Click for next page ( 75


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 74
4 Effective Instruction for Mildly Mentally Retarclec! Children In Chapter 1 we argued that the quality of instruction in special education programs is one of three key factors that determine whether disproportion should be considered problematic. Chapter 3 presented our view that the justification for assessment procedures derives from their contribution to effective teaching and learning. Thus, instruction for mildly mentally re- tarded children both the quality offered and the setting in which it is best provided are at the fore of the panel's concerns and recommenda- tions. This chapter begins by attempting to specify the characteristics of effec- tive education for mildly mentally retarded children. With these charac- teristics in mind we are able to address two core policy questions: (1) Are separate classes for mildly mentally retarded children needed, or can such children be as well or better served in the regular classroom? (2) Does the mentally retarded label as used in current practice specify unique instruc- tional programs, warranting a separate categorical grouping of children, or would a more general designation be just as useful in delivering educa- tional services? Our question then becomes: What is effective education for mildly men- tally retarded students? The apparent simplicity of this question is il- lusory, and the difficulty of arriving at a simple answer is in great measure The panel would like to thank Gaea Leinhardt, who helped gather evidence and who con- sulted extensively with us during the preparation of this chapter. 74

OCR for page 74
Effective Instruction 75 a function of the difficulty of deciding who is and who is not mildly men- tally retarded. At the very heart of the demand for special education is the assumption that all children do not prosper under identical educational programs. The aim of the enterprise, the reason for elaborate assessment and placement procedures, is to match children and treatments so that each child is treated optimally. By definition, then, what is good or effec- tive instruction is supposed to depend on the kind of children involved. This means that programs can only be evaluated with respect to a properly identified class of children. If an instructional program is not successful in a given case, it might not be a poor program but rather a misapplication to a given child or group of children. Some of the difficulties that we address here emerge from attempts to transform educational practices that were originally based on clinical practice for a highly select population into a special education program for a much wider range of students that must accommodate the bureau- cratic constraints of the public school. In the area of mental retardation, as in other areas of special education such as learning disabilities, many accepted principles of instruction have been based on careful observation and a tutorial type of instruction with highly atypical children. While this knowledge was being applied within the public school environment, changes were taking place in the identified population of exceptional children and in the educational practices that were functional within that setting (Cruickshank, 1967; Dunn, 1973~. For example, clinical popula- tions often include more severely handicapped individuals, while schools enroll children with mild or moderate handicapping conditions; clinical settings are usually able to provide individual tutorial instruction, while financial and organizational factors restrict schools to small-group instruc- tion or separate special classes; clinicians often identify unique diagnostic problems of individuals, while schools tend to recognize more general prob- lems of poor performance. It was not the original intention of special educators that all children with school problems or minor difficulties in adjustment or coping be eli- gible for special education services. In recent years, however, public sup- port for special education has been expanded to include a significant number of children with school problems or behavioral difficulties. Legal requirements and fiscal incentives have moved educators to identify and place more and more students in special education programs (U.S. De- partment of Health, Education, and Welfare, 1979a). As mentioned pre- viously, the jury is still out as to the most effective placement for these students. As we noted in Chapter 2, the variation and changes in the definition of educable mental retardation complicate the task of deciding what is effec

OCR for page 74
76 REPORT OF THE PANEL five education for mildly mentally retarded children, since it is unclear who should be classified as mentally retarded. The research on which we are able to draw has generally accepted the classifications made by school districts and sought statistically significant effects for groups of children with the mildly mentally retarded label. But these groups have usually been more heterogeneous than the common label implies. Since effective instruction for a given child is likely to depend more on his or her actual characteristics as a learner than on the classification as mentally retarded, the reliance on institutional labels to characterize children necessarily limits the kinds of conclusions that can be drawn from this research. A further limitation in the use of existing research concerns historical changes in labeling practices in the schools. There has been a sharp de- crease in the number of students classified as educable mentally retarded (EMR) in the past several years, with a concomitant increase in the num- ber of children labeled as learning disabled (U.S. Department of Educa- tion, 19801. Children who are currently in the EMR category, especially those in self-contained classrooms, may therefore be more disabled than their counterparts of previous years. Research on this older cohort may not be generalizable to the current group of EMR children. Special education for mildly mentally retarded children has grown from the widespread observation that children with generally low mental ability fare poorly in regular school programs. It is generally assumed that such children lack abilities, such as the ability to abstract or to transfer knowl- edge, that are assumed in regular instruction. These children are there- fore expected to profit from an adapted curriculum and teaching pro- cedures that make fewer assumptions of concept mastery; provide more explicit and more numerous examples through concrete experiences; allow more active participation in "hands-on" experiences by students; provide structure into which learners can insert specific information; and include specific efforts to build improved social cooperation skills, self- esteem, and work habits (Goldstein, 1974, 1975~. To accomplish these goals it is assumed that specially trained teachers and/or support staffs are required. Yet these special services can, at least in theory, be provided under a number of different institutional arrange- ments, including (1) the separate class structure, in which children are as- signed to a special EMR class conducted by a specially trained teacher who provides a unique curriculum for the children for a full school day; (2) the resource room structure, in which the basic assignment of the child is to a regular class, but the child is removed for special instruction by a specially trained teacher for a portion of the school day; and (3) the teacher/consul- tant model, in which a specialist advises the regular teacher on special tasks and lessons that can aid the exceptional child, but all instruction is

OCR for page 74
Effective Instruction 77 given in the regular classroom. We consider instruction under all of these arrangements as we attempt to define effective instruction for mildly men- tally retarded students. EVIDENCE ON EFFECTIVE INSTRUCTION FOR MILDLY MENTALLY RETARDED CHILDREN Despite several decades of research it remains difficult to gather definitive evidence on the nature of effective instruction for mildly mentally retarded children. In addition to problems created by shifting definitions of the EMR population? there are reasons for the lack of evidence that lie deep in the prevailing tradition of educational research, a tradition in which re- search on mental retardation has quite naturally shared. Much of this re- search has set out to test whether some new program is better than "stan- dard" practice. Groups of children in the new and the standard programs are compared on some outcome measures, but the programs themselves are not analyzed, nor is the actual functioning of children within them assessed. The result is a "black box" evaluation, comparing outcomes of differently labeled treatments without attempting to determine what fea- tures of the programs or treatments are responsible for the observed out- comes. Indeed, it is characteristic of most of these studies that only the most global descriptions of the educational treatments are offered. Typi- cally, we are able to learn of class size and something about the age and perhaps IQ distributions of the children in the classes. Little detail is of- fered concerning the actual curricula being used, nor are there usually ob- servations of how children interact with teachers, other children, or the curriculum materials. ~ Other methodological limitations in the bulk of the research on instruc- tion for mentally retarded children must also be noted. The most impor- tant are a failure to randomize treatment and control groups, so that sub- sequent comparisons of the effects of treatment can assume equality of initial status, and a tendency to rely on statistical significance between treatments even when differences are too small to reflect important differ- ences in educational outcomes. Like other education research, research on mental retardation has also suffered from a lack of appropriate outcome measures. In most instances, those domains have been measured that This lack of attention to curricula partially reflects the fact that few systematically devel- oped curricula have been available to teachers of EMR classes, forcing them to modify cur- ricula themselves or to develop their own. It was not until the late 1960s that the Office of Education invested in curriculum development for mildly mentally retarded students, and then only to a limited extent.

OCR for page 74
78 REPORT OF THE PANEL could be measured easily. This means that IQ and achievement scores are most often available, whereas changes in personality, behavior, or social processes, which are more difficult to define and measure, are neglected. With these limitations in mind we turn to a consideration of the research on effective instruction for mildly mentally retarded children. We consider first the pervading question of setting do mildly mentally retarded children fare best in separate classes, or do they do better when allowed to remain in regular classrooms? We then examine the specific features of instruction that appear to be helpful for EMR children. This "feature analysis" allows us to raise in a new light the question of whether separate labels for different categories of special students are useful in providing appropriate education for these children. THE QUE STION OF SETTING Until very recently, research on effective education for mentally retarded students generally addressed the question of the kind of administrative setting in which mentally retarded children would fare best. At stake in most studies was the practice of creating separate classes for children identified as mildly mentally retarded. Prom the 1930s to about 1970, most studies shared an initial hypothesis favoring such separate classes a hypothesis dictated by the widely shared belief that mentally retarded children needed both smaller classes and a distinctively different cur- riculum emphasis from that for "normal" children. A respectable number of studies accumulated data comparing the performance of mildly men- tally retarded children in self-contained classes with the performance of such children in regular classes. Several summaries of this literature (Kaufman and Alberto, 1976; MacMillan and Meyers, 1979; MacMillan et al., 1974) indicate that no clear judgment about the two settings for in- struction can be made on the basis of this research. With respect to academic performance (usually measured by standardized achievement tests), there is a slight favoring of regular class placement (e.g., Bennett, 1932; Cassidy and Stanton, 1959; Mullen and Itkin, 1961), but many studies showed no reliable differences between the two placements (e.g., Blatt, 1958; Goldstein et al., 1965~. Beginning in the early 1970s, professional and public opinion came to favor less segregation of the handicapped. The shift in opinion was proba- bly fueled only in small part by the disappointing performance of children in separate special classrooms. A more powerful impetus appears to have been the growing press for fuller participation of all kinds of "minority" groups including the handicapped- in the mainstream of public and social life. Whatever the impetus, the increasing interest in "mainstream

OCR for page 74
Effective Instruction 79 ing" of the handicapped led to a new round of studies, testing the hypoth- esis that mildly mentally retarded children would prosper if they spent all or some of their school time with their "normal" peers. These main- streamed students were not, however, to be left in ordinary classrooms to fare as they might. Instead they were to be identified as mentally retarded, and special services were to be provided either by the regular teacher sup- ported by a specialist or by a specialist teacher with whom the mentally re- tarded child spent part of the day. The separate classroom for mentally retarded children now became the "standard" practice on which main- streaming was to improve. As in the earlier round of research, findings concerning the academic effects of mainstreaming have been contradictory (Corman and Gottlieb, 1978; Heller, in this volume; Jones et al., 19781. There is no clear favoring of either separate classes or full-time mainstreaming; each showed more favorable effects in some studies and less favorable effects in others. The resource room, a special instructional environment to which children are assigned for a part of the day, spending the remainder in the regular class- room, often but not always shows favorable effects in comparison with separate classes and full-time placement in regular classes. But children sometimes do best in regular classrooms in which their own teachers are assisted in providing special instruction. There is some evidence that children with initially higher IQs do better in regular classrooms and that those with lower IQs fare best in separate classrooms. However, even this common sense conclusion cannot be asserted with confidence on the basis of the research to date. Our discussion thus far, like much of the research literature, has focused heavily on academic outcomes. As noted above, however, many studies have included one or more measures of social adjustment or self-concept. On the measures used, especially those assessing children's judgments of themselves, children in self-contained classrooms tended to rate them- selves somewhat better than did children who remained in regular class- rooms (see the review by Heller in this volume). Children in self-contained classrooms displayed more positive self-concepts. In more recent work that compares mainstream with separate class treatment, the results are more contradictory. In both bodies of research, there are major design problems that further confound any effort to decide what the real effects are. These include the problem of instrumentation there is little una- nimity in the field as to what a good self-concept is or how to measure it and the problem of finding appropriately matched groups for the various treatments. Often, the mainstreamed children were those who, in the judgment of various professionals, were more competent and were believed to be better able to function in the regular classroom. This makes

OCR for page 74
80 REPORT OF THE PANEL comparisons, particularly on measures of social adjustment and self- concept, virtually impossible; techniques of covariance may be difficult to apply because the measures do not meet necessary scaling assumptions. Finally, most of the research available that is relevant to this question is, like the research on achievement outcomes, a "black box" with respect to the actual treatment involved. FEATURES OF EFFECTIVE INSTRUCTION The most obvious conclusions from these kinds of inconclusive findings over several decades of research is that the instructional setting per se does not matter, that mildly mentally retarded children can do equally well or equally poorly in both kinds of settings. Yet this finding may mask some very real and important regularity in effects on children. Perhaps there are features of the educational treatment received by mentally retarded children that do systematically affect outcomes but that are not uniquely associated with any particular setting for instruction. Perhaps those studies that show a benefit for one setting or another were comparing programs with some specific features that are responsible for the effects. Reported as a com- parison between self-contained and regular class or mainstreamed settings, we learn nothing from these studies about what these features might be. Fortunately, a few recent studies offer descriptions of the educational process that are detailed enough to permit us to address the question of which features of instruction seem to be beneficial for mildly mentally retarded children. While the number of such studies is not large, there is substantial consistency in what has been found to be effective instruc- tional practice for children with the mildly mentally retarded label. Academic Outcomes Several studies have documented academic gains for EMR children through the use of individualized "behavioral" methods of instruction (Bradfield et al., 1973; Haring and Krug, 1975; Jenkins and Mayhall, 1976; Knight et al., 19811. In the instructional programs studied, work assignments were given on a daily basis so that the teacher rather than the child deter- mined the pace of work; a mastery learning approach was used in which detailed records and charts of progress (usually based on tests directly covering the curriculum content) were kept for each child; systematic rein- forcement was used, and significant amounts of one-to-one instruction, sometimes by peer tutors, were offered. In general, these procedures re- sulted in larger amounts of time spent on academic work and in a heavy over- lap between what was taught and what was tested in the instruments used to assess academic progress.

OCR for page 74
Effective Instruction 81 Although the practices described in these studies did not create academic stars of EMR-labeled children, clear learning benefits were achieved. It is striking that the settings in which these treatments were carried out varied from the self-contained EMR classrooms (Haring and Krug, 1975) to resource rooms (Jenkins and Mayhall, 1976) to regular classrooms (Bradfield et al., 1973; Knight et al., 1981~. This fact, al- though based on a small number of studies, offers striking confirmation of the conclusion reached by Corman and Gottlieb (1978:257~: "As a whole, these studies [of effectiveness] suggest that particular instructional tech- niques may be of greater relevance to improved achievement than the fact that these techniques are used in one of many possible integrated settings." A striking characteristic of the list of features associated with effective academic skill instruction for mildly mentally retarded students is its similarity to the features identified for other categories of children in aca- demic difficulty and, indeed, for the school population as a whole. In the "effective schools" research (e.g., Brookover and Lezotte, 1979; Venezky and Winfield, 1979; Weber, 1971) features of school organization that are associated with good academic performance among poor and minority children include an emphasis on the direct teaching of basic skills and the frequent assessment of progress. Both are also features of effective in- struction for mildly mentally retarded populations. A number of large-scale studies (e.g., the Beginning Teacher Evalua- tion Study [Fisher et al., 1978i, the Follow Through Evaluation [Stebbins et al., 1977], and the Instructional Dimensions Study [Cooley and Lein- hardt, 1980~) identify features of classroom organization and process that are associated with good academic performance in schools with high pro- portions of children receiving compensatory education (i.e., poor and mi- nority children). These studies, all conducted in large numbers of class- rooms, took advantage of naturally occurring variations in instruction, rather than attempting to use control groups, random assignments, and other characteristics of experimental designs that can only be approx- imated in real school settings. They converge on a set of descriptors of "direct instruction" (see Rosenshine and Berliner, 1978) that include high content overlap between learning activities and criterion (test) tasks, built- in formal assessment techniques, increased time on academic tasks, teacher pacing, and the use of motivating management systems (i.e., some form of contingent reward). Social Outcomes As we indicated earlier, the rationale for special education for mildly men- tally retarded students includes, even stresses, the social goals and out- comes that should be part of an educational plan for such children. The

OCR for page 74
82 REPORT OF THE PANEL theory has been that mentally retarded children. require special social en- vironments for two different but related reasons: (1) they tend to interact poorly with "normal" children, to experience rejection, and, in part as a result, to develop weaker self-concepts and (2) they lack certain specific social and adaptive skills and require special training in these that is not necessary for other children. A line of research that avoids the "black box" problem in that it is directly concerned with techniques for training social skills has been recently reviewed by Gresham (1981~. Gresham summarizes a large num- ber of studies that examined training techniques derived from social learning theory. The focus in these training efforts was on various aspects of social behavior as actually observed in the classroom and on social ac- ceptance by peers (using peer sociometric ratings and teacher ratings), with little attention to the less easily definable construct of self-concept. Many of the training techniques studied have been viewed as suitable or necessary only for the severely disabled or sometimes the institutionalized mentally retarded population. For this reason a large portion of the research has been conducted in separate classes rather than in main- stream settings as well as with populations not directly relevant to this panel's concern. A large segment of the research on what are termed social skills has really been directed at increasing "classroom appropriate" behavior (staying in one's seat, attending to the assigned task, not talking out, etc.) or at minimizing disruptive behavior, rather than at building social interaction skills or enhancing peer acceptance. Research dealing with mildly handicapped populations (including but not usually limited to EMR children) suggests that techniques such as ar- ranging game playing to include the handicapped child and having peers initiate social interaction can increase interaction and peer acceptance (Aloia et al., 1978; Ballard et al., 1977~. More direct teaching of social skills for example, by providing competent models, rewarding the models, having children rehearse the social skills, and providing feed- back has been shown in a few studies to build certain social skills (e.g., Bondy and Erickson, 1976; Cooke and Apolloni, 1976~. There is little evi- dence, however, for the generalization and maintenance of these skills be- yond the training setting. Furthermore, while a considerable body of re- search points toward the general effectiveness of behavioral and social learning methods, there is a paucity of demonstrations of effectiveness in actual classrooms for mildly mentally retarded children. A few comprehensive intervention programs for mildly mentally retarded students have focused on social skills. Perhaps the best known is Gold- stein's Social Learning Curriculum (Goldstein, 1974, 1975~. The principal focus of the curriculum is on the promotion of socially adaptive behavior,

OCR for page 74
Effective Instruction 83 accomplished by teaching children to think critically and to act in- dependently. The teacher transmits content through a special inductive teaching methodology (ITM), which aims to induce systematic and self- conscious problem-solving behavior. Although extensive field testing has been undertaken the Social Learning Curriculum has been introduced in approximately 300 classes in 29 states much of the work conducted during the field testing focused on testing the theoretical assumptions un- derlying the curriculum, developing new units, and revising others, and no summative evaluation data were collected. In general there does not seem to be as clear a set of conclusions to draw about the effective teaching of social skills and the promotion of social ac- ceptance of mildly mentally retarded children as there is for academic development. Cognitive Process Skills There is a line of instructional research on mentally retarded children that has been increasingly prominent in recent years and that may have impor- tant practical applications in the future. This is research on the direct training of those cognitive abilities that are thought to underlie the men- tally retarded person's difficulties in learning under ordinary school con- ditions. The first phase of research on process deficits in mentally retarded students largely served to identify specific processing skills that were weak in children with low IQs. Prominent among such skills were rehearsal and other techniques of memorizing that were shown to be spontaneously used by normal children but not by mentally retarded learners. Several in- vestigators showed that mentally retarded individuals could be trained to use various mnemonic techniques. However, in study after study it was found that the newly acquired learning skill was applied only to the specific task for which it had been trained, that there was little or no generalization and thus no general improvement in the cognitive function- ing of the trained individuals (e.g., Brown and Barclay, 1976; Butterfield et al., 1973; Engle and Nagle, 1979; Turnure et al., 19761. More recently, a few studies showing some generalization and maintenance of learning skills of various kinds have begun to accumulate (Belmont et al., 1980; Brown, 1978; Chipman et al., in press; Segal et al., in press), and there is new optimism in some quarters about the potential for actually improving the cognitive functioning of mentally retarded learners. Most of the studies clearly showing the acquisition of learning skills have been conducted with small samples under laboratory-like conditions rather than under normal school conditions. However, several programs currently being tested and refined in school settings have strong learning

OCR for page 74
84 REPORT OF THE PANEL skill/problem-solving orientations. These include the Instrumental En- richment Program (Feuerstein et al., 1980) and other programs that teach inductive problem-solving skills to mentally retarded learners (e.g., I. L. Smith, 1980~. A shared feature of the instruction in these programs is their focus on teaching children to monitor their own thinking and to plan strategies for learning and remembering as well as to solve social prob- lems. All the programs rely heavily on social interaction between the stu- dent and a highly skilled, specially trained teacher. Discussion and analysis of problems and learning tasks seem to be required but are difficult to ar- range in a self-study mode and seem to require the grouping of children ac- cording to their need for learning skill instruction. As research continues and as more extensive field data on these programs become available, learn- ing skill procedures may emerge as a supplement or alternative to individual mastery-oriented direct teaching of academic skills. CONCLU SIONS INSTRUCTIONAL SErrING What conclusions can be drawn from the research literature concerning the appropriate setting and instructional processes for mildly mentally retarded students? On the whole we are forced to conclude that adminis- trative setting, in and of itself, does not determine whether an educational program is effective or appropriate. Rather it is the things that go on in that setting that matter. In principle, any setting can serve as an ap- propriate educational environment for mentally retarded children if cer- tain principles of instruction are observed. Many observers agree that because of the belief that mentally retarded children cannot learn well, less is often demanded of them than might be. In classes for mentally retarded students there is little "cognitive press" (Leinhardt and Pallay, 1981) and often a sharply reduced curriculum, so that children in these classes are deprived of the opportunity to learn standard academic skills. There is no intrinsic reason why the cognitive press of a separate class for mentally retarded children cannot be increased. Nevertheless, a classroom of children bearing the label mentally retarded does not typically seem to evoke high expectations and, therefore, the academic demand on them may be reduced (Fine, 1967; Heintz, 1974; Meyen and Hieronymous, 1970; Salvia et al., 1973~. Unless this tendency can be overcome, it argues not only for reduced use of separate classes but also for reduced use of the label mentally retarded, since it seems likely that the tendency to lower cognitive demands would be applied to individual children as well as to groups. But

OCR for page 74
Effective Instruction 85 there may be other reasons for lack of cognitive press than lowered expec- tations. If, in a class of children even a small class all children require a great deal of teacher attention in order to stay "on task" and thus make reasonable cognitive gains, it may be difficult or impossible to set high standards for the rate of progress through a curriculum. This argues for either heterogeneous class grouping, in which only a few children need substantial and frequent attention, or a tutorial-like setting, in which a single child at a time can be attended to. The regular classroom provides a heterogeneous setting, but there is some evidence that, except in certain specially designed individualized settings, the great heterogeneity coupled with the larger class size often double that of the special classroom- makes it difficult for the special child to receive adequate attention. Some kind of identification of the child is required if he or she is to receive the special attention needed. A practical solution in some cases seems to lie in the resource room, a special teaching/learning laboratory to which the child identified as in need of special help is assigned for a limited period each day, in which instruction is given on a one-to-one basis or in very small groups, and adequate monitoring and rewarding by the teacher (or paraprofessional or peer tutor) is possible. However, to be assigned to this resource room, some kind of identification procedure is needed. Thus it appears that except perhaps in specially designed mainstream classrooms a complete absence of labeling would also imply an absence of the kind of special instructional treatment needed by the child. CATEGORICAL LABELING Some form of identification of children is likely to be required if they are to receive the kind of special education services that they need and to which the law entitles them. The identification of children in need of special services does not necessarily imply that distinct categories of hand- icaps need to be specified or that special education services should be delivered according to the categorical label that a child carries. Current special education practice as well as much theory divides children with academic difficulties into several categories, the most important of which are the mentally retarded and the learning-disabled (LD) categories. To what extent does the evidence on effective instruction support this prac- tice? That is, do EMR and LD children profit from distinctly different in- structional treatments, or do the same features of effective instruction ap- ply to both groups? An extensive body of theory discriminates LD children from mildly mentally retarded children (Cruickshank et al., 1961; Lerner, 1976;

OCR for page 74
86 REPORT OF THE PANEL Strauss and Lehtinen, 1947~. While the mentally retarded group is seen as being generally low in all aspects of mental functioning as well as having difficulties in social adaptation, LD children are expected to show uneven profiles of abilities (being strong in some areas and weak in others) and to have IQ scores higher than those of EMR children. Further, social skills are not identified as a major weakness in this category of children. The uneven profile of a LD child points, according to the theory, to an instruc- tional program that is specifically adaptive to particular areas of strength and weakness. A dominant instructional model for LD children involves differential diagnosis and prescriptive teaching aimed at weaknesses in such areas as psycholinguistic skills, perceptual skills, motor skills, and the like. The underlying theory is that, through correction of these cognitive skill deficits, the child's ability to learn school subjects will im- prove. A wide variety of programs designed to implement this instructional theory has been developed. It is difficult, however, to assemble strong evidence for the effectiveness of these programs in improving academic skills. While some of the identified cognitive subskills have been shown to be amenable to improvement through instruction (e.g., Kavale, 1981), there is little evidence to date that such training transfers to academic skills such as reading or mathematics or that teaching methods that adapt to skill deficits by making use of strong cognitive skills are more effective (Arter and Jenkins, 1981~. A small but respectable body of evidence is available suggesting that direct instruction in academic subjects is effec- tive for LD children (Baseman, 1979; Leinhardt and Pallay, 1981~. The key features of this direct instruction are shared with those identified as effective for mildly mentally retarded children. On the basis of documented effective practice in schools to date, it ap- pears that basically the same kind of instructional processes may be needed for LD children as for mildly mentally retarded children. It should be noted that there is at least one other large group of children with academic difficulties who do not acquire special education labels but who never- theless receive special instructional services in their schools. These are the children who by reason of low family income and poor performance on achievement tests are assigned to various compensatory education pro- grams usually in particular academic subjects for a part of each school day. The accumulating evidence about these children also suggests that the same features of direct, externally paced, and formally monitored in- struction in academic content that have been noted for mentally retarded children produce the best learning results (Leinhardt et al., in press). If these three theoretically distinct groups of children seem to-prosper best under the same kind of instruction, there is good reason for calling

OCR for page 74
Effective Instruction 87 into question the traditional system of categorical labeling within special education. At the very least, the burden of proof now seems to lie with those who would defend the traditional divisions within special education. SUMMARY AND SOME CAUTIONS The current evidence on instruction for mildly mentally retarded students seems to offer some clear directions for policy and for classroom practice. First, we can find little empirical justification for categorical labeling that discriminates mildly mentally retarded children from other children with academic difficulties, such as LD children or children receiving compen- satory education.2 Second, while there are fewer well-documented studies with clear results than we might wish, the weight of the evidence clearly points to a group of instructional practices that seem to benefit all of these types of children. Intense direct instructional methods, described earlier in this chapter, seem to be applicable in a variety of settings, from the separate special classroom to the mainstream classroom, and they are not different in spirit from the methods that appear to have been generally ef- fective in schools that serve children with poor prognoses for academic success. This similarity in the features of instructional treatments offers some hope that some proportion of the children now recognized as in need of special education might be reduced through the provision of more effec- tive regular instruction, especially in schools with high minority represen- tations. Of course, there is nothing in the evidence to date to suggest that an important subset of children who need more intensive attention, and thus more resources than the ordinary classroom is able to provide, will not continue to exist. Providing adequate services to these children will probably require some kind of identification and hence labeling. The labels need not categorize the children but can instead describe the types of special intensive instruction they need. The question of appropriate setting for instruction appears to be one of administrative manageability rather than one of instructional theory. In keeping with the general public sentiment favoring a minimum of social separation between different segments of the population, there should probably be some favoring of mainstream classroom or resource-room ar 2As we have noted, this statement refers explicitly to mildly mentally retarded children. Re- cent practice, responding in part to legal challenges to EMR placement for minority children, has in some states and local areas tended to reserve the EMR label for children who show very serious and sustained learning difficulties. The available research, by contrast, is based on a much more heterogeneous group of children that includes many with only mild dysfunctions.

OCR for page 74
88 REPORT OF THE PANEL rangements over separate classes. This does not mean, however, that children in need of intensive help should simply be put back in the regular classroom without recognized special status and without appropriate as- sistance for the classroom teacher. In planning instruction for the special child, primary attention should be directed to the specific features of the instructional treatments that have been identified as fostering academic progress in children with initial poor performance. Although these broad conclusions seem to be well supported by the evidence at hand, we believe it is important to point to some cautions that must be kept in mind in formulating a policy that may well have far-reach- ing and long-lasting effects on the kinds of educational opportunities and services offered to children. The caveats that should be kept in mind are discussed below. MASKING INDIVIDUAL DIFFERENCES At the beginning of this chapter we indicated that most research on in- struction for mildly mentally retarded students has proceeded as if the children with the EMR label were homogeneous with respect to cognitive capabilities and instructional needs. One possible effect of such research, which treats heterogeneous groups of children as if each had the same needs and capabilities and of examining only the group effects of in- struction is that so much error variance is produced that potentially real differences in program characteristics that benefit children are statisti- cally masked. This may be part of the reason for the preponderance of findings of no difference in the instructional effectiveness literature. If the definitions of mild mental retardation and learning disability were to be made tighter in future research- so that only individuals who were clearly those hypothesized to benefit most from a particular treatment were in- cluded in an evaluation we might begin to obtain a clearer picture of ef- fects. Such a trend in research findings would surely temper the conclu- sion that there is little basis for distinguishing between mildly mentally retarded children and others with academic difficulties. On the other hand, in the course of further specification of who is to be considered an EMR or an LD child, it is to be expected that potentially important changes in the current definitions of mild mental retardation and learning disability would be suggested. Thus, there is little likelihood that such research would end up supporting current categorical labeling practice, although it might provide confirmation of some of the theoretical distinctions that experts in special education now offer. In any event, what seems crucial is that any policy of decategorization adopted in response to the current scientific evidence should not be constructed so as to actively

OCR for page 74
Effective Instruction 89 prohibit the kinds of research on differential instruction that would be re- quired to arrive at relevant distinctions among children with academic dif- ficulties. UNKNOWN EFFECTS ON OTHER CHILDREN Integrating more special education students into regular programs may affect the achievement of the "normal" students. The limited data that are available on the effects of mainstreaming on children in regular pro- grams suffer from the same problems that apply to the literature we re- viewed on the effects of instructional setting on EMR children (see Heller in this volume for a review of the existing data on this topic). More critical from our standpoint are the effects of instructional processes that appear to benefit low-achieving children on students in the average or higher ranges. Along with research that identifies specific features of effective in- struction for the special child should be an equally direct look at the ef- fects of these features on other students in the classroom as well. Research in two related areas the effects of grouping by ability or "tracking" and aptitude-treatment interactions may shed some light on this issue. For reviews of the literature on those areas, see Calfee and Brown (1979), Cronbach and Snow (1977), and Esposito (1973~. BEHAVIORAL BIAS IN RE SEARCH Behaviorally oriented, direct instruction approaches have clearly emerged as the direction of effective practice in research to date, although there are reasons to remain open to changes in the weight of evidence in the longer run. First, for a variety of reasons rooted in both scientific and social value systems of the past two decades or so, behaviorally oriented researchers have focused more on academic skills and on clear outcome measurement and reporting than have other groups of researchers concerned with the same broad issues. For this reason, their work has had clearer, better doc- umented results than some potentially competing or supplementary ap- proaches. For example, the direct instruction approach as it has been used and documented to date favors a step-by-step, practice-oriented ap- proach to education. Approaches other than direct instruction have been less well analyzed and documented at the present time; nevertheless, they may also be effec- tive. For example, several programs (e.g., SEED, Renee Fuller's reading program for mentally retarded children) that claim strong results rely less on step-by-step methods and more on the general reasoning skills of students. Strong evidence other than the claims of those involved and of

OCR for page 74
Do REPORT OF THE PANEL occasional observers for the effectiveness of these programs with mildly mentally retarded students is not available at this time. Nevertheless, these programs and others like them deserve careful investigation. The results of such investigations may lead to clearer specification of when and for whom behavioral step-by-step methods are needed and when other ap- proaches which may have important "fringe benefits" in the kind of general adaptive capacities that they promote may be preferred. Simi- larly, the cognitive-process training programs discussed earlier in this chapter also suggest an alternative or supplement to direct step-by-step in- struction in academic skills. Again, no strong evaluative evidence is cur- rently available concerning these approaches. However, it is important that they continue to be investigated and that practical policy be for- mulated in a way that remains open to the implementation of the findings that emerge. EVALUATION CRITERIA As we have noted, the research evidence on which conclusions concerning effective education are based relies heavily on a particular set of outcome criteria. These can be characterized as oriented to "basic skills": the cen- tral basis for deciding what features constitute effective instruction has been their contribution to improved performance on mathematics and reading tests of various kinds. Neither social outcomes nor other kinds of learning e.g., the acquisition of knowledge relevant to functioning in a job or using various community resources has received an equivalent amount of attention. Similarly, certain characteristics of individuals for- merly educated in EMR programs, such as their employment, earnings, family lives, etc., may be sensitive indicators of the effectiveness of EMR programs yet remain at this time largely uncharted. A focus on basic academic skills as a criterion is appropriate for a popu- lation whose major reason for referral to special education is academic dif- ficulty. This is particularly true for younger children perhaps ages 8 through 12 or 13-when there is reason to hope that with intensive in- structional efforts the child can return to a regular classroom program with a competent level of basic skill performance. For children who con- tinue to have difficulty in acquiring basic skills, other educational goals and curricula, especially those related to specific vocational and social adaptive skills, take on increasing importance. It may well be the case that a differentiated set of outcomes for older mildly mentally retarded chil- dren will prescribe somewhat more education in separate classes than is necessary for younger children who have recently been identified as having academic problems. The vast majority of the research that we reviewed

OCR for page 74
Effective Instruction 91 has been conducted on children younger than age 12 or 13. The appro- priate instruction of and placement for special education students at the secondary level is a largely undiscussed issue without firm research under plnnlngs. CONCLUSION For all of these reasons the panel believes it is essential that a clear distinc- tion be made between recommendations for current "normal" practice and those for investigation that may eventually lead to changed views of "best" practice. While the educator facing the practical challenge of of- fering immediate educational services to children will do well to incor- porate the features of direct instruction that have been outlined here, the total educational system must continue to be open to efforts to determine still better procedures, even if these point toward complex revisions in cur- rent practice. Thus, we do not recommend any single structure for the or- ganization of special education. Rather we endorse a policy that allows for new approaches side by side with vigorous application of our best current knowledge about effective instruction.