No magic bullet for education

The “unschooling� movement of the 1970s featured open classrooms, in which children studied what they were most interested in, when they felt ready. That was followed by today’s back-to-basics, early-start model, in which students complete math worksheets in kindergarten and are supposed to take algebra by eighth grade at the latest. Under the “whole language� philosophy of the 1980s, children were expected to learn to read by having books read to them. By the late 1990s, reading lessons were dominated by phonics, with little time spent on the joys of what reading is all about — unlocking the world of stories and information.

A little more than a decade ago, educators bore no responsibility for their students’ failure; it was considered the fault of the students, their parents and unequal social circumstances. Now schools are held liable for whether students learn, regardless of the students’ lack of effort or previous preparation, and are held solely accountable for reaching unrealistic goals of achievement.

No wonder schools have a chronic case of educational whiplash. If there’s a single aspect of schooling that ought to end, it’s the decades of abrupt and destructive swings from one extreme to another. There is no magic in the magic-bullet approach to learning. Charters are neither evil nor saviors; they can be a useful complement to public schools, but they have not blazed a sure-fire path to student achievement. Decreeing that all students will be proficient in math and reading by 2014 hasn’t moved us dramatically closer to the mark.

Now consider the latest rush to extremes: teacher evaluations. In its effort to promote school reform with Race to the Top grants, the Obama administration rightly criticized state laws — such as one then in effect in California — that prohibited schools from making student test scores a part of teacher evaluations, and declared that such laws would preclude a state from qualifying for grants.

The firewall was obviously unreasonable. Part of what the public expects from schools is improvement over the years on test scores, which are clearly related to the quality of instruction. We have endorsed the idea of taking scores into account in teacher evaluations, while cautioning that standardized tests are just one of many factors — and not necessarily the most important one — that should go into a thoughtful and relevant performance evaluation.

But knowing that federal money was attached to this particular issue, states embraced the scores/evaluation connection as though it were the key to every high school senior mastering AP calculus. Tennessee, which won one of the first two Race to the Top grants, promised to make test scores count for half of a teacher’s evaluation; the governor of Colorado, which hopes to win a grant in the next round, signed legislation this month doing the same. The other winner in the first round, Delaware, vowed to come up with a definition of “satisfactory� improvement on scores; any teacher who failed to meet it could not be rated as effective, no matter what else he or she accomplished.

At least these states will measure how much each student improved over the course of the academic year, rather than judging from raw scores, which can depend on whether advanced or struggling students were assigned to a given teacher. But that still doesn’t justify the conclusion that at least half of a teacher’s worth lies in the largely multiple-choice tests given once a year.

Standardized tests have their uses. Over time, they provide a big-picture sense of whether a certain school or academic program is succeeding. They show what students still need to master. Even if teachers are “teaching to the test,� students will perform brilliantly only if they know their subject matter. Similarly, students who regularly score at the bottom clearly lack a firm grasp of the material. Judgments can be made about a teacher’s effectiveness when classes consistently make outstanding progress through the years, or consistently show none at all. But most fall somewhere between these extremes, where it’s harder to measure the effect of teaching quality.

Standardized tests were never designed to be the determinant of a teacher’s mettle. The creators of these tests have repeatedly warned against using the results for more than they were intended, but that has stopped no one from hopping on the latest educational bandwagon.

One problem with this is that statistical errors can be heightened when looking at the scores of a small group in a single classroom. Not uncommonly, test questions are poorly or ambiguously worded, or there’s more than one right answer. In middle and high school, students often put little effort into the tests, which don’t count toward their grades or graduation requirements.

Then there are the realities of the classroom that cannot be measured even by value-added scores. A teacher might have a couple of students with severe learning disabilities who require so much attention that progress for the entire class is held back. Or the improvement made in one year might not show up until the following year’s testing. Students might make great gains in areas that aren’t covered by the tests.

California has demolished the firewall between student scores and teacher evaluations, as it needed to, but has not replaced it with a formula to determine exactly how big a factor tests will be in judging teacher performance. We hope the state won’t join the fad of overrating scores as a measure. Revamping teacher evaluations — a task that both teachers and managers agree is overdue — should be done in a thoughtful and nuanced manner, not a political one.

This country’s tendency to see education reform in simplistic terms is what got us into eight years of trouble under the No Child Left Behind Act. Schools switched overnight from protecting teachers from any accountability for student progress to putting the entire burden on their shoulders, as though they could and should singlehandedly overcome uninvolved parents, troubled neighborhoods and indifferent students. They were given 12 years to bring all those students to proficiency according to a rigid formula that practically guaranteed failure.

We’re glad the Obama administration is proposing major changes to the law. But replacing the extremes of No Child Left Behind with new extremes tied to Race to the Top, with its glorification of charter schools and its emphasis on using scores to judge teachers, will not provide schools or students with meaningful education reform.