Research Use? Knowledge Use? School Improvement? Personal Reflections on the Last 40 Years….

Research Use? Knowledge Use? School Improvement? Personal Reflections on the Last 40 Years….

I Have “Positionality” in This Arena….

My career as a researcher began in a time that, like today, was consumed with an interest in school improvement and the way in which better knowledge could support it. In 1969 I was a graduate student in Sociology and Columbia University, and obtained my first research assistant position with Carol Weiss, helping her prepare an annotated bibliography that began her line of work in evaluation and research use among policy makers. Shortly thereafter, in 1970, I began work on a project that was one of the first to examine how research-based knowledge could be used in schools. The “Pilot State Dissemination Project” was funded by the federal government (then the Office of Education) to develop the capacity of state departments of education to promote the potential of the burgeoning (but difficult to access) data base in ERIC (Educational Research Information Centers). As is the case today, both scholars and educational professionals bemoaned the “gap” between research and practice, but back then there was almost no empirical understanding of why it existed.

The results of that mixed-method study (Sieber, Louis, & Metzger, 1972) suggested that the “problem” of research and knowledge use in schools was far more complex than the predominant research-development-dissemination-utilization (RDDU) models suggested. In particular, a closer examination of the data pointed to the need for intermediary structures to get the “right stuff” into the hands of educators who had a question that could be answered with research. The role of “linking agents” or brokers/intermediaries who could help educators in schools figure out how to use evidence-based solutions that were not pre-packaged was critical – second only to whether the information was deemed to be useful (Louis, 1977). This finding was confirmed a few years later in several other studies that looked at knowledge use in different programs (Louis, 1983). Between studies that focused on research use and a parallel line of work on program implementation (Berman & McLaughlin, 1978), the importance of the human factor in knowledge dissemination and use was incontrovertible.

At the time, the National Institute of Education (NIE) was deeply committed to understanding and supporting research use and invested its resources in understanding how it was connected to school improvement. It underwrote one of the first AERA SIGs (Special Interest Group on Research Utilization), and funded a motley group of researchers, professional developers, and “disseminators” to mix it up before the annual meeting. The Office of Education and the NIE also developed and funded a nascent dissemination system, which included an increased role for the relatively new Regional Laboratories, the funding of focused Research Centers in universities (whose work was expected to address applied problems), and dissemination systems like the National Diffusion Network, which maintained a facilitator in each state. Many states joined in, authorizing and funding regional service agencies to become part of the dissemination and professional development systems.[1]

Budget Cuts: Changing Priorities and Systems

Unfortunately, the enthusiasm for an alternative to RDDU eroded, even as a focus on research-driven use persisted:

The key assumption of a technocratic perspective is that if schools had more expertise at their fingertips, change efforts would be more potent or more lasting and would produce a better bottom line; The emphasis placed in federal policies was on dissemination that goes from federally funded research and development (R&D) centers and regional laboratories to schools, but not the other way (Louis, 1992, p.228).

The reasons for the decline are varied. University researchers bemoan the lack of immediate take-up of their important new ideas but scholars who fall into the category of “translational scholars” (concerned with how to apply their work) complain that they receive less recognition in the academy than those who publish in top-ranked journals. Shifting priorities in the federal government have also contributed to the narrowing of perspectives on dissemination and knowledge use. Much of the work of the Institute for Educational Sciences (and its predecessor, the National Institutes of Education) increasingly focused on creating a “gold standard” for educational knowledge (see Datta, 1994). Increased funding of experimental studies helped to convince Congress of the worth of educational research, but at the same time translational scholars report that they Screen Shot 2016-02-19 at 3.06.14 PM copycan’t get financial support to develop applications. Much “good research” remains inaccessible: Even when results are clear there are no easy ways to access them and there are few easy to use professional development materials.

Concurrent with the focus on “gold standard” research, the Department of Education’s emphasis on dissemination and knowledge use declined. In 1995 the federal government discontinued a 20 year-old program, The National Diffusion Network, that was focused on peer-to-peer sharing of and training to use “validated” (rigorously evaluated) innovation programs (Neill, 1981). At the same time, the Department of Education began to redefine the remit for Regional Laboratories to diminish direct engagement with schools and their original function as “developers” or translators of research into more useful practice. The current authorization for the Regional Laboratories narrows their dissemination work to partnerships “with school districts, state departments of education, and others to use data and research to improve academic outcomes for students. Fundamentally, the mission of the RELs is to provide support for a more evidence-reliant education system” ([2] Although the Department and its revived research arm, the Institute for Educational Sciences, has a proven capacity to survive in the face of continued political opposition, less attention has been paid to developing new systems that would encourage the wider flow of knowledge within the community of practitioners.

The emphasis on improving the quality of educational research under both NIE and IES has been (in my view) remarkably successful. In the 1960s, almost all legislators and practitioners believed that there was little educational research that was of high quality or usefulness.[3] That is no longer the case. Schools are generally aware of the value of research-based programs since the requirement to choose them is often embedded in both federal and state funding guidelines. Some successful research-based programs that have demonstrated impact on student outcomes are adopted by large numbers of school districts.

But supporting the demand side – the cadre of educators actively seeking information to improve their work – is still less well understood. Many lack access to resources: there is no national system that reaches school-based practitioners with information that will address their needs and, in most states, departments of education, informal networks and underfunded regional education centers cannot provide what is needed (Louis, Thomas, & Anderson, 2010; Thomas, 2011).

So, is all well?

The Problem of Knowledge Use at the School Level

We are emerging from several decades in which the dominant view that better research-based programs (comprehensive school reform models, for example, or Success for All) and more excellent research would change American education. Many asserted that the accumulation of research and data based evidence about the need for change would allow us to “turn around” low performing schools (or, alternatively, starve them by allowing more effective new schools to arise). At the same time, a concurrent stream of research was still looking a “bringing the user back in” including the idea that practitioners can also be knowledge producers (Louis, 2010, 2015).

We are facing the limits of that belief in the power of knowledge to change outcomes in the short run, just as the medical community long ago acknowledged that the “War on Cancer” would not be accomplished in a lifetime. What is next? I predict that Research4Schools/Center for Research on Knowledge Use will be part of a more nuanced perspective on school improvement that balances “knowledge push” with an understanding of the importance of contexts in which knowledge can be put to use.

Since my epiphany that better research and structures alone will not increase research use, I have focused my work on understanding how districts, schools, and other groups create settings that promote openness to new knowledge, and I have borrowed liberally from Carol Weiss’s ideas about “enlightenment” and “utility” as the screens that attract people to research. I have come to a number of conclusions over the many studies that I have conducted that address knowledge use.

  1. Incentives matter — but incentives come in a variety of different guises.

We cannot underestimate personal incentives – in particular, the desire on the part of teachers or administrators to improve their own work. An additional incentive noted in both education and other sectors is personal recognition, which is a more important reward than is often thought. But organizational incentives are just as important, and these include peer pressure (what others in the relevant group are doing) and mandates or requirements. Even more critical is agreement among peers that a collective “problem” needs to be “solved” – in other words, a sense of urgency.[4]

  1. Leadership matters – especially when using new knowledge requires difficult changes.

Where there is persistent encouragement and oversight by those who have responsibility for whatever improvements are being considered, knowledge use increases. Leaders have impacts not because they are powerful (it is difficult to mandate knowledge use, as suggested by the famous assertion that “you can’t mandate what matters” (McLaughlin, 1990). But principals, superintendents and teacher leaders have influence because they can create sustained conversations about practice and ideas. They also have influence because they help to develop school cultures that can encourages knowledge use (Louis & Lee, under review)

  1. Relationships matter – especially peer relationships among adults.

Whether you call it professional community, communities or practice, teacher networks or another term, teachers who are in significant relationship with other teachers are more likely to use new knowledge (Finnigan & Daly, 2012; Louis & Lee, under review). Knowledge use is social, whether changing individual classroom practice or collectively determining whether which dropout prevention initiatives will work best in a given setting. The number of citations that can be attached to this assertion is quite long, and also corresponds to classic ideas about what differentiates “professional bureaucracies” from other kinds of work settings.

  1. Context matters – there are differences among states, districts, and school in what is needed.

Knowledge use occurs most readily when it addresses a felt need or problem – and these vary widely among schools and districts. “Adoption” of research-based programs or practices is typically not associated with thoughtful strategic planning and environmental scanning in schools but rather with pressing questions that need to be answered quickly (Louis, 1980). We have known for a long time that without assistance, the “matching” of good (well researched) solutions to problems may be poor. Research that meets that kind of need in a poor district in in a low income/lower educational attainment state like Mississippi is unlikely to work as well in an urban district in a high income/high education state like Minnesota. Similarly, the differences in political culture between Mississippi and Minnesota are likely to affect how new ideas spread, and whose voice will carry weight (Louis, Febey, & Gordon, 2015). Major differences in context don’t just occur in disparate geographic areas. A major conclusion from a five-year case study investigation of a school improvement initiative in five schools in the same city is that some kinds of capacity – incentives of varied stakeholders, depth and experience of leaders, the degree to which teachers were already embedded in collaborative work – and the stability in the school had overwhelming impacts on what they could accomplish even when they all had more-or-less the same level of support.

A Coda

Screen Shot 2016-02-19 at 3.09.21 PM copyThe future is always ahead no matter how often we look in the rear view mirror. Thus, my comments are not offered cynically: We are not going around the same block again. Rather, it is exciting to think that we may be at a juncture where the availability of good research and educator’s capacities and enthusiasm for using new knowledge to improve practice are coming together. More importantly, this provides an opportunity to do what I have long advocated – to bring research/knowledge use scholarship together with what is known about school improvement.




Berman, P., & McLaughlin, M. (1978). Federal programs supporting educational change: Vol VII: Implementing and sustaining innovation. Santa Monica, CA: Rand Corporation

Datta, L. E. (1994). A Matter of Consensus. Working Paper. Retrieved from:

Finnigan, K. S., & Daly, A. J. (2012). Mind the gap: Organizational learning and improvement in an underperforming urban system. American Journal of Education, 119(1), 41-71.

Louis, K. S. (1977). Dissemination of information from bureaucracies to local schools: The role of the linking agent. Human Relations, 30(1), 25-42.

Louis, K. S. (1980). Linking R&D with Local Schools: Implications for School Administrators from the Study of the R&D Utilization Program. Paper presented at the Association of School Administrators, Chicago, IL.

Louis, K. S. (1983). Dissemination systems: Some lessons from programs of the past. In W. J. Paisley & M. Butler (Eds.), Knowledge systems in education (pp. 65-89). Beverly Hills: Sage.

Louis, K. S. (1992). Comparative perspectives on dissemination and knowledge use policies: Supporting school improvement. Knowledge: Creation, Diffusion, Utilization, 13(3), 287-304.

Louis, K. S. (2010). Better schools through better knowledge? New understanding, new uncertainty. In A. Hargreaves, D. Hopkins, M. Fullan, & A. Lieberman (Eds.), International handbook of educational change (2nd ed., pp. pp 3-28). NL: Kluwer.

Louis, K. S. (2015). Bringing people back in: Implications for praxis. Journal of Educational Administration, 53(1), null. doi:doi:10.1108/JEA-11-2014-0131

Louis, K. S., & Dentler, R. A. (1988). Knowledge use and school improvement. Curriculum Inquiry, 18(1), 33-62. doi:10.2307/1179560

Louis, K. S., Febey, K., & Gordon, M. F. (2015). Political cultures in education: Emerging perspectives. In B. S. Cooper, J. G. Cibulka, & L. D. Fusarelli (Eds.), Handbook of educational politics and polcy (pp. 118-147). New York: Routledge.

Louis, K. S., & Lee, M. (under review). School cultures and contexts that foster teachers’ capacity for organizational learning. School Effectiveness and School Improvement.

Louis, K. S., Thomas, E., & Anderson, S. E. (2010). How do states influence district leadership. Educational Leadership and Policy, 9(3), 328–366.

McLaughlin, M. (1990). Educational policy, impact on practice. In M. Alkin (Ed.), Encyclopedia of Educational Research (6th ed., pp. 375-382). New York: McMillan.

Neill, S. B. (1981). The National Diffusion Network. The Phi Delta Kappan, 62(10), 726-728. Retrieved from

Nelson, M. K. (1975). Adoption of innovation in urban schools. Final Report. Retrieved from New York:

Persell, C. H. (1971). The quality of research on education; An empirical study of researchers and their work. Final report. New York: Bureau of Applied Social Research, Columbia University.

Rothstein, R. (2004). Class and schools: Using social, economic, and educational reform to close the achievement gap. Washington, DC: Economic Policy Institute.

Sieber, S. D., Louis, K. S., & Metzger, L. (1972). The use of educational knowledge. Retrieved from

Thomas, E. (2011). New federalism and state leadership: The changing eole of state education agencies (SEAs) as quality assurance bureaus (Ph.D.), University of Minnesota, Minneapolis, MN.




[1] Most of these state service agencies still exist. Although little is known about them, they have a professional association, whose website indicates that “The Association of Educational Service Agencies (AESA) is a professional organization serving educational service agencies (ESAs) in 45 states; there are 553 agencies nationwide with hundreds of thousands of staff members. AESA is in the position to reach well over 80% of the public school districts, over 83% of the private schools, over 80% certified teachers, and more than 80% non-certified school employees, and well over 80% public and private school students. Annual budgets for ESAs come to $14.7 billion. ( Although eligible for federal funding, they are not part of any systemic initiatives.

[2] Note that the current authorization is quite different from the more free-ranging “responsive” perspective that dominated in the 1970s and 80’s, when the Laboratories saw themselves as serving schools within their regions (Louis & Dentler, 1988)

[3] There is evidence that this perception was correct (Persell, 1971). In addition, a later study suggested that the quality of innovations (judged by their research base) was inversely related to how often they were adopted (Nelson, 1975).

[4] Not all urgent problems have research-based solutions. For example, although we often hear that “we know how to close the achievement gap” most research points to factors that are associated with “the gap” but are not program evaluations; many successful programs raise achievement of all students without changing the gap. A search for “achievement gap” in the What Works Clearinghouse ( suggests a paucity of validated programs for k-12 education (See also Rothstein, 2004).

Published by