Patricia and Rodes Hart Professor of Educational Leadership and Policy
Mike McPherson asked me to provide some thinking on the general question of connections between education research and education practice. What follows is my response. While I am pleased to give some thoughts on the topic, the topic is enormous (as Mike acknowledges). What follows is not off the top of my head, since the topic is one that has interested me for some time. However, I did not attempt to write anything like a scholarly paper by way of response.
The first observation that comes to mind in thinking about connections between education research and education practice is that there is both a supply side and a demand side to be considered. Most people focus on the supply side. They ask, What kinds of research have an influence? Is there enough research? Is the research of high enough quality? They often ignore the demand side, which asks, Is the world of practice looking for “research-based solutions”? Do they see education research as a legitimate and important source of input? My own opinion is that too little attention is given to the demand side. For example, how does pre-service teacher education prepare teachers to be thoughtful users of education research? How does in-service teacher education do the same? Does the training of education leaders appropriately address the role that education research should play in education improvement? In the examples that follow, the demand side is shown to be enormously important.
Now for a few positive examples. First, consider the category of research methods. New insights and innovations in psychometrics get immediately put into practice. Why is that? Perhaps it is because the practitioners are few in number, highly trained, and hungry for better solutions. Many of these practitioners are employed by a handful of test companies, such as CTB, which brings me to an additional observation. Testing companies operate in an economically competitive environment, which may motivate their desire to have state-of-the-art solutions to the many psychometric problems they face. HLM is a second example. HLM has been significantly developed and expanded for use in social science research by such education researchers as Tony Bryk and Steve Raudenbush. Today, everyone is using HLM to analyze their hierarchically nested education data sets. This is a huge advance in an incredibly short period of time. A third example that fits the research method category, but makes a different point, concerns randomized field trials. Because IES has decided to only fund randomized field trials in their grants competitions, all of a sudden there are many more randomized field trials being conducted then ever before in the history of education research. Careful work on research methodology has long made clear that randomized field trials provide better evidence of cause and effect relationships then most other approaches. Still, randomized field trials weren’t common despite that research knowledge of their effectiveness until people were essentially “bribed” into using them.
Hypothesis 1: A well-educated and hungry field of practitioners will put relevant research knowledge into practice in a hurry.
Now consider a second category of examples. Smaller class size and school size are reforms that have been thought about for a long time. Neither reform took off until there was good education research to back it up. I’m thinking of the Tennessee Star Study on class size and Bryk and Lee’s work on school size.
Hypothesis 2: Research that documents the utility of education reforms that look easy to do from a distance, and which have lots of advocates for reasons that may not include effects on student achievement, can have a big and widespread effect on education practice (even if they are very expensive). One caveat is that when a reform is taken to scale, it may not have the same effect as found from research in more limited experimental trials. California’s class size reform is a good example.
For my third category of examples, I turn to reforms that deal directly with classroom instruction. Examples are Slavin’s Success for All and Carpenter and Fennema’s Cognitively Guided Instruction. Both are programs carefully and thoughtfully based on large bodies of education research. A substantial development effort was put into building each research-based program. Program effects were tested as well.
Hypothesis 3: For education research at the instruction level to find its way into practice, major development efforts are typically necessary.
From the world of policy, systemic reform/standards-based reform is a good example. The reform is based on considerable education research, including some of my own. But why did standards-based reform become a standard part of federal and state policy and practice? My guess is Mike Smith. Mike wrote a letter to Bassam Shakhashiri, who was then head of the NSF Education & Human Resources Directorate. Mike and Bassam had been colleagues together at the University of Wisconsin-Madison just two or three years earlier, but Mike had left to be Dean of the School of Education at Stanford and Bassam went to NSF. Mike’s letter laid out systemic reform; subsequent to that letter, Mike and Jennifer O’Day published a chapter on systemic reform; and then Bassam picked up on the idea and held a competition for states to engage in systemic reform. The state competition ended up being only the first step; there were several other competitions to encourage people to try systemic reform at various levels of the education hierarchy. Later, Mike joined the Clinton administration as Undersecretary of Education at the time the ESEA was being reauthorized. Not surprisingly, systemic reform/standards-based reform became a prominent part of the reauthorization.
Hypothesis 4: Charismatic individuals can have an enormous influence on the extent to which education research finds its way into education practice.
Now for an example of reform that surprisingly has not found its way into education practice. Technology has huge amounts of research behind it, though substantially less research about how it is best used. Short of the copy machine replacing dittos, technology has had a surprisingly modest effect on education practice at the classroom and school levels. Some may argue that technology requires more time to have effects. To them I say, look what’s happening in virtually every other sector of our society. Technology has completely revolutionized business, higher education, police, firefighters, and the like.
Hypothesis 5: Potential education reforms such as technology use don’t take off when they require practicing educators at the ground floor to invent fundamental new ways of doing their work, because there is not sufficient motivation for them to invest in the change.
My final observation is that education reforms that sweep the nation in a hurry are often not based on education research. Two recent examples are teacher professional development delivered through “coaches” and quarterly student achievement testing. While I have been unable to find a research basis for either reform, both have recently swept the nation (at least in large urban districts). Why is that?
Hypothesis 6: There are networks, both formal and informal, among leaders in education practice. When one influential leader tries a reform in her district or state, she talks about that reform to others in the various networks in which she is a member. Largely through word of mouth, these reform ideas pass quickly from one leader to another and on into implementation. Implementation is variable, because the documentation for each reform is largely missing, making strict replication impossible.
I’m pretty sure there is a literature on connections between research and practice in education. Thirty years ago the U.S. National Institute of Education had a unit on dissemination directed by Senta Raizen. That unit was charged with, among other things, doing research on how to better connect research and practice. Probably others have been doing research on these matters ever since; it’s not a body of research that I follow. Hopefully, good research has been done to test the hypotheses I offer (or better hypotheses than mine).