Christine Stuart photo
CEA President Sheila Cohen and Daniel Long, a Wesleyan Associate Professor of Sociology (Christine Stuart photo)

The state’s largest teachers union announced Thursday that it no longer supports the teacher evaluation method it participated in creating two years ago.

Sheila Cohen, president of the Connecticut Education Association, said she was not a member of the Performance Evaluation Advisory Council that made the recommendations regarding the evaluation system to the state Education Board. She pointed out that none of her current colleagues were on the PEAC council in 2012, either.

Former CEA Executive Director Mary Loftus Levine was a member of PEAC and so was Sharon Palmer, the former head of the second largest teacher’s union, who now heads the Labor Department. 

“It is my understanding that what happened there was not necessarily so much of a percentage, as the fact that we had an ideal and we set forth guidelines,” Cohen said. She explained that between drafting the guidelines and meeting for the last time that summer what the state Education Department put in place “was the interpretation of the state department rather than what was strictly according to the guidelines.”

The state Education Board adopted the guidelines for the System for Educator Evaluation and Development model on June 27, 2012.

Kelly Donnelly, a spokeswoman for the Education Department, said that CEA was an “integral member of PEAC throughout the process.” She said the union has participated “vocally and fully” in the process.

Under that system, standardized tests and other student indicators will make up about 45 percent of a teacher’s performance while the rest is made up by classroom observation, parent and peer surveys, and mutually agreed upon goals.

Cohen said she hopes the conversation about how much student performance should count toward a teacher’s evaluation continues to be part of the discussion.

“This didn’t happen overnight, it’s not going to be resolved overnight,” Cohen said. “But we’re looking for as much as we can do now to provide relief for our teachers but also for our students in the classroom this year.”

Already, the state has agreed to delay implementation of parts of the new teacher evaluation system offering districts some flexibility on how they conduct their evaluations. It also decoupled the evaluations from the new Common Core State Standards and the Smarter Balanced Assessment Consortium test, which will eventually replace the legacy Connecticut Mastery Test and Connecticut Academic Performance Test.

Daniel Long, an assistant sociology professor at Wesleyan University, said he believes the state relied heavily on one-sided academic research to come up with its evaluation system in 2012.

“The National Academy of Science study, which is sort of a summary of all of the kind of best research, from educators, economists, psychologists, sociologists and so on — wasn’t even cited,” Long said. “It would be useful to revisit this.”

Long believes the study he did for more than year in Hamden schools further proves a need to revise the teacher evaluation method.

Long tracked 11 teachers through four Hamden schools and concluded that the Connecticut Education Association’s teacher evaluation system, which relies less on quantitative measures such as standardized test scores, was superior to the teacher evaluation the state adopted.

Long said the field study they did with the CEA model gave teachers more time to focus on their students rather than spend endless hours on paperwork and compliance associated with the state method.

Rebecca Coven, one of Long’s former research associates, said that because educators were allowed to set their own goals and participate more in the evaluation process, their motivation and morale increased.

Hamden teacher David Abate said that through the CEA model of evaluation he was able to provide a complete “photo album of a student’s academic growth rather than just a snapshot in time.”

He went onto explain what he meant by the word photo album. If a student takes a pre-test and scores a 4 out of 20, but by the end of the lesson that same student scores a 12 out of 20 there’s no way to show that growth under the current model.

“The bottom line is this: Did the student show growth? The answer is yes. There is no way to show this on a standardized test,” Abate said. “The teacher and student have no way to show, in some cases great improvement holistically, instead both of us get scored as improvement needed, which is unfair and unjust.”

Cohen said going from 4 to 12 is a huge improvement, but it would mean under the current system that both the teacher and the student would be “deemed as needing improvement.”

Donnelly said the system already supports and requires “the use of non-standardized measures of student growth, which certainly can include the use of student work over time.”

“The evaluation is based on multiple factors and not solely on a standardized test score,” Donnelly said. “That said, we look forward to receiving a copy of the CEA’s study, which we welcome and which has the potential to make a contribution to our dialogue and our continuing collaboration.”