Skip to content

Usable Knowledge

Decrease text size Increase text size

Home > Decisions through Data > What can educators learn from the student-teacher relationship?

What can educators learn from the student-teacher relationship?
Professor Deborah Meier, NYU Steinhardt School of Education

Deborah Meier, senior scholar and adjunct professor at New York University's Steinhardt School of Education, urges education researchers to stop thinking about students as simply numbers in large-scale experiments, and to start thinking about them as individual learners. Arguing that randomized experiments in education are unethical, Professor Meier offers a sharp counterpoint to the views presented in a companion Usable Knowledge article by HGSE Professor Judith Singer and graduate Beth Gamse. From the Harvard Education Letter, May/June 2005.

The Student-Teacher Relationship: Our Most Important "Data" Source
by Deborah Meier

In "Lessons from the Red Sox Playbook" (HEL, January/February 2005), Beth C. Gamse and Judith D. Singer draw an excellent analogy between managing the World Series champion Boston Red Sox and trying to improve the quality of teaching in our schools. They describe how Red Sox manager Theo Epstein learns by closely studying the data of actual performance—what we educators might call work samples—in the context of his own intimate knowledge of the game, his players, and the opposition. It's a description worthy of what the best schoolteachers and principals do every day.

A good baseball manager, like a good teacher, takes into account a wide range of evidence, including some that comes in statistical form. He applies it on a trial-and-error basis, tailoring his strategies to the nuances of an always-changing set of circumstances. The same approach holds for the classroom, an environment in which circumstances change at a speed that makes baseball look stodgy indeed.

But the comparison breaks down when it is used to make the case for randomized experimentation in schools. No baseball manager—Red Sox, Yankee, or otherwise—ever agreed to be a control site for an experiment in baseball. As in teaching, the stakes are too high to leave the outcome to chance.

The analogy Gamse and Singer draw between randomized trials in education and in medicine is similarly flawed. When randomized trials take place in schools, they lack one of the essential features of medical research: voluntarism. It's true that districts have a choice whether or not to participate, but individual schools, teachers, and students generally do not. Nor can participants simply "drop out" of the study—as so many subjects of drug experiments do—unless they drop out of school altogether. Testing a new method of instruction on students is ethically akin to testing a new drug on prisoners—the subjects essentially have no choice but to participate in the study, regardless of whatever negative effects they might suffer from their participation.

Even in medicine, science has its limits—particularly when variations in human behavior are involved. In one recent study of four popular diet plans, all performed more or less equally well with controlled samples of dieters. But we all know people who have lost weight on one diet after failing at two or three others. The tricky part is figuring out which diet is most likely to work best for you, given your unique tastes, habits, and health history.

Studies of different educational programs present a similar dilemma. There may be a statistically "best" way to teach swimming, baseball, or spelling to the hypothetical "average" person. But what about all those people who aren't average? The imposition of a single solution on every individual is not and should not be the answer—not in sports, medicine, or education.

We ignore at our peril the importance of relationships in education—relationships that go far beyond dispensing a prescription or monitoring weekly caloric intake. The kids most left behind in our schools—those who flail and sputter while others, by comparison, float blithely along—need those relationships most of all. They need to feel that their teachers believe in them and will buoy them if they start to sink. The teacher who goes the extra mile to earn a child's trust, who offers a nuanced response to each learner and each classroom, may account for far more variation in children's performance than the exact method of instruction used. Relationships are the non-interchangeable part of teaching—a factor no study can control for.

When Central Park East, the New York City school I founded in 1985, expanded from an elementary to a K-12 school, some of the kids complained that they were "guinea pigs," as we experimented with innovations like two-hour time blocks, teaching science and math together, and assessments that resembled doctoral-level orals instead of standardized quizzes. Part of what made the experiment succeed, however, was the students' awareness that we were a community of caring adults who demanded the same things of ourselves that we did of them: We observed our own and one another's work closely, offered and accepted thoughtful critique, and never gave up. We had a deep personal investment in our students and they knew it, and as a result they gave us their best work and signed onto our grand experiment willingly.

Students learn a lot from the company they keep—including the intellectual habits of their teachers. We're never going to get kids to approach science or literature thoughtfully if their own teachers do not have the space or time for thoughtfulness, much less permission to practice it. Adults need to model the habits of mind they want their students to adopt—good judgment, the exercise of reason, respect for differences, a willingness to try new things, and the courage to ask hard questions. But teachers who are "just following orders"—implementing a one-size-fits-all program in accordance with an experimental protocol—are not helping their students learn these lessons.

Far more than baseball or medicine, schooling is about values. Not everything that will appear to "work" in the short term will meet the long-term goals that parents, teachers, and communities have for children and their learning—and some short-term "solutions" will be in sharp conflict with these goals. What is needed to generate better teaching practice is not simply more experimental research, but a combination of the very same principles that undergird scientific inquiry: an open marketplace of ideas, lots of opportunities for peer review, and access to shared data (including the anecdotal kind). If schools are short on learning from past experience, it's largely because we've structured schooling so that there is little time for observation, analysis, collaboration, and reflection. More data is not what we're lacking, but more time to understand the richest source of information already before our eyes: those unique, ornery, and ever-moving kids.

Deborah Meier is director of new ventures at the Mission Hill School in Roxbury, Massachusetts, and was the founding principal of the Central Park East schools in New York City. She is the author of a number of books and a founding member of the Coalition of Essential Schools, the National Board of Professional Teaching Standards, and the Forum for Education and Democracy.

For Further Information

D. Bensman. Central Park East and Its Graduates: "Learning by Heart." New York: Teachers College Press, 2000.
H.T. Johnson and A. Bröms. Profit Beyond Measure: Extraordinary Results through Attention to Work and People. New York: Free Press, 2000.
E.C. Lagemann. An Elusive Science: The Troubling History of Education Research. Chicago: University of Chicago Press, 2000.
D. Meier. In Schools We Trust: Creating Communities of Learning in an Era of Testing and Standardization. Boston: Beacon Press, 2003.

Reprinted with permission from Harvard Education Letter, May/June 2005.
Harvard Education Letter is an award-winning bimonthly newsletter dedicated to putting research on best educational practices in the hands of school administrators, teachers, and policymakers, with the goal of improving student performance. For more information visit

Harvard Education Letter cover

More Information

Copyright © 2009 The President and Fellows of Harvard College