School & District Management

Take It From Tennessee: Lessons on Teacher Evaluation

By Liana Loewus — November 27, 2012 3 min read
  • Save to favorites
  • Print

At a session on teacher effectiveness at yesterday’s Excellence in Action National Summit in Washington, Tennessee Education Commissioner Kevin Huffman offered some lessons learned for legislators and state leaders looking to implement revised teacher-evaluation systems. The information he presented may be instructive for educators as well, since Tennessee was one of the first states to go down this perilous road.

Tennessee is in its second year of using a new statewide system that bases 50 percent of a teacher’s evaluation on observation ratings, 35 percent on student-growth measures, and 15 percent on other measures of achievement. Last fall, Education Week reported on that system, as well as the backlash against it from overwhelmed teachers and principals. More recently, the Teacher Beat blog covered a report finding that teachers’ observation scores tend to be inflated under the new system.

At the conference, Huffman was candid about what he wishes he’d done differently that first year. First, he said, he and his team should have been more proactive about communicating directly with teachers about the new system, rather than filtering messages through districts and schools—a process that led to rampant rumors and misconceptions.

“When we took over the reins of communication, things got better,” he said. “One of the things we needed to communicate was that we would listen to teachers,” especially in looking to make changes in the second year of implementation.

The state should have done more to ease the burden for principals as well, Huffman said. “What we did was a lot more work for principals. I had principals say to me, ‘I’m spending so much time in classrooms that I can’t do my real job!’ ... The response to that is fairly obvious. But what is a legitimate complaint is that we didn’t take other things off their plate. We needed to take things they considered their ‘real job’ and distribute [them].”

Further, Huffman said allowing principals and teachers to agree on a quantitative measure for 15 percent of the evaluation simply “didn’t work.” Principals and teachers are choosing measures they know they can do well on—for instance, high school teachers can use graduation rates, which are reliably high since most students who will not finish drop out in the 9th or 10th grade and are not counted. “People are gaming the system,” he said.

In an interview after the session, Huffman said that for now, the 15-percent provision will stay, but he’s pushing for legislators to roll that percentage into the observation portion. Huffman also emphasized that he will hold fast on incorporating school-wide performance goals into evaluations (an issue he’s received consistent pushback on) because it encourages all teachers to incorporate literacy and math into their instruction and because he believes it’s a better option than trying to create tests for every grade level and subject.

During the session, Jill Hawley, Colorado Associate Commissioner for Achievement and Strategy, spoke about the pilot program for new teacher evaluations going on in her state as well. Colorado passed a law requiring changes to the evaluation system in 2010, but is behind Tennessee in implementation.

Perhaps the most notable feature of what’s happening in Colorado, based on Hawley’s description, is that so-called “content collaborative” teams are looking internationally—to places including New Zealand—for assessments in non-tested subjects. “We’re getting assessments from around the world in music, dance, theater,” she said. The content collaboratives evaluate whether assessments match the state’s standards, and “once they’ve passed that screen, they go up in an assessment resource bank that our schools and districts can use in the field.”

A version of this news article first appeared in the Teaching Now blog.