Most teacher professional development focuses on formal workshops or mentorships, but Tennessee hopes to help teachers improve by leveraging the power of a friendly conversation over coffee or a quick word in the hallways with teacher colleagues.
The state's Instructional Partnership Initiative pairs together educators who have complementary strengths and weaknesses so that they can help one another improve their teaching.
Jackson-Madison County schools participated in the original pilot of the initiative with fewer than 30 teachers, and has now expanded it to nearly 450 teachers in 11 schools.
The program is part of an ongoing project between the state and the Tennessee Education Research Alliance, a group of researchers testing out new ideas in teacher professional development. In a new statewide survey, 3 in 4 participating teachers said the evaluation and feedback improved their teaching, but nearly that many said they did not get professional development tailored to improve the skills that were identified.
The initiative is one of several pilots being scaled up around the state to provide more personalized support for teacher professional development.
"It's a low-cost program … but it builds on what's already taking place and the knowledge and skills that already exist in schools to allow everyone to improve their practice," said Nate Schwartz, Tennessee's chief research and strategy officer.
Make Me a Match
The high-poverty, 13,000-student district in West Tennessee has traditional teacher training workshops and mentorships for new teachers, but Rachel Lebo, the district's leader of professional development, said the partnership program has helped teachers grapple with practical daily problems more than the "traditional sit-and-get PD."
"We know job-embedded [professional development] is what moves our practice forward, and this is something that can be leveraged in that way," Lebo said.
The state began offering the program to all schools last year, and 100 to 150 of them have started implementation, with more than 1,250 teachers formally paired.
Tennessee evaluates teachers based on both their contributions to students' test scores and on classroom observations in three areas: instructional design and planning, learning environment, and instruction. Within those broad areas, a teacher is rated on 19 different skills, from lesson plans and managing student behavior to strategies for questioning and giving students appropriate academic feedback.
A 2016 pilot study led by John Papay, an assistant professor of education and economics at Brown University, found that more than 40 percent of the teachers who were studied initially scored less than 3 out of 5 in at least one of the skill areas on the evaluation, and among those, teachers had low scores in nearly six different skills on average.
Each July, the state uses an algorithm to match teachers in each school based on their observation scores in the prior year. A teacher with a score of 2 or 3 out of 5 in, say, strategies for questioning students but a score of 4 in planning student work would be matched with a teacher with stronger questioning skills but worse work planning. The state only sends a school potential matches if it can find at least three pairs of teachers that have complementary strengths and weaknesses, said Papay. About half of teachers in a participating school are given potential matches, and the principal and teachers ultimately decide whether to participate.
Teachers of any subject or grade level can be paired, but first-year teachers—who have not yet been evaluated—cannot be matched, said Jason Grissom, an associate professor of public policy and education at Vanderbilt University who is studying implementation of the program.
"It helps fill a need in some schools, because it can make matches across the teaching force," Grissom said, "and creates space for more experienced teachers and teachers in special subjects—art, music—to get some support."
Teachers meet initially and set their own goals based on the areas they each need to improve. There aren't many formal requirements, but Grissom said most teachers observe each others' classes, plan lessons together, and debrief after administrative evaluations. Often, he said, paired teachers may just chat for a few minutes to talk over an upcoming lesson.
"It's a very different model," Papay said. "This is more of a management practice, a tool to encourage professional learning through peer collaboration. It's not something where we say teachers need to meet once a month for two hours to do X, Y, Z. It's designed to have teachers decide how to walk their own path."
Teachers do not receive stipends for working together, though some schools do help teachers schedule time to collaborate.
Brandi Hodge, a 7th-grade math teacher at E. A. Cox Middle School, said in a presentation on the program that pairing up helped her connect with a teacher of the same subject in a different grade level. "We're like students in a way sometimes; we are creatures of habit and find our teachers that we know, so it kind of took us outside of our comfort zone," she said. In the 2016 pilot, students of initially low-performing teachers who participated in the partnership program performed significantly better on state math and language arts tests at the end of the year than students whose teachers had the same initial evaluation scores but did not participate—gains equal to the difference between being taught by an average teacher versus one in the lowest 25 percent.
Their growth was roughly equal to that of a student moving from a classroom taught by a teacher in the bottom 25 percent of performance to a classroom taught by an average teacher.
Also, veteran teachers improved a little more than those with less than 10 years experience, and the more closely teachers were matched on multiple target skills, the more they improved.
"Some people who seem initially like an unlikely pair, by the end actually were more successful, because they had such differing ideas," Lebo said. "As humans we tend to gravitate to the people who think very similar to the way we are, and what we found was when we put them with someone they wouldn't normally be with, the conversations remained very focused and they enjoyed the experience—and maybe even grew more than the others."
Teachers who participated in the partnership also developed a more positive overall impression of the state's teacher-evaluation system than teachers who did not.
Bethany Bowman, the director of professional learning for the Professional Educators of Tennessee, the state teachers' union, said the initiative "appears to be an effective program to promote growth and collaboration," but expects it may take several years to determine whether the scale-up is effective.
Papay, Grissom, and their colleagues who studied the original pilot have not finished evaluating the program at scale. While districts have conducted teacher evaluations, Tennessee canceled statewide testing for 2016 after problems with its testing contractor, preventing the evaluators from figuring out if changes in teachers' practices have led to improved student learning. Moreover, both Lebo and Bowman suggested the state will need to provide more support for teachers to leverage their new relationships.
"It's not just a random conversation; it's very goal-focused," Lebo said.