Abstract
Many K–Grade 12 students with disabilities (SWDs) in the United States spend a large portion of their school day in general classrooms because of a prevailing view that they are more appropriately challenging and beneficial than other educational placements. We questioned this belief by exploring a “dosage” hypothesis: The more time SWD are in general classrooms the better they do academically. We assembled a database spanning 1998 to 2015, inclusive. For 9 of these years, we found both Office of Special Education Program’s placement data and National Center for Education Statistics’ reading data. We ran multilevel growth models to describe trends across time for the placement and reading data. Findings indicated a steadily increasing trend for general class placement and a positive but decelerating trend for reading performance, which together produced a widening placement–performance gap after 2007. Among 10 states/jurisdictions with the strongest positive trends for general class placement, there was no uniform pattern of reading performance across years. In short, we found little corroboration of a dosage hypothesis.
Michael Yudin
By the time Michael Yudin won Senate confirmation in 2015 as assistant secretary of the Office of Special Education and Rehabilitation Services (OSERS) in the U.S. Department of Education, he had been acting assistant secretary of OSERS (2012–2015) and deputy assistant secretary for policy and strategic initiatives in the Office of Elementary and Secondary Education (2010–2012). He was known within and outside the Department as a savvy, articulate, and passionate advocate for people with disabilities and a loyal member of former President Obama’s Administration.
Expectations and access
In countless speeches, Yudin was by turns preacher, teacher, entertainer, and friend. Someone who never tired of explaining to parents and educators why they should expect more of special-needs children and youth and imploring them to do so. But his talks were more than Knute Rockne-like half-time exhortations. He would explain why context matters; how students with disabilities (SWDs) can prosper in mainstream settings because general educators are more likely than other educators to hold expectations both challenging and affirming and to ensure that they are accessing the same curriculum as their peers. The teachers make this happen, he would say, by relying in part on innovative curricula inspired by Universal Design for Learning, which, he would assure his audiences, is supported by great research (e.g., Crosby, 2017). A frequently heard refrain was “Thirty-five years of research show that kids with disabilities do better when they are held to high standards and can access the general education curriculum” (e.g., Yudin, 2014; also see Yudin & Musgrove, 2015).
Full inclusion
As suggested by the foregoing, Yudin used his leadership position as a bully pulpit to advance “full inclusion,” a policy that calls for the full-time participation of virtually all SWDs in general classrooms and the elimination of placements in special education settings like resource rooms, self-contained classrooms, and special schools. Full inclusion has been implemented by a small number of state education agencies (SEAs) and school districts, and endorsed by professional organizations (e.g., National Association of State Boards of Education, 1992), advocacy groups (e.g., TASH; see Agran et al., 2014), and academics (e.g., Sailor & McCart, 2014; Schifter & Hehir, 2018; Skrtic, 1991; Stainback & Stainback, 1984; Taylor, 1988). Full inclusion and its more moderate cousin “inclusion,” which does not call on educators to eliminate special education placements, have arguably been among the most consequential educational policies in the last four decades. And not just in the United States. For example, the Salamanca Statement and Framework for Action on Special Needs Education (UNESCO and Ministry of Education and Science, Spain, 1994), adopted by the 1994 World Congress on Special Needs Education in Salamanca, Spain, “challeng[ed] the . . . potent and commonly held idea that children with special educational needs [around the world] do not belong in mainstream schools . . .” (Florian, 2019, p. 691; see Supplemental File, Part A, Note 1).
Because of Yudin’s high-profile government position, his public support of full inclusion had to be more muted than full-throated, more implicit than explicit—a form of self-restraint, or reserve, no doubt reflecting recognition of a generally accepted view that full inclusion likely conflicts with the free appropriate public education requirements of the Individuals with Disabilities Education Act (IDEA; 20 U.S.C. § 14112[a][9]). More specifically, this law requires SEAs and school districts to ensure that SWDs are educated in the least restrictive environment (LRE): “[T]o the maximum extent appropriate, children with disabilities . . . are educated with children who are not disabled” and can be removed from the general classroom “only when the nature or severity of [their] disability . . . is such that education in regular classes with the use of supplementary aids and services cannot be achieved satisfactorily” (IDEA 20 U.S.C. § 14112[a][5]). Since passage of the Education for All Handicapped Children’s Act of 1975, which eventually became IDEA, most SEAs and school districts have attempted to comply with this LRE provision. To guarantee that each and every SWD is appropriately placed, public schools have established a continuum of alternative placements to meet their needs (IDEA Regulations, 34 C.F.R. § 300.115)—the very continuum that advocates of full inclusion want dismantled—in response to the law and also because of the widely held belief (advocates notwithstanding) that general classrooms cannot by themselves provide all SWDs the education to which they are entitled (e.g., Commission on the Education of the Deaf, 1988; Council for Exceptional Children, 1993; Kauffman, 1989; Kauffman & Hallahan, 2005; National Joint Committee on Learning Disabilities, 1993; Scruggs & Mastropieri, 1995; Zigmond, 2003; Zigmond et al., 1995).
Whereas Yudin spoke carefully in public about the virtues of full inclusion, he and his director of the Office of Special Education Programs (OSEP), Melody Musgrove, acted boldly in other ways to promote it. Perhaps the best example of this was in 2012 when OSEP awarded US $24.5 million to the University of Kansas—the largest grant in the university’s history (Monaco, 2012), and one of the largest OSERS discretionary grants of the time—to establish the Schoolwide Integrated Framework for Transformation (SWIFT) Center to assist 64 schools in 16 districts in five states to implement a model of fully inclusive education (Sailor & McCart, 2014; also see Supplemental File, Part A, Note 2).
The claim
So, when the first author took a seat with others to hear Yudin speak at OSEP’s 2015 Leadership Conference in Washington, DC, he was aware of Yudin’s position on full inclusion. Yet, the first author was still surprised when he heard him declare repeatedly and emphatically that “the more time SWD spend in general classrooms the better they do academically.” The first author had never heard this claim, let alone from an assistant secretary of education speaking to a room full of academics. The claim was remarkable on two more counts. First, it was expressed without qualification. Yudin did not say, “The more time SWD spend in general classrooms the better they do when their instruction is a blend of grade-appropriate standards and individually tailored curricula,” or “if their instruction is supplemented by frequent tutorials conducted by experts in the science of reading.” In other words, it seemed Yudin was trying to convince his audience of a general principal or immutable law: All (or virtually all) mainstream classrooms are preferable for SWD over other instructional placements. Second, he did not provide any evidence. Rather, he staked his claim with the surety of a guy who had all necessary documentation in his back pocket.
Because of the declamatory nature of the speech, the first author, while still listening to Yudin, began wondering if the “more time” statement was spun from whole cloth. Days later, he put the question to a long-time Washington, DC, reporter who had also heard Yudin speak at the OSEP conference. The reporter laughed and said he’d already asked the same question of an OSERS staffer. The staffer eventually gave the reporter four citations, which the reporter shared with the first author. None led to documents that had anything to do with Yudin’s assertion. Sometime after Yudin left government service, the first author made indirect contact with two persons who were in the Education Department when Yudin was assistant secretary. Both were aware of his claim and unaware of a factual basis for it. Indeed, one of the two remembers being sufficiently troubled by it to express concern to someone in his orbit.
In the years following Yudin’s OSEP talk, his “more time” declaration kept playing in the first author’s mind, partly because of the provocative and unsubstantiated way he expressed it, partly because similar pronouncements by others continue to serve as a rallying cry to place all SWD in general classrooms. As but one example, consider the Consortium for Citizens with Disabilities’ (2019) statement on placement in the LRE: When LRE requirements are met, students with disabilities participate to a greater extent in . . . general education classrooms. Research overwhelmingly show[s] that providing students with disabilities an education in the general education classroom has clear academic, social, and behavioral benefits for [them] and their peers without disabilities (p. 2).
Prior to writing this article, we read broadly from a research literature sometimes cited by full-inclusion advocates. As would be expected, it addresses many questions from theoretical to practical. But two stand out for their relevance to Yudin’s “more time” claim. The first question is whether there are indeed empirically validated activities, curricula, and materials that can strengthen the functioning of SWD in general education settings. In a word, the answer is “yes.”
Can General Classroom Interventions Strengthen SWD’s Academic Performance?
Beginning in the 1980s, OSEP earmarked monies for the development of interventions (a) to strengthen SWD’s performance in the domains of language, peer relations, behavior, and academics and (b) to be implemented primarily by general educators in preschool-to-high school mainstream settings. In part, OSEP’s investments were in response to The Regular Education Initiative (Kaufman et al., 1990; Will, 1986)—an influential policy that developed into full inclusion in the 1990s (D. Fuchs & Fuchs, 1994). The Regular Education Initiative’s basic assumption was that better resourced general classrooms, with teachers implementing evidence-based instructional programs, would benefit many SWD and many of their typically developing peers, thereby reducing the need of special education placements.
Many programs were developed for children and youth with low-incidence disabilities. Experimental and quasi-experimental evaluations of these and subsequent interventions for students with developmental disabilities, as an example, tended to show that these programs, when carefully designed and executed, strengthened expressive language, improved peer interactions, increased task engagement, and encouraged more normative school behavior (e.g., Buckley et al., 2006; Fisher & Meyer, 2002; Guralnick et al., 1996; Helmstetter et al., 1988 cited by the National Council on Disability, 2018).
OSEP-sponsored research and development also addressed the academic needs of the more numerous students with high-incidence disabilities (learning disabilities [LDs], language impairments, and behavior disorders) in general classrooms. Whereas some resulting programs and approaches proved ineffective (e.g., adaptive learning environments model; see D. Fuchs & Fuchs, 1988) or, to this day, remain unvalidated (e.g., Universal Design for Learning despite Yudin’s boosterism; see Edyburn, 2010), others, such as cooperative learning, have been beneficial. Cooperative learning makes use of small, heterogeneous groups of students who work together to achieve common learning goals (Johnson & Johnson, 1992). Comparatively effective versions require students to work both collectively and individually on structured curricula designed by the developers (e.g., Slavin, 1983, 1996); Research demonstrates that when cooperative learning is implemented with fidelity, it can accelerate the academic achievement of many SWD and their peers (e.g., Stevens & Slavin, 1991). The same may be said for other class-wide programs, such as class-wide peer tutoring (e.g., Delquadri et al., 1986), peer-assisted learning strategies (e.g., D. Fuchs et al., 1997), and direct instruction (e.g., Carnine et al., 2017).
Notwithstanding the demonstrable benefits of these programs, there is a necessary and important caveat. Not all SWD benefit from them. For example, McMaster and Fuchs (2002) reviewed 15 published studies of cooperative learning published from 1990 to 2000, in which authors reported cooperative learning effects on SWD’s academic performance. Among the 15 studies, eight were implemented in mainstream classrooms. No statistically significant effects were reported in four of these eight studies; mixed effects were found in three studies; and findings from one study were uninterpretable. One reasonable explanation of this seemingly inconsistent pattern of results is that SWD differ in the severity of their learning difficulties. Those with more serious learning problems require an intensity of intervention well beyond that which is offered by cooperative learning and other class-wide programs (see O’Connor & Jenkins, 1996, on the behavior and academic performance of SWD in cooperative learning classrooms; also see Supplemental File, Part A, Note 3).
Are General Classrooms Intrinsically Beneficial?
When designing their class-wide programs, the program developers’ attention was properly fixed on the programs themselves—how to make them practical and beneficial for teachers and students, respectively—not on the contexts in which they were developed. In other words, the independent variable in the cooperative learning studies, say, was the program itself, not the setting. This distinction between program and setting leads to a second key question asked by contributors to the full-inclusion literature: Do general classrooms, irrespective of the presence of evidence-based programs, improve SWD’s academic outcomes? Put differently, is there something inherently special about general classrooms?
This was surely Michael Yudin’s view and is the view today of many full-inclusion advocates (e.g., Schifter & Hehir, 2018). As we have explained, the “something special” about these placements for Yudin is their normalcy, their mix of students with and without disabilities, their putative culture of high expectation (see RTI Leadership Forum, 2010). Whereas he would often describe this supposed affirming, enabling culture as if it were “baked” into general classrooms, Yudin never suggested anything magical or mysterious about it. Rather, he would say it reflects a commitment to hold all students to high standards, a most important example of which is ensuring SWD access to the general curriculum (a belief to which we will return).
Which brings us to the question motivating the writing of this paper; namely, “Do general class placements indeed benefit SWD’s academic achievement apart from the particular interventions or programs that may be conducted there?” Two ambitious longitudinal studies run by researchers at SRI and funded by OSEP are pertinent to the question. They are the National Longitudinal Transition Study (NLTS; e.g., Wagner, 1991; Wagner et al., 2003, 2006) and Special Education Elementary Longitudinal Study (SEELS; e.g., Schiller et al., 2008).
NLTS and SEELS
The purpose of NLTS was to describe the school achievement of SWD in secondary school and their experiences while transitioning from high school to adult life. The hope was that these descriptions would generalize to SWD nationally in each of 12 disability categories. An important finding of this investigation was that the SWD taking three-fourths of their courses in general education settings were more than a full year closer to grade level than the students who took only one-fourth of their courses in general education (Wagner, 1991, pp. 4–12).
Researchers for SEELS (e.g., Schiller et al., 2008) conducted a complementary study. They followed a nationally representative sample of SWD as they moved from elementary school to middle school to high school from 2000 to 2006. The researchers found that SWD who spent 75% or more of the school day in general classrooms performed more strongly in reading and mathematics than SWD who were in general classrooms for only 25% of the school day. Gilmour (2018) describes similar studies conducted by the states of Washington (Theobald et al., 2017) and Massachusetts (Schifter, 2016), which produced results comparable to those reported in NLTS and SEELS.
Whereas findings from these national and state-wide efforts seem to support Yudin’s “more time” argument and a policy of full inclusion, they are compromised by selection bias. To explain, we turn to the “efficacy studies,” a historically important set of investigations in the 1950s and 1960s in which researchers explored the academic benefits of mainstreaming elementary-age students with intellectual disabilities then called “mental retardation.”
Selection bias and apples-oranges comparisons
The sine qua non of the efficacy studies was that the academic performance of two groups of special-needs students was compared against each other. One group obtained its instruction in general classrooms, the other in special education classes (see Kirk, 1964; Semmel et al., 1979). A recurring finding was that the performance of the SWD in general classrooms was superior, leading many to conclude then and afterwards that general class instruction was more effective. However, virtually every efficacy study reflected selection bias. Teachers deliberately chose students for instruction in general education and special education classes, which is to say SWD were with rare exception assigned randomly to the two settings. This should have raised suspicion that the mainstreamed students were viewed as more likely to profit from instruction there. Today, there is widespread agreement—at least among academics—that the efficacy studies illustrate case after case of an “apples versus oranges” comparison (see Zigmond, 2003, for discussion of the efficacy studies)
Similar concern exists about NLTS and SEELS and the state-wide data from Washington and Massachusetts. Using SEELS as an example, contrasts between SWD groups who were mainstreamed 75% and 25% of the school day are encumbered by the same issues that weakened the meaningfulness of the efficacy studies. Gilmour (2018) writes, Even in [more sophisticated] studies [like NLTS and SEELS] that account for students’ prior levels of academic achievement, the researchers may not have captured all . . . aspects of a student . . . that can influence both the setting in which he is placed and his future outcomes. A student’s educational placement is an IEP team decision and may be based on a host of factors not included in the administrative data sets to which researchers typically have access . . . [which] makes estimating the true causal effect of [general class placement] on student outcomes nearly impossible. (p. 12)
In sum, these national and state-wide efforts exploring an inherent educational benefit of mainstream placements are flawed because they do not parse setting effects from learner differences likely to characterize the SWD spending more and less time in general classrooms. Whereas these studies require cautious interpretation, full-inclusion advocates (e.g., Schifter & Hehir, 2018) have leaned on them for years as evidence of the superiority of general class placements, a superiority often attributed to an assumed transformative culture of high expectations.
Study Purpose
We have briefly reviewed two kinds of evidence often marshaled in support of the full inclusion of preschool and school-age students with low-incidence and high-incidence disabilities. The first comes from research on interventions developed for SWD in mainstream settings. Findings indicate that when these interventions are conducted with fidelity they benefit many (but not all) special-needs children and youth. A different group of studies has explored the importance of general class placement regardless of the use of evidence-based interventions. This latter set of studies is more directly related to the validity of Yudin’s “more time” claim and their outcomes are less persuasive.
The “more time” claim reflects a dose-response hypothesis. In simple form, it says that if an intervention or setting is beneficial, then more of it is better than less. If the general class is viewed as a desirable place for SWD, and study findings show that those mainstreamed for 75% of the school day outperform others mainstreamed for only a fraction of that time, then, on its face, the findings make sense. We have argued there is more to it than that. In the current study, we explore a dosage hypothesis like the NLTS and SEELS researchers before us. However, we do so in a manner that circumvents very difficult-to-interpret comparisons between different learners in different settings.
We used electronic and hardcopy documents from multiple sources to consolidate (i.e., bring together in one place) two national databases covering the same 18-year period. For nine of these 18 years, an OSEP database provides percentages of 6- to 11-year-olds with individual educational plans (IEPs) placed in general classrooms for at least 80% of the school day (OSEP’s operational definition of “inclusion”; see below). For the same 9 years, a second database from the National Center for Educational Statistics (Institute of Education Sciences, U.S. Department of Education) contains National Assessment of Educational Progress (NAEP) data on the proportion of fourth graders with IEPs or Section 504 plans reading at a “Basic-or-Better” level (explained below).
Displaying the trajectories of OSEP placement data and NAEP reading data on the same graph for the same 18-year period facilitated examination of the correspondence between the two over a considerable period. Increases in both trajectories would signify that more inclusion co-varied with more SWD reading above criterion. Such co-variation in turn would suggest, however tentatively, a dosage effect. We also explored if increases in SWD’s NAEP participation over time corresponded with increases or decreases in NAEP performance.
Such exploration was similar in important respects to the analyses described in NLTS and SEELS. We and the authors of these longitudinal studies, for example, ran correlational analyses on nationwide data. Unlike NLTS and SEELS, however, we supplemented our nationwide analyses with state-level study. For example, we looked at states with strongly positive inclusionary trajectories, reasoning that if a dose-response hypothesis has merit then we were most likely to observe it in these states. Supplementary analyses notwithstanding, we recognize the limits that correlational data impose on inferences like “placement in a mainstream classroom positively affects SWD academic performance.” Below, we discuss study features that complicate interpretations of this work in other ways as well. Nevertheless, we believe we provide a unique and legitimate look at important and enduring questions about whether and how educational placement benefits SWD’s school performance, however imperfect this look may be.
Before describing the assemblage of our database and related study methods, we make one more necessary introductory point. Because we explore the benefit of full inclusion in terms of reading performance on the NAEP, our study and related commentary are more focused on high-incidence disabilities (e.g., LDs) than on low-incidence disabilities (e.g., intellectual disabilities).
Method
Locating OSEP and NAEP Data
In putting together a multi-year database on SWD’s educational placement and reading performance, we had two objectives: to find both placement and performance data in as many years as possible and to find all available placement and performance data. This meant exploring different locations and sources (hardcopy and electronic) and going “back-and-forth” between the two objectives. In short, our search was iterative and interactive, not strictly linear.
We began by downloading OSEP state-level files on “Educational Environment” for 2005 through 2015, inclusive. We then went to the NAEP website to determine the maximum number of these years for which there might be both OSEP placement data and NAEP reading data. We learned that the NAEP had permitted accommodations for SWD beginning in 1998. So, this was the first year for which there were data on SWD’s reading performance. Thereafter, the NAEP reading exam was administered in 2000, 2002, 2003, 2005, 2007, 2009, 2011, 2013, and 2015. To make use of the NAEP data from 1998 to 2005, we needed to find OSEP placement data prior to 2005 (i.e., for 1998, 2000, 2002, and 2003).
However, OSEP placement data were not available online prior to 2005. In search of pre-2005 placement data, we spoke with the director of Westat’s IDEA Data Center, who in turn sent us to a co-manager of the Communication and Dissemination Activity Area for the IDEA Data Center. Because this co-manager had directed Westat’s production of OSEP’s Annual Reports to Congress (ARCs) for 10 years, she knew how and where the IDEA compliance data were collected, organized, and disseminated. She confirmed that the ARCs were the only published source of placement data prior to 2005. We then located these data for the 1998 school year in the 2000 ARC; the 2000 data in the 2002 ARC; and the 2002 data in the 2004 ARC. Placement data for the 2003 school year were not available.
OSEP’s Inclusion Data
Educational placement data were gathered from two public OSEP sources. Data before 2005 were found in the OSEP ARCs, which can be downloaded as PDF files. For each year, we selected the single page in the ARC that contained the table, “Percentage of children ages 6-11 served in different educational environments under IDEA, Part B: All Disabilities.” We converted the PDF of that page into an Excel file and selected the data describing the percentage of SWD educated “outside the regular class <21% of the day.” This became our de facto definition of “general class placement,” or “inclusion.”
Data from 2005 onward were accessed online (IDEA Section 618 Data Products, 2015). We downloaded the “Educational Environments” or “Child Count and Educational Environments” CSV file for each school year. The data were filtered to reflect students 6 to 11 years old across disabilities. We then used the number of these students who were served “inside regular class 80% or more of the day” as a numerator with the total number of students in the age range as a denominator to calculate a percentage educated in the general classroom for 80% or more of the day. This proportional metric matches the data in the ARC. The metric also permitted comparisons across years and states/territories and with the NAEP Reading data, which are reported as a percentage of fourth-grade SWD scoring at or above a given performance level per year.
NAEP’s Reading Data
Defining reading comprehension
The NAEP reading exam is a timed test of reading comprehension. It asks students to read grade appropriate materials and to answer questions based on what they have read. Grade appropriate materials include three types of literary texts (fiction, literary nonfiction, and poetry) and three categories of informational texts (exposition, argumentation, and persuasive text), and procedural texts and documents. Passages are between 200 and 800 words in length. Test items target three cognitive tasks: locate/recall, integrate/interpret, and critique/evaluate. At fourth grade, integrate/interpret items comprise approximately 50% of the questions, reflecting an emphasis on complex inferencing. Types of response include selected response and short and extended constructed response. Fourth-grade students are expected to spend approximately 50% and 40% of their time, respectively, on selected response and short constructed response item types. Students obtain a raw score which corresponds to one of four performance levels, two of which are proficient and basic.
A score in the proficient range indicates mastery of grade-level reading. A score at the basic level indicates partial mastery and is typically interpreted as one grade-level behind. Table 1 (online supplemental file, Part B) presents NAEP functional descriptions of proficient and basic reading levels for fourth-grade students. The NAEP performance framework changed in 2009 to assess a somewhat different set of skills. (See Table 1 in online Supplemental File B for descriptions of proficient and basic levels for 1992–2007 and 2009–2015). Despite differences in earlier and later frameworks, the NAEP concluded that scores from 2009 may be compared to scores from previous years.
Accommodations
SWDs are permitted accommodations on the NAEP reading exam as stipulated in a student’s IEP. Possible accommodations include but are not limited to extended testing time; small-group, or one-on-one, testing; directions read aloud by the examiner; and breaks during test administration. However, passage content and test questions may not be read aloud to SWD, regardless of disability.
Locating the data
We obtained the NAEP data from the Main NAEP Data Explorer (National Center for Education Statistics, 2017). Using menu options, we collected composite scale scores for students with and without disabilities on the Grade 4 reading assessment for all years during which accommodations were permitted. It was not possible in each year to disaggregate students with Section 504 plans (guaranteed by Section 504 of the Rehabilitation Act of 1973) from other SWD. So, SWD’s NAEP reading performance includes scores from students with Section 504 plans as well as those with IEPs.
Comparing SWDs in OSEP and NAEP Databases
Our intent was to match OSEP and NAEP datasets to permit relatively straightforward comparisons between the two across an 18-year period. Nevertheless, the two agencies, when developing their respective datasets, relied on different sampling methods. This led to a less-than-perfect match of definitions of disability, age groups, public and private schools, and regions of the United States.
Definitions of SWD
OSEP’s data collection efforts reflect the agency’s responsibility to monitor educators’ compliance regularly and systematically with IDEA; specifically, and most importantly, that SWD with IEPs obtain individualized instruction and frequent monitoring of their academic progress. All SWD in the OSEP database, therefore, have IEPs. None have Section 504 plans because by definition “504 students” do not have IEPs.
By contrast, and as mentioned, the NAEP’s database includes Section 504 students as well as SWD with IEPs. Whereas SWD with Section 504 plans do not require individualized instruction, they and their parents/guardians do require formal assurances that they will obtain equal access to public education and services. The NAEP’s more inclusive definition of SWD reflects its purpose, which is to provide educators and policymakers with accurate estimates of the academic achievement of students in the United States. The inclusion of students with Section 504 plans suggests that SWD in the NAEP dataset are more heterogeneous and less severely disabled than SWD in the OSEP dataset.
Another consideration when comparing the two datasets is that SWD were sometimes prevented from taking the NAEP Reading exam. The 2015 Technical Appendix for the NAEP Reading exam includes a table that, for each state, presents the number of fourth graders who were excluded from taking the test over the total number of fourth-grade SWD from 1992 to 2015. Participation rates vary significantly among states and across the years (see below).
Chronological age and public and private schools
The NAEP reading test is administered to students in Grades 4, 8, and 12. Because the test has most consistently been administered at Grade 4, we used Grade 4 reading data in the current study. Unlike the NAEP, OSEP has historically reported its IDEA placement data using age bands: 6 to 11, 12 to 17, 18 to 21, not grade levels. Because most fourth graders are 9 years old, we chose OSEP’s 6 to 11 age band as representing the closest match to the NAEP’s fourth graders.
The NAEP data in this study reflect only public school students, whereas the OSEP placement data reflect a mix of public and private school students. SWD “in the regular class for 80% or more of the day” refers to students in public schools only, but the total number of SWD in all educational environments includes students in public and private schools. Since the proportion of SWD in the most inclusive placement option compares public school students to a combination of public and private school students, this estimate is slightly deflated. But the vast majority of SWD receiving federal funds are educated in neighborhood public schools. In 1998, for example, only 2.22% of SWD were educated in a separate facility (including a public separate facility, private school, residential facility, or home). In 2015, this figure was 3.89%.
Definitions of the “United States.”
OSEP and NAEP datasets reflect more and less inclusive definitions of the “United States.” NAEP data include only students from the 50 states and the District of Columbia. OSEP data include students from the United States, outlying areas, and freely associated states, except for 2005, 2007, 2009, and 2011 when the students were from the 50 states and the District. We excluded the OSEP data from Puerto Rico, Guam, American Samoa, the U.S. Virgin Islands, Northern Marianas, the Republics of Palau and Marshall Islands, the Federated States of Micronesia, the Bureau of Indian Education, and the Department of Defense Education Activity. We did so because data from these regions were collected inconsistently or not at all.
Data Analysis
Data analyses were mostly descriptive in nature and conducted in two phases. In the first phase, we worked with the national data provided in the extant data files. We constructed graphs illustrating relations between general class placement and reading performance (see Figure 1). We then probed these relations by modeling state-level data from all 50 states and Washington, DC. We ran multilevel growth models to describe trends in correspondence between proportions of SWD (a) participating in general classrooms and (b) meeting a basic level of reading performance. The models produced estimates of the “average state” that represented the national trend as well as estimates of growth parameters for individual states.
National averages of the proportion of SWD in general classrooms 80% or more of the school day and the national proportion of SWD reading (a) 1 grade level below their assigned grade or better (basic or better) and (b) at grade level or better (proficient or better) on the NAEP between 1998 and 2015.
To compare placement and performance data, we compared linear and curvilinear growth models. In certain circumstances, it is advisable to model proportional, or percentage, data with a logistic regression model so that predicted values are bound between 0 and 1 [or 0% and 100%]. In this case, we chose a continuous model because we simply wanted to describe trends for the set time frame of 1998 to 2015; we did not make predictions about future years. Time was centered at 2005 both to make the intercept meaningful (i.e., 1 year after the reauthorization of IDEA in 2004) and to reduce collinearity among growth parameters.
Results
National Trends
Figure 1 displays proportions of 6- to 11-year-old students with IEPs who spent 80% or more of the school day in general classrooms in 9 of 18 years for which we had data. Averaged across the 50 states and Washington, D.C., these proportions ranged from just under 60% to 70%. Figure 1 also exhibits a national sample of SWD’s NAEP reading performance for the same 9 years. Reading performance is shown in the figure in two ways: the proportion of SWDs at and above grade level (proficient or better); and one grade level below or better (basic or better). Between 8% and 10% of SWD achieved proficiency across the years; 22% to 37% scored at a basic-or-better level.
Results from model comparison tests revealed that a linear model produced a best fit for rates of general classroom placement
Model estimates revealed that the 2005 national rate of general class placement was 63.5% and that it significantly and steadily increased by 0.8% from 1998 to 2015 annually. The percentage of SWD reading at a basic-or-better level in 2005 was estimated as 32.5%. Like the placement data, the trend for reading performance was positive. But unlike the placement data, it was decelerating, beginning between 2007 and 2009 (see Figure 1). So, the steadily increasing trend for general class placement and the positive but decelerating trend for reading performance produced a widening gap between the two.
We also looked at the relationship between SWD’s classroom placement and reading performance by calculating placement–performance correlations toward the start of the 18-year period (2002) and again toward its end (2015). Many states/jurisdictions had missing data in the early years, but 45 of 51 had both placement and reading data by 2002. By 2015, every state/jurisdiction had the two sets of data. In a preliminary scatterplot, the form of the placement–performance relationship appeared non-linear. So, we calculated Spearman’s rank correlation coefficient because it makes no distributional assumptions about the variables; rs was 0.22 in 2002 and 0.18 in 2015. Thus, general class placement and reading performance were only weakly associated at both the earlier and later time points, points separated by a 13-year interval.
State Trends
To further explore the validity of the national trends, we turned to state-level data; specifically, SWD’s NAEP scores in the 10 states/jurisdictions with the highest rates of (or strongest trends for) general class placement. Assuming advocates of full inclusion are right to believe that general class placements indeed benefit SWD’s academic achievement, we reasoned we were more likely to see evidence of corresponding reading improvement in these states/jurisdictions. Figure 2 depicts NAEP reading data for Washington, D.C. (
Proportions of students with disabilities (SWD) reading at a basic-or-better level on the National Assessment of Educational Progress (NAEP) in states/jurisdictions with the 10 highest trends of general class placements from 1998 to 2015.
In two of these 10 states/jurisdictions, there was a positive trend for NAEP reading performance (Washington, DC, Alabama); in three states, there was no evidence of growth (Mississippi, Rhode Island, Wisconsin); and in the five remaining states, there was a positive trend but it was decelerating over time. The states with curvilinear patterns had different maxima. Thus, for the 10 states/jurisdictions with highest rates of general class placement, there was no uniform pattern of reading performance. Figure 2 also shows a bolded line plotted with model estimates. It represents the national trend for the proportion of SWD with a basic-or-better performance.
Figure 3 presents general class placements and NAEP reading performance for the 10 states/jurisdictions with the strongest positive slopes of inclusion in mainstream settings. They have a wide range of intercepts, or absolute levels of SWD in general classrooms. The intercept represents the mean general class placement rate in 2005 and it ranges from .30 (Washington, DC) to .72 (Alabama). In Figure 3 (as in Figure 2) we did not see a commonly shared pattern of reading growth among these states/jurisdictions. Rather, we observed diverging trends in each state/jurisdiction for placement in mainstream classrooms and reading performance on the NAEP, a pattern paralleling the nationwide trend illustrated in Figure 1. Even in the states/jurisdictions showing a steady uptick of reading over time (i.e., Washington, DC, Alabama, and Georgia), rates of improvement were much lower than rates of SWD in general education 80% or more of the school day.
Proportions of students with disabilities (SWD) in general class placements who where reading at a basic-or-better level on the National Assessment of Educational Progress (NAEP) in 10 states/jurisdictions with highest slopes for general class placements from 1998 to 2015.
Does Excluding SWD from the NAEP Affect the Placement-Performance Relationship?
Figure 4 is an illustration of contemporaneous relations over time between the proportion of SWD (a) excluded from the NAEP, (b) reading at proficient and basic-or-better levels, and (c) spending 80% or more of the school day in general classrooms. The figure shows that, whereas the percentage of SWD excluded from the NAEP declined steadily over time, there was no corresponding change in percentages of SWD scoring at proficient or basic-or-better levels.
Rates of general education placements, NAEP scores at basic-or-better and proficient-or-better levels, and exclusion from NAEP testing for SWD from 1998 to 2015.
We also looked at six states with markedly decreasing rates of exclusion (see Figure 5). For each state, the figure shows relations over time between proportions of SWD excluded from the NAEP testing and scoring at a basic-or-better level. In Florida and Wyoming, as exclusion rates decreased, there was a concomitant, increasingly positive trend in reading performance. In Mississippi and West Virginia, there was neither a positive nor negative trend. And in Rhode Island and Vermont, there was a negative trend in performance. Exclusion trends for each state were calculated and plotted using ordinary least squares regression with linear and quadratic growth terms. So, among the states deliberately chosen for their decreasing rates of exclusion, there was no shared pattern of reading performance.
NAEP basic-or-better achievement rates and NAEP exclusion rates in six states with dramatically decreasing exclusion rates for SWD from 1998 to 2015.
Discussion
Findings and Caveats
Summary of findings
Whereas the proportion of SWDs in general classrooms for 80% or more of the school day was increasing steadily and significantly over time, the trend for the proportion of SWDs meeting or exceeding a basic level of performance, although positive, was decelerating such that the gap between the two was widening. Moreover, only 8% to 10% of SWD read with proficiency; just 22% to 37% read at a basic-or-better level. Among the 10 states/jurisdictions with strongest accelerating trends for general class placement, there was no uniform pattern of reading performance over the years. In Alabama and Washington, DC, there was a positive trend for NAEP scores; in Mississippi, Rhode Island, and Wisconsin, there was no growth in scores; and in Delaware, Georgia, Maryland, Texas, and Virginia, there was an initial positive trend that decelerated.
We also explored whether excluding SWD from the NAEP affected averaged scores over time. Nationwide, their exclusion steadily declined with no corresponding positive or negative change in percentages achieving proficient or basic-or-better levels—a finding that held for the six states with most notable decreases in exclusion.
These data on SWD’s generally very poor academic performance in mainstream classrooms do not stand alone. We have previously mentioned the well-known findings of the NLTS research program. However, our summary of these results is but a small part of the story and it fails to give a true indication of what the researchers discovered. In a 2007 email to the first author, Mary Wagner, the director and principal investigator, elaborated on her team’s findings. She wrote that high school students with LD in the 1990s (a) were taking more challenging academic classes (typically taught by general educators) than a decade earlier; (b) were earning higher grades (only 2% had “mostly As” in the 1980s; 19% in the 1990s); and (c) were more likely to graduate (56% in the 1980s; 74% in the 1990s). And yet, she continued, “When you look at what’s going on inside the classroom, you begin to see . . . that all is not well.” Her email described the infrequency with which students with LD participated in the life of classrooms—took part in class discussions, responded orally to teacher questions, and shared their work with peers (Wagner et al., 2003). Most importantly, she said, “They still can’t read.” Seventy-four percent placed in the bottom quartile of the normative group on the Passage Comprehension subtest of the Woodcock–Johnson III (Wagner et al., 2006). Wagner’s email ended this way: “So, more [students with LD] may be graduating from high school, but they have academic deficits that raise real questions about where they’re going next.”
Thus, our analysis indicated that SWD’s very low reading performance on the NAEP did not track the increasing amount of time they were spending in regular classrooms. There was little corroboration of a dosage hypothesis; scant evidence to support the truthfulness of Michael Yudin’s claim that virtually all SWD’s academic achievement is strengthened by their participation in mainstream settings. Moreover, as just mentioned, results corroborate the experiences of Wagner and her NLTS research team. All that said, important features of our archival analysis complicate interpretations of findings and these features require discussion and consideration when looking at the data.
Caveats and admonitions
Our first caveat draws attention to the differences between SWD in the OSEP and NAEP databases. As reported, the OSEP dataset includes SWD with IEPs. The NAEP database, by contrast, includes SWD with IEPs and Section 504 plans. Students with Section 504 plans usually demonstrate stronger academic performance. Also, the OSEP database pertains to 6- to 11-year-olds; the NAEP data were collected on fourth graders. Earlier, we expressed concern about “apples vs. oranges” comparisons implicit in the NLTS and SEELS longitudinal studies. Because “what’s good for the goose is good for the gander,” we recognize the inconsistencies between OSEP and NAEP databases in our analysis. We do not know whether and how such inconsistencies affected our findings.
Second, the apparent absence of a meaningful relation between OSEP placement data and NAEP reading data may be interpreted as an instance of “proving the null hypothesis.” It is widely acknowledged that proving the null hypothesis is an impossible proposition. Whereas there may indeed be an absence of a placement–performance correlation as we report, it may be present but unobserved for myriad reasons. Hence, another reason why we regard our findings as suggestive, not conclusive.
A third caveat concerns the NAEP. Despite that reading comprehension is a multi-dimensional, hard-to-measure construct (Clemens & Fuchs, 2021), we and others regard the NAEP as a generally credible test. That said, concern has been expressed about its use with SWD. Although it is designed to test a subset of students with and without disabilities, some states have deliberately excluded SWD. There has also been widespread confusion about whether and what test accommodations should be permitted. These considerations and others raise questions about the representativeness (generalizability) of SWD’s NAEP performance.
Fourth, our state-level analyses produced results that did not always jibe with national-level findings. For example, and as reported, Alabama and Washington, D.C. showed positive correlations over the years between general class placement and NAEP performance. Whereas an exploration of state- and local-level policies and practices might have helped us to understand these anomalies and to think with greater nuance about placement and performance, our archival analysis is a far cry from the kind of on-the-ground research necessary for such exploration.
The Quality of General Education Instruction
Because our focus to this point has been on SWD, we have not addressed how poorly many students without disabilities have been performing, a necessary consideration to understand the findings from our study and full inclusion more completely. Recent NAEP data show that 34% of fourth graders read below a basic level or, as we have explained, more than one grade level below where they sit. Reading below basic means one is unable to locate relevant information, draw simple inferences, and interpret meanings of words in text (NAEP, 2019). An additional 31% of fourth graders read above this performance level but still below grade level. Thus, nearly two-thirds of fourth graders nationwide read below their grade on the NAEP.
In addition, in a blog posted on a Brookings’ website, Rothwell (2016) reported no gains in literacy among 17-year-olds in the United States since the start of the NAEP in 1971; no progress in mathematics since 1990. Moreover, on the most recent NAEP, the literacy gap between 12th graders at the 90th percentile and 10th percentile was the greatest since 12th graders were first assessed on the NAEP in 1992 (see National Center on Education and the Economy, 2020). What explains the poor reading and mathematics scores of so many students with and without disabilities in general classrooms?
Under-funded schools and demoralized staff
A well-known explanation focuses on the under-funding of many public-school systems (e.g., Goldstein, 2018, 2019; Romero et al., 2018), a dramatic expression of which was the series of teacher strikes that began in the run-up to the 2018 midterm elections in six Republican-controlled swing states (Arizona, Colorado, Kentucky, North Carolina, Oklahoma, and West Virginia). Low salaries and a loss or degradation of instructional resources have arguably demoralized teachers, convincing many to leave the profession (e.g., Billingsley & Bettini, 2019). Darling-Hammond et al. (2018) have documented acute teacher shortages in California that are forcing many school districts to hire long-term substitutes or teachers with substandard credentials; to leave positions vacant; or to increase class sizes.
Unprepared teachers
A less obvious factor contributing to many students’ weak academic achievement is the large number of teachers unprepared for the challenges of the classroom. One reason for this unpreparedness, we argue, is that many schools of education in America’s colleges and universities employ faculty who refuse to teach scientifically validated methods of instruction to aspiring teachers. Consider reading. There is an impressive science of reading based on hundreds of studies conducted across decades (see National Institute of Child Health and Human Development [NICHD], 2000). This research explains why a good reader must learn to decode words and why many weak readers do not master this skill. The research has produced many demonstrably effective phonics programs. However, starting in the 1970s, the science of reading—and the phonics-based programs it inspired—was rejected by proponents of whole language. Frank Smith, a leader of the whole-language movement, wrote dismissively that it is “only through reading that children learn to read” and “teaching them the sounds of letters is literally a meaningless activity” (cited in Hanford, 2018).
Whole language—including the popular practice of cueing, which encourages children to “guess words using pictures and context instead of first sounding them out” (Hanford, 2020)—has been adopted by many faculty in the academy. Hanford (2018) writes, I interviewed a professor of literacy in Mississippi who told me she was “philosophically opposed” to phonics instruction. One of her colleagues told me she did not agree with the findings of reading scientists because “it’s their science” [not ours].
It should come as no surprise that surveys indicate many young teachers graduate from their formal preparation programs without this knowledge (Binks-Cantrell et al., 2012).
In short
There is incontrovertible evidence that many students with and without disabilities are performing poorly in school. Likely reasons are that K–Grade 12 schools have been under-funded, talented young people are avoiding the profession in favor of more attractive career options, and those motivated to teach are unready for the challenges of the classroom because schools of education have failed to prepare them properly. All this prompts the question, what is the logic—and where are the data—for thinking that virtually all SWD can obtain an appropriate education in the general classroom? Which brings us back to Michael Yudin’s frequently expressed belief in general education’s putative culture of high expectations and rigorous standards, the “active ingredients,” he might say, that make mainstream classrooms appropriate for all children. By connecting his support of full inclusion to the importance of rigorous standards, he was no doubt piggybacking on a larger initiative: standards-based reform.
Standards-Based Education
College- and career-ready standards (CCRS)
Standards-based reform has dominated education policy for more than two decades, beginning with the 1994 reauthorization of Title I of the Elementary and Secondary Education Act and strengthened in its 2001 reauthorization. It assumes that uniformly challenging standards are established for all children and youth; assessments are aligned with the standards; virtually all students (including most SWD) participate in the assessments; and their performance becomes a basis for district- and school-level accountability. In spring of 2009, the National Governors Association and Council of Chief State School Officers began work on Common Core State Standards (Porter et al. cited in Scruggs et al., 2013), a well-known version of CCRS. Teams of stakeholders produced rigorous grade-by-grade objectives for K–Grade 12 instruction in core subjects, believing that they would lead to more students graduating high school who would be “college and career ready.”
Center on Standards, Alignment, Instruction, and Learning (C-SAIL)
In 2015, the Institute of Education Sciences funded C-SAIL to support and to document state, district, and classroom efforts to implement CCRS. An additional charge was for C-SAIL to explore CCRS effects on teacher instruction and student performance. With respect to this last objective, Song et al. (2022) focused their evaluation on states that adopted CCRS in 2010, the year in which most of them adopted the more challenging standards. Song et al. divided the states into “treatment” and “comparison” states. Treatment states had relatively weak standards prior to CCRS adoption. Comparison states had stronger standards prior to adoption. Put differently, the C-SAIL researchers assumed CCRS adoption represented a stronger treatment in states whose prior standards were less rigorous than CCRS. Using eight consecutive years of state-level NAEP data in reading and mathematics at Grades 4 and 8, Song et al. generally found non-significant differences between treatment and comparison states across reading and math domains and grade levels. For SWD, most differences favored comparison states, especially in the later years.
The center’s researchers (e.g., Edgerton et al., 2020) also surveyed educators at three levels (state, district, and school) in three states (Kentucky, Ohio, and Texas) to determine if administrators and teachers had the same or different perspectives on various aspects of CCRS implementation, including whether the standards were appropriate for SWD. Respondents’ views were strongly influenced by (literally) how near or far they were from classrooms. State-level officials claimed CCRS benefited students with and without disabilities equally. District-level educators were less sure. Classroom teachers said they were unprepared to teach SWD to the standards.
In sum, C-SAIL researchers, using multiple research methods, found little evidence that standards-based reform—the establishment of uniformly challenging standards for all—accelerated the achievement of students with or without disabilities. Whereas these findings are consistent with others’ research, more encouraging data on the importance of standards have come from a small number of national- and district-level studies (see Bleiberg, 2021) There is a second and maybe more important caveat. Implicit in our judgment on the benefit of standards-based reform is an assumption that educators’ use of standards is all or nothing; that is, they either implement them or they do not. Yet, recent work with low-performing students shows that the use of some grade-level standards with some out-of-grade-level intensive instruction on prerequisite skills can accelerate learning (e.g., L. S. Fuchs et al., 2015; Shriner et al., 2013; Shriner et al., 2017).
To elaborate, consider a hypothetical involving Sofia and her special education teacher who targets “reading informational texts with understanding” as an appropriate grade-level standard for her. Sofia is taught strategies to help her preview and predict informational text; develop background knowledge; produce main ideas; make “in text” and “out-of-text” inferences; and recognize different text structures. At the same time, the teacher engages Sofia in out-of-grade fluency-building activities because she knows that it is counterproductive to ignore weaknesses in more basic skills if progress toward standards like CCRS is to be realized.
In trying to make sense of why millions of American students in K–Grade 12 schools have been performing so poorly, we are not suggesting that the current situation is hopeless. There are strong professional development programs; research-backed and practitioner-friendly instructional protocols and curricula in reading, math, writing, and the content areas; a network of clearinghouses that make such products accessible to skilled classroom teachers who interact daily with talented and dedicated leaders at the building level. America, indeed, has schools that work. For this to become the rule rather than the exception, we must invest in schools and teachers. We must make our children (all our children) a priority instead of a campaign slogan. If we do so, we can imagine more informed, resourceful, accommodating general education classrooms. What we cannot imagine is that general classrooms will become capable of teaching all SWDs. We cannot imagine this because we know it to be unrealistic and dangerously naïve. Many SWDs need something different than general classroom instruction.
Intensive Instruction
In a 2013 interview, Lou Danielson, former director of innovation and development at OSEP, reflected on the negligible influence of standards-based education on SWD: “There just has not been anywhere close to enough attention to improving instruction for these . . . kids, especially given that many of [them] require very intensive instruction” (Sherman, 2013). In 2015, L. S. Fuchs and colleagues published a study in which they contrasted two approaches to accelerating knowledge of math fractions in three groups of students: two very low-performing groups of SWD and a group of typically developing peers. The first low-performing group received intensive instruction from a trained tutor outside the classroom in groups of two to four students three times per week for 12 weeks. Adding to this intensity was (a) explicit instruction organized by a manual that included materials and directions for 36 lessons and (b) a narrow focus that emphasized the measurement interpretation of fractions and deemphasized calculation demands. The second low-performing group received fractions instruction in general classrooms with accommodations based on the principles of Universal Design for Learning.
In the final year of this multi-year randomized controlled trial, the average gap in fractions performance between low-performing SWD who were instructed in the general classroom and their non-disabled peers was more than twice as large as the gap between intensively instructed low-performing SWD and their peers. L. S. Fuchs et al. (2015) concluded that classroom accommodations like Universal Design for Learning are not likely by themselves to improve academic outcomes for many SWDs; that equating “access” and “inclusion” with placement reflects an unsubstantiated conviction that placement in general education curricula and exposure to putatively higher teacher expectations lead to academic gains for all SWDs.
Where to Go From Here?
Change is necessary to strengthen the education of many students with and without disabilities. And, in the past decade, there has indeed been encouraging change. There is growing recognition, for example, that SWD (and other at-risk youth) require more intensive forms of instruction. The National Center on Special Education Research (Institute of Education Sciences, U.S.D.E) and the Eunice Kennedy Shriver National Institute on Child Health and Human Development (National Institutes of Health) have funded research and development that has produced multiple reading and mathematics programs, requiring one-to-one tutoring or small-group instruction that has benefited students with serious learning disorders. Moreover, OSEP has created technical assistance centers (e.g., the National Center on Intensive Intervention) to support the use of these and other evidence-based programs. Such efforts reflect awareness that special education must focus more on the quality of its own placements and services.
And yet, this activity has had only modest impact on SWD’s academic achievement, partly because of a still popular belief in full inclusion—a zeitgeist, if you will, that often works against the delivery of intensive instruction. Prior to writing this article, we watched and listened to Michael Yudin online as he explained to his audiences his support of full inclusion in terms of general education’s presumed high expectations and rigorous standards. Whereas he touted unproven “innovations” like Universal Design for Learning and questionably promising reforms like Response to Intervention, he never spoke of the national-, state-, and district-level studies reporting the benefits of special education (e.g., Hanushek et al., 2002; also see Hurwitz et al., 2021; Hurwitz et al., 2020; Schwartz et al., 2019). Nor did he reference the more numerous experimental studies (e.g., L. S. Fuchs et al., 2015), documenting the superiority of intensive instruction over general instruction for many SWD. To ignore such evidence and their implications for practice is to ignore science. It is to pretend—in the face of data indicating otherwise—that there is but one solution to educating all SWD; namely, placing them in general classrooms.
It is past time for leaders “from the state house to the schoolhouse” to rethink the zeitgeist; to acknowledge that full inclusion has failed millions of SWD; to convene town halls and invited conferences; to sponsor blue ribbon committees and White Papers; all to reach a shared understanding of how “inclusion” should be defined and implemented so that it becomes a constructive force in the lives of all children and youth with disabilities. Such a process will require hard work, open-mindedness, and an ability to think outside the box. Most importantly, it will require leaders to speak truth to stakeholders. Truth-telling requires recognizing that many educational issues and problems—especially those that matter most—are inherently complex; and that they almost always require different solutions for different children and youth.
Supplemental Material
sj-docx-1-dps-10.1177_10442073221097713 – Supplemental material for Exploring the Truth of Michael Yudin’s Claim: The More Time Students With Disabilities Spend in General Classrooms, the Better They Do Academically
Supplemental material, sj-docx-1-dps-10.1177_10442073221097713 for Exploring the Truth of Michael Yudin’s Claim: The More Time Students With Disabilities Spend in General Classrooms, the Better They Do Academically by Douglas Fuchs, PhD, Hilary C. Mirowitz, MEd and Jennifer K. Gilbert, PhD in Journal of Disability Policy Studies
Footnotes
Acknowledgements
The second author collected the data reported in this paper for her thesis in partial fulfillment of the requirements of her master’s degree from Vanderbilt University. The authors thank Linda Lynch, Joy Markowitz, and Mark Sherman for helping us look for the Office of Special Education Program’s educational placement data; Debbie Shaver for providing information on SRI’s National Longitudinal Transition Study; Kylie Muccilli for our series of figures; and Alli Gilmour, Mengli Song, Lynn Fuchs, Dan Hallahan, Jim Kauffman, Dimitris Anastasiou, and the journal’s editors for their comments on earlier drafts of the paper. These individuals’ generous help should not be interpreted as support for anything we have written.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
This work was partially funded by Vanderbilt Kennedy Center grant support (EKS NICHD Grant #1P50HD103537-01).
Supplemental Material
Supplementary material for this article is available on the Journal of Disability Policy Studies website with the online version of this article.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
