Insiders Debate Teacher-Evaluation Approaches
Panelists at a K-12 education forum held in Washington last week sparred over how teachers' performance should be evaluated, presenting a microcosm of what has become a heated national debate.
The forum, hosted by the Atlantic on November 10, was organized around the general principle of strengthening schools and featured a kick-off discussion on "showcasing effective teaching."
Kevin Huffman, spokesperson for Teach for America, opened that discussion by explaining how his organization scrutinizes its best teachers—as determined by diverse measures of student achievement—to identify patterns in effective classrooms. In looking at the 28,000 teachers who have gone through the program, TFA has uncovered two factors that distinguish great teachers from the rest, Huffman explained. The most effective teachers 1) set big goals, and 2) invest students' parents and family in those goals. In these teachers' classes, students can verbalize what they're working toward, and the message permeates their home lives as well.
Because TFA, a program in which high-achieving recent college graduates commit two years to teaching in under-resourced schools, is not using the data to make high-stakes decisions on issues such as tenure, the group is in a unique position to "take on things that are tougher than the district can take on." "We're in the business of studying the efficacy of teachers, simply in the context of how we can make teachers better," Huffman said.
TFA also studies predictors of efficacy in order to refine its screening process. Huffman said the majority of attributes that predict classroom effectiveness are "clustered around the concept of leadership." Perseverance ranks toward the top of the critical predictors. Notably, past experience working with kids does not emerge as a predictive characteristic, he said.
The teacher evaluation rubric now used by the District of Columbia Public Schools known as IMPACT, bears similarities to the evaluations used by TFA, said Jason Kamras, a former National Teacher of the Year. Kamras, now a special assistant in the D.C. school district who helped design IMPACT, contended that evaluations in the past have been a "perfunctory exercise that's not particularly meaningful," with 95 percent of teachers receiving "satisfactory" or higher ratings. The new system, which uses a combination of value-added scores and observations to rate teachers, has stirred up controversy in the District. Many teachers who received positive ratings in previous years prior have scored as ineffective under IMPACT.
The discussion's moderator, Atlantic contributor Amanda Ripley, prodded panelists on how to overcome suspicions about these evaluations from teachers who see them as "punitive."
But Daniel Koretz, professor of education at Harvard University, contended that the suspicions about evaluations using value-added scores are warranted. "We can't avoid the need for better accountability," he said, but we "should be nervous about the emphasis on tests." Standardized tests fail to account for much of what comprises good teaching, Koretz said, and can identify the wrong teachers as effective. And state test scores do not tend to correlate to national assessments, which he said is an indication they can be unreliable.
This is particularly a problem in New York State, he said, where the media has picked up on huge disparities between state and national scores.
"Are you contending that teachers who have the highest value-added scores on NY state tests are not substantially different than the teachers getting the lowest scores?" Kamras shot back.
"In some cases," said Koretz. "Some teachers who are rated poorly are, in fact, really good."
Kamras argued that he notices a "palpable difference between classes" of teachers with the highest and lowest test scores. He conceded that "value added is not perfect. There are reasons to be cautious." That's why, he said, test scores are only one piece of the evaluation equation for IMPACT.
When asked whether teachers' value-added scores should be made public, in reference to the wrangling between unions and districts in New York and L.A., panelists had mixed reaction. As a former teacher and current team leader for teachers, having the information out in the open seems like it would be "quasi-disastrous," said Huffman. "But as a parent of DCPS kids," he asserted, "I want to know."
As it is, school policies can confuse parents about which classrooms are the most effective, said Kamras. Schools are required to send out letters to parents whose child has a teacher that is not considered "highly qualified" by the legal definition. But that has nothing to do with whether or not a teacher is highly effective, Kamras said. The release of value-added data would necessitate clear communication about what the scores signify.
Kamras, who also has children in the District of Columbia public schools, also argued that parents who want to know which teachers are "good" and "bad" have methods of finding out already—usually by asking around or visiting the school. Teacher performance, in other words, is not a big secret, Kamras suggested.