“Waiting to Fail” Redux: Understanding Inadequate Response to Intervention

Abstract

This introduction to the special issue provides an overview of the promise, but also the ongoing challenges, related to Response to Intervention (RTI) as a means of both prevention and identification of reading disabilities. We conclude by describing the articles in this special issue and considering their implications for future research.

The purpose of this special issue is to examine the promise, as well as concerns, about Response to Intervention (RTI) subsequent to the now decade-old amendments to Individuals with Disabilities Act (IDEA; 2004). These amendments, which allowed states to use RTI both for prevention and for identification of reading disabilities, were in large part driven by concern about several important issues among researchers, policy makers, practitioners, and parents that the IQ-Achievement discrepancy based formulas used to identify students with reading disabilities had become a “wait to fail” model (Fuchs & Fuchs, 1998; Lyon, 1995; Vellutino et al., 1996). We describe the reasons for these concerns below, and then discuss how RTI was designed to address these concerns

First, the mean age at which students were identified as having a reading disability, and were therefore eligible for intensive intervention (special education) was 10 years (or about third-fourth grade), despite converging evidence indicating that prevention was easier and more effective than remediation (e.g., Torgesen, 2000). Second, there were important psychometric concerns about inaccurate (over- and under-identification) classification and identification of reading disabilities for students from minority backgrounds, for students attending schools serving a high proportion of students from low socio-economic backgrounds, and for students with Limited English Proficiency (e.g., Hosp & Reschly, 2004; MacMillan & Reschly, 1998). Further complicating the matter, since instructional response was not formally part of the discrepancy model, it was unclear whether children had actually received and not responded to high quality Tier 1 or if they might not have received adequate instruction and thus could be considered as having an instructional disability (cf. Vellutino et al., 1996). Third, none of the tests used to determine the IQ-Achievement discrepancy provided adequate formative data-based guidance to help educators select and differentiate interventions (e.g., Fletcher, Francis, Shaywitz, Lyon, Foorman, Stuebing et al., 1998). Relatedly, research showed that students who were so-called garden-variety poor readers had similar growth trajectories as students who did manifest the discrepancy (Fletcher, Lyon, Fuchs, & Barnes, 2007).

Concern about these issues regarding the IQ-Achievement discrepancy culminated in a series of town meetings and a subsequent Learning Disabilities Summit, held in 2001, of leading researchers in the fields of special education and school psychology. These researchers argued in a series of white papers that there was a critical need to provide powerful early preventative reading interventions rather than waiting for children to fall far enough behind to qualify for help (far earlier than by third grade) (Bradley, Danielson, & Hallahan, 2002). There was evidence that reading trajectories were fairly stable, but at the same time, there was optimism (cf. National Institute of Child Health and Human Development, 2000) that early reading interventions that provided phonological and phonetic instruction could reduce reading problems. In addition, there was optimism that through RTI, it would be possible to differentiate students with true reading disabilities from those who had not received evidence-based reading instruction (Vellutino et al., 1996).

All in all, momentum gathered supporting RTI, which promised to incorporate the science of reading intervention and knowledge about screening and progress monitoring into general education classrooms (Tier 1) to reduce risk and prevent many reading problems. Beyond this critical foundation, RTI would immediately provide extra layers of increasingly intensive intervention. This tiered system would involve small group interventions with more frequent progress monitoring in Tier 2, and for those very few students who did not respond, more intensive Tier 3. As a field, professionals hoped that RTI approaches would reduce current special education case-loads and allow for more individualized reading instruction for students demonstrating the most persistently or chronically inadequate response, who we assumed would form a class of those with true reading disabilities (Vaughn, Moody, & Schumm, 1998). In sum, it was hoped RTI innovations would focus special education services on students who in the past were termed “treatment-resistors” or “nonresponders” (Blachman, 1994; Torgesen, 2000). In 2000, Torgesen estimated that only 2% to 6% of the school population might be truly unresponsive to generally effective intervention efforts (note that he based this estimate on students having access to powerful preventative instruction and intervention, but who had not acquired word reading skills above the 30^th percentile). At the time, his estimates were very similar to federal incidence figures for children with reading disabilities.

Multi-tier models are now used for prevention purposes in all 50 states, but there are explicit legal guidelines for identification and classifications in only 13 states (Zirkel & Thomas, 2010). The Institute of Education Sciences published a practice guide for RTI that identified five core components for the effective implementation of RTI (Gersten et al., 2009). These include universal screening, a high quality core reading program, progress monitoring, increasingly intensive tiers of intervention, and fidelity of implementation. These core components are salient in guidance documents for RTI such as practitioner guidebooks, state legislation, and technical assistance papers. Additionally, syntheses have been conducted to examine the efficacy of intensive interventions across the early grades, K-3 (Al Otaiba & Fuchs, 2002; Burns, Appleton & Stehouwer, 2005; Fuchs, Mock, Morgan & Young, 2003; Gersten et al., 2009; Hill, King, Lemons, & Partanen, 2012; Jimerson, Burns & VanDer Heyden, 2007; Nelson, Benner, & Gonzalez, 2003; Wanzek & Vaughn, 2007).

Despite the promise of RTI, and the hard work of many research and school teams, there remains ongoing concern about the limits of the current evidence base to guide RTI implementation. This concern begs the question: With the transition from an IQ-Achievement discrepancy approach to an RTI based approach, have the field simply transitioned from a model with known psychometric problems to a new system whose psychometric qualities are unknown and potentially problematic as well? Is the reality of RTI, as implemented in practice, potentially also a wait-to-fail model? Or are there inherent challenges in identifying the construct of reading disabilities regardless of the manner of identification? For example, how do we ensure that other exclusionary criteria are not responsible for a lack of response (e.g., language minority status, a lack of opportunity to learn due to poor or inadequate instruction or due to intervention implemented without fidelity)? The collection of articles in this special issue addresses several critical questions.

What are the characteristics of inadequate responders and to what degree has research incorporated the five essential components of RTI?

Lam and McMaster (this issue) conducted a review of the literature updating research on predicting RTI. They describe how responsiveness to Tier 2 and Tier 3 has been evaluated in the field and underscore the ongoing lack of consensus about a unified definition of RTI. Current practice still includes single measure criteria, multiple measure benchmarks, and benchmark and slope approaches. Furthermore, despite the importance of a well-implemented Tier 1 as the foundation for prevention, they report that to date in multi-tier research, Tier 1 has rarely been addressed to ensure its effectiveness. Furthermore, it is concerning that there remains so little research that has included all three tiers. Lam and McMaster found only two studies that have allowed fluent movement up or down tiers within a study year. They conclude by providing important directions for future research including improved screening, with the potential to fast-track students with the weakest initial skills to Tier 3, and the need to further understand predictors of response to Tier 3.

Within well-implemented Tier 1, how much time do beginning readers who are at risk for reading disability typically spend actively engaged in reading print?

Wanzek et al.'s (this issue) study addressed these questions in the context of a larger study that had successfully trained kindergarten teachers to individualize Tier 1 instruction, However, the results indicate that at-risk kindergarten students spent very little time actively engaged in print reading, and differences in amount of active print reading predicted year-end reading performance.

How stable and reliable are RTI-based classification decisions about the presence of reading disability?

Spencer et al's (this issue) study replicated existing findings of poor stability of classification decisions when evidence about the presence of reading disability is limited primarily to a single criterion. This was true whether the criterion is based on a traditional IQ-Achievement discrepancy or inadequate response to intervention. However, promising results were found for a hybrid model that combines RTI data with other assessment information.

Are there differences in patterns of strengths and weaknesses between adequate and inadequate responders to Tier 2?

Toste et al. (this issue) used profile analysis to compare responders and nonresponders to Tier 2 small-group intervention. They found both overall mean differences but also differences in the pattern of strengths and weaknesses between responders and nonresponders. Their results have implications for speeding up the provision of services for students who are likely to be unresponsive even to high-quality, Tier 2 intervention.

What are the strongest predictors of inadequate response to multi-tier RTI that includes Tier 3?

Greulich et al. (this issue) focused on inadequate responders from a nomothetic and quantitative perspective by examining a large set of predictors. They used a cut point of below a standard score of 90. In addition, they provide a more ideographic perspective by qualitatively exploring other potential emotional and behavioral characteristics displayed by inadequate responders during intervention.

Given a relative paucity of information about RTI for middle school, what issues face implementation for this population?

Solis and colleagues provide a great deal of information about issues that must be faced in implementing a multi-tier intervention framework with middle-school students with reading disability and/or poor reading comprehension. The results indicate that remediation at the middle-school level is more difficult than at the elementary-school level. However, when students are given from 1 to 3 years of intensive intervention, depending upon their needs and responsiveness to intervention, their outcomes are better than those of a business-as-usual control group.

To conclude, despite the vision that through RTI, children with reading disabilities would not have to wait for reading support until they failed, it is possible that Tier 2 constitutes another type of wait to fail, and we still lack agreement about defining responsiveness (or inadequate response), about how long and whether all students should receive Tier 2, and what interventions can close the gap in Tier 3. This is disappointing because we envisioned that student progress monitoring data would be used to fluidly move students up or down tiers of intervention and that special educators would have access to student progress monitoring reports during prevention, which would help them design more effective intensive interventions. Furthermore, there is woefully little research beyond second grade.

Now with the use of RTI models, more than ever, we understand how hard it may be to help schools develop and sustain sufficiently powerful interventions. We remain mindful that once students fall behind in reading, little evidence exists that even the most powerful remedial interventions make it possible to “close the gap,” particularly in terms of fluency and comprehension, even in studies that have come close to doing so in terms of word reading. This is troubling because we have known for a long while about negative consequences for early difficulties in reading that include relatively weak vocabulary growth (Cunningham & Stanovich, 1998), changes in attitude and motivation for reading (Wigfield & Guthrie, 1997), and loss of opportunities for developing increasingly sophisticated reading strategies (Brown, Palinscar, & Purcell, 1986).

One theme that emerges from the set of articles, corresponding to an area of needed research, is the absence of knowledge at critical interfaces in the RTI system. First, there is little knowledge about the most productive interface between the three tiers of intervention in an RTI model. For example, does coordination and synergy among levels enhance the effectiveness of tier 3 interventions for students with identified reading disabilities? Are there advantages to more flexible movement among tiers? Second, what is the most productive interface between data generated by RTI implementation and more traditional sources of data? For example, can traditional assessments be used following poor scores on a universal screening measure to identify students who would profit from moving immediately to the most intensive intervention? Is there a combination of data from RTI implementation and from traditional assessments that might result in more reliable and valid identification of students with reading disabilities?

In summary, the promise of RTI as a prevention model in the elementary grades has largely been supported, but it is clear that many children will need sustained intervention, and there remains a lack of substantially robust Tier 3 and special education interventions in place that will close the gap between students with reading disabilities and their peers. Moreover, schools are facing budgetary cuts that limit resources to provide these interventions. Additional research is sorely needed to document the promise of RTI in the later elementary, middle, and high school years. There remain challenges in using RTI as an identification model that are associated both with measurement error associated with measures used to operationalize reading disability identification and with cut-points used to define response. We hope this set of articles helps to clarify what we know and what we still need to know to efficaciously implement RTI in classrooms and to move us closer to achieving its promise.