Problems and Issues

The Political Context

Organizational Arrangements

Program Outcomes

An Agenda for Action

Conclusion

References

All children and youth are expected to be in school these days, at least through high school graduation.  The drive for universal enrollment has brought about increased diversity in characteristics of students, which in turn has led to changes in the schools to accommodate differences as well as producing strains all around.  Clearly, "bowling alley" school programs-just sending instruction "down the middle"-do not suffice today.  This publication tells about students at the margins, those most likely to be neglected unless quite extraordinary efforts are made.

 A common method of dealing with diversity is to create categorical programs that "set aside" to the margins students who stretch existing school programs.  Some of the categorizing is dimensional, that is, it is based on cutoff points along a continuum of differences among students.  For example, millions of children have been placed in special classes because their IQs are below 70 or 75.  As tests of "g" (general mental ability) came into common use, it was but a small step to categorize those with low IQs as mentally retarded and then set them aside.  Serious doubts are being raised today about such classification and grouping procedures, especially those that involve formal testing.  According to Glaser and Silver (1994, p. 397), "Tests are frequently used to sort students into instructional tracks that provide differential opportunities and differential expectations. . . .The general failure of this process to increase educational outcomes has been amply demonstrated." More acceptable are cutoff points that occur quite naturally in connection with curricular progress.  For example, enrollments in third-year Spanish are limited to students who completed second-year Spanish successfully, and enrollments in geometry are limited to those who completed the preceding class in algebra.

Other categorical programs are typological rather than dimensional.  The program in braille, for example, involves a distinct mode of instruction in reading for students who have extremely limited vision.  This may or may not involve a segregated class or school; the braille instructor may come to the regular classroom to offer instruction and related supports.  Instruction for students who are deaf necessarily circumvents aural modes of communication and involves distinct methods invented for this purpose.  School programs for students showing severe and profound limits in mental ability also require distinct curricular elements.  Programs for these three groups (blind, deaf, and severely retarded) were the first to be developed in special education, and they remain the clearest instances of distinct categories of special education.  The field of speech correction also has a number of distinct methods for treating specific conditions, such as cleft palate, and so may be regarded as mainly typological.  In these several instances, the special education categories meet the ATI (aptitude-treatment-interaction) test, that is, there are distinct treatments (or instructional alternatives) and ways of assessing individuals such that differential allocations to treatment can be made with validity.

A special panel created by the National Academy of Science (NAS) to study placement practices in special education proposed the following approach to student classifications: "It is the responsibility of a placement team that labels and places a child in a special program to demonstrate that any differential label used is related to a distinctive prescription for educational practices . . . that lead to improved outcomes" (Heller, Holtzman, & Messick, 1982, pp. 101-102).  This recommendation implies two standards for categorical school placement: first, that there be a distinctive aspect of the instruction prescribed for those in the category; and, second, that there be evidence that the special program "works" to enhance the life and learning of the students given the special placement.  Only a limited number of categorical school programs, such as those designed for meeting the needs of children who are blind, deaf, or severely retarded, meet those standards.  For at least three-fourths of special education students, mainly those with mild disabilities and those enrolled in early versions of the Chapter 1 program, current practices do not meet the NAS standards. (Note: Chapter 1 was recently reauthorized as Title I, and will hereafter be referred to as such.) This does not mean that such students do not have significant learning problems; rather, it is the validity of partitioning practices, the forming of student groups by category, that has become doubtful.
 

The U.S. Department of Education lists a number of categories of special education that are not linked to distinctive interventions that lead to improved learning outcomes.  Table 1 lists these categories and several others used by federal education authorities.  Many states have added still more categorical programs.  The Title I program for economically disadvantaged students, for example, is the largest federal educational program in terms of dollars spent, distributing about $6 billion per year.  But that figure is only marginally larger than the amount of money (more than $5 billion per year) spent on categorical education programs in just one state (California).  A variety of special programs are operated for children whose primary language is other than English.  Some states, such as California, Texas, and Florida, provide large numbers of language-related special programs, mainly because immigration to the United States in recent years has been higher than ever before in terms of raw numbers.

The situation of categorical programs becomes complicated when there are multiple sources of funding and control.  A recent report tells of 44 federal sources of funding for early intervention programs in the schools, and of 25 laws and programs addressing the same target population.  Without doubt, the myriad categorical programs and countless sources of support for them contribute to the fractionation of school organizations, especially those in inner-city schools and others that serve many students with special needs.
 

Some degree of classification of students is, of course, inevitable.  Students are not assembled randomly.  The most common approach is based on chronological age.  The age-grade system is applied universally and is not highly controversial, but other approaches are applied selectively and create problems.  Clearly, classification is both a scientific matter and a social act.  But often the processes involve too little science and doubtful social benefits.  Following is a brief discussion of some of the problems and issues of classification.

Taxons versus Classes

It is sometimes assumed that the categories of students as established in the schools are basic or taxonic, that is, "carving nature at its joints." The assumption is that the categories would exist whether or not the schools recognized them, but this is often untrue.  Some years ago a presidential advisory group published a report on the "six-hour retarded child," implying that some children are retarded only in school.  It is appropriate to see most school-related categories as classes of convenience rather than taxons.  Who may be regarded as mentally retarded, is decided somewhat inconsistently, although not in a totally arbitrary way, over time by committees of professionals.  Before a child is classified in the largest special education category, learning disabled, there must be evidence of a discrepancy between the child's expected achievement and actual achievement. Such discrepancies could be said to reflect an overprediction (and expectation), thus putting the onus on the psychologist who makes the prediction rather than on the child.  In any case, the category is not taxonic; even if it were, there might be no related and distinct treatments available for application.

Categories Beyond the Schools

Increasingly, schools are linking their work with that of social and health agencies.  But each agency has its own categories, often based in law or on the standards of the particular professionals they employ.  Mental health clinics, for example, may require classification of patients in accord with the diagnostic manual of the American Psychiatric Association.  Educators were surprised by a recent edition of such a manual in which the number of childhood mental disorders had tripled since the preceding edition; the manual included "mental disorders" relating to specific academic deficiencies such as in reading.  The problem is that the mental health committees, for example, may arrive at different formulations than do the school-related committees.  None of the diagnostic categories may be taxonic or useful in cross-professional work.  These are problems that will need to be settled as schools and other agencies seek to collaborate in providing services to children and families.

Reliability

Classifications of students, especially those who show only "mild" disabilities, are quite unreliable.  It has been estimated that 80% of all students could be classified as learning disabled (LD) according to one or more of the more than a dozen LD assessment procedures used in various school districts.  In the early 1970s, the numbers of students labeled mentally retarded declined by hundreds of thousands nationwide after a challenging and persuasive court case.  In the same period, the numbers of students labeled "learning disabled" increased sharply, showing how sensitive classification procedures are to the social climate and how unreliable they often are when considered only in their technical aspects.

Stigma

 Some student classifications result in labels for particular students that become public and stigmatic.  "Mental retardation" and "emotionally disturbed" tend to be especially stigmatic and are often resisted and resented by parents.  Unfortunately, rates of stigmatic labeling often correlate with racial differences, as do so many other negatives about schooling, such as suspension and expulsion, school dropouts, poor attendance, and low achievement.  This adds complex tensions and issues.
 

Categorical programs usually get started through political action by groups of parents and other advocates.  Much of the early energies for categorical programs came from the National Association for Retarded Children (NARC).  The earliest work of NARC focused on the federal courts; later attention turned to legislation.  NARC pointed to neglect of retarded children in the schools and then made a claim for legislation supporting new and expanded programs.  The political action was based only minimally on research evidence.  Virtually no data showed that efficient treatments were available.  As in the NARC case, schools often are pressured to organize programs that have no clear foundation in a knowledge base.  Equally, the process often results in distinctions among programs that relate to social issues but that have no related and valid distinctions in instructional practices.  This is the case, many observers believe, in the distinctions between Title I and learning disability programs.  In both cases, teachers are mostly engaged in teaching reading using a common knowledge base, but the two programs often operate separately.

In the case of the NARC advocates, most were relatively privileged middle-class people, but the consequences engendered by their work included the massive development of special classes in neighborhoods serving poor, minority families.  The parents of these children do not always welcome the labels given to their children, and the special classes do not seem to enhance their children's learning.  These findings tend to lead to complex court proceedings that stall, and sometimes reverse, program development.

There is growing interest in the concept of consequential validity, in which "the uses and interpretations of the results of an assessment . . . are validated rather than the assessment itself" (Linn, 1994, p. 6).  In the case of special classes for mildly retarded children, the assessment processes related to placement usually involve a construct of intelligence that has validity in a limited sense, but the processes lack validity in terms of consequences.  The emerging consensus on the importance of consequential validity requires that school psychologists give careful attention to the validity of programs as well as to the constructs represented by the assessment tools they use.

0nce categorical programs are established and money is flowing in support of them, there appears to be a hardening of the categories.  That is, it becomes harder to undo them, regardless of the evidence concerning their value.  However, at this time categorical programs are coming under close scrutiny, perhaps mainly because of their high and growing costs.  Many school administrators and political leaders feel that funding for categorical programs, often protected by rules and regulations, is causing retrenchments in regular school programs.
 

The Continuum

 Figure 1 (p. 10) is typical of a schematic used in special education to represent the organization of programs and the major policy.  At the base is the regular or general education classroom that is expected to serve many exceptional students.  According to a recent report by the U.S. Department of Education (1994), 34.9% of students identified as disabled were served mainly in regular classrooms.  This method might involve consultation with the regular teacher by special education teachers, the school psychologist, or others, or it might involve co-teaching in a common environment by the regular and special education teachers.  At a second level is the resource room, usually used only for part-time placements, for individual or small-group instruction.  The recent federal report shows that 36.3% of disabled students received their special education mainly in resource rooms.  Use of resource rooms entails much movement by students between regular classes and resource rooms, often on a prescribed schedule, which can cause disruptions of routines in both settings.  Figure 1 shows other organizational elements, the order of which depicts their degree of separation from ordinary school arrangements; the schematic also notes the rates of enrollment at each level.  As one moves up the continuum, management and instruction come increasingly under the control of specialists, students are increasingly isolated from non-disabled students, and programs become more expensive.

In the most thorough of cost studies (Moore, Strang, Schwartz, & Braddock, 1988), it was estimated that average annual expenditure per pupil for special education in resource rooms was $1,325; in special classes, $4,233; and in residential programs, $28,324.  Overall, the per-pupil costs for special education were found to be about 2.3 times the cost for non-disabled pupils.

In the 1990-91 school year, the estimated "regular" per-pupil expenditure for K-12 education across the nation was $5,266; the estimated excess cost for special education was $6,846?yielding a total per-pupil expenditure for special education of $12,112. Multiplying that cost by the 4.8 million students enrolled in special education in the same year yields a total expenditure for special education students of about $58 billion.

Along the sides of Figure 1 are notes expressing the least restrictive environment (LRE) principle.  That policy is expressed in law, but is qualified and made arguable by another principle that requires education to be individualized and appropriate.  The LRE idea is a familiar one in other fields, such as mental health, suggesting that as a public agent enters the life situation of an individual, it must do so in the least intrusive way feasible.  Similarly, a first duty of educators is to try to make appropriate arrangements for exceptional students in the regular class environment and to use other elements of the continuum only for compelling reasons and in time-limited ways.  Following the LRE principle through time, it should be expected that more students and programs would move down the continuum, thereby concentrating more special education in regular classes and schools.  Although that is occurring, many argue about how far and how fast the movement toward full inclusion in the mainstream should proceed.  The main focus of disagreements is on severely disabled students.  Should they be placed in regular classes?  If so, what are the consequences for the non-disabled students?
 

Outcomes of categorical programs may be considered at three different levels: (1) for each individual; (2) for each program at the local level; and (3) for each set of programs at some broader level, such as for an entire state or the nation.  In the case of special education, it is required that each student have an individualized plan for instruction and related services and that a review of progress under the plan be conducted periodically.  That is a valuable feature of special education operations, but it is not a focus of the present statement.

Even if broad evidence about a set of programs is positive, the local situation may be deficient.  Or, a particular local teacher may be especially competent and conduct a program of great value.  Thus, it is important that those who make decisions about placements of children be aware of the quality of local programs.  This is an aspect of the consequential validity concept; ultimately, all program validity is local.

It is useful to have outcome data at the third or broadest level, however, to help inform judgments about the general validity of each of the categorical programs and the policies that support them.  Achieving broad and valid evaluation studies is very difficult.  Experimental designs are preferable, but it is difficult to achieve random assignments of students to experimental and control conditions, all with the informed consent of parents.  Another difficulty concerns separation of focus on programs from focus on administrative arrangements.  Very often, only the administrative arrangement is examined.  One might, for example, conduct a braille program at any one of the levels represented in the continuum (see Figure 1).  Or, an LD program might be conducted in a resource room or by co-teaching approaches in a regular classroom. The inclusiveness of the organizational arrangement is one matter; the distinctiveness of the instructional program is something else.

Early meta-analytic studies of outcomes in special education, which focused mainly on classes for students with mental retardation, were extremely crude and do not merit review here.  A study of higher credibility, covering 11 special education studies reported from 1975 to 1984, produced an effect size of 0.44 for academic effects and 0.11 for social effects, both favoring inclusive organizational arrangements.  That is, regular class outcomes were superior to outcomes in special placements.  Several categories of special education, representing mild disabilities, were included.  A similar meta-analysis covering a set of 13 studies from 1983 to 1992 showed effect sizes of 0.08 and 0.28 for academic and social effects, respectively (Baker, Wang, & Walberg, 1995).  These results suggest small advantages for serving exceptional students in regular classes rather than in separately organized arrangements.  There is much debate about these matters, but no evidence of dramatic advantages either way.  Therefore, one might invoke the LRE principle and recommend, to the extent possible, retention of disabled students in regular classes.

But no placement can be judged satisfactory unless and until highly effective instruction that meets individual needs is provided.  Many regular teachers do not tailor instruction to meet individual differences.  In such circumstances, learners who have fallen behind tend to become passive or inactive learners, interacting with teachers at relatively low rates.  Gifted students may languish in boredom. All of that will need to change. By shifting some of the resources associated with separate categorical programs to the regular classrooms, perhaps the broader tailoring of instruction can be accelerated.  That is the idea advanced by advocates of the much-debated regular education initiative (REI).

Numerous evaluation studies of Title I programs have been conducted, many showing small positive short-term achievement effects.  Most such studies were of programs in which selected students were served in pullout programs that emphasize reading.  In early longitudinal studies, the early gains of Title I students often declined once students were returned to full-time instruction in regular classes.  Therefore, programs should be continually enriched in order to avoid the "washout" of earlier positive effects.  The most recent studies have shown modest evidence of long-term positive achievement effects but nothing like a true closing of gaps with more advantaged students.

There has been rising concern that the Title I programs have been framed too narrowly, often emphasizing decontextualized skills training.  One study group called for a transformation of the program "to one dedicated to spurring the kinds of educational change that would result in children born into poverty acquiring high-level knowledge and skills"; the call was for "deep change in the way whole school systems operate" (Commission on Chapter 1, 1992, pp. v-vi).  Recent revisions in law are in accord with that schoolwide orientation and provide for more program flexibility at the local level.  Programs of the future will be concentrated increasingly in schools serving many students who experience poverty, but they will be less categorical?that is, requiring less special within-school identification of students by category.  Title I programs may now be coordinated fully with other categorical programs and with regular education.  When required to teach low-achieving and economically poor students, regular educators will less often be able to say, "That's not my job"; instead, we hope, they will accept the challenge and the extra resources now available to effectively teach children of the poor within the regular school framework.  It is intended, also, that management concerns will move from system compliance to program improvement.

A similar shift is occurring in programs for language-minority students.  In the case of bilingual education, the evidence on programs is mixed.  The situation is very complex, involving varieties of approaches to instruction and factors other than language. When instruction takes into account cultural differences as well as language differences, there appear to be better academic outcomes.  But conducting programs within a whole-school framework is gaining in preference over non-inclusive arrangements (Garcia, 1995).

This section summarizes actions that can be suggested on the basis of evidence to help accommodate school programs to the wide diversity of student characteristics.  The suggestions are presented in two broad sets: those that involve organizational arrangements and those that concern instruction.  In no instance is the suggestion fully or absolutely confirmed by research; rather, the suggestions represent ideas and practices that are indicated in view of present knowledge.  Thus, continuing research and evaluation are important in all of the domains discussed.  Literature concerning each suggestion is quite readily available, providing both conceptual details and implementation approaches that have been tried and tested.

Organizational Arrangements

Inclusion.  There is not sufficient evidence to justify removal of mildly disabled students by category from regular education schools and classes.  Strong efforts should be made to be as inclusive as possible in regular school settings.  Even severely disabled students can usually be included in regular schools, although full inclusion in regular classes may often be doubtful as a practice.

Early Education.  Evidence suggests that high-quality early education programs that include children who are placed at risk by a variety of circumstances result in reductions in later referrals to special education.  Such programs tend to establish positive school-parent relations and to avoid the stigmatic labeling of children.

Intensive Early Preventive Programs.  Intensive instructional programs for students who show limited progress in early primary education (for example, in reading at the first-grade level) tend to prevent many later learning disabilities.  Making an early start in offering very intensive help to children who lag in the early phases of academic learning is exceedingly important.

Prereferral Interventions.  Teachers who become greatly concerned about pupil behavior and/or learning should be provided help in arranging prereferral interventions.  For example, a school psychologist, a special education teacher, or some other staff member should enter into a consulting relationship with the teacher.  One result is a reduced need for referral to special education, which can result in substantial savings.

Parents. A positive approach is to seek strong working relationships with parents, emphasizing mutual trust and commitments to healthy supports in the life of each child.  Through time, increased student involvement should be sought in school planning and in preparing for the transition to post-school life.

Cross-Category Programs.  Another arrangement involves the creation of broadly framed mixes of categorical programs rather than narrowly framed programs.  For example, many schools now combine programs for so-called learning disabled, mildly retarded, and emotionally disturbed children, with Title I programs.  Labels for students are dropped.  Such programs can be conducted within regular classrooms through co-teaching arrangements between special and regular teachers.

Mini-Schools.  Evidence favors formation of small schools or schools within schools, in which clusters of teachers and pupils remain together for more than one school year.  This fosters deeper and improved relationships between teachers and students and enhanced accountability (to one another) among teachers.  Such mini-schools should serve heterogeneous groups of students.  The procedure is consistent with important findings on childhood resilience.

Teacher Assistance Teams.  Small teams of regular teachers can be used to assist particular teachers at times of stress and uncertainty about the behavior of pupils or about instructional planning.  This approach builds on the idea of coordinated consultation, that is, using people whose situation is like yours in arranging help.  Not all help must come from specialists.  Ongoing teams of teachers can also offer leadership and support in working through change processes in their schools.

Intensive Special Education.  If resource rooms or other forms of special education that involve separate placements are used, they should be restricted to time-limited periods and offer very intensive help on particular topics or skills.  Speech-correction programs frequently operate in this way.

Funding. Funds to support categorical programs for mildly disabled students and those whose circumstances place them at risk should be distributed to individual schools that then pay for the special programs they use.  If the special program involves sending a student to a center beyond the local school, the costs could be quite extraordinary.  The idea is to create incentives for serving all students, even those with exceptional needs, in the schools they would ordinarily attend.  It might be necessary to arrange exceptions to this site-based funding system in the case of programs that meet the ATI test (i.e., for students who are deaf, blind, speech impaired, or severely/profoundly disabled).  In these instances, some districtwide arrangement might be made for operations and funding.  The distribution of funds to schools should be based on the number of students at each school who have special needs, with bonuses paid on evidence that the school program is successful in preventing and solving the problems of special-needs students.

Coalitions of Agencies. Schools should seek coalitions with social, health, and corrections agencies; with churches and other elements of the community; and with families as a means of meeting the special needs of students and families in cohesive and efficient ways.  The orientation should not be just "clinical," that is, to serve individual needs; the coordinating body should help serve the general needs of the community as it affects the life and learning of children and youth.

Waiver for Performance.  Schools should be enabled and encouraged to seek waivers from existing rules and regulations when they are barriers to trials and testing of new modes of operation.  The waivers should include provisions for melded use of categorical funds when, in fact, programs represent a melding of previously separated programs.  Waivered programs should always require strong evaluative components, that is, evidence of outcomes or "performance."

Demissions.  Today's schools require strong efforts to reduce all forms of demission: dropouts, suspensions, expulsions, exclusion and excuses.  Work on school climate, broadly framed judiciaries to deal with conflicts and unacceptable behavior, clear statements of expectations for student behavior, rewards for progress in "civilized behavior," strategic involvement of parents and community agencies, and more, are involved.  What is clear is that solid efforts do pay off in this domain.

Instruction

Several factors concerning instruction must be considered as attempts are made to bring categorical programs into more cohesive relations with regular education.  First, it is clear that the same principles apply to instruction in all programs (Reynolds, Wang, & Walberg, 1992). Thus, it is not necessary-at basic levels-to  separate teachers according to category for most of their initial preparation or continuing education.  There has been too much separation and narrowness in higher education in ways that correlate with separations in the schools.  This separation needs to be repaired.  Second, it is clear that the state of practice in schools falls far short of the state of the art in instruction.  There is a substantial knowledge base about effective instruction that should serve as an organizer for efforts to improve schools and to bring programs into a cohesive form.

A massive meta-review of the research literature on effective instruction was completed recently by Wang, Haertel, and Walberg (1993).  Discussed below are a very small sample of topics that emerged in that study.  For each topic, stated here in declarative terms, it is essential to make transformations to practical procedures (presumably of many kinds) that can be used effectively in teacher preparation and in student learning.  The examples should not be seen as bits of technical knowledge to be inserted in simple fashion in teaching-learning situations.  Rather, they are domains of knowledge to be considered by teachers, parents, and others as they perform their functions within the particular contexts where they meet with children and youth.  The topics considered below are important and reasonably well-confirmed elements of knowledge, worthy of awareness and reflection by all who share in the teaching of children.

Time.  Perhaps the clearest and most ubiquitous factor shown to be important in learning is time.  Active, conscious, and deliberate commitment to learning is, of course, of key importance.  But even if learning activities are well designed, time has an added independent importance.  Indeed, greater knowledge of how children use their time would likely yield solid predictions of what they are learning.  But here there is great failure for many children, especially for those who watch low-grade TV for four to six hours per day or who languish on unsafe streets.  There is much talk these days about lengthening the school day and year, each of them potentially positive in effects.  But carefully using the time already allocated to schools and to home life is essential.  We need efforts in every school and, in cooperation with parents, in every home to monitor the use of time by children and to improve its use.  The emerging coalitions of community agencies can come together to influence the available opportunities and the rates at which children use time constructively.  There is very little that is reassuring about how students use time now or about how communities make provisions for the controlled use of time.  This is prime territory for improvements.

Metacognition.  No topic has emerged more forcefully in recent years than that of helping students become more aware of their own learning and become self-managers in their education.  It may be easier for teachers to lay out all plans and to manage classes in detail, leaving compliance as the duty of the students.  But the evidence is clear that students who become reflective about their own learning and who gradually assume more of the management of it show the most advances.  Such management involves thinking, planning, judgment, and accountability.  Teachers who reveal openly their own thinking processes, who help students learn to keep track of their own learning, and who share with students the planning of learning activities are those who help most.  Having students help one another in learning provides another step in becoming aware of essential elements in learning and self-management.  All of these observations apply to students who are mildly disabled or whose circumstances place them at risk of educational failure.

Cooperative Grouping/Student Interdependence.  Evidence relating to students helping one another in classroom work has advanced rapidly in recent years.  The estimated effect size of cooperative grouping on learning outcomes is too high to be ignored by any teacher.  These arrangements engage students in awareness about individual differences and cultural diversity, but in ways that encourage appreciation rather than depreciation.  It is but a small step from cooperative grouping to procedures for reaching methods of nonviolent conflict resolution.  When students are expected to be interdependent, that is, to be useful and helpful to one another, schools become far different places than they were in decades past.

Frequent Feedback.  Students need frequent information on how their knowledge and skills are advancing in the various domains of the curriculum.  The realm of feedback is where the expertise of the teacher, coach, counselor, and parent can be helpful.  But students also can learn to chart their own progress.  Common sense and research agree that information about learning progress can be reinforcing.

Social Skills.  One of the most common reasons for suspending students from school or for referring them to special education is that their social behavior is unacceptable.  This unacceptable behavior can occur because students do not know what is acceptable.  Training in social skills is an area of rapid development to meet this need.  Although much evaluative work remains to be done, progress thus far is impressive.  School psychologists have taken a lead in this area.

Much more could be added about the possibilities for instructional improvement.  A more complete delineation of evidence on effective instruction is available quite readily (viz., Wang, Haertel, & Walberg, 1993).
 

To meet the challenge of the growing diversity in student characteristics, schools have created a variety of categorical programs.  Most school systems now have a dozen or more such programs.  Typically they have come about through socio-political processes with but little reference to research findings.  The segregation of many categorical programs probably has caused a neglect, or at least a delay, of efforts to expand the powers of general education programs to deal with individual differences among students.  Due to the disjointedness that categorical programs create within schools and the extremely high dollar cost of separate categorical programs, the situation is changing.  As efforts now go forward to create broadly systemic and inclusive schools that serve all students, it is suggested that the designs begin with consideration of the evidence now at hand concerning organizational arrangements and effective instruction.  In the main, students at the margins, those now so often set aside in categorical programs, need not receive a different kind of instruction, but only more intensive and effective instruction in inclusive settings.
 

Baker, E. T., Wang, M. C., & Walberg, H. J. (1995).  The effects of inclusion on learning.  Educational Leadership, 52(4), 33-35.  (See this entire issue of Educational Leadership for a variety of reports of evaluation of special education programs.)

Commission on Chapter 1 (1992).  Making schools work for children in poverty: A new framework prepared by the Commission on Chapter 1. Washington, DC: American Association for Higher Education.

Garcia, E. E. (1995). The impact of linguistic and cultural diversity on America's Schools: A need for new policy. In M. C. Wang & M. C. Reynolds (Eds.), Making a difference for students at risk: Trends and alternatives (pp. 156-180).  Thousand Oaks, CA: Corwin Press. 

Glaser, R., & Silver, E. (1994).  Assessment, testing and instruction: Retrospect and prospect.  Review of Research in Education, 20, 393-419.

Heller, K. A., Holtzman, W. H., & Messick, S. (Eds.). (1982). Placing children in special education: A strategy for equity.  Washington, DC: National Academy Press.

Linn, R. L. (1994).  Performance assessment: Policy promises and technical measurement standards. Educational Researcher, 23(9), 4-14.

Moore, M. T., Strang, E. W., Schwartz, M., & Braddock, M. (1988). Patterns in special education delivery and cost.  Washington, DC: Decision Resources Corporation.

Reynolds, M. C., Wang, M. C., & Walberg, H. J. (1992).  The knowledge bases for special and general education.  Remedial and Special Education, 13(5), 6-10.

U.S. Department of Education (1994). Sixteenth annual report to Congress on the implementation of the Individuals with Disabilities Education Act.  Washington, DC: Author.

Wang, M. C., Haertel, G. D., & Walberg, H. J. (1993).  Toward a knowledge base for school learning.  Review of Educational Research, 63(3), 249-294.