How Not to Grade Schools
The creation of school report cards with letter grades attached is an intriguing concept, which explains the enthusiasm by the editorial boards here in Gotham. At its center is a simple idea I advanced in this space more than five years ago, value added testing.
On October 4, 2002, I wrote, "The best schools are not necessarily those that score highest, but rather those that achieve the greatest improvement of their individual students. Only if we look at the schools by this measure can we evaluate the efficacy of the curriculum and teaching methods they employ."
Well, to whatever tiny extent my ideas may have contributed to the release this week of report cards on public schools, I apologize. My proposals were based on the work of William Sanders, then a professor of statistics at the University of Tennessee. He's a great sage. But this episode inclines me to the view of Grover Norquist — that the only way to restrain government is to starve it.
These last 15 or so years have been fat ones for the educrats running city schools. They have benefited from enormously increased educational expenditures by the city, state, and federal governments, just about doubling expenditures for schools without any significant rise of the number of students served. The system is flush with cash.
So the idea of adopting this more logical method of evaluating the progress made by students, teachers, principals, and, yes, chancellors, seemed to make sense. But mixed with tens of millions in cash, and a new $80-million computer system, the result — graded "school report cards" — is a recipe for disaster, on overblown grading system that already seems to be sinking from its own weight.
Assigning letter grades to schools may lend itself to press coverage, but does little to improve education. The value added concept, which could and should stand on its own, is now corrupted with a bagful of subjective adjustments, bonus points, and bureaucratic discretions. Once boiled down to the single familiar letter grade, we end up with nothing.
To make value added testing work requires appropriate testing devices. The tests we administer to children in grades three through eight are not designed to be used in this way. Entirely new tests, carefully calibrated to build one upon another, need to be used or created. So at its core the new system is a kludge.
Control over testing falls into the purview of the New York state education department, which has repeatedly demonstrated incompetence in the design, administration, grading, and security of its testing programs.
Chancellor Klein's initial proposal to administer his own tests in addition to the ones mandated by the state was quickly abandoned after parent groups protested over the number of "high stakes" tests to which their children were subjected.
Were I a principal, I would hesitate to bet my job on anything as poorly conceived or administered as the New York English language arts test, the math tests, or the high school Regents exams. Were I the mayor or a member of the city council, I would cringe at the thought of awarding bonuses (tax dollars for the rest of us) based on the results of these tests. Even were the tests adequate, using just two years of tests (only one period of growth to measure) is problematic, increasing the likelihood of anomalies. This is the most frequently heard criticism of the city's initiative by professionals in the testing field. It should have been rolled out after one more year, at minimum.
But most important to me is that in trying to boil everything down to a letter grade distorts the process. The weighting of the many factors that comprise the grade become political decisions, open to question after the fact.
Each datum could stand on its own. We should use value added test results to inform instructional decisions about individual students, and instructional strategies for the whole school and, indeed, the entire system.
Similarly, the opinion surveys of teachers, parents, and students can stand on their own. So can attendance figures and the dozens of other indicators that make up the score. Weighting all of this and distilling an artificial letter grade may be newsworthy, but not productive.
Finally, there's the question of the city administering, grading, and evaluating the school system it itself runs. The legislature should insist on turning these functions over to an independent entity, one that would ensure that the conclusions are objective, not part of an enterprise whose goal includes advancing the political fortunes of whoever happens to be mayor.
awolf@nysun.com
