Born of egalitarian instincts, the grand experiment of U.S. public education began over 200 years ago. The scope and complexity of its agenda is apparent:
to teach the fundamental skills of reading, writing, and arithmetic; to nurture critical thinking; to convey a general fund of knowledge; to develop creativity and aesthetic perception; to assist students in choosing and preparing for vocations in a highly complex economy; to inculcate ethical character and good citizenship; to develop physical and emotional well-being; and to nurture the ability, the intelligence, and the will to continue on with education as far as any particular individual wants to go (Cremin, 1990, p. 42).
The educational system is no less complex. Today the United States sends more than 45 million children to schools that are governed by 15,000 independent school districts in the 50 states (and territories); it boasts thousands of colleges and universities and myriad adult and informal learning centers. The nation takes pride in reaffirming the constitutional limitations on the federal role in education, yet recently has tentatively embraced the idea of national standards. The system is one of dualities: a national ethos with local control; commitment to excellence and aspiration to equality; and faith in tradition and appetite for innovation.
The context in which this system operates is also changing. The United States is no longer a manufacturing society in which people with little
formal education can find moderate- to high-paying jobs. It is now a service- and knowledge-driven economy in which high levels of literacy and numeracy are required of almost everyone to achieve a good standard of living (National Research Council, 1999a; Secretary’s Commission on Achieving Necessary Skills, 1991; Murnane and Levy, 1996; Judy and D’Amico, 1997; Packer, 1997). Moreover, to address the challenges of, for example, low-performing schools, the “achievement gap,” and language diversity, educators today require new knowledge to reengineer schools in effective ways.
To meet these new demands, rigorous, sustained, scientific research in education is needed. In today’s rapidly changing economic and technological environment, schooling cannot be improved by relying on folk wisdom about how students learn and how schools should be organized. No one would think of designing a rocket to the moon or wiping out a widespread disease by relying on untested hunches; likewise, one cannot expect to improve education without research.
Knowledge is needed on many topics, including: how to motivate children to succeed; how effective schools and classrooms are organized to foster learning; the roots of teenage alienation and violence; how human and economic resources can be used to support effective instruction; effective strategies for preparing teachers and school administrators; the interaction among what children learn in the context of their families, schools, colleges, and the media; the relationship between educational policy and the economic development of society; and the ways that the effects of schooling are moderated by culture and language. In order that society can learn how to improve its efforts to mount effective programs, rigorous evaluations of innovations must also be conducted. The education research community has produced important insights on many of these topics (we trace some of them in Chapter 2). However, in contrast to physics and other older sciences, many areas of education are relatively new domains for scientific study, and there is much work yet to do.
Everyone has opinions about schooling, because they were all once in school. But in this ever more complex world, in which educational problems tend to be portrayed with the urgency of national survival, there is (again) an understandable attraction to the rationality and disciplined style of science. Simply put, for some problems citizens, educators, administrators,
policy makers, and other concerned individuals want to hear about hard evidence, they want impartiality, and they want decisions to rest on reasonable, rigorous, and scientific deliberation. And how can the quality of science be judged? This is our topic.
To set the stage for this discussion, this chapter provides historical and philosophical background and describes how the current undertaking fits into that broader context.
HISTORICAL AND PHILOSOPHICAL CONTEXT
Education research in the United States is barely 100 years old, and its history is not a simple tale of progress. The study of education drew heavily on the emerging social sciences, which had found a place in research universities at the beginning of the twentieth century. That foothold was often tenuous, however, with intense debates about the essential character of these “sciences.” Many in academic circles sought to model the social sciences on the physical sciences, while others—regarding this as “physics envy”—insisted that broader accounts of the nature of science had to be adopted in order to encompass adequately the range of phenomena in these newer domains (Lagemann, 2000).
Education research began as a branch of psychology at a time when psychology was still a part of philosophy. In the first decade of the twentieth century, psychology was emerging as a distinct field, as were the budding fields of educational psychology, history of education, and educational administration. By the 1930s, subfields of work that centered on different subjects of the school curriculum—notably reading, mathematics, and social studies—had also emerged. As education research continued to develop new methods and questions and in response to developments in the social and behavioral sciences, research fields proliferated (Lagemann, 2000; Cronbach and Suppes, 1969).
From the beginning, the field has been plagued by skepticism concerning the value and validity of developing a “science of education.” This attitude was evident as long ago as the late nineteenth century, when universities began to establish departments and schools of education. A chorus of complaints arose from faculty in the arts and sciences concerning the inclusion of scholars intending to systematically study the organizational
and pedagogical aspects of schooling. Ellwood Patterson Cubberley, a school superintendent in San Diego who just before the end of the nineteenth century was appointed chair of the department of education (later the School of Education) at Stanford University, arrived on campus ready and eager to help improve education by generating studies of the history and current administration of the nation’s public schools. Despite his enthusiasm and extraordinary productivity, his colleagues refused to acknowledge that “the study of education could be validly considered either an art or a science.” On the opposite side of the country Paul Hanus, Harvard’s first scholar of education, faced similar skepticism. George Herbert Palmer liked to quip that when “Professor Hanus came to Cambridge, he bore the onus of his subject.” (quoted in Lagemann, 2000, p. 72). Indeed, a set of attitudes toward education research that one might call “anti-educationism” has been a constant to the present day.
Despite this skepticism, the enterprise grew apace. For example, by the end of the twentieth century, the American Educational Research Association (AERA) had well over 20,000 members (roughly 5,500 of whom report research as their primary professional responsibility), organized into 12 divisions (e.g., administration, curriculum, learning and instruction, teacher education), some with a number of subsections, and about 140 special interest groups (American Educational Research Association, 2000). This growth in the number of scholars has been notable because it occurred in the absence of a proportional increase in federal funding. And as a percentage of the total amount spent on public elementary and secondary education, the nation as a whole invested less than 0.1 percent in research (President’s Committee of Advisors on Science and Technology, 1997).
There are several reasons for the lack of public support for education research. Problems include research quality (Lagemann, 2000; Kaestle, 1993; Sroufe, 1997; Levin and O’Donnell, 1999), fragmentation of the effort (National Research Council, 1992), and oversimplified expectations about the role of research in education reform (National Research Council, 2001d). Another key problem has been the sharp divide between education research and scholarship and the practice of education in schools and other settings. This disconnect has several historic roots: researchers and practitioners have typically worked in different settings; most researchers
have been men, while most teachers have been women; and teacher education has typically relied on practical experience rather than research. Operating in different worlds, researchers and practitioners did not develop the kinds of cross fertilization that are necessary in fields where research and practice should develop reciprocally—medicine and agriculture faced similar problems in their early development (Lagemann, 2000; Mitchell and Haro, 1999).
The epistemology of education research—that is, understanding about its core nature as a scientific endeavor—has also evolved significantly since its early days (see Dewey  for an insightful early treatment). Five dimensions are particularly relevant to this report: the emergence of refined models of human nature; progress in understanding how scientific knowledge accumulates; recognition that education is a contested field of study; new developments in research designs and methods; and increased understanding of the nature of scientific rigor or quality. We comment briefly on each below and expand on several of them in the remaining chapters.
Models of Human Nature
In the decades when scientific research in education was gathering momentum, the most prevalent “models of man” and of human social life were derived from the mechanistic, positivistic sciences and philosophy of the nineteenth and twentieth centuries. The most famous example—the focus of numerous theoretical and methodological battles—was B.F. Skinner’s behaviorism (Skinner, 1953/1965, 1972). Following the work of the logical positivist philosophers, who believed that talking about entities that were not available for direct inspection (such as thoughts, values, ideals, and beliefs) was literally meaningless, Skinner’s research assumed that human behavior could be explained completely in terms of observable causes— for example, through schedules of reinforcement and punishment. Although Skinner’s work laid the foundation for modern theories of behavior (see National Research Council, 2001b), the behaviorist paradigm excluded important phenomena from inquiry at the outset of the study. Today, it is recognized that many phenomena of interest across the domains of the social sciences and education research result from voluntary human actions (or from the unintended or aggregate consequences of such actions) even
though direct measurement of such phenomena is typically not possible.1 Thus, research on human action must take into account individuals’ understandings, intentions, and values as well as their observable behavior (Phillips and Burbules, 2000; Phillips, 2000.)
The development of alternative perspectives on the nature of humans that are more inclusive than the once-dominant behaviorist perspective should be regarded as both highly promising and something of a cautionary tale for education research. The moral of the rise and at least partial fall of behaviorism warns the scientific community to resist the tendency to take a single model (whether behavioral, cognitive, or interpretive), derived in relation to a limited range of phenomena, and extrapolate it as appropriate across all the social and behavioral sciences. There is room in the mansion of science for more than one model, and also for the creative tension produced when rival models are deployed (see, for an example, Greeno et al., 1996).
Progress in Science
If appreciation for multiple perspectives on the nature of humans has enhanced efforts to develop scientific research, so has a better, more sophisticated awareness of what “progress” in science means and how it is achieved. Linear models of progress have been put aside in favor of more jagged ones. Mistakes are made as science moves forward. The process is not infallible (see Lakatos and Musgrave, 1970); science advances through professional criticism and self-correction. Indeed, we show in Chapter 2 that this jagged progression of scientific progress is typical across the range of physical and social sciences as well as education research.
A long history of the philosophy of science also teaches that there is no algorithm for scientific progress (and, consequently, we certainly do not attempt to offer one in this report). Despite its optimistic-sounding title, even Sir Karl Popper’s (1959) classic work, The Logic of Scientific Discovery, makes the point strongly that there is no logical process by which researchers
can make discoveries in the first place. Popper also argues that knowledge always remains conjectural and potentially revisable. Over time, erroneous theories and inaccurate findings are detected and eliminated, largely by the process of testing (seeking refutations) that Popper himself described (Popper, 1965; Newton-Smith, 1981).
Education—A Highly Contested Field
While knowledge in the physical and social sciences and education has accumulated over time, the highly contested nature of education has had an effect on the progress of scientific research (Lagemann, 1996). One reason education is highly contested is because values play a central role: people’s hopes and expectations for educating the nation’s young are integrally tied to their hopes and expectations about the direction of society and its development (Hirst and Peters, 1970; Dewey, 1916). Obviously, different people see these matters differently. As in other fields that have such a public character, social ideals inevitably influence the research that is done, the way it is framed and conducted, and the policies and practices that are based on research findings. And decisions about education are sometimes instituted with no scientific basis at all, but rather are derived directly from ideology or deeply held beliefs about social justice or the good of society in general.
A second reason that education is contested is that rarely, if ever, does an education intervention—one important focus of study in the broader domain of education research—have only one main effect. Both positive and negative unintended consequences are often important (Cronbach et al., 1980). Education interventions have costs—in money, time, and effort: making a judgment on the effectiveness of a treatment is complex and requires taking account of myriad factors.
In short, education research will inevitably reflect and have to face many different values, and it will as a consequence produce complex findings. Ultimately, policy makers and practicing educators will have to formulate specific policies and practices on the basis of values and practical wisdom as well as education research. Science-based education research will affect, but typically not solely determine, these policies and practices.
Research Design and Method
Research in education has been enhanced by the recent invention of methods: new observational techniques, new experimental designs, new methods of data gathering and analysis, and new software packages for managing and analyzing both quantitative and qualitative data. Rapid advances in computer technologies have also dramatically increased the capacity to store and analyze large data sets. As new methods are developed, they lead to the identification of new questions, and the investigation of these, in turn, can demand that new methods be devised. We illustrate this dynamic relationship between methods, theories, empirical findings, and problems in Chapter 2 and describe common designs and methods employed to address classes of research questions in Chapter 5.
Scientific Evidence and Rigor
In thinking about the ways that a research conjecture or hypothesis may be supported by evidence, many philosophers of science have found it fruitful to adopt a term that was featured in John Dewey’s (1938) treatise, Logic: The Theory of Inquiry (see, e.g., Phillips and Burbules, 2000). Dewey wrote of warrants for making assertions or knowledge claims. In science, measurements and experimental results, observational or interview data, and mathematical and logical analysis all can be part of the warrant—or case—that supports a theory, hypothesis, or judgment. However, warrants are always revocable depending on the findings of subsequent inquiry. Beliefs that are strongly warranted or supported at one time (e.g., the geocentric model of the solar system) may later need to be abandoned (for a heliocentric model). Evidence that is regarded as authoritative at one time (e.g., ice ages are caused by the eccentricity of the Earth’s orbit) can be shown later to be faulty (see Chapter 3). Science progresses both by advancing new theories or hypotheses and by eliminating theories, hypotheses, or previously accepted facts that have been refuted by newly acquired evidence judged to be definitive.
To make progress possible, then, theories, hypotheses, or conjectures must be stated in clear, unambiguous, and empirically testable terms. Evidence must be linked to them through a clear chain of reasoning. Moreover, the community of inquirers must be, in Karl Popper’s expres-
sion, “open societies” that encourage the free flow of critical comment. Researchers have an obligation to avoid seeking only such evidence that apparently supports their favored hypotheses; they also must seek evidence that is incompatible with these hypotheses even if such evidence, when found, would refute their ideas. Thus, it is the scientific community that enables scientific progress, not, as Nobel Prize-winning physicist Polykarp Kusch once declared, adherence to any one scientific method (Mills, 2000 [emphasis added]). We emphasize this notion of community in the scientific enterprise throughout this report.
These points about the nature of evidence constitute the essence of our account of rigor in inquiry; these ideas are fleshed out in the rest of this report. Importantly, our vision of scientific quality and rigor applies to the two forms of education research that have traditionally been labeled “quantitative” and “qualitative,” as well as to two forms of research that have been labeled “basic” and “applied.” These dichotomies have historically formed fault lines within and outside academia. As our brief discussion of the emergence of schools of education suggests, the perceived hierarchy of basic or “pure” science versus its messier cousin—applied research—has isolated the field of education research from other sciences. Similarly, sharp distinctions between quantitative and qualitative inquiry have divided the field. In particular, the current trend of schools of education to favor qualitative methods, often at the expense of quantitative methods, has invited criticism. Real problems stem from these “either/or” kinds of preferences, and we believe that both categorizations are neither well defined nor constructive. Thus, beyond a brief discussion that follows, we do not dwell on them in the report.
It is common to see quantitative and qualitative methods described as being fundamentally different modes of inquiry—even as being different paradigms embodying quite different epistemologies (Howe, 1988; Phillips, 1987). We regard this view as mistaken. Because we see quantitative and qualitative scientific inquiry as being epistemologically quite similar (King, Keohane, and Verba, 1994; Howe and Eisenhart, 1990), and as we recognize that both can be pursued rigorously, we do not distinguish between them as being different forms of inquiry. We believe the distinction is outmoded, and it does not map neatly in a one-to-one fashion onto any group or groupings of disciplines.
We also believe the distinction between basic and applied science has outlived its usefulness. This distinction often served to denigrate applied work (into which category education research was usually placed). But as Stokes (1997) in Pasteur’s Quadrant made clear, great scientific work has often been inspired by the desire to solve a pressing practical problem— much of the cutting-edge work of the scientist who inspired the book’s title had this origin. What makes research scientific is not the motive for carrying it out, but the manner in which it is carried out.
Finally, it is important to note that the question of what constitutes scientific rigor and quality has been the topic of much debate within the education research community itself since the nineteenth century. Two extreme views in the field’s complex history are worthy of brief elaboration. First, some extreme “postmodernists” have questioned whether there is any value in scientific evidence in education whatsoever (see the discussion of these issues in Gross, Levitt, and Lewis, 1997). At the other end of the spectrum, there are those who would define scientific research in education quite narrowly, suggesting that it is only quantitative measures and tight controls that unambiguously define science (see, e.g., Finn, 2001). We do not believe that either view is constructive, and in our estimation they have both compounded the “awful reputation” (Kaestle, 1993) of education research and diminished its promise.
PUBLIC AND PROFESSIONAL INTEREST IN EDUCATION RESEARCH
While federal funding for education research has waxed and (mostly) waned, the federal government has been clear and consistent in its call for scientific research into education. The Cooperative Research Act of 1954 first authorized the then Office of Education to fund education research (National Research Council, 1992). The National Institute of Education (NIE) was created in 1971 to provide “leadership in the conduct and support of scientific inquiry into education” (General Education Provisions Act, Sec. 405; cited in National Research Council, 1992). Likewise, as NIE was incorporated into the U.S. Office of Educational Research and Improvement (OERI), the quest for the scientific conduct of education research was front and center (Department of Education Organization Act, 1979; see National Research Council, 1992).
The federal government has not been alone in calling for scientific research into education. This call has been echoed in a series of reports and recommendations from the National Academies’ research arm, the National Research Council (NRC). In 1958, the NRC’s report, A Proposed Organization for Research in Education, recommended establishing a research organization for the advancement and improvement of education. A 1977 report, Fundamental Research and the Process of Education, called for fundamental research about educational processes. A 1986 report, Creating a Center for Education Statistics: A Time for Action, led to what many regard as the successful overhaul of the federal education statistical agency. And in the 1992 report, Research and Education Reform: Roles for the Office of Educational Research and Improvement, the NRC called for a complete overhaul of the federal research agency, criticizing its focus on “quick solutions to poorly understood problems” (National Research Council, 1992, p. viii). The report recommended creating an infrastructure that would support and foster scientific research into learning and cognitive processes underlying education, curriculum, teaching, and education reform.
What, then, warrants another NRC report on scientific research in education? First, as we argue above, the nation’s commitment to improve the education of all children requires continuing efforts to improve its research capacity. Questions concerning how to do this are currently being debated as Congress considers ways to organize a federal education research agency. Indeed, H.R. 4875—the so-called “Castle bill” to reauthorize OERI—has provided us with an opportunity to revisit historic questions about the “science of education” in a modern policy context. This bill includes definitions—crafted in the political milieu—of scientific concepts to be applied to education research, reflecting yet again a skepticism about the quality of current scholarship. (We discuss these definitions briefly in Chapter 6.) Our report is specifically intended to provide an articulation of the core nature of scientific inquiry in education from the research community.
The rapid growth of the education research community in recent years has resulted in the production of many studies, articles, journal publications, books and opinion pieces associated with academics, but that are not necessarily scientific in character. Moreover, the field of education researchers is itself a diverse mix of professionals with varying levels and types of research training, and they often bring quite different orientations
to their work. These multiple perspectives are in many ways indicative of the health of the enterprise, but they also render the development of a cohesive community with self-regulating norms difficult (Lagemann, 2000). In this spirit, we intend this report to provide a balanced account of scientific quality and rigor that sparks self-reflection within the research community about its roles and responsibilities for promoting scientific quality and advancing scientific understanding.
Finally, perhaps more than ever before, citizens, business leaders, politicians, and educators want credible information on which to evaluate and guide today’s reform and tomorrow’s education for all students. Driven by the performance goals inherent in standards-based reforms, they seek a working consensus on the challenges confronting education, on what works in what contexts and what doesn’t, and on why what works does work. Simply put, they seek trustworthy, scientific evidence on which to base decisions about education.
COMMITTEE CHARGE AND APPROACH
The committee was assembled in the fall of 2000 and was asked to complete its report by the fall of 2001. The charge from the committee’s sponsor, the National Educational Policy and Priorities Board of the U.S. Department of Education, was as follows:
This study will review and synthesize recent literature on the science and practice of scientific education research and consider how to support high quality science in a federal education research agency.
To organize its deliberations, the committee translated this mandate into three framing questions:
What are the principles of scientific quality in education research?
To address this question, the committee considered how the purposes, norms, methods, and traditions of scientific inquiry translated in the study of education. The committee also considered what scientific quality meant, both in individual research projects and in programs of research, to better
understand how knowledge could be organized, synthesized, and generalized. Furthermore, we sought to understand how scientific education research is similar to, and different from, other scientific endeavors.
In approaching this question, we recognize that existing education research has suffered from uneven quality. This statement is not very startling, because the same could be said about virtually every area of scientific research. Although it is clear that the reputation of education research is quite poor (Kaestle, 1993; Sroufe, 1997; H.R. 4875), we do not believe it is productive to attempt to catalogue “bad research.” Instead, we have found it useful to focus on constructive questions: How much good research has been produced? Why isn’t there more good research? How could more good research be generated? We address these kinds of questions in the report.
How can a federal research agency promote and protect scientific quality in the education research it supports?
The committee did not conduct an evaluation of OERI. Rather, the committee approached the general question of the federal role from the perspective of scientific quality and rigor. We sought to identify the key design principles for a federal agency charged with fostering the scientific integrity of the research it funds and with promoting the accumulation of science-based knowledge over time. Among the issues the committee explored were how research quality is affected by internal infrastructure mechanisms, such as peer review, as well as external forces, such as political influence and fiscal support, and how the federal role can build the capacity of the field to do high-quality scientific work.
Here again, our approach is constructive and forward looking. We attempt to strike a balance between understanding the realities of the federal bureaucracy and the history of an education research agency within it while avoiding the detailed prescriptions of previous and current proposals to reform the existing federal role. We hope to make a unique contribution by focusing on “first principles” that form the core of scientific education research at the federal level and providing guidance about how these principles might be implemented in practice. Some of our suggestions are already in place; some are not. Some will be easy to implement; others will
be more difficult. Our intent is to provide a set of principles that can serve as a guidepost for improvement over time.
How can research-based knowledge in education accumulate?
The committee believes that rigor in individual scientific investigations and a strong federal infrastructure for supporting such work are required for research in education to generate and nurture a robust knowledge base. Thus, in addressing this question, we focused on mechanisms that support the accumulation of knowledge from science-based education research—the organization and synthesis of knowledge generated from multiple investigations. The committee considered the roles of the professional research community, the practitioner communities, and the federal government. Since we view the accumulation of scientific knowledge as the ultimate goal of research, this issue weaves throughout the report.
Taking our cue from much of the historical and philosophical context we describe in this chapter, we make five core assumptions in approaching our work.
First, although science is often perceived as embodying a concise, unified view of research, the history of scientific inquiry attests to the fact there is no one method or process that unambiguously defines science. The committee has therefore taken an inclusive view of “the science of education” or “the educational sciences” in its work. This broad view, however, should not be misinterpreted to suggest “anything goes.” Indeed, the primary purpose of this report is to provide guidance for what constitutes rigorous scientific research in education. Thus, we identify a set of principles that apply to physical and social science research and to science-based education research (Chapter 3). In conjunction with a set of features that characterize education (Chapter 4), these principles help define the domain of scientific research in education, roughly delineating what is in the domain and what is not. We argue that education research, like research in the social, biological, and physical realms, faces—as a final “court of appeal”— the test of conceptual and empirical adequacy over time. An educational
hypothesis or conjecture must be judged in the light of the best array of relevant qualitative or quantitative data that can be garnered. If a hypothesis is insulated from such testing, then it cannot be considered as falling within the ambit of science.
A second assumption is that many scientific studies in education and other fields will not pan out. Research is like oil exploration—there are, on average, many dry holes for every successful well. This is not because initial decisions on where to dig were necessarily misguided. Competent oil explorers, like competent scientists, presumably used the best information available to conduct their work. Dry holes are found because there is considerable uncertainty in exploration of any kind. Sometimes exploration companies gain sufficient knowledge from a series of dry holes in an area to close it down. And in many cases, failure to find wells can shed light on why apparently productive holes turned out to be dry; in other words, the process of failing to make a grand discovery can itself be very instructive. Other times they doggedly pursue an area because the science suggests there is still a reasonable chance of success. Scientific progress advances in much the same way, as we describe in Chapter 2.
Third, we assume that it is possible to describe the physical and social world scientifically so that, for example, multiple observers can agree on what they see. Consequently, we reject the postmodernist school of thought when it posits that social science research can never generate objective or trustworthy knowledge.2 However, we simultaneously reject research that relies solely on the narrow tenets of behaviorism/positivism (see above) (National Research Council, 2001b) because we believe its view of human nature is too simplistic.
Fourth, the committee’s focus on the scientific underpinnings of research in education does not reflect a simplistic notion that scientific quality alone will improve the use of such research in school improvement efforts. Scientific quality and rigor are necessary, but not sufficient, conditions for improving the overall value of education research. There are major issues related to, for example, how the research enterprise should be
organized at the federal and local levels, how it should and can be connected to policy and practice (National Research Council, 1999d), and the nature of scientific knowledge in education (Weiss, 1999; Murnane and Nelson, 1984). Throughout this report, we treat these complementary issues with varying degrees of depth depending on their proximity to our focus on the scientific nature of the field. Indeed, over the course of our deliberations, we have become aware of several complementary efforts focused on improving education research (e.g., NRC’s Strategic Education Research Partnership, RAND panels, Education Quality Institute, Interagency Education Research Initiative, and National Academy of Education-Social Science Research Council Committee on Education Research).
Finally, and critically, the committee believes that scientific research in education is a form of scholarship that can uniquely contribute to understanding and improving education, especially when integrated with other approaches to studying human endeavors. For example, historical, philosophical, and literary scholarship can and should inform important questions of purpose and direction in education. Education is influenced by human ideals, ideologies, and judgments of value, and these things need to be subjected to rigorous—scientific and otherwise—examination.
Structure of Report
The remainder of this report moves from the general to the specific. We begin by describing the commonalities shared across all scientific endeavors, including education research. We then take up some of the specifics of education research by characterizing the nature of education and of studying it scientifically; describing a sampling of trusted research designs used to address key questions; and providing guidance on how a federal education research agency could best support high quality science. A description of the specific contents of each chapter follows.
In Chapter 2 we address the global question of whether scientific inquiry in education has generated useful insights for policy and practice. We describe and analyze several lines of work, both inside and outside of education, to compare the accumulation of knowledge in education to that of other fields. In doing so, we provide “existence proofs” of the
accumulation of knowledge in education and show that its progression is similar in many ways to other fields.
In Chapter 3 we provide a set of guiding principles that undergird all scientific endeavors. We argue that at its core, scientific inquiry in education is the same as in all other scientific disciplines and fields and provide examples from a range of fields to illustrate this common set of principles.
In Chapter 4 we describe how the unique set of features that characterize education shape the guiding principles of science in education research. We argue that it is this interaction between the principles of science and the features of education that makes scientific research in education specialized. We also describe some aspects of education research as a profession to further illuminate its character.
In Chapter 5, integrating our principles of science (Chapter 3) and the features of education (Chapter 4), we then take up the topic of the design of scientific education research. Recognizing that design must go hand in hand with the problem investigated, we examine education research design (and provide several examples) across three common types of research questions: What is happening? Is there a systematic effect? and How or why is it happening?
Finally, in Chapter 6 we offer a set of design principles for a federal education research agency charged with supporting the kind of scientific research in education we describe in this report. We argue that developing a strong scientific culture is the key to a successful agency and that all education stakeholders have a role to play in it.