Printer Friendly

Narrowing the gap between policy and practice.

Narrowing the Gap Between Policy andPractice

An appealing and persistent fantasy of many educational researchers is that policy change is based on empirical findings. The fantasy emerges from the analytic, rationalistic tradition, coupled with a genuine motivation by researchers to improve the state of the field. From this perspective policy should follow research, and change should be founded on evidence. Examination of major shifts in educational policy suggests quite a different scenario, however. Change results from social-political concerns, and most policy decisions precede rather than follow research. The sweeping changes in policy and practices contained in Public Law 94-142 emerged from civil rights issues and had a constitutional rather than an empirical foundation. Similarly, many of the changes proposed under the regular education initiative (REI) under derived from values and beliefs rather than from data.

That policy reflects social-political conditions, and that policy change precedes research, is in many ways a good thing. Given the slow pace of research, the penchant of researchers for precision of design and rigor of analysis, and the tendency of most researchers to be cautious in inference and generalization, it seems fair to say that we will not make major progress in educational reform if we wait for comprehensive data sets to drive our decisions. Indeed, ir regard to P.L. 94-142, based on research findings it is possible that we would still be amassing data documenting how many handicapped children are not provided educational services, or determining that some psychometric assessment techniques may be biased for certain subgroups of children.

There are appropriate roles for educational researchers, however, as exemplified in the programs of work discussed in this issue of Exceptional Children. These include efforts to test policy with evidence, to devise and evaluate methods of implementing policy, and to bridge the gap between policy and practice. This view of educational research does not demean the importance of the research contribution, but rather links research to social issues. There are, of course, a number of dangers in this model. Linking research to policy many diminish the importance of theoretical, more "basic" research> research continuity may be threatened as social issues change> policies may not address important concerns, or may ignore needed areas of work> and beliefs of policymakers may mitigate against acceptance of contrary evidence. At the same time, when focussed on educational reform, it is clear that one major responsibility of the research community is to study systematically and comprehensively the implementation of change. The other side of that coin is that policymakers must be willing to subject policy to test, and must provide support for analytic research directed at educational reform. In my view, the body of research reported here is consistent with both goals.

In their introductory article, Kaufman and his associates describe the structure of a 5-year program of research initiated by the Division of Innovation and Development of the Federal Office of Special Education and Rehabilitative Services (OSERS). The funded projects were directed to "...investigate instructional, organizational, and administrative issues relating to educating children with disabilities in the general classroom environment." Specific priorities included enhancing instructional program options, increasing teaching/learning efficiency, and developing school building models for providing services for students with disabilities in general education settings. The major goals and the specific priorities were consistent with policy defined in P.L. 94-142 and with the regular education initiative, yet the OSERS mandates were sufficiently broad to allow individual investigate groups considerable latitude in research objectives, design, content, and procedures.

The six studies reviewed here vary considerably from each other, yet all fit under the umbrella of the OSERS research mandates. As might be expected, the projects differ in specific goals, scope, and depth, although as a group they may be viewed as solid applied research. Several things stand out about the program of work as a whole. First, the research was conducted in schools with real teachers and real children. While that may sound unimportant, the fact that the projects were implemented in the field rather than inthe laboratory is relevant.

Researchers were faced with the problems and pressures that are part of school programs, including practical questions of schedules, time, and

curricular needs, as well as school professional' attitudes and beliefs, even some entrenched, vested interests. School are highly structured social systems, and change does not come easily. The importance of the findings and the generalizations which may be drawn rom them are immeasurably increased because the work was school based.

Second, some of the researchers adopted designs that went beyond the usual "experimental-control" format, instead testing multiple models or multiple variations of programs. Designs that test competing models yield more powerful and more useful data than designs that test single innovations, and in the long run, provide direction for improved services. The incorporation and test of various models, carried out in several of the projects, is to be commended. It is unfortunate that in some studies the tests of competing models, apparently included in the designs of the research, were not reported. I look forward to the results of the comparative analyses in future publications.

Third, and no unexpectedly, the projects tested and reported here were part of major programs of research which had been ongoing for some time. The content of the interventions and the implementation procedures were derived from reasonably clear theoretical orientations, and for the most part the procedures had already been refined. In one sense, then, providing support for these projects had a high probability of a solid payoff. On the other hand, support of already established programs of research almost inevitably means less supprt for new and perhaps innovative projects. In general, the programmatic changes implemented in these projects were conservative.

Taken as a whole, the projects may be viewed as timely, as addressing important "real world" questions, and as contributing to our understanding of mainstreaming. They were also expensive. It is important to ask, then, what have we learned, what has not been learned, and what do we need to know? Rather than review each project separately, I have attempted to respond to these questions by drawing on selective findings and/or generalizations from the various studies in the hope that they may inform both policy and practice.


The findings from the studies as a group underscore the well known generalizations that some intervention is better than no intervention. As example, each of the four interventions in the Schulte, et al. study led to gains in pre-postmeasures. Similarly, some success could be claimed for every program. The some intervention is better than none is so obvious as to be trivial, yet it is a point often overlooked when program implementors claim success. In the present studies an array of models and practices was tested: for example, consultant or direct services, assistance teams, curriculum-based measurement, and supplemental individualized instruction, each with specific variations. That all were successful, in part at least, is an important finding. Equally important, however, is that none could be considered wholly successful.

Program effects varied according to the outcome domains or measures selected. In some cases the interventions had a different degree of impact on academic domains than they had on social/behavioral domains. In general, social/behavioral outcomes were positive (e.g., fewer referrals for special services, more positive perceptions by teachers, etc.), whereas the findings for academic gains were mixed. The findings that pupils' behavior and adjustment were adequate in the regular class may alone be good reason to encourage regular class placement rather than placement in separate programs. The cautionary points is that mainstream placement in and of itself may not necessariy improve low achieving pupils' educational skills. As noted by Zigmond and Baker, "business as usual" is not enough. Indeed, the findings from several studies were discouraging in that there were few major gains in basic skills even with the considerable extra resources provided by the researchers. The point to be emphasized in that integrated placement models, like separate placement programs, are not automatic cure-alls for pupils' learning and social problems.

Differences in some effects were also associated with pupil characteristics. The OSERS research directive was to develop programs for "children with disabilities." This is a loosely defined target group, and the range of pupil characteristics and problems addressed in these studies was correspondingly broad. There were differences in chronological age as well as in subjects' attributes. Some pupils were already identified as learning disabled and in resource programs, others were nominated by regular class teachers as "difficult to teach," and still others were selected on the basis of review by prereferal teams. It should not surprise us, then, that the interventions were differentially effective. In my view this is a positive finding, underscoring the need for more detailed and precise study of integrated placements and of the pupils served in them. The point is made clear by Cooper and Speece, who found that certain skills or attributes were important if children were to benefit from particular classroom environments. Deno and his associates reported that low-achieving pupils (those with mild disabilities) did better in integrated than in resource programs> however, the special education pupils in their study did relatively poorly in both. The more precise and differentiated description of those who are program "successes" is an important next step in refining our intervention models.

Finally, and perhaps most important, it was clear in all of the models that the key person in program implementation is the teacher. Just as pupils differ, so do teachers, and it is likely that differences in teachers' skills, styles, attitudes, and beliefs in the interventions influenced the outcomes. Unfortunately, we are still uncertain as to which of these differences affect how teachers teach. Findings in these studies documented real differences among teachers in instructional practices. The study of effective and less effective teachers by Nowacek and her colleagues is instructive in this regard. Based on data gathered through on-site observations in classrooms, these investigators documented teachers' instructional behaviors (e.g., monitoring, questioning, and praise). Elementary teachers differed from intermediate and secondary teachers, and teachers of learning disabled students differed from general educators in both monitoring and praise. There were also some differences in the nature of affective interactions.

It is reasonable that differences in teachers' instructional techniques and styles must be considered when assessing program effects, especially when the range of pupil characteristics is taken into account. In this regard, it should be noted that even effective teachers, as traditionally defined, may need help in working with problem learners. Gersten, Walker, and Darch (1988), for example, found that teachers rated high in effectiveness in regular programs often lacked tolerance for pupils' deviant behaviors, and were troubled by inefficient use of academic time. The point that the teacher is a key variable in the integration of handicapped learners into regular education programs is not new. It does, however, have major implications for both pre- and in-service training programs and for the nature of support for teachers in mainstream classroom programs.

A point about research on innovative programs in schools deserves brief discussion. The innovative models described required a number of well trained persons to work with school faculty and staff. Nevertheless, problems of implementation were reported by several of the researchers. Some teachers felt burdened by extra paper work, time demands, and the like. Most school districts do not have, and cannot afford, equivalent support staff. A critical issue, then, relates to the feasibility of implementation and to the "maintainability" of programs over time, especially when the (presumed) positive presence of the researcher is no longer there. The real test of program success may be the ability of schools to continue to implement programs upon completion of the research.


Despite the insights gained from the work described in these studies there are still major gaps between policy and practice. These, too, deserve our consideration. First, we know considerably more about models of services for elementary school children than we do about those for intermediate or secondary school students. Intermediate and secondary level pupils were included in only one study among the six reported here. A relatively extensive literature suggests that the problems for mainstreaming high school students are different in kind and degree from those for younger children, and that elementary teachers are more willing than intermediate or secondary teachers to mainstream pupils (Ammer, 1984> Stephens & Braun, 1980). The emphasis on younger children in the present projects is understandable, given the complexity of the research task. No one investigator can do all things. Yet, the generalizations from these findings are limited and implications for practice and for school programs "across the boards" must be drawn cautiously.

Second, and closely related, the pupils included in these studies were almost exclusively mildly handicapped. Some students had not been identified for special education services, and where identified, were mostly in programs for learning disabilities. Many had already received considerable extra help from both resource and regular education personnel. From the studies reported here we have learned something about integrating pupils with relatively mild learning and behavior problems. We do not know as much about what are the best models and practices for integrating children with more severe problems. As mainstreaming children with a broad range of problems is an emerging policy, the study of models to accomplish this effectively is clearly an important goal for applied researchers.

Third, although major in scope, the studies were of relatively short duration, so we know little of the long-term effects or of the possible impact of innovative practices over time. Two aspects of the time issue may be important. One has to do with the effect of integrated placements on pupils over time and across grade levels. It is intuitively sensible that competence at one developmental period contributes to competence at subsequent levels, and that mastery of specific beginning skills provides the basis for success in learning more complex skills. Might we see a gradually accelerating improvement over time when learning handicapped children are in effective instructional and social settings early on? By the same token, we must consider the long-term impact of lack of success in mainstream placements. While well intentioned, it is possible that our instructional modifications may not work, and that over time pupils may have increasingly negative views of their own competencies> views enhanced by the social-comparison opportunities inherent in the regular classroom.

Another time-related concern has to do with the impact of innovative programs on teachers' attitudes and on their instructional practices. Just as success may breed success in children, so teachers, too, may change as their own competencies improve. We might expect that good instructional practices would become more integrated and more refined over time, and thus, would benefit a broader range of pupils in many areas of instruction. We might also expect that effective intervention and improved teaching would be associated with more positive teacher attitudes, an assumption that receives some support from the Fuchs, Fuchs, and Bahr study. Conversely, ineffective teaching techniques potentially lead to more severe problems for pupils, accompanied by increasingly negative attitudes of teachers. Reports from the studies described here, as well as a large literature, underscore the feelings of inadequacy regular class teachers feel about their skills in teaching exceptional learners. There is also some evidence of negative attitudes, although it is encouraging that change in attitude, at least in the short term, has been demonstrated (see Donaldson, 1980> Hudson, Reisberg, & Wolf, 1983). How these perceptions are affected by innovative practices is still uncertain. The question of change over time is testable and one which deserves study.


I began this discussion with the argument that educational change is based on beliefs and values, not on research evidence. Our belief in equity and our need to redress previous wrongs are strong reasons for initiating educational reforms. Now that the goals are established, the questions center around how to accomplish them, and it is here that educational researchers have a major role to play. The programs of work described in this issue of Exceptional Children are beginning efforts which provide interesting and useful information. The findings provide few definitive answers, but taken as a whole, they identify a number of important questions.

Readers of these reports will undoubtedly formulate somewhat different research needs, but as a start my list includes these questions and topics. What are the long term outcomes of different instructional programs? Is it the content or kind of intervention or is it the intensity of intervention which leads to successful outcomes? Are program innovations maintained and generalized when research support is no longer available? Are particular groups of pupils benefited (or not benefited) by particular programs? What are the situational or structural conditions (e.g., class size, numbers of mainstreamed pupils) which affect instruction? What are the social system, institutional conditions, and resources which are prerequisite to successful implementation of program practices? Do changes in teachers' attitudes toward exceptional children lead to improved instructional practices? What should be the content of pre-and in-service training programs and how can these programs be more effective? Answers to these questions require longitudinal strategies, consideration of multiple program options, and multiple study samples. They also require a comprehensive and consistent policy of research support.

In their book, Rousing Minds To Life, Tharp and Gallimore (1988) note that educational reformers tend to focus on "... matters remote from the practices of teaching and schooling, (or) on the daily experiences of teachers before or after they enter the profession. Ignoring such details and their effective implementation puts even the soundest of reforms at risk" (p. 3). This observation is relevant to the current reform efforts in special education. It is clear that major changes are needed in the delivery of services to problem learners, and that these services need to be the responsibility of regular as well as special educators. It is also clear that teachers are the central players in bringing about change in practice. It follows, then, that our greatest and most pressing challenge in the reform effort is to determine how to improve the quality of instruction at the classroom level. This is a formidable challenge that requires both creativity and hard work. It also forces us to examine the realities of linking policy and practices.


Ammer, J. (1984). The mechanics of mainstreaming: Considering the regular educators' perspective. Remedial and Special Education, 5, 15-20.

Donaldson, J. (1980). Changing attitudes toward handicapped persons: A review and analysis of research. Exceptional Children, 46(7), 504-514.

Gersten, R., Walker, H., & Darch, C. (1988). Relationships between teachers' effectiveness and their tolerance for handicapped students. Exceptional Children, 54, 433-438.

Hudson, F., Reisberg, L. E., & Wolf, R. (1983). Changing teachers' perceptions of mainstreaming. TEASE, 6(1), 18-24.

Stevens, T.J., & Braun, B.L. (1980). Measures of regular classroom teachers' attitudes toward handicapped children. Exceptional Children, 46, 292-294.

Tharp, R. G., & Gallimore, R. (1988). Rousing minds to life. Cambridge: Cambridge University Press.


BARBARA K. KEOGH is a Professor of Educational Psychology in the Graduate School of Education at the University of California, Los Angeles.
COPYRIGHT 1990 Council for Exceptional Children
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1990 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Special Issue: Enhancing the Education of Difficult-to-Teach Students in the Mainstream: Federally-Sponsored Research
Author:Keogh, Barbara K.
Publication:Exceptional Children
Date:Oct 1, 1990
Previous Article:Special education and the process of change: victim or master of educational reform?
Next Article:At-risk students in the fast lanes: let them through.

Related Articles
Tolerance and technology of instruction: implications for special education.
The consulting teacher model: risks and opportunities.
Special education and the process of change: victim or master of educational reform?
Teacher perceptions of the regular education initiative.
Mainstreaming during the early childhood years.
Special education in South Korea.
Inclusive education: policies, teachers' attitudes and perspectives.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters