Metro

Report finds potential problems with new state rating system for teachers

Teachers in classrooms with high numbers of poor or disabled students tended to get slightly lower ratings than their peers under the state’s new evaluation system, according to an outside review.

Similarly, teachers whose classes enrolled lots of already high-performing students were slightly more likely to earn better ratings on average, a report by the American Institute for Research found.

The trends came despite the fact that the evaluation system — which rates teachers based in part on the state test scores of their students — tries to account for the effects of poverty, disability, prior test scores and English proficiency on a student’s performance.

Experts say it’s because the new system doesn’t account for the effects of having large numbers of challenging or high-performing kids concentrated in one classroom.

“It certainly is a problem — although from what [the state has] presented so far, it isn’t apparent how big the problem is for individual teachers’ scores,” said Sean Corcoran, an associate professor of education economics at NYU.

The ratings, which are called “growth scores” because they look at a teacher’s ability to boost students’ test scores over prior years, account for only 20 percent of a teacher’s evaluation.

The state calculated those partial ratings for the first time this year – even though the city and dozens of other districts are still negotiating with teachers unions over the specific measures that will comprise the remaining 80 percent of a final rating.

The report by AIR, which only looked at the partial ratings that it helped create for the state, acknowledges that classroom characteristics shouldn’t correlate at all with a teacher’s growth score.

But both the reviewers and state officials sought to minimize the “minor” correlations that emerged by emphasizing that plenty of teachers serving challenging kids were able to earn good marks.

“The bottom line of it is that each teacher… has a roughly equal chance of being at the full range of performance,” State Education Commissioner John King told The Post. “There are teachers serving [classrooms of] 100% economically-disadvantaged kids who are getting… top ratings.”

State officials said the fact that teachers in high-performing classes got marginally better growth scores than their peers supports a recent study that found that the best teachers often work in top-notch classrooms or schools.

They said the uneven distribution of teacher talent would also explain why teachers in high-poverty schools — which have trouble attracting the best educators — tended to get slightly worse scores than their peers.

“It’s a question of, is this telling you something descriptive about where talent is placed? Or is it telling you something about the classroom effect [or] school effect of concentrations of students?” said King.

“This data alone can’t really answer that question, which is one of the reasons to have multiple measures — so that you have other information to inform your decision-making,” he added. “No one would say we should evaluate educators on growth scores alone. It’s a part of the picture, but it’s not the whole picture.”

State officials said they’re looking to control for classroom- and school-level student characteristics in coming years, using a more complex evaluation model known as “value-added.”

But Teachers College sociology and education professor Aaron Pallas said the failure to account for the effects of large concentrations of challenging kids could result in a biased review of certain teachers.

“The report tries to down-play this by saying that regardless of a teacher’s class characteristics, there can still be a wide range of [scores] — but that’s little consolation for the teacher who gets a lower score than she should have for reasons outside of her control,” he said.

Compared to teachers, there was an even higher correlation between principal ratings and the concentration of challenging or high-performing students in their schools.

The new principal evaluation system is similarly under negotiation between the principals’ union and the city.

The city has until January 17 to ink a deal with both unions or else forgo more than $250 million in state education aid.