HARVARD GAZETTE ARCHIVES

Educational research and policy analyst Michael Feuer presents the second in a series of three lectures at the HGSE on education policy. (Staff photo Justin Ide/Harvard News Office)

That question is at the core of the three-part Burton and Ingles Lecture Series at the Harvard Graduate School of Education, which has made the translation of solid educational research into educational practice one of its priorities under Dean Ellen Lagemann. Michael Feuer, a leading educational research and policy analyst, continued to probe "The Science of Rationality and the Rationality of Science" in the second of his three lectures on Feb. 24.

All three lectures by Feuer, who is executive director of the Division of Behavioral and Social Sciences and Education at the National Research Council of the National Academies, focus on cognitive psychology - "where theories of human rationality, ideas about rational decision-making reside" - and its applications on learning, teaching, organization theory, and even public policy.

Yet, said Feuer, there's an ironic twist. "Though cognitive psychology has had this important effect on teaching and learning, and some effect on organization theory, it has had almost no effect on the way we think about the organization of schooling and public policy about education," he said.

Feuer's thesis of how a science of rationality might shed light on schools and schooling also took into account procedural rationality - that is the process by which organizational decisions, for instance are made. He explained that a "purely rational approach" is insufficient for "problems of anything other than quite trivial dimensions."

Further, he argued, the decentralized system of school governance in the United States, where tens of thousands of local school boards hold enormous power over the processes of teaching and learning, compounds the complexity of finding any "quick fix" to education policy. Feuer offered an anecdote about the minister of education in France, who told a wide-eyed audience of American policy-makers that he decided in the spring to try a hands-on approach to science education, and by fall every school in France had implemented the new curriculum.

This sort of centralized decision-taking is impossible in the U.S. system. "There are very special features of the way the U.S. has decided or fallen into its system of governing education that rule out the possibility of anything like an optimal solution to this complex configuration of issues and goals and needs of our system," Feuer said.

The tensions of testing

To illustrate this systemic complexity, Feuer turned to tensions over the use of testing in the United States. While standardized testing has its roots in idealistic reform efforts of 19th century educators, who strove to ensure that all children were receiving equitable educations, relying on test scores for school reform has produced some unintended effects.

First, large-scale, machine-scored tests cannot yet work to provide a nuanced, in-depth analysis of the cognitive functions of teachers and learners. Further, he said, high-stakes tests that aim to change behavior with incentives produced the unintended consequence of either inflating scores or narrowing curriculum (that is, "teaching only to the test").

"We don't have the thermostat to raise the heat on student and teacher behavior just enough to get the performance we're looking for without so distorting the system and so distorting the results that we are no longer able to be confident in that which we are measuring," he said.

And finally, using tests for accountability without controlling for outside factors such as demographics amounts to what he called questionable science. On the other hand, when researchers do take into account such forces, it is sometimes interpreted as making excuses. "When you have a system that is driving towards some kind of a standards-based orientation, this conflict is profound," he said. "There is clearly no ... optimal configuration that can solve the problem."

Such examples, he said, have led him to ponder what would be the procedurally rational way to explore theories of education and education research.

"Maybe the evidence about costs, in this case, costs measured by the threats to validity of a testing program, is not sufficient as a basis to dismiss the program unless you can adequately estimate both the intended and perhaps the unintended benefits of the program," he said.

While randomized controlled trials are the current standard of evidence in educational research - no doubt a well-intentioned notion to boost the scientific nature of education research - Feuer cautioned against embracing such scientific values exclusively.

"If the effect is to so narrow education research that it crowds out historians, ethnographers, people doing seriously important qualitative inquiry, that would be a tragedy," he said. "Again, what we're missing here is the thermostat."

Michael Feuer's third Burton and Ingles Lecture at the Graduate School of Education will be April 21.