Teachers unions fight public performance rankings

A big legal battle is brewing between New York City's education department and the local teachers union over the recently announced plan to issue public rankings of teachers based on how their students perform on tests. Union leaders say the practice would be an intrusion on the privacy of rank-and-file members -- and are suing the city today to prevent the plan from going forward. Unless the union prevails, the education department will hand over the rankings of 12,000 teachers tomorrow morning to reporters who have requested the data.

The suit says teachers will face "harassment" from angry parents if the data is released, reports education blog Gothamschools.

"Such harassment could include demands for termination, discipline, and transfer of children out of teachers' classrooms, as well as threats to the persons of individual teachers," the suit reads.

So what's the big deal? It seems reasonable for parents to want to know if their child's teacher is improving classroom performance on standardized tests. Indeed, U.S. Education Secretary Arne Duncan says that all U.S. school districts should issue such public rankings, saying in a speech that "the truth is always hard to swallow, but it can only make us better, stronger and smarter."

But what if the scores aren't really the "truth"? That's what the United Federation of Teachers (UFT) contends, arguing the city's data is faulty and will wrongly stigmatize some teachers as bad. Those worries echo complaints from teachers unions across the country, which are fending off the call for more rigorous teacher evaluations in education reform circles.

"We've looked at the first 20 reports we had access to, and 13 of them have the wrong information in terms of they don't have the right students with the right teachers. So it's not good information," UFT President Michael Mulgrew told the cable news station NY1.

But the battle over what are known as "value-added" ratings goes beyond specific inaccuracies. For one thing, standardized tests can be flawed themselves; a recent study of New York tests, for example, found that they were too easy. For another, value-added analysis is often inaccurate if the rankings take into account only a few years of data. A study found that there's a one-in-four chance a teacher will be misidentified as bad if the sample only includes two years' worth of data. In addition, some argue that the measurements can never fully control for outside factors, such as a student's economic background.

"There are a lot of benefits to this approach, but the science of the methodology at this point isn't where it should be to attach teachers' names to it," Douglas Ready, a professor at Columbia University's Teachers College, told the Wall Street Journal.

This same battle played out in Los Angeles in August, when Los Angeles Times reporters released data that spanned seven years of teaching to rate 6,000 local teachers. Teachers union president A.J. Duffy called for a boycott of the paper in retaliation. He told teachers that the scores were "an irresponsible, offensive intrusion into your professional life that will do nothing to improve student learning."

The paper set up a database allowing readers to search for a teacher by name and establish his or her value-added score.

The reporters found that a student who spent a single year with a teacher ranked in the top 10 percent would score 17 percentile points higher on English and 25 points higher on standardized math tests compared with students who spent that year with a teacher in the bottom 10th percentile.

In New York, 25 percent of the 12,000 teachers received poor scores, another 25 percent high scores, and the rest of the teachers fell somewhere in the middle. According to Gothamschools, about 45 percent of New York teachers didn't download their own performance scores when they were made available online this year.

(Photo: Getty)