programs as Silver Bullets Rather Than Aides to Help Well Prepared Teachers Make Informed Instructional Decisions
-Confusing Awareness Training with Implementation Training
-Using Ineffective Practices to Train Teachers
-Underestimating the Magnitude of Change
-Taking on Too Many Grade Levels and Schools the First year
-Beginning the Implementation Without a Comprehensive Implementation Plan
-Failure to View the Implementation as a Systems-Wide Change
The good news is that districts and schools that have effective RTI programs in place know they can’t make the mistakes above and provide incentives that have trumped the traditional financial rewards that have potential for increasing referral rates.
I am sure that I have taken up too much space with details that may be of little interest to policymakers. But the details are what make RTI effective, and when RTI works, it really works. However, if you can’t deal with complexity, either-or concrete thinking, or have an allergic reaction to human and systems change, you might as well blow off trying to implement RTI.
Jay has done a service laying out the big picture issues. Implementing additional reforms to increase the probability that RTI can succeed is essential, as he has articulated. But the return on investment is only as good as understanding and addressing the amount of grunt work involved. It’s that “necessary but not sufficient” thing. I know this is self-evident, but we sure keep trying to do one without the other.
Jay P. Greene's Blog 
Great post Dr. Lyon. It sounds like those interested in RTI should make switching away from a bounty funding system a part of any statewide effort to improve diagnosis as a “necessary but not sufficient” step to create the correct incentives.
Under the bounty system you can be foregoing revenue when you correctly identify children. Add that to all the other problems you describe equals a recipe for disaster.
Interesting post, Mr. Lyon.
I suggested something along the lines of dyspedagogia in an article I wrote on traditional math. My argument is something along these lines. I’ve excerpted it here:
The mischaracterization of traditional math fails to capture the nuance of instruction and scaffolding that many teachers (including mine) engaged in.They didn’t just stand there and lecture; they asked us questions and goaded us into thinking, providing us with prompts and support through many worked examples.This is a form of direct instruction, and it was combined with cumulative reviews and mastery learning.It is interesting therefore to learn that direct instruction and mastery learning are recommended methods of teaching for students with learning disabilities. (Rosenberg, et al., 2008)It is also interesting to note that over the past two decades, the number of students with learning disabilities has increased.In 2006, approximately 2.6 million students were identified with learning disabilities, more than three times as manyas were identified in 1976-1977.Although one reason for this growth might be better means of diagnoses of specific disorders, there has still been growth.Between 1990 and 2004, 650,000 additional students were identified with learning disabilities, representing a 31% increase at a time when the overall student population grew by only 15%.(U.S. Department of Education, 2006).
The increase in the number of students with learning disabilities raises the interesting question (if not uncomfortable for some), of whether the older way of teaching (direct instruction and mastery learning) may have had unintended benefits. According to Rosenberg, et. al. (2008), one factor associated with the identification of students with learning disabilities is the lack of access to effective instruction.Rosenberg et. al, also note that up to 50% of students with learning disabilities have been shown to overcome their learning difficulties when given high-quality instruction. Is the shift toward inquiry-based teaching resulting in more students being identified with learning disabilities?Are these students who in earlier days would have swum with the rest of the pack?
(From http://ednews.org/articles/19414/1/It-Works-for-Me–An-Exploration-of-Traditional-Math-Part-III/Page1.html)
I hope this is looked into more seriously given the release of the National Math Panel’s report.
Barry: I believe you hit several nails on the head. The increase in referrals for special education are accounted for in part by ineffective instruction. RTI has taught us that referral rates can be reduced significantly when classroom teachers are provided with professional development that helps them identify academic strengths and weaknesses and to customize instruction based on continuous assessment data. There is overwhelming research evidence that instruction that is direct, explicit and systematic is significantly more effective than instruction that less intentional and systematic – particularly with kids who are having a tough time learning concepts whether it be reading, math, or science. But the gap between what we know works and what is actually implemented in schools and classrroms remains pretyy massive, no matter how good the research is. Trying to understand why we don’t employ what we know is effective has been a great deal harder than I ever imagined.
[…] Can RTI Work? We Went to The Moon, Right? […]