the processes of educational research, development, demonstration, evaluation and dissemination which will help get the best in material and procedure more quickly into practice by making them readily available in useful form to those who control, manage, and teach in our educational institutions. Accordingly, a special working group was set up to explore the ways in which we could go about identifying specific objectives relating to this goal, and to make recommendations for action. Even before the group began its work, judgments about research policy had in effect been made. For example, stating the goal constituted implicit judgments about educational policy issues and the state of the educational system, as well as a judgment that research and development regarding this problem either was or could be made sufficiently coherent to develop a structure for directed programming. Similarly, by deciding to engage in a managed R & D effort, an explicit decision was in effect being made that there might be a new decision structure in this area at least (that is, the Office of Education would accept the responsibility and perform whatever linkage and coordination functions were riquired to accomplish the management task in a responsible and acceptable manner).

Soon after the planning group undertook this particular effort it became clear that they would need to articulate their understandings about research models for education, and how R & D functions relate to operating educational programs, institutions, and agencies. Of particular interest to operations researchers, for example, was the conclusion that so long as we were being asked to produce a research and development program designed to assist the educational programs of the nation, then it might be more useful to view our mission less from the academic research perspective and more from an operations research perspective. The planning group became convinced that, if education were conceived as a total complex interactive system, then mission-oriented R & D could properly be seen as operations research writ large; the R & D capability for the system should be though of as intimately connected and linked to the operating whole. The planning group in arriving at this conclusion was not for a moment denying that specialized institutions like universities and research corporations will be needed. But they judged the potential effectiveness of specialized institutions to be much greater if schools, districts, colleges, universities, and states were to have their own research capabilities and carried out their daily activities as if they, too, were engaging in inquiry (5).

The working group's views of educational R & D shaped the planning effort in other ways. For example, we in OE concluded we did not know enough to carry out the complete job. But even if we had had sufficient expertise, we became convinced we shouldn't complete the job ourselves. Without engaging in a great deal of communication with researchers, policy makers, and practitioners during the program development process, no matter what we produced would have been viewed with skepticism and distrust. If the program were to be viable and used by the educational practitioners, everyone would have to be involved from the start. Again, this clearly reflected convictions regarding the political and social dimensions of educational R & D as an activity.

The development of strategies and tactics for a directed R & D program on educational organizations and systems provides further examples. How acceptable would certain approaches be on face value to the constituencies on whom we depend for either political support or for performing the work? What sorts of modifications might be desirable to further develop and enhance such support? Or. keeping a firm eye on social need, real or perceived, how much if anything should we sacrifice in scientific elegance in order to achieve some measurable benefit now rather than greater elegance and three times as much measurable benefit perhaps 3 years from now? This entailed some consideration of the strength of social demand, for example, relative to the quite different require. ments of the academic community. Here we see an explicit interaction of two different elements, social and political context and manpower and its location, an interaction made especially difficult, perhaps because of the operation of different reward structures and motivations.

EXPERIMENTAL SCHOOLS

A second illustration can be developed in connection with the request for funds to establish experimental schools.

Here is an example of a proposal founded on an understanding that educational practice and achievement are based on or employ, to good effect or bad, a very large number of variables and that many of these must be utilized

simultaneously before substantial positive effects on student achievement will be observed. It is an R & D program emergent from some understandings based on previous research and development.

Operating again within the framework of the third primary element, it is useful to note exactly what experimental schools are and what their purposes are. Are they to experiment with new products and techniques to see what effects they have when combined with one another in the context of an entire school? Are they to demonstrate new and tested products and techniques to show others how they work so that the innovations may become more widely diffused? Are they to test new departures in the conceptions which underlie our establishment of schools or, to borrow T. S. Kuhn's notion, to experiment with the very paradigms on which instructional and educational practices are based (6)? Are they to be designed to mount the first sophisticated cost-benefit studies of educational practice? Are they to find out what works for target groups and areas where schools are presently failing?

Each of these purposes is different from the others. Some are compatible with one another; others are not. Some will use certain kinds of people; others will use different ones Some will be very costly; others will be less so. Some will require very long periods of time for planning and community consent; others might require only a few months to initiate.

Decision structures will be involved. Where will initiative for experimental school proposals reside? What role might program managers here in Washington play? What will be the role of the community in which the school is ultimately established? Wat kinds of criteria will be required and who will develop them, within which project proposals and program designs are established and evaluated?

Consider manpower and its location. With one or another interpretation of purposes and definition, different kinds of manpower presently found in a variety of different places and institutions will be required. How can people be identified, located, interested, and employed? Or are they already in the schools where the experiments are to be mounted? And will the experiment be in schools or by them?

What of the larger social and political context? Where are the schools currently failing in the United States? What are the target groups that, as measured in terms of achievement, are not receiving an equal opportunity for education in this country? What roles do these areas or groups of people have in the experiments in deciding what should be done and perhaps whether they should be done? How will black militancy, community involvement, and demands for selfdetermination be an essential frame of reference for this program?

NATIONAL INSTITUTE OF EDUCATION

Finally, on a third and much larger scale, let me list briefly some of the questions the framework suggests regarding the proposed National Institute of Education (NIE). The Institute proposal envisions nothing less than a total reconstruction of the administrative structure for R & D efforts in support of education. Not only is it designed to create a new atmosphere and climate for educational research, but ultimately it will absorb many of the R & D activities presently administered by OE. It is intended to play an important coordinative role with other related R & D efforts scattered across the federal structure in such agencies as OEO, NIMH, NICHD, NSF, and the Defense Department.

Certainly one of the central points, if not the most central one, deals with the degree to which NIE can or ought to be thought of, in Moynihan's terms, as "modeled shamelessly on the National Institutes of Health" (7). Recall again the descriptive differences between educational research (and in fact all behavioral and social science research) and other kinds of research, say, in physics, natural science, or biomedicine. Educational research is interwoven with issues of choice and value. Any agency responsible for administering such research must, first of all, build the political decision structures which will constitute the necessary enabling condition for success (8). I am not suggesting that the proposed Institute cannot meet this requirement, but it must if it is to succeed. The policy framework suggests that care must be exercised in establishing the Institute so that linear models of research and development (with their hierarchical flavor stemming from the implied primacy of the importance of research) are not adopted to the exclusion of others. Such models carry obvious logical power. They are the ones most commonly used to describe how science contributes to technological and economic advance.

But for reasons associated with the peculiar characteristics of behavioral and social science research, I suspect that other models-for example, those emphasizing practitioner initiative and involvement-may well be far more important for understanding the role of science in fostering educational improvement. We speak, for example, of "an idea whose time has come." In social fields the "whose time has come" part of the phrase is far more important than the "idea." The conditions that create a readiness in a social field to accept an idea from science are more important as far as adoption is concerned than the idea itself. Hence we find a peculiar dual phenomenon in all social fields. On the one hand, we observe the nonadoption of strong ideas in the absence of readiness. On the other, we see faddism, which is nothing more than readiness to adopt, in the absence of knowledge, a readiness which is soon disappointed by the low power of the innovation. If education is to be improved by science, the conditions causing practitioners and policy makers to pay attention to the ideas emerging from science must be established parallel to and as part of the support of science itself.

Extending the argument a little further, then, it may well be that the type of research which most needs stimulation, development, and support is the kind of inquiry that must be conducted in the operating educational institution. This is the research that determines who the learners are (in all their richness and detail), what the schools' operations actually are, and what effect those operations have relative to intended accomplishments. In a word, operations research may be more important. The question then become how the Institute can foster this kind of research.

SUMMARY

A policy framework for analyzing educational research and development has been proposed. The framework consists of five primary elements focusing on the contexts in which educational R & D operates, its goals, its characteristics, its manpower, and its decision structures. These five primary elements interact to produce two secondary elements concerned with priorities and objectives, and strategies and tactics. The framework was applied to three current policy proposals in educational R & D including directed R & D programming in the U.S. Office of Education, the request for funds for experimental schools, and the proposed National Institute of Education. Rigorous application of the framework to major policy questions such as these three can help us examine issues before they arise. It can help prevent errors. Its application can provide greater assurance that primary, secondary, and tertiary consequences-those that arise from the interaction of the consequences we most directly perceive-will receive examination. This kind of consideration can help produce the desirable and intended effects and avoid the detrimental consequences of unanticipated impacts in unintended areas.

REFERENCES AND NOTES

1. The status study prepared in connection with this review is now available [Educational Research and Development in the United States, OE-12049 (Government Printing Office, Washington, D.C., 1970)]. Chapter 10 of the report presents a detailed summary of all but the most recent of these reviews. For a brief summary of the OECD policy review, see H. Gideonse, Educ. Res. 21. 5 (April 1970). Later in the year OECD will publish in one volume the status study, the report of the four examiners who were responsible for conducting the review, and a summary of the confrontation session between the examiners and an American team held 19-20 November 1969.

2. When I drafted this article on April 1970, I referred to the possible application of this framework to other behavioral and social science research programs. Since then I have had two opportunities to confront the possibility that research programs directed to social services other than educational (such as welfare, corrections, and housing) might also profit by using the framework. It probably is applicable, but I am not professionally familiar enough with these other areas to adduce the evidence. In the event that study should document the usefulness of this framework, it might be possible with appropriate content modifications and additions to see it as a framework for policy formulation in any mission-oriented behavioral and social science research program. 3. "American Education: Notes Toward a New History." AERA-Phi Delta Kappa Address, American Educational Research Association Annual Meeting at Minneapolis, Minnesota, 4 March 1970.

4. I have discussed these matters in chapter 6, Educational Research and Development in the United States (1).

5. Compare R. J. Schaefer, The School as a Center of Inquiry (Harper & Row, New York, 1967).

6. A number of examples of this approach were presented by John Mays of the Office of Science and Technology at the Annual Meeting of AERA in Minneapolis, 5 March 1970.

7. White House press conference, 3 March 1970, mimeographed, p. 1.

8. Borrowing an idea from Theodore Levitt, another way of saying the same thing is that the Institute must link itself closely to its several markets and not fall into the error of thinking, because it has command of such good science, that of course it has something to offer (sell) to education.

9. I thank Maurice Kogan, of the team of OECD examiners, for his help in the development of the framework presented here. This paper was written in the author's private capacity with no official support or endorsement by the U.S. Office of Education.

RESEARCH AND DEVELOPMENT FOR EDUCATION: A MARKET MODEL

(By Hendrik D. Gideonse)

0.0-INTRODUCTION

The second successive national administration-a Republican one this timehas embraced the cause of educational research and development. The Johnson Administration revised the basic authorizing legislation in 1965 and called for the establishment of a network of educational laboratories "large and significant, comparable in their way to the large-scale laboratories of the Defense or Atomic Energy establishments, . . . equal in size and scope to the major tasks they seek to accomplish." The form of the Nixon Administration's embrace has been the submission of legislation to create a National Institute of Education "modeled shamelessly on the National Institute of Health." The backing and the promise seems high. Why then are we in such a muddle?

2

Much speculation and discussion-and even some careful study and analysis— have been devoted to the role of research and development in improving educational practice. Major policy explorations have been completed, most recently the policy review conducted under the sponsorship of the Organization for Economic Cooperation and Development (OECD).3 Commissions, study groups, White House-appointed panels, inter-agency review committees at the Federal level, and other groups have addressed smaller or larger segments of the problems. The result of all this examination as far as policy is concerned seems to be a peculiar immobilizing self-consciousness.

Politically the field of educational research has been in a state of disarray. Under fire from the Congress, the Bureau of the Budget (now the Office of Management and Budget), the Office of Science and Technology, staff offices of the Department of Health, Education, and Welfare, other Associate Commissioners in USOE, the Chief State School Officers, the White House, and a swarm of self-appointed critics in and out of government, USOE's research programs have struggled, stuttered, and somehow survived, but each time at a slightly heightened level of exhaustion.

Now, barely four years after the last set was launched, new initiatives have been proposed. USOE has been pressed, for example, to move the management of the program away from the development of institutions to carry out R&D and toward the identification of specific research and development objectives which will govern the management of the program. The new proposal for a National Institute of Education which has been forwarded by the President to the Con

1 Letter from President Lyndon B. Johnson to Secretary John W. Gardner, July 5, 1966. 2 Daniel P. Moynihan, White House press conference, March 3, 1970.

8 For a brief summary of the OECD review see my "OECD Policy Review of U.S. Educational R&D." Educational Researcher, April, 1970. The status study I prepared in connec tion with the review has been published by the Government Printing Office and is available under the title Educational Research and Development in the United States (OE 12049) from the Superintendent of Documents, GPO.

For a report of this large number of studies, up-to-date through December, 1969, see Chapter X of Educational Research and Development in the United States. At this writing (March, 1971) two more reports should be added. One was prepared by the President's Commission on Instructional Technology and issued in August, 1969. The other was prepared by Roger Levien of the RAND Corporation under contract to United States Office of Education as a planning study for the proposed National Institute of Education.

gress envisages a complete, wholly unspecified, but much needed upgrading and reorganization of the management and administration of educational research and development at the Federal level. The companion proposal for a National Foundation for Higher Education, while it appears to be a clear overlap of the proposed new Institute's functions, deserves mention, too. Finally, a new program of experimental schools originally proposed by former Secretary Finch has just been launched. Its appropriation for the first year exceeds the sum used to launch the first ten educational laboratories in 1966.

All of this has the Congress, the Chief State School Officers, the various and sundry parts of the Federal establishment, and various constituencies attendant to educational research programs hyper-excited, skeptical, horrified, hopeful, confused, and groggy at one and the same time. It should hardly be surprising, therefore, that accepting the assignment to prepare this paper has taken on for me something of the character of "going to the mountain."

What better time could be found to take the most common, widely distributed, and basically linear and hierarchical notions about how research and development can improve educational practice and explicitly turn them on their head? I proposed in this paper that what is most likely to advance the field of educational R&D is not further worry and concern about the current state of the art in educational R&D or the nature and interrelationships of R&D functions and processes." What is needed instead is for us to develop a highly sophisticated, imaginative, and unyielding concern for the market being served or created by educational R&D, that is, for the consumers, the clients, and the users of the outcomes of such research.

1.0-A FEW WORDS ABOUT MODELING

Why develop models? What is their purpose? How can they be useful? A simple answer is that models help us understand; they give us a sense of order to the "blooming, buzzing confusion." They help sort out functions, ideas, or activities. They help clarify relationships among elements.

But this simple anwer is clearly not sufficient itself. It only raises the question why we want that understanding. And the answer, I think, relates to some kind of instrumental need, a desire to do or accomplish something. Thus, we model to understand in order that we can make better decisions, manage better, change, self-fulfill better prophecies, and so on. A critically important point, this means that models are constructed with an eye to some kind of purposive action. Because this is so it seemed to me important to try and sketch out my goal for the educational system as far as educational research was concerned. As I developed this paper, therefore, I tried at the same time to create a vision of the educational system as it might look if it were functioning with strong scientific support. This exercise, begun originally as part of my responsibilities as Director of Planning for USOE's National Center for Educational Research and Development, enabled me to move backwards and forwards in a continuing means/ends analysis. The conception I emerged with at the end of the appointed time for preparing this model can be found at the end of the paper. It is by no means complete, but I include it because it does provide the reader with some idea of the purpose I came to have in mind as I developed this particular model. A second point respecting models is the usefulness of pointing out the distinction between conceptual, logical, or ideal models and descriptive or empirical models. Models can describe present conditions or they can describe ideal states as possible future conditions. Both kinds of modeling are useful, but it is important to be clear which is being attempted. In the case of this paper, the model is future-oriented to a state of affairs conceived to be desirable.

A useful caveat is that being purposive, modeling itself is a contextually related activity. Consider, for example, a complex system like a "muscle car." The position such a mechanical marvel holds in the conceptual systems of a highway patrolman (e.g. a potential law violator or perpetrator of serious accidents), a twenty-two year-old male (e.g. an aid to amatory success), or an ecologist (e.g.

The basis for this statement is to be found in conclusions drawn by the four OECD examiners during the course of their policy review of American educational R&D. The full report of the review is unfortunately not yet published though more than a year has elapsed since the completion of the policy review.

This tremendous concern for R&D functions and processes seems not to be repeated in any other field and is, I suspect, an outcome of two related phenomena: the supreme and long-prevalent methodological consciousness of educational research and the basic insecurity about our pretensions to scientific status.