WSJ: New York City's New System for Rating Teachers Could Be Tougher Than Previous Approach

Glimpse of New Teacher Ratings Is Offered

New York City's New System for Rating Teachers Could Be Tougher Than Previous Approach

http://online.wsj.com/article/SB10001424127887324520904578549782294863350.html

By LISA FLEISHER

New York City's new system for rating teachers could be dramatically tougher on educators than the previous one, according to new data released by the city last week.

About 6% of the city's fourth- through eighth-grade teachers were rated ineffective based on their students' scores on state tests last year, the most controversial portion of the new evaluations.

The scores released on Friday won't count for city teachers, because the Department of Education and the teachers union, the United Federation of Teachers, couldn't reach a deal on the details of the new system in time for this year. The state Education Department imposed a system in June and it will take effect in the fall, helping the city and state qualify for hundreds of millions of dollars in federal funding.

But the rankings provided a quick snapshot of how the teacher-evaluation system imposed by the state in May could work in real life. Teachers could be fired if they are deemed ineffective on both state test scores released Friday and on a separate set of school-based tests for two consecutive years.
In the current system, less than 3% of teachers are flunked by their principals, and the firing process isn't as swift.

David Weiner, a deputy chancellor of city schools, said it was "impossible to tell at this point" how many teachers would be rated ineffective. But he said the new system would do a more fair and accurate job of "recognizing teachers that actually, probably shouldn't be teaching anymore."

City officials have said that, under the city's current teacher-performance system, far too few were flunked by principals, who had only two choices: a rating of satisfactory or unsatisfactory. In 2011-12, under the old rating system, 2.6% of the city's roughly 73,000 teachers received unsatisfactory ratings, and the rest were rated satisfactory. It was once common for less than 1% of teachers to be rated unsatisfactory.

The new system, which will take effect this fall, ranks teachers in four categories: ineffective, developing, effective and highly effective. The scores will generally be based 20% on state tests, 20% on school-based tests and 60% on classroom visits by administrators.
Teachers who are rated ineffective on the two test-score portions will flunk the evaluation, no matter how well they do during their classroom observations. Principals and the Department of Education will have some discretion in deciding whom to start termination proceedings against.
It is unclear how many teachers will ultimately receive the lowest rating, but data related to the 2011-12 state tests released Friday by the city Department of Education provides a glimpse of what could come.
The statistics are drawn from 10,544 city teachers who received what are called "growth scores"—a calculation of how well students did on state tests compared with similar students. About 6% were ranked ineffective, 10% were developing, 77% effective and 8% highly effective. The breakdown for the city's teachers was nearly identical to those across the state, though slightly more city teachers were ranked highly effective.
Michael Mulgrew, president of the United Federation of Teachers, said that just because 6% of teachers received the worst rating based on state test scores, doesn't mean that so many will be rated ineffective in the long run.
"You cannot extrapolate," he said. "Very few people will be ineffective…When it comes up, the numbers will bear it out."
Mr. Mulgrew said the teachers' scores were unreliable from one year to the next.
When the city experimented with a similar method of analyzing student test scores, teacher rankings swung wildly from year to year.
The growth scores are designed to show how a teacher's students perform on state tests from one year to the next, comparing students with others who have similar disability, poverty and English fluency status.
Experts on these types of statistics have said that the most accurate method of grading teachers would be to calculate an average of many years' scores over time.
Write to Lisa Fleisher at lisa.fleisher@wsj.com