In fall 1983, the late Pauline Gough, then managing editor of Phi Delta Kappan, asked me to write a column that would bridge the gap between researchers and practitioners and systematize Kappan's haphazard approach to reporting studies. agreed to try. Ironically, my maiden column appeared in the March 1984 edition that featured a debate between Elliott Eisner of Stanford University and Eva Baker of UCLA. topic: Can ever contribute anything to practice? Eisner took the negative and, thought, won easily. Although this characterization is in my words, not his, Eisner chastised the educational community for slavishly and inappropriately imitating experimental psychology, which, noted, had slavishly and inappropriately imitated physics. Psychology and then educational emphasized method without first deliberating and conceptualizing the phenomena it should study and how. Eisner had checked the duration of experiments in the American Educational Research Journal in 1981 and 1978. experiments lasted on average 72 minutes in 1981, a vast improvement over the 45 minutes of 1978. Nothing that brief can have a large or enduring impact, but that's what you get you model experiments after those with pigeons pecking ping-pong balls and white rats running through mazes. Eisner didn't actually say would never inform practice, but that the paradigm needed to change. By the column's 10th anniversary, thought things had much improved. Eisner thought so, too. When spoke to him, he mentioned trends like the reflective practitioner, lived experience, and action research. These, of course, weren't new. They reflected Dewey, though that name was seldom invoked in those days. Baker didn't respond to a letter or phone calls for that column. For the column's 20th anniversary, wrote a three-part series, The Trouble with Research. trouble with was that it had been subordinated to ideology. For the Bush Administration, science was just another instrument for accomplishing political and ideological goals, not a way to figure out how the world works. Speak out against testing three-year-olds, as Howard Gardner and colleagues did, and you're summarily disinvited to a conference on Head Start. Try to speak out on stem cell research, as Surgeon General Richard Carmona did, and your report is suppressed. U.S. Department of Education demanded scientifically based research even as the Administration declared war on science. Beliefs and Agendas People with beliefs and agendas aren't open to data that conflict with those beliefs and agendas. Gene Glass, in Fertilizers, Pills, and Magnetic Strips (Information Age, 2008), said he invented meta-analysis because I truly believed that the synthesis of dozens or even hundreds of studies into an aggregated, overall conclusion would command the attention and consent of all sides in a debate. Within a decade, he had, learned that no mass of data regardless of its size or its consistency resolves debates about the best way to educate children (p. xii). That's often true, but one problem is in the data meta-analysis analyzes, not in the imperviousness of beliefs. We need to rethink the nature of research, at least as it relates to practice. We need to reassess the role of the teacher in the classroom and in research. At the present time, the teacher is often precluded from both. It's necessary to move on many fronts at once. In Tested: One School Struggles to Make the Grade (Holt, 2008), Linda Perlstein quotes a teacher whose district had just imposed a scripted curriculum: did spend four years in school? Why am getting my master's? It feels like anyone could walk off the street and have my job. In some places, that's the point. But a thinking citizenry should realize the truth in Craig Kridel and Robert Bullough's conclusion in Stories of the Eight-Year Study (SUNY Press, 2007), when ends are taken for granted and means dominate educational discourse as they currently do, teachers will rarely be in control of their work and the reasons for taking one or another course of action will become increasingly bureaucratic and unsatisfying. …