Response-to-Intervention: Separating the Rhetoric of Self- Congratulation From the Reality of Specific Learning Disability Identification
By Kavale, Kenneth A Kauffman, James M; Bachmeier, Randy J; LeFever, Gretchen B
Abstract. The policies underlying the specific learning disability (SLD) regulations in the 2004 reauthorization of the Individuals with Disabilities Education Act (IDEA) are analyzed. The analysis focuses on the Response-to-Intervention (RTI) provision (“if the child responds to scientific, research-based intervention”) as a diagnostic model, revealing that it is conceptually flawed, practically inadequate, and politically rather than scientifically motivated. It is argued that RTI is best described as a model for providing remedial reading (prereferral) services for students experiencing early reading failure. To achieve a reliable and valid diagnosis of SLD under IDEA, it is argued that a comprehensive psychometric assessment is required. With a model that combines RTI and cognitive assessments, it is possible to provide an identification process that closely aligns with the best current conceptualizations of SLD.
The field of special education is facing a significant policy debate about the identification procedures used for specific learning disability (SLD). The debate surrounds a process termed Response-to-intervention (RTI), which is a “comprehensive assessment and intervention process utilizing a problem-solving framework to identify and address student academic difficulties using effective, efficient researched-based instruction” (Cortiella, 2006, p. 2). Although proposed in the IDEA provisions for SLD identification, implementation of RTI appears to have transcended the boundaries of special education to include significant implications for general education. Consequently, perhaps up to one third of students in the public schools will be directly affected by RTI, with more indirect implications for the remaining two thirds.
Kavale and Forness (2000) demonstrated how inclusion escalated into debate, not only about students with disabilities, but about the education of all students. Kavale and Forness also demonstrated that support for inclusion was primarily ideological and political rather than strictly empirical. As such, the debate becomes what Sowell (2002) termed a “conflict of visions” – different views of how the world works that set the agenda for both thought and action. Sowell suggested that a vision “is what we sense or feel before we have constructed any systematic reasoning that could be called a theory … a vision is a sense of causation” (pp. 4-6).
According to Sowell (1995), at times, one vision may come to predominate over others to such an extent that it represents the prevailing vision for that time and place, usually because of the presumed authority position of an elite intelligentsia. This is termed the “vision of the anointed,” while all others are said to hold the “vision of the benighted.” Because of its ascendancy, the vision of the anointed is preferred in formulating policy, while the vision of the benighted is to be ignored because it is “at best ‘perceptions,’ more often ‘stereotypes,’ and more bluntly ‘false consciousness’” (p. 187).
Particularly for special education, the vision of the anointed is to be preferred because it also includes assumptions about compassion and caring, which are viewed as their special province. Thus, the vision of the anointed is “not merely factually correct but morally on a higher plane” (p. 23). In contrast, the vision of the benighted should not be considered because “the benighted are to be made ‘aware,’ to have their ‘consciousness raised,’ and the wistful hope is held out that they will ‘grow’” (p. 3).
Kavale and Forness (2000) demonstrated how the inclusion debate was dominated by the vision of the anointed. Beginning with the least restrictive environment articulated in PL 94-142, the Regular Education Initiative where education was viewed as a “shared responsibility” (Will, 1986), through “full inclusion” (Gartner & Lysky, 1989), the vision of the anointed dominated position statements of professional organizations (e.g., Association of Persons with Severe Handicaps, 1992; Council of Chief State School Officers, 1992; National Association of State Boards of Education, 1992). Hence, inclusion was the preferred policy because it was viewed as a “good thing” (just, essential, democratic, liberating), even though “‘reality’ painted a less sanguine picture about the general education classroom being the sole placement option for educating students with disabilities” (Kavale & Mostert, 2003, p. 203).
The RTI debate is also predicated on it being a “good thing.” As Sowell (1995) suggested, the vision of the anointed “is a vision of differential rectitude [where] problems exist because others are not as wise or as virtuous as the anointed” (p. 5). Consequently, any different rendering of reality must be dismissed as either uninformed or irresponsible, and superseded by the vision of the anointed, usually through the power of government.
The purpose of this paper is to analyze the RTI debate as a conflict of visions. Proponents of RTI will be assumed to hold the vision of the anointed, while the authors will represent the vision of the benighted.
RESPONSE-TO-INTERVENTION: THE WHAT AND HOW
The foundation of the vision of the anointed surrounds RTI, the core concepts of which include the systematic (a) application of scientific, research-based interventions in general education settings, (b) measurement of a student’s response to those interventions, and (c) use of data to inform instruction (Mellard, 2004). Basically, RTI serves an accountability function to answer the question: Does instruction result in increased learning and acceptable progress?
The RTI concept grew out of concerns expressed about SLD identification (e.g., misidentification, over-identification, problematic criteria) (Bradley, Danielson, & Hallahan, 2002; Donovan & Cross, 2002; Finn, Rotherham, & Hokanson, 2001; Learning Disabilities Roundtable, 2005; Pasternak, 2002; President’s Commission on Excellence in Special Education, 2002). The primary concern focused on problems with the discrepancy model, which led to a “paradigm shift” (i.e., a new vision) (Reschly, 2004) centering on the concept of treatment validity, whereby it is possible “to simultaneously inform, foster and document the necessity for and effectiveness of special treatment” (L. Fuchs & D. Fuchs, 1998). With respect to SLD classification, Berninger and Abbot (1994) explored
going beyond defining learning disability as a discrepancy between achievement and ability (based on static, one-shot assessment of either IQ or listening comprehension) to a broader view of learning disability based on dynamic assessment failure to respond, over time, to validated intervention protocols. (p. 165)
Gresham (2002) reinforced this view at the LD Summit by suggesting that “children who fail to respond to empirically validated treatments implemented with integrity might be identified as LD” (p. 499). The Learning Disabilities Roundtable (2005) affirmed the importance of RTI and recommended that IDEA regulations “require that the essential core concepts of a responsiveness-to- intervention process have been employed.” Accordingly, the reauthorization of IDEA 2004 (PL 108-446) indicated that: “a local educational agency may use a process that determines if the child responds to scientific research-based intervention as a part of the evaluation procedures.” Thus, the reauthorization of IDEA in 2004 described RTI as an intervention that provides (a) data for more effective and early identification of students with SLD, and (b) a systematic way to ensure that students experiencing educational difficulties receive more timely and effective support (Boardman & Vaughn, 2007; National Joint Committee on Learning Disabilities [NJCLD], 2005).
To aid implementation of RTI procedures for SLD identification, the National Research Center on Learning Disabilities (NRCLD) was established (e.g., Deshler, Mellard, Tollefson, & Byrd, 2005; D. Fuchs, Deshler, & Reschly, 2004; Mellard, Byrd, Johnson, Tollefson, & Boesche, 2004). The NRCLD was charged with addressing the following questions: (a) How is RTI used in the process of SLD identification? and (b) Does RTI enhance SLD identification? Another stipulated purpose of RTI was to “redefine” SLD as inadequate response to instruction that would permit (a) identification of students using a risk rather than a deficit model; (b) early identification and instruction of students with SLD; (c) reduction of identification bias (i.e., higher probability that identified students are truly those with greatest academic needs); and (d) connecting identification assessment with instructional planning and progress monitoring (Vaughn & L. Fuchs, 2003, pp. 139-141).
Although originally focused on SLD identification as outlined in IDEA 2004, RTI was soon viewed as a means whereby “schools don’t wait for formal identification of a learning disability, but instead start providing targeted interventions early on” (West Ed, 2006, p. 1). The School Social Work Association of America (SSWAA, 2006) expanded the scope of RTI by terming it “a systematic, multi-tiered approach to helping all [italics added] students achieve school success” (p. 1). The National Association of School Psychologists endorsed this view of the RTI process by indicating that it is a “provision of scientific research-based instruction and interventions in general education [that] provides an improved process and structure for school teams in designing, implementing, and evaluating educational interventions [that may be] part [italics added] of the evaluation procedures for special education eligibility” (Klotz & Canter, 2006, pp. 1-2). Finally, a group of 13 national organizations (Collaborative Project, 2006) issued a report entitled “New roles in response to intervention: Creating success for schools and children” (2006), wherein RTI was described as follows, “To meet the needs of all students, the educational system must use its collective resources to intervene early and provide appropriate interventions and supports to prevent learning and behavioral problems from becoming larger issues” (p. 2). When the vision of the anointed is examined, it appears that the primary intent of RTI is not perceived uniformly. From one perspective, the purpose of the NRCLD was to provide “assistance on issues in the area of identification and assessment of children with learning disabilities” (Deshler et al., 2005, p. 484). Bradley, Davidson, and Doolittle (2005) endorsed this view by indicating that the Learning Disabilities Initiative “has been focused on a more efficient and effective process for determining specific learning disability eligibility” (p. 485). Yet, statements from major organizations suggest another perspective, whereby SLD identification is not viewed as a major purpose of RTI. For example, in a joint paper by the National Association of State Directors of Special Education (NASDSE) and the Council of Administrators of Special Education (case) (2006), the stated goal of RTI is “to engage the general education community in conversations and strategies to provide knowledge and technical assistance to help implement this successful approach [i.e., RTI] to teaching all children, including students with disabilities” (p. 1). The focus shifted to general education where the NASDSE/case paper represents “a call from the special education community to the general education community to join together to commit to a uniform system of education, where RTI plays a key role in identifying and working with struggling learners in any setting …” (p. 2). In fact, NASDSE (2006) clearly stated that, “Special education eligibility decisions can be a product of these efforts, but is not the primary goal” (p. 1).
The vision of the anointed appears not to be uniform with respect to the goal of RTI. On one hand, the goal of RTI is to deliver evidence-based interventions and use students’ response to determine educational needs (e.g., NASDSE, 2006) while, on the other hand, the goal is to redefine SLD through operationalizing RTI procedures (Vaughn & L. Fuchs, 2003). With different views about the purpose of RTI, emerging models may differ significantly in form and function and thus may not be compatible. Will agencies have to choose the purpose they prefer? If the delivery of instruction is primary, how will SLD be identified? If SLD identification is primary, then why include all children in RTI, even students with disabilities, who presumably have already been identified?
We believe that enhanced SLD identification is not a goal of the RTI anointed. Support for this view is found in statements noting the presumed advantages of RTI: “Most significant is that the focus shifts from eligibility to concerns about providing effective instruction” (Fletcher, Coulter, Reschly, & Vaughn, 2004, p. 311; see also Boardman & Vaughn, 2007). Although the goal of providing effective instruction is meritorious, we believe the real goal of RTI is to foster greater collaboration with general education and essentially create a unified system without an independent special education (see Batsche et al., 2005). Hollenbeck (2007) stressed the general education aspect of RTI: “With its emphasis on improving general education for all learners, RTI has the potential to positively impact students across educational settings” (p. 144). Fletcher et al. (2004) endorsed the general education emphasis in RTI by suggesting that RTI “approaches facilitate the integration of general and special education around instruction, line up IDEA with the laudatory goals of NCLB, and lead to federal regulations and conceptual models of LD consistent with our best research about teaching and learning” (p. 327). IDEA appears to be referenced only because RTI must be perceived as a special education initiative in which the enduring problems associated with SLD identification provide convenient cover for deforming the SLD category and altering long-standing special education roles and responsibilities.
FEDERAL LEGISLATION AND RESPONSE-TO-INTERVENTION
We believe that the impetus for promoting RTI is not found in IDEA but rather in No Child Left Behind (NCLB) (2001) legislation (PL 107-110). The NCLB foundation provides the reason for general education embracing RTI when historically many special education initiatives (e.g., inclusion) have not been so readily embraced (see Kauffman & Hallahan, 2005). Implementation of NCLB implicitly assumes that all children will succeed, or that no child will be left behind, which will require progress levels that many students simply will not attain. Although commendable in its goal, NCLB, because it ignores important realities, seems doomed to failure and may do more harm than good (Kauffman, 2004, 2005; Kauffman & Konold, 2007).
Inevitably, schools must deal with less proficient students. NCLB legislation mandates that schools failing to make adequate yearly progress (AYP) must provide students the opportunity to attend a better performing school. If this change is not successful and subsequent AYP standards are not met, increasingly severe sanctions may be leveled. Kauffman and Konold (2007) commented on the counterproductive nature of removing resources from students who fail to meet AYP goals:
These schools and the students that comprise them are placed in even greater danger to fail, thereby contributing to a widening achievement gap, rather than a narrowing of it. Moreover, the ‘successful’ schools to which students are transferred from ‘failing’ schools will inevitably encounter the very same statistical realities and will predictably, fail in future years. The very policy itself is internally destructive to the goals it has laid out. (p. 79)
The RTI process ostensibly provides a means to achieve the unrealistic requirement that all students achieve a minimum standard regardless of their inherent limitations: “RTI is partly a reflection of a greater commitment to the philosophical ideal that all children can learn” (Cruey, 2006, p. 1). Thus, the vision of the anointed seems to endorse instructing less proficient students in general education in order to be consistent with NCLB mandates, not IDEA regulations (e.g., Barnett, Daly, Jones, & Lentz, 2004; Fletcher et al., 2004; D. Fuchs & L. Fuchs, 2006).
For example, the claim by Sternberg and Grigorenko (2002) that “by excluding from remediation those poor readers who do not show IQ- based discrepancies, we effectively take away their only chance to become better readers” (p. 80) appears logically flawed. First, the word “remediation” implies a cause of failure other than disability (e.g., inadequate instruction), which places both the origin of the problem and its resolution within the realm of general education, not special education. Second, it is assumed (falsely) that changing the operational definition of SLD is the “only” way to help non- discrepant struggling readers, which obviously it is not. Rather, if the purpose of RTI is “to identify a subset of children at risk for poor outcomes due to their unresponsiveness” (Vaughn & L. Fuchs, 2003, p. 138), we have a general education activity that will be funded with special education monies since IDEA permits an LEA to allocate 15% of its federal funding to develop and implement early intervention services (i.e., RTI). Without the promised “full funding” for special education, the diverted monies mean that special education funds will be used for non-special educational activities. How is lost funding a benefit for students truly needing or already receiving special education services?
Since 1975, special education policy has been guided by IDEA but, upon examination, its provisions appear incompatible with the objectives of NCLB. Johns (2004) described it as follows.
It is most difficult over the long term to be both ‘equal’ and ‘unequal’ at the same time. IDEA allowed (even demanded) unequal treatment. It demanded individualization – not one size fits all. NCLB demands equal treatment with once-a-year tests in reading and math as the measuring instruments. IDEA focuses entirely on the individual. NCLB focuses entirely on the group. (p. 89)
The lack of individualization in both instruction and assessment means that the essence of special education is lost, and illustrates the danger in trying to create a single system where NCLB gains priority over IDEA.
PROFESSIONAL REPORTS AND THEIR INFLUENCE ON RESPONSE-TO- INTERVENTION
The President’s Commission on Excellence in Special Education and RTI
A number of reports (e.g., Learning Disabilities Summit, President’s Commission on Excellence in Education) stressed the problems associated with the discrepancy criterion and focused attention on RTI as a viable alternative. The RTI anointed point out that their views are based on “participation in three recent consensus reports in special education, each of which addresses the identification of SLD. The reports represent diverse groups of researchers, policy makers, practitioners, and advocates …” (Fletcher & Reschly, 2005, p. 11). We believe the reports disguised the real purpose of RTI and represent examples of how the vision of the anointed is assumed to be both factually correct and on a higher moral plane. Thus, the reports appear to provide a false rectitude because “they reveal more about foregone conclusion and ideological dogma than evidence-based recommendation” (Kavale, Kaufman, Naglieri, & Hale, 2005, p. 23). The vision of the anointed refers to the President’s Commission on Excellence in Special Education (PCESE) (2002) as a guiding report, but this effort has been criticized (Kauffman, 2004; Kauffman & Wiley, 2004). The PCESE attempted to reconcile IDEA and NCLB by calling for a school climate in which accountability is primary. The PCESE emphasized the need to close the achievement gap, which may mean either (a) the gap between the achievement of every student in special education and the achievement of the typical student in general education, or (b) the gap between the average achievement of students in special education and the average achievement of students in general education. These distinctions are without merit because, “Either way, there is a huge chasm between the expressed desire and the nature of disability. The gap to which the PCESE refers is simply not closable for reasons obvious to anyone with a rudimentary understanding of statistics and disability” (Kauffman, 2004, p. 309).
Kauffman and Wiley (2004) also questioned the PCESE recommendations regarding identification of students with disabilities and prevention. The PCESE (2000) report concluded that present identification methods lack validity and, that consequently misidentification is rampant: “Many children who are placed in special education are essentially instructional casualties and not students with disabilities” (p. 26).
Misidentification may be avoided if special education were to adopt a model based on prevention; that is, identification precedes failure. Although a noble goal, it may not be possible to achieve without some basis for anticipating failure: Who are the children we should identify in order to prevent disability? Problems associated with identification (e.g., “false positives”) are not resolved, and may be made worse because “a prevention model increases false positives – always, inevitably, with mathematical certainty” (Kauffman, 2004, p. 313; see also Kauffman & Konold, 2007). Thus, the two ideas, that is, assertions about misidentifications and adoption of a prevention model, “are logical contradictions. The logical conclusion is that the PCESE hasn’t thought through its findings or its recommendations…” (p. 314). In fact, it is possible that the earlier the prevention efforts take place, the less precise the identification process becomes, which creates the potential for yielding even more false positives (Gersten & Dimino, 2006).
In concluding their analyses, Kauffman and Wiley (2004) suggested that the PCESE “systematically devalues what special education is and what special education does” (p. 5). We, therefore, see no reason to be impressed by the PCESE and its support for RTI. In evaluating the validity of PCESE recommendations, Kauffman (2004) warned that, “The truth of a statement is not determined by who makes it” (p. 308). Although the anointed (e.g., Fletcher & Reschly, 2005) influenced the PCESE, we side with those who question the wisdom of the PCESE report.
The Summit on Learning Disabilities and RTI
The Summit on Learning Disabilities sponsored by the Office of Special Education Programs (see Bradley et al., 2002) is included in the vision of the anointed to support recommendations about RTI for SLD identification. The title of the Summit report alludes to an enduring problem surrounding terminology. In IDEA, the category defined is “specific learning disability,” connoting a singular and particular condition. Yet, the vision of the anointed continually refers to the plural form “learning disabilities,” a designation that moves the condition from the specific to the general. As SLD becomes increasingly generalized, the construct tends to lose meaning until a point is reached where the term “learning disabilities” may or may not include SLD.
The vision of the anointed assumes that there are many “learning disabilities” because of the problems created by the heterogeneity found in the SLD population (Kavale & Forness, 1995). In reality, the “problem” of heterogeneity is also found in other high- incidence, mild disability groups (e.g., MR and EBD), and should be taken to simply mean that no two students with SLD are alike. The fiction surrounding heterogeneity is necessary in order for the special education category of SLD to include students who may not be SLD.
Consequently, the vision of the anointed seems misinformed by suggesting that, “LD is rarely conceptualized as a single disorder, but instead is represented as a general category composed of disabilities in any one or combination of several academic domains” (Fletcher et al., 2002, p. 206). This misleading statement is readily negated if SLD is properly conceptualized as a single condition likely to manifest a variety of symptoms. The very designation “specific” indicates that SLD is not a general category, but for the vision of the anointed to be achieved, SLD needs to be transformed into a generic and ill-defined entity that may include any form of academic failure under the SLD label (Kavale & Forness, 1998).
Because of the way RTI has focused almost exclusively on reading achievement, the SLD concept has essentially morphed into reading disability (RD). The problem is that reading difficulties represent only one potential symptom, or indicator, of SLD. Although a large proportion of students with SLD manifest reading difficulties, there is little justification for focusing identification efforts solely on assessment of reading progress. According to Occam’s Razor (“No more things should be presumed to exist than are absolutely necessary”), discussion about RD should not be a primary factor in deliberations about SLD status. Although RD itself is a legitimate concept, it is not what makes SLD what it is. The presence of RD does not represent the essence of SLD, which is better articulated in, for example, constructs like unexpected school failure and processing deficits (Kavale & Forness, 2000). If SLD and RD are viewed as equivalent, only one designation would be necessary. The emphasis on reading in the vision of the anointed suggests that RD is the favored concept, but then the practical problem arises about how to provide special education services for students with RD, which is not a category defined in IDEA.
Because of problems surrounding definition, the reality of SLD has often been called into question by terming it a “myth” (McKnight, 1982) or “imaginary disease” (Finlan, 1993). With the integrity of SLD undermined, it becomes possible to blur distinctions between SLD and RD, thus providing students with RD convenient entry to special education. In the vision of the anointed, SLD and RD appear interchangeable, as evidenced in a paper entitled “Identifying reading disabilities [italics added] by responsiveness-to-instruction: Specifying measures and criteria” (D. Fuchs, L. Fuchs, & Compton, 2004), whose abstract begins: “We describe two types of assessment (problem solving and standard treatment protocol) within a ‘responsiveness-to-instruction’ framework to identify learning disabilities” [italics added] (p. 216). Thus, the vision of the anointed appears to view SLD and RD as equivalent, but such a counterfactual conception is inconvenienced by a simple actuality: RD is not equivalent to SLD (Kavale & Forness, 1995).
The LD Summit report included several consensus statements, some of which are important for discussions about RTI. One statement concerned the SLD concept and indicated that, “Strong converging evidence supports the validity of the concept of specific learning disabilities (SLD)” (Bradley et al., 2002, p. 792). Although a seemingly positive statement, it illustrates the tendency to use the plural (“disabilities”) when referring to what should be the singular (“disability”). IDEA clearly defines “specific learning disability,” but the persistent use of the plural, although linguistically easier, moves the construct conceptually in a direction that makes it less precise and open to varying interpretations. The use of the plural form appears to support the view that meaning resides, not in language itself, but in individuals who interpret language based upon their own unique perspective (i.e., vision of the anointed) (Grenz, 1996). Consequently, when the vision of the anointed supports the use of the term SLD “to emphasize the difference between children with SLD and those with general learning difficulties” (p. 793), it seems disingenuous since the vision of the anointed moves the SLD construct away from its original conceptualization in order to accommodate a new class of students who possess primarily reading problems but are not otherwise SLD in any significant sense. In essence, “SLD will be SLD in name only. … The RTI process would move SLD in a direction that further removes it from any of its conceptual foundations” (Batsche, Kavale, & Kovaleski, 2006, pp. 24- 26).
Although attempts to obscure categorical distinctions would be denied in the vision of the anointed (e.g., Fletcher et al., 2004; D. Fuchs & L. Fuchs, 2006), Mastropieri and Scruggs (2005) contended that,
If elimination of the category of [S]LD is sought, then this specifically should be the topic of discussion. If it is not, then discussion is needed that demonstrates how RTI identification procedures will preserve the category of [S]LD while improving identification of students with [S]LD. (p. 529)
A major focus of the LD Summit was on the ability-achievement discrepancy, the primary criterion for SLD identification since 1977. The vision of the anointed included a polemical campaign to eliminate discrepancy from SLD identification procedures (e.g., Aaron, 1997; Stanovich, 2005; Sternberg & Grigorenko, 2002). One LD Summit paper, however, concluded that ability-achievement discrepancy was a legitimate construct that was psychometrically defensible (Kavale, 2002). As the operational definition of underachievement, discrepancy represents a critical marker for SLD determination that permits its differentiation from low achievement (see Kavale & Forness, 2000). The vision of the anointed included a consensus statement indicating that, “IQ/achievement discrepancy is neither necessary nor sufficient for identifying individuals with SLD” (Bradley et al., 2002, p. 796). However, it was accompanied by a minority statement indicating that, “Aptitude/achievement discrepancy is an appropriate marker of SLD” (p. 796). Consequently, discrepancy was not eliminated in the reauthorized IDEA but modified to indicate that “a local educational agency shall not be required to take into consideration whether a child has a severe discrepancy between achievement and intellectual ability” (PL 108-446, Sec. 614 (b)(6)). With ability-achievement discrepancy not eliminated, the vision of the anointed incorporated contorted forms of discrepancy in order to maintain the concept but alter its traditional meaning (Fletcher, Denton, & Francis, 2005). For example, Peterson and Shinn (2002) described an “absolute achievement discrepancy” defined by severe low achievement “operationalized as a discrepancy between average national achievement; e.g., 50th percentile and actual student achievement on commercially available norm-referenced achievement tests” (p. 462), and a “relative achievement discrepancy,” defined by a severe achievement discrepancy from a local achievement standard (p. 463). In another iteration, L. Fuchs (2003) described a “dual discrepancy” model that assessed both intervention efficacy and progress in reading growth. Children who are below peers in both intervention outcome level and slope of reading progress would be considered SLD (e.g., Burns & Senesac, 2005; D. Fuchs, L. Fuchs, McMaster, & Al Otaiba, 2003; Speece, Case, & Molloy, 2003). In reality, however, the intended meaning and purpose of the ability-achievement discrepancy criterion is not captured in any of these alternative forms (Kavale, 2002).
In evaluating the potential of RTI, the LD Summit indicated that, “Response to quality intervention is the most promising method of alternative identification and can promote effective practices in schools and help to close the gap between identification and treatment” (Bradley et al., 2002, p. 798). Descriptions of RTI emphasize its instructional aspects but, at the same time, indicate that RTI represents “one way of conceptualizing learning disabilities (LD) is to apply research-validated interventions and then identify the small subset of children who do not respond as having LD” (L. Fuchs, 2003, p. 172).
D. Fuchs and L. Fuchs (2005) illustrated the RTI identification process by presenting four case studies. After describing reading instruction at Tier 1 and Tier 2, and the measure (curriculum-based measurement – word identification fluency), one student was not at risk, two students were at risk but responsive, while one student was unresponsive and deemed as having a disability. After excluding MR and EBD with brief screening measures, the student was classified as SLD. However, questions arise about the validity of the SLD classification: What essential markers of SLD were evaluated? What is the basis for SLD classification? How was RTI diagnostic of anything other than early reading failure? As suggested by Kavale (2005),
In reality, there is [no basis for SLD classification] unless there is some legerdemain in which all RD magically transforms itself into SLD. The real problem with the RTI model lies not in the procedures themselves but rather in the leap of faith necessary for unresponsiveness to become SLD. (p. 559)
The vision of the anointed includes a favorable view of RTI because the primary concern appears not to be with valid SLD identification but to ensure that (a) students receive good instruction (something that all of us would like to see), (b) students who are not performing as desired get extra instruction immediately (something that no one can argue against), and (c) the SLD category becomes the convenient home for those who otherwise might be left behind (something that perverts the category and is consistent only with the fantasy world of NCLB). Two out of three may seem like a convincing case, but as Kauffman and Konold (2007) argued, only one proposition not grounded in reality is sufficient to undermine an otherwise good policy.
THE ROLE OF RTI IN THE IDENTIFICATION PROCESS
Defining Response-to-intervention and Specific Learning Disability
If the vision of the anointed was really focused on enhancing SLD identification, attempts to resolve the long-standing problem of SLD definition would have been attempted (see Kavale & Forness, 2000). The number of SLD definitions proposed over the past decades suggests that no single effort has achieved consensus status. Consequently, the definition first proposed in 1968 and codified in 1975 has been maintained almost unchanged in federal legislation.
In commenting on the definition of SLD, Willis and Dumont (2006) asked, “Has Congress provided us with a better definition leading to more appropriate diagnosis of SLD, or has the definition just gotten worse?” (p. 907). With no substantive change in the SLD definition, neither view seems correct, and it appears that the SLD definition will continue to fail to provide “two critical elements: understanding – a clear and unobscured sense of SLD – and explanation – a rational exposition of the reasons why a particular student is SLD” (Kavale & Forness, 2000, p. 240). As suggested by Kavale (2005), “The present SLD definition is too broad to be wrong and too vague to be complete” (p. 553). Consequently, it seems curious that the LD Summit did not provide a rationale for either modifying or maintaining the present contentious definition while, at the same time, sanctioning the SLD concept. What SLD construct was there consensus about at the LD Summit?
Technically, it seems incorrect to suggest that RTI may be the basis for “redefining learning disabilities” (see Vaughn & L. Fuchs, 2003). To be used in practice, a formal definition must be translated into an operational definition: rules stipulating how a concept is to apply in a particular case if specified actions yield specified characteristic results (Bergmann, 1961). In a scientific sense, an operational definition possesses a logical and rational relationship to elements described in the formal definition. Otherwise, the risk of proposing spurious operations is substantial. Consider the example provided by Kavale and Forness (2000) of the Learning Disability Coefficient (LDC) operationally defined as white blood cell count multiplied by body weight divided by hat size. The LDC is absurd because,
it would possess little meaning or significance because a good deal is known about LD, and the LDC does not ‘fit’ with any of the existing knowledge. Thus, it is possible to operationalize anything, but whether it ‘makes sense’ is a different matter. (p. 247)
For SLD, however, the process of developing an operational definition appears to have been neither logically nor rationally sufficient to maintain scientific integrity. For example, the federal SLD definition contains no reference to underachievement, which would have provided a valid basis for including its operational definition of discrepancy. For a complex phenomenon like SLD, it seems unlikely that any single element (e.g., discrepancy) can capture that complexity (Lloyd, 2002). Thus, statements suggesting that SLD identification “hinges on an implicit classification of underachievers into those with LD and those with other forms of low achievement” (Francis et al., 2005, p. 103) demonstrate significant confusion: SLD is the only defined special education category that includes the concept of underachievement.
The replacement of discrepancy with RTI as the operational definition of SLD will not resolve the problem of basing a classification on a single indicator not specifically articulated in the definition. Consequently, it is difficult to discern how the vision of the anointed can suggest that, “the deceptive ease of simple univariate (i.e., low achievement) and bivariate (IQ- achievement discrepancy) endpoint approaches to the identification of LD cannot be mistaken for validity or conceptual clarity” (Francis et al., 2005, p. 104), while that same vision can also suggest that “nonresponders represent a better indicator of the construct of LD” (p. 107).
In an unintended miscalculation, the vision of the anointed was correct in concluding that “IQ and achievement scores are not sufficient” (p. 95) for SLD identification. The simple fact is that discrepancy is not a proxy for SLD and represents a necessary but not sufficient criterion for valid SLD classification (Kavale, 2002). Likewise, RTI is not a proxy for SLD but, unlike discrepancy, which validates the presence or absence of an accepted construct (i.e., underachievement), RTI can only validate the self-evident fact that a student is experiencing reading problems. The NRCLD (Johnson, Mellard, D. Fuchs, & McKnight, 2006) acknowledged this view by indicating that “RTI should be considered to be one important element within the larger context of the SLD determination process. RTI as one component of SLD determination is insufficient as a sole criterion for accurately determining SLD” (p. 3).
The inadequacy of an approach focusing exclusively on a low achievement criterion is seen in the diagnostic criteria for SLD proposed by Dombrowski, Kamphaus, and Reynolds (2004) in After the demise of discrepancy: Proposed learning disabilities diagnostic criteria. Why is it inadequate? First, SLD is given a new name – developmental learning delay; this is not the way to fortify the existing SLD definition. second, the proposed diagnostic framework includes only a one standard achievement score cutoff (i.e., 85 or below) with no appraisal of the presence or absence of either underachievement or processing deficits. Essentially, this is an absolute low-achievement definition that fails to capture consequential elements of SLD. As suggested by Mather and Gregg (2006), “The Dombrowski et al. diagnostic framework does not take into account and incorporate findings from current research or historical and clinical perspectives regarding identification of individuals with LD” (p. 99). What Is Response-to-intervention?
RTI is best viewed as an instructional model, not an identification model. Consequently, it should not be the basis for SLD identification even though the vision of the anointed indicates that RTI has “the potential to be as good as traditional measures in regard to eligibility determination” (Reschly, 1988, p. 498). In reality, classroom teachers have always been aware of students experiencing difficulties in learning to read and have usually sought accommodations for those students.
Since the 1980s, support for general education interventions as an alternative to special education has been termed “prereferral,” which has been articulated in several models (e.g., school-based intervention teams, teacher assistance teams, instructional support teams) (e.g., Buck, Polloway, Smith-Thomas, & Cook, 2003; Truscott, Cohen, Sams, Sanborn, & Frank, 2005). In general, prereferral activities have produced positive outcomes (Burns & Symington, 2002), but problems have been noted surrounding limited guidance about professional roles and responsibilities, lack of validated intervention strategies, inconsistent progress monitoring, and increased costs (e.g., Buck et al., 2003; Ross, 1995; Truscott et al., 2005).
Although the mechanics of RTI as an SLD identification process remain vague, most descriptions of RTI connote resemblance to prereferral activities, whose potential benefits include “increased achievement school wide, because struggling children are identified proactively and immediate help is provided to determine whether or not more intensive help is needed, as well as a coherent and flexible system of services” (VanDerHeyden, Witt, & Barnett, 2005, p. 339). In fact, VanDerHeyden, Witt, and Gilbertson (2007) demonstrated the positive effects of an RTI model, whose purpose “was to identify early those students at-risk for academic problems [and] to enable teams to more accurately determine who should be referred for evaluation and eligibility determination” (p. 249). Thus, “The RTI model appears to be prereferral writ large with greater specification about the types of reading interventions and outcome assessments that should be used” (Kavale, 2005, p. 557), which should serve to improve consistency in implementation (see Burns, Vanderwood, & Ruby, 2005). Yet, NASDSE (2006) indicated that a vision that suggests RTI as only prereferral is mythical because, “RTI is more than prereferal [sic] services; it is a comprehensive service delivery system that requires significant changes in how a school serves all students” (p. 2). Unfortunately, those changes involve undermining special education and contorting the nature of the SLD category.
Even if viewed as a prereferral process, RTI is not seen as being related to SLD identification or special education, as suggested by NASDSE (2006): “When thought of as a prereferral system, [RTI] remains the province of special education and the desired integration of general education and special education services around the goal of enhanced outcomes for all [italics added] students will not [italics added] be achieved” (p. 2). Thus, the emphasis on all students suggests that the goal of enhanced SLD identification is a limited concern.
Could the emphasis on service for all children explain the dearth of information about RTI as an identification process? Although the special education literature is replete with discussions about instructional aspects of RTI (e.g., Denton, Vaughn, & Fletcher, 2003; Noell & Gansle, 2006; Olinghouse, Lambert, & Compton, 2006), technical aspects (e.g., measurement) of RTI (e.g., Barnett et al., 2006; D. Fuchs, L. Fuchs, & Compton, 2004; L. Fuchs, 2003), and development of RTI instructional models (e.g., Ardoin, Witt, Connell, & Koenig, 2005; McMaster, D. Fuchs, L. Fuchs, & Compton, 2005; Vaughn, Linan-Thompson, & Hickman, 2003), descriptions of how RTI functions as an identification procedure are lacking. Even the NCRLD says little about RTI as an SLD identification process (e.g., Mellard et al., 2004; Mellard, Deshler, & Barth, 2004). The Collaborative Project (2006), in “New roles in response to intervention,” although indicating that “RTI may be used as part of a process to identify students with [SLD] rather than relying on the use of a discrepancy model as a means of identification” (p. 2), offers no insight into how RTI functions as an identification procedure.
Instead, the vision of the anointed places emphasis on how “the educational system must use its collective resources to intervene early and provide appropriate interventions and supports to prevent learning and behavioral problems from becoming larger issues” (Collaborative Project, 2006, p. 2). We agree that early intervention is important, but the RTI model cannot deliver the defining feature of special education – individualized instruction. Instead, RTI delivers a “one size fits all” approach that focuses on methods for determining treatment validity and says little about selecting interventions, ensuring treatment integrity, or evaluating whether interventions are indeed effective or generalizable. As suggested by Kavale et al. (2005), “scientific, research-based intervention [in RTI] translates into Try something, anything, try to measure it well, make sure the teacher does what might or might not help, and if the child doesn’t get better, then he’s [sic] SLD’” (p. 21).
Response-to-intervention and Identification of SLD
With the vision of the anointed providing little insight into the mechanics of RTI in SLD identification, why should it be discussed as a diagnostic procedure? Although nonresponsiveness may possess “ecological validity” (see Dean, Burns, Grialov, & Varro, 2006), because RTI models “demonstrated strong effects in improving student learning and systematic variables” (Burns, Appleton, & Stehouwer, 2005, p. 389), the failure to respond is of little import for SLD identification unless it is deemed that nonresponsiveness equals SLD. Ecological validity only serves to support RTI as a prereferral activity.
The absurdity of a “diagnosis by fiat” scenario in SLD identification can be readily avoided if RTI activities are properly viewed as prereferral. With its rigorous and systematic procedures, RTI can enhance the prereferral process and achieve the aim of reducing the number of unnecessary referrals. As prereferral, RTI models can avoid having to answer a question for which few answers presently exist: What do practitioners do after the RTI process concludes that a student is a treatment resister? Although the student is deemed to have an SLD, this conclusion must be placed in the context of having neither uniform teaching techniques, assessment tools, expected outcomes, nor formal timelines, but only the assumption that the teaching has been “scientifically validated.” Consequently, RTI is not falsifiable – no one can demonstrate that RTI methods were ineffective. We can, therefore, blame any child’s failure to respond on the child because instruction is held blameless (Kavale et al., 2005).
To avoid a situation where a student is simply declared to have SLD, RTI procedures should be combined with psychometric testing. Wodrich, Spencer, and Daley (2006) provided reasons why RTI needs to be combined with psychoeducational assessments. Specifically, use of RTI alone makes it difficult to (a) distinguish SLD from mild mental retardation, (b) distinguish students with SLD from slow learners, (c) identify intra-individual differences, (d) determine the meaning of a positive RTI, and (e) identify the best means to implement effective interventions. Models that combine RTI and psychometric assessment have been described (e.g., Flanagan, Ortiz, Alfonso, & Dynda, 2006; Kavale & Flanagan, 2007), and are necessary because, “An RTI model without a comprehensive evaluation cannot identify SLD because it is not aligned with the construct of SLD” (Ofiesh, 2006, p. 887).
Cognitive processing assessment aligns diagnostic procedures with a clearly articulated SLD definitional component: a disorder in one or more of the basic psychological processes. Willis and Dumont (2006) noted that the federal government has paid scant attention to disorders of processing even though such disorders define SLD: “Practitioners have been given the regulatory definition of what a specific learning disability is, but then, in essence instructed to ignore it” (p. 907). With “intrinsic processing weaknesses” as an important component of SLD (see Torgesen, 2002), “We believe children with ‘true’ SLD have cognitive deficits and integrities in the basic psychological processes, which often lead to academic failure” (Kavale et al., 2005, p. 17). A major problem is in understanding how RTI can identify processing disorders but can be avoided if SLD identification includes procedures to “identify the specific cognitive and/or linguistic correlates that appear to be related to the identified area of underachievement or relative difficulty” (Mather & Gregg, 2006, p. 103).
The process is aided by modern conceptualizations of cognitive processing that have moved well beyond outmoded perceptual-motor theories (e.g., those of Newell Kephart or Marianne Frostig) and now provide far more dynamic views of cognitive ability than previous conceptions, which were based on a singular global ability (g). A. Kaufman and N. Kaufman (2001) detailed an array of well-validated cognitive ability tests that have incorporated advances in multifactorial intelligence theory and neuropsychological assessment practices. Although the vision of the anointed suggests intelligence tests are irrelevant (D. Fuchs & Young, 2006), the value of modern tests of cognitive processing is found in their ability both to predict academic performance (Naglieri & Bornstein, 2003), and to identify processing strengths and weaknesses that have relevance for both identification and intervention (Hale & Fiorello, 2004; Naglieri, 2003). A comprehensive assessment also offers the possibility of distinguishing students with SLD from those who are failing to achieve for other reasons (e.g., MR, EBD). Although pervasive criticism has created the perception that SLD does not exist, “LD is real and … describes problems that are distinct from other conditions subsumed under the broad category of problems in learning and achievement” (Keogh, 2005, p. 107). As Bateman (2007) noted, IDEA is built on the presumption that the disabilities listed in the law are real and can be defined and identified. Consequently, it is misleading to suggest that students with SLD are simply “victims of ‘poor teaching’ [or] that many children identified as LD are ‘teaching disabled’” (Wright & Wright, 2007, p. 1).
Without a comprehensive assessment, it would not be possible to document the presence or absence of under-achievement, the critical SLD marker operationally defined by discrepancy. Without the ability- achievement discrepancy criterion, the notion of SLD as unexpected learning failure in the presence of average or above-average cognitive ability cannot be verified. Although the vision of the anointed does not view discrepancy as a necessary element (see Speece & Shekitka, 2002), valid SLD identification demands that discrepancy remain an essential component for documenting unexpected learning failure (Keogh, 1987; Mastropieri & Scruggs, 2002). How is one to estimate what a student should be able to achieve if not by using a test of ability? Are we simply to make the obviously absurd assumption that all children have the same ability to achieve at an average level, although some do not? Or are we to rely simply on teacher estimates with all their known subjectivity and vagaries?
The RTI approach, with its nonresponsiveness criterion and absence of cognitive ability testing, essentially excludes unexpected learning failure as a parameter of SLD. In reality, efforts by the RTI anointed to attenuate the SLD category have led to a situation in which “the logical relation shifts from All students with SLD have learning problems to All students with learning problems have SLD” (Kavale, 2005, p. 554). With the shift to generalized learning problems instead of SLD, the RTI anointed are satisfied in suggesting that,
If RTI is done thoroughly and correctly there should not be a need for a comprehensive evaluation. In fact your ‘hit rate’ for students in need of special education services will be better under a data based RTI approach than if using the traditional method of discrepancy. (LD Talk, 2007, p. 10)
Although the “hit rate” for students needing special education will be much improved, the “hit rate” for SLD will not be known.
When a student does not meet the discrepancy criterion and, therefore, cannot be deemed an underachiever, there is the strong possibility that the student is a “slow learner” (SL; i.e., a student with an IQ level between about 70 and 85). About 14% of the school population may be deemed SL, but this group does not demonstrate unexpected learning failure, but rather an achievement level consonant with IQ level. Although NCLB makes such low achievement problematic, slow learner has never been a special education category, and “What should not happen is that a designation of SLD be given to a slow learner” (Kavale, 2005, p. 555).
The discrepancy criterion provides the means to differentiate SLD and SL. RTI offers no such mechanism. Consequently, RTI cannot answer a critical question posed by Mastropieri and Scruggs (2005), “If RTI cannot discriminate, how can it classify?” (p. 528). In fact, RTI cannot classify, but the vision of the anointed finds SLD “a tempting target for their goal of creating a category for children experiencing academic failure who, without additional instruction might be ‘left behind’” (Kavale et al., 2005, p. 21). “If your child’s IQ is between 76 and 90 (that’s between 15% and 20% of all kids), the Response to Intervention model probably brings them help they’re not getting now” (Cruey, 2006, p. 1). The humanistic plea for additional academic support in the vision of the anointed has led to RTI being improperly viewed “as a means of reallocating resources – away from discrepant, middle class children of dubious disability to nondiscrepant, low-socioeconomic-status, low-achieving students [with no disability]” (D. Fuchs & L. Fuchs, 2006, p. 96).
The predominance of the vision of the anointed has fostered poor diagnostic practice, resulting in the de facto inclusion of both SLD and SL individuals in the same category. Instead of continuing to subvert the SLD category, policy “should be directed at establishing SLD as the particular disability class originally intended when first introduced into special education” (Kavale et al., 2005, p. 21).
RTI: Questions and Answers
A special issue of Assessment for Effective Intervention (Burns & VanDerHeyden, 2006) was devoted to the topic “Using response to intervention to assess learning disabilities.” One article presented a dialogue where “Kovaleski poses key questions about the meaningfulness and utility of RTI as a diagnostic tool. Batsche and Kavale each respond, representing dichotomous views [i.e., a conflict of visions]” (pp. 3-4). The questions covered several topics, including basic premises of RTI, research base for RTI, implications of RTI for the SLD construct, and impact of RTI in the general-special education system. Although a majority of responses revealed disagreement, Kovaleski also noted that,
Both Batsche and Kavale appear to agree that the RTI procedures should be adopted in general education to help structure the support system for improved learning for all students [i.e., prereferral]. The difference of opinion emerged when RTI was proposed as a basis for diagnosing SLD. (Batsche et al., 2006, p. 17)
Kovaleski went on to suggest that, “the paradigmatic differences between the positions illustrated in this article are perhaps irresolvable, in that underlying the two positions are decidedly different viewpoints regarding the basic nature of SLD” (Batsche et al., 2006, p. 17).
The different viewpoints are illustrated in a response to the conversation offered by Vaughn and L. Fuchs (2006), who subtitled their paper “Why response to intervention is necessary but not sufficient for identifying students with learning disabilities [italics added].” We believe the focus should be on identifying students with specific learning disability (SLD), the category defined in the law; students with (general) learning disabilities are relatively easy to identify but those with a SLD are not.
The widespread use of the term “learning disabilities” seems predicated on a conventional wisdom that questions the validity of SLD as a distinct classification. The origins are found in research conducted by Ysseldyke, Algozzine, Shinn, and McGue (1982) who, citing significant group overlap on psychoeducational assessments, suggested that SLD had become primarily a category of low achievement (LA): “The results of this investigation raise serious concerns regarding the differential classification of poorly achieving students as either LD or non-LD” (p. 82). However, Kavale, D. Fuchs, and Scruggs (1994) demonstrated, using the Ysseldyke et al. data, that students with LD and LA could be reliably distinguished from each other, with the LD group being the lowest of the low achievers but equivalent to the LA group on ability (i.e., IQ). Consequently, SLD and LA “represent two distinct populations … defined by an ability-achievement distinction” (Kavale, 1995, p. 146). Thus, failure of the vision of the anointed to recognize SLD as a distinct classification is necessary because RTI can only identify achievement differences when, in fact, valid SLD classification requires both achievement and cognitive ability information.
Vaughn and L. Fuchs (2006) suggested that the conflict of visions may be related to what they term “connectedness”:
Batsche and other RTI proponents seem primarily concerned about RTI as a prevention mechanism, and this is where the bulk of the RTI evidence resides. By contrast, Kavale and fellow opponents seem to focus their attention on how RTI will affect the integrity of the LD classification, and this is where much less research has been conducted, (p. 60)
We agree with the above and believe it is, therefore, incumbent on RTI proponents to demonstrate how RTI can function as a diagnostic procedure for SLD as defined in IDEA, and not focus on general, non-specific learning problems. Although SLD identification continues to be a major RTI theme, in reality, identification appears to be, at best, a secondary concern while the goal of prevention seems primary.
IDENTIFICATION OF SPECIFIC LEARNING DISABILITY AND RESPONSE-TO- INTERVENTION
The dearth of research literature on RTI as an identification process suggests either that RTI is difficult to conceptualize as a diagnostic model or that there is limited interest in doing so. Both explanations seem reasonable considering the lack of research on RTI as a diagnostic process and the focus on prevention that aligns RTI with NCLB instead of IDEA. Given the present status of RTI research, and to resolve the conflict of visions surrounding RTI, we make the following recommendations: 1. Reform RTI into what it is: prereferral intervention.
2. Make RTI activities the exclusive province of general education.
3. Involve special education only after there is RTI failure, when the process shifts from prevention to identification.
4. Base identification on a comprehensive psychometric assessment that is designed to provide both diagnostic and instructional data.
5. Change regulations to (a) require use of abilityachievement discrepancy as the first (but not only) marker for SLD, (b) require use of “a process that determines if the child responds to scientific, research-based intervention” before SLD evaluation procedures, and (c) eliminate the 15% special education funding for RTI development and implementation.
The recommendations for policy change offered here provide a means whereby prevention is followed by increased confidence in the validity of SLD classification (Zach, 2005). A focus on enhanced SLD identification will eliminate potential “diagnostic chaos [in which] the number of false positives and false negatives may increase significantly because of a failure to know what a true positive should be” (Kavale, 2005, p. 560). A revised policy that incorporates the best of both visions (anointed and benighted) eliminates a polarizing either/or perspective about the value of RTI and cognitive testing where “never the twain shall meet” (Willis & Dumont, 2006, p. 901).
Scruggs and Mastropieri (2002) provided validity criteria for identification procedures that are not met when RTI is used as the exclusive identification procedure. Similarly, Keogh (2005) discussed three criteria for determining the adequacy and utility of identification procedures: (a) homogeneity, similarity of those included in a category; (b) reliability, degree of agreement when placing individuals in a category; and (c) validity, how well category membership informs treatment.
In reality, RTI can only meet these criteria when combined with a comprehensive psychometric evaluation. As Keogh (2005) suggested, cogent identification procedures “require careful and critical thinking and discussion, and a great deal of empirical work” (p. 102). Presently, RTI does not possess the requisite research base: “the research base on RTI is presently very limited, especially compared to the expansive claims being made for the procedures” (Scruggs & Mastropieri, 2006, p. 61). Fuchs and Deshler (2007) argued against accepting a “we-know-all-we-need-to-know message about RTI and then discussed the importance “of knowing what we don’t know” (p. 130). They also cautioned against characterizing “those raising questions about RTI [the benighted] as uninformed or (worse) temporizing or (much worse) attempting to obstruct wider use of RTI through passive-aggressive intellectualizing” (p. 129). Like the “ideology of full inclusion [that] influenced policy and practice disproportionately to its claims of efficacy” (Kavale & Mostert, 2003, p. 191), RTI must not transcend its ability to support recommendations about SLD identification.
Resolution of the conflict of visions would necessitate the RTI anointed relinquishing the following: exclusive focus on intervention activities, engaging in general education instructional responsibilities, endorsing a more discrete and independent status for the SLD category, and embracing cognitive ability assessment. Because their vision assumes “a special state of grace for those who believe it [and] those who disagree with the prevailing vision are seen as being not merely in error, but in sin” (Sowell, 1995, pp. 2- 3), the RTI anointed are unlikely to abandon any part of their vision, and instead conclude that, “It is unlikely that the minds of these ‘true believers’ [i.e., the benighted] will be changed in the near future” (Reschly, 2005, p. 513). Consequently, recommendations from the benighted are dismissed and “characterized as the opinion of four people with vested interests in current practices who formed what they described as an ad hoc committee that met over a weekend” (Fletcher & Reschly, 2005, p. 14). Any disagreement with the RTI anointed’s vision is viewed as “outright hostility [from] those who personally profit substantially from a traditional refer-testplace system. … The biases and vested interests of these critics should be considered in evaluating their claims” (Reschly, 2004, p. 412).
We urge similar scrutiny of the vision of the anointed for anyone concerned with the probity of special education. The problem surrounds the vision of the anointed failing to recognize that “our collective ‘vested interest’ is upon promoting the accurate identification and timely treatment of individuals with SLD” (Mather & Kaufman, 2006, p. 751). By not acknowledging the vision of the benighted, the vision of the anointed becomes perilous, “because insulation from evidence virtually guarantees a never-ending supply of policies and practices fatally independent of reality” (Sowell, 1995, p. 241). Under the circumstances, the vision of the anointed is “a badge of honor and a proclamation of identity: To affirm it is to be one of us and to oppose it is to be one of them” (p. 241).
Such a divide is unnecessary and should be replaced by a “vision of the rational” (Kavale & Forness, 2000, p. 290) that incorporates facets of both visions (anointed and benighted) into a single vision where extreme positions are avoided. The vision of the rational can create a model where “RTI has us look through a wide-lens telescope at the entire school population whereas cognitive assessments provide a microscope with a direct intensive focus on an individual’s specific needs” (Mather & Kaufman, 2006, p. 751).
As individual states respond to the RTI mandate, a rational vision creates a situation where “both RTI and cognitive assessments can serve to meet the eligibility guideline