Wetenschappelijk aangetoond dat leerlingen met beperking beter presteren in inclusieve klassen? Geenszins! 

1          Woord Vooraf

Inge Wagemans (UA en auteur van de Koppen-enquête van 4 december j.l. over het M-decreet) beweerde op 5 december in ‘De Ochtend’ dat wetenschappelijk bewezen is dat leerlingen met een beperking beter presteren in een gewone klas.  Ook in publicaties van de Unesco en van onze beleidsmakers wordt die indruk veelal gewekt. Zo lezen we in de recentste brochure (verspreid via Klasse): “Inclusief onderwijs moet de eerste optie worden. Het M-decreet stuurt ons onderwijs in deze richting. Uit de enquête van Wagemans en Koppen bleek wel dat meer dan 80% van de leerkrachten inclusief onderwijs niet haalbaar vond, maar de inclusie-hardliners zijn niet geneigd rekening te houden met de opinie en ervaringswijsheid  van de praktijkmensen  en met de getuigenissen van veel ouders met een kind in het buitengewoon onderwijs. Voor de hardliners inzake inclusie hebben de die leerkrachten en ouders het verkeerd voor.

De leerkrachten hebben in principe wel geen bezwaren tegen de inclusie van leerlingen met een beperking die het grootste deel van de lessen en de leerplannen kunnen volgen. Voor veel van die leerlingen is dit overigens nu al het geval. Er is voor de inclusie van dit soort leerlingen geen  M-decreet nodig. De leerkrachten klagen wel terecht over het feit dat ze voor de goede integratie van die inclusieleerlingen al te weinig steun ondervinden via GON e.d.  Dit bleek ook uit de getuigenissen van de bso-leerkracht tijdens het programma ‘De zevende Dag’  van 7 december j.l. Aan de hand van  twee bekende overzichtsstudies  willen we in deze bijdrage even weerleggen dat volgens onderzoek incusief onderwijs veruit  de beste optie is  voor leerlingen die voor een groot deel de gewone lessen niet kunnen volgen en dus ook niet echt geïntegreerd zijn in het leerproces.  Beide  studies staan ook integraal op het Internet.)

2          Reviewstudie van Geoff Lindsay

De effectiviteit van inclusief onderwijs is volgens die recente overzichtsstudie van Geoff Linsay (op basis van meer dan 100 studies) niet aangetoond! De conclusie luidt: "The evidence  from this review does not provide a clear endorsement for the positive effects of inclusion. There is a lack of evidence from appropriate studies and, where evidence does exist, the balance was only marginally positive. It is argued that the policy has been driven by a concern for children’s rights.”   

Zijn erbij aansluitende conclusie luidt: “The important task now is to research more thoroughly the mediators and moderators that support the optimal education for children with SEN (=specific educational needs) and disabilities and, as a consequence, develop an evidence-based approach to these children’s education." (Educational psychology and the effectiveness of inclusive education/mainstreaming -Reviewstudy  van Geoff Lindsay* Centre for Educational Development, University of Warwick, 2013)

De onderzoeker betreurt in zijn studie ook dat de meeste studies uitgevoerd werden door vurige aanhangers van inclusief onderwijs en dus minder betrouwbaar zijn.  Dit was overigens ook het probleem met de samenstelling van het KOPPEN-programma en de erbij horende enquête van 3 december j.l.

We citeren nog een paar passages uit de overzichtsstudie van Geoff Linsay. “Overall, the weight of evidence reviewed in this paper cannot be said to provide a clear endorsement for the positive effects of inclusion (see also Zigmond, 2003). Just 1% of over 1300 studies published 2000–2005 reviewed addressed effectiveness and the results from these studies were only marginally positive overall, although comparability between outcomes for SEN and TD children could be interpreted as positive rather than non-difference.”

“Furthermore, the studies cover a range of ages and methods of inclusion; used a variety of methods and produced evidence on a number of different outcome variables. Taken as a whole, and with the pre-2000 evidence, which presents a similar picture, there is a lack of a firm research base for inclusive education to support either whether this is a preferable approach in terms of outcomes, or how inclusion should be implemented. The review has highlighted the importance of interaction hence the need to examine moderators and mediators affecting outcomes. This amounts to a clear indication of the power of policies argued to be supporting children’s rights (to be included) rather than of evidence of optimal practice. The research base is more helpful in identifying processes that facilitate inclusion. The differential benefit of training TAs is a case in point. However, there is a lack of rigorous studies that demonstrate positive child outcomes rather than improved teaching processes.”

3    Reviewstudie van Naomi Zigmond

Where Should Students with Disabilities receive Special Education Services? Is One Place Better Than Another? Onderzoeker: Naomi Zigmond, University of Pittsburgh (THE JOURNAL OF SPECIAL EDUCATION VOL. 37/NO. 3/2003/PP. 193–199). De studie staat ook op het Internet.

3.1       Abstracts en eindconclusie

Abstract 1: The question of where special education students should be educated is not new. In this article, the author reviews research studies and research reviews that address this question. She argues that research evidence on the relative efficacy of one special education placement over another is scarce, methodologically flawed, and inconclusive. She also states that “Where should students with disabilities be educated?” is the wrong question to ask, that it is antithetical to the kind of individualized planning that should be embodied in decision making for and with students with disabilities, and that it fails to specify where, for what, and for whom.

Abstract 2 :  90% van de leerlingen kan de gewone lessen grotendeels volgen en daar profijt uit halen. For many of the remaining 10% of students, however, a different orientation will probably be needed. These students need to learn something different because they are clearly not learning what everyone else is learning. Interventions that might be effective for this group of students require a considerable investment of time and effort, as well as extensive support. Special education in a pull-out setting, with its emphasis on empirically validated practices and its use of data-based decision making to tailor instruction to the individual students’ needs, might be better for teaching these students.

Eindconclusie empirisch onderzoek 

“There is no simple and straightforward answer to the question of where students with disabilities should receive their special education instruction. The efficacy research reviewed here, which spans more than 3 decades, provides no compelling research evidence that place is the critical factor in the academic or social progress of students with mild/moderate disabilities. There are probably many reasons for reaching this conclusion, but I suggest only two. The first has to do with the body of research evidence itself. The second has to do with the appropriateness of the question.”

3.2       Explanation 1: Research Base Is Insufficient

Despite the fact that the efficacy research literature on the places where special education services are provided spans more than 3 decades and that dozens of studies have been reported in refereed special education journals, Murawski and Swanson (2002) are right to ask where the data are. Studies worthy of consideration in a meta-analysis or narrative literature review, with appropriate controls and appropriate dependent measures, are few and far between. Of course, research on the efficacy of special education placements is very hard to conduct at all, let alone to conduct well. For example, definitions of service-delivery models or settings vary from researcher to researcher, and descriptions of the treatments being implemented in those models or settings are woefully inadequate. Random assignment of students to treatments is seldom an option, and appropriately matched (sufficiently alike) samples of experimental and control students and teachers are rare. As a result, where special education occurs is not a phenomenon that lends itself to precise investigation, and funding for research studies and publication of results in refereed journals are difficult to achieve.

3.3       Methodologically Flawed Research.

Research designs used to explore the effectiveness of different service-delivery models often employ pre–post treatment group designs. The limitations of these research designs for studying the efficacy of special education have been reported in numerous previous research reviews, most notably in Kirk (1964) and Semmel, Gottleib, and Robinson (1979). Some studies use control groups, often samples of students experiencing “traditional” programs (sometimes referred to as “business as usual” programs) in nonexperimental schools.

In some studies, the researchers manage to achieve random assignment of students to treatments, but most use intact groups of students assigned to the teacher or the school building who volunteered to participate in the experimental treatment program. Often the experimental treatment is well described, although degree of implementation is not. Descriptions of the control treatment and its degree of implementation (if indeed a control group is used) are rarely provided.

Most often, replication is hindered by inadequate descriptions of the treatments and insufficient monitoring of treatment implementation. Thus, even if reliable achievement changes are demonstrated in one research study, difficulty in identifying critical treatment variables makes replicability impossible in virtually all cases. Achievementgains, or lack there of, cannot be related to replicable interventions, and the fundamental question of whether Place A is better than Place B cannot actually be answered.

3.4       Inconclusive Research

The accumulated evidence to date has produced only one unequivocal finding: Languishing in a general education class where nothing changes and no one pays you any attention is not as useful to students with mild/moderate learning and behavior disorders as is getting some help, although it does not seem to matter for students with mild mental retardation. All other evidence on whether students with disabilities learn more, academically or socially, and are happier in one school setting or another is at best inconclusive. Resource programs are more effective for some students with disabilities than are self-contained special education classes or self-contained general education classes, but they are less effective for other students with similar disabilities.

Fully inclusive programs are superior for some students with disabilities on some measures of academic or social skills development and inferior for other students or on other measures. The empirical research not only does not identify one best place but also often finds equivalent progress being made by students with disabilities across settings; that is, the research reports non significant differences in outcomes. Interpreting non significant findings can be tricky. Do we conclude that the proverbial cup is half full or half empty? Do we acknowledge that it does not matter where students receive their special education services and allow parents or school personnel wide berth in making choices? Or do we proclaim that one setting is preferred over another for philosophical or moral reasons with empirical evidence that it “doesn’t hurt”?

3.5  Explanation 2: Efficacy Studies Have Been Asking the Wrong Question Failure to Specify ‘Best for Whom?’

Special education has evolved as a means of providing specialized interventions to students with disabilities based on individual student progress on individualized objectives. The bedrock of special education is instruction focused on individual needs. The very concept of “one best place” contradicts this commitment to individualization. Furthermore, results of research on how groups of students respond to treatment settings does not help the researcher or practitioner make an individualized decision for an individual student’s plan. A better question to ask, if we dare, is “best for whom?” or best for which individual students with which individual profiles of characteristics and needs? Answering this question requires that we abandon the rhetoric in which we call for all students to do this, or all students to learn that, or all students be educated in a certain place.

3.6 Special educators understand about individual differences.

Special educators understand that no matter how hard they try or how well they are taught, there are some students who will never be able to learn on the same schedule as most others, who will take so long to learn some things that they will have to forego learning other things, or who will need to be taught curricular content that is not ordinarily taught. Special educators understood this when they fought hard for the legal requirement of the Individualized Educational Program for children with disabilities, to permit formulation of unique programs of instruction to meet unique individual needs. By continuing to ask, “What is the best place?” we are ignoring what we know.

Restating the question as “best for whom?” would also require new research designs and data analysis. A first step in that direction might be to reanalyze group design data at the individual student level. For example, Zigmond et al. (1995) collected achievement test data for 145 students with LD in three full inclusion programs and for many of these students’ nondisabled classmates. Rather than reporting average growth of the students with LD, the researchers reported the number and percentage of students with LD who made reliably significant gains (i.e., gains exceeded the standard error of measurement of the reading test) during the experimental year. They also reported on the number and percentage of students with LD whose reading gains matched or exceeded the average gain of their grade-level peers. Finally, they reported on the number and percentage of students with LD whose achievement status (i.e., their relative standing in the grade-level peer group) had improved during the school year. These analytic techniques allowed for exploration of setting effects at an individual level. Waldron and McLeskey (1998) followed this same tactic. Unfortunately, neither group of researchers took the final step of describing individual participants in enough detail to permit generalization of the findings or extrapolation of the findings to the individual case. Nevertheless, this approach seems more promising than the traditional approaches that have been used to date in terms of answering the question “best for whom?”

3.7 Failure to Specify “Best for What?”

Different settings offer different opportunities for teaching and learning. The general education classroom provides students with disabilities with access to students who do not have disabilities; access to the curricula and textbooks to which most other students are exposed; access to instruction from a general education teacher whose training and expertise are quite different from those of a special education teacher; access to subject matter content taught by a subject matter specialist; and access to all of the stresses and strains associated with the preparation for, taking of, and passing or failing of the statewide assessments. If the goal is to have students learn content subject information or how to interact with nondisabled peers, the general education setting is the best place.

Pull-out settings allow for smaller teacher–student ratios and flexibility in the selection of texts, choice of curricular objectives, pacing of instruction, scheduling of examinations, and assignment of grades. Special education pull-out settings allow students to learn different content in different ways and on a different schedule. A pull-out special education setting may be most appropriate if students need (a) intensive instruction in basic academic skills well beyond the grade level at which nondisabled peers are learning how to read or do basic mathematics, (b) explicit instruction in controlling behavior or interacting with peers and adults, or (c) to learn anything that is not customarily taught to everyone else.

If educators value education that is different and special and want to preserve that feature of special education, it is legitimate to ask whether the general education classroom can be transformed to support this desire. Or, as Fuchs and Fuchs (1995) asked, “Can general education become special education?” (p. 528) Their experience (and mine) strongly suggests that the answer to this question is “no.” Attempts to transport teaching methods that were developed and validated in special education to general education settings have not been successful.

 Instructional practices that focus on individual decision making for individual students and improve outcomes of students with severe learning problems are not easily transposed into practices that can survive in a general education classroom. General educators will make instructional adaptations in response to students’ persistent failure to learn, but the accommodations are typically oriented to the group, not to the individual, and are relatively minor in substance, with little chance for helping students with chronically poor learning histories (Zigmond & Baker, 1995).

Over and over again, researchers and staff development personnel have come to recognize that general education teachers have a different set of assumptions about the form and function of education than do special educators. General educators cannot imagine focusing intensively on individual students to the extent that different instructional activities for different students are being implemented at the same time. This is simply impractical in a classroom of 25 to 35 students. Moreover, special education’s most basic article of faith, that instruction must be individualized to be truly effective, is rarely contemplated, let alone observed, in most general education classrooms. Mainstream teachers must consider the good of the group and the extent to which the learning activities they present maintain classroom flow, orderliness, and cooperation. In addition, they generally formulate teaching plans that result in a productive learning environment for 90% or more of their students. General education settings are best for learning what most students need to learn. For many of the remaining 10% of students, however, a different orientation will probably be needed. These students need to learn something different because they are clearly not learning what everyone else is learning. Interventions that might be effective for this group of students require a considerable investment of time and effort, as well as extensive support. Special education in a pull-out setting, with its emphasis on empirically validated practices and its use of data-based decision making to tailor instruction to the individual students’ needs, might be better for teaching these students.

3.8  Conclusion

As early as 1979, federal monitoring of state programs was put into place to guard against not only too much segregation of students with disabilities but also “inappropriate mainstreaming” (U.S. Department of Health, Education and Welfare, 1979, p. 39). Although most would agree that students with mild and moderate disabilities should spend a large proportion of the school day with peers without disabilities, research does not support the superiority of any one service-delivery model over another. Furthermore, effectiveness depends not only on the characteristics and needs of a particular student but also on the quality of the program’s implementation. A poorly run model with limited resources will seldom be superior to a model in which there is a heavy investment of time, energy, and money.

Good programs can be developed in any setting, as can bad ones. The setting itself is less important than what is going on in the setting. Reflecting on the 35 years of efficacy research on the settings in which special education is delivered that I have reviewed in this article, what do we know? We know that what goes on in a place, not the location itself, is what makes a difference. We know that you learn what you spend time on and that most students with disabilities will not learn to read or to write or to calculate if they are not explicitly taught these skills. We know that some instructional practices are easier to implement and more likely to occur in some settings than in others. We know that we need more research that asks better and more focused questions about who learns what best where. In addition, we know that we need to explore new research designs and new data analysis techniques that will help us bridge the gap between efficacy findings and decision making on placements for individual students.

In response to the query of what is special about special education, I can say with some certainty that place is not what makes special education “special” or effective. Effective teaching strategies and an individualized approach are the more critical ingredients in special education, and neither of these is associated solely with one particular environment. Educators must also remember that research has shown that typical general education environments are not supportive places in which to implement what we know to be effective teaching strategies for students with disabilities (e.g., Zigmond, 1996).

Considering the research evidence to date, it is clear that placement decisions must continue to be made by determining whether a particular placement option will support the effective instructional practices that are required for a particular child to achieve his or her individual objectives and goals. The search for the best place in which to receive special education services has tended to be fueled by passion and principle, rather than by reason and rationality. Until educators are ready to say that receiving special education services in a particular setting is good for some students with disabilities but not for others, that different educational environments are more conducive to different forms of teaching and learning, that different students need to learn different things in different ways, and that traditional group research designs may not capture these individual differences in useful ways, we may never get beyond the equivocal findings reported here. We may even fail to realize that, in terms of the best place to receive special education and related services, we have probably been asking the wrong questions.