I never seem to tire of quoting Diane Ravitch. In her latest piece at the New York Review of Books, she tackles the ongoing fight between Mayor Bloomberg and the teachers’ union in New York City over teacher evaluation. (For background, see this NYT article and editorial from a couple of weeks ago.)
As Ravitch explains, the background to the New York stalemate is the Race to the Top program of President Obama and his secretary of education, Arne Duncan, with its emphasis on evaluating teachers through student test performance. What’s wrong with that? Well, let Ravitch explain:
Many researchers and testing experts have cautioned that evaluating teachers by the test scores of their students—called value-added assessment—is fraught with problems. Linda Darling-Hammond, a prominent scholar at Stanford University and one of the nation’s leading authorities on issues of teacher quality, has written that the measures say more about which students are in the classroom than about the competence of the teacher.
The National Academy of Education and the American Educational Research Association issued a joint statement saying the same thing. Those who teach students with disabilities, English-language learners, and low-performing students are likely to get smaller gains in test scores than those who teach students from affluent homes in well-funded schools. Using test scores to rate teachers will penalize those who teach the students in greatest need. Over time, teachers will avoid the students who jeopardize their jobs and their reputations. This will be harmful to the students who need talented and experienced teachers most urgently.
It is simply wrong to devise a measure of teacher quality based on standardized tests. The tests are not yardsticks. They are not scientific instruments. They are social constructions, and quite apart from how contingent their results are on the social and economic background of the students being tested, they are also subject to human error, sampling error, random error, and other errors. It is true that the cleanliness of restaurants can be given a letter grade (another of Bloomberg’s test-oriented innovations in New York City), and agribusiness can be measured by crop yields, and corporations can be measured by their profits. But to apply a letter grade or a numerical ranking to a professional is to radically misunderstand the complex set of qualities that make someone good at what they do. It is an effort by economists and statisticians to quantify activities that are at heart matters of judgment, not productivity. Professionals must be judged by other professionals, by their peers. Nowhere is this more true than among educators, whose success at teaching character, wisdom, and judgment cannot be measured by standardized tests.