Certifying the intangible: no one really knows what makes a good teacher

Do we need good teachers? Don't be silly. Of course we do. We can all recall a teacher who made a big difference in our lives. And now we have research, as reviewed in Dan Goldhaber's Feature essay, which shows more clearly than ever before that the quality of the teacher has more influence over student learning than anything else at school.

But what makes for a good teacher? Aye, there's the rub, Certainly, it's nothing easily put into a box and counted. Whether a teacher has a master's or other advanced degree appears to count only in science and math. A few years of experience in the classroom makes some difference, but less than one would think, A teacher's command of the English language seems to be an important predictor of a teacher's effectiveness, at least in the United States. But even this, together with all the other measurable elements, leaves more than 90 percent of a teacher's influence unexplained.

When so little is known with certainty, how should we recruit teachers? Most states presently insist that teachers take a specific curriculum taught in schools of education in order to be certified to work in public school classrooms. Unless one follows this well-worn path or gets an emergency exemption, David Ruenzel tells us in his Feature essay, the maze that a prospective teacher must negotiate is filled with dead ends, Yet the hoops and hurdles of certification persist-despite the fact that alternative paths have considerable promise. So show Margaret Raymond and Stephen Fletcher in their careful Research on the classroom performance of Teach for America recruits in Houston. Even more to the point, Kate Walsh Checks the Facts underlying the claims that certification boosts teacher performance, and demonstrates that traditional education coursework, typically required by certification standards, has never been shown-even by its most ardent champions-to have a payoff in the classroom.

Yet certification laws still have their defenders, including Mary Diez, who contends in our Forum on teacher certification that the new standards for teacher education will ensure that teachers are "well equipped to work effectively with learners' In response, James Fraser and Frederick Hess propose alternatives that would allow superintendents and principals, rather than state boards and schools of education, to decide who is qualified to teach.

Key to granting such freedoms to local educators is finding legitimate ways to hold them accountable for their results. The new federal law demands that schools meet certain performance goals or risk sanctions as severe as having their entire staffs fired and forced to reapply for their jobs, a measure known as "reconstitution:' In enacting it, Congress took note of the Research presented here by Thomas Kane and his colleagues, which shows just how complicated it is to judge schools by changes in their test scores. That problem also gives bite, in our Forum on accountability, to Richard Elmore's tough critique of the bipartisan federal initiative, In response, Andrew Rotherham tells us not to rush to judgment too quickly. Testing, he argues, has had positive consequences in some states, and the wrinkles in the system can be ironed out in the years ahead.

Phi Delta Kappa (PDK), however, isn't even trying to iron out the wrinkles in its widely cited surveys of America's attitudes toward the public schools, Terry Moe's powerful Check the Facts shows that PDK, using the venerable Gallup label, routinely fixes its survey questions to get the answers it wants, at least on the issue of school vouchers. PDK neglected to change its format, even after Gallup itself proved that its survey questions on vouchers were biased. We imagine that Mark Twain would use PDK's results as Exhibit A if his famous line on lies and statistics were ever on trial.

THE EDITORS