Education Matters When Data Can Be Used Against Teachers Too Fire Them

Two stories once again highlight how the supposedly data-driven corporate-style education reform is actually driven by ideology, with no data too shoddy and suspect to be held up against teachers or as a victory for "reform," and no amount of questionable data enough to sink a champion of "reform" like Michelle Rhee. The New York Times' Michael Winerip asks why, with the Office of the Inspector General in the Department of Education "investigating whether Washington school officials cheated to raise test scores during Ms. Rhee’s tenure," Secretary of Education Arne Duncan would think it appropriate to join a panel also featuring Rhee to discuss, of all things, "the use of education data." Mr. Duncan doesn’t think [there's a problem], according to his spokesman, Justin Hamilton. “It’s irresponsible for a New York Times columnist to presume guilt before we have all the facts,” Mr. Hamilton wrote in an e-mail. “Our inspector general is investigating the cheating issue in D.C. public schools, and we should all let the findings speak for themselves.” And while the investigation is ongoing, the Secretary of Education should by all means validate the authority of the person who led the district during the time currently under investigation, declined to order a real investigation on behalf of the district, and has refused to comment ever since, though she's kept a very high profile all along. While Rhee is celebrated for a record that was dubious even before evidence of cheating surfaced, New York has released reams of shoddy data evaluating teachers based on test scores, and naming them. It has been widely reported that the data have high margins of error, but teachers nonetheless have been separated into categories—low, below average, average, above average, or high—based on it. Additionally, teachers are rated not just based on the test scores their students received, but in comparison with the scores a model expected the students to receive, so that: The data, for example, showed 73 cases in which teachers whose students produced consistently outstanding test scores — at or above the 84th percentile citywide — were nonetheless tagged as below average. The reason? The formula expected even better results, based on the demographics and past performance of the students. [...] In one extreme case, the formula assigned an eighth-grade math teacher at the prestigious Anderson School on the Upper West Side the lowest possible rating, a zero, even though her students posted test scores 1.22 standard deviations above the mean — normally good enough to rank in the 89th percentile. Her problem? The formula expected her high-achieving students to be 1.84 standard deviations higher than the average — roughly the 97th percentile. Finding ways to account for the fact that success means different things for different students is important, but expecting a teacher to achieve the 97th percentile as measured by flawed, imprecise tests is outrageous. Pinning responsibility not just for teaching but for lousy tests, lousy treatment, and structural problems facing students and schools, though, falls in ideological line with the story the corporate reformers want to tell about education, so that data matters. But Michelle Rhee? She'll keep raking in the big bucks and being chummy with the Secretary of Education no matter what the data actually says about her.