Research Support for RTI -- Are We There Yet?

By Joseph Kovaleski, Indiana University of Pennsylvania

At a recent meeting, a colleague opined that the research support for RTI was “just not there yet.” That comment got me thinking about what type of research support is needed for schools to adopt RTI and, in particular, the three-tier model. In this blog post, I would like to take a look at some of the research findings related to RTI. Because of space constraints inherent to blogs, I will not be able to provide citations at this time; however, I would be happy to provide them upon request.

It is useful to start with a distinction. As I described in previous posts, the contemporary three-tier model includes many features that were not components of precursor models, such as problem-solving teams. These new features are typically research-based practices that logically should improve the effectiveness of problem-solving teams. In addition, the three-tier model uses the problem-solving process as its central operating procedure.

So to evaluate the effectiveness of RTI, one should look at the effects of problem-solving models as well as the newly added features. We should also examine the two stated purposes of RTI: 1) to provide a framework of general education supports with the goal of bringing high percentages of students to proficiency in basic skills and 2) to serve an alternative procedure for identifying students with specific learning disabilities.

With both precursor models and contemporary multi-tier service delivery systems, there is a great deal of research indicating that students who are provided with interventions through team-based problem-solving processes improve academic performance in terms of numbers of academic goals attained and actual measures of academic performance. A number of studies of schools using the three-tier model have indicated that students in general have improved in academic skills on the basis of school-wide measures of performance.

One would expect that these types of outcomes will be enhanced if schools use a team-based problem-solving process that has good research support, and then add to the model other research-based practices such as universal screening, progress monitoring, and robust, Tier 2 interventions in reading, written language, mathematics, and behavior. Of course, as we found with our study of instructional support teams, positive outcomes only pertain when the system is delivered with a high degree of fidelity. So, in my view, I believe there is ample support for the implementation of multi-tier support systems in the schools.

Perhaps those who question the research base of RTI are more fundamentally concerned with the second aspect of its use, the assessment of students with potential learning disabilities. One common finding of both precursor problem-solving models and contemporary multi-tier models is that special-education identification rates decrease after the implementation of team-based problem-solving. Researchers also demonstrated that these procedures result in higher percentages of referred students being identified as eligible for special education, and at least one study indicated that the use of instructional consultation teams reduces minority overrepresentation in special education. Of course, demonstrating decreases in special-education identification rates is good news only when they are accompanied by increases in overall student performance. It is good to see that most evaluations of multi-tier service delivery systems now include both measures in their program evaluation efforts.

What is more difficult to determine through research is whether RTI procedures identify students who are “truly learning disabled.” However, this may be of a conceptual question than an empirical one. The critical research question is whether the “right” students are being identified as having learning disabilities. It appears from RTI implementation sites that roughly the same number of students are identified using RTI as with the ability-achievement discrepancy approach, although, as indicated previously, some sites report decreases in identifications.

It could reasonably be concluded from these studies that RTI decreases the identification of those students who have been termed “curriculum casualties,” as Tier 1 supports and Tier 2 interventions should put these students on the road to appropriate levels of academic attainments. The students who will be identified as having learning disabilities using RTI procedures are those who display significant deficits as compared with state standards and who fail to progress in response to robust, scientifically based instruction and who need special education to make acceptable academic progress.

Given the efficacy of a well-delivered three-tier system of supports, this should be a fairly small number of students. Some would suggest that this winnowed-down pool of students should then receive further testing (e.g., psychological processing) to determine the “real LDs.” My concern with such an approach is that we would then potentially eliminate from eligibility those students who display significant academic concerns but fail to display deficient psychological processes. In my opinion, this would be a disservice to these students, who, in my mind, are just as disabled as those who display other deficits.

Tell me, readers:  Do you think we're "there yet"?

Comments

[Mike]

We are certainly not there yet. I think that the way a school/group of teachers and administrators use the system has a major impact on all of the factors mentioned.

[RShaw]

You asked, "What is more difficult to determine through research is whether RTI procedures identify students who are truly learning disabled." The premise of your question falsely suggests that, in part, RTI is or should be a diagnostic process. I would argue that for the practicing school psychologist it is an eligibility process. This means you should ask, "Can RTI procedures appropriately help determine whether students are eligible for special education as LD according to state and federal criteria."

School psychologists don’t diagnose LD. Let’s stop suggesting that they do. School psychologists, as part of a multidisciplinary team, determine eligibility for special education according to criteria only remotely associated with the diagnosis of LD. After 25 years of practice, I can’t tell you whether a student is “truly learning disabled,” but I can tell with a reasonable degree of confidence whether a student is eligible for special education as LD. And, I can also tell you that I am more confident about my eligibility decisions using an RTI approach than I ever was using the Severe Discrepancy approach.

[Joe Kovaleski]

RShaw,

Excellent points. I agree and regret if I suggested that school psychologists diagnose SLD. You're correct that that it is a team decision regarding eligibility, and not diagnosis per se. I am very comfortable with your clarification. Thanks.

Joe Kovaleski

[Joe Kovaleski]

Mike, You're right. How RTI is interpreted has everything to do with treatment fidelity, which is why so many of us emphasize the need for building a school-based system that is capable of delivering that level of integrity.

Joe Kovaleski

[sheryl nagel]

I think that as we look at RTI and this three tier model, it is another assessment tool and help that we can give to our strugglers. I feel that any additional help that we could give to these children is beneficial if the right teachers and personnel are giving it their all.

[Sandra Brown]

I am encouraged by the RTI model use as a tool to help determine eligibilty for learning disability. Unfortunately I have witness occssions where the evaluation process has been misused to ensure that students are eligible for special services. The RTI model may provide more accurate data that can be used to assess students performance, so that interventions are put into place before a referral for special education evaluations.

[Regina]

We actually started discussing and implementing an RTI model for our school and district 3 years ago. However, last year we go a new administrator, and we are not as informed of these processes or as involved. We are using this model to some extent, but it is unclear as to exactly how we have implemented.